diff --git a/drivers/Kconfig b/drivers/Kconfig index 48bbdbe..906e837 100644 --- a/drivers/Kconfig +++ b/drivers/Kconfig @@ -113,4 +113,6 @@ source "drivers/xen/Kconfig" source "drivers/staging/Kconfig" source "drivers/platform/Kconfig" + +source "drivers/xenpvdrivers/Kconfig" endmenu diff --git a/drivers/Makefile b/drivers/Makefile index a3c07b1..0f4a0c9 100644 --- a/drivers/Makefile +++ b/drivers/Makefile @@ -5,6 +5,9 @@ # Rewritten to use lists instead of if-statements. # +# Put this at the top so pvdrivers load before normal drivers +obj-$(CONFIG_XEN_PVDRIVERS) += xenpvdrivers/ + obj-y += gpio/ obj-$(CONFIG_PCI) += pci/ obj-$(CONFIG_PARISC) += parisc/ diff --git a/drivers/xenpvdrivers/Kconfig b/drivers/xenpvdrivers/Kconfig new file mode 100644 index 0000000..0c483df --- /dev/null +++ b/drivers/xenpvdrivers/Kconfig @@ -0,0 +1,19 @@ +config XEN_PVDRIVERS + tristate "Xen pvdrivers" + default n + help + The Xen pvdrivers provide PV functionality in XEN HVM mode. This option + is incompatible with CONFIG_XEN (turn off through Paravirtualised Guest + support) + +config XEN_PVDRIVERS_SCRUB_PAGES + bool "Scrub pages before returning them to system" + depends on XEN_PVDRIVERS + default y + help + Scrub pages before returning them to the system for reuse by + other domains. This makes sure that any confidential data + is not accidentally visible to other domains. Is it more + secure, but slightly less efficient. + If in doubt, say yes. + diff --git a/drivers/xenpvdrivers/Makefile b/drivers/xenpvdrivers/Makefile new file mode 100644 index 0000000..c069208 --- /dev/null +++ b/drivers/xenpvdrivers/Makefile @@ -0,0 +1,81 @@ +# Hack: we need to use the config which was used to build the kernel, +# except that that won't have the right headers etc., so duplicate +# some of the mach-xen infrastructure in here. +# +# (i.e. we need the native config for things like -mregparm, but +# a Xen kernel to find the right headers) +_XEN_CPPFLAGS += -D__XEN_INTERFACE_VERSION__=0x00030205 +_XEN_CPPFLAGS += -DCONFIG_XEN_COMPAT=0xffffff +_XEN_CPPFLAGS += -DHAVE_XEN_PLATFORM_COMPAT_H +ifeq ($(ARCH),ia64) + _XEN_CPPFLAGS += -DCONFIG_VMX_GUEST +endif + +# Xen code does all sorts of unnecessary weird stuff it is isn't built with MODULE +# set that seems to break compile. So the code is cahnged to look at XENPV_COMPAT_MODULE +# instead, which we set true, so it compiles as if for a module, even though MODULE +# isn't set +_XEN_CPPFLAGS += -DXENPV_COMPAT_MODULE + +_XEN_CPPFLAGS += -include $(objtree)/include/linux/autoconf.h + +EXTRA_CFLAGS += $(_XEN_CPPFLAGS) +EXTRA_AFLAGS += $(_XEN_CPPFLAGS) + + +nostackp := $(call cc-option, -fno-stack-protector) +CFLAGS_features.o := $(nostackp) + +# balloon drivers +obj-$(CONFIG_XEN_PVDRIVERS) += xen-balloon.o +xen-balloon-y := balloon.o sysfs.o +xen-balloon-$(CONFIG_XEN_PVDRIVERS_SCRUB_PAGES) += scrub.o + +# blkfront +obj-$(CONFIG_XEN_PVDRIVERS) += xen-vbd.o +xen-vbd-objs := blkfront.o +xen-vbd-objs += vbd.o +xen-vbd-objs += vcd.o + +# netfront +obj-$(CONFIG_XEN_PVDRIVERS) += xen-vnif.o +xen-vnif-objs := netfront.o +xen-vnif-objs += accel.o + +# platform-pci and xenbus +obj-$(CONFIG_XEN_PVDRIVERS) += xen-platform-pci.o +xen-platform-pci-objs := evtchn.o platform-pci.o gnttab.o xen_support.o +xen-platform-pci-objs += features.o platform-compat.o +xen-platform-pci-objs += reboot.o machine_reboot.o +xen-platform-pci-objs += panic-handler.o +# Apparently as a module, we want to link these in more than once, which causes +# problems as a built in +ifeq ($(CONFIG_XEN_PVDRIVERS),m) + xen-platform-pci-objs += xenbus_comms.o + xen-platform-pci-objs += xenbus_xs.o + xen-platform-pci-objs += xenbus_probe.o + xen-platform-pci-objs += xenbus_client.o +endif +xen-platform-pci-objs += xenbus_dev.o +xen-platform-pci-objs += xen_proc.o + +# Can we do better ? +ifeq ($(ARCH),ia64) + xen-platform-pci-objs += xencomm.o xencomm_arch.o xcom_hcall.o xcom_asm.o +endif + +# scsifront +obj-$(CONFIG_XEN_PVDRIVERS) += xen-scsi.o +xen-scsi-objs := scsifront.o xenbus.o + +# xenbus +obj-y += xenbus_client.o xenbus_comms.o xenbus_xs.o xenbus_probe.o +obj-$(CONFIG_XEN_BACKEND) += xenbus_be.o + +xenbus_be-objs = +xenbus_be-objs += xenbus_backend_client.o + +# FIXME-AMB use different conditional +xenbus-$(CONFIG_XEN_BACKEND) += xenbus_probe_backend.o +obj-y += $(xenbus-y) $(xenbus-m) +#obj-$(CONFIG_XEN_XENBUS_DEV) += xenbus_dev.o diff --git a/drivers/xenpvdrivers/accel.c b/drivers/xenpvdrivers/accel.c new file mode 100644 index 0000000..c0519e9 --- /dev/null +++ b/drivers/xenpvdrivers/accel.c @@ -0,0 +1,828 @@ +/****************************************************************************** + * Virtual network driver for conversing with remote driver backends. + * + * Copyright (C) 2007 Solarflare Communications, Inc. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include +#include +#include + +#include "netfront.h" + +#define DPRINTK(fmt, args...) \ + pr_debug("netfront/accel (%s:%d) " fmt, \ + __FUNCTION__, __LINE__, ##args) +#define IPRINTK(fmt, args...) \ + printk(KERN_INFO "netfront/accel: " fmt, ##args) +#define WPRINTK(fmt, args...) \ + printk(KERN_WARNING "netfront/accel: " fmt, ##args) + +static int netfront_remove_accelerator(struct netfront_info *np, + struct xenbus_device *dev); +static int netfront_load_accelerator(struct netfront_info *np, + struct xenbus_device *dev, + const char *frontend); + +static void netfront_accelerator_remove_watch(struct netfront_info *np); + +/* + * List of all netfront accelerator plugin modules available. Each + * list entry is of type struct netfront_accelerator. + */ +static struct list_head accelerators_list; + +/* Workqueue to process acceleration configuration changes */ +struct workqueue_struct *accel_watch_workqueue; + +/* Mutex to prevent concurrent loads and suspends, etc. */ +DEFINE_MUTEX(accelerator_mutex); + +void netif_init_accel(void) +{ + INIT_LIST_HEAD(&accelerators_list); + + accel_watch_workqueue = create_workqueue("net_accel"); +} + +void netif_exit_accel(void) +{ + struct netfront_accelerator *accelerator, *tmp; + + flush_workqueue(accel_watch_workqueue); + destroy_workqueue(accel_watch_workqueue); + + /* No lock required as everything else should be quiet by now */ + list_for_each_entry_safe(accelerator, tmp, &accelerators_list, link) { + BUG_ON(!list_empty(&accelerator->vif_states)); + + list_del(&accelerator->link); + kfree(accelerator->frontend); + kfree(accelerator); + } +} + + +/* + * Watch the configured accelerator and change plugin if it's modified + */ +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20) +static void accel_watch_work(struct work_struct *context) +#else +static void accel_watch_work(void *context) +#endif +{ +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20) + struct netfront_accel_vif_state *vif_state = + container_of(context, struct netfront_accel_vif_state, + accel_work); +#else + struct netfront_accel_vif_state *vif_state = + (struct netfront_accel_vif_state *)context; +#endif + struct netfront_info *np = vif_state->np; + char *accel_frontend; + int accel_len, rc = -1; + + mutex_lock(&accelerator_mutex); + + accel_frontend = xenbus_read(XBT_NIL, np->xbdev->otherend, + "accel-frontend", &accel_len); + if (IS_ERR(accel_frontend)) { + accel_frontend = NULL; + netfront_remove_accelerator(np, np->xbdev); + } else { + /* If this is the first time, request the accelerator, + otherwise only request one if it has changed */ + if (vif_state->accel_frontend == NULL) { + rc = netfront_load_accelerator(np, np->xbdev, + accel_frontend); + } else { + if (strncmp(vif_state->accel_frontend, accel_frontend, + accel_len)) { + netfront_remove_accelerator(np, np->xbdev); + rc = netfront_load_accelerator(np, np->xbdev, + accel_frontend); + } + } + } + + /* Get rid of previous state and replace with the new name */ + if (vif_state->accel_frontend != NULL) + kfree(vif_state->accel_frontend); + vif_state->accel_frontend = accel_frontend; + + mutex_unlock(&accelerator_mutex); + + if (rc == 0) { + DPRINTK("requesting module %s\n", accel_frontend); + request_module("%s", accel_frontend); + /* + * Module should now call netfront_accelerator_loaded() once + * it's up and running, and we can continue from there + */ + } +} + + +static void accel_watch_changed(struct xenbus_watch *watch, + const char **vec, unsigned int len) +{ + struct netfront_accel_vif_state *vif_state = + container_of(watch, struct netfront_accel_vif_state, + accel_watch); + queue_work(accel_watch_workqueue, &vif_state->accel_work); +} + + +void netfront_accelerator_add_watch(struct netfront_info *np) +{ + int err; + + /* + * If old watch exists, e.g. from before suspend/resume, + * remove it now + */ + netfront_accelerator_remove_watch(np); + + /* Get a watch on the accelerator plugin */ + err = xenbus_watch_path2(np->xbdev, np->xbdev->otherend, + "accel-frontend", + &np->accel_vif_state.accel_watch, + accel_watch_changed); + if (err) { + DPRINTK("%s: Failed to register accel watch: %d\n", + __FUNCTION__, err); + np->accel_vif_state.accel_watch.node = NULL; + } +} + + +static void +netfront_accelerator_purge_watch(struct netfront_accel_vif_state *vif_state) +{ + flush_workqueue(accel_watch_workqueue); + + /* Clean up any state left from watch */ + if (vif_state->accel_frontend != NULL) { + kfree(vif_state->accel_frontend); + vif_state->accel_frontend = NULL; + } +} + + +static +void netfront_accelerator_remove_watch(struct netfront_info *np) +{ + struct netfront_accel_vif_state *vif_state = &np->accel_vif_state; + + /* Get rid of watch on accelerator plugin */ + if (vif_state->accel_watch.node != NULL) { + unregister_xenbus_watch(&vif_state->accel_watch); + kfree(vif_state->accel_watch.node); + vif_state->accel_watch.node = NULL; + + netfront_accelerator_purge_watch(vif_state); + } +} + + +/* + * Initialise the accel_vif_state field in the netfront state + */ +void init_accelerator_vif(struct netfront_info *np, + struct xenbus_device *dev) +{ + np->accelerator = NULL; + + /* It's assumed that these things don't change */ + np->accel_vif_state.np = np; + np->accel_vif_state.dev = dev; + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20) + INIT_WORK(&np->accel_vif_state.accel_work, accel_watch_work); +#else + INIT_WORK(&np->accel_vif_state.accel_work, accel_watch_work, + &np->accel_vif_state); +#endif +} + + +/* + * Compare a frontend description string against an accelerator to see + * if they match. Would ultimately be nice to replace the string with + * a unique numeric identifier for each accelerator. + */ +static int match_accelerator(const char *frontend, + struct netfront_accelerator *accelerator) +{ + return strcmp(frontend, accelerator->frontend) == 0; +} + + +/* + * Add a frontend vif to the list of vifs that is using a netfront + * accelerator plugin module. Must be called with the accelerator + * mutex held. + */ +static void add_accelerator_vif(struct netfront_accelerator *accelerator, + struct netfront_info *np) +{ + if (np->accelerator == NULL) { + np->accelerator = accelerator; + + list_add(&np->accel_vif_state.link, &accelerator->vif_states); + } else { + /* + * May get here legitimately if suspend_cancel is + * called, but in that case configuration should not + * have changed + */ + BUG_ON(np->accelerator != accelerator); + } +} + + +/* + * Initialise the state to track an accelerator plugin module. + * + * Must be called with the accelerator mutex held. + */ +static int init_accelerator(const char *frontend, + struct netfront_accelerator **result, + struct netfront_accel_hooks *hooks) +{ + struct netfront_accelerator *accelerator = + kmalloc(sizeof(struct netfront_accelerator), GFP_KERNEL); + int frontend_len; + + if (!accelerator) { + DPRINTK("no memory for accelerator\n"); + return -ENOMEM; + } + + frontend_len = strlen(frontend) + 1; + accelerator->frontend = kmalloc(frontend_len, GFP_KERNEL); + if (!accelerator->frontend) { + DPRINTK("no memory for accelerator\n"); + kfree(accelerator); + return -ENOMEM; + } + strlcpy(accelerator->frontend, frontend, frontend_len); + + INIT_LIST_HEAD(&accelerator->vif_states); + spin_lock_init(&accelerator->vif_states_lock); + + accelerator->hooks = hooks; + + list_add(&accelerator->link, &accelerators_list); + + *result = accelerator; + + return 0; +} + + +/* + * Modify the hooks stored in the per-vif state to match that in the + * netfront accelerator's state. + * + * Takes the vif_states_lock spinlock and may sleep. + */ +static void +accelerator_set_vif_state_hooks(struct netfront_accel_vif_state *vif_state) +{ + struct netfront_accelerator *accelerator; + unsigned long flags; + + DPRINTK("%p\n",vif_state); + + /* Make sure there are no data path operations going on */ + napi_disable(&vif_state->np->napi); + netif_tx_lock_bh(vif_state->np->netdev); + + accelerator = vif_state->np->accelerator; + spin_lock_irqsave(&accelerator->vif_states_lock, flags); + vif_state->hooks = accelerator->hooks; + spin_unlock_irqrestore(&accelerator->vif_states_lock, flags); + + netif_tx_unlock_bh(vif_state->np->netdev); + napi_enable(&vif_state->np->napi); +} + + +/* + * Must be called with the accelerator mutex held. Takes the + * vif_states_lock spinlock. + */ +static void accelerator_probe_new_vif(struct netfront_info *np, + struct xenbus_device *dev, + struct netfront_accelerator *accelerator) +{ + struct netfront_accel_hooks *hooks; + + DPRINTK("\n"); + + /* Include this frontend device on the accelerator's list */ + add_accelerator_vif(accelerator, np); + + hooks = accelerator->hooks; + + if (hooks && hooks->new_device(np->netdev, dev) == 0) + accelerator_set_vif_state_hooks(&np->accel_vif_state); + + return; +} + + +/* + * Request that a particular netfront accelerator plugin is loaded. + * Usually called as a result of the vif configuration specifying + * which one to use. + * + * Must be called with accelerator_mutex held. Takes the + * vif_states_lock spinlock. + */ +static int netfront_load_accelerator(struct netfront_info *np, + struct xenbus_device *dev, + const char *frontend) +{ + struct netfront_accelerator *accelerator; + int rc = 0; + + DPRINTK(" %s\n", frontend); + + /* + * Look at list of loaded accelerators to see if the requested + * one is already there + */ + list_for_each_entry(accelerator, &accelerators_list, link) { + if (match_accelerator(frontend, accelerator)) { + accelerator_probe_new_vif(np, dev, accelerator); + return 0; + } + } + + /* Couldn't find it, so create a new one and load the module */ + if ((rc = init_accelerator(frontend, &accelerator, NULL)) < 0) { + return rc; + } + + /* Include this frontend device on the accelerator's list */ + add_accelerator_vif(accelerator, np); + + return rc; +} + + +/* + * Go through all the netfront vifs and see if they have requested + * this accelerator. Notify the accelerator plugin of the relevant + * device if so. Called when an accelerator plugin module is first + * loaded and connects to netfront. + * + * Must be called with accelerator_mutex held. Takes the + * vif_states_lock spinlock. + */ +static void +accelerator_probe_vifs(struct netfront_accelerator *accelerator, + struct netfront_accel_hooks *hooks) +{ + struct netfront_accel_vif_state *vif_state, *tmp; + + DPRINTK("%p\n", accelerator); + + /* + * Store the hooks for future calls to probe a new device, and + * to wire into the vif_state once the accelerator plugin is + * ready to accelerate each vif + */ + BUG_ON(hooks == NULL); + accelerator->hooks = hooks; + + /* Holds accelerator_mutex to iterate list */ + list_for_each_entry_safe(vif_state, tmp, &accelerator->vif_states, + link) { + struct netfront_info *np = vif_state->np; + + if (hooks->new_device(np->netdev, vif_state->dev) == 0) + accelerator_set_vif_state_hooks(vif_state); + } +} + + +/* + * Called by the netfront accelerator plugin module when it has + * loaded. + * + * Takes the accelerator_mutex and vif_states_lock spinlock. + */ +int netfront_accelerator_loaded(int version, const char *frontend, + struct netfront_accel_hooks *hooks) +{ + struct netfront_accelerator *accelerator; + + if (is_initial_xendomain()) + return -EINVAL; + + if (version != NETFRONT_ACCEL_VERSION) { + if (version > NETFRONT_ACCEL_VERSION) { + /* Caller has higher version number, leave it + up to them to decide whether to continue. + They can re-call with a lower number if + they're happy to be compatible with us */ + return NETFRONT_ACCEL_VERSION; + } else { + /* We have a more recent version than caller. + Currently reject, but may in future be able + to be backwardly compatible */ + return -EPROTO; + } + } + + mutex_lock(&accelerator_mutex); + + /* + * Look through list of accelerators to see if it has already + * been requested + */ + list_for_each_entry(accelerator, &accelerators_list, link) { + if (match_accelerator(frontend, accelerator)) { + accelerator_probe_vifs(accelerator, hooks); + goto out; + } + } + + /* + * If it wasn't in the list, add it now so that when it is + * requested the caller will find it + */ + DPRINTK("Couldn't find matching accelerator (%s)\n", + frontend); + + init_accelerator(frontend, &accelerator, hooks); + + out: + mutex_unlock(&accelerator_mutex); + return 0; +} +EXPORT_SYMBOL_GPL(netfront_accelerator_loaded); + + +/* + * Remove the hooks from a single vif state. + * + * Takes the vif_states_lock spinlock and may sleep. + */ +static void +accelerator_remove_single_hook(struct netfront_accelerator *accelerator, + struct netfront_accel_vif_state *vif_state) +{ + unsigned long flags; + + /* Make sure there are no data path operations going on */ + napi_disable(&vif_state->np->napi); + netif_tx_lock_bh(vif_state->np->netdev); + + spin_lock_irqsave(&accelerator->vif_states_lock, flags); + + /* + * Remove the hooks, but leave the vif_state on the + * accelerator's list as that signifies this vif is + * interested in using that accelerator if it becomes + * available again + */ + vif_state->hooks = NULL; + + spin_unlock_irqrestore(&accelerator->vif_states_lock, flags); + + netif_tx_unlock_bh(vif_state->np->netdev); + napi_enable(&vif_state->np->napi); +} + + +/* + * Safely remove the accelerator function hooks from a netfront state. + * + * Must be called with the accelerator mutex held. Takes the + * vif_states_lock spinlock. + */ +static void accelerator_remove_hooks(struct netfront_accelerator *accelerator) +{ + struct netfront_accel_vif_state *vif_state, *tmp; + unsigned long flags; + + /* Mutex is held to iterate list */ + list_for_each_entry_safe(vif_state, tmp, + &accelerator->vif_states, + link) { + if(vif_state->hooks) { + spin_lock_irqsave(&accelerator->vif_states_lock, flags); + + /* Last chance to get statistics from the accelerator */ + vif_state->hooks->get_stats(vif_state->np->netdev, + &vif_state->np->stats); + + spin_unlock_irqrestore(&accelerator->vif_states_lock, + flags); + + accelerator_remove_single_hook(accelerator, vif_state); + + accelerator->hooks->remove(vif_state->dev); + } + } + + accelerator->hooks = NULL; +} + + +/* + * Called by a netfront accelerator when it is unloaded. This safely + * removes the hooks into the plugin and blocks until all devices have + * finished using it, so on return it is safe to unload. + * + * Takes the accelerator mutex, and vif_states_lock spinlock. + */ +void netfront_accelerator_stop(const char *frontend) +{ + struct netfront_accelerator *accelerator; + + mutex_lock(&accelerator_mutex); + + list_for_each_entry(accelerator, &accelerators_list, link) { + if (match_accelerator(frontend, accelerator)) { + accelerator_remove_hooks(accelerator); + goto out; + } + } + out: + mutex_unlock(&accelerator_mutex); +} +EXPORT_SYMBOL_GPL(netfront_accelerator_stop); + + +/* + * Helper for call_remove and do_suspend + * + * Must be called with the accelerator mutex held. Takes the + * vif_states_lock spinlock. + */ +static int do_remove(struct netfront_info *np, struct xenbus_device *dev) +{ + struct netfront_accelerator *accelerator = np->accelerator; + unsigned long flags; + int rc = 0; + + if (np->accel_vif_state.hooks) { + spin_lock_irqsave(&accelerator->vif_states_lock, flags); + + /* Last chance to get statistics from the accelerator */ + np->accel_vif_state.hooks->get_stats(np->netdev, &np->stats); + + spin_unlock_irqrestore(&accelerator->vif_states_lock, + flags); + + /* + * Try and do the opposite of accelerator_probe_new_vif + * to ensure there's no state pointing back at the + * netdev + */ + accelerator_remove_single_hook(accelerator, + &np->accel_vif_state); + + rc = accelerator->hooks->remove(dev); + } + + return rc; +} + + +/* + * Must be called with the accelerator mutex held. Takes the + * vif_states_lock spinlock + */ +static int netfront_remove_accelerator(struct netfront_info *np, + struct xenbus_device *dev) +{ + struct netfront_accelerator *accelerator; + struct netfront_accel_vif_state *tmp_vif_state; + int rc = 0; + + /* Check that we've got a device that was accelerated */ + if (np->accelerator == NULL) + return rc; + + accelerator = np->accelerator; + + list_for_each_entry(tmp_vif_state, &accelerator->vif_states, + link) { + if (tmp_vif_state == &np->accel_vif_state) { + list_del(&np->accel_vif_state.link); + break; + } + } + + rc = do_remove(np, dev); + + np->accelerator = NULL; + + return rc; +} + + +/* + * No lock pre-requisites. Takes the accelerator mutex and the + * vif_states_lock spinlock. + */ +int netfront_accelerator_call_remove(struct netfront_info *np, + struct xenbus_device *dev) +{ + int rc; + netfront_accelerator_remove_watch(np); + mutex_lock(&accelerator_mutex); + rc = netfront_remove_accelerator(np, dev); + mutex_unlock(&accelerator_mutex); + return rc; +} + + +/* + * No lock pre-requisites. Takes the accelerator mutex and the + * vif_states_lock spinlock. + */ +int netfront_accelerator_suspend(struct netfront_info *np, + struct xenbus_device *dev) +{ + int rc = 0; + + mutex_lock(&accelerator_mutex); + + /* Check that we've got a device that was accelerated */ + if (np->accelerator == NULL) + goto out; + + /* + * Call the remove accelerator hook, but leave the vif_state + * on the accelerator's list in case there is a suspend_cancel. + */ + rc = do_remove(np, dev); + out: + mutex_unlock(&accelerator_mutex); + return rc; +} + + +int netfront_accelerator_suspend_cancel(struct netfront_info *np, + struct xenbus_device *dev) +{ + netfront_accelerator_purge_watch(&np->accel_vif_state); + + /* + * Gratuitously fire the watch handler to reinstate the + * configured accelerator + */ + if (dev->state == XenbusStateConnected) + queue_work(accel_watch_workqueue, + &np->accel_vif_state.accel_work); + + return 0; +} + + +/* + * No lock pre-requisites. Takes the accelerator mutex + */ +void netfront_accelerator_resume(struct netfront_info *np, + struct xenbus_device *dev) +{ + struct netfront_accel_vif_state *accel_vif_state = NULL; + + mutex_lock(&accelerator_mutex); + + /* Check that we've got a device that was accelerated */ + if(np->accelerator == NULL) + goto out; + + /* Find the vif_state from the accelerator's list */ + list_for_each_entry(accel_vif_state, &np->accelerator->vif_states, + link) { + if (accel_vif_state->dev == dev) { + BUG_ON(accel_vif_state != &np->accel_vif_state); + + /* + * Remove it from the accelerator's list so + * state is consistent for probing new vifs + * when they get connected + */ + list_del(&accel_vif_state->link); + np->accelerator = NULL; + + break; + } + } + + out: + mutex_unlock(&accelerator_mutex); + return; +} + + +/* + * No lock pre-requisites. Takes the vif_states_lock spinlock + */ +int netfront_check_accelerator_queue_ready(struct net_device *dev, + struct netfront_info *np) +{ + struct netfront_accelerator *accelerator; + int rc = 1; + unsigned long flags; + + accelerator = np->accelerator; + + /* Call the check_ready accelerator hook. */ + if (np->accel_vif_state.hooks && accelerator) { + spin_lock_irqsave(&accelerator->vif_states_lock, flags); + if (np->accel_vif_state.hooks && + np->accelerator == accelerator) + rc = np->accel_vif_state.hooks->check_ready(dev); + spin_unlock_irqrestore(&accelerator->vif_states_lock, flags); + } + + return rc; +} + + +/* + * No lock pre-requisites. Takes the vif_states_lock spinlock + */ +void netfront_accelerator_call_stop_napi_irq(struct netfront_info *np, + struct net_device *dev) +{ + struct netfront_accelerator *accelerator; + unsigned long flags; + + accelerator = np->accelerator; + + /* Call the stop_napi_interrupts accelerator hook. */ + if (np->accel_vif_state.hooks && accelerator != NULL) { + spin_lock_irqsave(&accelerator->vif_states_lock, flags); + if (np->accel_vif_state.hooks && + np->accelerator == accelerator) + np->accel_vif_state.hooks->stop_napi_irq(dev); + spin_unlock_irqrestore(&accelerator->vif_states_lock, flags); + } +} + + +/* + * No lock pre-requisites. Takes the vif_states_lock spinlock + */ +int netfront_accelerator_call_get_stats(struct netfront_info *np, + struct net_device *dev) +{ + struct netfront_accelerator *accelerator; + unsigned long flags; + int rc = 0; + + accelerator = np->accelerator; + + /* Call the get_stats accelerator hook. */ + if (np->accel_vif_state.hooks && accelerator != NULL) { + spin_lock_irqsave(&accelerator->vif_states_lock, flags); + if (np->accel_vif_state.hooks && + np->accelerator == accelerator) + rc = np->accel_vif_state.hooks->get_stats(dev, + &np->stats); + spin_unlock_irqrestore(&accelerator->vif_states_lock, flags); + } + return rc; +} + diff --git a/drivers/xenpvdrivers/balloon-common.h b/drivers/xenpvdrivers/balloon-common.h new file mode 100644 index 0000000..0a53f7a --- /dev/null +++ b/drivers/xenpvdrivers/balloon-common.h @@ -0,0 +1,57 @@ +/****************************************************************************** + * balloon/common.h + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __XEN_BALLOON_COMMON_H__ +#define __XEN_BALLOON_COMMON_H__ + +#define PAGES2KB(_p) ((_p)<<(PAGE_SHIFT-10)) + +struct balloon_stats { + /* We aim for 'current allocation' == 'target allocation'. */ + unsigned long current_pages; + unsigned long target_pages; + /* + * Drivers may alter the memory reservation independently, but they + * must inform the balloon driver so we avoid hitting the hard limit. + */ + unsigned long driver_pages; + /* Number of pages in high- and low-memory balloons. */ + unsigned long balloon_low; + unsigned long balloon_high; +}; + +extern struct balloon_stats balloon_stats; +#define bs balloon_stats + +int balloon_sysfs_init(void); +void balloon_sysfs_exit(void); + +void balloon_set_new_target(unsigned long target); +unsigned long balloon_minimum_target(void); + +#endif /* __XEN_BALLOON_COMMON_H__ */ diff --git a/drivers/xenpvdrivers/balloon.c b/drivers/xenpvdrivers/balloon.c new file mode 100644 index 0000000..f396123 --- /dev/null +++ b/drivers/xenpvdrivers/balloon.c @@ -0,0 +1,836 @@ +/****************************************************************************** + * balloon.c + * + * Xen balloon driver - enables returning/claiming memory to/from Xen. + * + * Copyright (c) 2003, B Dragovic + * Copyright (c) 2003-2004, M Williamson, K Fraser + * Copyright (c) 2005 Dan M. Smith, IBM Corporation + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "balloon-common.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#ifdef CONFIG_PROC_FS +static struct proc_dir_entry *balloon_pde; +#endif + +static DEFINE_MUTEX(balloon_mutex); + +/* + * Protects atomic reservation decrease/increase against concurrent increases. + * Also protects non-atomic updates of current_pages and driver_pages, and + * balloon lists. + */ +DEFINE_SPINLOCK(balloon_lock); + +#ifndef XENPV_COMPAT_MODULE +#include +static struct pagevec free_pagevec; +#endif + +struct balloon_stats balloon_stats; + +/* We increase/decrease in batches which fit in a page */ +static unsigned long frame_list[PAGE_SIZE / sizeof(unsigned long)]; + +#ifdef CONFIG_HIGHMEM +#define inc_totalhigh_pages() (totalhigh_pages++) +#define dec_totalhigh_pages() (totalhigh_pages--) +#else +#define inc_totalhigh_pages() ((void)0) +#define dec_totalhigh_pages() ((void)0) +#endif + +#ifndef CONFIG_XEN +/* + * In HVM guests accounting here uses the Xen visible values, but the kernel + * determined totalram_pages value shouldn't get altered. Since totalram_pages + * includes neither the kernel static image nor any memory allocated prior to + * or from the bootmem allocator, we have to synchronize the two values. + */ +static unsigned long __read_mostly totalram_bias; +#else +#define totalram_bias 0 +#endif + +/* List of ballooned pages, threaded through the mem_map array. */ +static LIST_HEAD(ballooned_pages); + +/* Main work function, always executed in process context. */ +static void balloon_process(struct work_struct *unused); +static DECLARE_WORK(balloon_worker, balloon_process); +static struct timer_list balloon_timer; + +/* When ballooning out (allocating memory to return to Xen) we don't really + want the kernel to try too hard since that can trigger the oom killer. */ +#define GFP_BALLOON \ + (GFP_HIGHUSER|__GFP_NOWARN|__GFP_NORETRY|__GFP_NOMEMALLOC|__GFP_COLD) + +#define PAGE_TO_LIST(p) (&(p)->lru) +#define LIST_TO_PAGE(l) list_entry((l), struct page, lru) +#define UNLIST_PAGE(p) \ + do { \ + list_del(PAGE_TO_LIST(p)); \ + PAGE_TO_LIST(p)->next = NULL; \ + PAGE_TO_LIST(p)->prev = NULL; \ + } while(0) + +#define IPRINTK(fmt, args...) \ + printk(KERN_INFO "xen_mem: " fmt, ##args) +#define WPRINTK(fmt, args...) \ + printk(KERN_WARNING "xen_mem: " fmt, ##args) + +/* balloon_append: add the given page to the balloon. */ +static void balloon_append(struct page *page, int account) +{ + unsigned long pfn; + + /* Lowmem is re-populated first, so highmem pages go at list tail. */ + if (PageHighMem(page)) { + list_add_tail(PAGE_TO_LIST(page), &ballooned_pages); + bs.balloon_high++; + if (account) + dec_totalhigh_pages(); + } else { + list_add(PAGE_TO_LIST(page), &ballooned_pages); + bs.balloon_low++; + } + + pfn = page_to_pfn(page); + if (account) { + SetPageReserved(page); + set_phys_to_machine(pfn, INVALID_P2M_ENTRY); + page_zone(page)->present_pages--; + } else { + BUG_ON(!PageReserved(page)); + WARN_ON_ONCE(phys_to_machine_mapping_valid(pfn)); + } +} + +/* balloon_retrieve: rescue a page from the balloon, if it is not empty. */ +static struct page *balloon_retrieve(int *was_empty) +{ + struct page *page; + struct zone *zone; + + if (list_empty(&ballooned_pages)) + return NULL; + + page = LIST_TO_PAGE(ballooned_pages.next); + UNLIST_PAGE(page); + BUG_ON(!PageReserved(page)); + + if (PageHighMem(page)) { + bs.balloon_high--; + inc_totalhigh_pages(); + } + else + bs.balloon_low--; + zone = page_zone(page); + *was_empty |= !populated_zone(zone); + zone->present_pages++; + + return page; +} + +static struct page *balloon_first_page(void) +{ + if (list_empty(&ballooned_pages)) + return NULL; + return LIST_TO_PAGE(ballooned_pages.next); +} + +static struct page *balloon_next_page(struct page *page) +{ + struct list_head *next = PAGE_TO_LIST(page)->next; + if (next == &ballooned_pages) + return NULL; + return LIST_TO_PAGE(next); +} + +static inline void balloon_free_page(struct page *page) +{ +#ifndef XENPV_COMPAT_MODULE + if (put_page_testzero(page) && !pagevec_add(&free_pagevec, page)) { + __pagevec_free(&free_pagevec); + pagevec_reinit(&free_pagevec); + } +#else + /* pagevec interface is not being exported. */ + __free_page(page); +#endif +} + +static inline void balloon_free_and_unlock(unsigned long flags) +{ +#ifndef XENPV_COMPAT_MODULE + if (pagevec_count(&free_pagevec)) { + __pagevec_free(&free_pagevec); + pagevec_reinit(&free_pagevec); + } +#endif + balloon_unlock(flags); +} + +static void balloon_alarm(unsigned long unused) +{ + schedule_work(&balloon_worker); +} + +static unsigned long current_target(void) +{ + unsigned long target = bs.target_pages; + if (target > (bs.current_pages + bs.balloon_low + bs.balloon_high)) + target = bs.current_pages + bs.balloon_low + bs.balloon_high; + return target; +} + +unsigned long balloon_minimum_target(void) +{ +#ifndef CONFIG_XEN +#define max_pfn num_physpages +#endif + unsigned long min_pages, curr_pages = current_target(); + +#define MB2PAGES(mb) ((mb) << (20 - PAGE_SHIFT)) + /* Simple continuous piecewiese linear function: + * max MiB -> min MiB gradient + * 0 0 + * 16 16 + * 32 24 + * 128 72 (1/2) + * 512 168 (1/4) + * 2048 360 (1/8) + * 8192 552 (1/32) + * 32768 1320 + * 131072 4392 + */ + if (max_pfn < MB2PAGES(128)) + min_pages = MB2PAGES(8) + (max_pfn >> 1); + else if (max_pfn < MB2PAGES(512)) + min_pages = MB2PAGES(40) + (max_pfn >> 2); + else if (max_pfn < MB2PAGES(2048)) + min_pages = MB2PAGES(104) + (max_pfn >> 3); + else + min_pages = MB2PAGES(296) + (max_pfn >> 5); +#undef MB2PAGES + + /* Don't enforce growth */ + return min(min_pages, curr_pages); +#ifndef CONFIG_XEN +#undef max_pfn +#endif +} + +static int increase_reservation(unsigned long nr_pages) +{ + unsigned long pfn, i, flags; + struct page *page; + long rc; + int need_zonelists_rebuild = 0; + struct xen_memory_reservation reservation = { + .address_bits = 0, + .extent_order = 0, + .domid = DOMID_SELF + }; + + if (nr_pages > ARRAY_SIZE(frame_list)) + nr_pages = ARRAY_SIZE(frame_list); + + balloon_lock(flags); + + page = balloon_first_page(); + for (i = 0; i < nr_pages; i++) { + BUG_ON(page == NULL); + frame_list[i] = page_to_pfn(page);; + page = balloon_next_page(page); + } + + set_xen_guest_handle(reservation.extent_start, frame_list); + reservation.nr_extents = nr_pages; + rc = HYPERVISOR_memory_op(XENMEM_populate_physmap, &reservation); + if (rc < 0) + goto out; + + for (i = 0; i < rc; i++) { + page = balloon_retrieve(&need_zonelists_rebuild); + BUG_ON(page == NULL); + + pfn = page_to_pfn(page); + BUG_ON(!xen_feature(XENFEAT_auto_translated_physmap) && + phys_to_machine_mapping_valid(pfn)); + + set_phys_to_machine(pfn, frame_list[i]); + +#ifdef CONFIG_XEN + /* Link back into the page tables if not highmem. */ + if (pfn < max_low_pfn) { + int ret; + ret = HYPERVISOR_update_va_mapping( + (unsigned long)__va(pfn << PAGE_SHIFT), + pfn_pte_ma(frame_list[i], PAGE_KERNEL), + 0); + BUG_ON(ret); + } +#endif + + /* Relinquish the page back to the allocator. */ + ClearPageReserved(page); + init_page_count(page); + balloon_free_page(page); + } + + bs.current_pages += rc; + totalram_pages = bs.current_pages - totalram_bias; + + out: + balloon_free_and_unlock(flags); + +#ifndef XENPV_COMPAT_MODULE + setup_per_zone_wmarks(); + if (rc > 0) + kswapd_run(0); + if (need_zonelists_rebuild) + build_all_zonelists(); + else + vm_total_pages = nr_free_pagecache_pages(); +#endif + + return rc < 0 ? rc : rc != nr_pages; +} + +static int decrease_reservation(unsigned long nr_pages) +{ + unsigned long pfn, i, flags; + struct page *page; + void *v; + int need_sleep = 0; + int ret; + struct xen_memory_reservation reservation = { + .address_bits = 0, + .extent_order = 0, + .domid = DOMID_SELF + }; + + if (nr_pages > ARRAY_SIZE(frame_list)) + nr_pages = ARRAY_SIZE(frame_list); + + for (i = 0; i < nr_pages; i++) { + if ((page = alloc_page(GFP_BALLOON)) == NULL) { + nr_pages = i; + need_sleep = 1; + break; + } + + pfn = page_to_pfn(page); + frame_list[i] = pfn_to_mfn(pfn); + + if (!PageHighMem(page)) { + v = phys_to_virt(pfn << PAGE_SHIFT); + scrub_pages(v, 1); +#ifdef CONFIG_XEN + ret = HYPERVISOR_update_va_mapping( + (unsigned long)v, __pte_ma(0), 0); + BUG_ON(ret); +#endif + } +#ifdef CONFIG_XEN_PVDRIVERS_SCRUB_PAGES + else { + v = kmap(page); + scrub_pages(v, 1); + kunmap(page); + } +#endif + } + +#ifdef CONFIG_XEN + /* Ensure that ballooned highmem pages don't have kmaps. */ + kmap_flush_unused(); + flush_tlb_all(); +#endif + + balloon_lock(flags); + + /* No more mappings: invalidate P2M and add to balloon. */ + for (i = 0; i < nr_pages; i++) { + pfn = mfn_to_pfn(frame_list[i]); + balloon_append(pfn_to_page(pfn), 1); + } + + set_xen_guest_handle(reservation.extent_start, frame_list); + reservation.nr_extents = nr_pages; + ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation); + BUG_ON(ret != nr_pages); + + bs.current_pages -= nr_pages; + totalram_pages = bs.current_pages - totalram_bias; + + balloon_unlock(flags); + + return need_sleep; +} + +/* + * We avoid multiple worker processes conflicting via the balloon mutex. + * We may of course race updates of the target counts (which are protected + * by the balloon lock), or with changes to the Xen hard limit, but we will + * recover from these in time. + */ +static void balloon_process(struct work_struct *unused) +{ + int need_sleep = 0; + long credit; + + mutex_lock(&balloon_mutex); + + do { + credit = current_target() - bs.current_pages; + if (credit > 0) + need_sleep = (increase_reservation(credit) != 0); + if (credit < 0) + need_sleep = (decrease_reservation(-credit) != 0); + +#ifndef CONFIG_PREEMPT + if (need_resched()) + schedule(); +#endif + } while ((credit != 0) && !need_sleep); + + /* Schedule more work if there is some still to be done. */ + if (current_target() != bs.current_pages) + mod_timer(&balloon_timer, jiffies + HZ); + + mutex_unlock(&balloon_mutex); +} + +/* Resets the Xen limit, sets new target, and kicks off processing. */ +void balloon_set_new_target(unsigned long target) +{ + /* No need for lock. Not read-modify-write updates. */ + bs.target_pages = max(target, balloon_minimum_target()); + schedule_work(&balloon_worker); +} + +static struct xenbus_watch target_watch = +{ + .node = "memory/target" +}; + +/* React to a change in the target key */ +static void watch_target(struct xenbus_watch *watch, + const char **vec, unsigned int len) +{ + unsigned long long new_target; + int err; + + err = xenbus_scanf(XBT_NIL, "memory", "target", "%llu", &new_target); + if (err != 1) { + /* This is ok (for domain0 at least) - so just return */ + return; + } + + /* The given memory/target value is in KiB, so it needs converting to + * pages. PAGE_SHIFT converts bytes to pages, hence PAGE_SHIFT - 10. + */ + balloon_set_new_target(new_target >> (PAGE_SHIFT - 10)); +} + +static int balloon_init_watcher(struct notifier_block *notifier, + unsigned long event, + void *data) +{ + int err; + + err = register_xenbus_watch(&target_watch); + if (err) + printk(KERN_ERR "Failed to set balloon watcher\n"); + + return NOTIFY_DONE; +} + +#ifdef CONFIG_PROC_FS +static int balloon_write(struct file *file, const char __user *buffer, + unsigned long count, void *data) +{ + char memstring[64], *endchar; + unsigned long long target_bytes; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (count <= 1) + return -EBADMSG; /* runt */ + if (count > sizeof(memstring)) + return -EFBIG; /* too long */ + + if (copy_from_user(memstring, buffer, count)) + return -EFAULT; + memstring[sizeof(memstring)-1] = '\0'; + + target_bytes = memparse(memstring, &endchar); + balloon_set_new_target(target_bytes >> PAGE_SHIFT); + + return count; +} + +static int balloon_read(char *page, char **start, off_t off, + int count, int *eof, void *data) +{ + int len; + + len = sprintf( + page, + "Current allocation: %8lu kB\n" + "Requested target: %8lu kB\n" + "Minimum target: %8lu kB\n" + "Maximum target: %8lu kB\n" + "Low-mem balloon: %8lu kB\n" + "High-mem balloon: %8lu kB\n" + "Driver pages: %8lu kB\n", + PAGES2KB(bs.current_pages), PAGES2KB(bs.target_pages), + PAGES2KB(balloon_minimum_target()), PAGES2KB(num_physpages), + PAGES2KB(bs.balloon_low), PAGES2KB(bs.balloon_high), + PAGES2KB(bs.driver_pages)); + + + *eof = 1; + return len; +} +#endif + +static struct notifier_block xenstore_notifier; + +static int __init balloon_init(void) +{ +#if !defined(CONFIG_XEN) +# ifndef XENMEM_get_pod_target +# define XENMEM_get_pod_target 17 + typedef struct xen_pod_target { + uint64_t target_pages; + uint64_t tot_pages; + uint64_t pod_cache_pages; + uint64_t pod_entries; + domid_t domid; + } xen_pod_target_t; +# endif + xen_pod_target_t pod_target = { .domid = DOMID_SELF }; + int rc; +#elif defined(CONFIG_X86) + unsigned long pfn; + struct page *page; +#endif + + if (!is_running_on_xen()) + return -ENODEV; + + IPRINTK("Initialising balloon driver.\n"); + +#ifdef CONFIG_XEN + pagevec_init(&free_pagevec, true); + bs.current_pages = min(xen_start_info->nr_pages, max_pfn); + totalram_pages = bs.current_pages; +#else + rc = HYPERVISOR_memory_op(XENMEM_get_pod_target, &pod_target); + /* + * Xen prior to 3.4.0 masks the memory_op command to 4 bits, thus + * converting XENMEM_get_pod_target to XENMEM_decrease_reservation. + * Fortunately this results in a request with all input fields zero, + * but (due to the way bit 4 and upwards get interpreted) a starting + * extent of 1. When start_extent > nr_extents (>= in newer Xen), we + * simply get start_extent returned. + */ + totalram_bias = HYPERVISOR_memory_op(rc != -ENOSYS && rc != 1 + ? XENMEM_maximum_reservation : XENMEM_current_reservation, + &pod_target.domid); + if ((long)totalram_bias != -ENOSYS) { + BUG_ON(totalram_bias < totalram_pages); + bs.current_pages = totalram_bias; + totalram_bias -= totalram_pages; + } else { + totalram_bias = 0; + bs.current_pages = totalram_pages; + } +#endif + bs.target_pages = bs.current_pages; + bs.balloon_low = 0; + bs.balloon_high = 0; + bs.driver_pages = 0UL; + + init_timer(&balloon_timer); + balloon_timer.data = 0; + balloon_timer.function = balloon_alarm; + +#ifdef CONFIG_PROC_FS + if ((balloon_pde = create_xen_proc_entry("balloon", 0644)) == NULL) { + WPRINTK("Unable to create /proc/xen/balloon.\n"); + return -1; + } + + balloon_pde->read_proc = balloon_read; + balloon_pde->write_proc = balloon_write; +#endif + balloon_sysfs_init(); + +#if defined(CONFIG_X86) && defined(CONFIG_XEN) + /* Initialise the balloon with excess memory space. */ + for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) { + page = pfn_to_page(pfn); + if (!PageReserved(page)) { + SetPageReserved(page); + set_phys_to_machine(pfn, INVALID_P2M_ENTRY); + balloon_append(page, 0); + } + } +#endif + + target_watch.callback = watch_target; + xenstore_notifier.notifier_call = balloon_init_watcher; + + register_xenstore_notifier(&xenstore_notifier); + + return 0; +} + +subsys_initcall(balloon_init); + +static void __exit balloon_exit(void) +{ + balloon_sysfs_exit(); + /* XXX - release balloon here */ +} + +module_exit(balloon_exit); + +void balloon_update_driver_allowance(long delta) +{ + unsigned long flags; + + balloon_lock(flags); + bs.driver_pages += delta; + balloon_unlock(flags); +} +EXPORT_SYMBOL_GPL(balloon_update_driver_allowance); + +#if defined(CONFIG_XEN_BACKEND) || defined(CONFIG_XEN_BACKEND_MODULE) + +#ifdef CONFIG_XEN +static int dealloc_pte_fn( + pte_t *pte, struct page *pmd_page, unsigned long addr, void *data) +{ + unsigned long pfn, mfn = pte_mfn(*pte); + int ret; + struct xen_memory_reservation reservation = { + .nr_extents = 1, + .extent_order = 0, + .domid = DOMID_SELF + }; + set_xen_guest_handle(reservation.extent_start, &mfn); + set_pte_at(&init_mm, addr, pte, __pte_ma(0)); + pfn = __pa(addr) >> PAGE_SHIFT; + set_phys_to_machine(pfn, INVALID_P2M_ENTRY); + SetPageReserved(pfn_to_page(pfn)); + ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation); + BUG_ON(ret != 1); + return 0; +} +#endif + +struct page **alloc_empty_pages_and_pagevec(int nr_pages) +{ + unsigned long flags; + void *v; + struct page *page, **pagevec; + int i, ret; + + pagevec = kmalloc(sizeof(page) * nr_pages, GFP_KERNEL); + if (pagevec == NULL) + return NULL; + + for (i = 0; i < nr_pages; i++) { + balloon_lock(flags); + page = balloon_first_page(); + if (page && !PageHighMem(page)) { + UNLIST_PAGE(page); + bs.balloon_low--; + balloon_unlock(flags); + pagevec[i] = page; + continue; + } + balloon_unlock(flags); + + page = pagevec[i] = alloc_page(GFP_KERNEL|__GFP_COLD); + if (page == NULL) + goto err; + + v = page_address(page); + scrub_pages(v, 1); + + balloon_lock(flags); + + if (xen_feature(XENFEAT_auto_translated_physmap)) { + unsigned long gmfn = page_to_pfn(page); + struct xen_memory_reservation reservation = { + .nr_extents = 1, + .extent_order = 0, + .domid = DOMID_SELF + }; + set_xen_guest_handle(reservation.extent_start, &gmfn); + ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, + &reservation); + if (ret == 1) + ret = 0; /* success */ + } else { +#ifdef CONFIG_XEN + ret = apply_to_page_range(&init_mm, (unsigned long)v, + PAGE_SIZE, dealloc_pte_fn, + NULL); +#else + /* Cannot handle non-auto translate mode. */ + ret = 1; +#endif + } + + if (ret != 0) { + balloon_free_page(page); + balloon_free_and_unlock(flags); + goto err; + } + + totalram_pages = --bs.current_pages - totalram_bias; + if (PageHighMem(page)) + dec_totalhigh_pages(); + page_zone(page)->present_pages--; + + balloon_unlock(flags); + } + + out: + schedule_work(&balloon_worker); +#ifdef CONFIG_XEN + flush_tlb_all(); +#endif + return pagevec; + + err: + balloon_lock(flags); + while (--i >= 0) + balloon_append(pagevec[i], 0); + balloon_unlock(flags); + kfree(pagevec); + pagevec = NULL; + goto out; +} +EXPORT_SYMBOL_GPL(alloc_empty_pages_and_pagevec); + +#endif /* CONFIG_XEN_BACKEND */ + +#ifdef CONFIG_XEN +static void _free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages, + bool free_vec) +{ + unsigned long flags; + int i; + + if (pagevec == NULL) + return; + + balloon_lock(flags); + for (i = 0; i < nr_pages; i++) { + BUG_ON(page_count(pagevec[i]) != 1); + balloon_append(pagevec[i], !free_vec); + } + if (!free_vec) { + bs.current_pages -= nr_pages; + totalram_pages = bs.current_pages - totalram_bias; + } + balloon_unlock(flags); + + if (free_vec) + kfree(pagevec); + + schedule_work(&balloon_worker); +} + +void free_empty_pages(struct page **pagevec, int nr_pages) +{ + _free_empty_pages_and_pagevec(pagevec, nr_pages, false); +} +#endif + +#if defined(CONFIG_XEN_BACKEND) || defined(CONFIG_XEN_BACKEND_MODULE) +void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages) +{ + _free_empty_pages_and_pagevec(pagevec, nr_pages, true); +} +EXPORT_SYMBOL_GPL(free_empty_pages_and_pagevec); +#endif + +void balloon_release_driver_page(struct page *page) +{ + unsigned long flags; + + balloon_lock(flags); + balloon_append(page, 1); + totalram_pages = --bs.current_pages - totalram_bias; + bs.driver_pages--; + balloon_unlock(flags); + + schedule_work(&balloon_worker); +} +EXPORT_SYMBOL_GPL(balloon_release_driver_page); + +MODULE_LICENSE("Dual BSD/GPL"); diff --git a/drivers/xenpvdrivers/blkfront.c b/drivers/xenpvdrivers/blkfront.c new file mode 100644 index 0000000..6c357cd --- /dev/null +++ b/drivers/xenpvdrivers/blkfront.c @@ -0,0 +1,981 @@ +/****************************************************************************** + * blkfront.c + * + * XenLinux virtual block-device driver. + * + * Copyright (c) 2003-2004, Keir Fraser & Steve Hand + * Modifications by Mark A. Williamson are (c) Intel Research Cambridge + * Copyright (c) 2004, Christian Limpach + * Copyright (c) 2004, Andrew Warfield + * Copyright (c) 2005, Christopher Clark + * Copyright (c) 2005, XenSource Ltd + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include "block.h" +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#define BLKIF_STATE_DISCONNECTED 0 +#define BLKIF_STATE_CONNECTED 1 +#define BLKIF_STATE_SUSPENDED 2 + +#define MAXIMUM_OUTSTANDING_BLOCK_REQS \ + (BLKIF_MAX_SEGMENTS_PER_REQUEST * BLK_RING_SIZE) +#define GRANT_INVALID_REF 0 + +static void connect(struct blkfront_info *); +static void blkfront_closing(struct blkfront_info *); +static int blkfront_remove(struct xenbus_device *); +static int talk_to_backend(struct xenbus_device *, struct blkfront_info *); +static int setup_blkring(struct xenbus_device *, struct blkfront_info *); + +static void kick_pending_request_queues(struct blkfront_info *); + +static irqreturn_t blkif_int(int irq, void *dev_id); +static void blkif_restart_queue(struct work_struct *arg); +static void blkif_recover(struct blkfront_info *); +static void blkif_completion(struct blk_shadow *); +static void blkif_free(struct blkfront_info *, int); + + +/** + * Entry point to this code when a new device is created. Allocate the basic + * structures and the ring buffer for communication with the backend, and + * inform the backend of the appropriate details for those. Switch to + * Initialised state. + */ +static int blkfront_probe(struct xenbus_device *dev, + const struct xenbus_device_id *id) +{ + int err, vdevice, i; + struct blkfront_info *info; + + /* FIXME: Use dynamic device id if this is not set. */ + err = xenbus_scanf(XBT_NIL, dev->nodename, + "virtual-device", "%i", &vdevice); + if (err != 1) { + /* go looking in the extended area instead */ + err = xenbus_scanf(XBT_NIL, dev->nodename, "virtual-device-ext", + "%i", &vdevice); + if (err != 1) { + xenbus_dev_fatal(dev, err, "reading virtual-device"); + return err; + } + } + + info = kzalloc(sizeof(*info), GFP_KERNEL); + if (!info) { + xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure"); + return -ENOMEM; + } + + info->xbdev = dev; + info->vdevice = vdevice; + info->connected = BLKIF_STATE_DISCONNECTED; + INIT_WORK(&info->work, blkif_restart_queue); + + for (i = 0; i < BLK_RING_SIZE; i++) + info->shadow[i].req.id = i+1; + info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff; + + /* Front end dir is a number, which is used as the id. */ + info->handle = simple_strtoul(strrchr(dev->nodename,'/')+1, NULL, 0); + dev_set_drvdata(&dev->dev, info); + + err = talk_to_backend(dev, info); + if (err) { + kfree(info); + dev_set_drvdata(&dev->dev, NULL); + return err; + } + + return 0; +} + + +/** + * We are reconnecting to the backend, due to a suspend/resume, or a backend + * driver restart. We tear down our blkif structure and recreate it, but + * leave the device-layer structures intact so that this is transparent to the + * rest of the kernel. + */ +static int blkfront_resume(struct xenbus_device *dev) +{ + struct blkfront_info *info = dev_get_drvdata(&dev->dev); + int err; + + DPRINTK("blkfront_resume: %s\n", dev->nodename); + + blkif_free(info, info->connected == BLKIF_STATE_CONNECTED); + + err = talk_to_backend(dev, info); + if (info->connected == BLKIF_STATE_SUSPENDED && !err) + blkif_recover(info); + + return err; +} + + +/* Common code used when first setting up, and when resuming. */ +static int talk_to_backend(struct xenbus_device *dev, + struct blkfront_info *info) +{ + const char *message = NULL; + struct xenbus_transaction xbt; + int err; + + /* Create shared ring, alloc event channel. */ + err = setup_blkring(dev, info); + if (err) + goto out; + +again: + err = xenbus_transaction_start(&xbt); + if (err) { + xenbus_dev_fatal(dev, err, "starting transaction"); + goto destroy_blkring; + } + + err = xenbus_printf(xbt, dev->nodename, + "ring-ref","%u", info->ring_ref); + if (err) { + message = "writing ring-ref"; + goto abort_transaction; + } + err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u", + irq_to_evtchn_port(info->irq)); + if (err) { + message = "writing event-channel"; + goto abort_transaction; + } + err = xenbus_printf(xbt, dev->nodename, "protocol", "%s", + XEN_IO_PROTO_ABI_NATIVE); + if (err) { + message = "writing protocol"; + goto abort_transaction; + } + + err = xenbus_transaction_end(xbt, 0); + if (err) { + if (err == -EAGAIN) + goto again; + xenbus_dev_fatal(dev, err, "completing transaction"); + goto destroy_blkring; + } + + xenbus_switch_state(dev, XenbusStateInitialised); + + return 0; + + abort_transaction: + xenbus_transaction_end(xbt, 1); + if (message) + xenbus_dev_fatal(dev, err, "%s", message); + destroy_blkring: + blkif_free(info, 0); + out: + return err; +} + + +static int setup_blkring(struct xenbus_device *dev, + struct blkfront_info *info) +{ + blkif_sring_t *sring; + int err; + + info->ring_ref = GRANT_INVALID_REF; + + sring = (blkif_sring_t *)__get_free_page(GFP_NOIO | __GFP_HIGH); + if (!sring) { + xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring"); + return -ENOMEM; + } + SHARED_RING_INIT(sring); + FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE); + + sg_init_table(info->sg, BLKIF_MAX_SEGMENTS_PER_REQUEST); + + err = xenbus_grant_ring(dev, virt_to_mfn(info->ring.sring)); + if (err < 0) { + free_page((unsigned long)sring); + info->ring.sring = NULL; + goto fail; + } + info->ring_ref = err; + + err = bind_listening_port_to_irqhandler( + dev->otherend_id, blkif_int, IRQF_SAMPLE_RANDOM, "blkif", info); + if (err <= 0) { + xenbus_dev_fatal(dev, err, + "bind_listening_port_to_irqhandler"); + goto fail; + } + info->irq = err; + + return 0; +fail: + blkif_free(info, 0); + return err; +} + + +/** + * Callback received when the backend's state changes. + */ +static void backend_changed(struct xenbus_device *dev, + enum xenbus_state backend_state) +{ + struct blkfront_info *info = dev_get_drvdata(&dev->dev); + struct block_device *bd; + + DPRINTK("blkfront:backend_changed.\n"); + + switch (backend_state) { + case XenbusStateInitialising: + case XenbusStateInitWait: + case XenbusStateInitialised: + case XenbusStateReconfiguring: + case XenbusStateReconfigured: + case XenbusStateUnknown: + case XenbusStateClosed: + break; + + case XenbusStateConnected: + connect(info); + break; + + case XenbusStateClosing: + if (!info->gd) { + xenbus_frontend_closed(dev); + break; + } + bd = bdget_disk(info->gd, 0); + if (bd == NULL) { + xenbus_dev_fatal(dev, -ENODEV, "bdget failed"); + break; + } + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) + down(&bd->bd_sem); +#else + mutex_lock(&bd->bd_mutex); +#endif + if (info->users > 0) + xenbus_dev_error(dev, -EBUSY, + "Device in use; refusing to close"); + else + blkfront_closing(info); +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) + up(&bd->bd_sem); +#else + mutex_unlock(&bd->bd_mutex); +#endif + bdput(bd); + break; + } +} + + +/* ** Connection ** */ + + +/* + * Invoked when the backend is finally 'ready' (and has told produced + * the details about the physical device - #sectors, size, etc). + */ +static void connect(struct blkfront_info *info) +{ + unsigned long long sectors; + unsigned long sector_size; + unsigned int binfo; + int err; + + switch (info->connected) { + case BLKIF_STATE_CONNECTED: + /* + * Potentially, the back-end may be signalling + * a capacity change; update the capacity. + */ + err = xenbus_scanf(XBT_NIL, info->xbdev->otherend, + "sectors", "%Lu", §ors); + if (XENBUS_EXIST_ERR(err)) + return; + printk(KERN_INFO "Setting capacity to %Lu\n", + sectors); + set_capacity(info->gd, sectors); + revalidate_disk(info->gd); + + /* fall through */ + case BLKIF_STATE_SUSPENDED: + return; + } + + DPRINTK("blkfront.c:connect:%s.\n", info->xbdev->otherend); + + err = xenbus_gather(XBT_NIL, info->xbdev->otherend, + "sectors", "%Lu", §ors, + "info", "%u", &binfo, + "sector-size", "%lu", §or_size, + NULL); + if (err) { + xenbus_dev_fatal(info->xbdev, err, + "reading backend fields at %s", + info->xbdev->otherend); + return; + } + + err = xenbus_gather(XBT_NIL, info->xbdev->otherend, + "feature-barrier", "%lu", &info->feature_barrier, + NULL); + if (err) + info->feature_barrier = 0; + + err = xlvbd_add(sectors, info->vdevice, binfo, sector_size, info); + if (err) { + xenbus_dev_fatal(info->xbdev, err, "xlvbd_add at %s", + info->xbdev->otherend); + return; + } + + err = xlvbd_sysfs_addif(info); + if (err) { + xenbus_dev_fatal(info->xbdev, err, "xlvbd_sysfs_addif at %s", + info->xbdev->otherend); + return; + } + + (void)xenbus_switch_state(info->xbdev, XenbusStateConnected); + + /* Kick pending requests. */ + spin_lock_irq(&blkif_io_lock); + info->connected = BLKIF_STATE_CONNECTED; + kick_pending_request_queues(info); + spin_unlock_irq(&blkif_io_lock); + + add_disk(info->gd); + + info->is_ready = 1; + + register_vcd(info); +} + +/** + * Handle the change of state of the backend to Closing. We must delete our + * device-layer structures now, to ensure that writes are flushed through to + * the backend. Once is this done, we can switch to Closed in + * acknowledgement. + */ +static void blkfront_closing(struct blkfront_info *info) +{ + unsigned long flags; + + DPRINTK("blkfront_closing: %d removed\n", info->vdevice); + + if (info->rq == NULL) + goto out; + + spin_lock_irqsave(&blkif_io_lock, flags); + /* No more blkif_request(). */ + blk_stop_queue(info->rq); + /* No more gnttab callback work. */ + gnttab_cancel_free_callback(&info->callback); + spin_unlock_irqrestore(&blkif_io_lock, flags); + + /* Flush gnttab callback work. Must be done with no locks held. */ + flush_scheduled_work(); + + xlvbd_sysfs_delif(info); + + unregister_vcd(info); + + xlvbd_del(info); + + out: + if (info->xbdev) + xenbus_frontend_closed(info->xbdev); +} + + +static int blkfront_remove(struct xenbus_device *dev) +{ + struct blkfront_info *info = dev_get_drvdata(&dev->dev); + + DPRINTK("blkfront_remove: %s removed\n", dev->nodename); + + blkif_free(info, 0); + + if(info->users == 0) + kfree(info); + else + info->xbdev = NULL; + + return 0; +} + + +static inline int GET_ID_FROM_FREELIST( + struct blkfront_info *info) +{ + unsigned long free = info->shadow_free; + BUG_ON(free >= BLK_RING_SIZE); + info->shadow_free = info->shadow[free].req.id; + info->shadow[free].req.id = 0x0fffffee; /* debug */ + return free; +} + +static inline void ADD_ID_TO_FREELIST( + struct blkfront_info *info, unsigned long id) +{ + info->shadow[id].req.id = info->shadow_free; + info->shadow[id].request = 0; + info->shadow_free = id; +} + +static inline void flush_requests(struct blkfront_info *info) +{ + int notify; + + RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->ring, notify); + + if (notify) + notify_remote_via_irq(info->irq); +} + +static void kick_pending_request_queues(struct blkfront_info *info) +{ + if (!RING_FULL(&info->ring)) { + /* Re-enable calldowns. */ + blk_start_queue(info->rq); + /* Kick things off immediately. */ + do_blkif_request(info->rq); + } +} + +static void blkif_restart_queue(struct work_struct *arg) +{ + struct blkfront_info *info = container_of(arg, struct blkfront_info, work); + spin_lock_irq(&blkif_io_lock); + if (info->connected == BLKIF_STATE_CONNECTED) + kick_pending_request_queues(info); + spin_unlock_irq(&blkif_io_lock); +} + +static void blkif_restart_queue_callback(void *arg) +{ + struct blkfront_info *info = (struct blkfront_info *)arg; + schedule_work(&info->work); +} + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) +int blkif_open(struct inode *inode, struct file *filep) +{ + struct block_device *bd = inode->i_bdev; +#else +int blkif_open(struct block_device *bd, fmode_t mode) +{ +#endif + struct blkfront_info *info = bd->bd_disk->private_data; + + if (!info->xbdev) + return -ENODEV; + info->users++; + return 0; +} + + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) +int blkif_release(struct inode *inode, struct file *filep) +{ + struct gendisk *disk = inode->i_bdev->bd_disk; +#else +int blkif_release(struct gendisk *disk, fmode_t mode) +{ +#endif + struct blkfront_info *info = disk->private_data; + + info->users--; + if (info->users == 0) { + /* Check whether we have been instructed to close. We will + have ignored this request initially, as the device was + still mounted. */ + struct xenbus_device * dev = info->xbdev; + + if (!dev) { + blkfront_closing(info); + kfree(info); + } else if (xenbus_read_driver_state(dev->otherend) + == XenbusStateClosing && info->is_ready) + blkfront_closing(info); + } + return 0; +} + + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) +int blkif_ioctl(struct inode *inode, struct file *filep, + unsigned command, unsigned long argument) +{ + struct block_device *bd = inode->i_bdev; +#else +int blkif_ioctl(struct block_device *bd, fmode_t mode, + unsigned command, unsigned long argument) +{ +#endif + int i; + + DPRINTK_IOCTL("command: 0x%x, argument: 0x%lx, dev: 0x%04x\n", + command, (long)argument, inode->i_rdev); + + switch (command) { +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16) + case HDIO_GETGEO: { + struct hd_geometry geo; + int ret; + + if (!argument) + return -EINVAL; + + geo.start = get_start_sect(bd); + ret = blkif_getgeo(bd, &geo); + if (ret) + return ret; + + if (copy_to_user((struct hd_geometry __user *)argument, &geo, + sizeof(geo))) + return -EFAULT; + + return 0; + } +#endif + case CDROMMULTISESSION: + DPRINTK("FIXME: support multisession CDs later\n"); + for (i = 0; i < sizeof(struct cdrom_multisession); i++) + if (put_user(0, (char __user *)(argument + i))) + return -EFAULT; + return 0; + + case CDROM_GET_CAPABILITY: { + struct blkfront_info *info = bd->bd_disk->private_data; + struct gendisk *gd = info->gd; + if (gd->flags & GENHD_FL_CD) + return 0; + return -EINVAL; + } + default: + /*printk(KERN_ALERT "ioctl %08x not supported by Xen blkdev\n", + command);*/ + return -EINVAL; /* same return as native Linux */ + } + + return 0; +} + + +int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg) +{ + /* We don't have real geometry info, but let's at least return + values consistent with the size of the device */ + sector_t nsect = get_capacity(bd->bd_disk); + sector_t cylinders = nsect; + + hg->heads = 0xff; + hg->sectors = 0x3f; + sector_div(cylinders, hg->heads * hg->sectors); + hg->cylinders = cylinders; + if ((sector_t)(hg->cylinders + 1) * hg->heads * hg->sectors < nsect) + hg->cylinders = 0xffff; + return 0; +} + + +/* + * blkif_queue_request + * + * request block io + * + * id: for guest use only. + * operation: BLKIF_OP_{READ,WRITE,PROBE} + * buffer: buffer to read/write into. this should be a + * virtual address in the guest os. + */ +static int blkif_queue_request(struct request *req) +{ + struct blkfront_info *info = req->rq_disk->private_data; + unsigned long buffer_mfn; + blkif_request_t *ring_req; + unsigned long id; + unsigned int fsect, lsect; + int i, ref; + grant_ref_t gref_head; + struct scatterlist *sg; + + if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) + return 1; + + if (gnttab_alloc_grant_references( + BLKIF_MAX_SEGMENTS_PER_REQUEST, &gref_head) < 0) { + gnttab_request_free_callback( + &info->callback, + blkif_restart_queue_callback, + info, + BLKIF_MAX_SEGMENTS_PER_REQUEST); + return 1; + } + + /* Fill out a communications ring structure. */ + ring_req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt); + id = GET_ID_FROM_FREELIST(info); + info->shadow[id].request = (unsigned long)req; + + ring_req->id = id; + ring_req->sector_number = (blkif_sector_t)blk_rq_pos(req); + ring_req->handle = info->handle; + + ring_req->operation = rq_data_dir(req) ? + BLKIF_OP_WRITE : BLKIF_OP_READ; + if (blk_barrier_rq(req)) + ring_req->operation = BLKIF_OP_WRITE_BARRIER; + if (blk_pc_request(req)) + ring_req->operation = BLKIF_OP_PACKET; + + ring_req->nr_segments = blk_rq_map_sg(req->q, req, info->sg); + BUG_ON(ring_req->nr_segments > BLKIF_MAX_SEGMENTS_PER_REQUEST); + for_each_sg(info->sg, sg, ring_req->nr_segments, i) { + buffer_mfn = page_to_phys(sg_page(sg)) >> PAGE_SHIFT; + fsect = sg->offset >> 9; + lsect = fsect + (sg->length >> 9) - 1; + /* install a grant reference. */ + ref = gnttab_claim_grant_reference(&gref_head); + BUG_ON(ref == -ENOSPC); + + gnttab_grant_foreign_access_ref( + ref, + info->xbdev->otherend_id, + buffer_mfn, + rq_data_dir(req) ? GTF_readonly : 0 ); + + info->shadow[id].frame[i] = mfn_to_pfn(buffer_mfn); + ring_req->seg[i] = + (struct blkif_request_segment) { + .gref = ref, + .first_sect = fsect, + .last_sect = lsect }; + } + + info->ring.req_prod_pvt++; + + /* Keep a private copy so we can reissue requests when recovering. */ + info->shadow[id].req = *ring_req; + + gnttab_free_grant_references(gref_head); + + return 0; +} + +/* + * do_blkif_request + * read a block; request is in a request queue + */ +void do_blkif_request(struct request_queue *rq) +{ + struct blkfront_info *info = NULL; + struct request *req; + int queued; + + DPRINTK("Entered do_blkif_request\n"); + + queued = 0; + + while ((req = blk_peek_request(rq)) != NULL) { + info = req->rq_disk->private_data; + + if (RING_FULL(&info->ring)) + goto wait; + + blk_start_request(req); + + if (!blk_fs_request(req) && !blk_pc_request(req)) { + __blk_end_request_all(req, -EIO); + continue; + } + + DPRINTK("do_blk_req %p: cmd %p, sec %llx, " + "(%u/%u) buffer:%p [%s]\n", + req, req->cmd, (long long)blk_rq_pos(req), + blk_rq_cur_sectors(req), blk_rq_sectors(req), + req->buffer, rq_data_dir(req) ? "write" : "read"); + + if (blkif_queue_request(req)) { + blk_requeue_request(rq, req); + wait: + /* Avoid pointless unplugs. */ + blk_stop_queue(rq); + break; + } + + queued++; + } + + if (queued != 0) + flush_requests(info); +} + + +static irqreturn_t blkif_int(int irq, void *dev_id) +{ + struct request *req; + blkif_response_t *bret; + RING_IDX i, rp; + unsigned long flags; + struct blkfront_info *info = (struct blkfront_info *)dev_id; + + spin_lock_irqsave(&blkif_io_lock, flags); + + if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) { + spin_unlock_irqrestore(&blkif_io_lock, flags); + return IRQ_HANDLED; + } + + again: + rp = info->ring.sring->rsp_prod; + rmb(); /* Ensure we see queued responses up to 'rp'. */ + + for (i = info->ring.rsp_cons; i != rp; i++) { + unsigned long id; + int ret; + + bret = RING_GET_RESPONSE(&info->ring, i); + id = bret->id; + req = (struct request *)info->shadow[id].request; + + blkif_completion(&info->shadow[id]); + + ADD_ID_TO_FREELIST(info, id); + + ret = bret->status == BLKIF_RSP_OKAY ? 0 : -EIO; + switch (bret->operation) { + case BLKIF_OP_WRITE_BARRIER: + if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) { + printk("blkfront: %s: write barrier op failed\n", + info->gd->disk_name); + ret = -EOPNOTSUPP; + info->feature_barrier = 0; + xlvbd_barrier(info); + } + /* fall through */ + case BLKIF_OP_READ: + case BLKIF_OP_WRITE: + case BLKIF_OP_PACKET: + if (unlikely(bret->status != BLKIF_RSP_OKAY)) + DPRINTK("Bad return from blkdev data " + "request: %x\n", bret->status); + + __blk_end_request_all(req, ret); + break; + default: + BUG(); + } + } + + info->ring.rsp_cons = i; + + if (i != info->ring.req_prod_pvt) { + int more_to_do; + RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do); + if (more_to_do) + goto again; + } else + info->ring.sring->rsp_event = i + 1; + + kick_pending_request_queues(info); + + spin_unlock_irqrestore(&blkif_io_lock, flags); + + return IRQ_HANDLED; +} + +static void blkif_free(struct blkfront_info *info, int suspend) +{ + /* Prevent new requests being issued until we fix things up. */ + spin_lock_irq(&blkif_io_lock); + info->connected = suspend ? + BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED; + /* No more blkif_request(). */ + if (info->rq) + blk_stop_queue(info->rq); + /* No more gnttab callback work. */ + gnttab_cancel_free_callback(&info->callback); + spin_unlock_irq(&blkif_io_lock); + + /* Flush gnttab callback work. Must be done with no locks held. */ + flush_scheduled_work(); + + /* Free resources associated with old device channel. */ + if (info->ring_ref != GRANT_INVALID_REF) { + gnttab_end_foreign_access(info->ring_ref, + (unsigned long)info->ring.sring); + info->ring_ref = GRANT_INVALID_REF; + info->ring.sring = NULL; + } + if (info->irq) + unbind_from_irqhandler(info->irq, info); + info->irq = 0; +} + +static void blkif_completion(struct blk_shadow *s) +{ + int i; + for (i = 0; i < s->req.nr_segments; i++) + gnttab_end_foreign_access(s->req.seg[i].gref, 0UL); +} + +static void blkif_recover(struct blkfront_info *info) +{ + int i; + blkif_request_t *req; + struct blk_shadow *copy; + int j; + + /* Stage 1: Make a safe copy of the shadow state. */ + copy = kmalloc(sizeof(info->shadow), GFP_NOIO | __GFP_NOFAIL | __GFP_HIGH); + memcpy(copy, info->shadow, sizeof(info->shadow)); + + /* Stage 2: Set up free list. */ + memset(&info->shadow, 0, sizeof(info->shadow)); + for (i = 0; i < BLK_RING_SIZE; i++) + info->shadow[i].req.id = i+1; + info->shadow_free = info->ring.req_prod_pvt; + info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff; + + /* Stage 3: Find pending requests and requeue them. */ + for (i = 0; i < BLK_RING_SIZE; i++) { + /* Not in use? */ + if (copy[i].request == 0) + continue; + + /* Grab a request slot and copy shadow state into it. */ + req = RING_GET_REQUEST( + &info->ring, info->ring.req_prod_pvt); + *req = copy[i].req; + + /* We get a new request id, and must reset the shadow state. */ + req->id = GET_ID_FROM_FREELIST(info); + memcpy(&info->shadow[req->id], ©[i], sizeof(copy[i])); + + /* Rewrite any grant references invalidated by susp/resume. */ + for (j = 0; j < req->nr_segments; j++) + gnttab_grant_foreign_access_ref( + req->seg[j].gref, + info->xbdev->otherend_id, + pfn_to_mfn(info->shadow[req->id].frame[j]), + rq_data_dir((struct request *) + info->shadow[req->id].request) ? + GTF_readonly : 0); + info->shadow[req->id].req = *req; + + info->ring.req_prod_pvt++; + } + + kfree(copy); + + (void)xenbus_switch_state(info->xbdev, XenbusStateConnected); + + spin_lock_irq(&blkif_io_lock); + + /* Now safe for us to use the shared ring */ + info->connected = BLKIF_STATE_CONNECTED; + + /* Send off requeued requests */ + flush_requests(info); + + /* Kick any other new requests queued since we resumed */ + kick_pending_request_queues(info); + + spin_unlock_irq(&blkif_io_lock); +} + +int blkfront_is_ready(struct xenbus_device *dev) +{ + struct blkfront_info *info = dev_get_drvdata(&dev->dev); + + return info->is_ready && info->xbdev; +} + + +/* ** Driver Registration ** */ + + +static const struct xenbus_device_id blkfront_ids[] = { + { "vbd" }, + { "" } +}; +MODULE_ALIAS("xen:vbd"); + +static struct xenbus_driver blkfront = { + .name = "vbd", + .ids = blkfront_ids, + .probe = blkfront_probe, + .remove = blkfront_remove, + .resume = blkfront_resume, + .otherend_changed = backend_changed, + .is_ready = blkfront_is_ready, +}; + + +static int __init xlblk_init(void) +{ + if (!is_running_on_xen()) + return -ENODEV; + + return xenbus_register_frontend(&blkfront); +} +module_init(xlblk_init); + + +static void __exit xlblk_exit(void) +{ + return xenbus_unregister_driver(&blkfront); +} +module_exit(xlblk_exit); + +MODULE_LICENSE("Dual BSD/GPL"); diff --git a/drivers/xenpvdrivers/block.h b/drivers/xenpvdrivers/block.h new file mode 100644 index 0000000..b81cccc --- /dev/null +++ b/drivers/xenpvdrivers/block.h @@ -0,0 +1,170 @@ +/****************************************************************************** + * block.h + * + * Shared definitions between all levels of XenLinux Virtual block devices. + * + * Copyright (c) 2003-2004, Keir Fraser & Steve Hand + * Modifications by Mark A. Williamson are (c) Intel Research Cambridge + * Copyright (c) 2004-2005, Christian Limpach + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __XEN_DRIVERS_BLOCK_H__ +#define __XEN_DRIVERS_BLOCK_H__ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define DPRINTK(_f, _a...) pr_debug(_f, ## _a) + +#if 0 +#define DPRINTK_IOCTL(_f, _a...) printk(KERN_ALERT _f, ## _a) +#else +#define DPRINTK_IOCTL(_f, _a...) ((void)0) +#endif + +struct xlbd_type_info +{ + int partn_shift; + int disks_per_major; + char *devname; + char *diskname; +}; + +struct xlbd_major_info +{ + int major; + int index; + int usage; + struct xlbd_type_info *type; + struct xlbd_minor_state *minors; +}; + +struct blk_shadow { + blkif_request_t req; + unsigned long request; + unsigned long frame[BLKIF_MAX_SEGMENTS_PER_REQUEST]; +}; + +#define BLK_RING_SIZE __CONST_RING_SIZE(blkif, PAGE_SIZE) + +/* + * We have one of these per vbd, whether ide, scsi or 'other'. They + * hang in private_data off the gendisk structure. We may end up + * putting all kinds of interesting stuff here :-) + */ +struct blkfront_info +{ + struct xenbus_device *xbdev; + struct gendisk *gd; + int vdevice; + blkif_vdev_t handle; + int connected; + int ring_ref; + blkif_front_ring_t ring; + struct scatterlist sg[BLKIF_MAX_SEGMENTS_PER_REQUEST]; + unsigned int irq; + struct xlbd_major_info *mi; + struct request_queue *rq; + struct work_struct work; + struct gnttab_free_callback callback; + struct blk_shadow shadow[BLK_RING_SIZE]; + unsigned long shadow_free; + int feature_barrier; + int is_ready; + + /** + * The number of people holding this device open. We won't allow a + * hot-unplug unless this is 0. + */ + int users; +}; + +extern spinlock_t blkif_io_lock; + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) +extern int blkif_open(struct inode *inode, struct file *filep); +extern int blkif_release(struct inode *inode, struct file *filep); +extern int blkif_ioctl(struct inode *inode, struct file *filep, + unsigned command, unsigned long argument); +#else +extern int blkif_open(struct block_device *bdev, fmode_t mode); +extern int blkif_release(struct gendisk *disk, fmode_t mode); +extern int blkif_ioctl(struct block_device *bdev, fmode_t mode, + unsigned command, unsigned long argument); +#endif +extern int blkif_getgeo(struct block_device *, struct hd_geometry *); +extern int blkif_check(dev_t dev); +extern int blkif_revalidate(dev_t dev); +extern void do_blkif_request (struct request_queue *rq); + +/* Virtual block-device subsystem. */ +/* Note that xlvbd_add doesn't call add_disk for you: you're expected + to call add_disk on info->gd once the disk is properly connected + up. */ +int xlvbd_add(blkif_sector_t capacity, int device, + u16 vdisk_info, u16 sector_size, struct blkfront_info *info); +void xlvbd_del(struct blkfront_info *info); +int xlvbd_barrier(struct blkfront_info *info); + +#ifdef CONFIG_SYSFS +int xlvbd_sysfs_addif(struct blkfront_info *info); +void xlvbd_sysfs_delif(struct blkfront_info *info); +#else +static inline int xlvbd_sysfs_addif(struct blkfront_info *info) +{ + return 0; +} + +static inline void xlvbd_sysfs_delif(struct blkfront_info *info) +{ + ; +} +#endif + +/* Virtual cdrom block-device */ +extern void register_vcd(struct blkfront_info *info); +extern void unregister_vcd(struct blkfront_info *info); + +#endif /* __XEN_DRIVERS_BLOCK_H__ */ diff --git a/drivers/xenpvdrivers/evtchn.c b/drivers/xenpvdrivers/evtchn.c new file mode 100644 index 0000000..6cbef9d --- /dev/null +++ b/drivers/xenpvdrivers/evtchn.c @@ -0,0 +1,356 @@ +/****************************************************************************** + * evtchn.c + * + * A simplified event channel for para-drivers in unmodified linux + * + * Copyright (c) 2002-2005, K A Fraser + * Copyright (c) 2005, Intel Corporation + * + * This file may be distributed separately from the Linux kernel, or + * incorporated into other software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include +#include +#include "platform-pci.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +void *shared_info_area; + +#define is_valid_evtchn(x) ((x) != 0) +#define evtchn_from_irq(x) (irq_evtchn[irq].evtchn) + +static struct { + spinlock_t lock; + irq_handler_t handler; + void *dev_id; + int evtchn; + int close:1; /* close on unbind_from_irqhandler()? */ + int inuse:1; + int in_handler:1; +} irq_evtchn[256]; +static int evtchn_to_irq[NR_EVENT_CHANNELS] = { + [0 ... NR_EVENT_CHANNELS-1] = -1 }; + +static DEFINE_SPINLOCK(irq_alloc_lock); + +static int alloc_xen_irq(void) +{ + static int warned; + int irq; + + spin_lock(&irq_alloc_lock); + + for (irq = 1; irq < ARRAY_SIZE(irq_evtchn); irq++) { + if (irq_evtchn[irq].inuse) + continue; + irq_evtchn[irq].inuse = 1; + spin_unlock(&irq_alloc_lock); + return irq; + } + + if (!warned) { + warned = 1; + printk(KERN_WARNING "No available IRQ to bind to: " + "increase irq_evtchn[] size in evtchn.c.\n"); + } + + spin_unlock(&irq_alloc_lock); + + return -ENOSPC; +} + +static void free_xen_irq(int irq) +{ + spin_lock(&irq_alloc_lock); + irq_evtchn[irq].inuse = 0; + spin_unlock(&irq_alloc_lock); +} + +int irq_to_evtchn_port(int irq) +{ + return irq_evtchn[irq].evtchn; +} +EXPORT_SYMBOL(irq_to_evtchn_port); + +void mask_evtchn(int port) +{ + shared_info_t *s = shared_info_area; + synch_set_bit(port, &s->evtchn_mask[0]); +} +EXPORT_SYMBOL(mask_evtchn); + +void unmask_evtchn(int port) +{ + evtchn_unmask_t op = { .port = port }; + VOID(HYPERVISOR_event_channel_op(EVTCHNOP_unmask, &op)); +} +EXPORT_SYMBOL(unmask_evtchn); + +int bind_listening_port_to_irqhandler( + unsigned int remote_domain, + irq_handler_t handler, + unsigned long irqflags, + const char *devname, + void *dev_id) +{ + struct evtchn_alloc_unbound alloc_unbound; + int err, irq; + + irq = alloc_xen_irq(); + if (irq < 0) + return irq; + + spin_lock_irq(&irq_evtchn[irq].lock); + + alloc_unbound.dom = DOMID_SELF; + alloc_unbound.remote_dom = remote_domain; + err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, + &alloc_unbound); + if (err) { + spin_unlock_irq(&irq_evtchn[irq].lock); + free_xen_irq(irq); + return err; + } + + irq_evtchn[irq].handler = handler; + irq_evtchn[irq].dev_id = dev_id; + irq_evtchn[irq].evtchn = alloc_unbound.port; + irq_evtchn[irq].close = 1; + + evtchn_to_irq[alloc_unbound.port] = irq; + + unmask_evtchn(alloc_unbound.port); + + spin_unlock_irq(&irq_evtchn[irq].lock); + + return irq; +} +EXPORT_SYMBOL(bind_listening_port_to_irqhandler); + +int bind_caller_port_to_irqhandler( + unsigned int caller_port, + irq_handler_t handler, + unsigned long irqflags, + const char *devname, + void *dev_id) +{ + int irq; + + irq = alloc_xen_irq(); + if (irq < 0) + return irq; + + spin_lock_irq(&irq_evtchn[irq].lock); + + irq_evtchn[irq].handler = handler; + irq_evtchn[irq].dev_id = dev_id; + irq_evtchn[irq].evtchn = caller_port; + irq_evtchn[irq].close = 0; + + evtchn_to_irq[caller_port] = irq; + + unmask_evtchn(caller_port); + + spin_unlock_irq(&irq_evtchn[irq].lock); + + return irq; +} +EXPORT_SYMBOL(bind_caller_port_to_irqhandler); + +void unbind_from_irqhandler(unsigned int irq, void *dev_id) +{ + int evtchn; + + spin_lock_irq(&irq_evtchn[irq].lock); + + evtchn = evtchn_from_irq(irq); + + if (is_valid_evtchn(evtchn)) { + evtchn_to_irq[evtchn] = -1; + mask_evtchn(evtchn); + if (irq_evtchn[irq].close) { + struct evtchn_close close = { .port = evtchn }; + if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close)) + BUG(); + } + } + + irq_evtchn[irq].handler = NULL; + irq_evtchn[irq].evtchn = 0; + + spin_unlock_irq(&irq_evtchn[irq].lock); + + while (irq_evtchn[irq].in_handler) + cpu_relax(); + + free_xen_irq(irq); +} +EXPORT_SYMBOL(unbind_from_irqhandler); + +void notify_remote_via_irq(int irq) +{ + int evtchn; + + evtchn = evtchn_from_irq(irq); + if (is_valid_evtchn(evtchn)) + notify_remote_via_evtchn(evtchn); +} +EXPORT_SYMBOL(notify_remote_via_irq); + +static DEFINE_PER_CPU(unsigned int, last_processed_l1i) = { BITS_PER_LONG - 1 }; +static DEFINE_PER_CPU(unsigned int, last_processed_l2i) = { BITS_PER_LONG - 1 }; + +static inline unsigned long active_evtchns(unsigned int cpu, shared_info_t *sh, + unsigned int idx) +{ + return (sh->evtchn_pending[idx] & ~sh->evtchn_mask[idx]); +} + +static irqreturn_t evtchn_interrupt(int irq, void *dev_id +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19) + , struct pt_regs *regs +#else +# define handler(irq, dev_id, regs) handler(irq, dev_id) +#endif + ) +{ + unsigned int l1i, l2i, port; + unsigned long masked_l1, masked_l2; + /* XXX: All events are bound to vcpu0 but irq may be redirected. */ + int cpu = 0; /*smp_processor_id();*/ + irq_handler_t handler; + shared_info_t *s = shared_info_area; + vcpu_info_t *v = &s->vcpu_info[cpu]; + unsigned long l1, l2; + + v->evtchn_upcall_pending = 0; + +#ifndef CONFIG_X86 /* No need for a barrier -- XCHG is a barrier on x86. */ + /* Clear master flag /before/ clearing selector flag. */ + wmb(); +#endif + l1 = xchg(&v->evtchn_pending_sel, 0); + + l1i = per_cpu(last_processed_l1i, cpu); + l2i = per_cpu(last_processed_l2i, cpu); + + while (l1 != 0) { + + l1i = (l1i + 1) % BITS_PER_LONG; + masked_l1 = l1 & ((~0UL) << l1i); + + if (masked_l1 == 0) { /* if we masked out all events, wrap around to the beginning */ + l1i = BITS_PER_LONG - 1; + l2i = BITS_PER_LONG - 1; + continue; + } + l1i = __ffs(masked_l1); + + do { + l2 = active_evtchns(cpu, s, l1i); + + l2i = (l2i + 1) % BITS_PER_LONG; + masked_l2 = l2 & ((~0UL) << l2i); + + if (masked_l2 == 0) { /* if we masked out all events, move on */ + l2i = BITS_PER_LONG - 1; + break; + } + l2i = __ffs(masked_l2); + + /* process port */ + port = (l1i * BITS_PER_LONG) + l2i; + synch_clear_bit(port, &s->evtchn_pending[0]); + + irq = evtchn_to_irq[port]; + if (irq < 0) + continue; + + spin_lock(&irq_evtchn[irq].lock); + handler = irq_evtchn[irq].handler; + dev_id = irq_evtchn[irq].dev_id; + if (unlikely(handler == NULL)) { + printk("Xen IRQ%d (port %d) has no handler!\n", + irq, port); + spin_unlock(&irq_evtchn[irq].lock); + continue; + } + irq_evtchn[irq].in_handler = 1; + spin_unlock(&irq_evtchn[irq].lock); + + local_irq_enable(); + handler(irq, irq_evtchn[irq].dev_id, regs); + local_irq_disable(); + + spin_lock(&irq_evtchn[irq].lock); + irq_evtchn[irq].in_handler = 0; + spin_unlock(&irq_evtchn[irq].lock); + + /* if this is the final port processed, we'll pick up here+1 next time */ + per_cpu(last_processed_l1i, cpu) = l1i; + per_cpu(last_processed_l2i, cpu) = l2i; + + } while (l2i != BITS_PER_LONG - 1); + + l2 = active_evtchns(cpu, s, l1i); + if (l2 == 0) /* we handled all ports, so we can clear the selector bit */ + l1 &= ~(1UL << l1i); + } + + return IRQ_HANDLED; +} + +void irq_resume(void) +{ + int evtchn, irq; + + for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++) { + mask_evtchn(evtchn); + evtchn_to_irq[evtchn] = -1; + } + + for (irq = 0; irq < ARRAY_SIZE(irq_evtchn); irq++) + irq_evtchn[irq].evtchn = 0; +} + +int xen_irq_init(struct pci_dev *pdev) +{ + int irq; + + for (irq = 0; irq < ARRAY_SIZE(irq_evtchn); irq++) + spin_lock_init(&irq_evtchn[irq].lock); + + return request_irq(pdev->irq, evtchn_interrupt, +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22) + SA_SHIRQ | SA_SAMPLE_RANDOM | SA_INTERRUPT, +#else + IRQF_SHARED | IRQF_SAMPLE_RANDOM | IRQF_DISABLED, +#endif + "xen-platform-pci", pdev); +} diff --git a/drivers/xenpvdrivers/features.c b/drivers/xenpvdrivers/features.c new file mode 100644 index 0000000..70031ee --- /dev/null +++ b/drivers/xenpvdrivers/features.c @@ -0,0 +1,40 @@ +/****************************************************************************** + * features.c + * + * Xen feature flags. + * + * Copyright (c) 2006, Ian Campbell, XenSource Inc. + */ +#include +#include +#include + +#ifdef CONFIG_PARAVIRT_XEN +#include +#else +#include +#endif +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#include +#include +#include + +u8 xen_features[XENFEAT_NR_SUBMAPS * 32] __read_mostly; +EXPORT_SYMBOL(xen_features); + +void xen_setup_features(void) +{ + struct xen_feature_info fi; + int i, j; + + for (i = 0; i < XENFEAT_NR_SUBMAPS; i++) { + fi.submap_idx = i; + if (HYPERVISOR_xen_version(XENVER_get_features, &fi) < 0) + break; + for (j = 0; j < 32; j++) + xen_features[i * 32 + j] = !!(fi.submap & 1< +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +/* External tools reserve first few grant table entries. */ +#define NR_RESERVED_ENTRIES 8 +#define GNTTAB_LIST_END 0xffffffff +#define ENTRIES_PER_GRANT_FRAME (PAGE_SIZE / sizeof(grant_entry_t)) + +static grant_ref_t **gnttab_list; +static unsigned int nr_grant_frames; +static unsigned int boot_max_nr_grant_frames; +static int gnttab_free_count; +static grant_ref_t gnttab_free_head; +static DEFINE_SPINLOCK(gnttab_list_lock); + +static struct grant_entry *shared; + +static struct gnttab_free_callback *gnttab_free_callback_list; + +static int gnttab_expand(unsigned int req_entries); + +#define RPP (PAGE_SIZE / sizeof(grant_ref_t)) +#define gnttab_entry(entry) (gnttab_list[(entry) / RPP][(entry) % RPP]) + +#define nr_freelist_frames(grant_frames) \ + (((grant_frames) * ENTRIES_PER_GRANT_FRAME + RPP - 1) / RPP) + +static int get_free_entries(int count) +{ + unsigned long flags; + int ref, rc; + grant_ref_t head; + + spin_lock_irqsave(&gnttab_list_lock, flags); + + if ((gnttab_free_count < count) && + ((rc = gnttab_expand(count - gnttab_free_count)) < 0)) { + spin_unlock_irqrestore(&gnttab_list_lock, flags); + return rc; + } + + ref = head = gnttab_free_head; + gnttab_free_count -= count; + while (count-- > 1) + head = gnttab_entry(head); + gnttab_free_head = gnttab_entry(head); + gnttab_entry(head) = GNTTAB_LIST_END; + + spin_unlock_irqrestore(&gnttab_list_lock, flags); + + return ref; +} + +#define get_free_entry() get_free_entries(1) + +static void do_free_callbacks(void) +{ + struct gnttab_free_callback *callback, *next; + + callback = gnttab_free_callback_list; + gnttab_free_callback_list = NULL; + + while (callback != NULL) { + next = callback->next; + if (gnttab_free_count >= callback->count) { + callback->next = NULL; + callback->queued = 0; + callback->fn(callback->arg); + } else { + callback->next = gnttab_free_callback_list; + gnttab_free_callback_list = callback; + } + callback = next; + } +} + +static inline void check_free_callbacks(void) +{ + if (unlikely(gnttab_free_callback_list)) + do_free_callbacks(); +} + +static void put_free_entry(grant_ref_t ref) +{ + unsigned long flags; + spin_lock_irqsave(&gnttab_list_lock, flags); + gnttab_entry(ref) = gnttab_free_head; + gnttab_free_head = ref; + gnttab_free_count++; + check_free_callbacks(); + spin_unlock_irqrestore(&gnttab_list_lock, flags); +} + +/* + * Public grant-issuing interface functions + */ + +int gnttab_grant_foreign_access(domid_t domid, unsigned long frame, + int flags) +{ + int ref; + + if (unlikely((ref = get_free_entry()) < 0)) + return -ENOSPC; + + shared[ref].frame = frame; + shared[ref].domid = domid; + wmb(); + BUG_ON(flags & (GTF_accept_transfer | GTF_reading | GTF_writing)); + shared[ref].flags = GTF_permit_access | flags; + + return ref; +} +EXPORT_SYMBOL_GPL(gnttab_grant_foreign_access); + +void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid, + unsigned long frame, int flags) +{ + shared[ref].frame = frame; + shared[ref].domid = domid; + wmb(); + BUG_ON(flags & (GTF_accept_transfer | GTF_reading | GTF_writing)); + shared[ref].flags = GTF_permit_access | flags; +} +EXPORT_SYMBOL_GPL(gnttab_grant_foreign_access_ref); + + +int gnttab_query_foreign_access(grant_ref_t ref) +{ + u16 nflags; + + nflags = shared[ref].flags; + + return (nflags & (GTF_reading|GTF_writing)); +} +EXPORT_SYMBOL_GPL(gnttab_query_foreign_access); + +int gnttab_end_foreign_access_ref(grant_ref_t ref) +{ + u16 flags, nflags; + + nflags = shared[ref].flags; + do { + if ((flags = nflags) & (GTF_reading|GTF_writing)) { + printk(KERN_DEBUG "WARNING: g.e. still in use!\n"); + return 0; + } + } while ((nflags = synch_cmpxchg_subword(&shared[ref].flags, flags, 0)) != + flags); + + return 1; +} +EXPORT_SYMBOL_GPL(gnttab_end_foreign_access_ref); + +void gnttab_end_foreign_access(grant_ref_t ref, unsigned long page) +{ + if (gnttab_end_foreign_access_ref(ref)) { + put_free_entry(ref); + if (page != 0) + free_page(page); + } else { + /* XXX This needs to be fixed so that the ref and page are + placed on a list to be freed up later. */ + printk(KERN_DEBUG + "WARNING: leaking g.e. and page still in use!\n"); + } +} +EXPORT_SYMBOL_GPL(gnttab_end_foreign_access); + +int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn) +{ + int ref; + + if (unlikely((ref = get_free_entry()) < 0)) + return -ENOSPC; + gnttab_grant_foreign_transfer_ref(ref, domid, pfn); + + return ref; +} +EXPORT_SYMBOL_GPL(gnttab_grant_foreign_transfer); + +void gnttab_grant_foreign_transfer_ref(grant_ref_t ref, domid_t domid, + unsigned long pfn) +{ + shared[ref].frame = pfn; + shared[ref].domid = domid; + wmb(); + shared[ref].flags = GTF_accept_transfer; +} +EXPORT_SYMBOL_GPL(gnttab_grant_foreign_transfer_ref); + +unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref) +{ + unsigned long frame; + u16 flags; + + /* + * If a transfer is not even yet started, try to reclaim the grant + * reference and return failure (== 0). + */ + while (!((flags = shared[ref].flags) & GTF_transfer_committed)) { + if (synch_cmpxchg_subword(&shared[ref].flags, flags, 0) == flags) + return 0; + cpu_relax(); + } + + /* If a transfer is in progress then wait until it is completed. */ + while (!(flags & GTF_transfer_completed)) { + flags = shared[ref].flags; + cpu_relax(); + } + + /* Read the frame number /after/ reading completion status. */ + rmb(); + frame = shared[ref].frame; + BUG_ON(frame == 0); + + return frame; +} +EXPORT_SYMBOL_GPL(gnttab_end_foreign_transfer_ref); + +unsigned long gnttab_end_foreign_transfer(grant_ref_t ref) +{ + unsigned long frame = gnttab_end_foreign_transfer_ref(ref); + put_free_entry(ref); + return frame; +} +EXPORT_SYMBOL_GPL(gnttab_end_foreign_transfer); + +void gnttab_free_grant_reference(grant_ref_t ref) +{ + put_free_entry(ref); +} +EXPORT_SYMBOL_GPL(gnttab_free_grant_reference); + +void gnttab_free_grant_references(grant_ref_t head) +{ + grant_ref_t ref; + unsigned long flags; + int count = 1; + if (head == GNTTAB_LIST_END) + return; + spin_lock_irqsave(&gnttab_list_lock, flags); + ref = head; + while (gnttab_entry(ref) != GNTTAB_LIST_END) { + ref = gnttab_entry(ref); + count++; + } + gnttab_entry(ref) = gnttab_free_head; + gnttab_free_head = head; + gnttab_free_count += count; + check_free_callbacks(); + spin_unlock_irqrestore(&gnttab_list_lock, flags); +} +EXPORT_SYMBOL_GPL(gnttab_free_grant_references); + +int gnttab_alloc_grant_references(u16 count, grant_ref_t *head) +{ + int h = get_free_entries(count); + + if (h < 0) + return -ENOSPC; + + *head = h; + + return 0; +} +EXPORT_SYMBOL_GPL(gnttab_alloc_grant_references); + +int gnttab_empty_grant_references(const grant_ref_t *private_head) +{ + return (*private_head == GNTTAB_LIST_END); +} +EXPORT_SYMBOL_GPL(gnttab_empty_grant_references); + +int gnttab_claim_grant_reference(grant_ref_t *private_head) +{ + grant_ref_t g = *private_head; + if (unlikely(g == GNTTAB_LIST_END)) + return -ENOSPC; + *private_head = gnttab_entry(g); + return g; +} +EXPORT_SYMBOL_GPL(gnttab_claim_grant_reference); + +void gnttab_release_grant_reference(grant_ref_t *private_head, + grant_ref_t release) +{ + gnttab_entry(release) = *private_head; + *private_head = release; +} +EXPORT_SYMBOL_GPL(gnttab_release_grant_reference); + +void gnttab_request_free_callback(struct gnttab_free_callback *callback, + void (*fn)(void *), void *arg, u16 count) +{ + unsigned long flags; + spin_lock_irqsave(&gnttab_list_lock, flags); + if (callback->queued) + goto out; + callback->fn = fn; + callback->arg = arg; + callback->count = count; + callback->queued = 1; + callback->next = gnttab_free_callback_list; + gnttab_free_callback_list = callback; + check_free_callbacks(); +out: + spin_unlock_irqrestore(&gnttab_list_lock, flags); +} +EXPORT_SYMBOL_GPL(gnttab_request_free_callback); + +void gnttab_cancel_free_callback(struct gnttab_free_callback *callback) +{ + struct gnttab_free_callback **pcb; + unsigned long flags; + + spin_lock_irqsave(&gnttab_list_lock, flags); + for (pcb = &gnttab_free_callback_list; *pcb; pcb = &(*pcb)->next) { + if (*pcb == callback) { + *pcb = callback->next; + callback->queued = 0; + break; + } + } + spin_unlock_irqrestore(&gnttab_list_lock, flags); +} +EXPORT_SYMBOL_GPL(gnttab_cancel_free_callback); + +static int grow_gnttab_list(unsigned int more_frames) +{ + unsigned int new_nr_grant_frames, extra_entries, i; + unsigned int nr_glist_frames, new_nr_glist_frames; + + new_nr_grant_frames = nr_grant_frames + more_frames; + extra_entries = more_frames * ENTRIES_PER_GRANT_FRAME; + + nr_glist_frames = nr_freelist_frames(nr_grant_frames); + new_nr_glist_frames = nr_freelist_frames(new_nr_grant_frames); + for (i = nr_glist_frames; i < new_nr_glist_frames; i++) { + gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_ATOMIC); + if (!gnttab_list[i]) + goto grow_nomem; + } + + for (i = ENTRIES_PER_GRANT_FRAME * nr_grant_frames; + i < ENTRIES_PER_GRANT_FRAME * new_nr_grant_frames - 1; i++) + gnttab_entry(i) = i + 1; + + gnttab_entry(i) = gnttab_free_head; + gnttab_free_head = ENTRIES_PER_GRANT_FRAME * nr_grant_frames; + gnttab_free_count += extra_entries; + + nr_grant_frames = new_nr_grant_frames; + + check_free_callbacks(); + + return 0; + +grow_nomem: + for ( ; i >= nr_glist_frames; i--) + free_page((unsigned long) gnttab_list[i]); + return -ENOMEM; +} + +static unsigned int __max_nr_grant_frames(void) +{ + struct gnttab_query_size query; + int rc; + + query.dom = DOMID_SELF; + + rc = HYPERVISOR_grant_table_op(GNTTABOP_query_size, &query, 1); + if ((rc < 0) || (query.status != GNTST_okay)) + return 4; /* Legacy max supported number of frames */ + + return query.max_nr_frames; +} + +static inline unsigned int max_nr_grant_frames(void) +{ + unsigned int xen_max = __max_nr_grant_frames(); + + if (xen_max > boot_max_nr_grant_frames) + return boot_max_nr_grant_frames; + return xen_max; +} + +#ifdef CONFIG_XEN + +#ifdef CONFIG_X86 +static int map_pte_fn(pte_t *pte, struct page *pmd_page, + unsigned long addr, void *data) +{ + unsigned long **frames = (unsigned long **)data; + + set_pte_at(&init_mm, addr, pte, pfn_pte_ma((*frames)[0], PAGE_KERNEL)); + (*frames)++; + return 0; +} + +#ifdef CONFIG_PM_SLEEP +static int unmap_pte_fn(pte_t *pte, struct page *pmd_page, + unsigned long addr, void *data) +{ + + set_pte_at(&init_mm, addr, pte, __pte(0)); + return 0; +} +#endif + +void *arch_gnttab_alloc_shared(unsigned long *frames) +{ + struct vm_struct *area; + area = alloc_vm_area(PAGE_SIZE * max_nr_grant_frames()); + BUG_ON(area == NULL); + return area->addr; +} +#endif /* CONFIG_X86 */ + +static int gnttab_map(unsigned int start_idx, unsigned int end_idx) +{ + struct gnttab_setup_table setup; + unsigned long *frames; + unsigned int nr_gframes = end_idx + 1; + int rc; + + frames = kmalloc(nr_gframes * sizeof(unsigned long), GFP_ATOMIC); + if (!frames) + return -ENOMEM; + + setup.dom = DOMID_SELF; + setup.nr_frames = nr_gframes; + set_xen_guest_handle(setup.frame_list, frames); + + rc = HYPERVISOR_grant_table_op(GNTTABOP_setup_table, &setup, 1); + if (rc == -ENOSYS) { + kfree(frames); + return -ENOSYS; + } + + BUG_ON(rc || setup.status); + + if (shared == NULL) + shared = arch_gnttab_alloc_shared(frames); + +#ifdef CONFIG_X86 + rc = apply_to_page_range(&init_mm, (unsigned long)shared, + PAGE_SIZE * nr_gframes, + map_pte_fn, &frames); + BUG_ON(rc); + frames -= nr_gframes; /* adjust after map_pte_fn() */ +#endif /* CONFIG_X86 */ + + kfree(frames); + + return 0; +} + +#if defined(CONFIG_XEN_BACKEND) || defined(CONFIG_XEN_BACKEND_MODULE) + +static DEFINE_SEQLOCK(gnttab_dma_lock); + +static void gnttab_page_free(struct page *page, unsigned int order) +{ + BUG_ON(order); + ClearPageForeign(page); + gnttab_reset_grant_page(page); + ClearPageReserved(page); + put_page(page); +} + +/* + * Must not be called with IRQs off. This should only be used on the + * slow path. + * + * Copy a foreign granted page to local memory. + */ +int gnttab_copy_grant_page(grant_ref_t ref, struct page **pagep) +{ + struct gnttab_unmap_and_replace unmap; + mmu_update_t mmu; + struct page *page; + struct page *new_page; + void *new_addr; + void *addr; + paddr_t pfn; + maddr_t mfn; + maddr_t new_mfn; + int err; + + page = *pagep; + if (!get_page_unless_zero(page)) + return -ENOENT; + + err = -ENOMEM; + new_page = alloc_page(GFP_ATOMIC | __GFP_NOWARN); + if (!new_page) + goto out; + + new_addr = page_address(new_page); + addr = page_address(page); + memcpy(new_addr, addr, PAGE_SIZE); + + pfn = page_to_pfn(page); + mfn = pfn_to_mfn(pfn); + new_mfn = virt_to_mfn(new_addr); + + write_seqlock_bh(&gnttab_dma_lock); + + /* Make seq visible before checking page_mapped. */ + smp_mb(); + + /* Has the page been DMA-mapped? */ + if (unlikely(page_mapped(page))) { + write_sequnlock_bh(&gnttab_dma_lock); + put_page(new_page); + err = -EBUSY; + goto out; + } + + if (!xen_feature(XENFEAT_auto_translated_physmap)) + set_phys_to_machine(pfn, new_mfn); + + gnttab_set_replace_op(&unmap, (unsigned long)addr, + (unsigned long)new_addr, ref); + + err = HYPERVISOR_grant_table_op(GNTTABOP_unmap_and_replace, + &unmap, 1); + BUG_ON(err); + BUG_ON(unmap.status); + + write_sequnlock_bh(&gnttab_dma_lock); + + if (!xen_feature(XENFEAT_auto_translated_physmap)) { + set_phys_to_machine(page_to_pfn(new_page), INVALID_P2M_ENTRY); + + mmu.ptr = (new_mfn << PAGE_SHIFT) | MMU_MACHPHYS_UPDATE; + mmu.val = pfn; + err = HYPERVISOR_mmu_update(&mmu, 1, NULL, DOMID_SELF); + BUG_ON(err); + } + + new_page->mapping = page->mapping; + new_page->index = page->index; + set_bit(PG_foreign, &new_page->flags); + if (PageReserved(page)) + SetPageReserved(new_page); + *pagep = new_page; + + SetPageForeign(page, gnttab_page_free); + page->mapping = NULL; + +out: + put_page(page); + return err; +} +EXPORT_SYMBOL_GPL(gnttab_copy_grant_page); + +void gnttab_reset_grant_page(struct page *page) +{ + init_page_count(page); + reset_page_mapcount(page); +} +EXPORT_SYMBOL_GPL(gnttab_reset_grant_page); + +/* + * Keep track of foreign pages marked as PageForeign so that we don't + * return them to the remote domain prematurely. + * + * PageForeign pages are pinned down by increasing their mapcount. + * + * All other pages are simply returned as is. + */ +void __gnttab_dma_map_page(struct page *page) +{ + unsigned int seq; + + if (!is_running_on_xen() || !PageForeign(page)) + return; + + do { + seq = read_seqbegin(&gnttab_dma_lock); + + if (gnttab_dma_local_pfn(page)) + break; + + atomic_set(&page->_mapcount, 0); + + /* Make _mapcount visible before read_seqretry. */ + smp_mb(); + } while (unlikely(read_seqretry(&gnttab_dma_lock, seq))); +} + +#endif /* CONFIG_XEN_BACKEND */ + +#ifdef __HAVE_ARCH_PTE_SPECIAL + +static unsigned int GNTMAP_pte_special; + +bool gnttab_pre_map_adjust(unsigned int cmd, struct gnttab_map_grant_ref *map, + unsigned int count) +{ + unsigned int i; + + if (unlikely(cmd != GNTTABOP_map_grant_ref)) + count = 0; + + for (i = 0; i < count; ++i, ++map) { + if (!(map->flags & GNTMAP_host_map) + || !(map->flags & GNTMAP_application_map)) + continue; + if (GNTMAP_pte_special) + map->flags |= GNTMAP_pte_special; + else { + BUG_ON(xen_feature(XENFEAT_auto_translated_physmap)); + return true; + } + } + + return false; +} +EXPORT_SYMBOL(gnttab_pre_map_adjust); + +#if CONFIG_XEN_COMPAT < 0x030400 +int gnttab_post_map_adjust(const struct gnttab_map_grant_ref *map, unsigned int count) +{ + unsigned int i; + int rc = 0; + + for (i = 0; i < count && rc == 0; ++i, ++map) { + pte_t pte; + + if (!(map->flags & GNTMAP_host_map) + || !(map->flags & GNTMAP_application_map)) + continue; + +#ifdef CONFIG_X86 + pte = __pte_ma((map->dev_bus_addr | _PAGE_PRESENT | _PAGE_USER + | _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_NX + | _PAGE_SPECIAL) + & __supported_pte_mask); +#else +#error Architecture not yet supported. +#endif + if (!(map->flags & GNTMAP_readonly)) + pte = pte_mkwrite(pte); + + if (map->flags & GNTMAP_contains_pte) { + mmu_update_t u; + + u.ptr = map->host_addr; + u.val = __pte_val(pte); + rc = HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF); + } else + rc = HYPERVISOR_update_va_mapping(map->host_addr, pte, 0); + } + + return rc; +} +EXPORT_SYMBOL(gnttab_post_map_adjust); +#endif + +#endif /* __HAVE_ARCH_PTE_SPECIAL */ + +static int gnttab_resume(struct sys_device *dev) +{ + if (max_nr_grant_frames() < nr_grant_frames) + return -ENOSYS; + return gnttab_map(0, nr_grant_frames - 1); +} +#define gnttab_resume() gnttab_resume(NULL) + +#ifdef CONFIG_PM_SLEEP +#ifdef CONFIG_X86 +static int gnttab_suspend(struct sys_device *dev, pm_message_t state) +{ + apply_to_page_range(&init_mm, (unsigned long)shared, + PAGE_SIZE * nr_grant_frames, + unmap_pte_fn, NULL); + return 0; +} +#else +#define gnttab_suspend NULL +#endif + +static struct sysdev_class gnttab_sysclass = { + .name = "gnttab", + .resume = gnttab_resume, + .suspend = gnttab_suspend, +}; + +static struct sys_device device_gnttab = { + .id = 0, + .cls = &gnttab_sysclass, +}; +#endif + +#else /* !CONFIG_XEN */ + +#include "platform-pci.h" + +static unsigned long resume_frames; + +static int gnttab_map(unsigned int start_idx, unsigned int end_idx) +{ + struct xen_add_to_physmap xatp; + unsigned int i = end_idx; + + /* Loop backwards, so that the first hypercall has the largest index, + * ensuring that the table will grow only once. + */ + do { + xatp.domid = DOMID_SELF; + xatp.idx = i; + xatp.space = XENMAPSPACE_grant_table; + xatp.gpfn = (resume_frames >> PAGE_SHIFT) + i; + if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp)) + BUG(); + } while (i-- > start_idx); + + return 0; +} + +int gnttab_resume(void) +{ + unsigned int max_nr_gframes, nr_gframes; + + nr_gframes = nr_grant_frames; + max_nr_gframes = max_nr_grant_frames(); + if (max_nr_gframes < nr_gframes) + return -ENOSYS; + + if (!resume_frames) { + resume_frames = alloc_xen_mmio(PAGE_SIZE * max_nr_gframes); + shared = ioremap(resume_frames, PAGE_SIZE * max_nr_gframes); + if (shared == NULL) { + printk("error to ioremap gnttab share frames\n"); + return -1; + } + } + + gnttab_map(0, nr_gframes - 1); + + return 0; +} + +#endif /* !CONFIG_XEN */ + +static int gnttab_expand(unsigned int req_entries) +{ + int rc; + unsigned int cur, extra; + + cur = nr_grant_frames; + extra = ((req_entries + (ENTRIES_PER_GRANT_FRAME-1)) / + ENTRIES_PER_GRANT_FRAME); + if (cur + extra > max_nr_grant_frames()) + return -ENOSPC; + + if ((rc = gnttab_map(cur, cur + extra - 1)) == 0) + rc = grow_gnttab_list(extra); + + return rc; +} + +int __devinit gnttab_init(void) +{ + int i; + unsigned int max_nr_glist_frames, nr_glist_frames; + unsigned int nr_init_grefs; + + if (!is_running_on_xen()) + return -ENODEV; + +#if defined(CONFIG_XEN) && defined(CONFIG_PM_SLEEP) + if (!is_initial_xendomain()) { + int err = sysdev_class_register(&gnttab_sysclass); + + if (!err) + err = sysdev_register(&device_gnttab); + if (err) + return err; + } +#endif + + nr_grant_frames = 1; + boot_max_nr_grant_frames = __max_nr_grant_frames(); + + /* Determine the maximum number of frames required for the + * grant reference free list on the current hypervisor. + */ + max_nr_glist_frames = nr_freelist_frames(boot_max_nr_grant_frames); + + gnttab_list = kmalloc(max_nr_glist_frames * sizeof(grant_ref_t *), + GFP_KERNEL); + if (gnttab_list == NULL) + return -ENOMEM; + + nr_glist_frames = nr_freelist_frames(nr_grant_frames); + for (i = 0; i < nr_glist_frames; i++) { + gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_KERNEL); + if (gnttab_list[i] == NULL) + goto ini_nomem; + } + + if (gnttab_resume() < 0) + return -ENODEV; + + nr_init_grefs = nr_grant_frames * ENTRIES_PER_GRANT_FRAME; + + for (i = NR_RESERVED_ENTRIES; i < nr_init_grefs - 1; i++) + gnttab_entry(i) = i + 1; + + gnttab_entry(nr_init_grefs - 1) = GNTTAB_LIST_END; + gnttab_free_count = nr_init_grefs - NR_RESERVED_ENTRIES; + gnttab_free_head = NR_RESERVED_ENTRIES; + +#if defined(CONFIG_XEN) && defined(__HAVE_ARCH_PTE_SPECIAL) + if (!xen_feature(XENFEAT_auto_translated_physmap) + && xen_feature(XENFEAT_gnttab_map_avail_bits)) { +#ifdef CONFIG_X86 + GNTMAP_pte_special = (__pte_val(pte_mkspecial(__pte_ma(0))) + >> _PAGE_BIT_UNUSED1) << _GNTMAP_guest_avail0; +#else +#error Architecture not yet supported. +#endif + } +#endif + + return 0; + + ini_nomem: + for (i--; i >= 0; i--) + free_page((unsigned long)gnttab_list[i]); + kfree(gnttab_list); + return -ENOMEM; +} + +#ifdef CONFIG_XEN +core_initcall(gnttab_init); +#endif diff --git a/drivers/xenpvdrivers/machine_reboot.c b/drivers/xenpvdrivers/machine_reboot.c new file mode 100644 index 0000000..7dc5d99 --- /dev/null +++ b/drivers/xenpvdrivers/machine_reboot.c @@ -0,0 +1,114 @@ +#include +#include +#include +#include +#include +#include "platform-pci.h" +#include + +struct ap_suspend_info { + int do_spin; + atomic_t nr_spinning; +}; + +#ifdef CONFIG_SMP + +/* + * Spinning prevents, for example, APs touching grant table entries while + * the shared grant table is not mapped into the address space imemdiately + * after resume. + */ +static void ap_suspend(void *_info) +{ + struct ap_suspend_info *info = _info; + + BUG_ON(!irqs_disabled()); + + atomic_inc(&info->nr_spinning); + mb(); + + while (info->do_spin) + cpu_relax(); + + mb(); + atomic_dec(&info->nr_spinning); +} + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27) +#define initiate_ap_suspend(i) smp_call_function(ap_suspend, i, 0, 0) +#else +#define initiate_ap_suspend(i) smp_call_function(ap_suspend, i, 0) +#endif + +#else /* !defined(CONFIG_SMP) */ + +#define initiate_ap_suspend(i) 0 + +#endif + +static int bp_suspend(void) +{ + int suspend_cancelled; + + BUG_ON(!irqs_disabled()); + + suspend_cancelled = HYPERVISOR_suspend(0); + + if (!suspend_cancelled) { + platform_pci_resume(); + gnttab_resume(); + irq_resume(); + } + + return suspend_cancelled; +} + +int __xen_suspend(int fast_suspend, void (*resume_notifier)(int)) +{ + int err, suspend_cancelled, nr_cpus; + struct ap_suspend_info info; + + xenbus_suspend(); + + preempt_disable(); + + /* Prevent any races with evtchn_interrupt() handler. */ + disable_irq(xen_platform_pdev->irq); + + info.do_spin = 1; + atomic_set(&info.nr_spinning, 0); + smp_mb(); + + nr_cpus = num_online_cpus() - 1; + + err = initiate_ap_suspend(&info); + if (err < 0) { + preempt_enable(); + xenbus_suspend_cancel(); + return err; + } + + while (atomic_read(&info.nr_spinning) != nr_cpus) + cpu_relax(); + + local_irq_disable(); + suspend_cancelled = bp_suspend(); + resume_notifier(suspend_cancelled); + local_irq_enable(); + + smp_mb(); + info.do_spin = 0; + while (atomic_read(&info.nr_spinning) != 0) + cpu_relax(); + + enable_irq(xen_platform_pdev->irq); + + preempt_enable(); + + if (!suspend_cancelled) + xenbus_resume(); + else + xenbus_suspend_cancel(); + + return 0; +} diff --git a/drivers/xenpvdrivers/netfront.c b/drivers/xenpvdrivers/netfront.c new file mode 100644 index 0000000..60379db --- /dev/null +++ b/drivers/xenpvdrivers/netfront.c @@ -0,0 +1,2245 @@ +/****************************************************************************** + * Virtual network driver for conversing with remote driver backends. + * + * Copyright (c) 2002-2005, K A Fraser + * Copyright (c) 2005, XenSource Ltd + * Copyright (C) 2007 Solarflare Communications, Inc. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +struct netfront_cb { + struct page *page; + unsigned offset; +}; + +#define NETFRONT_SKB_CB(skb) ((struct netfront_cb *)((skb)->cb)) + +#include "netfront.h" + +/* + * Mutually-exclusive module options to select receive data path: + * rx_copy : Packets are copied by network backend into local memory + * rx_flip : Page containing packet data is transferred to our ownership + * For fully-virtualised guests there is no option - copying must be used. + * For paravirtualised guests, flipping is the default. + */ +#ifdef CONFIG_XEN +static int MODPARM_rx_copy = 0; +module_param_named(rx_copy, MODPARM_rx_copy, bool, 0); +MODULE_PARM_DESC(rx_copy, "Copy packets from network card (rather than flip)"); +static int MODPARM_rx_flip = 0; +module_param_named(rx_flip, MODPARM_rx_flip, bool, 0); +MODULE_PARM_DESC(rx_flip, "Flip packets from network card (rather than copy)"); +#else +static const int MODPARM_rx_copy = 1; +static const int MODPARM_rx_flip = 0; +#endif + +#define RX_COPY_THRESHOLD 256 + +/* If we don't have GSO, fake things up so that we never try to use it. */ +#if defined(NETIF_F_GSO) +#define HAVE_GSO 1 +#define HAVE_TSO 1 /* TSO is a subset of GSO */ +#define HAVE_CSUM_OFFLOAD 1 +static inline void dev_disable_gso_features(struct net_device *dev) +{ + /* Turn off all GSO bits except ROBUST. */ + dev->features &= ~NETIF_F_GSO_MASK; + dev->features |= NETIF_F_GSO_ROBUST; +} +#elif defined(NETIF_F_TSO) +#define HAVE_GSO 0 +#define HAVE_TSO 1 + +/* Some older kernels cannot cope with incorrect checksums, + * particularly in netfilter. I'm not sure there is 100% correlation + * with the presence of NETIF_F_TSO but it appears to be a good first + * approximiation. + */ +#define HAVE_CSUM_OFFLOAD 0 + +#define gso_size tso_size +#define gso_segs tso_segs +static inline void dev_disable_gso_features(struct net_device *dev) +{ + /* Turn off all TSO bits. */ + dev->features &= ~NETIF_F_TSO; +} +static inline int skb_is_gso(const struct sk_buff *skb) +{ + return skb_shinfo(skb)->tso_size; +} +static inline int skb_gso_ok(struct sk_buff *skb, int features) +{ + return (features & NETIF_F_TSO); +} + +static inline int netif_needs_gso(struct net_device *dev, struct sk_buff *skb) +{ + return skb_is_gso(skb) && + (!skb_gso_ok(skb, dev->features) || + unlikely(skb->ip_summed != CHECKSUM_PARTIAL)); +} +#else +#define HAVE_GSO 0 +#define HAVE_TSO 0 +#define HAVE_CSUM_OFFLOAD 0 +#define netif_needs_gso(dev, skb) 0 +#define dev_disable_gso_features(dev) ((void)0) +#define ethtool_op_set_tso(dev, data) (-ENOSYS) +#endif + +#define GRANT_INVALID_REF 0 + +struct netfront_rx_info { + struct netif_rx_response rx; + struct netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1]; +}; + +/* + * Implement our own carrier flag: the network stack's version causes delays + * when the carrier is re-enabled (in particular, dev_activate() may not + * immediately be called, which can cause packet loss). + */ +#define netfront_carrier_on(netif) ((netif)->carrier = 1) +#define netfront_carrier_off(netif) ((netif)->carrier = 0) +#define netfront_carrier_ok(netif) ((netif)->carrier) + +/* + * Access macros for acquiring freeing slots in tx_skbs[]. + */ + +static inline void add_id_to_freelist(struct sk_buff **list, unsigned short id) +{ + list[id] = list[0]; + list[0] = (void *)(unsigned long)id; +} + +static inline unsigned short get_id_from_freelist(struct sk_buff **list) +{ + unsigned int id = (unsigned int)(unsigned long)list[0]; + list[0] = list[id]; + return id; +} + +static inline int xennet_rxidx(RING_IDX idx) +{ + return idx & (NET_RX_RING_SIZE - 1); +} + +static inline struct sk_buff *xennet_get_rx_skb(struct netfront_info *np, + RING_IDX ri) +{ + int i = xennet_rxidx(ri); + struct sk_buff *skb = np->rx_skbs[i]; + np->rx_skbs[i] = NULL; + return skb; +} + +static inline grant_ref_t xennet_get_rx_ref(struct netfront_info *np, + RING_IDX ri) +{ + int i = xennet_rxidx(ri); + grant_ref_t ref = np->grant_rx_ref[i]; + np->grant_rx_ref[i] = GRANT_INVALID_REF; + return ref; +} + +#define DPRINTK(fmt, args...) \ + pr_debug("netfront (%s:%d) " fmt, \ + __FUNCTION__, __LINE__, ##args) +#define IPRINTK(fmt, args...) \ + printk(KERN_INFO "netfront: " fmt, ##args) +#define WPRINTK(fmt, args...) \ + printk(KERN_WARNING "netfront: " fmt, ##args) + +static int setup_device(struct xenbus_device *, struct netfront_info *); +static struct net_device *create_netdev(struct xenbus_device *); + +static void end_access(int, void *); +static void netif_disconnect_backend(struct netfront_info *); + +static int network_connect(struct net_device *); +static void network_tx_buf_gc(struct net_device *); +static void network_alloc_rx_buffers(struct net_device *); +static void send_fake_arp(struct net_device *); + +static irqreturn_t netif_int(int irq, void *dev_id); + +#ifdef CONFIG_SYSFS +static int xennet_sysfs_addif(struct net_device *netdev); +static void xennet_sysfs_delif(struct net_device *netdev); +#else /* !CONFIG_SYSFS */ +#define xennet_sysfs_addif(dev) (0) +#define xennet_sysfs_delif(dev) do { } while(0) +#endif + +static inline int xennet_can_sg(struct net_device *dev) +{ + return dev->features & NETIF_F_SG; +} + +/** + * Entry point to this code when a new device is created. Allocate the basic + * structures and the ring buffers for communication with the backend, and + * inform the backend of the appropriate details for those. + */ +static int __devinit netfront_probe(struct xenbus_device *dev, + const struct xenbus_device_id *id) +{ + int err; + struct net_device *netdev; + struct netfront_info *info; + + netdev = create_netdev(dev); + if (IS_ERR(netdev)) { + err = PTR_ERR(netdev); + xenbus_dev_fatal(dev, err, "creating netdev"); + return err; + } + + info = netdev_priv(netdev); + dev_set_drvdata(&dev->dev, info); + + err = register_netdev(info->netdev); + if (err) { + printk(KERN_WARNING "%s: register_netdev err=%d\n", + __FUNCTION__, err); + goto fail; + } + + err = xennet_sysfs_addif(info->netdev); + if (err) { + unregister_netdev(info->netdev); + printk(KERN_WARNING "%s: add sysfs failed err=%d\n", + __FUNCTION__, err); + goto fail; + } + + return 0; + + fail: + free_netdev(netdev); + dev_set_drvdata(&dev->dev, NULL); + return err; +} + +static int __devexit netfront_remove(struct xenbus_device *dev) +{ + struct netfront_info *info = dev_get_drvdata(&dev->dev); + + DPRINTK("%s\n", dev->nodename); + + netfront_accelerator_call_remove(info, dev); + + netif_disconnect_backend(info); + + del_timer_sync(&info->rx_refill_timer); + + xennet_sysfs_delif(info->netdev); + + unregister_netdev(info->netdev); + + free_netdev(info->netdev); + + return 0; +} + + +static int netfront_suspend(struct xenbus_device *dev) +{ + struct netfront_info *info = dev_get_drvdata(&dev->dev); + return netfront_accelerator_suspend(info, dev); +} + + +static int netfront_suspend_cancel(struct xenbus_device *dev) +{ + struct netfront_info *info = dev_get_drvdata(&dev->dev); + return netfront_accelerator_suspend_cancel(info, dev); +} + + +/** + * We are reconnecting to the backend, due to a suspend/resume, or a backend + * driver restart. We tear down our netif structure and recreate it, but + * leave the device-layer structures intact so that this is transparent to the + * rest of the kernel. + */ +static int netfront_resume(struct xenbus_device *dev) +{ + struct netfront_info *info = dev_get_drvdata(&dev->dev); + + DPRINTK("%s\n", dev->nodename); + + netfront_accelerator_resume(info, dev); + + netif_disconnect_backend(info); + return 0; +} + +static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[]) +{ + char *s, *e, *macstr; + int i; + + macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL); + if (IS_ERR(macstr)) + return PTR_ERR(macstr); + + for (i = 0; i < ETH_ALEN; i++) { + mac[i] = simple_strtoul(s, &e, 16); + if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) { + kfree(macstr); + return -ENOENT; + } + s = e+1; + } + + kfree(macstr); + return 0; +} + +/* Common code used when first setting up, and when resuming. */ +static int talk_to_backend(struct xenbus_device *dev, + struct netfront_info *info) +{ + const char *message; + struct xenbus_transaction xbt; + int err; + + /* Read mac only in the first setup. */ + if (!is_valid_ether_addr(info->mac)) { + err = xen_net_read_mac(dev, info->mac); + if (err) { + xenbus_dev_fatal(dev, err, "parsing %s/mac", + dev->nodename); + goto out; + } + } + + /* Create shared ring, alloc event channel. */ + err = setup_device(dev, info); + if (err) + goto out; + + /* This will load an accelerator if one is configured when the + * watch fires */ + netfront_accelerator_add_watch(info); + +again: + err = xenbus_transaction_start(&xbt); + if (err) { + xenbus_dev_fatal(dev, err, "starting transaction"); + goto destroy_ring; + } + + err = xenbus_printf(xbt, dev->nodename, "tx-ring-ref","%u", + info->tx_ring_ref); + if (err) { + message = "writing tx ring-ref"; + goto abort_transaction; + } + err = xenbus_printf(xbt, dev->nodename, "rx-ring-ref","%u", + info->rx_ring_ref); + if (err) { + message = "writing rx ring-ref"; + goto abort_transaction; + } + err = xenbus_printf(xbt, dev->nodename, + "event-channel", "%u", + irq_to_evtchn_port(info->irq)); + if (err) { + message = "writing event-channel"; + goto abort_transaction; + } + + err = xenbus_printf(xbt, dev->nodename, "request-rx-copy", "%u", + info->copying_receiver); + if (err) { + message = "writing request-rx-copy"; + goto abort_transaction; + } + + err = xenbus_printf(xbt, dev->nodename, "feature-rx-notify", "%d", 1); + if (err) { + message = "writing feature-rx-notify"; + goto abort_transaction; + } + + err = xenbus_printf(xbt, dev->nodename, "feature-no-csum-offload", + "%d", !HAVE_CSUM_OFFLOAD); + if (err) { + message = "writing feature-no-csum-offload"; + goto abort_transaction; + } + + err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", 1); + if (err) { + message = "writing feature-sg"; + goto abort_transaction; + } + + err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4", "%d", + HAVE_TSO); + if (err) { + message = "writing feature-gso-tcpv4"; + goto abort_transaction; + } + + err = xenbus_transaction_end(xbt, 0); + if (err) { + if (err == -EAGAIN) + goto again; + xenbus_dev_fatal(dev, err, "completing transaction"); + goto destroy_ring; + } + + return 0; + + abort_transaction: + xenbus_transaction_end(xbt, 1); + xenbus_dev_fatal(dev, err, "%s", message); + destroy_ring: + netfront_accelerator_call_remove(info, dev); + netif_disconnect_backend(info); + out: + return err; +} + +static int setup_device(struct xenbus_device *dev, struct netfront_info *info) +{ + struct netif_tx_sring *txs; + struct netif_rx_sring *rxs; + int err; + struct net_device *netdev = info->netdev; + + info->tx_ring_ref = GRANT_INVALID_REF; + info->rx_ring_ref = GRANT_INVALID_REF; + info->rx.sring = NULL; + info->tx.sring = NULL; + info->irq = 0; + + txs = (struct netif_tx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH); + if (!txs) { + err = -ENOMEM; + xenbus_dev_fatal(dev, err, "allocating tx ring page"); + goto fail; + } + SHARED_RING_INIT(txs); + FRONT_RING_INIT(&info->tx, txs, PAGE_SIZE); + + err = xenbus_grant_ring(dev, virt_to_mfn(txs)); + if (err < 0) { + free_page((unsigned long)txs); + goto fail; + } + info->tx_ring_ref = err; + + rxs = (struct netif_rx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH); + if (!rxs) { + err = -ENOMEM; + xenbus_dev_fatal(dev, err, "allocating rx ring page"); + goto fail; + } + SHARED_RING_INIT(rxs); + FRONT_RING_INIT(&info->rx, rxs, PAGE_SIZE); + + err = xenbus_grant_ring(dev, virt_to_mfn(rxs)); + if (err < 0) { + free_page((unsigned long)rxs); + goto fail; + } + info->rx_ring_ref = err; + + memcpy(netdev->dev_addr, info->mac, ETH_ALEN); + + err = bind_listening_port_to_irqhandler( + dev->otherend_id, netif_int, IRQF_SAMPLE_RANDOM, netdev->name, + netdev); + if (err < 0) + goto fail; + info->irq = err; + + return 0; + + fail: + return err; +} + +/** + * Callback received when the backend's state changes. + */ +static void backend_changed(struct xenbus_device *dev, + enum xenbus_state backend_state) +{ + struct netfront_info *np = dev_get_drvdata(&dev->dev); + struct net_device *netdev = np->netdev; + + DPRINTK("%s\n", xenbus_strstate(backend_state)); + + switch (backend_state) { + case XenbusStateInitialising: + case XenbusStateInitialised: + case XenbusStateConnected: + case XenbusStateReconfiguring: + case XenbusStateReconfigured: + case XenbusStateUnknown: + case XenbusStateClosed: + break; + + case XenbusStateInitWait: + if (dev->state != XenbusStateInitialising) + break; + if (network_connect(netdev) != 0) + break; + xenbus_switch_state(dev, XenbusStateConnected); + send_fake_arp(netdev); + break; + + case XenbusStateClosing: + xenbus_frontend_closed(dev); + break; + } +} + +/** Send a packet on a net device to encourage switches to learn the + * MAC. We send a fake ARP request. + * + * @param dev device + * @return 0 on success, error code otherwise + */ +static void send_fake_arp(struct net_device *dev) +{ +#ifdef CONFIG_INET + struct sk_buff *skb; + u32 src_ip, dst_ip; + + dst_ip = INADDR_BROADCAST; + src_ip = inet_select_addr(dev, dst_ip, RT_SCOPE_LINK); + + /* No IP? Then nothing to do. */ + if (src_ip == 0) + return; + + skb = arp_create(ARPOP_REPLY, ETH_P_ARP, + dst_ip, dev, src_ip, + /*dst_hw*/ NULL, /*src_hw*/ NULL, + /*target_hw*/ dev->dev_addr); + if (skb == NULL) + return; + + dev_queue_xmit(skb); +#endif +} + +static inline int netfront_tx_slot_available(struct netfront_info *np) +{ + return ((np->tx.req_prod_pvt - np->tx.rsp_cons) < + (TX_MAX_TARGET - MAX_SKB_FRAGS - 2)); +} + + +static inline void network_maybe_wake_tx(struct net_device *dev) +{ + struct netfront_info *np = netdev_priv(dev); + + if (unlikely(netif_queue_stopped(dev)) && + netfront_tx_slot_available(np) && + likely(netif_running(dev)) && + netfront_check_accelerator_queue_ready(dev, np)) + netif_wake_queue(dev); +} + + +int netfront_check_queue_ready(struct net_device *dev) +{ + struct netfront_info *np = netdev_priv(dev); + + return unlikely(netif_queue_stopped(dev)) && + netfront_tx_slot_available(np) && + likely(netif_running(dev)); +} +EXPORT_SYMBOL(netfront_check_queue_ready); + + +static int network_open(struct net_device *dev) +{ + struct netfront_info *np = netdev_priv(dev); + + memset(&np->stats, 0, sizeof(np->stats)); + napi_enable(&np->napi); + + spin_lock_bh(&np->rx_lock); + if (netfront_carrier_ok(np)) { + network_alloc_rx_buffers(dev); + np->rx.sring->rsp_event = np->rx.rsp_cons + 1; + if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx)){ + netfront_accelerator_call_stop_napi_irq(np, dev); + + napi_schedule(&np->napi); + } + } + spin_unlock_bh(&np->rx_lock); + + netif_start_queue(dev); + + return 0; +} + +static void network_tx_buf_gc(struct net_device *dev) +{ + RING_IDX cons, prod; + unsigned short id; + struct netfront_info *np = netdev_priv(dev); + struct sk_buff *skb; + + BUG_ON(!netfront_carrier_ok(np)); + + do { + prod = np->tx.sring->rsp_prod; + rmb(); /* Ensure we see responses up to 'rp'. */ + + for (cons = np->tx.rsp_cons; cons != prod; cons++) { + struct netif_tx_response *txrsp; + + txrsp = RING_GET_RESPONSE(&np->tx, cons); + if (txrsp->status == NETIF_RSP_NULL) + continue; + + id = txrsp->id; + skb = np->tx_skbs[id]; + if (unlikely(gnttab_query_foreign_access( + np->grant_tx_ref[id]) != 0)) { + printk(KERN_ALERT "network_tx_buf_gc: warning " + "-- grant still in use by backend " + "domain.\n"); + BUG(); + } + gnttab_end_foreign_access_ref(np->grant_tx_ref[id]); + gnttab_release_grant_reference( + &np->gref_tx_head, np->grant_tx_ref[id]); + np->grant_tx_ref[id] = GRANT_INVALID_REF; + add_id_to_freelist(np->tx_skbs, id); + dev_kfree_skb_irq(skb); + } + + np->tx.rsp_cons = prod; + + /* + * Set a new event, then check for race with update of tx_cons. + * Note that it is essential to schedule a callback, no matter + * how few buffers are pending. Even if there is space in the + * transmit ring, higher layers may be blocked because too much + * data is outstanding: in such cases notification from Xen is + * likely to be the only kick that we'll get. + */ + np->tx.sring->rsp_event = + prod + ((np->tx.sring->req_prod - prod) >> 1) + 1; + mb(); + } while ((cons == prod) && (prod != np->tx.sring->rsp_prod)); + + network_maybe_wake_tx(dev); +} + +static void rx_refill_timeout(unsigned long data) +{ + struct net_device *dev = (struct net_device *)data; + struct netfront_info *np = netdev_priv(dev); + + netfront_accelerator_call_stop_napi_irq(np, dev); + + napi_schedule(&np->napi); +} + +static void network_alloc_rx_buffers(struct net_device *dev) +{ + unsigned short id; + struct netfront_info *np = netdev_priv(dev); + struct sk_buff *skb; + struct page *page; + int i, batch_target, notify; + RING_IDX req_prod = np->rx.req_prod_pvt; + struct xen_memory_reservation reservation; + grant_ref_t ref; + unsigned long pfn; + void *vaddr; + int nr_flips; + netif_rx_request_t *req; + + if (unlikely(!netfront_carrier_ok(np))) + return; + + /* + * Allocate skbuffs greedily, even though we batch updates to the + * receive ring. This creates a less bursty demand on the memory + * allocator, so should reduce the chance of failed allocation requests + * both for ourself and for other kernel subsystems. + */ + batch_target = np->rx_target - (req_prod - np->rx.rsp_cons); + for (i = skb_queue_len(&np->rx_batch); i < batch_target; i++) { + /* + * Allocate an skb and a page. Do not use __dev_alloc_skb as + * that will allocate page-sized buffers which is not + * necessary here. + * 16 bytes added as necessary headroom for netif_receive_skb. + */ + skb = alloc_skb(RX_COPY_THRESHOLD + 16 + NET_IP_ALIGN, + GFP_ATOMIC | __GFP_NOWARN); + if (unlikely(!skb)) + goto no_skb; + + page = alloc_page(GFP_ATOMIC | __GFP_NOWARN); + if (!page) { + kfree_skb(skb); +no_skb: + /* Any skbuffs queued for refill? Force them out. */ + if (i != 0) + goto refill; + /* Could not allocate any skbuffs. Try again later. */ + mod_timer(&np->rx_refill_timer, + jiffies + (HZ/10)); + break; + } + + skb_reserve(skb, 16 + NET_IP_ALIGN); /* mimic dev_alloc_skb() */ + skb_shinfo(skb)->frags[0].page = page; + skb_shinfo(skb)->nr_frags = 1; + __skb_queue_tail(&np->rx_batch, skb); + } + + /* Is the batch large enough to be worthwhile? */ + if (i < (np->rx_target/2)) { + if (req_prod > np->rx.sring->req_prod) + goto push; + return; + } + + /* Adjust our fill target if we risked running out of buffers. */ + if (((req_prod - np->rx.sring->rsp_prod) < (np->rx_target / 4)) && + ((np->rx_target *= 2) > np->rx_max_target)) + np->rx_target = np->rx_max_target; + + refill: + for (nr_flips = i = 0; ; i++) { + if ((skb = __skb_dequeue(&np->rx_batch)) == NULL) + break; + + skb->dev = dev; + + id = xennet_rxidx(req_prod + i); + + BUG_ON(np->rx_skbs[id]); + np->rx_skbs[id] = skb; + + ref = gnttab_claim_grant_reference(&np->gref_rx_head); + BUG_ON((signed short)ref < 0); + np->grant_rx_ref[id] = ref; + + pfn = page_to_pfn(skb_shinfo(skb)->frags[0].page); + vaddr = page_address(skb_shinfo(skb)->frags[0].page); + + req = RING_GET_REQUEST(&np->rx, req_prod + i); + if (!np->copying_receiver) { + gnttab_grant_foreign_transfer_ref(ref, + np->xbdev->otherend_id, + pfn); + np->rx_pfn_array[nr_flips] = pfn_to_mfn(pfn); + if (!xen_feature(XENFEAT_auto_translated_physmap)) { + /* Remove this page before passing + * back to Xen. */ + set_phys_to_machine(pfn, INVALID_P2M_ENTRY); + MULTI_update_va_mapping(np->rx_mcl+i, + (unsigned long)vaddr, + __pte(0), 0); + } + nr_flips++; + } else { + gnttab_grant_foreign_access_ref(ref, + np->xbdev->otherend_id, + pfn_to_mfn(pfn), + 0); + } + + req->id = id; + req->gref = ref; + } + + if ( nr_flips != 0 ) { + /* Tell the ballon driver what is going on. */ + balloon_update_driver_allowance(i); + + set_xen_guest_handle(reservation.extent_start, + np->rx_pfn_array); + reservation.nr_extents = nr_flips; + reservation.extent_order = 0; + reservation.address_bits = 0; + reservation.domid = DOMID_SELF; + + if (!xen_feature(XENFEAT_auto_translated_physmap)) { + /* After all PTEs have been zapped, flush the TLB. */ + np->rx_mcl[i-1].args[MULTI_UVMFLAGS_INDEX] = + UVMF_TLB_FLUSH|UVMF_ALL; + + /* Give away a batch of pages. */ + np->rx_mcl[i].op = __HYPERVISOR_memory_op; + np->rx_mcl[i].args[0] = XENMEM_decrease_reservation; + np->rx_mcl[i].args[1] = (unsigned long)&reservation; + + /* Zap PTEs and give away pages in one big + * multicall. */ + if (unlikely(HYPERVISOR_multicall(np->rx_mcl, i+1))) + BUG(); + + /* Check return status of HYPERVISOR_memory_op(). */ + if (unlikely(np->rx_mcl[i].result != i)) + panic("Unable to reduce memory reservation\n"); + while (nr_flips--) + BUG_ON(np->rx_mcl[nr_flips].result); + } else { + if (HYPERVISOR_memory_op(XENMEM_decrease_reservation, + &reservation) != i) + panic("Unable to reduce memory reservation\n"); + } + } else { + wmb(); + } + + /* Above is a suitable barrier to ensure backend will see requests. */ + np->rx.req_prod_pvt = req_prod + i; + push: + RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&np->rx, notify); + if (notify) + notify_remote_via_irq(np->irq); +} + +static void xennet_make_frags(struct sk_buff *skb, struct net_device *dev, + struct netif_tx_request *tx) +{ + struct netfront_info *np = netdev_priv(dev); + char *data = skb->data; + unsigned long mfn; + RING_IDX prod = np->tx.req_prod_pvt; + int frags = skb_shinfo(skb)->nr_frags; + unsigned int offset = offset_in_page(data); + unsigned int len = skb_headlen(skb); + unsigned int id; + grant_ref_t ref; + int i; + + while (len > PAGE_SIZE - offset) { + tx->size = PAGE_SIZE - offset; + tx->flags |= NETTXF_more_data; + len -= tx->size; + data += tx->size; + offset = 0; + + id = get_id_from_freelist(np->tx_skbs); + np->tx_skbs[id] = skb_get(skb); + tx = RING_GET_REQUEST(&np->tx, prod++); + tx->id = id; + ref = gnttab_claim_grant_reference(&np->gref_tx_head); + BUG_ON((signed short)ref < 0); + + mfn = virt_to_mfn(data); + gnttab_grant_foreign_access_ref(ref, np->xbdev->otherend_id, + mfn, GTF_readonly); + + tx->gref = np->grant_tx_ref[id] = ref; + tx->offset = offset; + tx->size = len; + tx->flags = 0; + } + + for (i = 0; i < frags; i++) { + skb_frag_t *frag = skb_shinfo(skb)->frags + i; + + tx->flags |= NETTXF_more_data; + + id = get_id_from_freelist(np->tx_skbs); + np->tx_skbs[id] = skb_get(skb); + tx = RING_GET_REQUEST(&np->tx, prod++); + tx->id = id; + ref = gnttab_claim_grant_reference(&np->gref_tx_head); + BUG_ON((signed short)ref < 0); + + mfn = pfn_to_mfn(page_to_pfn(frag->page)); + gnttab_grant_foreign_access_ref(ref, np->xbdev->otherend_id, + mfn, GTF_readonly); + + tx->gref = np->grant_tx_ref[id] = ref; + tx->offset = frag->page_offset; + tx->size = frag->size; + tx->flags = 0; + } + + np->tx.req_prod_pvt = prod; +} + +static int network_start_xmit(struct sk_buff *skb, struct net_device *dev) +{ + unsigned short id; + struct netfront_info *np = netdev_priv(dev); + struct netif_tx_request *tx; + struct netif_extra_info *extra; + char *data = skb->data; + RING_IDX i; + grant_ref_t ref; + unsigned long mfn; + int notify; + int frags = skb_shinfo(skb)->nr_frags; + unsigned int offset = offset_in_page(data); + unsigned int len = skb_headlen(skb); + + /* Check the fast path, if hooks are available */ + if (np->accel_vif_state.hooks && + np->accel_vif_state.hooks->start_xmit(skb, dev)) { + /* Fast path has sent this packet */ + return NETDEV_TX_OK; + } + + frags += DIV_ROUND_UP(offset + len, PAGE_SIZE); + if (unlikely(frags > MAX_SKB_FRAGS + 1)) { + printk(KERN_ALERT "xennet: skb rides the rocket: %d frags\n", + frags); + dump_stack(); + goto drop; + } + + spin_lock_irq(&np->tx_lock); + + if (unlikely(!netfront_carrier_ok(np) || + (frags > 1 && !xennet_can_sg(dev)) || + netif_needs_gso(dev, skb))) { + spin_unlock_irq(&np->tx_lock); + goto drop; + } + + i = np->tx.req_prod_pvt; + + id = get_id_from_freelist(np->tx_skbs); + np->tx_skbs[id] = skb; + + tx = RING_GET_REQUEST(&np->tx, i); + + tx->id = id; + ref = gnttab_claim_grant_reference(&np->gref_tx_head); + BUG_ON((signed short)ref < 0); + mfn = virt_to_mfn(data); + gnttab_grant_foreign_access_ref( + ref, np->xbdev->otherend_id, mfn, GTF_readonly); + tx->gref = np->grant_tx_ref[id] = ref; + tx->offset = offset; + tx->size = len; + + tx->flags = 0; + extra = NULL; + + if (skb->ip_summed == CHECKSUM_PARTIAL) /* local packet? */ + tx->flags |= NETTXF_csum_blank | NETTXF_data_validated; +#ifdef CONFIG_XEN + if (skb->proto_data_valid) /* remote but checksummed? */ + tx->flags |= NETTXF_data_validated; +#endif + +#if HAVE_TSO + if (skb_shinfo(skb)->gso_size) { + struct netif_extra_info *gso = (struct netif_extra_info *) + RING_GET_REQUEST(&np->tx, ++i); + + if (extra) + extra->flags |= XEN_NETIF_EXTRA_FLAG_MORE; + else + tx->flags |= NETTXF_extra_info; + + gso->u.gso.size = skb_shinfo(skb)->gso_size; + gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4; + gso->u.gso.pad = 0; + gso->u.gso.features = 0; + + gso->type = XEN_NETIF_EXTRA_TYPE_GSO; + gso->flags = 0; + extra = gso; + } +#endif + + np->tx.req_prod_pvt = i + 1; + + xennet_make_frags(skb, dev, tx); + tx->size = skb->len; + + RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&np->tx, notify); + if (notify) + notify_remote_via_irq(np->irq); + + np->stats.tx_bytes += skb->len; + np->stats.tx_packets++; + dev->trans_start = jiffies; + + /* Note: It is not safe to access skb after network_tx_buf_gc()! */ + network_tx_buf_gc(dev); + + if (!netfront_tx_slot_available(np)) + netif_stop_queue(dev); + + spin_unlock_irq(&np->tx_lock); + + return NETDEV_TX_OK; + + drop: + np->stats.tx_dropped++; + dev_kfree_skb(skb); + return NETDEV_TX_OK; +} + +static irqreturn_t netif_int(int irq, void *dev_id) +{ + struct net_device *dev = dev_id; + struct netfront_info *np = netdev_priv(dev); + unsigned long flags; + + spin_lock_irqsave(&np->tx_lock, flags); + + if (likely(netfront_carrier_ok(np))) { + network_tx_buf_gc(dev); + /* Under tx_lock: protects access to rx shared-ring indexes. */ + if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx)) { + netfront_accelerator_call_stop_napi_irq(np, dev); + + napi_schedule(&np->napi); + } + } + + spin_unlock_irqrestore(&np->tx_lock, flags); + + return IRQ_HANDLED; +} + +static void xennet_move_rx_slot(struct netfront_info *np, struct sk_buff *skb, + grant_ref_t ref) +{ + int new = xennet_rxidx(np->rx.req_prod_pvt); + + BUG_ON(np->rx_skbs[new]); + np->rx_skbs[new] = skb; + np->grant_rx_ref[new] = ref; + RING_GET_REQUEST(&np->rx, np->rx.req_prod_pvt)->id = new; + RING_GET_REQUEST(&np->rx, np->rx.req_prod_pvt)->gref = ref; + np->rx.req_prod_pvt++; +} + +int xennet_get_extras(struct netfront_info *np, + struct netif_extra_info *extras, RING_IDX rp) + +{ + struct netif_extra_info *extra; + RING_IDX cons = np->rx.rsp_cons; + int err = 0; + + do { + struct sk_buff *skb; + grant_ref_t ref; + + if (unlikely(cons + 1 == rp)) { + if (net_ratelimit()) + WPRINTK("Missing extra info\n"); + err = -EBADR; + break; + } + + extra = (struct netif_extra_info *) + RING_GET_RESPONSE(&np->rx, ++cons); + + if (unlikely(!extra->type || + extra->type >= XEN_NETIF_EXTRA_TYPE_MAX)) { + if (net_ratelimit()) + WPRINTK("Invalid extra type: %d\n", + extra->type); + err = -EINVAL; + } else { + memcpy(&extras[extra->type - 1], extra, + sizeof(*extra)); + } + + skb = xennet_get_rx_skb(np, cons); + ref = xennet_get_rx_ref(np, cons); + xennet_move_rx_slot(np, skb, ref); + } while (extra->flags & XEN_NETIF_EXTRA_FLAG_MORE); + + np->rx.rsp_cons = cons; + return err; +} + +static int xennet_get_responses(struct netfront_info *np, + struct netfront_rx_info *rinfo, RING_IDX rp, + struct sk_buff_head *list, + int *pages_flipped_p) +{ + int pages_flipped = *pages_flipped_p; + struct mmu_update *mmu; + struct multicall_entry *mcl; + struct netif_rx_response *rx = &rinfo->rx; + struct netif_extra_info *extras = rinfo->extras; + RING_IDX cons = np->rx.rsp_cons; + struct sk_buff *skb = xennet_get_rx_skb(np, cons); + grant_ref_t ref = xennet_get_rx_ref(np, cons); + int max = MAX_SKB_FRAGS + (rx->status <= RX_COPY_THRESHOLD); + int frags = 1; + int err = 0; + unsigned long ret; + + if (rx->flags & NETRXF_extra_info) { + err = xennet_get_extras(np, extras, rp); + cons = np->rx.rsp_cons; + } + + for (;;) { + unsigned long mfn; + + if (unlikely(rx->status < 0 || + rx->offset + rx->status > PAGE_SIZE)) { + if (net_ratelimit()) + WPRINTK("rx->offset: %x, size: %u\n", + rx->offset, rx->status); + xennet_move_rx_slot(np, skb, ref); + err = -EINVAL; + goto next; + } + + /* + * This definitely indicates a bug, either in this driver or in + * the backend driver. In future this should flag the bad + * situation to the system controller to reboot the backed. + */ + if (ref == GRANT_INVALID_REF) { + if (net_ratelimit()) + WPRINTK("Bad rx response id %d.\n", rx->id); + err = -EINVAL; + goto next; + } + + if (!np->copying_receiver) { + /* Memory pressure, insufficient buffer + * headroom, ... */ + if (!(mfn = gnttab_end_foreign_transfer_ref(ref))) { + if (net_ratelimit()) + WPRINTK("Unfulfilled rx req " + "(id=%d, st=%d).\n", + rx->id, rx->status); + xennet_move_rx_slot(np, skb, ref); + err = -ENOMEM; + goto next; + } + + if (!xen_feature(XENFEAT_auto_translated_physmap)) { + /* Remap the page. */ + struct page *page = + skb_shinfo(skb)->frags[0].page; + unsigned long pfn = page_to_pfn(page); + void *vaddr = page_address(page); + + mcl = np->rx_mcl + pages_flipped; + mmu = np->rx_mmu + pages_flipped; + + MULTI_update_va_mapping(mcl, + (unsigned long)vaddr, + pfn_pte_ma(mfn, + PAGE_KERNEL), + 0); + mmu->ptr = ((maddr_t)mfn << PAGE_SHIFT) + | MMU_MACHPHYS_UPDATE; + mmu->val = pfn; + + set_phys_to_machine(pfn, mfn); + } + pages_flipped++; + } else { + ret = gnttab_end_foreign_access_ref(ref); + BUG_ON(!ret); + } + + gnttab_release_grant_reference(&np->gref_rx_head, ref); + + __skb_queue_tail(list, skb); + +next: + if (!(rx->flags & NETRXF_more_data)) + break; + + if (cons + frags == rp) { + if (net_ratelimit()) + WPRINTK("Need more frags\n"); + err = -ENOENT; + break; + } + + rx = RING_GET_RESPONSE(&np->rx, cons + frags); + skb = xennet_get_rx_skb(np, cons + frags); + ref = xennet_get_rx_ref(np, cons + frags); + frags++; + } + + if (unlikely(frags > max)) { + if (net_ratelimit()) + WPRINTK("Too many frags\n"); + err = -E2BIG; + } + + if (unlikely(err)) + np->rx.rsp_cons = cons + frags; + + *pages_flipped_p = pages_flipped; + + return err; +} + +static RING_IDX xennet_fill_frags(struct netfront_info *np, + struct sk_buff *skb, + struct sk_buff_head *list) +{ + struct skb_shared_info *shinfo = skb_shinfo(skb); + int nr_frags = shinfo->nr_frags; + RING_IDX cons = np->rx.rsp_cons; + skb_frag_t *frag = shinfo->frags + nr_frags; + struct sk_buff *nskb; + + while ((nskb = __skb_dequeue(list))) { + struct netif_rx_response *rx = + RING_GET_RESPONSE(&np->rx, ++cons); + + frag->page = skb_shinfo(nskb)->frags[0].page; + frag->page_offset = rx->offset; + frag->size = rx->status; + + skb->data_len += rx->status; + + skb_shinfo(nskb)->nr_frags = 0; + kfree_skb(nskb); + + frag++; + nr_frags++; + } + + shinfo->nr_frags = nr_frags; + return cons; +} + +static int xennet_set_skb_gso(struct sk_buff *skb, + struct netif_extra_info *gso) +{ + if (!gso->u.gso.size) { + if (net_ratelimit()) + WPRINTK("GSO size must not be zero.\n"); + return -EINVAL; + } + + /* Currently only TCPv4 S.O. is supported. */ + if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4) { + if (net_ratelimit()) + WPRINTK("Bad GSO type %d.\n", gso->u.gso.type); + return -EINVAL; + } + +#if HAVE_TSO + skb_shinfo(skb)->gso_size = gso->u.gso.size; +#if HAVE_GSO + skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4; + + /* Header must be checked, and gso_segs computed. */ + skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY; +#endif + skb_shinfo(skb)->gso_segs = 0; + + return 0; +#else + if (net_ratelimit()) + WPRINTK("GSO unsupported by this kernel.\n"); + return -EINVAL; +#endif +} + +static int netif_poll(struct napi_struct *napi, int budget) +{ + struct netfront_info *np = container_of(napi, struct netfront_info, napi); + struct net_device *dev = np->netdev; + struct sk_buff *skb; + struct netfront_rx_info rinfo; + struct netif_rx_response *rx = &rinfo.rx; + struct netif_extra_info *extras = rinfo.extras; + RING_IDX i, rp; + struct multicall_entry *mcl; + int work_done, more_to_do = 1, accel_more_to_do = 1; + struct sk_buff_head rxq; + struct sk_buff_head errq; + struct sk_buff_head tmpq; + unsigned long flags; + unsigned int len; + int pages_flipped = 0; + int err; + + spin_lock(&np->rx_lock); /* no need for spin_lock_bh() in ->poll() */ + + if (unlikely(!netfront_carrier_ok(np))) { + spin_unlock(&np->rx_lock); + return 0; + } + + skb_queue_head_init(&rxq); + skb_queue_head_init(&errq); + skb_queue_head_init(&tmpq); + + rp = np->rx.sring->rsp_prod; + rmb(); /* Ensure we see queued responses up to 'rp'. */ + + i = np->rx.rsp_cons; + work_done = 0; + while ((i != rp) && (work_done < budget)) { + memcpy(rx, RING_GET_RESPONSE(&np->rx, i), sizeof(*rx)); + memset(extras, 0, sizeof(rinfo.extras)); + + err = xennet_get_responses(np, &rinfo, rp, &tmpq, + &pages_flipped); + + if (unlikely(err)) { +err: + while ((skb = __skb_dequeue(&tmpq))) + __skb_queue_tail(&errq, skb); + np->stats.rx_errors++; + i = np->rx.rsp_cons; + continue; + } + + skb = __skb_dequeue(&tmpq); + + if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) { + struct netif_extra_info *gso; + gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1]; + + if (unlikely(xennet_set_skb_gso(skb, gso))) { + __skb_queue_head(&tmpq, skb); + np->rx.rsp_cons += skb_queue_len(&tmpq); + goto err; + } + } + + NETFRONT_SKB_CB(skb)->page = skb_shinfo(skb)->frags[0].page; + NETFRONT_SKB_CB(skb)->offset = rx->offset; + + len = rx->status; + if (len > RX_COPY_THRESHOLD) + len = RX_COPY_THRESHOLD; + skb_put(skb, len); + + if (rx->status > len) { + skb_shinfo(skb)->frags[0].page_offset = + rx->offset + len; + skb_shinfo(skb)->frags[0].size = rx->status - len; + skb->data_len = rx->status - len; + } else { + skb_shinfo(skb)->frags[0].page = NULL; + skb_shinfo(skb)->nr_frags = 0; + } + + i = xennet_fill_frags(np, skb, &tmpq); + + /* + * Truesize must approximates the size of true data plus + * any supervisor overheads. Adding hypervisor overheads + * has been shown to significantly reduce achievable + * bandwidth with the default receive buffer size. It is + * therefore not wise to account for it here. + * + * After alloc_skb(RX_COPY_THRESHOLD), truesize is set to + * RX_COPY_THRESHOLD + the supervisor overheads. Here, we + * add the size of the data pulled in xennet_fill_frags(). + * + * We also adjust for any unused space in the main data + * area by subtracting (RX_COPY_THRESHOLD - len). This is + * especially important with drivers which split incoming + * packets into header and data, using only 66 bytes of + * the main data area (see the e1000 driver for example.) + * On such systems, without this last adjustement, our + * achievable receive throughout using the standard receive + * buffer size was cut by 25%(!!!). + */ + skb->truesize += skb->data_len - (RX_COPY_THRESHOLD - len); + skb->len += skb->data_len; + + /* + * Old backends do not assert data_validated but we + * can infer it from csum_blank so test both flags. + */ + if (rx->flags & (NETRXF_data_validated|NETRXF_csum_blank)) + skb->ip_summed = CHECKSUM_UNNECESSARY; + else + skb->ip_summed = CHECKSUM_NONE; +#ifdef CONFIG_XEN + skb->proto_data_valid = (skb->ip_summed != CHECKSUM_NONE); + skb->proto_csum_blank = !!(rx->flags & NETRXF_csum_blank); +#endif + np->stats.rx_packets++; + np->stats.rx_bytes += skb->len; + + __skb_queue_tail(&rxq, skb); + + np->rx.rsp_cons = ++i; + work_done++; + } + + if (pages_flipped) { + /* Some pages are no longer absent... */ + balloon_update_driver_allowance(-pages_flipped); + + /* Do all the remapping work and M2P updates. */ + if (!xen_feature(XENFEAT_auto_translated_physmap)) { + mcl = np->rx_mcl + pages_flipped; + mcl->op = __HYPERVISOR_mmu_update; + mcl->args[0] = (unsigned long)np->rx_mmu; + mcl->args[1] = pages_flipped; + mcl->args[2] = 0; + mcl->args[3] = DOMID_SELF; + err = HYPERVISOR_multicall_check(np->rx_mcl, + pages_flipped + 1, + NULL); + BUG_ON(err); + } + } + + __skb_queue_purge(&errq); + + while ((skb = __skb_dequeue(&rxq)) != NULL) { + struct page *page = NETFRONT_SKB_CB(skb)->page; + void *vaddr = page_address(page); + unsigned offset = NETFRONT_SKB_CB(skb)->offset; + + memcpy(skb->data, vaddr + offset, skb_headlen(skb)); + + if (page != skb_shinfo(skb)->frags[0].page) + __free_page(page); + + /* Ethernet work: Delayed to here as it peeks the header. */ + skb->protocol = eth_type_trans(skb, dev); + + /* Pass it up. */ + netif_receive_skb(skb); + } + + /* If we get a callback with very few responses, reduce fill target. */ + /* NB. Note exponential increase, linear decrease. */ + if (((np->rx.req_prod_pvt - np->rx.sring->rsp_prod) > + ((3*np->rx_target) / 4)) && + (--np->rx_target < np->rx_min_target)) + np->rx_target = np->rx_min_target; + + network_alloc_rx_buffers(dev); + + if (work_done < budget) { + /* there's some spare capacity, try the accelerated path */ + int accel_budget = budget - work_done; + int accel_budget_start = accel_budget; + + if (np->accel_vif_state.hooks) { + accel_more_to_do = + np->accel_vif_state.hooks->netdev_poll + (dev, &accel_budget); + work_done += (accel_budget_start - accel_budget); + } else + accel_more_to_do = 0; + } + + if (work_done < budget) { + local_irq_save(flags); + + RING_FINAL_CHECK_FOR_RESPONSES(&np->rx, more_to_do); + + if (!more_to_do && !accel_more_to_do && + np->accel_vif_state.hooks) { + /* + * Slow path has nothing more to do, see if + * fast path is likewise + */ + accel_more_to_do = + np->accel_vif_state.hooks->start_napi_irq(dev); + } + + if (!more_to_do && !accel_more_to_do) + __napi_complete(napi); + + local_irq_restore(flags); + } + + spin_unlock(&np->rx_lock); + + return work_done; +} + +static void netif_release_tx_bufs(struct netfront_info *np) +{ + struct sk_buff *skb; + int i; + + for (i = 1; i <= NET_TX_RING_SIZE; i++) { + if ((unsigned long)np->tx_skbs[i] < PAGE_OFFSET) + continue; + + skb = np->tx_skbs[i]; + gnttab_end_foreign_access_ref(np->grant_tx_ref[i]); + gnttab_release_grant_reference( + &np->gref_tx_head, np->grant_tx_ref[i]); + np->grant_tx_ref[i] = GRANT_INVALID_REF; + add_id_to_freelist(np->tx_skbs, i); + dev_kfree_skb_irq(skb); + } +} + +static void netif_release_rx_bufs_flip(struct netfront_info *np) +{ + struct mmu_update *mmu = np->rx_mmu; + struct multicall_entry *mcl = np->rx_mcl; + struct sk_buff_head free_list; + struct sk_buff *skb; + unsigned long mfn; + int xfer = 0, noxfer = 0, unused = 0; + int id, ref, rc; + + skb_queue_head_init(&free_list); + + spin_lock_bh(&np->rx_lock); + + for (id = 0; id < NET_RX_RING_SIZE; id++) { + if ((ref = np->grant_rx_ref[id]) == GRANT_INVALID_REF) { + unused++; + continue; + } + + skb = np->rx_skbs[id]; + mfn = gnttab_end_foreign_transfer_ref(ref); + gnttab_release_grant_reference(&np->gref_rx_head, ref); + np->grant_rx_ref[id] = GRANT_INVALID_REF; + add_id_to_freelist(np->rx_skbs, id); + + if (0 == mfn) { + struct page *page = skb_shinfo(skb)->frags[0].page; + balloon_release_driver_page(page); + skb_shinfo(skb)->nr_frags = 0; + dev_kfree_skb(skb); + noxfer++; + continue; + } + + if (!xen_feature(XENFEAT_auto_translated_physmap)) { + /* Remap the page. */ + struct page *page = skb_shinfo(skb)->frags[0].page; + unsigned long pfn = page_to_pfn(page); + void *vaddr = page_address(page); + + MULTI_update_va_mapping(mcl, (unsigned long)vaddr, + pfn_pte_ma(mfn, PAGE_KERNEL), + 0); + mcl++; + mmu->ptr = ((maddr_t)mfn << PAGE_SHIFT) + | MMU_MACHPHYS_UPDATE; + mmu->val = pfn; + mmu++; + + set_phys_to_machine(pfn, mfn); + } + __skb_queue_tail(&free_list, skb); + xfer++; + } + + DPRINTK("%s: %d xfer, %d noxfer, %d unused\n", + __FUNCTION__, xfer, noxfer, unused); + + if (xfer) { + /* Some pages are no longer absent... */ + balloon_update_driver_allowance(-xfer); + + if (!xen_feature(XENFEAT_auto_translated_physmap)) { + /* Do all the remapping work and M2P updates. */ + mcl->op = __HYPERVISOR_mmu_update; + mcl->args[0] = (unsigned long)np->rx_mmu; + mcl->args[1] = mmu - np->rx_mmu; + mcl->args[2] = 0; + mcl->args[3] = DOMID_SELF; + mcl++; + rc = HYPERVISOR_multicall_check( + np->rx_mcl, mcl - np->rx_mcl, NULL); + BUG_ON(rc); + } + } + + __skb_queue_purge(&free_list); + + spin_unlock_bh(&np->rx_lock); +} + +static void netif_release_rx_bufs_copy(struct netfront_info *np) +{ + struct sk_buff *skb; + int i, ref; + int busy = 0, inuse = 0; + + spin_lock_bh(&np->rx_lock); + + for (i = 0; i < NET_RX_RING_SIZE; i++) { + ref = np->grant_rx_ref[i]; + + if (ref == GRANT_INVALID_REF) + continue; + + inuse++; + + skb = np->rx_skbs[i]; + + if (!gnttab_end_foreign_access_ref(ref)) + { + busy++; + continue; + } + + gnttab_release_grant_reference(&np->gref_rx_head, ref); + np->grant_rx_ref[i] = GRANT_INVALID_REF; + add_id_to_freelist(np->rx_skbs, i); + + dev_kfree_skb(skb); + } + + if (busy) + DPRINTK("%s: Unable to release %d of %d inuse grant references out of %ld total.\n", + __FUNCTION__, busy, inuse, NET_RX_RING_SIZE); + + spin_unlock_bh(&np->rx_lock); +} + +static int network_close(struct net_device *dev) +{ + struct netfront_info *np = netdev_priv(dev); + netif_stop_queue(np->netdev); + napi_disable(&np->napi); + return 0; +} + + +static struct net_device_stats *network_get_stats(struct net_device *dev) +{ + struct netfront_info *np = netdev_priv(dev); + + netfront_accelerator_call_get_stats(np, dev); + return &np->stats; +} + +static int xennet_set_mac_address(struct net_device *dev, void *p) +{ + struct netfront_info *np = netdev_priv(dev); + struct sockaddr *addr = p; + + if (netif_running(dev)) + return -EBUSY; + + if (!is_valid_ether_addr(addr->sa_data)) + return -EADDRNOTAVAIL; + + memcpy(dev->dev_addr, addr->sa_data, dev->addr_len); + memcpy(np->mac, addr->sa_data, ETH_ALEN); + + return 0; +} + +static int xennet_change_mtu(struct net_device *dev, int mtu) +{ + int max = xennet_can_sg(dev) ? 65535 - ETH_HLEN : ETH_DATA_LEN; + + if (mtu > max) + return -EINVAL; + dev->mtu = mtu; + return 0; +} + +static int xennet_set_sg(struct net_device *dev, u32 data) +{ + if (data) { + struct netfront_info *np = netdev_priv(dev); + int val; + + if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, "feature-sg", + "%d", &val) < 0) + val = 0; + if (!val) + return -ENOSYS; + } else if (dev->mtu > ETH_DATA_LEN) + dev->mtu = ETH_DATA_LEN; + + return ethtool_op_set_sg(dev, data); +} + +static int xennet_set_tso(struct net_device *dev, u32 data) +{ + if (data) { + struct netfront_info *np = netdev_priv(dev); + int val; + + if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, + "feature-gso-tcpv4", "%d", &val) < 0) + val = 0; + if (!val) + return -ENOSYS; + } + + return ethtool_op_set_tso(dev, data); +} + +static void xennet_set_features(struct net_device *dev) +{ + dev_disable_gso_features(dev); + xennet_set_sg(dev, 0); + + /* We need checksum offload to enable scatter/gather and TSO. */ + if (!(dev->features & NETIF_F_IP_CSUM)) + return; + + if (xennet_set_sg(dev, 1)) + return; + + /* Before 2.6.9 TSO seems to be unreliable so do not enable it + * on older kernels. + */ + if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,9)) + xennet_set_tso(dev, 1); +} + +static void netfront_get_drvinfo(struct net_device *dev, + struct ethtool_drvinfo *info) +{ + strcpy(info->driver, "netfront"); + strcpy(info->bus_info, dev_name(dev->dev.parent)); +} + +static int network_connect(struct net_device *dev) +{ + struct netfront_info *np = netdev_priv(dev); + int i, requeue_idx, err; + struct sk_buff *skb; + grant_ref_t ref; + netif_rx_request_t *req; + unsigned int feature_rx_copy, feature_rx_flip; + + err = xenbus_scanf(XBT_NIL, np->xbdev->otherend, + "feature-rx-copy", "%u", &feature_rx_copy); + if (err != 1) + feature_rx_copy = 0; + err = xenbus_scanf(XBT_NIL, np->xbdev->otherend, + "feature-rx-flip", "%u", &feature_rx_flip); + if (err != 1) + feature_rx_flip = 1; + + /* + * Copy packets on receive path if: + * (a) This was requested by user, and the backend supports it; or + * (b) Flipping was requested, but this is unsupported by the backend. + */ + np->copying_receiver = ((MODPARM_rx_copy && feature_rx_copy) || + (MODPARM_rx_flip && !feature_rx_flip)); + + err = talk_to_backend(np->xbdev, np); + if (err) + return err; + + xennet_set_features(dev); + + DPRINTK("device %s has %sing receive path.\n", + dev->name, np->copying_receiver ? "copy" : "flipp"); + + spin_lock_bh(&np->rx_lock); + spin_lock_irq(&np->tx_lock); + + /* + * Recovery procedure: + * NB. Freelist index entries are always going to be less than + * PAGE_OFFSET, whereas pointers to skbs will always be equal or + * greater than PAGE_OFFSET: we use this property to distinguish + * them. + */ + + /* Step 1: Discard all pending TX packet fragments. */ + netif_release_tx_bufs(np); + + /* Step 2: Rebuild the RX buffer freelist and the RX ring itself. */ + for (requeue_idx = 0, i = 0; i < NET_RX_RING_SIZE; i++) { + if (!np->rx_skbs[i]) + continue; + + skb = np->rx_skbs[requeue_idx] = xennet_get_rx_skb(np, i); + ref = np->grant_rx_ref[requeue_idx] = xennet_get_rx_ref(np, i); + req = RING_GET_REQUEST(&np->rx, requeue_idx); + + if (!np->copying_receiver) { + gnttab_grant_foreign_transfer_ref( + ref, np->xbdev->otherend_id, + page_to_pfn(skb_shinfo(skb)->frags->page)); + } else { + gnttab_grant_foreign_access_ref( + ref, np->xbdev->otherend_id, + pfn_to_mfn(page_to_pfn(skb_shinfo(skb)-> + frags->page)), + 0); + } + req->gref = ref; + req->id = requeue_idx; + + requeue_idx++; + } + + np->rx.req_prod_pvt = requeue_idx; + + /* + * Step 3: All public and private state should now be sane. Get + * ready to start sending and receiving packets and give the driver + * domain a kick because we've probably just requeued some + * packets. + */ + netfront_carrier_on(np); + notify_remote_via_irq(np->irq); + network_tx_buf_gc(dev); + network_alloc_rx_buffers(dev); + + spin_unlock_irq(&np->tx_lock); + spin_unlock_bh(&np->rx_lock); + + return 0; +} + +static void netif_uninit(struct net_device *dev) +{ + struct netfront_info *np = netdev_priv(dev); + netif_release_tx_bufs(np); + if (np->copying_receiver) + netif_release_rx_bufs_copy(np); + else + netif_release_rx_bufs_flip(np); + gnttab_free_grant_references(np->gref_tx_head); + gnttab_free_grant_references(np->gref_rx_head); +} + +static const struct ethtool_ops network_ethtool_ops = +{ + .get_drvinfo = netfront_get_drvinfo, + .get_tx_csum = ethtool_op_get_tx_csum, + .set_tx_csum = ethtool_op_set_tx_csum, + .get_sg = ethtool_op_get_sg, + .set_sg = xennet_set_sg, +#if HAVE_TSO + .get_tso = ethtool_op_get_tso, + .set_tso = xennet_set_tso, +#endif + .get_link = ethtool_op_get_link, +}; + +#ifdef CONFIG_SYSFS +static ssize_t show_rxbuf_min(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct netfront_info *info = netdev_priv(to_net_dev(dev)); + + return sprintf(buf, "%u\n", info->rx_min_target); +} + +static ssize_t store_rxbuf_min(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t len) +{ + struct net_device *netdev = to_net_dev(dev); + struct netfront_info *np = netdev_priv(netdev); + char *endp; + unsigned long target; + + if (!capable(CAP_NET_ADMIN)) + return -EPERM; + + target = simple_strtoul(buf, &endp, 0); + if (endp == buf) + return -EBADMSG; + + if (target < RX_MIN_TARGET) + target = RX_MIN_TARGET; + if (target > RX_MAX_TARGET) + target = RX_MAX_TARGET; + + spin_lock_bh(&np->rx_lock); + if (target > np->rx_max_target) + np->rx_max_target = target; + np->rx_min_target = target; + if (target > np->rx_target) + np->rx_target = target; + + network_alloc_rx_buffers(netdev); + + spin_unlock_bh(&np->rx_lock); + return len; +} + +static ssize_t show_rxbuf_max(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct netfront_info *info = netdev_priv(to_net_dev(dev)); + + return sprintf(buf, "%u\n", info->rx_max_target); +} + +static ssize_t store_rxbuf_max(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t len) +{ + struct net_device *netdev = to_net_dev(dev); + struct netfront_info *np = netdev_priv(netdev); + char *endp; + unsigned long target; + + if (!capable(CAP_NET_ADMIN)) + return -EPERM; + + target = simple_strtoul(buf, &endp, 0); + if (endp == buf) + return -EBADMSG; + + if (target < RX_MIN_TARGET) + target = RX_MIN_TARGET; + if (target > RX_MAX_TARGET) + target = RX_MAX_TARGET; + + spin_lock_bh(&np->rx_lock); + if (target < np->rx_min_target) + np->rx_min_target = target; + np->rx_max_target = target; + if (target < np->rx_target) + np->rx_target = target; + + network_alloc_rx_buffers(netdev); + + spin_unlock_bh(&np->rx_lock); + return len; +} + +static ssize_t show_rxbuf_cur(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct netfront_info *info = netdev_priv(to_net_dev(dev)); + + return sprintf(buf, "%u\n", info->rx_target); +} + +static struct device_attribute xennet_attrs[] = { + __ATTR(rxbuf_min, S_IRUGO|S_IWUSR, show_rxbuf_min, store_rxbuf_min), + __ATTR(rxbuf_max, S_IRUGO|S_IWUSR, show_rxbuf_max, store_rxbuf_max), + __ATTR(rxbuf_cur, S_IRUGO, show_rxbuf_cur, NULL), +}; + +static int xennet_sysfs_addif(struct net_device *netdev) +{ + int i; + int error = 0; + + for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) { + error = device_create_file(&netdev->dev, + &xennet_attrs[i]); + if (error) + goto fail; + } + return 0; + + fail: + while (--i >= 0) + device_remove_file(&netdev->dev, &xennet_attrs[i]); + return error; +} + +static void xennet_sysfs_delif(struct net_device *netdev) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) + device_remove_file(&netdev->dev, &xennet_attrs[i]); +} + +#endif /* CONFIG_SYSFS */ + + +/* + * Nothing to do here. Virtual interface is point-to-point and the + * physical interface is probably promiscuous anyway. + */ +static void network_set_multicast_list(struct net_device *dev) +{ +} + +static const struct net_device_ops xennet_netdev_ops = { + .ndo_uninit = netif_uninit, + .ndo_open = network_open, + .ndo_stop = network_close, + .ndo_start_xmit = network_start_xmit, + .ndo_set_multicast_list = network_set_multicast_list, + .ndo_set_mac_address = xennet_set_mac_address, + .ndo_validate_addr = eth_validate_addr, + .ndo_change_mtu = xennet_change_mtu, + .ndo_get_stats = network_get_stats, +}; + +static struct net_device * __devinit create_netdev(struct xenbus_device *dev) +{ + int i, err = 0; + struct net_device *netdev = NULL; + struct netfront_info *np = NULL; + + netdev = alloc_etherdev(sizeof(struct netfront_info)); + if (!netdev) { + printk(KERN_WARNING "%s> alloc_etherdev failed.\n", + __FUNCTION__); + return ERR_PTR(-ENOMEM); + } + + np = netdev_priv(netdev); + np->xbdev = dev; + + spin_lock_init(&np->tx_lock); + spin_lock_init(&np->rx_lock); + + init_accelerator_vif(np, dev); + + skb_queue_head_init(&np->rx_batch); + np->rx_target = RX_DFL_MIN_TARGET; + np->rx_min_target = RX_DFL_MIN_TARGET; + np->rx_max_target = RX_MAX_TARGET; + + init_timer(&np->rx_refill_timer); + np->rx_refill_timer.data = (unsigned long)netdev; + np->rx_refill_timer.function = rx_refill_timeout; + + /* Initialise {tx,rx}_skbs as a free chain containing every entry. */ + for (i = 0; i <= NET_TX_RING_SIZE; i++) { + np->tx_skbs[i] = (void *)((unsigned long) i+1); + np->grant_tx_ref[i] = GRANT_INVALID_REF; + } + + for (i = 0; i < NET_RX_RING_SIZE; i++) { + np->rx_skbs[i] = NULL; + np->grant_rx_ref[i] = GRANT_INVALID_REF; + } + + /* A grant for every tx ring slot */ + if (gnttab_alloc_grant_references(TX_MAX_TARGET, + &np->gref_tx_head) < 0) { + printk(KERN_ALERT "#### netfront can't alloc tx grant refs\n"); + err = -ENOMEM; + goto exit; + } + /* A grant for every rx ring slot */ + if (gnttab_alloc_grant_references(RX_MAX_TARGET, + &np->gref_rx_head) < 0) { + printk(KERN_ALERT "#### netfront can't alloc rx grant refs\n"); + err = -ENOMEM; + goto exit_free_tx; + } + + netdev->netdev_ops = &xennet_netdev_ops; + netif_napi_add(netdev, &np->napi, netif_poll, 64); + netdev->features = NETIF_F_IP_CSUM; + + SET_ETHTOOL_OPS(netdev, &network_ethtool_ops); + SET_NETDEV_DEV(netdev, &dev->dev); + + np->netdev = netdev; + + netfront_carrier_off(np); + + return netdev; + + exit_free_tx: + gnttab_free_grant_references(np->gref_tx_head); + exit: + free_netdev(netdev); + return ERR_PTR(err); +} + +#ifdef CONFIG_INET +/* + * We use this notifier to send out a fake ARP reply to reset switches and + * router ARP caches when an IP interface is brought up on a VIF. + */ +static int +inetdev_notify(struct notifier_block *this, unsigned long event, void *ptr) +{ + struct in_ifaddr *ifa = (struct in_ifaddr *)ptr; + struct net_device *dev = ifa->ifa_dev->dev; + + /* UP event and is it one of our devices? */ + if (event == NETDEV_UP && dev->netdev_ops->ndo_open == network_open) + send_fake_arp(dev); + + return NOTIFY_DONE; +} + +static struct notifier_block notifier_inetdev = { + .notifier_call = inetdev_notify, + .next = NULL, + .priority = 0 +}; +#endif + + +static void netif_disconnect_backend(struct netfront_info *info) +{ + /* Stop old i/f to prevent errors whilst we rebuild the state. */ + spin_lock_bh(&info->rx_lock); + spin_lock_irq(&info->tx_lock); + netfront_carrier_off(info); + spin_unlock_irq(&info->tx_lock); + spin_unlock_bh(&info->rx_lock); + + if (info->irq) + unbind_from_irqhandler(info->irq, info->netdev); + info->irq = 0; + + end_access(info->tx_ring_ref, info->tx.sring); + end_access(info->rx_ring_ref, info->rx.sring); + info->tx_ring_ref = GRANT_INVALID_REF; + info->rx_ring_ref = GRANT_INVALID_REF; + info->tx.sring = NULL; + info->rx.sring = NULL; +} + + +static void end_access(int ref, void *page) +{ + if (ref != GRANT_INVALID_REF) + gnttab_end_foreign_access(ref, (unsigned long)page); +} + + +/* ** Driver registration ** */ + + +static const struct xenbus_device_id netfront_ids[] = { + { "vif" }, + { "" } +}; +MODULE_ALIAS("xen:vif"); + + +static struct xenbus_driver netfront_driver = { + .name = "vif", + .ids = netfront_ids, + .probe = netfront_probe, + .remove = __devexit_p(netfront_remove), + .suspend = netfront_suspend, + .suspend_cancel = netfront_suspend_cancel, + .resume = netfront_resume, + .otherend_changed = backend_changed, +}; + + +static int __init netif_init(void) +{ + int err; + + if (!is_running_on_xen()) + return -ENODEV; + +#ifdef CONFIG_XEN + if (MODPARM_rx_flip && MODPARM_rx_copy) { + WPRINTK("Cannot specify both rx_copy and rx_flip.\n"); + return -EINVAL; + } + + if (!MODPARM_rx_flip && !MODPARM_rx_copy) + MODPARM_rx_flip = 1; /* Default is to flip. */ +#endif + + netif_init_accel(); + + IPRINTK("Initialising virtual ethernet driver.\n"); + +#ifdef CONFIG_INET + (void)register_inetaddr_notifier(¬ifier_inetdev); +#endif + + err = xenbus_register_frontend(&netfront_driver); + if (err) { +#ifdef CONFIG_INET + unregister_inetaddr_notifier(¬ifier_inetdev); +#endif + } + return err; +} +module_init(netif_init); + + +static void __exit netif_exit(void) +{ +#ifdef CONFIG_INET + unregister_inetaddr_notifier(¬ifier_inetdev); +#endif + xenbus_unregister_driver(&netfront_driver); + + netif_exit_accel(); +} +module_exit(netif_exit); + +MODULE_LICENSE("Dual BSD/GPL"); diff --git a/drivers/xenpvdrivers/netfront.h b/drivers/xenpvdrivers/netfront.h new file mode 100644 index 0000000..3c13436 --- /dev/null +++ b/drivers/xenpvdrivers/netfront.h @@ -0,0 +1,276 @@ +/****************************************************************************** + * Virtual network driver for conversing with remote driver backends. + * + * Copyright (c) 2002-2005, K A Fraser + * Copyright (c) 2005, XenSource Ltd + * Copyright (C) 2007 Solarflare Communications, Inc. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef NETFRONT_H +#define NETFRONT_H + +#include +#include +#include +#include + +#define NET_TX_RING_SIZE __CONST_RING_SIZE(netif_tx, PAGE_SIZE) +#define NET_RX_RING_SIZE __CONST_RING_SIZE(netif_rx, PAGE_SIZE) + +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +/* + * Function pointer table for hooks into a network acceleration + * plugin. These are called at appropriate points from the netfront + * driver + */ +struct netfront_accel_hooks { + /* + * new_device: Accelerator hook to ask the plugin to support a + * new network interface + */ + int (*new_device)(struct net_device *net_dev, struct xenbus_device *dev); + /* + * remove: Opposite of new_device + */ + int (*remove)(struct xenbus_device *dev); + /* + * The net_device is being polled, check the accelerated + * hardware for any pending packets + */ + int (*netdev_poll)(struct net_device *dev, int *pbudget); + /* + * start_xmit: Used to give the accelerated plugin the option + * of sending a packet. Returns non-zero if has done so, or + * zero to decline and force the packet onto normal send + * path + */ + int (*start_xmit)(struct sk_buff *skb, struct net_device *dev); + /* + * start/stop_napi_interrupts Used by netfront to indicate + * when napi interrupts should be enabled or disabled + */ + int (*start_napi_irq)(struct net_device *dev); + void (*stop_napi_irq)(struct net_device *dev); + /* + * Called before re-enabling the TX queue to check the fast + * path has slots too + */ + int (*check_ready)(struct net_device *dev); + /* + * Get the fastpath network statistics + */ + int (*get_stats)(struct net_device *dev, + struct net_device_stats *stats); +}; + + +/* Version of API/protocol for communication between netfront and + acceleration plugin supported */ +#define NETFRONT_ACCEL_VERSION 0x00010003 + +/* + * Per-netfront device state for the accelerator. This is used to + * allow efficient per-netfront device access to the accelerator + * hooks + */ +struct netfront_accel_vif_state { + struct list_head link; + + struct xenbus_device *dev; + struct netfront_info *np; + struct netfront_accel_hooks *hooks; + + /* Watch on the accelerator configuration value */ + struct xenbus_watch accel_watch; + /* Work item to process change in accelerator */ + struct work_struct accel_work; + /* The string from xenbus last time accel_watch fired */ + char *accel_frontend; +}; + +/* + * Per-accelerator state stored in netfront. These form a list that + * is used to track which devices are accelerated by which plugins, + * and what plugins are available/have been requested + */ +struct netfront_accelerator { + /* Used to make a list */ + struct list_head link; + /* ID of the accelerator */ + int id; + /* + * String describing the accelerator. Currently this is the + * name of the accelerator module. This is provided by the + * backend accelerator through xenstore + */ + char *frontend; + /* The hooks into the accelerator plugin module */ + struct netfront_accel_hooks *hooks; + + /* + * List of per-netfront device state (struct + * netfront_accel_vif_state) for each netfront device that is + * using this accelerator + */ + struct list_head vif_states; + spinlock_t vif_states_lock; +}; + +struct netfront_info { + struct list_head list; + struct net_device *netdev; + + struct net_device_stats stats; + + struct netif_tx_front_ring tx; + struct netif_rx_front_ring rx; + + spinlock_t tx_lock; + spinlock_t rx_lock; + + struct napi_struct napi; + + unsigned int irq; + unsigned int copying_receiver; + unsigned int carrier; + + /* Receive-ring batched refills. */ +#define RX_MIN_TARGET 8 +#define RX_DFL_MIN_TARGET 64 +#define RX_MAX_TARGET min_t(int, NET_RX_RING_SIZE, 256) + unsigned rx_min_target, rx_max_target, rx_target; + struct sk_buff_head rx_batch; + + struct timer_list rx_refill_timer; + + /* + * {tx,rx}_skbs store outstanding skbuffs. The first entry in tx_skbs + * is an index into a chain of free entries. + */ + struct sk_buff *tx_skbs[NET_TX_RING_SIZE+1]; + struct sk_buff *rx_skbs[NET_RX_RING_SIZE]; + +#define TX_MAX_TARGET min_t(int, NET_RX_RING_SIZE, 256) + grant_ref_t gref_tx_head; + grant_ref_t grant_tx_ref[NET_TX_RING_SIZE + 1]; + grant_ref_t gref_rx_head; + grant_ref_t grant_rx_ref[NET_RX_RING_SIZE]; + + struct xenbus_device *xbdev; + int tx_ring_ref; + int rx_ring_ref; + u8 mac[ETH_ALEN]; + + unsigned long rx_pfn_array[NET_RX_RING_SIZE]; + struct multicall_entry rx_mcl[NET_RX_RING_SIZE+1]; + struct mmu_update rx_mmu[NET_RX_RING_SIZE]; + + /* Private pointer to state internal to accelerator module */ + void *accel_priv; + /* The accelerator used by this netfront device */ + struct netfront_accelerator *accelerator; + /* The accelerator state for this netfront device */ + struct netfront_accel_vif_state accel_vif_state; +}; + + +/* Exported Functions */ + +/* + * Called by an accelerator plugin module when it has loaded. + * + * frontend: the string describing the accelerator, currently the module name + * hooks: the hooks for netfront to use to call into the accelerator + * version: the version of API between frontend and plugin requested + * + * return: 0 on success, <0 on error, >0 (with version supported) on + * version mismatch + */ +extern int netfront_accelerator_loaded(int version, const char *frontend, + struct netfront_accel_hooks *hooks); + +/* + * Called by an accelerator plugin module when it is about to unload. + * + * frontend: the string describing the accelerator. Must match the + * one passed to netfront_accelerator_loaded() + */ +extern void netfront_accelerator_stop(const char *frontend); + +/* + * Called by an accelerator before waking the net device's TX queue to + * ensure the slow path has available slots. Returns true if OK to + * wake, false if still busy + */ +extern int netfront_check_queue_ready(struct net_device *net_dev); + + +/* Internal-to-netfront Functions */ + +/* + * Call into accelerator and check to see if it has tx space before we + * wake the net device's TX queue. Returns true if OK to wake, false + * if still busy + */ +extern +int netfront_check_accelerator_queue_ready(struct net_device *dev, + struct netfront_info *np); +extern +int netfront_accelerator_call_remove(struct netfront_info *np, + struct xenbus_device *dev); +extern +int netfront_accelerator_suspend(struct netfront_info *np, + struct xenbus_device *dev); +extern +int netfront_accelerator_suspend_cancel(struct netfront_info *np, + struct xenbus_device *dev); +extern +void netfront_accelerator_resume(struct netfront_info *np, + struct xenbus_device *dev); +extern +void netfront_accelerator_call_stop_napi_irq(struct netfront_info *np, + struct net_device *dev); +extern +int netfront_accelerator_call_get_stats(struct netfront_info *np, + struct net_device *dev); +extern +void netfront_accelerator_add_watch(struct netfront_info *np); + +extern +void netif_init_accel(void); +extern +void netif_exit_accel(void); + +extern +void init_accelerator_vif(struct netfront_info *np, + struct xenbus_device *dev); +#endif /* NETFRONT_H */ diff --git a/drivers/xenpvdrivers/overrides.mk b/drivers/xenpvdrivers/overrides.mk new file mode 100644 index 0000000..a0e9730 --- /dev/null +++ b/drivers/xenpvdrivers/overrides.mk @@ -0,0 +1,17 @@ +# Hack: we need to use the config which was used to build the kernel, +# except that that won't have the right headers etc., so duplicate +# some of the mach-xen infrastructure in here. +# +# (i.e. we need the native config for things like -mregparm, but +# a Xen kernel to find the right headers) +_XEN_CPPFLAGS += -D__XEN_INTERFACE_VERSION__=0x00030205 +_XEN_CPPFLAGS += -DCONFIG_XEN_COMPAT=0xffffff +_XEN_CPPFLAGS += -DHAVE_XEN_PLATFORM_COMPAT_H +ifeq ($(ARCH),ia64) + _XEN_CPPFLAGS += -DCONFIG_VMX_GUEST +endif + +_XEN_CPPFLAGS += -include $(objtree)/include/linux/autoconf.h + +EXTRA_CFLAGS += $(_XEN_CPPFLAGS) +EXTRA_AFLAGS += $(_XEN_CPPFLAGS) diff --git a/drivers/xenpvdrivers/panic-handler.c b/drivers/xenpvdrivers/panic-handler.c new file mode 100644 index 0000000..8aaead7 --- /dev/null +++ b/drivers/xenpvdrivers/panic-handler.c @@ -0,0 +1,42 @@ +#include +#include +#include +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +MODULE_LICENSE("GPL"); + +#ifdef __ia64__ +static void +xen_panic_hypercall(struct unw_frame_info *info, void *arg) +{ + current->thread.ksp = (__u64)info->sw - 16; + HYPERVISOR_shutdown(SHUTDOWN_crash); + /* we're never actually going to get here... */ +} +#endif + +static int +xen_panic_event(struct notifier_block *this, unsigned long event, void *ptr) +{ +#ifdef __ia64__ + unw_init_running(xen_panic_hypercall, NULL); +#else /* !__ia64__ */ + HYPERVISOR_shutdown(SHUTDOWN_crash); +#endif + /* we're never actually going to get here... */ + return NOTIFY_DONE; +} + +static struct notifier_block xen_panic_block = { + .notifier_call = xen_panic_event +}; + +int xen_panic_handler_init(void) +{ + atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block); + return 0; +} diff --git a/drivers/xenpvdrivers/platform-compat.c b/drivers/xenpvdrivers/platform-compat.c new file mode 100644 index 0000000..3103c19 --- /dev/null +++ b/drivers/xenpvdrivers/platform-compat.c @@ -0,0 +1,147 @@ +#include + +#include +#include +#include +#include + +#include + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,7) +static int system_state = 1; +EXPORT_SYMBOL(system_state); +#endif + +void xenpvdrivers_ctrl_alt_del(void) +{ +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27) + kill_proc(1, SIGINT, 1); /* interrupt init */ +#else + kill_cad_pid(SIGINT, 1); +#endif +} + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,8) +size_t strcspn(const char *s, const char *reject) +{ + const char *p; + const char *r; + size_t count = 0; + + for (p = s; *p != '\0'; ++p) { + for (r = reject; *r != '\0'; ++r) { + if (*p == *r) + return count; + } + ++count; + } + + return count; +} +EXPORT_SYMBOL(strcspn); +#endif + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,10) +/* + * Map a vmalloc()-space virtual address to the physical page frame number. + */ +unsigned long vmalloc_to_pfn(void * vmalloc_addr) +{ + return page_to_pfn(vmalloc_to_page(vmalloc_addr)); +} +EXPORT_SYMBOL(vmalloc_to_pfn); +#endif + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,11) +unsigned long wait_for_completion_timeout(struct completion *x, unsigned long timeout) +{ + might_sleep(); + + spin_lock_irq(&x->wait.lock); + if (!x->done) { + DECLARE_WAITQUEUE(wait, current); + + wait.flags |= WQ_FLAG_EXCLUSIVE; + __add_wait_queue_tail(&x->wait, &wait); + do { + __set_current_state(TASK_UNINTERRUPTIBLE); + spin_unlock_irq(&x->wait.lock); + timeout = schedule_timeout(timeout); + spin_lock_irq(&x->wait.lock); + if (!timeout) { + __remove_wait_queue(&x->wait, &wait); + goto out; + } + } while (!x->done); + __remove_wait_queue(&x->wait, &wait); + } + x->done--; +out: + spin_unlock_irq(&x->wait.lock); + return timeout; +} +EXPORT_SYMBOL(wait_for_completion_timeout); +#endif + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,12) +/* + fake do_exit using complete_and_exit + */ +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,10) +asmlinkage NORET_TYPE void do_exit(long code) +#else +fastcall NORET_TYPE void do_exit(long code) +#endif +{ + complete_and_exit(NULL, code); +} +EXPORT_SYMBOL_GPL(do_exit); +#endif + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14) +signed long schedule_timeout_interruptible(signed long timeout) +{ + __set_current_state(TASK_INTERRUPTIBLE); + return schedule_timeout(timeout); +} +EXPORT_SYMBOL(schedule_timeout_interruptible); +#endif + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14) +/** + * kzalloc - allocate memory. The memory is set to zero. + * @size: how many bytes of memory are required. + * @flags: the type of memory to allocate. + */ +void *kzalloc(size_t size, int flags) +{ + void *ret = kmalloc(size, flags); + if (ret) + memset(ret, 0, size); + return ret; +} +EXPORT_SYMBOL(kzalloc); +#endif + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18) +/* Simplified asprintf. */ +char *kasprintf(gfp_t gfp, const char *fmt, ...) +{ + va_list ap; + unsigned int len; + char *p, dummy[1]; + + va_start(ap, fmt); + len = vsnprintf(dummy, 0, fmt, ap); + va_end(ap); + + p = kmalloc(len + 1, gfp); + if (!p) + return NULL; + va_start(ap, fmt); + vsprintf(p, fmt, ap); + va_end(ap); + return p; +} +EXPORT_SYMBOL(kasprintf); +#endif diff --git a/drivers/xenpvdrivers/platform-pci.c b/drivers/xenpvdrivers/platform-pci.c new file mode 100644 index 0000000..4aa6dfb --- /dev/null +++ b/drivers/xenpvdrivers/platform-pci.c @@ -0,0 +1,490 @@ +/****************************************************************************** + * platform-pci.c + * + * Xen platform PCI device driver + * Copyright (c) 2005, Intel Corporation. + * Copyright (c) 2007, XenSource Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#ifdef __ia64__ +#include +#endif + +#include "platform-pci.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#define DRV_NAME "xen-platform-pci" +#define DRV_VERSION "0.10" +#define DRV_RELDATE "03/03/2005" + +static int max_hypercall_stub_pages, nr_hypercall_stub_pages; +char *hypercall_stubs; +EXPORT_SYMBOL(hypercall_stubs); + +MODULE_AUTHOR("ssmith@xensource.com"); +MODULE_DESCRIPTION("Xen platform PCI device"); +MODULE_LICENSE("GPL"); + +/* NB. [aux-]ide-disks options do not unplug IDE CD-ROM drives. */ +/* NB. aux-ide-disks is equiv to ide-disks except ignores primary master. */ +static char *dev_unplug; +module_param(dev_unplug, charp, 0644); +MODULE_PARM_DESC(dev_unplug, "Emulated devices to unplug: " + "[all,][ide-disks,][aux-ide-disks,][nics]\n"); + +struct pci_dev *xen_platform_pdev; + +static unsigned long shared_info_frame; +static uint64_t callback_via; + +static int __devinit init_xen_info(void) +{ + struct xen_add_to_physmap xatp; + extern void *shared_info_area; + +#ifdef __ia64__ + xencomm_initialize(); +#endif + + setup_xen_features(); + + shared_info_frame = alloc_xen_mmio(PAGE_SIZE) >> PAGE_SHIFT; + xatp.domid = DOMID_SELF; + xatp.idx = 0; + xatp.space = XENMAPSPACE_shared_info; + xatp.gpfn = shared_info_frame; + if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp)) + BUG(); + + shared_info_area = + ioremap(shared_info_frame << PAGE_SHIFT, PAGE_SIZE); + if (shared_info_area == NULL) + panic("can't map shared info\n"); + + return 0; +} + +static unsigned long platform_mmio; +static unsigned long platform_mmio_alloc; +static unsigned long platform_mmiolen; + +unsigned long alloc_xen_mmio(unsigned long len) +{ + unsigned long addr; + + addr = platform_mmio + platform_mmio_alloc; + platform_mmio_alloc += len; + BUG_ON(platform_mmio_alloc > platform_mmiolen); + + return addr; +} + +#ifndef __ia64__ + +static uint32_t xen_cpuid_base(void) +{ + uint32_t base, eax, ebx, ecx, edx; + char signature[13]; + + for (base = 0x40000000; base < 0x40010000; base += 0x100) { + cpuid(base, &eax, &ebx, &ecx, &edx); + *(uint32_t*)(signature + 0) = ebx; + *(uint32_t*)(signature + 4) = ecx; + *(uint32_t*)(signature + 8) = edx; + signature[12] = 0; + + if (!strcmp("XenVMMXenVMM", signature) && ((eax - base) >= 2)) + return base; + } + + return 0; +} + +static int init_hypercall_stubs(void) +{ + uint32_t eax, ebx, ecx, edx, pages, msr, i, base; + + base = xen_cpuid_base(); + if (base == 0) { + printk(KERN_WARNING + "Detected Xen platform device but not Xen VMM?\n"); + return -EINVAL; + } + + cpuid(base + 1, &eax, &ebx, &ecx, &edx); + + printk(KERN_INFO "Xen version %d.%d.\n", eax >> 16, eax & 0xffff); + + /* + * Find largest supported number of hypercall pages. + * We'll create as many as possible up to this number. + */ + cpuid(base + 2, &pages, &msr, &ecx, &edx); + + /* + * Use __vmalloc() because vmalloc_exec() is not an exported symbol. + * PAGE_KERNEL_EXEC also is not exported, hence we use PAGE_KERNEL. + * hypercall_stubs = vmalloc_exec(pages * PAGE_SIZE); + */ + while (pages > 0) { + hypercall_stubs = __vmalloc( + pages * PAGE_SIZE, + GFP_KERNEL | __GFP_HIGHMEM, + __pgprot(__PAGE_KERNEL & ~_PAGE_NX)); + if (hypercall_stubs != NULL) + break; + pages--; /* vmalloc failed: try one fewer pages */ + } + + if (hypercall_stubs == NULL) + return -ENOMEM; + + for (i = 0; i < pages; i++) { + unsigned long pfn; + pfn = vmalloc_to_pfn((char *)hypercall_stubs + i*PAGE_SIZE); + wrmsrl(msr, ((u64)pfn << PAGE_SHIFT) + i); + } + + nr_hypercall_stub_pages = pages; + max_hypercall_stub_pages = pages; + + printk(KERN_INFO "Hypercall area is %u pages.\n", pages); + + return 0; +} + +static void resume_hypercall_stubs(void) +{ + uint32_t base, ecx, edx, pages, msr, i; + + base = xen_cpuid_base(); + BUG_ON(base == 0); + + cpuid(base + 2, &pages, &msr, &ecx, &edx); + + if (pages > max_hypercall_stub_pages) + pages = max_hypercall_stub_pages; + + for (i = 0; i < pages; i++) { + unsigned long pfn; + pfn = vmalloc_to_pfn((char *)hypercall_stubs + i*PAGE_SIZE); + wrmsrl(msr, ((u64)pfn << PAGE_SHIFT) + i); + } + + nr_hypercall_stub_pages = pages; +} + +#else /* __ia64__ */ + +#define init_hypercall_stubs() (0) +#define resume_hypercall_stubs() ((void)0) + +#endif + +static uint64_t get_callback_via(struct pci_dev *pdev) +{ + u8 pin; + int irq; + +#ifdef __ia64__ + for (irq = 0; irq < 16; irq++) { + if (isa_irq_to_vector(irq) == pdev->irq) + return irq; /* ISA IRQ */ + } +#else /* !__ia64__ */ + irq = pdev->irq; + if (irq < 16) + return irq; /* ISA IRQ */ +#endif + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16) + pin = pdev->pin; +#else + pci_read_config_byte(pdev, PCI_INTERRUPT_PIN, &pin); +#endif + + /* We don't know the GSI. Specify the PCI INTx line instead. */ + return (((uint64_t)0x01 << 56) | /* PCI INTx identifier */ + ((uint64_t)pci_domain_nr(pdev->bus) << 32) | + ((uint64_t)pdev->bus->number << 16) | + ((uint64_t)(pdev->devfn & 0xff) << 8) | + ((uint64_t)(pin - 1) & 3)); +} + +static int set_callback_via(uint64_t via) +{ + struct xen_hvm_param a; + + a.domid = DOMID_SELF; + a.index = HVM_PARAM_CALLBACK_IRQ; + a.value = via; + return HYPERVISOR_hvm_op(HVMOP_set_param, &a); +} + +int xen_irq_init(struct pci_dev *pdev); +int xenbus_init(void); +int xen_reboot_init(void); +int xen_panic_handler_init(void); +int gnttab_init(void); + +#define XEN_IOPORT_BASE 0x10 + +#define XEN_IOPORT_PLATFLAGS (XEN_IOPORT_BASE + 0) /* 1 byte access (R/W) */ +#define XEN_IOPORT_MAGIC (XEN_IOPORT_BASE + 0) /* 2 byte access (R) */ +#define XEN_IOPORT_UNPLUG (XEN_IOPORT_BASE + 0) /* 2 byte access (W) */ +#define XEN_IOPORT_DRVVER (XEN_IOPORT_BASE + 0) /* 4 byte access (W) */ + +#define XEN_IOPORT_SYSLOG (XEN_IOPORT_BASE + 2) /* 1 byte access (W) */ +#define XEN_IOPORT_PROTOVER (XEN_IOPORT_BASE + 2) /* 1 byte access (R) */ +#define XEN_IOPORT_PRODNUM (XEN_IOPORT_BASE + 2) /* 2 byte access (W) */ + +#define XEN_IOPORT_MAGIC_VAL 0x49d2 +#define XEN_IOPORT_LINUX_PRODNUM 0xffff /* NB: register a proper one */ +#define XEN_IOPORT_LINUX_DRVVER ((LINUX_VERSION_CODE << 8) + 0x0) + +#define UNPLUG_ALL_IDE_DISKS 1 +#define UNPLUG_ALL_NICS 2 +#define UNPLUG_AUX_IDE_DISKS 4 +#define UNPLUG_ALL 7 + +static int check_platform_magic(struct device *dev, long ioaddr, long iolen) +{ + short magic, unplug = 0; + char protocol, *p, *q, *err; + + for (p = dev_unplug; p; p = q) { + q = strchr(dev_unplug, ','); + if (q) + *q++ = '\0'; + if (!strcmp(p, "all")) + unplug |= UNPLUG_ALL; + else if (!strcmp(p, "ide-disks")) + unplug |= UNPLUG_ALL_IDE_DISKS; + else if (!strcmp(p, "aux-ide-disks")) + unplug |= UNPLUG_AUX_IDE_DISKS; + else if (!strcmp(p, "nics")) + unplug |= UNPLUG_ALL_NICS; + else + dev_warn(dev, "unrecognised option '%s' " + "in module parameter 'dev_unplug'\n", p); + } + + if (iolen < 0x16) { + err = "backend too old"; + goto no_dev; + } + + magic = inw(XEN_IOPORT_MAGIC); + + if (magic != XEN_IOPORT_MAGIC_VAL) { + err = "unrecognised magic value"; + goto no_dev; + } + + protocol = inb(XEN_IOPORT_PROTOVER); + + dev_info(dev, "I/O protocol version %d\n", protocol); + + switch (protocol) { + case 1: + outw(XEN_IOPORT_LINUX_PRODNUM, XEN_IOPORT_PRODNUM); + outl(XEN_IOPORT_LINUX_DRVVER, XEN_IOPORT_DRVVER); + if (inw(XEN_IOPORT_MAGIC) != XEN_IOPORT_MAGIC_VAL) { + dev_err(dev, "blacklisted by host\n"); + return -ENODEV; + } + /* Fall through */ + case 0: + outw(unplug, XEN_IOPORT_UNPLUG); + break; + default: + err = "unknown I/O protocol version"; + goto no_dev; + } + + return 0; + + no_dev: + dev_warn(dev, "failed backend handshake: %s\n", err); + if (!unplug) + return 0; + dev_err(dev, "failed to execute specified dev_unplug options!\n"); + return -ENODEV; +} + +static int __devinit platform_pci_init(struct pci_dev *pdev, + const struct pci_device_id *ent) +{ + int i, ret; + long ioaddr, iolen; + long mmio_addr, mmio_len; + + if (xen_platform_pdev) + return -EBUSY; + xen_platform_pdev = pdev; + + i = pci_enable_device(pdev); + if (i) + return i; + + ioaddr = pci_resource_start(pdev, 0); + iolen = pci_resource_len(pdev, 0); + + mmio_addr = pci_resource_start(pdev, 1); + mmio_len = pci_resource_len(pdev, 1); + + callback_via = get_callback_via(pdev); + + if (mmio_addr == 0 || ioaddr == 0 || callback_via == 0) { + printk(KERN_WARNING DRV_NAME ":no resources found\n"); + return -ENOENT; + } + + if (request_mem_region(mmio_addr, mmio_len, DRV_NAME) == NULL) { + printk(KERN_ERR ":MEM I/O resource 0x%lx @ 0x%lx busy\n", + mmio_addr, mmio_len); + return -EBUSY; + } + + if (request_region(ioaddr, iolen, DRV_NAME) == NULL) { + printk(KERN_ERR DRV_NAME ":I/O resource 0x%lx @ 0x%lx busy\n", + iolen, ioaddr); + release_mem_region(mmio_addr, mmio_len); + return -EBUSY; + } + + platform_mmio = mmio_addr; + platform_mmiolen = mmio_len; + + ret = init_hypercall_stubs(); + if (ret < 0) + goto out; + + ret = check_platform_magic(&pdev->dev, ioaddr, iolen); + if (ret < 0) + goto out; + + if ((ret = init_xen_info())) + goto out; + + if ((ret = gnttab_init())) + goto out; + + if ((ret = xen_irq_init(pdev))) + goto out; + + if ((ret = set_callback_via(callback_via))) + goto out; + + if ((ret = xenbus_init())) + goto out; + + if ((ret = xen_reboot_init())) + goto out; + + if ((ret = xen_panic_handler_init())) + goto out; + + out: + if (ret) { + release_mem_region(mmio_addr, mmio_len); + release_region(ioaddr, iolen); + } + + return ret; +} + +#define XEN_PLATFORM_VENDOR_ID 0x5853 +#define XEN_PLATFORM_DEVICE_ID 0x0001 +static struct pci_device_id platform_pci_tbl[] __devinitdata = { + {XEN_PLATFORM_VENDOR_ID, XEN_PLATFORM_DEVICE_ID, + PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, + /* Continue to recognise the old ID for now */ + {0xfffd, 0x0101, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, + {0,} +}; + +MODULE_DEVICE_TABLE(pci, platform_pci_tbl); + +static struct pci_driver platform_driver = { + name: DRV_NAME, + probe: platform_pci_init, + id_table: platform_pci_tbl, +}; + +static int pci_device_registered; + +void platform_pci_resume(void) +{ + struct xen_add_to_physmap xatp; + + resume_hypercall_stubs(); + + xatp.domid = DOMID_SELF; + xatp.idx = 0; + xatp.space = XENMAPSPACE_shared_info; + xatp.gpfn = shared_info_frame; + if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp)) + BUG(); + + if (set_callback_via(callback_via)) + printk("platform_pci_resume failure!\n"); +} + +static int __init platform_pci_module_init(void) +{ + int rc; + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,10) + rc = pci_module_init(&platform_driver); +#else + rc = pci_register_driver(&platform_driver); +#endif + if (rc) { + printk(KERN_INFO DRV_NAME + ": No platform pci device model found\n"); + return rc; + } + + pci_device_registered = 1; + return 0; +} + +module_init(platform_pci_module_init); diff --git a/drivers/xenpvdrivers/platform-pci.h b/drivers/xenpvdrivers/platform-pci.h new file mode 100644 index 0000000..2537213 --- /dev/null +++ b/drivers/xenpvdrivers/platform-pci.h @@ -0,0 +1,32 @@ +/****************************************************************************** + * platform-pci.h + * + * Xen platform PCI device driver + * Copyright (c) 2004, Intel Corporation. + * Copyright (c) 2007, XenSource Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + */ + +#ifndef _XEN_PLATFORM_PCI_H +#define _XEN_PLATFORM_PCI_H + +#include + +unsigned long alloc_xen_mmio(unsigned long len); +void platform_pci_resume(void); + +extern struct pci_dev *xen_platform_pdev; + +#endif /* _XEN_PLATFORM_PCI_H */ diff --git a/drivers/xenpvdrivers/reboot.c b/drivers/xenpvdrivers/reboot.c new file mode 100644 index 0000000..b1b57d4 --- /dev/null +++ b/drivers/xenpvdrivers/reboot.c @@ -0,0 +1,349 @@ +#define __KERNEL_SYSCALLS__ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#undef handle_sysrq +#endif + +MODULE_LICENSE("Dual BSD/GPL"); + +#define SHUTDOWN_INVALID -1 +#define SHUTDOWN_POWEROFF 0 +#define SHUTDOWN_SUSPEND 2 +#define SHUTDOWN_RESUMING 3 +#define SHUTDOWN_HALT 4 + +/* Ignore multiple shutdown requests. */ +static int shutting_down = SHUTDOWN_INVALID; + +/* Can we leave APs online when we suspend? */ +static int fast_suspend; + +static void __shutdown_handler(struct work_struct *unused); +static DECLARE_DELAYED_WORK(shutdown_work, __shutdown_handler); + +int __xen_suspend(int fast_suspend, void (*resume_notifier)(int)); + +static int shutdown_process(void *__unused) +{ + static char *envp[] = { "HOME=/", "TERM=linux", + "PATH=/sbin:/usr/sbin:/bin:/usr/bin", NULL }; + static char *poweroff_argv[] = { "/sbin/poweroff", NULL }; + + extern asmlinkage long sys_reboot(int magic1, int magic2, + unsigned int cmd, void *arg); + + if ((shutting_down == SHUTDOWN_POWEROFF) || + (shutting_down == SHUTDOWN_HALT)) { + if (call_usermodehelper("/sbin/poweroff", poweroff_argv, + envp, 0) < 0) { +#ifdef CONFIG_XEN + sys_reboot(LINUX_REBOOT_MAGIC1, + LINUX_REBOOT_MAGIC2, + LINUX_REBOOT_CMD_POWER_OFF, + NULL); +#endif /* CONFIG_XEN */ + } + } + + shutting_down = SHUTDOWN_INVALID; /* could try again */ + + return 0; +} + +#ifdef CONFIG_PM_SLEEP + +static int setup_suspend_evtchn(void); + +/* Was last suspend request cancelled? */ +static int suspend_cancelled; + +static void xen_resume_notifier(int _suspend_cancelled) +{ + int old_state = xchg(&shutting_down, SHUTDOWN_RESUMING); + BUG_ON(old_state != SHUTDOWN_SUSPEND); + suspend_cancelled = _suspend_cancelled; +} + +static int xen_suspend(void *__unused) +{ + int err, old_state; + + daemonize("suspend"); + err = set_cpus_allowed_ptr(current, cpumask_of(0)); + if (err) { + printk(KERN_ERR "Xen suspend can't run on CPU0 (%d)\n", err); + goto fail; + } + + do { + err = __xen_suspend(fast_suspend, xen_resume_notifier); + if (err) { + printk(KERN_ERR "Xen suspend failed (%d)\n", err); + goto fail; + } + if (!suspend_cancelled) + setup_suspend_evtchn(); + old_state = cmpxchg( + &shutting_down, SHUTDOWN_RESUMING, SHUTDOWN_INVALID); + } while (old_state == SHUTDOWN_SUSPEND); + + switch (old_state) { + case SHUTDOWN_INVALID: + case SHUTDOWN_SUSPEND: + BUG(); + case SHUTDOWN_RESUMING: + break; + default: + schedule_delayed_work(&shutdown_work, 0); + break; + } + + return 0; + + fail: + old_state = xchg(&shutting_down, SHUTDOWN_INVALID); + BUG_ON(old_state != SHUTDOWN_SUSPEND); + return 0; +} + +#else +# define xen_suspend NULL +#endif + +static void switch_shutdown_state(int new_state) +{ + int prev_state, old_state = SHUTDOWN_INVALID; + + /* We only drive shutdown_state into an active state. */ + if (new_state == SHUTDOWN_INVALID) + return; + + do { + /* We drop this transition if already in an active state. */ + if ((old_state != SHUTDOWN_INVALID) && + (old_state != SHUTDOWN_RESUMING)) + return; + /* Attempt to transition. */ + prev_state = old_state; + old_state = cmpxchg(&shutting_down, old_state, new_state); + } while (old_state != prev_state); + + /* Either we kick off the work, or we leave it to xen_suspend(). */ + if (old_state == SHUTDOWN_INVALID) + schedule_delayed_work(&shutdown_work, 0); + else + BUG_ON(old_state != SHUTDOWN_RESUMING); +} + +static void __shutdown_handler(struct work_struct *unused) +{ + int err; + + err = kernel_thread((shutting_down == SHUTDOWN_SUSPEND) ? + xen_suspend : shutdown_process, + NULL, CLONE_FS | CLONE_FILES); + + if (err < 0) { + printk(KERN_WARNING "Error creating shutdown process (%d): " + "retrying...\n", -err); + schedule_delayed_work(&shutdown_work, HZ/2); + } +} + +static void shutdown_handler(struct xenbus_watch *watch, + const char **vec, unsigned int len) +{ + extern void xenpvdrivers_ctrl_alt_del(void); + char *str; + struct xenbus_transaction xbt; + int err, new_state = SHUTDOWN_INVALID; + + if ((shutting_down != SHUTDOWN_INVALID) && + (shutting_down != SHUTDOWN_RESUMING)) + return; + + again: + err = xenbus_transaction_start(&xbt); + if (err) + return; + + str = (char *)xenbus_read(xbt, "control", "shutdown", NULL); + /* Ignore read errors and empty reads. */ + if (XENBUS_IS_ERR_READ(str)) { + xenbus_transaction_end(xbt, 1); + return; + } + + xenbus_write(xbt, "control", "shutdown", ""); + + err = xenbus_transaction_end(xbt, 0); + if (err == -EAGAIN) { + kfree(str); + goto again; + } + + if (strcmp(str, "poweroff") == 0) + new_state = SHUTDOWN_POWEROFF; + else if (strcmp(str, "reboot") == 0) + xenpvdrivers_ctrl_alt_del(); +#ifdef CONFIG_PM_SLEEP + else if (strcmp(str, "suspend") == 0) + new_state = SHUTDOWN_SUSPEND; +#endif + else if (strcmp(str, "halt") == 0) + new_state = SHUTDOWN_HALT; + else + printk("Ignoring shutdown request: %s\n", str); + + switch_shutdown_state(new_state); + + kfree(str); +} + +static void sysrq_handler(struct xenbus_watch *watch, const char **vec, + unsigned int len) +{ + char sysrq_key = '\0'; + struct xenbus_transaction xbt; + int err; + + again: + err = xenbus_transaction_start(&xbt); + if (err) + return; + if (!xenbus_scanf(xbt, "control", "sysrq", "%c", &sysrq_key)) { + printk(KERN_ERR "Unable to read sysrq code in " + "control/sysrq\n"); + xenbus_transaction_end(xbt, 1); + return; + } + + if (sysrq_key != '\0') + xenbus_printf(xbt, "control", "sysrq", "%c", '\0'); + + err = xenbus_transaction_end(xbt, 0); + if (err == -EAGAIN) + goto again; + +#ifdef CONFIG_MAGIC_SYSRQ + if (sysrq_key != '\0') + handle_sysrq(sysrq_key, NULL); +#endif +} + +static struct xenbus_watch shutdown_watch = { + .node = "control/shutdown", + .callback = shutdown_handler +}; + +static struct xenbus_watch sysrq_watch = { + .node = "control/sysrq", + .callback = sysrq_handler +}; + +#ifdef CONFIG_PM_SLEEP +static irqreturn_t suspend_int(int irq, void* dev_id) +{ + switch_shutdown_state(SHUTDOWN_SUSPEND); + return IRQ_HANDLED; +} + +static int setup_suspend_evtchn(void) +{ + static int irq; + int port; + char portstr[16]; + + if (irq > 0) + unbind_from_irqhandler(irq, NULL); + + irq = bind_listening_port_to_irqhandler(0, suspend_int, 0, "suspend", + NULL); + if (irq <= 0) + return -1; + + port = irq_to_evtchn_port(irq); + printk(KERN_INFO "suspend: event channel %d\n", port); + sprintf(portstr, "%d", port); + xenbus_write(XBT_NIL, "device/suspend", "event-channel", portstr); + + return 0; +} +#else +#define setup_suspend_evtchn() 0 +#endif + +static int setup_shutdown_watcher(void) +{ + int err; + + xenbus_scanf(XBT_NIL, "control", + "platform-feature-multiprocessor-suspend", + "%d", &fast_suspend); + + err = register_xenbus_watch(&shutdown_watch); + if (err) { + printk(KERN_ERR "Failed to set shutdown watcher\n"); + return err; + } + + err = register_xenbus_watch(&sysrq_watch); + if (err) { + printk(KERN_ERR "Failed to set sysrq watcher\n"); + return err; + } + + /* suspend event channel */ + err = setup_suspend_evtchn(); + if (err) { + printk(KERN_ERR "Failed to register suspend event channel\n"); + return err; + } + + return 0; +} + +#ifdef CONFIG_XEN + +static int shutdown_event(struct notifier_block *notifier, + unsigned long event, + void *data) +{ + setup_shutdown_watcher(); + return NOTIFY_DONE; +} + +static int __init setup_shutdown_event(void) +{ + static struct notifier_block xenstore_notifier = { + .notifier_call = shutdown_event + }; + register_xenstore_notifier(&xenstore_notifier); + + return 0; +} + +subsys_initcall(setup_shutdown_event); + +#else /* !defined(CONFIG_XEN) */ + +int xen_reboot_init(void) +{ + return setup_shutdown_watcher(); +} + +#endif /* !defined(CONFIG_XEN) */ diff --git a/drivers/xenpvdrivers/scrub.c b/drivers/xenpvdrivers/scrub.c new file mode 100644 index 0000000..f333ae8 --- /dev/null +++ b/drivers/xenpvdrivers/scrub.c @@ -0,0 +1,21 @@ +#include +#include +#include + +void scrub_pages(void *v, unsigned int count) +{ + if (likely(cpu_has_xmm2)) { + unsigned long n = count * (PAGE_SIZE / sizeof(long) / 4); + + for (; n--; v += sizeof(long) * 4) + asm("movnti %1,(%0)\n\t" + "movnti %1,%c2(%0)\n\t" + "movnti %1,2*%c2(%0)\n\t" + "movnti %1,3*%c2(%0)\n\t" + : : "r" (v), "r" (0L), "i" (sizeof(long)) + : "memory"); + asm volatile("sfence" : : : "memory"); + } else + for (; count--; v += PAGE_SIZE) + clear_page(v); +} diff --git a/drivers/xenpvdrivers/scsifront-common.h b/drivers/xenpvdrivers/scsifront-common.h new file mode 100644 index 0000000..8b29f9c --- /dev/null +++ b/drivers/xenpvdrivers/scsifront-common.h @@ -0,0 +1,135 @@ +/* + * Xen SCSI frontend driver + * + * Copyright (c) 2008, FUJITSU Limited + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __XEN_DRIVERS_SCSIFRONT_H__ +#define __XEN_DRIVERS_SCSIFRONT_H__ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#define GRANT_INVALID_REF 0 +#define VSCSI_IN_ABORT 1 +#define VSCSI_IN_RESET 2 + +/* tuning point*/ +#define VSCSIIF_DEFAULT_CMD_PER_LUN 10 +#define VSCSIIF_MAX_TARGET 64 +#define VSCSIIF_MAX_LUN 255 + +#define VSCSIIF_RING_SIZE __CONST_RING_SIZE(vscsiif, PAGE_SIZE) +#define VSCSIIF_MAX_REQS VSCSIIF_RING_SIZE + +struct vscsifrnt_shadow { + uint16_t next_free; + + /* command between backend and frontend + * VSCSIIF_ACT_SCSI_CDB or VSCSIIF_ACT_SCSI_RESET */ + unsigned char act; + + /* do reset function */ + wait_queue_head_t wq_reset; /* reset work queue */ + int wait_reset; /* reset work queue condition */ + int32_t rslt_reset; /* reset response status */ + /* (SUCESS or FAILED) */ + + /* for DMA_TO_DEVICE(1), DMA_FROM_DEVICE(2), DMA_NONE(3) + requests */ + unsigned int sc_data_direction; + + /* Number of pieces of scatter-gather */ + unsigned int nr_segments; + + /* requested struct scsi_cmnd is stored from kernel */ + unsigned long req_scsi_cmnd; + int gref[VSCSIIF_SG_TABLESIZE]; +}; + +struct vscsifrnt_info { + struct xenbus_device *dev; + + struct Scsi_Host *host; + + spinlock_t io_lock; + spinlock_t shadow_lock; + unsigned int evtchn; + unsigned int irq; + + grant_ref_t ring_ref; + struct vscsiif_front_ring ring; + struct vscsiif_response ring_res; + + struct vscsifrnt_shadow shadow[VSCSIIF_MAX_REQS]; + uint32_t shadow_free; + + struct task_struct *kthread; + wait_queue_head_t wq; + unsigned int waiting_resp; + +}; + +#define DPRINTK(_f, _a...) \ + pr_debug("(file=%s, line=%d) " _f, \ + __FILE__ , __LINE__ , ## _a ) + +int scsifront_xenbus_init(void); +void scsifront_xenbus_unregister(void); +int scsifront_schedule(void *data); +irqreturn_t scsifront_intr(int irq, void *dev_id); +int scsifront_cmd_done(struct vscsifrnt_info *info); + + +#endif /* __XEN_DRIVERS_SCSIFRONT_H__ */ diff --git a/drivers/xenpvdrivers/scsifront.c b/drivers/xenpvdrivers/scsifront.c new file mode 100644 index 0000000..2b6a78d --- /dev/null +++ b/drivers/xenpvdrivers/scsifront.c @@ -0,0 +1,478 @@ +/* + * Xen SCSI frontend driver + * + * Copyright (c) 2008, FUJITSU Limited + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + + +#include +#include "scsifront-common.h" + +static int get_id_from_freelist(struct vscsifrnt_info *info) +{ + unsigned long flags; + uint32_t free; + + spin_lock_irqsave(&info->shadow_lock, flags); + + free = info->shadow_free; + BUG_ON(free > VSCSIIF_MAX_REQS); + info->shadow_free = info->shadow[free].next_free; + info->shadow[free].next_free = 0x0fff; + + info->shadow[free].wait_reset = 0; + + spin_unlock_irqrestore(&info->shadow_lock, flags); + + return free; +} + +static void add_id_to_freelist(struct vscsifrnt_info *info, uint32_t id) +{ + unsigned long flags; + + spin_lock_irqsave(&info->shadow_lock, flags); + + info->shadow[id].next_free = info->shadow_free; + info->shadow[id].req_scsi_cmnd = 0; + info->shadow_free = id; + + spin_unlock_irqrestore(&info->shadow_lock, flags); +} + + +struct vscsiif_request * scsifront_pre_request(struct vscsifrnt_info *info) +{ + struct vscsiif_front_ring *ring = &(info->ring); + vscsiif_request_t *ring_req; + uint32_t id; + + ring_req = RING_GET_REQUEST(&(info->ring), ring->req_prod_pvt); + + ring->req_prod_pvt++; + + id = get_id_from_freelist(info); /* use id by response */ + ring_req->rqid = (uint16_t)id; + + return ring_req; +} + + +static void scsifront_notify_work(struct vscsifrnt_info *info) +{ + info->waiting_resp = 1; + wake_up(&info->wq); +} + + +static void scsifront_do_request(struct vscsifrnt_info *info) +{ + struct vscsiif_front_ring *ring = &(info->ring); + unsigned int irq = info->irq; + int notify; + + RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(ring, notify); + if (notify) + notify_remote_via_irq(irq); +} + +irqreturn_t scsifront_intr(int irq, void *dev_id) +{ + scsifront_notify_work((struct vscsifrnt_info *)dev_id); + return IRQ_HANDLED; +} + + +static void scsifront_gnttab_done(struct vscsifrnt_shadow *s, uint32_t id) +{ + int i; + + if (s->sc_data_direction == DMA_NONE) + return; + + if (s->nr_segments) { + for (i = 0; i < s->nr_segments; i++) { + if (unlikely(gnttab_query_foreign_access( + s->gref[i]) != 0)) { + printk(KERN_ALERT "scsifront: " + "grant still in use by backend.\n"); + BUG(); + } + gnttab_end_foreign_access(s->gref[i], 0UL); + } + } + + return; +} + + +static void scsifront_cdb_cmd_done(struct vscsifrnt_info *info, + vscsiif_response_t *ring_res) +{ + struct scsi_cmnd *sc; + uint32_t id; + uint8_t sense_len; + + id = ring_res->rqid; + sc = (struct scsi_cmnd *)info->shadow[id].req_scsi_cmnd; + + if (sc == NULL) + BUG(); + + scsifront_gnttab_done(&info->shadow[id], id); + add_id_to_freelist(info, id); + + sc->result = ring_res->rslt; + scsi_set_resid(sc, ring_res->residual_len); + + if (ring_res->sense_len > VSCSIIF_SENSE_BUFFERSIZE) + sense_len = VSCSIIF_SENSE_BUFFERSIZE; + else + sense_len = ring_res->sense_len; + + if (sense_len) + memcpy(sc->sense_buffer, ring_res->sense_buffer, sense_len); + + sc->scsi_done(sc); + + return; +} + + +static void scsifront_sync_cmd_done(struct vscsifrnt_info *info, + vscsiif_response_t *ring_res) +{ + uint16_t id = ring_res->rqid; + unsigned long flags; + + spin_lock_irqsave(&info->shadow_lock, flags); + info->shadow[id].wait_reset = 1; + info->shadow[id].rslt_reset = ring_res->rslt; + spin_unlock_irqrestore(&info->shadow_lock, flags); + + wake_up(&(info->shadow[id].wq_reset)); +} + + +int scsifront_cmd_done(struct vscsifrnt_info *info) +{ + vscsiif_response_t *ring_res; + + RING_IDX i, rp; + int more_to_do = 0; + unsigned long flags; + + spin_lock_irqsave(&info->io_lock, flags); + + rp = info->ring.sring->rsp_prod; + rmb(); + for (i = info->ring.rsp_cons; i != rp; i++) { + + ring_res = RING_GET_RESPONSE(&info->ring, i); + + if (info->shadow[ring_res->rqid].act == VSCSIIF_ACT_SCSI_CDB) + scsifront_cdb_cmd_done(info, ring_res); + else + scsifront_sync_cmd_done(info, ring_res); + } + + info->ring.rsp_cons = i; + + if (i != info->ring.req_prod_pvt) { + RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do); + } else { + info->ring.sring->rsp_event = i + 1; + } + + spin_unlock_irqrestore(&info->io_lock, flags); + + + /* Yield point for this unbounded loop. */ + cond_resched(); + + return more_to_do; +} + + + + +int scsifront_schedule(void *data) +{ + struct vscsifrnt_info *info = (struct vscsifrnt_info *)data; + + while (!kthread_should_stop()) { + wait_event_interruptible( + info->wq, + info->waiting_resp || kthread_should_stop()); + + info->waiting_resp = 0; + smp_mb(); + + if (scsifront_cmd_done(info)) + info->waiting_resp = 1; + } + + return 0; +} + + + +static int map_data_for_request(struct vscsifrnt_info *info, + struct scsi_cmnd *sc, vscsiif_request_t *ring_req, uint32_t id) +{ + grant_ref_t gref_head; + struct page *page; + int err, ref, ref_cnt = 0; + int write = (sc->sc_data_direction == DMA_TO_DEVICE); + unsigned int i, nr_pages, off, len, bytes; + unsigned long buffer_pfn; + + if (sc->sc_data_direction == DMA_NONE) + return 0; + + err = gnttab_alloc_grant_references(VSCSIIF_SG_TABLESIZE, &gref_head); + if (err) { + printk(KERN_ERR "scsifront: gnttab_alloc_grant_references() error\n"); + return -ENOMEM; + } + + if (scsi_bufflen(sc)) { + /* quoted scsi_lib.c/scsi_req_map_sg . */ + struct scatterlist *sg, *sgl = scsi_sglist(sc); + unsigned int data_len = scsi_bufflen(sc); + + nr_pages = (data_len + sgl->offset + PAGE_SIZE - 1) >> PAGE_SHIFT; + if (nr_pages > VSCSIIF_SG_TABLESIZE) { + printk(KERN_ERR "scsifront: Unable to map request_buffer for command!\n"); + ref_cnt = (-E2BIG); + goto big_to_sg; + } + + for_each_sg (sgl, sg, scsi_sg_count(sc), i) { + page = sg_page(sg); + off = sg->offset; + len = sg->length; + + buffer_pfn = page_to_phys(page) >> PAGE_SHIFT; + + while (len > 0 && data_len > 0) { + /* + * sg sends a scatterlist that is larger than + * the data_len it wants transferred for certain + * IO sizes + */ + bytes = min_t(unsigned int, len, PAGE_SIZE - off); + bytes = min(bytes, data_len); + + ref = gnttab_claim_grant_reference(&gref_head); + BUG_ON(ref == -ENOSPC); + + gnttab_grant_foreign_access_ref(ref, info->dev->otherend_id, + buffer_pfn, write); + + info->shadow[id].gref[ref_cnt] = ref; + ring_req->seg[ref_cnt].gref = ref; + ring_req->seg[ref_cnt].offset = (uint16_t)off; + ring_req->seg[ref_cnt].length = (uint16_t)bytes; + + buffer_pfn++; + len -= bytes; + data_len -= bytes; + off = 0; + ref_cnt++; + } + } + } + +big_to_sg: + + gnttab_free_grant_references(gref_head); + + return ref_cnt; +} + +static int scsifront_queuecommand(struct scsi_cmnd *sc, + void (*done)(struct scsi_cmnd *)) +{ + struct vscsifrnt_info *info = + (struct vscsifrnt_info *) sc->device->host->hostdata; + vscsiif_request_t *ring_req; + int ref_cnt; + uint16_t rqid; + + if (RING_FULL(&info->ring)) { + goto out_host_busy; + } + + sc->scsi_done = done; + sc->result = 0; + + ring_req = scsifront_pre_request(info); + rqid = ring_req->rqid; + ring_req->act = VSCSIIF_ACT_SCSI_CDB; + + ring_req->id = sc->device->id; + ring_req->lun = sc->device->lun; + ring_req->channel = sc->device->channel; + ring_req->cmd_len = sc->cmd_len; + + BUG_ON(sc->cmd_len > VSCSIIF_MAX_COMMAND_SIZE); + + if ( sc->cmd_len ) + memcpy(ring_req->cmnd, sc->cmnd, sc->cmd_len); + else + memset(ring_req->cmnd, 0, VSCSIIF_MAX_COMMAND_SIZE); + + ring_req->sc_data_direction = (uint8_t)sc->sc_data_direction; + ring_req->timeout_per_command = (sc->request->timeout / HZ); + + info->shadow[rqid].req_scsi_cmnd = (unsigned long)sc; + info->shadow[rqid].sc_data_direction = sc->sc_data_direction; + info->shadow[rqid].act = ring_req->act; + + ref_cnt = map_data_for_request(info, sc, ring_req, rqid); + if (ref_cnt < 0) { + add_id_to_freelist(info, rqid); + if (ref_cnt == (-ENOMEM)) + goto out_host_busy; + else { + sc->result = (DID_ERROR << 16); + goto out_fail_command; + } + } + + ring_req->nr_segments = (uint8_t)ref_cnt; + info->shadow[rqid].nr_segments = ref_cnt; + + scsifront_do_request(info); + + return 0; + +out_host_busy: + return SCSI_MLQUEUE_HOST_BUSY; + +out_fail_command: + done(sc); + return 0; +} + + +static int scsifront_eh_abort_handler(struct scsi_cmnd *sc) +{ + return (FAILED); +} + +/* vscsi supports only device_reset, because it is each of LUNs */ +static int scsifront_dev_reset_handler(struct scsi_cmnd *sc) +{ + struct Scsi_Host *host = sc->device->host; + struct vscsifrnt_info *info = + (struct vscsifrnt_info *) sc->device->host->hostdata; + + vscsiif_request_t *ring_req; + uint16_t rqid; + int err; + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,12) + spin_lock_irq(host->host_lock); +#endif + + ring_req = scsifront_pre_request(info); + ring_req->act = VSCSIIF_ACT_SCSI_RESET; + + rqid = ring_req->rqid; + info->shadow[rqid].act = VSCSIIF_ACT_SCSI_RESET; + + ring_req->channel = sc->device->channel; + ring_req->id = sc->device->id; + ring_req->lun = sc->device->lun; + ring_req->cmd_len = sc->cmd_len; + + if ( sc->cmd_len ) + memcpy(ring_req->cmnd, sc->cmnd, sc->cmd_len); + else + memset(ring_req->cmnd, 0, VSCSIIF_MAX_COMMAND_SIZE); + + ring_req->sc_data_direction = (uint8_t)sc->sc_data_direction; + ring_req->timeout_per_command = (sc->request->timeout / HZ); + ring_req->nr_segments = 0; + + scsifront_do_request(info); + + spin_unlock_irq(host->host_lock); + wait_event_interruptible(info->shadow[rqid].wq_reset, + info->shadow[rqid].wait_reset); + spin_lock_irq(host->host_lock); + + err = info->shadow[rqid].rslt_reset; + + add_id_to_freelist(info, rqid); + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,12) + spin_unlock_irq(host->host_lock); +#endif + return (err); +} + + +struct scsi_host_template scsifront_sht = { + .module = THIS_MODULE, + .name = "Xen SCSI frontend driver", + .queuecommand = scsifront_queuecommand, + .eh_abort_handler = scsifront_eh_abort_handler, + .eh_device_reset_handler= scsifront_dev_reset_handler, + .cmd_per_lun = VSCSIIF_DEFAULT_CMD_PER_LUN, + .can_queue = VSCSIIF_MAX_REQS, + .this_id = -1, + .sg_tablesize = VSCSIIF_SG_TABLESIZE, + .use_clustering = DISABLE_CLUSTERING, + .proc_name = "scsifront", +}; + + +static int __init scsifront_init(void) +{ + int err; + + if (!is_running_on_xen()) + return -ENODEV; + + err = scsifront_xenbus_init(); + + return err; +} + +static void __exit scsifront_exit(void) +{ + scsifront_xenbus_unregister(); +} + +module_init(scsifront_init); +module_exit(scsifront_exit); + +MODULE_DESCRIPTION("Xen SCSI frontend driver"); +MODULE_LICENSE("GPL"); diff --git a/drivers/xenpvdrivers/sysfs.c b/drivers/xenpvdrivers/sysfs.c new file mode 100644 index 0000000..2292329 --- /dev/null +++ b/drivers/xenpvdrivers/sysfs.c @@ -0,0 +1,204 @@ +/****************************************************************************** + * balloon/sysfs.c + * + * Xen balloon driver - sysfs interfaces. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include "balloon-common.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#define BALLOON_CLASS_NAME "xen_memory" + +#define BALLOON_SHOW(name, format, args...) \ + static ssize_t show_##name(struct sys_device *dev, \ + struct sysdev_attribute *attr, \ + char *buf) \ + { \ + return sprintf(buf, format, ##args); \ + } \ + static SYSDEV_ATTR(name, S_IRUGO, show_##name, NULL) + +BALLOON_SHOW(current_kb, "%lu\n", PAGES2KB(bs.current_pages)); +BALLOON_SHOW(min_kb, "%lu\n", PAGES2KB(balloon_minimum_target())); +BALLOON_SHOW(max_kb, "%lu\n", PAGES2KB(num_physpages)); +BALLOON_SHOW(low_kb, "%lu\n", PAGES2KB(bs.balloon_low)); +BALLOON_SHOW(high_kb, "%lu\n", PAGES2KB(bs.balloon_high)); +BALLOON_SHOW(driver_kb, "%lu\n", PAGES2KB(bs.driver_pages)); + +static ssize_t show_target_kb(struct sys_device *dev, + struct sysdev_attribute *attr, char *buf) +{ + return sprintf(buf, "%lu\n", PAGES2KB(bs.target_pages)); +} + +static ssize_t store_target_kb(struct sys_device *dev, + struct sysdev_attribute *attr, + const char *buf, size_t count) +{ + char *endchar; + unsigned long long target_bytes; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (count <= 1) + return -EBADMSG; /* runt */ + + target_bytes = simple_strtoull(buf, &endchar, 0) << 10; + balloon_set_new_target(target_bytes >> PAGE_SHIFT); + + return count; +} + +static SYSDEV_ATTR(target_kb, S_IRUGO | S_IWUSR, + show_target_kb, store_target_kb); + +static ssize_t show_target(struct sys_device *dev, + struct sysdev_attribute *attr, char *buf) +{ + return sprintf(buf, "%llu\n", + (unsigned long long)balloon_stats.target_pages + << PAGE_SHIFT); +} + +static ssize_t store_target(struct sys_device *dev, + struct sysdev_attribute *attr, + const char *buf, + size_t count) +{ + char *endchar; + unsigned long long target_bytes; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (count <= 1) + return -EBADMSG; /* runt */ + + target_bytes = memparse(buf, &endchar); + balloon_set_new_target(target_bytes >> PAGE_SHIFT); + + return count; +} + +static SYSDEV_ATTR(target, S_IRUGO | S_IWUSR, + show_target, store_target); + +static struct sysdev_attribute *balloon_attrs[] = { + &attr_target_kb, + &attr_target, +}; + +static struct attribute *balloon_info_attrs[] = { + &attr_current_kb.attr, + &attr_min_kb.attr, + &attr_max_kb.attr, + &attr_low_kb.attr, + &attr_high_kb.attr, + &attr_driver_kb.attr, + NULL +}; + +static struct attribute_group balloon_info_group = { + .name = "info", + .attrs = balloon_info_attrs, +}; + +static struct sysdev_class balloon_sysdev_class = { + .name = BALLOON_CLASS_NAME, +}; + +static struct sys_device balloon_sysdev; + +static int __init register_balloon(struct sys_device *sysdev) +{ + int i, error; + + error = sysdev_class_register(&balloon_sysdev_class); + if (error) + return error; + + sysdev->id = 0; + sysdev->cls = &balloon_sysdev_class; + + error = sysdev_register(sysdev); + if (error) { + sysdev_class_unregister(&balloon_sysdev_class); + return error; + } + + for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++) { + error = sysdev_create_file(sysdev, balloon_attrs[i]); + if (error) + goto fail; + } + + error = sysfs_create_group(&sysdev->kobj, &balloon_info_group); + if (error) + goto fail; + + return 0; + + fail: + while (--i >= 0) + sysdev_remove_file(sysdev, balloon_attrs[i]); + sysdev_unregister(sysdev); + sysdev_class_unregister(&balloon_sysdev_class); + return error; +} + +static __exit void unregister_balloon(struct sys_device *sysdev) +{ + int i; + + sysfs_remove_group(&sysdev->kobj, &balloon_info_group); + for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++) + sysdev_remove_file(sysdev, balloon_attrs[i]); + sysdev_unregister(sysdev); + sysdev_class_unregister(&balloon_sysdev_class); +} + +int __init balloon_sysfs_init(void) +{ + return register_balloon(&balloon_sysdev); +} + +void __exit balloon_sysfs_exit(void) +{ + unregister_balloon(&balloon_sysdev); +} diff --git a/drivers/xenpvdrivers/vbd.c b/drivers/xenpvdrivers/vbd.c new file mode 100644 index 0000000..7e522a1 --- /dev/null +++ b/drivers/xenpvdrivers/vbd.c @@ -0,0 +1,542 @@ +/****************************************************************************** + * vbd.c + * + * XenLinux virtual block-device driver (xvd). + * + * Copyright (c) 2003-2004, Keir Fraser & Steve Hand + * Modifications by Mark A. Williamson are (c) Intel Research Cambridge + * Copyright (c) 2004-2005, Christian Limpach + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include "block.h" +#include +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#define BLKIF_MAJOR(dev) ((dev)>>8) +#define BLKIF_MINOR(dev) ((dev) & 0xff) + +#define EXT_SHIFT 28 +#define EXTENDED (1<= KERNEL_VERSION(2,6,16) + .getgeo = blkif_getgeo +#endif +}; + +DEFINE_SPINLOCK(blkif_io_lock); + +static struct xlbd_major_info * +xlbd_alloc_major_info(int major, int minor, int index) +{ + struct xlbd_major_info *ptr; + struct xlbd_minor_state *minors; + int do_register; + + ptr = kzalloc(sizeof(struct xlbd_major_info), GFP_KERNEL); + if (ptr == NULL) + return NULL; + + ptr->major = major; + minors = kmalloc(sizeof(*minors), GFP_KERNEL); + if (minors == NULL) { + kfree(ptr); + return NULL; + } + + minors->bitmap = kzalloc(BITS_TO_LONGS(256) * sizeof(*minors->bitmap), + GFP_KERNEL); + if (minors->bitmap == NULL) { + kfree(minors); + kfree(ptr); + return NULL; + } + + spin_lock_init(&minors->lock); + minors->nr = 256; + do_register = 1; + + switch (index) { + case XLBD_MAJOR_IDE_RANGE: + ptr->type = &xlbd_ide_type; + ptr->index = index - XLBD_MAJOR_IDE_START; + break; + case XLBD_MAJOR_SCSI_RANGE: + ptr->type = &xlbd_scsi_type; + ptr->index = index - XLBD_MAJOR_SCSI_START; + break; + case XLBD_MAJOR_VBD_RANGE: + ptr->index = 0; + if ((index - XLBD_MAJOR_VBD_START) == 0) + ptr->type = &xlbd_vbd_type; + else + ptr->type = &xlbd_vbd_type_ext; + + /* + * if someone already registered block major 202, + * don't try to register it again + */ + if (major_info[XLBD_MAJOR_VBD_ALT(index)] != NULL) { + kfree(minors->bitmap); + kfree(minors); + minors = major_info[XLBD_MAJOR_VBD_ALT(index)]->minors; + do_register = 0; + } + break; + } + + if (do_register) { + if (register_blkdev(ptr->major, ptr->type->devname)) { + kfree(minors->bitmap); + kfree(minors); + kfree(ptr); + return NULL; + } + + printk("xen-vbd: registered block device major %i\n", ptr->major); + } + + ptr->minors = minors; + major_info[index] = ptr; + return ptr; +} + +static struct xlbd_major_info * +xlbd_get_major_info(int major, int minor, int vdevice) +{ + struct xlbd_major_info *mi; + int index; + + switch (major) { + case IDE0_MAJOR: index = 0; break; + case IDE1_MAJOR: index = 1; break; + case IDE2_MAJOR: index = 2; break; + case IDE3_MAJOR: index = 3; break; + case IDE4_MAJOR: index = 4; break; + case IDE5_MAJOR: index = 5; break; + case IDE6_MAJOR: index = 6; break; + case IDE7_MAJOR: index = 7; break; + case IDE8_MAJOR: index = 8; break; + case IDE9_MAJOR: index = 9; break; + case SCSI_DISK0_MAJOR: index = 10; break; + case SCSI_DISK1_MAJOR ... SCSI_DISK7_MAJOR: + index = 11 + major - SCSI_DISK1_MAJOR; + break; + case SCSI_DISK8_MAJOR ... SCSI_DISK15_MAJOR: + index = 18 + major - SCSI_DISK8_MAJOR; + break; + case SCSI_CDROM_MAJOR: index = 26; break; + default: + if (!VDEV_IS_EXTENDED(vdevice)) + index = 27; + else + index = 28; + break; + } + + mi = ((major_info[index] != NULL) ? major_info[index] : + xlbd_alloc_major_info(major, minor, index)); + if (mi) + mi->usage++; + return mi; +} + +static void +xlbd_put_major_info(struct xlbd_major_info *mi) +{ + mi->usage--; + /* XXX: release major if 0 */ +} + +static int +xlbd_reserve_minors(struct xlbd_major_info *mi, unsigned int minor, + unsigned int nr_minors) +{ + struct xlbd_minor_state *ms = mi->minors; + unsigned int end = minor + nr_minors; + int rc; + + if (end > ms->nr) { + unsigned long *bitmap, *old; + + bitmap = kzalloc(BITS_TO_LONGS(end) * sizeof(*bitmap), + GFP_KERNEL); + if (bitmap == NULL) + return -ENOMEM; + + spin_lock(&ms->lock); + if (end > ms->nr) { + old = ms->bitmap; + memcpy(bitmap, ms->bitmap, + BITS_TO_LONGS(ms->nr) * sizeof(*bitmap)); + ms->bitmap = bitmap; + ms->nr = BITS_TO_LONGS(end) * BITS_PER_LONG; + } else + old = bitmap; + spin_unlock(&ms->lock); + kfree(old); + } + + spin_lock(&ms->lock); + if (find_next_bit(ms->bitmap, end, minor) >= end) { + for (; minor < end; ++minor) + __set_bit(minor, ms->bitmap); + rc = 0; + } else + rc = -EBUSY; + spin_unlock(&ms->lock); + + return rc; +} + +static void +xlbd_release_minors(struct xlbd_major_info *mi, unsigned int minor, + unsigned int nr_minors) +{ + struct xlbd_minor_state *ms = mi->minors; + unsigned int end = minor + nr_minors; + + BUG_ON(end > ms->nr); + spin_lock(&ms->lock); + for (; minor < end; ++minor) + __clear_bit(minor, ms->bitmap); + spin_unlock(&ms->lock); +} + +static int +xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size) +{ + struct request_queue *rq; + + rq = blk_init_queue(do_blkif_request, &blkif_io_lock); + if (rq == NULL) + return -1; + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,29) + queue_flag_set_unlocked(QUEUE_FLAG_VIRT, rq); +#endif + + /* Hard sector size and max sectors impersonate the equiv. hardware. */ + blk_queue_logical_block_size(rq, sector_size); + blk_queue_max_sectors(rq, 512); + + /* Each segment in a request is up to an aligned page in size. */ + blk_queue_segment_boundary(rq, PAGE_SIZE - 1); + blk_queue_max_segment_size(rq, PAGE_SIZE); + + /* Ensure a merged request will fit in a single I/O ring slot. */ + blk_queue_max_phys_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST); + blk_queue_max_hw_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST); + + /* Make sure buffer addresses are sector-aligned. */ + blk_queue_dma_alignment(rq, 511); + + /* Make sure we don't use bounce buffers. */ + blk_queue_bounce_limit(rq, BLK_BOUNCE_ANY); + + gd->queue = rq; + + return 0; +} + +int +xlvbd_add(blkif_sector_t capacity, int vdevice, u16 vdisk_info, + u16 sector_size, struct blkfront_info *info) +{ + int major, minor; + struct gendisk *gd; + struct xlbd_major_info *mi; + int nr_minors = 1; + int err = -ENODEV; + unsigned int offset; + + if ((vdevice>>EXT_SHIFT) > 1) { + /* this is above the extended range; something is wrong */ + printk(KERN_WARNING "blkfront: vdevice 0x%x is above the extended range; ignoring\n", vdevice); + return -ENODEV; + } + + if (!VDEV_IS_EXTENDED(vdevice)) { + major = BLKIF_MAJOR(vdevice); + minor = BLKIF_MINOR(vdevice); + } + else { + major = 202; + minor = BLKIF_MINOR_EXT(vdevice); + } + + BUG_ON(info->gd != NULL); + BUG_ON(info->mi != NULL); + BUG_ON(info->rq != NULL); + + mi = xlbd_get_major_info(major, minor, vdevice); + if (mi == NULL) + goto out; + info->mi = mi; + + if (!(vdisk_info & VDISK_CDROM) && + (minor & ((1 << mi->type->partn_shift) - 1)) == 0) + nr_minors = 1 << mi->type->partn_shift; + + err = xlbd_reserve_minors(mi, minor, nr_minors); + if (err) + goto out; + err = -ENODEV; + + gd = alloc_disk(nr_minors); + if (gd == NULL) + goto release; + + offset = mi->index * mi->type->disks_per_major + + (minor >> mi->type->partn_shift); + if (nr_minors > 1 || (vdisk_info & VDISK_CDROM)) { + if (offset < 26) { + sprintf(gd->disk_name, "%s%c", + mi->type->diskname, 'a' + offset ); + } + else { + sprintf(gd->disk_name, "%s%c%c", + mi->type->diskname, + 'a' + ((offset/26)-1), 'a' + (offset%26) ); + } + } + else { + if (offset < 26) { + sprintf(gd->disk_name, "%s%c%d", + mi->type->diskname, + 'a' + offset, + minor & ((1 << mi->type->partn_shift) - 1)); + } + else { + sprintf(gd->disk_name, "%s%c%c%d", + mi->type->diskname, + 'a' + ((offset/26)-1), 'a' + (offset%26), + minor & ((1 << mi->type->partn_shift) - 1)); + } + } + + gd->major = mi->major; + gd->first_minor = minor; + gd->fops = &xlvbd_block_fops; + gd->private_data = info; + gd->driverfs_dev = &(info->xbdev->dev); + set_capacity(gd, capacity); + + if (xlvbd_init_blk_queue(gd, sector_size)) { + del_gendisk(gd); + goto release; + } + + info->rq = gd->queue; + info->gd = gd; + + if (info->feature_barrier) + xlvbd_barrier(info); + + if (vdisk_info & VDISK_READONLY) + set_disk_ro(gd, 1); + + if (vdisk_info & VDISK_REMOVABLE) + gd->flags |= GENHD_FL_REMOVABLE; + + if (vdisk_info & VDISK_CDROM) + gd->flags |= GENHD_FL_CD; + + return 0; + + release: + xlbd_release_minors(mi, minor, nr_minors); + out: + if (mi) + xlbd_put_major_info(mi); + info->mi = NULL; + return err; +} + +void +xlvbd_del(struct blkfront_info *info) +{ + unsigned int minor, nr_minors; + + if (info->mi == NULL) + return; + + BUG_ON(info->gd == NULL); + minor = info->gd->first_minor; + nr_minors = info->gd->minors; + del_gendisk(info->gd); + put_disk(info->gd); + info->gd = NULL; + + xlbd_release_minors(info->mi, minor, nr_minors); + xlbd_put_major_info(info->mi); + info->mi = NULL; + + BUG_ON(info->rq == NULL); + blk_cleanup_queue(info->rq); + info->rq = NULL; +} + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16) +int +xlvbd_barrier(struct blkfront_info *info) +{ + int err; + + err = blk_queue_ordered(info->rq, + info->feature_barrier ? QUEUE_ORDERED_DRAIN : QUEUE_ORDERED_NONE, NULL); + if (err) + return err; + printk(KERN_INFO "blkfront: %s: barriers %s\n", + info->gd->disk_name, info->feature_barrier ? "enabled" : "disabled"); + return 0; +} +#else +int +xlvbd_barrier(struct blkfront_info *info) +{ + printk(KERN_INFO "blkfront: %s: barriers disabled\n", info->gd->disk_name); + return -ENOSYS; +} +#endif + +#ifdef CONFIG_SYSFS +static ssize_t show_media(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct xenbus_device *xendev = to_xenbus_device(dev); + struct blkfront_info *info = dev_get_drvdata(&xendev->dev); + + if (info->gd->flags & GENHD_FL_CD) + return sprintf(buf, "cdrom\n"); + return sprintf(buf, "disk\n"); +} + +static struct device_attribute xlvbd_attrs[] = { + __ATTR(media, S_IRUGO, show_media, NULL), +}; + +int xlvbd_sysfs_addif(struct blkfront_info *info) +{ + int i; + int error = 0; + + for (i = 0; i < ARRAY_SIZE(xlvbd_attrs); i++) { + error = device_create_file(info->gd->driverfs_dev, + &xlvbd_attrs[i]); + if (error) + goto fail; + } + return 0; + +fail: + while (--i >= 0) + device_remove_file(info->gd->driverfs_dev, &xlvbd_attrs[i]); + return error; +} + +void xlvbd_sysfs_delif(struct blkfront_info *info) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(xlvbd_attrs); i++) + device_remove_file(info->gd->driverfs_dev, &xlvbd_attrs[i]); +} + +#endif /* CONFIG_SYSFS */ diff --git a/drivers/xenpvdrivers/vcd.c b/drivers/xenpvdrivers/vcd.c new file mode 100644 index 0000000..cc311b4 --- /dev/null +++ b/drivers/xenpvdrivers/vcd.c @@ -0,0 +1,509 @@ +/******************************************************************************* + * vcd.c + * + * Implements CDROM cmd packet passing between frontend guest and backend driver. + * + * Copyright (c) 2008, Pat Campell plc@novell.com + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#define REVISION "$Revision: 1.0 $" + +#include +#include +#include +#include +#include +#include "block.h" + +/* List of cdrom_device_info, can have as many as blkfront supports */ +struct vcd_disk { + struct list_head vcd_entry; + struct cdrom_device_info vcd_cdrom_info; + spinlock_t vcd_cdrom_info_lock; +}; +static LIST_HEAD(vcd_disks); +static DEFINE_SPINLOCK(vcd_disks_lock); + +static struct vcd_disk *xencdrom_get_list_entry(struct gendisk *disk) +{ + struct vcd_disk *ret_vcd = NULL; + struct vcd_disk *vcd; + + spin_lock(&vcd_disks_lock); + list_for_each_entry(vcd, &vcd_disks, vcd_entry) { + if (vcd->vcd_cdrom_info.disk == disk) { + spin_lock(&vcd->vcd_cdrom_info_lock); + ret_vcd = vcd; + break; + } + } + spin_unlock(&vcd_disks_lock); + return ret_vcd; +} + +static void submit_message(struct blkfront_info *info, void *sp) +{ + struct request *req = NULL; + + req = blk_get_request(info->rq, READ, __GFP_WAIT); + if (blk_rq_map_kern(info->rq, req, sp, PAGE_SIZE, __GFP_WAIT)) + goto out; + + req->rq_disk = info->gd; +#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,18) + req->cmd_type = REQ_TYPE_BLOCK_PC; + req->cmd_flags |= REQ_NOMERGE; +#else + req->flags |= REQ_BLOCK_PC; +#endif + req->__sector = 0; + req->__data_len = PAGE_SIZE; + req->timeout = 60*HZ; + + blk_execute_rq(req->q, info->gd, req, 1); + +out: + blk_put_request(req); +} + +static int submit_cdrom_cmd(struct blkfront_info *info, + struct packet_command *cgc) +{ + int ret = 0; + struct page *page; + size_t size; + union xen_block_packet *sp; + struct xen_cdrom_packet *xcp; + struct vcd_generic_command *vgc; + + if (cgc->buffer && cgc->buflen > MAX_PACKET_DATA) { + printk(KERN_WARNING "%s() Packet buffer length is to large \n", __func__); + return -EIO; + } + + page = alloc_page(GFP_NOIO); + if (!page) { + printk(KERN_CRIT "%s() Unable to allocate page\n", __func__); + return -ENOMEM; + } + + size = PAGE_SIZE; + memset(page_address(page), 0, PAGE_SIZE); + sp = page_address(page); + xcp = &(sp->xcp); + xcp->type = XEN_TYPE_CDROM_PACKET; + xcp->payload_offset = PACKET_PAYLOAD_OFFSET; + + vgc = (struct vcd_generic_command *)((char *)sp + xcp->payload_offset); + memcpy(vgc->cmd, cgc->cmd, CDROM_PACKET_SIZE); + vgc->stat = cgc->stat; + vgc->data_direction = cgc->data_direction; + vgc->quiet = cgc->quiet; + vgc->timeout = cgc->timeout; + if (cgc->sense) { + vgc->sense_offset = PACKET_SENSE_OFFSET; + memcpy((char *)sp + vgc->sense_offset, cgc->sense, sizeof(struct request_sense)); + } + if (cgc->buffer) { + vgc->buffer_offset = PACKET_BUFFER_OFFSET; + memcpy((char *)sp + vgc->buffer_offset, cgc->buffer, cgc->buflen); + vgc->buflen = cgc->buflen; + } + + submit_message(info,sp); + + if (xcp->ret) + ret = xcp->err; + + if (cgc->sense) { + memcpy(cgc->sense, (char *)sp + PACKET_SENSE_OFFSET, sizeof(struct request_sense)); + } + if (cgc->buffer && cgc->buflen) { + memcpy(cgc->buffer, (char *)sp + PACKET_BUFFER_OFFSET, cgc->buflen); + } + + __free_page(page); + return ret; +} + + +static int xencdrom_open(struct cdrom_device_info *cdi, int purpose) +{ + int ret = 0; + struct page *page; + struct blkfront_info *info; + union xen_block_packet *sp; + struct xen_cdrom_open *xco; + + info = cdi->disk->private_data; + + if (!info->xbdev) + return -ENODEV; + + if (strlen(info->xbdev->otherend) > MAX_PACKET_DATA) { + return -EIO; + } + + page = alloc_page(GFP_NOIO); + if (!page) { + printk(KERN_CRIT "%s() Unable to allocate page\n", __func__); + return -ENOMEM; + } + + memset(page_address(page), 0, PAGE_SIZE); + sp = page_address(page); + xco = &(sp->xco); + xco->type = XEN_TYPE_CDROM_OPEN; + xco->payload_offset = sizeof(struct xen_cdrom_open); + strcpy((char *)sp + xco->payload_offset, info->xbdev->otherend); + + submit_message(info,sp); + + if (xco->ret) { + ret = xco->err; + goto out; + } + + if (xco->media_present) + set_capacity(cdi->disk, xco->sectors); + +out: + __free_page(page); + return ret; +} + +static void xencdrom_release(struct cdrom_device_info *cdi) +{ +} + +static int xencdrom_media_changed(struct cdrom_device_info *cdi, int disc_nr) +{ + int ret; + struct page *page; + struct blkfront_info *info; + union xen_block_packet *sp; + struct xen_cdrom_media_changed *xcmc; + + info = cdi->disk->private_data; + + page = alloc_page(GFP_NOIO); + if (!page) { + printk(KERN_CRIT "%s() Unable to allocate page\n", __func__); + return -ENOMEM; + } + + memset(page_address(page), 0, PAGE_SIZE); + sp = page_address(page); + xcmc = &(sp->xcmc); + xcmc->type = XEN_TYPE_CDROM_MEDIA_CHANGED; + submit_message(info,sp); + ret = xcmc->media_changed; + + __free_page(page); + + return ret; +} + +static int xencdrom_tray_move(struct cdrom_device_info *cdi, int position) +{ + int ret; + struct packet_command cgc; + struct blkfront_info *info; + + info = cdi->disk->private_data; + init_cdrom_command(&cgc, NULL, 0, CGC_DATA_NONE); + cgc.cmd[0] = GPCMD_START_STOP_UNIT; + if (position) + cgc.cmd[4] = 2; + else + cgc.cmd[4] = 3; + ret = submit_cdrom_cmd(info, &cgc); + return ret; +} + +static int xencdrom_lock_door(struct cdrom_device_info *cdi, int lock) +{ + int ret = 0; + struct blkfront_info *info; + struct packet_command cgc; + + info = cdi->disk->private_data; + init_cdrom_command(&cgc, NULL, 0, CGC_DATA_NONE); + cgc.cmd[0] = GPCMD_PREVENT_ALLOW_MEDIUM_REMOVAL; + cgc.cmd[4] = lock; + ret = submit_cdrom_cmd(info, &cgc); + return ret; +} + +static int xencdrom_packet(struct cdrom_device_info *cdi, + struct packet_command *cgc) +{ + int ret = -EIO; + struct blkfront_info *info; + + info = cdi->disk->private_data; + ret = submit_cdrom_cmd(info, cgc); + cgc->stat = ret; + return ret; +} + +static int xencdrom_audio_ioctl(struct cdrom_device_info *cdi, unsigned int cmd, + void *arg) +{ + return -EINVAL; +} + +/* Query backend to see if CDROM packets are supported */ +static int xencdrom_supported(struct blkfront_info *info) +{ + struct page *page; + union xen_block_packet *sp; + struct xen_cdrom_support *xcs; + + page = alloc_page(GFP_NOIO); + if (!page) { + printk(KERN_CRIT "%s() Unable to allocate page\n", __func__); + return -ENOMEM; + } + + memset(page_address(page), 0, PAGE_SIZE); + sp = page_address(page); + xcs = &(sp->xcs); + xcs->type = XEN_TYPE_CDROM_SUPPORT; + submit_message(info,sp); + return xcs->supported; +} + +static struct cdrom_device_ops xencdrom_dops = { + .open = xencdrom_open, + .release = xencdrom_release, + .media_changed = xencdrom_media_changed, + .tray_move = xencdrom_tray_move, + .lock_door = xencdrom_lock_door, + .generic_packet = xencdrom_packet, + .audio_ioctl = xencdrom_audio_ioctl, + .capability = (CDC_CLOSE_TRAY | CDC_OPEN_TRAY | CDC_LOCK | \ + CDC_MEDIA_CHANGED | CDC_GENERIC_PACKET | CDC_DVD | \ + CDC_CD_R), + .n_minors = 1, +}; + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) +static int xencdrom_block_open(struct inode *inode, struct file *file) +{ + struct block_device *bd = inode->i_bdev; +#else +static int xencdrom_block_open(struct block_device *bd, fmode_t mode) +{ +#endif + struct blkfront_info *info = bd->bd_disk->private_data; + struct vcd_disk *vcd; + int ret = 0; + + if (!info->xbdev) + return -ENODEV; + + if ((vcd = xencdrom_get_list_entry(info->gd))) { +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) + ret = cdrom_open(&vcd->vcd_cdrom_info, inode, file); +#else + ret = cdrom_open(&vcd->vcd_cdrom_info, bd, mode); +#endif + info->users = vcd->vcd_cdrom_info.use_count; + spin_unlock(&vcd->vcd_cdrom_info_lock); + } + return ret; +} + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) +static int xencdrom_block_release(struct inode *inode, struct file *file) +{ + struct gendisk *gd = inode->i_bdev->bd_disk; +#else +static int xencdrom_block_release(struct gendisk *gd, fmode_t mode) +{ +#endif + struct blkfront_info *info = gd->private_data; + struct vcd_disk *vcd; + int ret = 0; + + if ((vcd = xencdrom_get_list_entry(info->gd))) { +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) + ret = cdrom_release(&vcd->vcd_cdrom_info, file); +#else + cdrom_release(&vcd->vcd_cdrom_info, mode); +#endif + spin_unlock(&vcd->vcd_cdrom_info_lock); + if (vcd->vcd_cdrom_info.use_count == 0) { + info->users = 1; +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) + blkif_release(inode, file); +#else + blkif_release(gd, mode); +#endif + } + } + return ret; +} + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28) +static int xencdrom_block_ioctl(struct inode *inode, struct file *file, + unsigned cmd, unsigned long arg) +{ + struct block_device *bd = inode->i_bdev; +#else +static int xencdrom_block_ioctl(struct block_device *bd, fmode_t mode, + unsigned cmd, unsigned long arg) +{ +#endif + struct blkfront_info *info = bd->bd_disk->private_data; + struct vcd_disk *vcd; + int ret = 0; + + if (!(vcd = xencdrom_get_list_entry(info->gd))) + goto out; + + switch (cmd) { + case 2285: /* SG_IO */ + ret = -ENOSYS; + break; + case CDROMEJECT: + ret = xencdrom_tray_move(&vcd->vcd_cdrom_info, 1); + break; + case CDROMCLOSETRAY: + ret = xencdrom_tray_move(&vcd->vcd_cdrom_info, 0); + break; + case CDROM_GET_CAPABILITY: + ret = vcd->vcd_cdrom_info.ops->capability & ~vcd->vcd_cdrom_info.mask; + break; + case CDROM_SET_OPTIONS: + ret = vcd->vcd_cdrom_info.options; + break; + case CDROM_SEND_PACKET: + ret = submit_cdrom_cmd(info, (struct packet_command *)arg); + break; + default: + /* Not supported, augment supported above if necessary */ + printk("%s():%d Unsupported IOCTL:%x \n", __func__, __LINE__, cmd); + ret = -ENOTTY; + break; + } + spin_unlock(&vcd->vcd_cdrom_info_lock); +out: + return ret; +} + +/* Called as result of cdrom_open, vcd_cdrom_info_lock already held */ +static int xencdrom_block_media_changed(struct gendisk *disk) +{ + struct vcd_disk *vcd; + struct vcd_disk *ret_vcd = NULL; + int ret = 0; + + spin_lock(&vcd_disks_lock); + list_for_each_entry(vcd, &vcd_disks, vcd_entry) { + if (vcd->vcd_cdrom_info.disk == disk) { + ret_vcd = vcd; + break; + } + } + spin_unlock(&vcd_disks_lock); + if (ret_vcd) { + ret = cdrom_media_changed(&ret_vcd->vcd_cdrom_info); + } + return ret; +} + +static const struct block_device_operations xencdrom_bdops = +{ + .owner = THIS_MODULE, + .open = xencdrom_block_open, + .release = xencdrom_block_release, + .ioctl = xencdrom_block_ioctl, + .media_changed = xencdrom_block_media_changed, +}; + +void register_vcd(struct blkfront_info *info) +{ + struct gendisk *gd = info->gd; + struct vcd_disk *vcd; + + /* Make sure this is for a CD device */ + if (!(gd->flags & GENHD_FL_CD)) + goto out; + + /* Make sure we have backend support */ + if (!xencdrom_supported(info)) { + goto out; + } + + /* Create new vcd_disk and fill in cdrom_info */ + vcd = (struct vcd_disk *)kzalloc(sizeof(struct vcd_disk), GFP_KERNEL); + if (!vcd) { + printk(KERN_INFO "%s(): Unable to allocate vcd struct!\n", __func__); + goto out; + } + spin_lock_init(&vcd->vcd_cdrom_info_lock); + + vcd->vcd_cdrom_info.ops = &xencdrom_dops; + vcd->vcd_cdrom_info.speed = 4; + vcd->vcd_cdrom_info.capacity = 1; + vcd->vcd_cdrom_info.options = 0; + strcpy(vcd->vcd_cdrom_info.name, gd->disk_name); + vcd->vcd_cdrom_info.mask = (CDC_CD_RW | CDC_DVD_R | CDC_DVD_RAM | + CDC_SELECT_DISC | CDC_SELECT_SPEED | + CDC_MRW | CDC_MRW_W | CDC_RAM); + + if (register_cdrom(&(vcd->vcd_cdrom_info)) != 0) { + printk(KERN_WARNING "%s() Cannot register blkdev as a cdrom %d!\n", __func__, + gd->major); + goto err_out; + } + gd->fops = &xencdrom_bdops; + vcd->vcd_cdrom_info.disk = gd; + + spin_lock(&vcd_disks_lock); + list_add(&(vcd->vcd_entry), &vcd_disks); + spin_unlock(&vcd_disks_lock); +out: + return; +err_out: + kfree(vcd); +} + +void unregister_vcd(struct blkfront_info *info) { + struct gendisk *gd = info->gd; + struct vcd_disk *vcd; + + spin_lock(&vcd_disks_lock); + list_for_each_entry(vcd, &vcd_disks, vcd_entry) { + if (vcd->vcd_cdrom_info.disk == gd) { + spin_lock(&vcd->vcd_cdrom_info_lock); + unregister_cdrom(&vcd->vcd_cdrom_info); + list_del(&vcd->vcd_entry); + spin_unlock(&vcd->vcd_cdrom_info_lock); + kfree(vcd); + break; + } + } + spin_unlock(&vcd_disks_lock); +} + diff --git a/drivers/xenpvdrivers/xen_proc.c b/drivers/xenpvdrivers/xen_proc.c new file mode 100644 index 0000000..a056972 --- /dev/null +++ b/drivers/xenpvdrivers/xen_proc.c @@ -0,0 +1,23 @@ + +#include +#include +#include + +static struct proc_dir_entry *xen_base; + +struct proc_dir_entry *create_xen_proc_entry(const char *name, mode_t mode) +{ + if ( xen_base == NULL ) + if ( (xen_base = proc_mkdir("xen", NULL)) == NULL ) + panic("Couldn't create /proc/xen"); + return create_proc_entry(name, mode, xen_base); +} + +EXPORT_SYMBOL_GPL(create_xen_proc_entry); + +void remove_xen_proc_entry(const char *name) +{ + remove_proc_entry(name, xen_base); +} + +EXPORT_SYMBOL_GPL(remove_xen_proc_entry); diff --git a/drivers/xenpvdrivers/xen_support.c b/drivers/xenpvdrivers/xen_support.c new file mode 100644 index 0000000..ab50aac --- /dev/null +++ b/drivers/xenpvdrivers/xen_support.c @@ -0,0 +1,74 @@ +/****************************************************************************** + * support.c + * Xen module support functions. + * Copyright (C) 2004, Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + +#include +#include +#include +#include +#include +#include +#include "platform-pci.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#if defined (__ia64__) +unsigned long __hypercall(unsigned long a1, unsigned long a2, + unsigned long a3, unsigned long a4, + unsigned long a5, unsigned long cmd) +{ + unsigned long __res; + __asm__ __volatile__ (";;\n" + "mov r2=%1\n" + "break 0x1000 ;;\n" + "mov %0=r8 ;;\n" + : "=r"(__res) : "r"(cmd) : "r2", "r8", "memory"); + + return __res; +} +EXPORT_SYMBOL(__hypercall); + +int HYPERVISOR_grant_table_op(unsigned int cmd, void *uop, unsigned int count) +{ + return xencomm_hypercall_grant_table_op(cmd, uop, count); +} +EXPORT_SYMBOL(HYPERVISOR_grant_table_op); + +/* without using balloon driver on PV-on-HVM for ia64 */ +void balloon_update_driver_allowance(long delta) +{ + /* nothing */ +} +EXPORT_SYMBOL_GPL(balloon_update_driver_allowance); + +void balloon_release_driver_page(struct page *page) +{ + /* nothing */ +} +EXPORT_SYMBOL_GPL(balloon_release_driver_page); +#endif /* __ia64__ */ + +void xen_machphys_update(unsigned long mfn, unsigned long pfn) +{ + BUG(); +} +EXPORT_SYMBOL(xen_machphys_update); + diff --git a/drivers/xenpvdrivers/xenbus.c b/drivers/xenpvdrivers/xenbus.c new file mode 100644 index 0000000..7cde152 --- /dev/null +++ b/drivers/xenpvdrivers/xenbus.c @@ -0,0 +1,420 @@ +/* + * Xen SCSI frontend driver + * + * Copyright (c) 2008, FUJITSU Limited + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + + +#include +#include "scsifront-common.h" + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,11) + #define DEFAULT_TASK_COMM_LEN 16 +#else + #define DEFAULT_TASK_COMM_LEN TASK_COMM_LEN +#endif + +extern struct scsi_host_template scsifront_sht; + +static void scsifront_free(struct vscsifrnt_info *info) +{ + struct Scsi_Host *host = info->host; + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,14) + if (host->shost_state != SHOST_DEL) { +#else + if (!test_bit(SHOST_DEL, &host->shost_state)) { +#endif + scsi_remove_host(info->host); + } + + if (info->ring_ref != GRANT_INVALID_REF) { + gnttab_end_foreign_access(info->ring_ref, + (unsigned long)info->ring.sring); + info->ring_ref = GRANT_INVALID_REF; + info->ring.sring = NULL; + } + + if (info->irq) + unbind_from_irqhandler(info->irq, info); + info->irq = 0; + + scsi_host_put(info->host); +} + + +static int scsifront_alloc_ring(struct vscsifrnt_info *info) +{ + struct xenbus_device *dev = info->dev; + struct vscsiif_sring *sring; + int err = -ENOMEM; + + + info->ring_ref = GRANT_INVALID_REF; + + /***** Frontend to Backend ring start *****/ + sring = (struct vscsiif_sring *) __get_free_page(GFP_KERNEL); + if (!sring) { + xenbus_dev_fatal(dev, err, "fail to allocate shared ring (Front to Back)"); + return err; + } + SHARED_RING_INIT(sring); + FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE); + + err = xenbus_grant_ring(dev, virt_to_mfn(sring)); + if (err < 0) { + free_page((unsigned long) sring); + info->ring.sring = NULL; + xenbus_dev_fatal(dev, err, "fail to grant shared ring (Front to Back)"); + goto free_sring; + } + info->ring_ref = err; + + err = bind_listening_port_to_irqhandler( + dev->otherend_id, scsifront_intr, + IRQF_SAMPLE_RANDOM, "scsifront", info); + + if (err <= 0) { + xenbus_dev_fatal(dev, err, "bind_listening_port_to_irqhandler"); + goto free_sring; + } + info->irq = err; + + return 0; + +/* free resource */ +free_sring: + scsifront_free(info); + + return err; +} + + +static int scsifront_init_ring(struct vscsifrnt_info *info) +{ + struct xenbus_device *dev = info->dev; + struct xenbus_transaction xbt; + int err; + + DPRINTK("%s\n",__FUNCTION__); + + err = scsifront_alloc_ring(info); + if (err) + return err; + DPRINTK("%u %u\n", info->ring_ref, info->evtchn); + +again: + err = xenbus_transaction_start(&xbt); + if (err) { + xenbus_dev_fatal(dev, err, "starting transaction"); + } + + err = xenbus_printf(xbt, dev->nodename, "ring-ref", "%u", + info->ring_ref); + if (err) { + xenbus_dev_fatal(dev, err, "%s", "writing ring-ref"); + goto fail; + } + + err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u", + irq_to_evtchn_port(info->irq)); + + if (err) { + xenbus_dev_fatal(dev, err, "%s", "writing event-channel"); + goto fail; + } + + err = xenbus_transaction_end(xbt, 0); + if (err) { + if (err == -EAGAIN) + goto again; + xenbus_dev_fatal(dev, err, "completing transaction"); + goto free_sring; + } + + return 0; + +fail: + xenbus_transaction_end(xbt, 1); +free_sring: + /* free resource */ + scsifront_free(info); + + return err; +} + + +static int scsifront_probe(struct xenbus_device *dev, + const struct xenbus_device_id *id) +{ + struct vscsifrnt_info *info; + struct Scsi_Host *host; + int i, err = -ENOMEM; + char name[DEFAULT_TASK_COMM_LEN]; + + host = scsi_host_alloc(&scsifront_sht, sizeof(*info)); + if (!host) { + xenbus_dev_fatal(dev, err, "fail to allocate scsi host"); + return err; + } + info = (struct vscsifrnt_info *) host->hostdata; + info->host = host; + + + dev_set_drvdata(&dev->dev, info); + info->dev = dev; + + for (i = 0; i < VSCSIIF_MAX_REQS; i++) { + info->shadow[i].next_free = i + 1; + init_waitqueue_head(&(info->shadow[i].wq_reset)); + info->shadow[i].wait_reset = 0; + } + info->shadow[VSCSIIF_MAX_REQS - 1].next_free = 0x0fff; + + err = scsifront_init_ring(info); + if (err) { + scsi_host_put(host); + return err; + } + + init_waitqueue_head(&info->wq); + spin_lock_init(&info->io_lock); + spin_lock_init(&info->shadow_lock); + + snprintf(name, DEFAULT_TASK_COMM_LEN, "vscsiif.%d", info->host->host_no); + + info->kthread = kthread_run(scsifront_schedule, info, name); + if (IS_ERR(info->kthread)) { + err = PTR_ERR(info->kthread); + info->kthread = NULL; + printk(KERN_ERR "scsifront: kthread start err %d\n", err); + goto free_sring; + } + + host->max_id = VSCSIIF_MAX_TARGET; + host->max_channel = 0; + host->max_lun = VSCSIIF_MAX_LUN; + host->max_sectors = (VSCSIIF_SG_TABLESIZE - 1) * PAGE_SIZE / 512; + + err = scsi_add_host(host, &dev->dev); + if (err) { + printk(KERN_ERR "scsifront: fail to add scsi host %d\n", err); + goto free_sring; + } + + xenbus_switch_state(dev, XenbusStateInitialised); + + return 0; + +free_sring: + /* free resource */ + scsifront_free(info); + return err; +} + +static int scsifront_remove(struct xenbus_device *dev) +{ + struct vscsifrnt_info *info = dev_get_drvdata(&dev->dev); + + DPRINTK("%s: %s removed\n",__FUNCTION__ ,dev->nodename); + + if (info->kthread) { + kthread_stop(info->kthread); + info->kthread = NULL; + } + + scsifront_free(info); + + return 0; +} + + +static int scsifront_disconnect(struct vscsifrnt_info *info) +{ + struct xenbus_device *dev = info->dev; + struct Scsi_Host *host = info->host; + + DPRINTK("%s: %s disconnect\n",__FUNCTION__ ,dev->nodename); + + /* + When this function is executed, all devices of + Frontend have been deleted. + Therefore, it need not block I/O before remove_host. + */ + + scsi_remove_host(host); + xenbus_frontend_closed(dev); + + return 0; +} + +#define VSCSIFRONT_OP_ADD_LUN 1 +#define VSCSIFRONT_OP_DEL_LUN 2 + +static void scsifront_do_lun_hotplug(struct vscsifrnt_info *info, int op) +{ + struct xenbus_device *dev = info->dev; + int i, err = 0; + char str[64], state_str[64]; + char **dir; + unsigned int dir_n = 0; + unsigned int device_state; + unsigned int hst, chn, tgt, lun; + struct scsi_device *sdev; + + dir = xenbus_directory(XBT_NIL, dev->otherend, "vscsi-devs", &dir_n); + if (IS_ERR(dir)) + return; + + for (i = 0; i < dir_n; i++) { + /* read status */ + snprintf(str, sizeof(str), "vscsi-devs/%s/state", dir[i]); + err = xenbus_scanf(XBT_NIL, dev->otherend, str, "%u", + &device_state); + if (XENBUS_EXIST_ERR(err)) + continue; + + /* virtual SCSI device */ + snprintf(str, sizeof(str), "vscsi-devs/%s/v-dev", dir[i]); + err = xenbus_scanf(XBT_NIL, dev->otherend, str, + "%u:%u:%u:%u", &hst, &chn, &tgt, &lun); + if (XENBUS_EXIST_ERR(err)) + continue; + + /* front device state path */ + snprintf(state_str, sizeof(state_str), "vscsi-devs/%s/state", dir[i]); + + switch (op) { + case VSCSIFRONT_OP_ADD_LUN: + if (device_state == XenbusStateInitialised) { + sdev = scsi_device_lookup(info->host, chn, tgt, lun); + if (sdev) { + printk(KERN_ERR "scsifront: Device already in use.\n"); + scsi_device_put(sdev); + xenbus_printf(XBT_NIL, dev->nodename, + state_str, "%d", XenbusStateClosed); + } else { + scsi_add_device(info->host, chn, tgt, lun); + xenbus_printf(XBT_NIL, dev->nodename, + state_str, "%d", XenbusStateConnected); + } + } + break; + case VSCSIFRONT_OP_DEL_LUN: + if (device_state == XenbusStateClosing) { + sdev = scsi_device_lookup(info->host, chn, tgt, lun); + if (sdev) { + scsi_remove_device(sdev); + scsi_device_put(sdev); + xenbus_printf(XBT_NIL, dev->nodename, + state_str, "%d", XenbusStateClosed); + } + } + break; + default: + break; + } + } + + kfree(dir); + return; +} + + + + +static void scsifront_backend_changed(struct xenbus_device *dev, + enum xenbus_state backend_state) +{ + struct vscsifrnt_info *info = dev_get_drvdata(&dev->dev); + + DPRINTK("%p %u %u\n", dev, dev->state, backend_state); + + switch (backend_state) { + case XenbusStateUnknown: + case XenbusStateInitialising: + case XenbusStateInitWait: + case XenbusStateClosed: + break; + + case XenbusStateInitialised: + break; + + case XenbusStateConnected: + if (xenbus_read_driver_state(dev->nodename) == + XenbusStateInitialised) { + scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_ADD_LUN); + } + + if (dev->state == XenbusStateConnected) + break; + + xenbus_switch_state(dev, XenbusStateConnected); + break; + + case XenbusStateClosing: + scsifront_disconnect(info); + break; + + case XenbusStateReconfiguring: + scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_DEL_LUN); + xenbus_switch_state(dev, XenbusStateReconfiguring); + break; + + case XenbusStateReconfigured: + scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_ADD_LUN); + xenbus_switch_state(dev, XenbusStateConnected); + break; + } +} + + +static struct xenbus_device_id scsifront_ids[] = { + { "vscsi" }, + { "" } +}; +MODULE_ALIAS("xen:vscsi"); + +static struct xenbus_driver scsifront_driver = { + .name = "vscsi", + .ids = scsifront_ids, + .probe = scsifront_probe, + .remove = scsifront_remove, +/* .resume = scsifront_resume, */ + .otherend_changed = scsifront_backend_changed, +}; + +int scsifront_xenbus_init(void) +{ + return xenbus_register_frontend(&scsifront_driver); +} + +void scsifront_xenbus_unregister(void) +{ + xenbus_unregister_driver(&scsifront_driver); +} + diff --git a/drivers/xenpvdrivers/xenbus_backend_client.c b/drivers/xenpvdrivers/xenbus_backend_client.c new file mode 100644 index 0000000..4df6b5f --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_backend_client.c @@ -0,0 +1,154 @@ +/****************************************************************************** + * Backend-client-facing interface for the Xenbus driver. In other words, the + * interface between the Xenbus and the device-specific code in the backend + * driver. + * + * Copyright (C) 2005-2006 XenSource Ltd + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include + +/* Based on Rusty Russell's skeleton driver's map_page */ +struct vm_struct *xenbus_map_ring_valloc(struct xenbus_device *dev, int gnt_ref) +{ + struct gnttab_map_grant_ref op; + struct vm_struct *area; + + area = alloc_vm_area(PAGE_SIZE); + if (!area) + return ERR_PTR(-ENOMEM); + + gnttab_set_map_op(&op, (unsigned long)area->addr, GNTMAP_host_map, + gnt_ref, dev->otherend_id); + + do { + if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1)) + BUG(); + msleep(10); + } while(op.status == GNTST_eagain); + + if (op.status != GNTST_okay) { + free_vm_area(area); + xenbus_dev_fatal(dev, op.status, + "mapping in shared page %d from domain %d", + gnt_ref, dev->otherend_id); + BUG_ON(!IS_ERR(ERR_PTR(op.status))); + return ERR_PTR(op.status); + } + + /* Stuff the handle in an unused field */ + area->phys_addr = (unsigned long)op.handle; + + return area; +} +EXPORT_SYMBOL_GPL(xenbus_map_ring_valloc); + + +int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref, + grant_handle_t *handle, void *vaddr) +{ + struct gnttab_map_grant_ref op; + + gnttab_set_map_op(&op, (unsigned long)vaddr, GNTMAP_host_map, + gnt_ref, dev->otherend_id); + do { + if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1)) + BUG(); + msleep(10); + } while(op.status == GNTST_eagain); + + if (op.status != GNTST_okay) { + xenbus_dev_fatal(dev, op.status, + "mapping in shared page %d from domain %d", + gnt_ref, dev->otherend_id); + } else + *handle = op.handle; + + return op.status; +} +EXPORT_SYMBOL_GPL(xenbus_map_ring); + + +/* Based on Rusty Russell's skeleton driver's unmap_page */ +int xenbus_unmap_ring_vfree(struct xenbus_device *dev, struct vm_struct *area) +{ + struct gnttab_unmap_grant_ref op; + + gnttab_set_unmap_op(&op, (unsigned long)area->addr, GNTMAP_host_map, + (grant_handle_t)area->phys_addr); + + if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1)) + BUG(); + + if (op.status == GNTST_okay) + free_vm_area(area); + else + xenbus_dev_error(dev, op.status, + "unmapping page at handle %d error %d", + (int16_t)area->phys_addr, op.status); + + return op.status; +} +EXPORT_SYMBOL_GPL(xenbus_unmap_ring_vfree); + + +int xenbus_unmap_ring(struct xenbus_device *dev, + grant_handle_t handle, void *vaddr) +{ + struct gnttab_unmap_grant_ref op; + + gnttab_set_unmap_op(&op, (unsigned long)vaddr, GNTMAP_host_map, + handle); + if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1)) + BUG(); + + if (op.status != GNTST_okay) + xenbus_dev_error(dev, op.status, + "unmapping page at handle %d error %d", + handle, op.status); + + return op.status; +} +EXPORT_SYMBOL_GPL(xenbus_unmap_ring); + +int xenbus_dev_is_online(struct xenbus_device *dev) +{ + int rc, val; + + rc = xenbus_scanf(XBT_NIL, dev->nodename, "online", "%d", &val); + if (rc != 1) + val = 0; /* no online node present */ + + return val; +} +EXPORT_SYMBOL_GPL(xenbus_dev_is_online); + +MODULE_LICENSE("Dual BSD/GPL"); diff --git a/drivers/xenpvdrivers/xenbus_client.c b/drivers/xenpvdrivers/xenbus_client.c new file mode 100644 index 0000000..e7a9cc0 --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_client.c @@ -0,0 +1,610 @@ +/****************************************************************************** + * Client-facing interface for the Xenbus driver. In other words, the + * interface between the Xenbus and the device-specific code, be it the + * frontend or the backend of that driver. + * + * Copyright (C) 2005 XenSource Ltd + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +#include +#include +#include +#include +#else +#include +#include +#include +#include +#include +#include +#include +#endif +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +const char *xenbus_strstate(enum xenbus_state state) +{ + static const char *const name[] = { + [ XenbusStateUnknown ] = "Unknown", + [ XenbusStateInitialising ] = "Initialising", + [ XenbusStateInitWait ] = "InitWait", + [ XenbusStateInitialised ] = "Initialised", + [ XenbusStateConnected ] = "Connected", + [ XenbusStateClosing ] = "Closing", + [ XenbusStateClosed ] = "Closed", + [ XenbusStateReconfiguring ] = "Reconfiguring", + [ XenbusStateReconfigured ] = "Reconfigured", + }; + return (state < ARRAY_SIZE(name)) ? name[state] : "INVALID"; +} +EXPORT_SYMBOL_GPL(xenbus_strstate); + +/** + * xenbus_watch_path - register a watch + * @dev: xenbus device + * @path: path to watch + * @watch: watch to register + * @callback: callback to register + * + * Register a @watch on the given path, using the given xenbus_watch structure + * for storage, and the given @callback function as the callback. Return 0 on + * success, or -errno on error. On success, the given @path will be saved as + * @watch->node, and remains the caller's to free. On error, @watch->node will + * be NULL, the device will switch to %XenbusStateClosing, and the error will + * be saved in the store. + */ +int xenbus_watch_path(struct xenbus_device *dev, const char *path, + struct xenbus_watch *watch, + void (*callback)(struct xenbus_watch *, + const char **, unsigned int)) +{ + int err; + + watch->node = path; + watch->callback = callback; + + err = register_xenbus_watch(watch); + + if (err) { + watch->node = NULL; + watch->callback = NULL; + xenbus_dev_fatal(dev, err, "adding watch on %s", path); + } + + return err; +} +EXPORT_SYMBOL_GPL(xenbus_watch_path); + + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +int xenbus_watch_path2(struct xenbus_device *dev, const char *path, + const char *path2, struct xenbus_watch *watch, + void (*callback)(struct xenbus_watch *, + const char **, unsigned int)) +{ + int err; + char *state = kasprintf(GFP_NOIO | __GFP_HIGH, "%s/%s", path, path2); + if (!state) { + xenbus_dev_fatal(dev, -ENOMEM, "allocating path for watch"); + return -ENOMEM; + } + err = xenbus_watch_path(dev, state, watch, callback); + + if (err) + kfree(state); + return err; +} +EXPORT_SYMBOL_GPL(xenbus_watch_path2); +#else +/** + * xenbus_watch_pathfmt - register a watch on a sprintf-formatted path + * @dev: xenbus device + * @watch: watch to register + * @callback: callback to register + * @pathfmt: format of path to watch + * + * Register a watch on the given @path, using the given xenbus_watch + * structure for storage, and the given @callback function as the callback. + * Return 0 on success, or -errno on error. On success, the watched path + * (@path/@path2) will be saved as @watch->node, and becomes the caller's to + * kfree(). On error, watch->node will be NULL, so the caller has nothing to + * free, the device will switch to %XenbusStateClosing, and the error will be + * saved in the store. + */ +int xenbus_watch_pathfmt(struct xenbus_device *dev, + struct xenbus_watch *watch, + void (*callback)(struct xenbus_watch *, + const char **, unsigned int), + const char *pathfmt, ...) +{ + int err; + va_list ap; + char *path; + + va_start(ap, pathfmt); + path = kvasprintf(GFP_NOIO | __GFP_HIGH, pathfmt, ap); + va_end(ap); + + if (!path) { + xenbus_dev_fatal(dev, -ENOMEM, "allocating path for watch"); + return -ENOMEM; + } + err = xenbus_watch_path(dev, path, watch, callback); + + if (err) + kfree(path); + return err; +} +EXPORT_SYMBOL_GPL(xenbus_watch_pathfmt); +#endif + + +/** + * xenbus_switch_state + * @dev: xenbus device + * @state: new state + * + * Advertise in the store a change of the given driver to the given new_state. + * Return 0 on success, or -errno on error. On error, the device will switch + * to XenbusStateClosing, and the error will be saved in the store. + */ +int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state state) +{ + /* We check whether the state is currently set to the given value, and + if not, then the state is set. We don't want to unconditionally + write the given state, because we don't want to fire watches + unnecessarily. Furthermore, if the node has gone, we don't write + to it, as the device will be tearing down, and we don't want to + resurrect that directory. + + Note that, because of this cached value of our state, this function + will not work inside a Xenstore transaction (something it was + trying to in the past) because dev->state would not get reset if + the transaction was aborted. + + */ + + int current_state; + int err; + + if (state == dev->state) + return 0; + + err = xenbus_scanf(XBT_NIL, dev->nodename, "state", "%d", + ¤t_state); + if (err != 1) + return 0; + + err = xenbus_printf(XBT_NIL, dev->nodename, "state", "%d", state); + if (err) { + if (state != XenbusStateClosing) /* Avoid looping */ + xenbus_dev_fatal(dev, err, "writing new state"); + return err; + } + + dev->state = state; + + return 0; +} +EXPORT_SYMBOL_GPL(xenbus_switch_state); + +int xenbus_frontend_closed(struct xenbus_device *dev) +{ + xenbus_switch_state(dev, XenbusStateClosed); + complete(&dev->down); + return 0; +} +EXPORT_SYMBOL_GPL(xenbus_frontend_closed); + +/** + * Return the path to the error node for the given device, or NULL on failure. + * If the value returned is non-NULL, then it is the caller's to kfree. + */ +static char *error_path(struct xenbus_device *dev) +{ + return kasprintf(GFP_KERNEL, "error/%s", dev->nodename); +} + + +static void _dev_error(struct xenbus_device *dev, int err, + const char *fmt, va_list ap) +{ + int ret; + unsigned int len; + char *printf_buffer = NULL, *path_buffer = NULL; + +#define PRINTF_BUFFER_SIZE 4096 + printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_KERNEL); + if (printf_buffer == NULL) + goto fail; + + len = sprintf(printf_buffer, "%i ", -err); + ret = vsnprintf(printf_buffer+len, PRINTF_BUFFER_SIZE-len, fmt, ap); + + BUG_ON(len + ret > PRINTF_BUFFER_SIZE-1); + + dev_err(&dev->dev, "%s\n", printf_buffer); + + path_buffer = error_path(dev); + + if (path_buffer == NULL) { + dev_err(&dev->dev, + "xenbus: failed to write error node for %s (%s)\n", + dev->nodename, printf_buffer); + goto fail; + } + + if (xenbus_write(XBT_NIL, path_buffer, "error", printf_buffer) != 0) { + dev_err(&dev->dev, + "xenbus: failed to write error node for %s (%s)\n", + dev->nodename, printf_buffer); + goto fail; + } + +fail: + kfree(printf_buffer); + kfree(path_buffer); +} + + +/** + * xenbus_dev_error + * @dev: xenbus device + * @err: error to report + * @fmt: error message format + * + * Report the given negative errno into the store, along with the given + * formatted message. + */ +void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt, + ...) +{ + va_list ap; + + va_start(ap, fmt); + _dev_error(dev, err, fmt, ap); + va_end(ap); +} +EXPORT_SYMBOL_GPL(xenbus_dev_error); + + +/** + * xenbus_dev_fatal + * @dev: xenbus device + * @err: error to report + * @fmt: error message format + * + * Equivalent to xenbus_dev_error(dev, err, fmt, args), followed by + * xenbus_switch_state(dev, XenbusStateClosing) to schedule an orderly + * closedown of this driver and its peer. + */ +void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt, + ...) +{ + va_list ap; + + va_start(ap, fmt); + _dev_error(dev, err, fmt, ap); + va_end(ap); + + xenbus_switch_state(dev, XenbusStateClosing); +} +EXPORT_SYMBOL_GPL(xenbus_dev_fatal); + + +/** + * xenbus_grant_ring + * @dev: xenbus device + * @ring_mfn: mfn of ring to grant + * + * Grant access to the given @ring_mfn to the peer of the given device. Return + * 0 on success, or -errno on error. On error, the device will switch to + * XenbusStateClosing, and the error will be saved in the store. + */ +int xenbus_grant_ring(struct xenbus_device *dev, unsigned long ring_mfn) +{ + int err = gnttab_grant_foreign_access(dev->otherend_id, ring_mfn, 0); + if (err < 0) + xenbus_dev_fatal(dev, err, "granting access to ring page"); + return err; +} +EXPORT_SYMBOL_GPL(xenbus_grant_ring); + + +/** + * Allocate an event channel for the given xenbus_device, assigning the newly + * created local port to *port. Return 0 on success, or -errno on error. On + * error, the device will switch to XenbusStateClosing, and the error will be + * saved in the store. + */ +int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port) +{ + struct evtchn_alloc_unbound alloc_unbound; + int err; + + alloc_unbound.dom = DOMID_SELF; + alloc_unbound.remote_dom = dev->otherend_id; + + err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, + &alloc_unbound); + if (err) + xenbus_dev_fatal(dev, err, "allocating event channel"); + else + *port = alloc_unbound.port; + + return err; +} +EXPORT_SYMBOL_GPL(xenbus_alloc_evtchn); + + +#if 0 /* !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) */ +/** + * Bind to an existing interdomain event channel in another domain. Returns 0 + * on success and stores the local port in *port. On error, returns -errno, + * switches the device to XenbusStateClosing, and saves the error in XenStore. + */ +int xenbus_bind_evtchn(struct xenbus_device *dev, int remote_port, int *port) +{ + struct evtchn_bind_interdomain bind_interdomain; + int err; + + bind_interdomain.remote_dom = dev->otherend_id; + bind_interdomain.remote_port = remote_port; + + err = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain, + &bind_interdomain); + if (err) + xenbus_dev_fatal(dev, err, + "binding to event channel %d from domain %d", + remote_port, dev->otherend_id); + else + *port = bind_interdomain.local_port; + + return err; +} +EXPORT_SYMBOL_GPL(xenbus_bind_evtchn); +#endif + + +/** + * Free an existing event channel. Returns 0 on success or -errno on error. + */ +int xenbus_free_evtchn(struct xenbus_device *dev, int port) +{ + struct evtchn_close close; + int err; + + close.port = port; + + err = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close); + if (err) + xenbus_dev_error(dev, err, "freeing event channel %d", port); + + return err; +} +EXPORT_SYMBOL_GPL(xenbus_free_evtchn); + + +#if 0 /* !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) */ +/** + * xenbus_map_ring_valloc + * @dev: xenbus device + * @gnt_ref: grant reference + * @vaddr: pointer to address to be filled out by mapping + * + * Based on Rusty Russell's skeleton driver's map_page. + * Map a page of memory into this domain from another domain's grant table. + * xenbus_map_ring_valloc allocates a page of virtual address space, maps the + * page to that address, and sets *vaddr to that address. + * Returns 0 on success, and GNTST_* (see xen/include/interface/grant_table.h) + * or -ENOMEM on error. If an error is returned, device will switch to + * XenbusStateClosing and the error message will be saved in XenStore. + */ +int xenbus_map_ring_valloc(struct xenbus_device *dev, int gnt_ref, void **vaddr) +{ + struct gnttab_map_grant_ref op = { + .flags = GNTMAP_host_map, + .ref = gnt_ref, + .dom = dev->otherend_id, + }; + struct vm_struct *area; + + *vaddr = NULL; + + area = xen_alloc_vm_area(PAGE_SIZE); + if (!area) + return -ENOMEM; + + op.host_addr = (unsigned long)area->addr; + + if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1)) + BUG(); + + if (op.status != GNTST_okay) { + xen_free_vm_area(area); + xenbus_dev_fatal(dev, op.status, + "mapping in shared page %d from domain %d", + gnt_ref, dev->otherend_id); + return op.status; + } + + /* Stuff the handle in an unused field */ + area->phys_addr = (unsigned long)op.handle; + + *vaddr = area->addr; + return 0; +} +EXPORT_SYMBOL_GPL(xenbus_map_ring_valloc); + + +/** + * xenbus_map_ring + * @dev: xenbus device + * @gnt_ref: grant reference + * @handle: pointer to grant handle to be filled + * @vaddr: address to be mapped to + * + * Map a page of memory into this domain from another domain's grant table. + * xenbus_map_ring does not allocate the virtual address space (you must do + * this yourself!). It only maps in the page to the specified address. + * Returns 0 on success, and GNTST_* (see xen/include/interface/grant_table.h) + * or -ENOMEM on error. If an error is returned, device will switch to + * XenbusStateClosing and the error message will be saved in XenStore. + */ +int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref, + grant_handle_t *handle, void *vaddr) +{ + struct gnttab_map_grant_ref op = { + .host_addr = (unsigned long)vaddr, + .flags = GNTMAP_host_map, + .ref = gnt_ref, + .dom = dev->otherend_id, + }; + + if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1)) + BUG(); + + if (op.status != GNTST_okay) { + xenbus_dev_fatal(dev, op.status, + "mapping in shared page %d from domain %d", + gnt_ref, dev->otherend_id); + } else + *handle = op.handle; + + return op.status; +} +EXPORT_SYMBOL_GPL(xenbus_map_ring); + + +/** + * xenbus_unmap_ring_vfree + * @dev: xenbus device + * @vaddr: addr to unmap + * + * Based on Rusty Russell's skeleton driver's unmap_page. + * Unmap a page of memory in this domain that was imported from another domain. + * Use xenbus_unmap_ring_vfree if you mapped in your memory with + * xenbus_map_ring_valloc (it will free the virtual address space). + * Returns 0 on success and returns GNTST_* on error + * (see xen/include/interface/grant_table.h). + */ +int xenbus_unmap_ring_vfree(struct xenbus_device *dev, void *vaddr) +{ + struct vm_struct *area; + struct gnttab_unmap_grant_ref op = { + .host_addr = (unsigned long)vaddr, + }; + + /* It'd be nice if linux/vmalloc.h provided a find_vm_area(void *addr) + * method so that we don't have to muck with vmalloc internals here. + * We could force the user to hang on to their struct vm_struct from + * xenbus_map_ring_valloc, but these 6 lines considerably simplify + * this API. + */ + read_lock(&vmlist_lock); + for (area = vmlist; area != NULL; area = area->next) { + if (area->addr == vaddr) + break; + } + read_unlock(&vmlist_lock); + + if (!area) { + xenbus_dev_error(dev, -ENOENT, + "can't find mapped virtual address %p", vaddr); + return GNTST_bad_virt_addr; + } + + op.handle = (grant_handle_t)area->phys_addr; + + if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1)) + BUG(); + + if (op.status == GNTST_okay) + xen_free_vm_area(area); + else + xenbus_dev_error(dev, op.status, + "unmapping page at handle %d error %d", + (int16_t)area->phys_addr, op.status); + + return op.status; +} +EXPORT_SYMBOL_GPL(xenbus_unmap_ring_vfree); + + +/** + * xenbus_unmap_ring + * @dev: xenbus device + * @handle: grant handle + * @vaddr: addr to unmap + * + * Unmap a page of memory in this domain that was imported from another domain. + * Returns 0 on success and returns GNTST_* on error + * (see xen/include/interface/grant_table.h). + */ +int xenbus_unmap_ring(struct xenbus_device *dev, + grant_handle_t handle, void *vaddr) +{ + struct gnttab_unmap_grant_ref op = { + .host_addr = (unsigned long)vaddr, + .handle = handle, + }; + + if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1)) + BUG(); + + if (op.status != GNTST_okay) + xenbus_dev_error(dev, op.status, + "unmapping page at handle %d error %d", + handle, op.status); + + return op.status; +} +EXPORT_SYMBOL_GPL(xenbus_unmap_ring); +#endif + + +/** + * xenbus_read_driver_state + * @path: path for driver + * + * Return the state of the driver rooted at the given store path, or + * XenbusStateUnknown if no state can be read. + */ +enum xenbus_state xenbus_read_driver_state(const char *path) +{ + enum xenbus_state result; + int err = xenbus_gather(XBT_NIL, path, "state", "%d", &result, NULL); + if (err) + result = XenbusStateUnknown; + + return result; +} +EXPORT_SYMBOL_GPL(xenbus_read_driver_state); diff --git a/drivers/xenpvdrivers/xenbus_comms.c b/drivers/xenpvdrivers/xenbus_comms.c new file mode 100644 index 0000000..6ced74e --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_comms.c @@ -0,0 +1,281 @@ +/****************************************************************************** + * xenbus_comms.c + * + * Low level code to talks to Xen Store: ringbuffer and event channel. + * + * Copyright (C) 2005 Rusty Russell, IBM Corporation + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +#include +#include +#else +#include +#include +#include +#endif + +#include "xenbus_comms.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +static int xenbus_irq; + +extern void xenbus_probe(struct work_struct *); +static DECLARE_WORK(probe_work, xenbus_probe); + +static DECLARE_WAIT_QUEUE_HEAD(xb_waitq); + +static irqreturn_t wake_waiting(int irq, void *unused) +{ + int old, new; + + old = atomic_read(&xenbus_xsd_state); + switch (old) { + case XENBUS_XSD_UNCOMMITTED: + BUG(); + return IRQ_HANDLED; + + case XENBUS_XSD_FOREIGN_INIT: + new = XENBUS_XSD_FOREIGN_READY; + break; + + case XENBUS_XSD_LOCAL_INIT: + new = XENBUS_XSD_LOCAL_READY; + break; + + case XENBUS_XSD_FOREIGN_READY: + case XENBUS_XSD_LOCAL_READY: + default: + goto wake; + } + + old = atomic_cmpxchg(&xenbus_xsd_state, old, new); + if (old != new) + schedule_work(&probe_work); + +wake: + wake_up(&xb_waitq); + return IRQ_HANDLED; +} + +static int check_indexes(XENSTORE_RING_IDX cons, XENSTORE_RING_IDX prod) +{ + return ((prod - cons) <= XENSTORE_RING_SIZE); +} + +static void *get_output_chunk(XENSTORE_RING_IDX cons, + XENSTORE_RING_IDX prod, + char *buf, uint32_t *len) +{ + *len = XENSTORE_RING_SIZE - MASK_XENSTORE_IDX(prod); + if ((XENSTORE_RING_SIZE - (prod - cons)) < *len) + *len = XENSTORE_RING_SIZE - (prod - cons); + return buf + MASK_XENSTORE_IDX(prod); +} + +static const void *get_input_chunk(XENSTORE_RING_IDX cons, + XENSTORE_RING_IDX prod, + const char *buf, uint32_t *len) +{ + *len = XENSTORE_RING_SIZE - MASK_XENSTORE_IDX(cons); + if ((prod - cons) < *len) + *len = prod - cons; + return buf + MASK_XENSTORE_IDX(cons); +} + +/** + * xb_write - low level write + * @data: buffer to send + * @len: length of buffer + * + * Returns 0 on success, error otherwise. + */ +int xb_write(const void *data, unsigned len) +{ + struct xenstore_domain_interface *intf = xen_store_interface; + XENSTORE_RING_IDX cons, prod; + int rc; + + while (len != 0) { + void *dst; + unsigned int avail; + + rc = wait_event_interruptible( + xb_waitq, + (intf->req_prod - intf->req_cons) != + XENSTORE_RING_SIZE); + if (rc < 0) + return rc; + + /* Read indexes, then verify. */ + cons = intf->req_cons; + prod = intf->req_prod; + if (!check_indexes(cons, prod)) { + intf->req_cons = intf->req_prod = 0; + return -EIO; + } + + dst = get_output_chunk(cons, prod, intf->req, &avail); + if (avail == 0) + continue; + if (avail > len) + avail = len; + + /* Must write data /after/ reading the consumer index. */ + mb(); + + memcpy(dst, data, avail); + data += avail; + len -= avail; + + /* Other side must not see new producer until data is there. */ + wmb(); + intf->req_prod += avail; + + /* Implies mb(): other side will see the updated producer. */ + notify_remote_via_evtchn(xen_store_evtchn); + } + + return 0; +} + +int xb_data_to_read(void) +{ + struct xenstore_domain_interface *intf = xen_store_interface; + return (intf->rsp_cons != intf->rsp_prod); +} + +int xb_wait_for_data_to_read(void) +{ + return wait_event_interruptible(xb_waitq, xb_data_to_read()); +} + +int xb_read(void *data, unsigned len) +{ + struct xenstore_domain_interface *intf = xen_store_interface; + XENSTORE_RING_IDX cons, prod; + int rc; + + while (len != 0) { + unsigned int avail; + const char *src; + + rc = xb_wait_for_data_to_read(); + if (rc < 0) + return rc; + + /* Read indexes, then verify. */ + cons = intf->rsp_cons; + prod = intf->rsp_prod; + if (!check_indexes(cons, prod)) { + intf->rsp_cons = intf->rsp_prod = 0; + return -EIO; + } + + src = get_input_chunk(cons, prod, intf->rsp, &avail); + if (avail == 0) + continue; + if (avail > len) + avail = len; + + /* Must read data /after/ reading the producer index. */ + rmb(); + + memcpy(data, src, avail); + data += avail; + len -= avail; + + /* Other side must not see free space until we've copied out */ + mb(); + intf->rsp_cons += avail; + + pr_debug("Finished read of %i bytes (%i to go)\n", avail, len); + + /* Implies mb(): other side will see the updated consumer. */ + notify_remote_via_evtchn(xen_store_evtchn); + } + + return 0; +} + +/** + * xb_init_comms - Set up interrupt handler off store event channel. + */ +int xb_init_comms(void) +{ + struct xenstore_domain_interface *intf = xen_store_interface; + int err; + + if (intf->req_prod != intf->req_cons) + printk(KERN_ERR "XENBUS request ring is not quiescent " + "(%08x:%08x)!\n", intf->req_cons, intf->req_prod); + + if (intf->rsp_prod != intf->rsp_cons) { + printk(KERN_WARNING "XENBUS response ring is not quiescent " + "(%08x:%08x): fixing up\n", + intf->rsp_cons, intf->rsp_prod); + intf->rsp_cons = intf->rsp_prod; + } + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + if (xenbus_irq) + unbind_from_irqhandler(xenbus_irq, &xb_waitq); + + err = bind_caller_port_to_irqhandler( + xen_store_evtchn, wake_waiting, + 0, "xenbus", &xb_waitq); + if (err <= 0) { + printk(KERN_ERR "XENBUS request irq failed %i\n", err); + return err; + } + + xenbus_irq = err; +#else + if (xenbus_irq) { + /* Already have an irq; assume we're resuming */ + rebind_evtchn_irq(xen_store_evtchn, xenbus_irq); + } else { + err = bind_evtchn_to_irqhandler(xen_store_evtchn, wake_waiting, + 0, "xenbus", &xb_waitq); + if (err <= 0) { + printk(KERN_ERR "XENBUS request irq failed %i\n", err); + return err; + } + xenbus_irq = err; + } +#endif + + return 0; +} diff --git a/drivers/xenpvdrivers/xenbus_comms.h b/drivers/xenpvdrivers/xenbus_comms.h new file mode 100644 index 0000000..476cf89 --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_comms.h @@ -0,0 +1,62 @@ +/* + * Private include for xenbus communications. + * + * Copyright (C) 2005 Rusty Russell, IBM Corporation + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef _XENBUS_COMMS_H +#define _XENBUS_COMMS_H + +int xs_init(void); +int xb_init_comms(void); + +/* Low level routines. */ +int xb_write(const void *data, unsigned len); +int xb_read(void *data, unsigned len); +int xb_data_to_read(void); +int xb_wait_for_data_to_read(void); +int xs_input_avail(void); +extern struct xenstore_domain_interface *xen_store_interface; +extern int xen_store_evtchn; + +/* For xenbus internal use. */ +enum { + XENBUS_XSD_UNCOMMITTED = 0, + XENBUS_XSD_FOREIGN_INIT, + XENBUS_XSD_FOREIGN_READY, + XENBUS_XSD_LOCAL_INIT, + XENBUS_XSD_LOCAL_READY, +}; +extern atomic_t xenbus_xsd_state; + +static inline int is_xenstored_ready(void) +{ + int s = atomic_read(&xenbus_xsd_state); + return s == XENBUS_XSD_FOREIGN_READY || s == XENBUS_XSD_LOCAL_READY; +} + +#endif /* _XENBUS_COMMS_H */ diff --git a/drivers/xenpvdrivers/xenbus_dev.c b/drivers/xenpvdrivers/xenbus_dev.c new file mode 100644 index 0000000..f140a9a --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_dev.c @@ -0,0 +1,461 @@ +/* + * xenbus_dev.c + * + * Driver giving user-space access to the kernel's xenbus connection + * to xenstore. + * + * Copyright (c) 2005, Christian Limpach + * Copyright (c) 2005, Rusty Russell, IBM Corporation + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "xenbus_comms.h" + +#include +#include +#include +#include +#include + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#include + +struct xenbus_dev_transaction { + struct list_head list; + struct xenbus_transaction handle; +}; + +struct read_buffer { + struct list_head list; + unsigned int cons; + unsigned int len; + char msg[]; +}; + +struct xenbus_dev_data { + /* In-progress transaction. */ + struct list_head transactions; + + /* Active watches. */ + struct list_head watches; + + /* Partial request. */ + unsigned int len; + union { + struct xsd_sockmsg msg; + char buffer[PAGE_SIZE]; + } u; + + /* Response queue. */ + struct list_head read_buffers; + wait_queue_head_t read_waitq; + + struct mutex reply_mutex; +}; + +static struct proc_dir_entry *xenbus_dev_intf; + +static ssize_t xenbus_dev_read(struct file *filp, + char __user *ubuf, + size_t len, loff_t *ppos) +{ + struct xenbus_dev_data *u = filp->private_data; + struct read_buffer *rb; + int i, ret; + + if (!is_xenstored_ready()) + return -ENODEV; + + mutex_lock(&u->reply_mutex); + while (list_empty(&u->read_buffers)) { + mutex_unlock(&u->reply_mutex); + ret = wait_event_interruptible(u->read_waitq, + !list_empty(&u->read_buffers)); + if (ret) + return ret; + mutex_lock(&u->reply_mutex); + } + + rb = list_entry(u->read_buffers.next, struct read_buffer, list); + for (i = 0; i < len;) { + put_user(rb->msg[rb->cons], ubuf + i); + i++; + rb->cons++; + if (rb->cons == rb->len) { + list_del(&rb->list); + kfree(rb); + if (list_empty(&u->read_buffers)) + break; + rb = list_entry(u->read_buffers.next, + struct read_buffer, list); + } + } + mutex_unlock(&u->reply_mutex); + + return i; +} + +static void queue_reply(struct xenbus_dev_data *u, + char *data, unsigned int len) +{ + struct read_buffer *rb; + + if (len == 0) + return; + + rb = kmalloc(sizeof(*rb) + len, GFP_KERNEL); + BUG_ON(rb == NULL); + + rb->cons = 0; + rb->len = len; + + memcpy(rb->msg, data, len); + + list_add_tail(&rb->list, &u->read_buffers); + + wake_up(&u->read_waitq); +} + +struct watch_adapter +{ + struct list_head list; + struct xenbus_watch watch; + struct xenbus_dev_data *dev_data; + char *token; +}; + +static void free_watch_adapter (struct watch_adapter *watch) +{ + kfree(watch->watch.node); + kfree(watch->token); + kfree(watch); +} + +static void watch_fired(struct xenbus_watch *watch, + const char **vec, + unsigned int len) +{ + struct watch_adapter *adap = + container_of(watch, struct watch_adapter, watch); + struct xsd_sockmsg hdr; + const char *path, *token; + int path_len, tok_len, body_len, data_len = 0; + + path = vec[XS_WATCH_PATH]; + token = adap->token; + + path_len = strlen(path) + 1; + tok_len = strlen(token) + 1; + if (len > 2) + data_len = vec[len] - vec[2] + 1; + body_len = path_len + tok_len + data_len; + + hdr.type = XS_WATCH_EVENT; + hdr.len = body_len; + + mutex_lock(&adap->dev_data->reply_mutex); + queue_reply(adap->dev_data, (char *)&hdr, sizeof(hdr)); + queue_reply(adap->dev_data, (char *)path, path_len); + queue_reply(adap->dev_data, (char *)token, tok_len); + if (len > 2) + queue_reply(adap->dev_data, (char *)vec[2], data_len); + mutex_unlock(&adap->dev_data->reply_mutex); +} + +static LIST_HEAD(watch_list); + +static ssize_t xenbus_dev_write(struct file *filp, + const char __user *ubuf, + size_t len, loff_t *ppos) +{ + struct xenbus_dev_data *u = filp->private_data; + struct xenbus_dev_transaction *trans = NULL; + uint32_t msg_type; + void *reply; + char *path, *token; + struct watch_adapter *watch, *tmp_watch; + int err, rc = len; + + if (!is_xenstored_ready()) + return -ENODEV; + + if ((len + u->len) > sizeof(u->u.buffer)) { + rc = -EINVAL; + goto out; + } + + if (copy_from_user(u->u.buffer + u->len, ubuf, len) != 0) { + rc = -EFAULT; + goto out; + } + + u->len += len; + if ((u->len < sizeof(u->u.msg)) || + (u->len < (sizeof(u->u.msg) + u->u.msg.len))) + return rc; + + msg_type = u->u.msg.type; + + switch (msg_type) { + case XS_WATCH: + case XS_UNWATCH: { + static const char *XS_RESP = "OK"; + struct xsd_sockmsg hdr; + + path = u->u.buffer + sizeof(u->u.msg); + token = memchr(path, 0, u->u.msg.len); + if (token == NULL) { + rc = -EILSEQ; + goto out; + } + token++; + + if (msg_type == XS_WATCH) { + watch = kzalloc(sizeof(*watch), GFP_KERNEL); + watch->watch.node = kmalloc(strlen(path)+1, + GFP_KERNEL); + strcpy((char *)watch->watch.node, path); + watch->watch.callback = watch_fired; + watch->token = kmalloc(strlen(token)+1, GFP_KERNEL); + strcpy(watch->token, token); + watch->dev_data = u; + + err = register_xenbus_watch(&watch->watch); + if (err) { + free_watch_adapter(watch); + rc = err; + goto out; + } + + list_add(&watch->list, &u->watches); + } else { + list_for_each_entry_safe(watch, tmp_watch, + &u->watches, list) { + if (!strcmp(watch->token, token) && + !strcmp(watch->watch.node, path)) + { + unregister_xenbus_watch(&watch->watch); + list_del(&watch->list); + free_watch_adapter(watch); + break; + } + } + } + + hdr.type = msg_type; + hdr.len = strlen(XS_RESP) + 1; + mutex_lock(&u->reply_mutex); + queue_reply(u, (char *)&hdr, sizeof(hdr)); + queue_reply(u, (char *)XS_RESP, hdr.len); + mutex_unlock(&u->reply_mutex); + break; + } + + default: + if (msg_type == XS_TRANSACTION_START) { + trans = kmalloc(sizeof(*trans), GFP_KERNEL); + if (!trans) { + rc = -ENOMEM; + goto out; + } + } + + reply = xenbus_dev_request_and_reply(&u->u.msg); + if (IS_ERR(reply)) { + kfree(trans); + rc = PTR_ERR(reply); + goto out; + } + + if (msg_type == XS_TRANSACTION_START) { + trans->handle.id = simple_strtoul(reply, NULL, 0); + list_add(&trans->list, &u->transactions); + } else if (msg_type == XS_TRANSACTION_END) { + list_for_each_entry(trans, &u->transactions, list) + if (trans->handle.id == u->u.msg.tx_id) + break; + BUG_ON(&trans->list == &u->transactions); + list_del(&trans->list); + kfree(trans); + } + mutex_lock(&u->reply_mutex); + queue_reply(u, (char *)&u->u.msg, sizeof(u->u.msg)); + queue_reply(u, (char *)reply, u->u.msg.len); + mutex_unlock(&u->reply_mutex); + kfree(reply); + break; + } + + out: + u->len = 0; + return rc; +} + +static int xenbus_dev_open(struct inode *inode, struct file *filp) +{ + struct xenbus_dev_data *u; + + if (xen_store_evtchn == 0) + return -ENOENT; + + nonseekable_open(inode, filp); + + u = kzalloc(sizeof(*u), GFP_KERNEL); + if (u == NULL) + return -ENOMEM; + + INIT_LIST_HEAD(&u->transactions); + INIT_LIST_HEAD(&u->watches); + INIT_LIST_HEAD(&u->read_buffers); + init_waitqueue_head(&u->read_waitq); + + mutex_init(&u->reply_mutex); + + filp->private_data = u; + + return 0; +} + +static int xenbus_dev_release(struct inode *inode, struct file *filp) +{ + struct xenbus_dev_data *u = filp->private_data; + struct xenbus_dev_transaction *trans, *tmp; + struct watch_adapter *watch, *tmp_watch; + + list_for_each_entry_safe(trans, tmp, &u->transactions, list) { + xenbus_transaction_end(trans->handle, 1); + list_del(&trans->list); + kfree(trans); + } + + list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) { + unregister_xenbus_watch(&watch->watch); + list_del(&watch->list); + free_watch_adapter(watch); + } + + kfree(u); + + return 0; +} + +static unsigned int xenbus_dev_poll(struct file *file, poll_table *wait) +{ + struct xenbus_dev_data *u = file->private_data; + + if (!is_xenstored_ready()) + return -ENODEV; + + poll_wait(file, &u->read_waitq, wait); + if (!list_empty(&u->read_buffers)) + return POLLIN | POLLRDNORM; + return 0; +} + +#ifdef HAVE_UNLOCKED_IOCTL +static long xenbus_dev_ioctl(struct file *file, + unsigned int cmd, unsigned long data) +{ + extern int xenbus_conn(domid_t remote_dom, int *grant_ref, + evtchn_port_t *local_port); + void __user *udata = (void __user *) data; + int ret = -ENOTTY; + + if (!is_initial_xendomain()) + return -ENODEV; + + + switch (cmd) { + case IOCTL_XENBUS_ALLOC: { + xenbus_alloc_t xa; + int old; + + old = atomic_cmpxchg(&xenbus_xsd_state, + XENBUS_XSD_UNCOMMITTED, + XENBUS_XSD_FOREIGN_INIT); + if (old != XENBUS_XSD_UNCOMMITTED) + return -EBUSY; + + if (copy_from_user(&xa, udata, sizeof(xa))) { + ret = -EFAULT; + atomic_set(&xenbus_xsd_state, XENBUS_XSD_UNCOMMITTED); + break; + } + + ret = xenbus_conn(xa.dom, &xa.grant_ref, &xa.port); + if (ret != 0) { + atomic_set(&xenbus_xsd_state, XENBUS_XSD_UNCOMMITTED); + break; + } + + if (copy_to_user(udata, &xa, sizeof(xa))) { + ret = -EFAULT; + atomic_set(&xenbus_xsd_state, XENBUS_XSD_UNCOMMITTED); + break; + } + } + break; + + default: + break; + } + + return ret; +} +#endif + +static const struct file_operations xenbus_dev_file_ops = { + .read = xenbus_dev_read, + .write = xenbus_dev_write, + .open = xenbus_dev_open, + .release = xenbus_dev_release, + .poll = xenbus_dev_poll, +#ifdef HAVE_UNLOCKED_IOCTL + .unlocked_ioctl = xenbus_dev_ioctl +#endif +}; + +int xenbus_dev_init(void) +{ + xenbus_dev_intf = create_xen_proc_entry("xenbus", 0400); + if (xenbus_dev_intf) + xenbus_dev_intf->proc_fops = &xenbus_dev_file_ops; + + return 0; +} diff --git a/drivers/xenpvdrivers/xenbus_probe.c b/drivers/xenpvdrivers/xenbus_probe.c new file mode 100644 index 0000000..593c808 --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_probe.c @@ -0,0 +1,1369 @@ +/****************************************************************************** + * Talks to Xen Store to figure out what devices we have. + * + * Copyright (C) 2005 Rusty Russell, IBM Corporation + * Copyright (C) 2005 Mike Wray, Hewlett-Packard + * Copyright (C) 2005, 2006 XenSource Ltd + * Copyright (C) 2007 Solarflare Communications, Inc. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#define DPRINTK(fmt, args...) \ + pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \ + __FUNCTION__, __LINE__, ##args) + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +#include +#include +#include +#include +#include +#include +#ifdef XENPV_COMPAT_MODULE +#include +#endif +#else +#include +#include +#include +#include +#endif + +#include "xenbus_comms.h" +#include "xenbus_probe.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +int xen_store_evtchn; +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) +EXPORT_SYMBOL(xen_store_evtchn); +#endif + +struct xenstore_domain_interface *xen_store_interface; +static unsigned long xen_store_mfn; + +extern struct mutex xenwatch_mutex; + +static BLOCKING_NOTIFIER_HEAD(xenstore_chain); + +static void wait_for_devices(struct xenbus_driver *xendrv); + +static int xenbus_probe_frontend(const char *type, const char *name); + +static void xenbus_dev_shutdown(struct device *_dev); + +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) +static int xenbus_dev_suspend(struct device *dev, pm_message_t state); +static int xenbus_dev_resume(struct device *dev); +#endif + +/* If something in array of ids matches this device, return it. */ +static const struct xenbus_device_id * +match_device(const struct xenbus_device_id *arr, struct xenbus_device *dev) +{ + for (; *arr->devicetype != '\0'; arr++) { + if (!strcmp(arr->devicetype, dev->devicetype)) + return arr; + } + return NULL; +} + +int xenbus_match(struct device *_dev, struct device_driver *_drv) +{ + struct xenbus_driver *drv = to_xenbus_driver(_drv); + + if (!drv->ids) + return 0; + + return match_device(drv->ids, to_xenbus_device(_dev)) != NULL; +} + +/* device// => - */ +static int frontend_bus_id(char bus_id[XEN_BUS_ID_SIZE], const char *nodename) +{ + nodename = strchr(nodename, '/'); + if (!nodename || strlen(nodename + 1) >= XEN_BUS_ID_SIZE) { + printk(KERN_WARNING "XENBUS: bad frontend %s\n", nodename); + return -EINVAL; + } + + strlcpy(bus_id, nodename + 1, XEN_BUS_ID_SIZE); + if (!strchr(bus_id, '/')) { + printk(KERN_WARNING "XENBUS: bus_id %s no slash\n", bus_id); + return -EINVAL; + } + *strchr(bus_id, '/') = '-'; + return 0; +} + + +static void free_otherend_details(struct xenbus_device *dev) +{ + kfree(dev->otherend); + dev->otherend = NULL; +} + + +static void free_otherend_watch(struct xenbus_device *dev) +{ + if (dev->otherend_watch.node) { + unregister_xenbus_watch(&dev->otherend_watch); + kfree(dev->otherend_watch.node); + dev->otherend_watch.node = NULL; + } +} + + +int read_otherend_details(struct xenbus_device *xendev, + char *id_node, char *path_node) +{ + int err = xenbus_gather(XBT_NIL, xendev->nodename, + id_node, "%i", &xendev->otherend_id, + path_node, NULL, &xendev->otherend, + NULL); + if (err) { + xenbus_dev_fatal(xendev, err, + "reading other end details from %s", + xendev->nodename); + return err; + } + if (strlen(xendev->otherend) == 0 || + !xenbus_exists(XBT_NIL, xendev->otherend, "")) { + xenbus_dev_fatal(xendev, -ENOENT, + "unable to read other end from %s. " + "missing or inaccessible.", + xendev->nodename); + free_otherend_details(xendev); + return -ENOENT; + } + + return 0; +} + + +static int read_backend_details(struct xenbus_device *xendev) +{ + return read_otherend_details(xendev, "backend-id", "backend"); +} + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16) +static int xenbus_uevent_frontend(struct device *dev, struct kobj_uevent_env *env) +{ + struct xenbus_device *xdev; + + if (dev == NULL) + return -ENODEV; + xdev = to_xenbus_device(dev); + if (xdev == NULL) + return -ENODEV; + + /* stuff we want to pass to /sbin/hotplug */ +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + add_uevent_var(env, "XENBUS_TYPE=%s", xdev->devicetype); + add_uevent_var(env, "XENBUS_PATH=%s", xdev->nodename); +#endif + add_uevent_var(env, "MODALIAS=xen:%s", xdev->devicetype); + + return 0; +} +#endif + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,29) +static struct device_attribute xenbus_dev_attrs[] = { + __ATTR_NULL +}; +#endif + +/* Bus type for frontend drivers. */ +static struct xen_bus_type xenbus_frontend = { + .root = "device", + .levels = 2, /* device/type/ */ + .get_bus_id = frontend_bus_id, + .probe = xenbus_probe_frontend, + .error = -ENODEV, + .bus = { + .name = "xen", + .match = xenbus_match, +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16) + .probe = xenbus_dev_probe, + .remove = xenbus_dev_remove, + .shutdown = xenbus_dev_shutdown, + .uevent = xenbus_uevent_frontend, +#endif +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,29) + .dev_attrs = xenbus_dev_attrs, +#endif +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) + .suspend = xenbus_dev_suspend, + .resume = xenbus_dev_resume, +#endif + }, +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + .dev = { + .init_name = "xen", + }, +#endif +}; + +static void otherend_changed(struct xenbus_watch *watch, + const char **vec, unsigned int len) +{ + struct xenbus_device *dev = + container_of(watch, struct xenbus_device, otherend_watch); + struct xenbus_driver *drv = to_xenbus_driver(dev->dev.driver); + enum xenbus_state state; + + /* Protect us against watches firing on old details when the otherend + details change, say immediately after a resume. */ + if (!dev->otherend || + strncmp(dev->otherend, vec[XS_WATCH_PATH], + strlen(dev->otherend))) { + dev_dbg(&dev->dev, "Ignoring watch at %s", vec[XS_WATCH_PATH]); + return; + } + + state = xenbus_read_driver_state(dev->otherend); + + dev_dbg(&dev->dev, "state is %d (%s), %s, %s", + state, xenbus_strstate(state), dev->otherend_watch.node, + vec[XS_WATCH_PATH]); + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16) + /* + * Ignore xenbus transitions during shutdown. This prevents us doing + * work that can fail e.g., when the rootfs is gone. + */ + if (system_state > SYSTEM_RUNNING) { + struct xen_bus_type *bus = bus; + bus = container_of(dev->dev.bus, struct xen_bus_type, bus); + /* If we're frontend, drive the state machine to Closed. */ + /* This should cause the backend to release our resources. */ + if ((bus == &xenbus_frontend) && (state == XenbusStateClosing)) + xenbus_frontend_closed(dev); + return; + } +#endif + + if (drv->otherend_changed) + drv->otherend_changed(dev, state); +} + + +static int talk_to_otherend(struct xenbus_device *dev) +{ + struct xenbus_driver *drv = to_xenbus_driver(dev->dev.driver); + + free_otherend_watch(dev); + free_otherend_details(dev); + + return drv->read_otherend_details(dev); +} + + +static int watch_otherend(struct xenbus_device *dev) +{ +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + return xenbus_watch_path2(dev, dev->otherend, "state", + &dev->otherend_watch, otherend_changed); +#else + return xenbus_watch_pathfmt(dev, &dev->otherend_watch, otherend_changed, + "%s/%s", dev->otherend, "state"); +#endif +} + + +int xenbus_dev_probe(struct device *_dev) +{ + struct xenbus_device *dev = to_xenbus_device(_dev); + struct xenbus_driver *drv = to_xenbus_driver(_dev->driver); + const struct xenbus_device_id *id; + int err; + + DPRINTK("%s", dev->nodename); + + if (!drv->probe) { + err = -ENODEV; + goto fail; + } + + id = match_device(drv->ids, dev); + if (!id) { + err = -ENODEV; + goto fail; + } + + err = talk_to_otherend(dev); + if (err) { + dev_warn(&dev->dev, + "xenbus_probe: talk_to_otherend on %s failed.\n", + dev->nodename); + return err; + } + + err = drv->probe(dev, id); + if (err) + goto fail; + + err = watch_otherend(dev); + if (err) { + dev_warn(&dev->dev, + "xenbus_probe: watch_otherend on %s failed.\n", + dev->nodename); + return err; + } + + return 0; +fail: + xenbus_dev_error(dev, err, "xenbus_dev_probe on %s", dev->nodename); + xenbus_switch_state(dev, XenbusStateClosed); + return -ENODEV; +} + +int xenbus_dev_remove(struct device *_dev) +{ + struct xenbus_device *dev = to_xenbus_device(_dev); + struct xenbus_driver *drv = to_xenbus_driver(_dev->driver); + + DPRINTK("%s", dev->nodename); + + free_otherend_watch(dev); + free_otherend_details(dev); + + if (drv->remove) + drv->remove(dev); + + xenbus_switch_state(dev, XenbusStateClosed); + return 0; +} + +static void xenbus_dev_shutdown(struct device *_dev) +{ + struct xenbus_device *dev = to_xenbus_device(_dev); + unsigned long timeout = 5*HZ; + + DPRINTK("%s", dev->nodename); + +/* Commented out since xenstored stubdom is now minios based not linux based +#define XENSTORE_DOMAIN_SHARES_THIS_KERNEL +*/ +#ifndef XENSTORE_DOMAIN_SHARES_THIS_KERNEL + if (is_initial_xendomain()) +#endif + return; + + get_device(&dev->dev); + if (dev->state != XenbusStateConnected) { + dev_info(&dev->dev, "%s: %s: %s != Connected, skipping\n", __FUNCTION__, + dev->nodename, xenbus_strstate(dev->state)); + goto out; + } + xenbus_switch_state(dev, XenbusStateClosing); + + if (!strcmp(dev->devicetype, "vfb")) + goto out; + + timeout = wait_for_completion_timeout(&dev->down, timeout); + if (!timeout) + dev_info(&dev->dev, "%s: %s timeout closing device\n", + __FUNCTION__, dev->nodename); + out: + put_device(&dev->dev); +} + +int xenbus_register_driver_common(struct xenbus_driver *drv, + struct xen_bus_type *bus, + struct module *owner, + const char *mod_name) +{ + int ret; + + if (bus->error) + return bus->error; + + drv->driver.name = drv->name; + drv->driver.bus = &bus->bus; +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,10) + drv->driver.owner = owner; +#endif +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,21) + drv->driver.mod_name = mod_name; +#endif +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16) + drv->driver.probe = xenbus_dev_probe; + drv->driver.remove = xenbus_dev_remove; + drv->driver.shutdown = xenbus_dev_shutdown; +#endif + + mutex_lock(&xenwatch_mutex); + ret = driver_register(&drv->driver); + mutex_unlock(&xenwatch_mutex); + return ret; +} + +int __xenbus_register_frontend(struct xenbus_driver *drv, + struct module *owner, const char *mod_name) +{ + int ret; + + drv->read_otherend_details = read_backend_details; + + ret = xenbus_register_driver_common(drv, &xenbus_frontend, + owner, mod_name); + if (ret) + return ret; + + /* If this driver is loaded as a module wait for devices to attach. */ + wait_for_devices(drv); + + return 0; +} +EXPORT_SYMBOL_GPL(__xenbus_register_frontend); + +void xenbus_unregister_driver(struct xenbus_driver *drv) +{ + driver_unregister(&drv->driver); +} +EXPORT_SYMBOL_GPL(xenbus_unregister_driver); + +struct xb_find_info +{ + struct xenbus_device *dev; + const char *nodename; +}; + +static int cmp_dev(struct device *dev, void *data) +{ + struct xenbus_device *xendev = to_xenbus_device(dev); + struct xb_find_info *info = data; + + if (!strcmp(xendev->nodename, info->nodename)) { + info->dev = xendev; + get_device(dev); + return 1; + } + return 0; +} + +struct xenbus_device *xenbus_device_find(const char *nodename, + struct bus_type *bus) +{ + struct xb_find_info info = { .dev = NULL, .nodename = nodename }; + + bus_for_each_dev(bus, NULL, &info, cmp_dev); + return info.dev; +} + +static int cleanup_dev(struct device *dev, void *data) +{ + struct xenbus_device *xendev = to_xenbus_device(dev); + struct xb_find_info *info = data; + int len = strlen(info->nodename); + + DPRINTK("%s", info->nodename); + + /* Match the info->nodename path, or any subdirectory of that path. */ + if (strncmp(xendev->nodename, info->nodename, len)) + return 0; + + /* If the node name is longer, ensure it really is a subdirectory. */ + if ((strlen(xendev->nodename) > len) && (xendev->nodename[len] != '/')) + return 0; + + info->dev = xendev; + get_device(dev); + return 1; +} + +static void xenbus_cleanup_devices(const char *path, struct bus_type *bus) +{ + struct xb_find_info info = { .nodename = path }; + + do { + info.dev = NULL; + bus_for_each_dev(bus, NULL, &info, cleanup_dev); + if (info.dev) { + device_unregister(&info.dev->dev); + put_device(&info.dev->dev); + } + } while (info.dev); +} + +static void xenbus_dev_release(struct device *dev) +{ + if (dev) + kfree(to_xenbus_device(dev)); +} + +static ssize_t xendev_show_nodename(struct device *dev, +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13) + struct device_attribute *attr, +#endif + char *buf) +{ + return sprintf(buf, "%s\n", to_xenbus_device(dev)->nodename); +} +static DEVICE_ATTR(nodename, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_nodename, NULL); + +static ssize_t xendev_show_devtype(struct device *dev, +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13) + struct device_attribute *attr, +#endif + char *buf) +{ + return sprintf(buf, "%s\n", to_xenbus_device(dev)->devicetype); +} +static DEVICE_ATTR(devtype, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_devtype, NULL); + +static ssize_t xendev_show_modalias(struct device *dev, +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13) + struct device_attribute *attr, +#endif + char *buf) +{ + return sprintf(buf, "xen:%s\n", to_xenbus_device(dev)->devicetype); +} +static DEVICE_ATTR(modalias, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_modalias, NULL); + +int xenbus_probe_node(struct xen_bus_type *bus, + const char *type, + const char *nodename) +{ + int err; + struct xenbus_device *xendev; + size_t stringlen; + char *tmpstring; + + enum xenbus_state state = xenbus_read_driver_state(nodename); + + if (bus->error) + return bus->error; + + if (state != XenbusStateInitialising) { + /* Device is not new, so ignore it. This can happen if a + device is going away after switching to Closed. */ + return 0; + } + + stringlen = strlen(nodename) + 1 + strlen(type) + 1; + xendev = kzalloc(sizeof(*xendev) + stringlen, GFP_KERNEL); + if (!xendev) + return -ENOMEM; + + xendev->state = XenbusStateInitialising; + + /* Copy the strings into the extra space. */ + + tmpstring = (char *)(xendev + 1); + strcpy(tmpstring, nodename); + xendev->nodename = tmpstring; + + tmpstring += strlen(tmpstring) + 1; + strcpy(tmpstring, type); + xendev->devicetype = tmpstring; + init_completion(&xendev->down); + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + xendev->dev.parent = &bus->dev; +#endif + xendev->dev.bus = &bus->bus; + xendev->dev.release = xenbus_dev_release; + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,26) + { + char devname[XEN_BUS_ID_SIZE]; + + err = bus->get_bus_id(devname, xendev->nodename); + if (!err) + dev_set_name(&xendev->dev, devname); + } +#else + err = bus->get_bus_id(xendev->dev.bus_id, xendev->nodename); +#endif + if (err) + goto fail; + + /* Register with generic device framework. */ + err = device_register(&xendev->dev); + if (err) + goto fail; + + err = device_create_file(&xendev->dev, &dev_attr_nodename); + if (err) + goto fail_unregister; + + err = device_create_file(&xendev->dev, &dev_attr_devtype); + if (err) + goto fail_remove_nodename; + + err = device_create_file(&xendev->dev, &dev_attr_modalias); + if (err) + goto fail_remove_devtype; + + return 0; +fail_remove_devtype: + device_remove_file(&xendev->dev, &dev_attr_devtype); +fail_remove_nodename: + device_remove_file(&xendev->dev, &dev_attr_nodename); +fail_unregister: + device_unregister(&xendev->dev); +fail: + kfree(xendev); + return err; +} + +/* device// */ +static int xenbus_probe_frontend(const char *type, const char *name) +{ + char *nodename; + int err; + + if (!strcmp(type, "console")) + return 0; + + nodename = kasprintf(GFP_KERNEL, "%s/%s/%s", + xenbus_frontend.root, type, name); + if (!nodename) + return -ENOMEM; + + DPRINTK("%s", nodename); + + err = xenbus_probe_node(&xenbus_frontend, type, nodename); + kfree(nodename); + return err; +} + +static int xenbus_probe_device_type(struct xen_bus_type *bus, const char *type) +{ + int err = 0; + char **dir; + unsigned int dir_n = 0; + int i; + + dir = xenbus_directory(XBT_NIL, bus->root, type, &dir_n); + if (IS_ERR(dir)) + return PTR_ERR(dir); + + for (i = 0; i < dir_n; i++) { + err = bus->probe(type, dir[i]); + if (err) + break; + } + kfree(dir); + return err; +} + +int xenbus_probe_devices(struct xen_bus_type *bus) +{ + int err = 0; + char **dir; + unsigned int i, dir_n; + + if (bus->error) + return bus->error; + + dir = xenbus_directory(XBT_NIL, bus->root, "", &dir_n); + if (IS_ERR(dir)) + return PTR_ERR(dir); + + for (i = 0; i < dir_n; i++) { + err = xenbus_probe_device_type(bus, dir[i]); + if (err) + break; + } + kfree(dir); + return err; +} + +static unsigned int char_count(const char *str, char c) +{ + unsigned int i, ret = 0; + + for (i = 0; str[i]; i++) + if (str[i] == c) + ret++; + return ret; +} + +static int strsep_len(const char *str, char c, unsigned int len) +{ + unsigned int i; + + for (i = 0; str[i]; i++) + if (str[i] == c) { + if (len == 0) + return i; + len--; + } + return (len == 0) ? i : -ERANGE; +} + +void xenbus_dev_changed(const char *node, struct xen_bus_type *bus) +{ + int exists, rootlen; + struct xenbus_device *dev; + char type[XEN_BUS_ID_SIZE]; + const char *p, *root; + + if (bus->error || char_count(node, '/') < 2) + return; + + exists = xenbus_exists(XBT_NIL, node, ""); + if (!exists) { + xenbus_cleanup_devices(node, &bus->bus); + return; + } + + /* backend//... or device//... */ + p = strchr(node, '/') + 1; + snprintf(type, XEN_BUS_ID_SIZE, "%.*s", (int)strcspn(p, "/"), p); + type[XEN_BUS_ID_SIZE-1] = '\0'; + + rootlen = strsep_len(node, '/', bus->levels); + if (rootlen < 0) + return; + root = kasprintf(GFP_KERNEL, "%.*s", rootlen, node); + if (!root) + return; + + dev = xenbus_device_find(root, &bus->bus); + if (!dev) + xenbus_probe_node(bus, type, root); + else + put_device(&dev->dev); + + kfree(root); +} +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) +EXPORT_SYMBOL_GPL(xenbus_dev_changed); +#endif + +static void frontend_changed(struct xenbus_watch *watch, + const char **vec, unsigned int len) +{ + DPRINTK(""); + + xenbus_dev_changed(vec[XS_WATCH_PATH], &xenbus_frontend); +} + +/* We watch for devices appearing and vanishing. */ +static struct xenbus_watch fe_watch = { + .node = "device", + .callback = frontend_changed, +}; + +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) +static int xenbus_dev_suspend(struct device *dev, pm_message_t state) +#else +static int suspend_dev(struct device *dev, void *data) +#endif +{ + int err = 0; + struct xenbus_driver *drv; + struct xenbus_device *xdev; + + DPRINTK(""); + + if (dev->driver == NULL) + return 0; + drv = to_xenbus_driver(dev->driver); + xdev = container_of(dev, struct xenbus_device, dev); + if (drv->suspend) +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) + err = drv->suspend(xdev, state); +#else + err = drv->suspend(xdev); +#endif + if (err) + printk(KERN_WARNING + "xenbus: suspend %s failed: %i\n", dev_name(dev), err); + return 0; +} + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +static int suspend_cancel_dev(struct device *dev, void *data) +{ + int err = 0; + struct xenbus_driver *drv; + struct xenbus_device *xdev; + + DPRINTK(""); + + if (dev->driver == NULL) + return 0; + drv = to_xenbus_driver(dev->driver); + xdev = container_of(dev, struct xenbus_device, dev); + if (drv->suspend_cancel) + err = drv->suspend_cancel(xdev); + if (err) + printk(KERN_WARNING + "xenbus: suspend_cancel %s failed: %i\n", + dev_name(dev), err); + return 0; +} +#endif + +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) +static int xenbus_dev_resume(struct device *dev) +#else +static int resume_dev(struct device *dev, void *data) +#endif +{ + int err; + struct xenbus_driver *drv; + struct xenbus_device *xdev; + + DPRINTK(""); + + if (dev->driver == NULL) + return 0; + + drv = to_xenbus_driver(dev->driver); + xdev = container_of(dev, struct xenbus_device, dev); + + err = talk_to_otherend(xdev); + if (err) { + printk(KERN_WARNING + "xenbus: resume (talk_to_otherend) %s failed: %i\n", + dev_name(dev), err); + return err; + } + + xdev->state = XenbusStateInitialising; + + if (drv->resume) { + err = drv->resume(xdev); + if (err) { + printk(KERN_WARNING + "xenbus: resume %s failed: %i\n", + dev_name(dev), err); + return err; + } + } + + err = watch_otherend(xdev); + if (err) { + printk(KERN_WARNING + "xenbus_probe: resume (watch_otherend) %s failed: " + "%d.\n", dev_name(dev), err); + return err; + } + + return 0; +} + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +void xenbus_suspend(void) +{ + DPRINTK(""); + + if (!xenbus_frontend.error) + bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_dev); + xenbus_backend_suspend(suspend_dev); + xs_suspend(); +} +EXPORT_SYMBOL_GPL(xenbus_suspend); + +void xenbus_resume(void) +{ + xb_init_comms(); + xs_resume(); + if (!xenbus_frontend.error) + bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, resume_dev); + xenbus_backend_resume(resume_dev); +} +EXPORT_SYMBOL_GPL(xenbus_resume); + +void xenbus_suspend_cancel(void) +{ + xs_suspend_cancel(); + if (!xenbus_frontend.error) + bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_cancel_dev); + xenbus_backend_resume(suspend_cancel_dev); +} +EXPORT_SYMBOL_GPL(xenbus_suspend_cancel); +#endif + +/* A flag to determine if xenstored is 'ready' (i.e. has started) */ +atomic_t xenbus_xsd_state = ATOMIC_INIT(XENBUS_XSD_UNCOMMITTED); + + +int register_xenstore_notifier(struct notifier_block *nb) +{ + int ret = 0; + + if (is_xenstored_ready()) + ret = nb->notifier_call(nb, 0, NULL); + else + blocking_notifier_chain_register(&xenstore_chain, nb); + + return ret; +} +EXPORT_SYMBOL_GPL(register_xenstore_notifier); + +void unregister_xenstore_notifier(struct notifier_block *nb) +{ + blocking_notifier_chain_unregister(&xenstore_chain, nb); +} +EXPORT_SYMBOL_GPL(unregister_xenstore_notifier); + + +void xenbus_probe(struct work_struct *unused) +{ + BUG_ON(!is_xenstored_ready()); + + /* Enumerate devices in xenstore and watch for changes. */ + xenbus_probe_devices(&xenbus_frontend); + register_xenbus_watch(&fe_watch); + xenbus_backend_probe_and_watch(); + + /* Notify others that xenstore is up */ + blocking_notifier_call_chain(&xenstore_chain, 0, NULL); +} + + +#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST) +static struct file_operations xsd_kva_fops; +static struct proc_dir_entry *xsd_kva_intf; +static struct proc_dir_entry *xsd_port_intf; + +static int xsd_kva_mmap(struct file *file, struct vm_area_struct *vma) +{ + size_t size = vma->vm_end - vma->vm_start; + int old; + int rc; + + old = atomic_cmpxchg(&xenbus_xsd_state, + XENBUS_XSD_UNCOMMITTED, + XENBUS_XSD_LOCAL_INIT); + switch (old) { + case XENBUS_XSD_UNCOMMITTED: + rc = xb_init_comms(); + if (rc != 0) + return rc; + break; + + case XENBUS_XSD_FOREIGN_INIT: + case XENBUS_XSD_FOREIGN_READY: + return -EBUSY; + + case XENBUS_XSD_LOCAL_INIT: + case XENBUS_XSD_LOCAL_READY: + default: + break; + } + + if ((size > PAGE_SIZE) || (vma->vm_pgoff != 0)) + return -EINVAL; + + if (remap_pfn_range(vma, vma->vm_start, mfn_to_pfn(xen_store_mfn), + size, vma->vm_page_prot)) + return -EAGAIN; + + return 0; +} + +static int xsd_kva_read(char *page, char **start, off_t off, + int count, int *eof, void *data) +{ + int len; + + len = sprintf(page, "0x%p", xen_store_interface); + *eof = 1; + return len; +} + +static int xsd_port_read(char *page, char **start, off_t off, + int count, int *eof, void *data) +{ + int len; + + len = sprintf(page, "%d", xen_store_evtchn); + *eof = 1; + return len; +} +#endif + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +static int xb_free_port(evtchn_port_t port) +{ + struct evtchn_close close; + close.port = port; + return HYPERVISOR_event_channel_op(EVTCHNOP_close, &close); +} + +int xenbus_conn(domid_t remote_dom, unsigned long *grant_ref, evtchn_port_t *local_port) +{ + struct evtchn_alloc_unbound alloc_unbound; + int rc, rc2; + + BUG_ON(atomic_read(&xenbus_xsd_state) != XENBUS_XSD_FOREIGN_INIT); + BUG_ON(!is_initial_xendomain()); + +#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST) + remove_xen_proc_entry("xsd_kva"); + remove_xen_proc_entry("xsd_port"); +#endif + + rc = xb_free_port(xen_store_evtchn); + if (rc != 0) + goto fail0; + + alloc_unbound.dom = DOMID_SELF; + alloc_unbound.remote_dom = remote_dom; + rc = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, + &alloc_unbound); + if (rc != 0) + goto fail0; + *local_port = xen_store_evtchn = alloc_unbound.port; + + /* keep the old page (xen_store_mfn, xen_store_interface) */ + rc = gnttab_grant_foreign_access(remote_dom, xen_store_mfn, + GTF_permit_access); + if (rc < 0) + goto fail1; + *grant_ref = rc; + + rc = xb_init_comms(); + if (rc != 0) + goto fail1; + + return 0; + +fail1: + rc2 = xb_free_port(xen_store_evtchn); + if (rc2 != 0) + printk(KERN_WARNING + "XENBUS: Error freeing xenstore event channel: %d\n", + rc2); +fail0: + xen_store_evtchn = -1; + return rc; +} +#endif + +#ifndef XENPV_COMPAT_MODULE +static int __init xenbus_probe_init(void) +#else +static int __devinit xenbus_probe_init(void) +#endif +{ + int err = 0; +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + unsigned long page = 0; +#endif + + DPRINTK(""); + + if (!is_running_on_xen()) + return -ENODEV; + + /* Register ourselves with the kernel bus subsystem */ + xenbus_frontend.error = bus_register(&xenbus_frontend.bus); + if (xenbus_frontend.error) + printk(KERN_WARNING + "XENBUS: Error registering frontend bus: %i\n", + xenbus_frontend.error); + xenbus_backend_bus_register(); + + /* + * Domain0 doesn't have a store_evtchn or store_mfn yet. + */ + if (is_initial_xendomain()) { +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + struct evtchn_alloc_unbound alloc_unbound; + + /* Allocate page. */ + page = get_zeroed_page(GFP_KERNEL); + if (!page) + return -ENOMEM; + + xen_store_mfn = xen_start_info->store_mfn = + pfn_to_mfn(virt_to_phys((void *)page) >> + PAGE_SHIFT); + + /* Next allocate a local port which xenstored can bind to */ + alloc_unbound.dom = DOMID_SELF; + alloc_unbound.remote_dom = DOMID_SELF; + + err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, + &alloc_unbound); + if (err == -ENOSYS) + goto err; + BUG_ON(err); + xen_store_evtchn = xen_start_info->store_evtchn = + alloc_unbound.port; + +#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST) + /* And finally publish the above info in /proc/xen */ + xsd_kva_intf = create_xen_proc_entry("xsd_kva", 0600); + if (xsd_kva_intf) { + memcpy(&xsd_kva_fops, xsd_kva_intf->proc_fops, + sizeof(xsd_kva_fops)); + xsd_kva_fops.mmap = xsd_kva_mmap; + xsd_kva_intf->proc_fops = &xsd_kva_fops; + xsd_kva_intf->read_proc = xsd_kva_read; + } + xsd_port_intf = create_xen_proc_entry("xsd_port", 0400); + if (xsd_port_intf) + xsd_port_intf->read_proc = xsd_port_read; +#endif +#else + /* dom0 not yet supported */ +#endif + xen_store_interface = mfn_to_virt(xen_store_mfn); + } else { + atomic_set(&xenbus_xsd_state, XENBUS_XSD_FOREIGN_READY); +#ifndef XENPV_COMPAT_MODULE + xen_store_evtchn = xen_start_info->store_evtchn; + xen_store_mfn = xen_start_info->store_mfn; + xen_store_interface = mfn_to_virt(xen_store_mfn); +#else + xen_store_evtchn = hvm_get_parameter(HVM_PARAM_STORE_EVTCHN); + xen_store_mfn = hvm_get_parameter(HVM_PARAM_STORE_PFN); + xen_store_interface = ioremap(xen_store_mfn << PAGE_SHIFT, + PAGE_SIZE); +#endif + /* Initialize the shared memory rings to talk to xenstored */ + err = xb_init_comms(); + if (err) + goto err; + } + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + xenbus_dev_init(); +#endif + + /* Initialize the interface to xenstore. */ + err = xs_init(); + if (err) { + printk(KERN_WARNING + "XENBUS: Error initializing xenstore comms: %i\n", err); + goto err; + } + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + /* Register ourselves with the kernel device subsystem */ + if (!xenbus_frontend.error) { + xenbus_frontend.error = device_register(&xenbus_frontend.dev); + if (xenbus_frontend.error) { + bus_unregister(&xenbus_frontend.bus); + printk(KERN_WARNING + "XENBUS: Error registering frontend device: %i\n", + xenbus_frontend.error); + } + } +#endif + xenbus_backend_device_register(); + + if (!is_initial_xendomain()) + xenbus_probe(NULL); + +#if defined(CONFIG_XEN_COMPAT_XENFS) && !defined(XENPV_COMPAT_MODULE) + /* + * Create xenfs mountpoint in /proc for compatibility with + * utilities that expect to find "xenbus" under "/proc/xen". + */ + proc_mkdir("xen", NULL); +#endif + + return 0; + + err: +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + if (page) + free_page(page); +#endif + + /* + * Do not unregister the xenbus front/backend buses here. The buses + * must exist because front/backend drivers will use them when they are + * registered. + */ + + return err; +} + +#ifndef XENPV_COMPAT_MODULE +postcore_initcall(xenbus_probe_init); +#ifdef CONFIG_XEN +MODULE_LICENSE("Dual BSD/GPL"); +#else +MODULE_LICENSE("GPL"); +#endif +#else +int __devinit xenbus_init(void) +{ + return xenbus_probe_init(); +} +#endif + +static int is_device_connecting(struct device *dev, void *data) +{ + struct xenbus_device *xendev = to_xenbus_device(dev); + struct device_driver *drv = data; + struct xenbus_driver *xendrv; + + /* + * A device with no driver will never connect. We care only about + * devices which should currently be in the process of connecting. + */ + if (!dev->driver) + return 0; + + /* Is this search limited to a particular driver? */ + if (drv && (dev->driver != drv)) + return 0; + + xendrv = to_xenbus_driver(dev->driver); + return (xendev->state < XenbusStateConnected || + (xendev->state == XenbusStateConnected && + xendrv->is_ready && !xendrv->is_ready(xendev))); +} + +static int exists_connecting_device(struct device_driver *drv) +{ + if (xenbus_frontend.error) + return xenbus_frontend.error; + return bus_for_each_dev(&xenbus_frontend.bus, NULL, drv, + is_device_connecting); +} + +static int print_device_status(struct device *dev, void *data) +{ + struct xenbus_device *xendev = to_xenbus_device(dev); + struct device_driver *drv = data; + struct xenbus_driver *xendrv; + + /* Is this operation limited to a particular driver? */ + if (drv && (dev->driver != drv)) + return 0; + + if (!dev->driver) { + /* Information only: is this too noisy? */ + printk(KERN_INFO "XENBUS: Device with no driver: %s\n", + xendev->nodename); + return 0; + } + + if (xendev->state < XenbusStateConnected) { + enum xenbus_state rstate = XenbusStateUnknown; + if (xendev->otherend) + rstate = xenbus_read_driver_state(xendev->otherend); + printk(KERN_WARNING "XENBUS: Timeout connecting " + "to device: %s (local state %d, remote state %d)\n", + xendev->nodename, xendev->state, rstate); + } + + xendrv = to_xenbus_driver(dev->driver); + if (xendrv->is_ready && !xendrv->is_ready(xendev)) + printk(KERN_WARNING "XENBUS: Device not ready: %s\n", + xendev->nodename); + + return 0; +} + +/* We only wait for device setup after most initcalls have run. */ +static int ready_to_wait_for_devices; + +/* + * On a 5-minute timeout, wait for all devices currently configured. We need + * to do this to guarantee that the filesystems and / or network devices + * needed for boot are available, before we can allow the boot to proceed. + * + * This needs to be on a late_initcall, to happen after the frontend device + * drivers have been initialised, but before the root fs is mounted. + * + * A possible improvement here would be to have the tools add a per-device + * flag to the store entry, indicating whether it is needed at boot time. + * This would allow people who knew what they were doing to accelerate their + * boot slightly, but of course needs tools or manual intervention to set up + * those flags correctly. + */ +static void wait_for_devices(struct xenbus_driver *xendrv) +{ + unsigned long start = jiffies; + struct device_driver *drv = xendrv ? &xendrv->driver : NULL; + unsigned int seconds_waited = 0; + + if (!ready_to_wait_for_devices || !is_running_on_xen()) + return; + + while (exists_connecting_device(drv)) { + if (time_after(jiffies, start + (seconds_waited+5)*HZ)) { + if (!seconds_waited) + printk(KERN_WARNING "XENBUS: Waiting for " + "devices to initialise: "); + seconds_waited += 5; + printk("%us...", 300 - seconds_waited); + if (seconds_waited == 300) + break; + } + + schedule_timeout_interruptible(HZ/10); + } + + if (seconds_waited) + printk("\n"); + + bus_for_each_dev(&xenbus_frontend.bus, NULL, drv, + print_device_status); +} + +#ifndef XENPV_COMPAT_MODULE +static int __init boot_wait_for_devices(void) +{ + if (!xenbus_frontend.error) { + ready_to_wait_for_devices = 1; + wait_for_devices(NULL); + } + return 0; +} + +late_initcall(boot_wait_for_devices); +#endif + +int xenbus_for_each_frontend(void *arg, int (*fn)(struct device *, void *)) +{ + return bus_for_each_dev(&xenbus_frontend.bus, NULL, arg, fn); +} +EXPORT_SYMBOL_GPL(xenbus_for_each_frontend); diff --git a/drivers/xenpvdrivers/xenbus_probe.h b/drivers/xenpvdrivers/xenbus_probe.h new file mode 100644 index 0000000..6156d31 --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_probe.h @@ -0,0 +1,93 @@ +/****************************************************************************** + * xenbus_probe.h + * + * Talks to Xen Store to figure out what devices we have. + * + * Copyright (C) 2005 Rusty Russell, IBM Corporation + * Copyright (C) 2005 XenSource Ltd. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef _XENBUS_PROBE_H +#define _XENBUS_PROBE_H + +#ifndef BUS_ID_SIZE +#define XEN_BUS_ID_SIZE 20 +#else +#define XEN_BUS_ID_SIZE BUS_ID_SIZE +#endif + +#ifdef CONFIG_PARAVIRT_XEN +#define is_running_on_xen() xen_domain() +#define is_initial_xendomain() xen_initial_domain() +#endif + +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26) +#define dev_name(dev) ((dev)->bus_id) +#endif + +#if defined(CONFIG_XEN_BACKEND) || defined(CONFIG_XEN_BACKEND_MODULE) +extern void xenbus_backend_suspend(int (*fn)(struct device *, void *)); +extern void xenbus_backend_resume(int (*fn)(struct device *, void *)); +extern void xenbus_backend_probe_and_watch(void); +extern void xenbus_backend_bus_register(void); +extern void xenbus_backend_device_register(void); +#else +static inline void xenbus_backend_suspend(int (*fn)(struct device *, void *)) {} +static inline void xenbus_backend_resume(int (*fn)(struct device *, void *)) {} +static inline void xenbus_backend_probe_and_watch(void) {} +static inline void xenbus_backend_bus_register(void) {} +static inline void xenbus_backend_device_register(void) {} +#endif + +struct xen_bus_type +{ + char *root; + int error; + unsigned int levels; + int (*get_bus_id)(char bus_id[XEN_BUS_ID_SIZE], const char *nodename); + int (*probe)(const char *type, const char *dir); + struct bus_type bus; +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + struct device dev; +#endif +}; + +extern int xenbus_match(struct device *_dev, struct device_driver *_drv); +extern int xenbus_dev_probe(struct device *_dev); +extern int xenbus_dev_remove(struct device *_dev); +extern int xenbus_register_driver_common(struct xenbus_driver *drv, + struct xen_bus_type *bus, + struct module *owner, + const char *mod_name); +extern int xenbus_probe_node(struct xen_bus_type *bus, + const char *type, + const char *nodename); +extern int xenbus_probe_devices(struct xen_bus_type *bus); + +extern void xenbus_dev_changed(const char *node, struct xen_bus_type *bus); + +#endif diff --git a/drivers/xenpvdrivers/xenbus_probe_backend.c b/drivers/xenpvdrivers/xenbus_probe_backend.c new file mode 100644 index 0000000..9bd6aa3 --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_probe_backend.c @@ -0,0 +1,285 @@ +/****************************************************************************** + * Talks to Xen Store to figure out what devices we have (backend half). + * + * Copyright (C) 2005 Rusty Russell, IBM Corporation + * Copyright (C) 2005 Mike Wray, Hewlett-Packard + * Copyright (C) 2005, 2006 XenSource Ltd + * Copyright (C) 2007 Solarflare Communications, Inc. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#define DPRINTK(fmt, args...) \ + pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \ + __FUNCTION__, __LINE__, ##args) + +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "xenbus_comms.h" +#include "xenbus_probe.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +static int xenbus_uevent_backend(struct device *dev, struct kobj_uevent_env *env); +static int xenbus_probe_backend(const char *type, const char *domid); + +extern int read_otherend_details(struct xenbus_device *xendev, + char *id_node, char *path_node); + +static int read_frontend_details(struct xenbus_device *xendev) +{ + return read_otherend_details(xendev, "frontend-id", "frontend"); +} + +/* backend/// => -- */ +static int backend_bus_id(char bus_id[XEN_BUS_ID_SIZE], const char *nodename) +{ + int domid, err; + const char *devid, *type, *frontend; + unsigned int typelen; + + type = strchr(nodename, '/'); + if (!type) + return -EINVAL; + type++; + typelen = strcspn(type, "/"); + if (!typelen || type[typelen] != '/') + return -EINVAL; + + devid = strrchr(nodename, '/') + 1; + + err = xenbus_gather(XBT_NIL, nodename, "frontend-id", "%i", &domid, + "frontend", NULL, &frontend, + NULL); + if (err) + return err; + if (strlen(frontend) == 0) + err = -ERANGE; + if (!err && !xenbus_exists(XBT_NIL, frontend, "")) + err = -ENOENT; + kfree(frontend); + + if (err) + return err; + + if (snprintf(bus_id, XEN_BUS_ID_SIZE, "%.*s-%i-%s", + typelen, type, domid, devid) >= XEN_BUS_ID_SIZE) + return -ENOSPC; + return 0; +} + +static struct device_attribute xenbus_backend_attrs[] = { + __ATTR_NULL +}; + +static struct xen_bus_type xenbus_backend = { + .root = "backend", + .levels = 3, /* backend/type// */ + .get_bus_id = backend_bus_id, + .probe = xenbus_probe_backend, + .error = -ENODEV, + .bus = { + .name = "xen-backend", + .match = xenbus_match, + .probe = xenbus_dev_probe, + .remove = xenbus_dev_remove, +// .shutdown = xenbus_dev_shutdown, + .uevent = xenbus_uevent_backend, + .dev_attrs = xenbus_backend_attrs, + }, + .dev = { + .init_name = "xen-backend", + }, +}; + +static int xenbus_uevent_backend(struct device *dev, struct kobj_uevent_env *env) +{ + struct xenbus_device *xdev; + struct xenbus_driver *drv; + + DPRINTK(""); + + if (dev == NULL) + return -ENODEV; + + xdev = to_xenbus_device(dev); + if (xdev == NULL) + return -ENODEV; + + /* stuff we want to pass to /sbin/hotplug */ + add_uevent_var(env, "XENBUS_TYPE=%s", xdev->devicetype); + + add_uevent_var(env, "XENBUS_PATH=%s", xdev->nodename); + + add_uevent_var(env, "XENBUS_BASE_PATH=%s", xenbus_backend.root); + + if (dev->driver) { + drv = to_xenbus_driver(dev->driver); + if (drv && drv->uevent) + return drv->uevent(xdev, env); + } + + return 0; +} + +int __xenbus_register_backend(struct xenbus_driver *drv, + struct module *owner, const char *mod_name) +{ + drv->read_otherend_details = read_frontend_details; + + return xenbus_register_driver_common(drv, &xenbus_backend, + owner, mod_name); +} +EXPORT_SYMBOL_GPL(__xenbus_register_backend); + +/* backend/// */ +static int xenbus_probe_backend_unit(const char *dir, + const char *type, + const char *name) +{ + char *nodename; + int err; + + nodename = kasprintf(GFP_KERNEL, "%s/%s", dir, name); + if (!nodename) + return -ENOMEM; + + DPRINTK("%s\n", nodename); + + err = xenbus_probe_node(&xenbus_backend, type, nodename); + kfree(nodename); + return err; +} + +/* backend// */ +static int xenbus_probe_backend(const char *type, const char *domid) +{ + char *nodename; + int err = 0; + char **dir; + unsigned int i, dir_n = 0; + + DPRINTK(""); + + nodename = kasprintf(GFP_KERNEL, "%s/%s/%s", xenbus_backend.root, type, domid); + if (!nodename) + return -ENOMEM; + + dir = xenbus_directory(XBT_NIL, nodename, "", &dir_n); + if (IS_ERR(dir)) { + kfree(nodename); + return PTR_ERR(dir); + } + + for (i = 0; i < dir_n; i++) { + err = xenbus_probe_backend_unit(nodename, type, dir[i]); + if (err) + break; + } + kfree(dir); + kfree(nodename); + return err; +} + +static void backend_changed(struct xenbus_watch *watch, + const char **vec, unsigned int len) +{ + DPRINTK(""); + + xenbus_dev_changed(vec[XS_WATCH_PATH], &xenbus_backend); +} + +static struct xenbus_watch be_watch = { + .node = "backend", + .callback = backend_changed, +}; + +void xenbus_backend_suspend(int (*fn)(struct device *, void *)) +{ + DPRINTK(""); + if (!xenbus_backend.error) + bus_for_each_dev(&xenbus_backend.bus, NULL, NULL, fn); +} + +void xenbus_backend_resume(int (*fn)(struct device *, void *)) +{ + DPRINTK(""); + if (!xenbus_backend.error) + bus_for_each_dev(&xenbus_backend.bus, NULL, NULL, fn); +} + +void xenbus_backend_probe_and_watch(void) +{ + xenbus_probe_devices(&xenbus_backend); + register_xenbus_watch(&be_watch); +} + +void xenbus_backend_bus_register(void) +{ + xenbus_backend.error = bus_register(&xenbus_backend.bus); + if (xenbus_backend.error) + printk(KERN_WARNING + "XENBUS: Error registering backend bus: %i\n", + xenbus_backend.error); +} + +void xenbus_backend_device_register(void) +{ + if (xenbus_backend.error) + return; + + xenbus_backend.error = device_register(&xenbus_backend.dev); + if (xenbus_backend.error) { + bus_unregister(&xenbus_backend.bus); + printk(KERN_WARNING + "XENBUS: Error registering backend device: %i\n", + xenbus_backend.error); + } +} + +int xenbus_for_each_backend(void *arg, int (*fn)(struct device *, void *)) +{ + return bus_for_each_dev(&xenbus_backend.bus, NULL, arg, fn); +} +EXPORT_SYMBOL_GPL(xenbus_for_each_backend); diff --git a/drivers/xenpvdrivers/xenbus_xs.c b/drivers/xenpvdrivers/xenbus_xs.c new file mode 100644 index 0000000..bb7c3f3 --- /dev/null +++ b/drivers/xenpvdrivers/xenbus_xs.c @@ -0,0 +1,956 @@ +/****************************************************************************** + * xenbus_xs.c + * + * This is the kernel equivalent of the "xs" library. We don't need everything + * and we use xenbus_comms for communication. + * + * Copyright (C) 2005 Rusty Russell, IBM Corporation + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "xenbus_comms.h" + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#ifndef PF_NOFREEZE /* Old kernel (pre-2.6.6). */ +#define PF_NOFREEZE 0 +#endif + +struct xs_stored_msg { + struct list_head list; + + struct xsd_sockmsg hdr; + + union { + /* Queued replies. */ + struct { + char *body; + } reply; + + /* Queued watch events. */ + struct { + struct xenbus_watch *handle; + char **vec; + unsigned int vec_size; + } watch; + } u; +}; + +struct xs_handle { + /* A list of replies. Currently only one will ever be outstanding. */ + struct list_head reply_list; + spinlock_t reply_lock; + wait_queue_head_t reply_waitq; + + /* + * Mutex ordering: transaction_mutex -> watch_mutex -> request_mutex. + * response_mutex is never taken simultaneously with the other three. + * + * transaction_mutex must be held before incrementing + * transaction_count. The mutex is held when a suspend is in + * progress to prevent new transactions starting. + * + * When decrementing transaction_count to zero the wait queue + * should be woken up, the suspend code waits for count to + * reach zero. + */ + + /* One request at a time. */ + struct mutex request_mutex; + + /* Protect xenbus reader thread against save/restore. */ + struct mutex response_mutex; + + /* Protect transactions against save/restore. */ + struct mutex transaction_mutex; + atomic_t transaction_count; + wait_queue_head_t transaction_wq; + + /* Protect watch (de)register against save/restore. */ + struct rw_semaphore watch_mutex; +}; + +static struct xs_handle xs_state; + +/* List of registered watches, and a lock to protect it. */ +static LIST_HEAD(watches); +static DEFINE_SPINLOCK(watches_lock); + +/* List of pending watch callback events, and a lock to protect it. */ +static LIST_HEAD(watch_events); +static DEFINE_SPINLOCK(watch_events_lock); + +/* + * Details of the xenwatch callback kernel thread. The thread waits on the + * watch_events_waitq for work to do (queued on watch_events list). When it + * wakes up it acquires the xenwatch_mutex before reading the list and + * carrying out work. + */ +static pid_t xenwatch_pid; +/* static */ DEFINE_MUTEX(xenwatch_mutex); +static DECLARE_WAIT_QUEUE_HEAD(watch_events_waitq); + +static int get_error(const char *errorstring) +{ + unsigned int i; + + for (i = 0; strcmp(errorstring, xsd_errors[i].errstring) != 0; i++) { + if (i == ARRAY_SIZE(xsd_errors) - 1) { + printk(KERN_WARNING + "XENBUS xen store gave: unknown error %s", + errorstring); + return EINVAL; + } + } + return xsd_errors[i].errnum; +} + +static void *read_reply(enum xsd_sockmsg_type *type, unsigned int *len) +{ + struct xs_stored_msg *msg; + char *body; + + spin_lock(&xs_state.reply_lock); + + while (list_empty(&xs_state.reply_list)) { + spin_unlock(&xs_state.reply_lock); + /* XXX FIXME: Avoid synchronous wait for response here. */ + wait_event(xs_state.reply_waitq, + !list_empty(&xs_state.reply_list)); + spin_lock(&xs_state.reply_lock); + } + + msg = list_entry(xs_state.reply_list.next, + struct xs_stored_msg, list); + list_del(&msg->list); + + spin_unlock(&xs_state.reply_lock); + + *type = msg->hdr.type; + if (len) + *len = msg->hdr.len; + body = msg->u.reply.body; + + kfree(msg); + + return body; +} + +static void transaction_start(void) +{ + mutex_lock(&xs_state.transaction_mutex); + atomic_inc(&xs_state.transaction_count); + mutex_unlock(&xs_state.transaction_mutex); +} + +static void transaction_end(void) +{ + if (atomic_dec_and_test(&xs_state.transaction_count)) + wake_up(&xs_state.transaction_wq); +} + +static void transaction_suspend(void) +{ + mutex_lock(&xs_state.transaction_mutex); + wait_event(xs_state.transaction_wq, + atomic_read(&xs_state.transaction_count) == 0); +} + +static void transaction_resume(void) +{ + mutex_unlock(&xs_state.transaction_mutex); +} + +void *xenbus_dev_request_and_reply(struct xsd_sockmsg *msg) +{ + void *ret; + struct xsd_sockmsg req_msg = *msg; + int err; + + if (req_msg.type == XS_TRANSACTION_START) + transaction_start(); + + mutex_lock(&xs_state.request_mutex); + + err = xb_write(msg, sizeof(*msg) + msg->len); + if (err) { + msg->type = XS_ERROR; + ret = ERR_PTR(err); + } else + ret = read_reply(&msg->type, &msg->len); + + mutex_unlock(&xs_state.request_mutex); + + if ((req_msg.type == XS_TRANSACTION_END) || + ((req_msg.type == XS_TRANSACTION_START) && + (msg->type == XS_ERROR))) + transaction_end(); + + return ret; +} +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) +EXPORT_SYMBOL(xenbus_dev_request_and_reply); +#endif + +/* Send message to xs, get kmalloc'ed reply. ERR_PTR() on error. */ +static void *xs_talkv(struct xenbus_transaction t, + enum xsd_sockmsg_type type, + const struct kvec *iovec, + unsigned int num_vecs, + unsigned int *len) +{ + struct xsd_sockmsg msg; + void *ret = NULL; + unsigned int i; + int err; + + msg.tx_id = t.id; + msg.req_id = 0; + msg.type = type; + msg.len = 0; + for (i = 0; i < num_vecs; i++) + msg.len += iovec[i].iov_len; + + mutex_lock(&xs_state.request_mutex); + + err = xb_write(&msg, sizeof(msg)); + if (err) { + mutex_unlock(&xs_state.request_mutex); + return ERR_PTR(err); + } + + for (i = 0; i < num_vecs; i++) { + err = xb_write(iovec[i].iov_base, iovec[i].iov_len); + if (err) { + mutex_unlock(&xs_state.request_mutex); + return ERR_PTR(err); + } + } + + ret = read_reply(&msg.type, len); + + mutex_unlock(&xs_state.request_mutex); + + if (IS_ERR(ret)) + return ret; + + if (msg.type == XS_ERROR) { + err = get_error(ret); + kfree(ret); + return ERR_PTR(-err); + } + + if (msg.type != type) { + if (printk_ratelimit()) + printk(KERN_WARNING + "XENBUS unexpected type [%d], expected [%d]\n", + msg.type, type); + kfree(ret); + return ERR_PTR(-EINVAL); + } + return ret; +} + +/* Simplified version of xs_talkv: single message. */ +static void *xs_single(struct xenbus_transaction t, + enum xsd_sockmsg_type type, + const char *string, + unsigned int *len) +{ + struct kvec iovec; + + iovec.iov_base = (void *)string; + iovec.iov_len = strlen(string) + 1; + return xs_talkv(t, type, &iovec, 1, len); +} + +/* Many commands only need an ack, don't care what it says. */ +static int xs_error(char *reply) +{ + if (IS_ERR(reply)) + return PTR_ERR(reply); + kfree(reply); + return 0; +} + +static unsigned int count_strings(const char *strings, unsigned int len) +{ + unsigned int num; + const char *p; + + for (p = strings, num = 0; p < strings + len; p += strlen(p) + 1) + num++; + + return num; +} + +/* Return the path to dir with /name appended. Buffer must be kfree()'ed. */ +static char *join(const char *dir, const char *name) +{ + char *buffer; + + if (strlen(name) == 0) + buffer = kasprintf(GFP_NOIO | __GFP_HIGH, "%s", dir); + else + buffer = kasprintf(GFP_NOIO | __GFP_HIGH, "%s/%s", dir, name); + return (!buffer) ? ERR_PTR(-ENOMEM) : buffer; +} + +static char **split(char *strings, unsigned int len, unsigned int *num) +{ + char *p, **ret; + + /* Count the strings. */ + *num = count_strings(strings, len) + 1; + + /* Transfer to one big alloc for easy freeing. */ + ret = kmalloc(*num * sizeof(char *) + len, GFP_NOIO | __GFP_HIGH); + if (!ret) { + kfree(strings); + return ERR_PTR(-ENOMEM); + } + memcpy(&ret[*num], strings, len); + kfree(strings); + + strings = (char *)&ret[*num]; + for (p = strings, *num = 0; p < strings + len; p += strlen(p) + 1) + ret[(*num)++] = p; + ret[*num] = strings + len; + + return ret; +} + +char **xenbus_directory(struct xenbus_transaction t, + const char *dir, const char *node, unsigned int *num) +{ + char *strings, *path; + unsigned int len; + + path = join(dir, node); + if (IS_ERR(path)) + return (char **)path; + + strings = xs_single(t, XS_DIRECTORY, path, &len); + kfree(path); + if (IS_ERR(strings)) + return (char **)strings; + + return split(strings, len, num); +} +EXPORT_SYMBOL_GPL(xenbus_directory); + +/* Check if a path exists. Return 1 if it does. */ +int xenbus_exists(struct xenbus_transaction t, + const char *dir, const char *node) +{ + char **d; + int dir_n; + + d = xenbus_directory(t, dir, node, &dir_n); + if (IS_ERR(d)) + return 0; + kfree(d); + return 1; +} +EXPORT_SYMBOL_GPL(xenbus_exists); + +/* Get the value of a single file. + * Returns a kmalloced value: call free() on it after use. + * len indicates length in bytes. + */ +void *xenbus_read(struct xenbus_transaction t, + const char *dir, const char *node, unsigned int *len) +{ + char *path; + void *ret; + + path = join(dir, node); + if (IS_ERR(path)) + return (void *)path; + + ret = xs_single(t, XS_READ, path, len); + kfree(path); + return ret; +} +EXPORT_SYMBOL_GPL(xenbus_read); + +/* Write the value of a single file. + * Returns -err on failure. + */ +int xenbus_write(struct xenbus_transaction t, + const char *dir, const char *node, const char *string) +{ + const char *path; + struct kvec iovec[2]; + int ret; + + path = join(dir, node); + if (IS_ERR(path)) + return PTR_ERR(path); + + iovec[0].iov_base = (void *)path; + iovec[0].iov_len = strlen(path) + 1; + iovec[1].iov_base = (void *)string; + iovec[1].iov_len = strlen(string); + + ret = xs_error(xs_talkv(t, XS_WRITE, iovec, ARRAY_SIZE(iovec), NULL)); + kfree(path); + return ret; +} +EXPORT_SYMBOL_GPL(xenbus_write); + +/* Create a new directory. */ +int xenbus_mkdir(struct xenbus_transaction t, + const char *dir, const char *node) +{ + char *path; + int ret; + + path = join(dir, node); + if (IS_ERR(path)) + return PTR_ERR(path); + + ret = xs_error(xs_single(t, XS_MKDIR, path, NULL)); + kfree(path); + return ret; +} +EXPORT_SYMBOL_GPL(xenbus_mkdir); + +/* Destroy a file or directory (directories must be empty). */ +int xenbus_rm(struct xenbus_transaction t, const char *dir, const char *node) +{ + char *path; + int ret; + + path = join(dir, node); + if (IS_ERR(path)) + return PTR_ERR(path); + + ret = xs_error(xs_single(t, XS_RM, path, NULL)); + kfree(path); + return ret; +} +EXPORT_SYMBOL_GPL(xenbus_rm); + +/* Start a transaction: changes by others will not be seen during this + * transaction, and changes will not be visible to others until end. + */ +int xenbus_transaction_start(struct xenbus_transaction *t) +{ + char *id_str; + + transaction_start(); + + id_str = xs_single(XBT_NIL, XS_TRANSACTION_START, "", NULL); + if (IS_ERR(id_str)) { + transaction_end(); + return PTR_ERR(id_str); + } + + t->id = simple_strtoul(id_str, NULL, 0); + kfree(id_str); + return 0; +} +EXPORT_SYMBOL_GPL(xenbus_transaction_start); + +/* End a transaction. + * If abandon is true, transaction is discarded instead of committed. + */ +int xenbus_transaction_end(struct xenbus_transaction t, int abort) +{ + char abortstr[2]; + int err; + + if (abort) + strcpy(abortstr, "F"); + else + strcpy(abortstr, "T"); + + err = xs_error(xs_single(t, XS_TRANSACTION_END, abortstr, NULL)); + + transaction_end(); + + return err; +} +EXPORT_SYMBOL_GPL(xenbus_transaction_end); + +/* Single read and scanf: returns -errno or num scanned. */ +int xenbus_scanf(struct xenbus_transaction t, + const char *dir, const char *node, const char *fmt, ...) +{ + va_list ap; + int ret; + char *val; + + val = xenbus_read(t, dir, node, NULL); + if (IS_ERR(val)) + return PTR_ERR(val); + + va_start(ap, fmt); + ret = vsscanf(val, fmt, ap); + va_end(ap); + kfree(val); + /* Distinctive errno. */ + if (ret == 0) + return -ERANGE; + return ret; +} +EXPORT_SYMBOL_GPL(xenbus_scanf); + +/* Single printf and write: returns -errno or 0. */ +int xenbus_printf(struct xenbus_transaction t, + const char *dir, const char *node, const char *fmt, ...) +{ + va_list ap; + int ret; +#define PRINTF_BUFFER_SIZE 4096 + char *printf_buffer; + + printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_NOIO | __GFP_HIGH); + if (printf_buffer == NULL) + return -ENOMEM; + + va_start(ap, fmt); + ret = vsnprintf(printf_buffer, PRINTF_BUFFER_SIZE, fmt, ap); + va_end(ap); + + BUG_ON(ret > PRINTF_BUFFER_SIZE-1); + ret = xenbus_write(t, dir, node, printf_buffer); + + kfree(printf_buffer); + + return ret; +} +EXPORT_SYMBOL_GPL(xenbus_printf); + +/* Takes tuples of names, scanf-style args, and void **, NULL terminated. */ +int xenbus_gather(struct xenbus_transaction t, const char *dir, ...) +{ + va_list ap; + const char *name; + int ret = 0; + + va_start(ap, dir); + while (ret == 0 && (name = va_arg(ap, char *)) != NULL) { + const char *fmt = va_arg(ap, char *); + void *result = va_arg(ap, void *); + char *p; + + p = xenbus_read(t, dir, name, NULL); + if (IS_ERR(p)) { + ret = PTR_ERR(p); + break; + } + if (fmt) { + if (sscanf(p, fmt, result) == 0) + ret = -EINVAL; + kfree(p); + } else + *(char **)result = p; + } + va_end(ap); + return ret; +} +EXPORT_SYMBOL_GPL(xenbus_gather); + +static int xs_watch(const char *path, const char *token) +{ + struct kvec iov[2]; + + iov[0].iov_base = (void *)path; + iov[0].iov_len = strlen(path) + 1; + iov[1].iov_base = (void *)token; + iov[1].iov_len = strlen(token) + 1; + + return xs_error(xs_talkv(XBT_NIL, XS_WATCH, iov, + ARRAY_SIZE(iov), NULL)); +} + +static int xs_unwatch(const char *path, const char *token) +{ + struct kvec iov[2]; + + iov[0].iov_base = (char *)path; + iov[0].iov_len = strlen(path) + 1; + iov[1].iov_base = (char *)token; + iov[1].iov_len = strlen(token) + 1; + + return xs_error(xs_talkv(XBT_NIL, XS_UNWATCH, iov, + ARRAY_SIZE(iov), NULL)); +} + +static struct xenbus_watch *find_watch(const char *token) +{ + struct xenbus_watch *i, *cmp; + + cmp = (void *)simple_strtoul(token, NULL, 16); + + list_for_each_entry(i, &watches, list) + if (i == cmp) + return i; + + return NULL; +} + +/* Register callback to watch this node. */ +int register_xenbus_watch(struct xenbus_watch *watch) +{ + /* Pointer in ascii is the token. */ + char token[sizeof(watch) * 2 + 1]; + int err; + + sprintf(token, "%lX", (long)watch); + + down_read(&xs_state.watch_mutex); + + spin_lock(&watches_lock); + BUG_ON(find_watch(token)); + list_add(&watch->list, &watches); + spin_unlock(&watches_lock); + + err = xs_watch(watch->node, token); + + /* Ignore errors due to multiple registration. */ + if ((err != 0) && (err != -EEXIST)) { + spin_lock(&watches_lock); + list_del(&watch->list); + spin_unlock(&watches_lock); + } + + up_read(&xs_state.watch_mutex); + + return err; +} +EXPORT_SYMBOL_GPL(register_xenbus_watch); + +void unregister_xenbus_watch(struct xenbus_watch *watch) +{ + struct xs_stored_msg *msg, *tmp; + char token[sizeof(watch) * 2 + 1]; + int err; + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + BUG_ON(watch->flags & XBWF_new_thread); +#endif + + sprintf(token, "%lX", (long)watch); + + down_read(&xs_state.watch_mutex); + + spin_lock(&watches_lock); + BUG_ON(!find_watch(token)); + list_del(&watch->list); + spin_unlock(&watches_lock); + + err = xs_unwatch(watch->node, token); + if (err) + printk(KERN_WARNING + "XENBUS Failed to release watch %s: %i\n", + watch->node, err); + + up_read(&xs_state.watch_mutex); + + /* Make sure there are no callbacks running currently (unless + its us) */ + if (current->pid != xenwatch_pid) + mutex_lock(&xenwatch_mutex); + + /* Cancel pending watch events. */ + spin_lock(&watch_events_lock); + list_for_each_entry_safe(msg, tmp, &watch_events, list) { + if (msg->u.watch.handle != watch) + continue; + list_del(&msg->list); + kfree(msg->u.watch.vec); + kfree(msg); + } + spin_unlock(&watch_events_lock); + + if (current->pid != xenwatch_pid) + mutex_unlock(&xenwatch_mutex); +} +EXPORT_SYMBOL_GPL(unregister_xenbus_watch); + +void xs_suspend(void) +{ + transaction_suspend(); + down_write(&xs_state.watch_mutex); + mutex_lock(&xs_state.request_mutex); + mutex_lock(&xs_state.response_mutex); +} + +void xs_resume(void) +{ + struct xenbus_watch *watch; + char token[sizeof(watch) * 2 + 1]; + +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) + xb_init_comms(); +#endif + + mutex_unlock(&xs_state.response_mutex); + mutex_unlock(&xs_state.request_mutex); + transaction_resume(); + + /* No need for watches_lock: the watch_mutex is sufficient. */ + list_for_each_entry(watch, &watches, list) { + sprintf(token, "%lX", (long)watch); + xs_watch(watch->node, token); + } + + up_write(&xs_state.watch_mutex); +} + +void xs_suspend_cancel(void) +{ + mutex_unlock(&xs_state.response_mutex); + mutex_unlock(&xs_state.request_mutex); + up_write(&xs_state.watch_mutex); + mutex_unlock(&xs_state.transaction_mutex); +} + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +static int xenwatch_handle_callback(void *data) +{ + struct xs_stored_msg *msg = data; + + msg->u.watch.handle->callback(msg->u.watch.handle, + (const char **)msg->u.watch.vec, + msg->u.watch.vec_size); + + kfree(msg->u.watch.vec); + kfree(msg); + + /* Kill this kthread if we were spawned just for this callback. */ + if (current->pid != xenwatch_pid) + do_exit(0); + + return 0; +} +#endif + +static int xenwatch_thread(void *unused) +{ + struct list_head *ent; + struct xs_stored_msg *msg; + + current->flags |= PF_NOFREEZE; + for (;;) { + wait_event_interruptible(watch_events_waitq, + !list_empty(&watch_events)); + + if (kthread_should_stop()) + break; + + mutex_lock(&xenwatch_mutex); + + spin_lock(&watch_events_lock); + ent = watch_events.next; + if (ent != &watch_events) + list_del(ent); + spin_unlock(&watch_events_lock); + + if (ent == &watch_events) { + mutex_unlock(&xenwatch_mutex); + continue; + } + + msg = list_entry(ent, struct xs_stored_msg, list); + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) + /* + * Unlock the mutex before running an XBWF_new_thread + * handler. kthread_run can block which can deadlock + * against unregister_xenbus_watch() if we need to + * unregister other watches in order to make + * progress. This can occur on resume before the swap + * device is attached. + */ + if (msg->u.watch.handle->flags & XBWF_new_thread) { + mutex_unlock(&xenwatch_mutex); + kthread_run(xenwatch_handle_callback, + msg, "xenwatch_cb"); + } else { + xenwatch_handle_callback(msg); + mutex_unlock(&xenwatch_mutex); + } +#else + msg->u.watch.handle->callback( + msg->u.watch.handle, + (const char **)msg->u.watch.vec, + msg->u.watch.vec_size); + mutex_unlock(&xenwatch_mutex); + kfree(msg->u.watch.vec); + kfree(msg); +#endif + } + + return 0; +} + +static int process_msg(void) +{ + struct xs_stored_msg *msg; + char *body; + int err; + + /* + * We must disallow save/restore while reading a xenstore message. + * A partial read across s/r leaves us out of sync with xenstored. + */ + for (;;) { + err = xb_wait_for_data_to_read(); + if (err) + return err; + mutex_lock(&xs_state.response_mutex); + if (xb_data_to_read()) + break; + /* We raced with save/restore: pending data 'disappeared'. */ + mutex_unlock(&xs_state.response_mutex); + } + + + msg = kmalloc(sizeof(*msg), GFP_NOIO | __GFP_HIGH); + if (msg == NULL) { + err = -ENOMEM; + goto out; + } + + err = xb_read(&msg->hdr, sizeof(msg->hdr)); + if (err) { + kfree(msg); + goto out; + } + + body = kmalloc(msg->hdr.len + 1, GFP_NOIO | __GFP_HIGH); + if (body == NULL) { + kfree(msg); + err = -ENOMEM; + goto out; + } + + err = xb_read(body, msg->hdr.len); + if (err) { + kfree(body); + kfree(msg); + goto out; + } + body[msg->hdr.len] = '\0'; + + if (msg->hdr.type == XS_WATCH_EVENT) { + msg->u.watch.vec = split(body, msg->hdr.len, + &msg->u.watch.vec_size); + if (IS_ERR(msg->u.watch.vec)) { + err = PTR_ERR(msg->u.watch.vec); + kfree(msg); + goto out; + } + + spin_lock(&watches_lock); + msg->u.watch.handle = find_watch( + msg->u.watch.vec[XS_WATCH_TOKEN]); + if (msg->u.watch.handle != NULL) { + spin_lock(&watch_events_lock); + list_add_tail(&msg->list, &watch_events); + wake_up(&watch_events_waitq); + spin_unlock(&watch_events_lock); + } else { + kfree(msg->u.watch.vec); + kfree(msg); + } + spin_unlock(&watches_lock); + } else { + msg->u.reply.body = body; + spin_lock(&xs_state.reply_lock); + list_add_tail(&msg->list, &xs_state.reply_list); + spin_unlock(&xs_state.reply_lock); + wake_up(&xs_state.reply_waitq); + } + + out: + mutex_unlock(&xs_state.response_mutex); + return err; +} + +static int xenbus_thread(void *unused) +{ + int err; + + current->flags |= PF_NOFREEZE; + for (;;) { + err = process_msg(); + if (err) + printk(KERN_WARNING "XENBUS error %d while reading " + "message\n", err); + if (kthread_should_stop()) + break; + } + + return 0; +} + +int xs_init(void) +{ + struct task_struct *task; + + INIT_LIST_HEAD(&xs_state.reply_list); + spin_lock_init(&xs_state.reply_lock); + init_waitqueue_head(&xs_state.reply_waitq); + + mutex_init(&xs_state.request_mutex); + mutex_init(&xs_state.response_mutex); + mutex_init(&xs_state.transaction_mutex); + init_rwsem(&xs_state.watch_mutex); + atomic_set(&xs_state.transaction_count, 0); + init_waitqueue_head(&xs_state.transaction_wq); + + task = kthread_run(xenwatch_thread, NULL, "xenwatch"); + if (IS_ERR(task)) + return PTR_ERR(task); + xenwatch_pid = task->pid; + + task = kthread_run(xenbus_thread, NULL, "xenbus"); + if (IS_ERR(task)) + return PTR_ERR(task); + + return 0; +} diff --git a/include/xenpvdrivers/asm/gnttab_dma.h b/include/xenpvdrivers/asm/gnttab_dma.h new file mode 100644 index 0000000..fd7197c --- /dev/null +++ b/include/xenpvdrivers/asm/gnttab_dma.h @@ -0,0 +1,41 @@ +/* + * Copyright (c) 2007 Herbert Xu + * Copyright (c) 2007 Isaku Yamahata + * VA Linux Systems Japan K.K. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef _ASM_I386_GNTTAB_DMA_H +#define _ASM_I386_GNTTAB_DMA_H + +static inline int gnttab_dma_local_pfn(struct page *page) +{ + /* Has it become a local MFN? */ + return pfn_valid(mfn_to_local_pfn(pfn_to_mfn(page_to_pfn(page)))); +} + +static inline maddr_t gnttab_dma_map_page(struct page *page) +{ + __gnttab_dma_map_page(page); + return ((maddr_t)pfn_to_mfn(page_to_pfn(page)) << PAGE_SHIFT); +} + +static inline void gnttab_dma_unmap_page(maddr_t maddr) +{ + __gnttab_dma_unmap_page(virt_to_page(bus_to_virt(maddr))); +} + +#endif /* _ASM_I386_GNTTAB_DMA_H */ diff --git a/include/xenpvdrivers/asm/hypercall.h b/include/xenpvdrivers/asm/hypercall.h new file mode 100644 index 0000000..f9362ca --- /dev/null +++ b/include/xenpvdrivers/asm/hypercall.h @@ -0,0 +1,430 @@ +/****************************************************************************** + * hypercall.h + * + * Linux-specific hypervisor handling. + * + * Copyright (c) 2002-2004, K A Fraser + * + * 64-bit updates: + * Benjamin Liu + * Jun Nakajima + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __HYPERCALL_H__ +#define __HYPERCALL_H__ + +#ifndef __HYPERVISOR_H__ +# error "please don't include this file directly" +#endif + +#if CONFIG_XEN_COMPAT <= 0x030002 +# include /* memcpy() */ +#endif + +#ifdef CONFIG_XEN +#define HYPERCALL_ASM_OPERAND "%c" +#define HYPERCALL_LOCATION(op) (hypercall_page + (op) * 32) +#define HYPERCALL_C_OPERAND(name) "i" (HYPERCALL_LOCATION(__HYPERVISOR_##name)) +#else +#define HYPERCALL_ASM_OPERAND "*%" +#define HYPERCALL_LOCATION(op) (hypercall_stubs + (op) * 32) +#define HYPERCALL_C_OPERAND(name) "g" (HYPERCALL_LOCATION(__HYPERVISOR_##name)) +#endif + +#define HYPERCALL_ARG(arg, n) \ + register typeof((arg)+0) __arg##n asm(HYPERCALL_arg##n) = (arg) + +#define _hypercall0(type, name) \ +({ \ + type __res; \ + asm volatile ( \ + "call " HYPERCALL_ASM_OPERAND "1" \ + : "=a" (__res) \ + : HYPERCALL_C_OPERAND(name) \ + : "memory" ); \ + __res; \ +}) + +#define _hypercall1(type, name, arg) \ +({ \ + type __res; \ + HYPERCALL_ARG(arg, 1); \ + asm volatile ( \ + "call " HYPERCALL_ASM_OPERAND "2" \ + : "=a" (__res), "+r" (__arg1) \ + : HYPERCALL_C_OPERAND(name) \ + : "memory" ); \ + __res; \ +}) + +#define _hypercall2(type, name, a1, a2) \ +({ \ + type __res; \ + HYPERCALL_ARG(a1, 1); \ + HYPERCALL_ARG(a2, 2); \ + asm volatile ( \ + "call " HYPERCALL_ASM_OPERAND "3" \ + : "=a" (__res), "+r" (__arg1), "+r" (__arg2) \ + : HYPERCALL_C_OPERAND(name) \ + : "memory" ); \ + __res; \ +}) + +#define _hypercall3(type, name, a1, a2, a3) \ +({ \ + type __res; \ + HYPERCALL_ARG(a1, 1); \ + HYPERCALL_ARG(a2, 2); \ + HYPERCALL_ARG(a3, 3); \ + asm volatile ( \ + "call " HYPERCALL_ASM_OPERAND "4" \ + : "=a" (__res), "+r" (__arg1), \ + "+r" (__arg2), "+r" (__arg3) \ + : HYPERCALL_C_OPERAND(name) \ + : "memory" ); \ + __res; \ +}) + +#define _hypercall4(type, name, a1, a2, a3, a4) \ +({ \ + type __res; \ + HYPERCALL_ARG(a1, 1); \ + HYPERCALL_ARG(a2, 2); \ + HYPERCALL_ARG(a3, 3); \ + HYPERCALL_ARG(a4, 4); \ + asm volatile ( \ + "call " HYPERCALL_ASM_OPERAND "5" \ + : "=a" (__res), "+r" (__arg1), "+r" (__arg2), \ + "+r" (__arg3), "+r" (__arg4) \ + : HYPERCALL_C_OPERAND(name) \ + : "memory" ); \ + __res; \ +}) + +#define _hypercall5(type, name, a1, a2, a3, a4, a5) \ +({ \ + type __res; \ + HYPERCALL_ARG(a1, 1); \ + HYPERCALL_ARG(a2, 2); \ + HYPERCALL_ARG(a3, 3); \ + HYPERCALL_ARG(a4, 4); \ + HYPERCALL_ARG(a5, 5); \ + asm volatile ( \ + "call " HYPERCALL_ASM_OPERAND "6" \ + : "=a" (__res), "+r" (__arg1), "+r" (__arg2), \ + "+r" (__arg3), "+r" (__arg4), "+r" (__arg5) \ + : HYPERCALL_C_OPERAND(name) \ + : "memory" ); \ + __res; \ +}) + +#define _hypercall(type, op, a1, a2, a3, a4, a5) \ +({ \ + type __res; \ + HYPERCALL_ARG(a1, 1); \ + HYPERCALL_ARG(a2, 2); \ + HYPERCALL_ARG(a3, 3); \ + HYPERCALL_ARG(a4, 4); \ + HYPERCALL_ARG(a5, 5); \ + asm volatile ( \ + "call *%6" \ + : "=a" (__res), "+r" (__arg1), "+r" (__arg2), \ + "+r" (__arg3), "+r" (__arg4), "+r" (__arg5) \ + : "g" (HYPERCALL_LOCATION(op)) \ + : "memory" ); \ + __res; \ +}) + +#ifdef CONFIG_X86_32 +# include "hypercall_32.h" +#else +# include "hypercall_64.h" +#endif + +static inline int __must_check +HYPERVISOR_set_trap_table( + const trap_info_t *table) +{ + return _hypercall1(int, set_trap_table, table); +} + +static inline int __must_check +HYPERVISOR_mmu_update( + mmu_update_t *req, unsigned int count, unsigned int *success_count, + domid_t domid) +{ + if (arch_use_lazy_mmu_mode()) + return xen_multi_mmu_update(req, count, success_count, domid); + return _hypercall4(int, mmu_update, req, count, success_count, domid); +} + +static inline int __must_check +HYPERVISOR_mmuext_op( + struct mmuext_op *op, unsigned int count, unsigned int *success_count, + domid_t domid) +{ + if (arch_use_lazy_mmu_mode()) + return xen_multi_mmuext_op(op, count, success_count, domid); + return _hypercall4(int, mmuext_op, op, count, success_count, domid); +} + +static inline int __must_check +HYPERVISOR_set_gdt( + unsigned long *frame_list, unsigned int entries) +{ + return _hypercall2(int, set_gdt, frame_list, entries); +} + +static inline int __must_check +HYPERVISOR_stack_switch( + unsigned long ss, unsigned long esp) +{ + return _hypercall2(int, stack_switch, ss, esp); +} + +static inline int +HYPERVISOR_fpu_taskswitch( + int set) +{ + return _hypercall1(int, fpu_taskswitch, set); +} + +#if CONFIG_XEN_COMPAT <= 0x030002 +static inline int __must_check +HYPERVISOR_sched_op_compat( + int cmd, unsigned long arg) +{ + return _hypercall2(int, sched_op_compat, cmd, arg); +} +#endif + +static inline int __must_check +HYPERVISOR_sched_op( + int cmd, void *arg) +{ + return _hypercall2(int, sched_op, cmd, arg); +} + +static inline int __must_check +HYPERVISOR_platform_op( + struct xen_platform_op *platform_op) +{ + platform_op->interface_version = XENPF_INTERFACE_VERSION; + return _hypercall1(int, platform_op, platform_op); +} + +struct xen_mc; +static inline int __must_check +HYPERVISOR_mca( + struct xen_mc *mc_op) +{ + mc_op->interface_version = XEN_MCA_INTERFACE_VERSION; + return _hypercall1(int, mca, mc_op); +} + +static inline int __must_check +HYPERVISOR_set_debugreg( + unsigned int reg, unsigned long value) +{ + return _hypercall2(int, set_debugreg, reg, value); +} + +static inline unsigned long __must_check +HYPERVISOR_get_debugreg( + unsigned int reg) +{ + return _hypercall1(unsigned long, get_debugreg, reg); +} + +static inline int __must_check +HYPERVISOR_memory_op( + unsigned int cmd, void *arg) +{ + if (arch_use_lazy_mmu_mode()) + xen_multicall_flush(false); + return _hypercall2(int, memory_op, cmd, arg); +} + +static inline int __must_check +HYPERVISOR_multicall( + multicall_entry_t *call_list, unsigned int nr_calls) +{ + return _hypercall2(int, multicall, call_list, nr_calls); +} + +static inline int __must_check +HYPERVISOR_event_channel_op( + int cmd, void *arg) +{ + int rc = _hypercall2(int, event_channel_op, cmd, arg); + +#if CONFIG_XEN_COMPAT <= 0x030002 + if (unlikely(rc == -ENOSYS)) { + struct evtchn_op op; + op.cmd = cmd; + memcpy(&op.u, arg, sizeof(op.u)); + rc = _hypercall1(int, event_channel_op_compat, &op); + memcpy(arg, &op.u, sizeof(op.u)); + } +#endif + + return rc; +} + +static inline int __must_check +HYPERVISOR_xen_version( + int cmd, void *arg) +{ + return _hypercall2(int, xen_version, cmd, arg); +} + +static inline int __must_check +HYPERVISOR_console_io( + int cmd, unsigned int count, char *str) +{ + return _hypercall3(int, console_io, cmd, count, str); +} + +static inline int __must_check +HYPERVISOR_physdev_op( + int cmd, void *arg) +{ + int rc = _hypercall2(int, physdev_op, cmd, arg); + +#if CONFIG_XEN_COMPAT <= 0x030002 + if (unlikely(rc == -ENOSYS)) { + struct physdev_op op; + op.cmd = cmd; + memcpy(&op.u, arg, sizeof(op.u)); + rc = _hypercall1(int, physdev_op_compat, &op); + memcpy(arg, &op.u, sizeof(op.u)); + } +#endif + + return rc; +} + +static inline int __must_check +HYPERVISOR_grant_table_op( + unsigned int cmd, void *uop, unsigned int count) +{ + bool fixup = false; + int rc; + + if (arch_use_lazy_mmu_mode()) + xen_multicall_flush(false); +#ifdef GNTTABOP_map_grant_ref + if (cmd == GNTTABOP_map_grant_ref) +#endif + fixup = gnttab_pre_map_adjust(cmd, uop, count); + rc = _hypercall3(int, grant_table_op, cmd, uop, count); + if (rc == 0 && fixup) + rc = gnttab_post_map_adjust(uop, count); + return rc; +} + +static inline int __must_check +HYPERVISOR_vm_assist( + unsigned int cmd, unsigned int type) +{ + return _hypercall2(int, vm_assist, cmd, type); +} + +static inline int __must_check +HYPERVISOR_vcpu_op( + int cmd, unsigned int vcpuid, void *extra_args) +{ + return _hypercall3(int, vcpu_op, cmd, vcpuid, extra_args); +} + +static inline int __must_check +HYPERVISOR_suspend( + unsigned long srec) +{ + struct sched_shutdown sched_shutdown = { + .reason = SHUTDOWN_suspend + }; + + int rc = _hypercall3(int, sched_op, SCHEDOP_shutdown, + &sched_shutdown, srec); + +#if CONFIG_XEN_COMPAT <= 0x030002 + if (rc == -ENOSYS) + rc = _hypercall3(int, sched_op_compat, SCHEDOP_shutdown, + SHUTDOWN_suspend, srec); +#endif + + return rc; +} + +#if CONFIG_XEN_COMPAT <= 0x030002 +static inline int +HYPERVISOR_nmi_op( + unsigned long op, void *arg) +{ + return _hypercall2(int, nmi_op, op, arg); +} +#endif + +#ifndef CONFIG_XEN +static inline unsigned long __must_check +HYPERVISOR_hvm_op( + int op, void *arg) +{ + return _hypercall2(unsigned long, hvm_op, op, arg); +} +#endif + +static inline int __must_check +HYPERVISOR_callback_op( + int cmd, const void *arg) +{ + return _hypercall2(int, callback_op, cmd, arg); +} + +static inline int __must_check +HYPERVISOR_xenoprof_op( + int op, void *arg) +{ + return _hypercall2(int, xenoprof_op, op, arg); +} + +static inline int __must_check +HYPERVISOR_kexec_op( + unsigned long op, void *args) +{ + return _hypercall2(int, kexec_op, op, args); +} + +static inline int __must_check +HYPERVISOR_tmem_op( + struct tmem_op *op) +{ + return _hypercall1(int, tmem_op, op); +} + +#endif /* __HYPERCALL_H__ */ diff --git a/include/xenpvdrivers/asm/hypercall_32.h b/include/xenpvdrivers/asm/hypercall_32.h new file mode 100644 index 0000000..3987b2e --- /dev/null +++ b/include/xenpvdrivers/asm/hypercall_32.h @@ -0,0 +1,62 @@ +#define HYPERCALL_arg1 "ebx" +#define HYPERCALL_arg2 "ecx" +#define HYPERCALL_arg3 "edx" +#define HYPERCALL_arg4 "esi" +#define HYPERCALL_arg5 "edi" + +#if CONFIG_XEN_COMPAT <= 0x030002 +static inline int __must_check +HYPERVISOR_set_callbacks( + unsigned long event_selector, unsigned long event_address, + unsigned long failsafe_selector, unsigned long failsafe_address) +{ + return _hypercall4(int, set_callbacks, + event_selector, event_address, + failsafe_selector, failsafe_address); +} +#endif + +static inline long __must_check +HYPERVISOR_set_timer_op( + u64 timeout) +{ + return _hypercall2(long, set_timer_op, + (unsigned long)timeout, + (unsigned long)(timeout>>32)); +} + +static inline int __must_check +HYPERVISOR_update_descriptor( + u64 ma, u64 desc) +{ + return _hypercall4(int, update_descriptor, + (unsigned long)ma, (unsigned long)(ma>>32), + (unsigned long)desc, (unsigned long)(desc>>32)); +} + +static inline int __must_check +HYPERVISOR_update_va_mapping( + unsigned long va, pte_t new_val, unsigned long flags) +{ + unsigned long pte_hi = 0; + + if (arch_use_lazy_mmu_mode()) + return xen_multi_update_va_mapping(va, new_val, flags); +#ifdef CONFIG_X86_PAE + pte_hi = new_val.pte_high; +#endif + return _hypercall4(int, update_va_mapping, va, + new_val.pte_low, pte_hi, flags); +} + +static inline int __must_check +HYPERVISOR_update_va_mapping_otherdomain( + unsigned long va, pte_t new_val, unsigned long flags, domid_t domid) +{ + unsigned long pte_hi = 0; +#ifdef CONFIG_X86_PAE + pte_hi = new_val.pte_high; +#endif + return _hypercall5(int, update_va_mapping_otherdomain, va, + new_val.pte_low, pte_hi, flags, domid); +} diff --git a/include/xenpvdrivers/asm/hypercall_64.h b/include/xenpvdrivers/asm/hypercall_64.h new file mode 100644 index 0000000..97d9445 --- /dev/null +++ b/include/xenpvdrivers/asm/hypercall_64.h @@ -0,0 +1,54 @@ +#define HYPERCALL_arg1 "rdi" +#define HYPERCALL_arg2 "rsi" +#define HYPERCALL_arg3 "rdx" +#define HYPERCALL_arg4 "r10" +#define HYPERCALL_arg5 "r8" + +#if CONFIG_XEN_COMPAT <= 0x030002 +static inline int __must_check +HYPERVISOR_set_callbacks( + unsigned long event_address, unsigned long failsafe_address, + unsigned long syscall_address) +{ + return _hypercall3(int, set_callbacks, + event_address, failsafe_address, syscall_address); +} +#endif + +static inline long __must_check +HYPERVISOR_set_timer_op( + u64 timeout) +{ + return _hypercall1(long, set_timer_op, timeout); +} + +static inline int __must_check +HYPERVISOR_update_descriptor( + unsigned long ma, unsigned long word) +{ + return _hypercall2(int, update_descriptor, ma, word); +} + +static inline int __must_check +HYPERVISOR_update_va_mapping( + unsigned long va, pte_t new_val, unsigned long flags) +{ + if (arch_use_lazy_mmu_mode()) + return xen_multi_update_va_mapping(va, new_val, flags); + return _hypercall3(int, update_va_mapping, va, new_val.pte, flags); +} + +static inline int __must_check +HYPERVISOR_update_va_mapping_otherdomain( + unsigned long va, pte_t new_val, unsigned long flags, domid_t domid) +{ + return _hypercall4(int, update_va_mapping_otherdomain, va, + new_val.pte, flags, domid); +} + +static inline int __must_check +HYPERVISOR_set_segment_base( + int reg, unsigned long value) +{ + return _hypercall2(int, set_segment_base, reg, value); +} diff --git a/include/xenpvdrivers/asm/hypervisor.h b/include/xenpvdrivers/asm/hypervisor.h new file mode 100644 index 0000000..54ee922 --- /dev/null +++ b/include/xenpvdrivers/asm/hypervisor.h @@ -0,0 +1,390 @@ +/****************************************************************************** + * hypervisor.h + * + * Linux-specific hypervisor handling. + * + * Copyright (c) 2002-2004, K A Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __HYPERVISOR_H__ +#define __HYPERVISOR_H__ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +extern shared_info_t *HYPERVISOR_shared_info; + +#ifdef CONFIG_XEN_VCPU_INFO_PLACEMENT +DECLARE_PER_CPU(struct vcpu_info, vcpu_info); +#define vcpu_info(cpu) (&per_cpu(vcpu_info, cpu)) +#define current_vcpu_info() (&__get_cpu_var(vcpu_info)) +#define vcpu_info_read(fld) percpu_read(vcpu_info.fld) +#define vcpu_info_write(fld, val) percpu_write(vcpu_info.fld, val) +#define vcpu_info_xchg(fld, val) percpu_xchg(vcpu_info.fld, val) +void setup_vcpu_info(unsigned int cpu); +void adjust_boot_vcpu_info(void); +#else +#define vcpu_info(cpu) (HYPERVISOR_shared_info->vcpu_info + (cpu)) +#ifdef CONFIG_SMP +#define current_vcpu_info() vcpu_info(smp_processor_id()) +#else +#define current_vcpu_info() vcpu_info(0) +#endif +#define vcpu_info_read(fld) (current_vcpu_info()->fld) +#define vcpu_info_write(fld, val) (current_vcpu_info()->fld = (val)) +static inline void setup_vcpu_info(unsigned int cpu) {} +#endif + +#ifdef CONFIG_X86_32 +extern unsigned long hypervisor_virt_start; +#endif + +/* arch/xen/i386/kernel/setup.c */ +extern start_info_t *xen_start_info; +#ifdef CONFIG_XEN_PRIVILEGED_GUEST +#define is_initial_xendomain() (xen_start_info->flags & SIF_INITDOMAIN) +#else +#define is_initial_xendomain() 0 +#endif + +#define init_hypervisor(c) ((void)((c)->x86_hyper_vendor = X86_HYPER_VENDOR_XEN)) +#define init_hypervisor_platform() init_hypervisor(&boot_cpu_data) + +struct vcpu_runstate_info *setup_runstate_area(unsigned int cpu); + +/* arch/xen/kernel/evtchn.c */ +/* Force a proper event-channel callback from Xen. */ +void force_evtchn_callback(void); + +/* arch/xen/kernel/process.c */ +void xen_cpu_idle (void); + +/* arch/xen/i386/kernel/hypervisor.c */ +void do_hypervisor_callback(struct pt_regs *regs); + +/* arch/xen/i386/mm/hypervisor.c */ +/* + * NB. ptr values should be PHYSICAL, not MACHINE. 'vals' should be already + * be MACHINE addresses. + */ + +void xen_pt_switch(pgd_t *); +void xen_new_user_pt(pgd_t *); /* x86_64 only */ +void xen_load_gs(unsigned int selector); /* x86_64 only */ +void xen_tlb_flush(void); +void xen_invlpg(unsigned long ptr); + +void xen_l1_entry_update(pte_t *ptr, pte_t val); +void xen_l2_entry_update(pmd_t *ptr, pmd_t val); +void xen_l3_entry_update(pud_t *ptr, pud_t val); /* x86_64/PAE */ +void xen_l4_entry_update(pgd_t *ptr, pgd_t val); /* x86_64 only */ +void xen_pgd_pin(pgd_t *); +void xen_pgd_unpin(pgd_t *); + +void xen_init_pgd_pin(void); + +void xen_set_ldt(const void *ptr, unsigned int ents); + +#ifdef CONFIG_SMP +#include +void xen_tlb_flush_all(void); +void xen_invlpg_all(unsigned long ptr); +void xen_tlb_flush_mask(const cpumask_t *mask); +void xen_invlpg_mask(const cpumask_t *mask, unsigned long ptr); +#else +#define xen_tlb_flush_all xen_tlb_flush +#define xen_invlpg_all xen_invlpg +#endif + +/* Returns zero on success else negative errno. */ +int xen_create_contiguous_region( + unsigned long vstart, unsigned int order, unsigned int address_bits); +void xen_destroy_contiguous_region( + unsigned long vstart, unsigned int order); +int early_create_contiguous_region(unsigned long pfn, unsigned int order, + unsigned int address_bits); + +struct page; + +int xen_limit_pages_to_max_mfn( + struct page *pages, unsigned int order, unsigned int address_bits); + +/* Turn jiffies into Xen system time. */ +u64 jiffies_to_st(unsigned long jiffies); + +#ifdef CONFIG_XEN_SCRUB_PAGES +void scrub_pages(void *, unsigned int); +#else +#define scrub_pages(_p,_n) ((void)0) +#endif + +#if defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) + +DECLARE_PER_CPU(bool, xen_lazy_mmu); + +void xen_multicall_flush(bool); + +int __must_check xen_multi_update_va_mapping(unsigned long va, pte_t, + unsigned long flags); +int __must_check xen_multi_mmu_update(mmu_update_t *, unsigned int count, + unsigned int *success_count, domid_t); +int __must_check xen_multi_mmuext_op(struct mmuext_op *, unsigned int count, + unsigned int *success_count, domid_t); + +#define __HAVE_ARCH_ENTER_LAZY_MMU_MODE +static inline void arch_enter_lazy_mmu_mode(void) +{ + percpu_write(xen_lazy_mmu, true); +} + +static inline void arch_leave_lazy_mmu_mode(void) +{ + percpu_write(xen_lazy_mmu, false); + xen_multicall_flush(false); +} + +#define arch_use_lazy_mmu_mode() unlikely(percpu_read(xen_lazy_mmu)) + +#if 0 /* All uses are in places potentially called asynchronously, but + * asynchronous code should rather not make use of lazy mode at all. + * Therefore, all uses of this function get commented out, proper + * detection of asynchronous invocations is added whereever needed, + * and this function is disabled to catch any new (improper) uses. + */ +static inline void arch_flush_lazy_mmu_mode(void) +{ + if (arch_use_lazy_mmu_mode()) + xen_multicall_flush(false); +} +#endif + +#else /* !CONFIG_XEN || XENPV_COMPAT_MODULE */ + +static inline void xen_multicall_flush(bool ignore) {} +#define arch_use_lazy_mmu_mode() false +#define xen_multi_update_va_mapping(...) ({ BUG(); -ENOSYS; }) +#define xen_multi_mmu_update(...) ({ BUG(); -ENOSYS; }) +#define xen_multi_mmuext_op(...) ({ BUG(); -ENOSYS; }) + +#endif /* CONFIG_XEN && !XENPV_COMPAT_MODULE */ + +#ifdef CONFIG_XEN + +struct gnttab_map_grant_ref; +bool gnttab_pre_map_adjust(unsigned int cmd, struct gnttab_map_grant_ref *, + unsigned int count); +#if CONFIG_XEN_COMPAT < 0x030400 +int gnttab_post_map_adjust(const struct gnttab_map_grant_ref *, unsigned int); +#else +static inline int gnttab_post_map_adjust(const struct gnttab_map_grant_ref *m, + unsigned int count) +{ + BUG(); + return -ENOSYS; +} +#endif + +#else /* !CONFIG_XEN */ + +#define gnttab_pre_map_adjust(...) false +#define gnttab_post_map_adjust(...) ({ BUG(); -ENOSYS; }) + +#endif /* CONFIG_XEN */ + +#if defined(CONFIG_X86_64) +#define MULTI_UVMFLAGS_INDEX 2 +#define MULTI_UVMDOMID_INDEX 3 +#else +#define MULTI_UVMFLAGS_INDEX 3 +#define MULTI_UVMDOMID_INDEX 4 +#endif + +#ifdef CONFIG_XEN +#define is_running_on_xen() 1 +extern char hypercall_page[PAGE_SIZE]; +#else +extern char *hypercall_stubs; +#define is_running_on_xen() (!!hypercall_stubs) +#endif + +#include + +static inline int +HYPERVISOR_yield( + void) +{ + int rc = HYPERVISOR_sched_op(SCHEDOP_yield, NULL); + +#if CONFIG_XEN_COMPAT <= 0x030002 + if (rc == -ENOSYS) + rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0); +#endif + + return rc; +} + +static inline int +HYPERVISOR_block( + void) +{ + int rc = HYPERVISOR_sched_op(SCHEDOP_block, NULL); + +#if CONFIG_XEN_COMPAT <= 0x030002 + if (rc == -ENOSYS) + rc = HYPERVISOR_sched_op_compat(SCHEDOP_block, 0); +#endif + + return rc; +} + +static inline void __noreturn +HYPERVISOR_shutdown( + unsigned int reason) +{ + struct sched_shutdown sched_shutdown = { + .reason = reason + }; + + VOID(HYPERVISOR_sched_op(SCHEDOP_shutdown, &sched_shutdown)); +#if CONFIG_XEN_COMPAT <= 0x030002 + VOID(HYPERVISOR_sched_op_compat(SCHEDOP_shutdown, reason)); +#endif + /* Don't recurse needlessly. */ + BUG_ON(reason != SHUTDOWN_crash); + for(;;); +} + +static inline int __must_check +HYPERVISOR_poll( + evtchn_port_t *ports, unsigned int nr_ports, u64 timeout) +{ + int rc; + struct sched_poll sched_poll = { + .nr_ports = nr_ports, + .timeout = jiffies_to_st(timeout) + }; + set_xen_guest_handle(sched_poll.ports, ports); + + rc = HYPERVISOR_sched_op(SCHEDOP_poll, &sched_poll); +#if CONFIG_XEN_COMPAT <= 0x030002 + if (rc == -ENOSYS) + rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0); +#endif + + return rc; +} + +static inline int __must_check +HYPERVISOR_poll_no_timeout( + evtchn_port_t *ports, unsigned int nr_ports) +{ + int rc; + struct sched_poll sched_poll = { + .nr_ports = nr_ports + }; + set_xen_guest_handle(sched_poll.ports, ports); + + rc = HYPERVISOR_sched_op(SCHEDOP_poll, &sched_poll); +#if CONFIG_XEN_COMPAT <= 0x030002 + if (rc == -ENOSYS) + rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0); +#endif + + return rc; +} + +#ifdef CONFIG_XEN + +static inline void +MULTI_update_va_mapping( + multicall_entry_t *mcl, unsigned long va, + pte_t new_val, unsigned long flags) +{ + mcl->op = __HYPERVISOR_update_va_mapping; + mcl->args[0] = va; +#if defined(CONFIG_X86_64) + mcl->args[1] = new_val.pte; +#elif defined(CONFIG_X86_PAE) + mcl->args[1] = new_val.pte_low; + mcl->args[2] = new_val.pte_high; +#else + mcl->args[1] = new_val.pte_low; + mcl->args[2] = 0; +#endif + mcl->args[MULTI_UVMFLAGS_INDEX] = flags; +} + +static inline void +MULTI_mmu_update(multicall_entry_t *mcl, mmu_update_t *req, + unsigned int count, unsigned int *success_count, + domid_t domid) +{ + mcl->op = __HYPERVISOR_mmu_update; + mcl->args[0] = (unsigned long)req; + mcl->args[1] = count; + mcl->args[2] = (unsigned long)success_count; + mcl->args[3] = domid; +} + +static inline void +MULTI_grant_table_op(multicall_entry_t *mcl, unsigned int cmd, + void *uop, unsigned int count) +{ + mcl->op = __HYPERVISOR_grant_table_op; + mcl->args[0] = cmd; + mcl->args[1] = (unsigned long)uop; + mcl->args[2] = count; +} + +#else /* !defined(CONFIG_XEN) */ + +/* Multicalls not supported for HVM guests. */ +#define MULTI_update_va_mapping(a,b,c,d) ((void)0) +#define MULTI_grant_table_op(a,b,c,d) ((void)0) + +#endif + +#define uvm_multi(cpumask) ((unsigned long)cpumask_bits(cpumask) | UVMF_MULTI) + +#ifdef LINUX +/* drivers/staging/ use Windows-style types, including VOID */ +#undef VOID +#endif + +#endif /* __HYPERVISOR_H__ */ diff --git a/include/xenpvdrivers/asm/maddr.h b/include/xenpvdrivers/asm/maddr.h new file mode 100644 index 0000000..934054c --- /dev/null +++ b/include/xenpvdrivers/asm/maddr.h @@ -0,0 +1,5 @@ +#ifdef CONFIG_X86_32 +# include "maddr_32.h" +#else +# include "maddr_64.h" +#endif diff --git a/include/xenpvdrivers/asm/maddr_32.h b/include/xenpvdrivers/asm/maddr_32.h new file mode 100644 index 0000000..a792cf3 --- /dev/null +++ b/include/xenpvdrivers/asm/maddr_32.h @@ -0,0 +1,182 @@ +#ifndef _I386_MADDR_H +#define _I386_MADDR_H + +#include +#include +#include + +/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/ +#define INVALID_P2M_ENTRY (~0UL) +#define FOREIGN_FRAME_BIT (1UL<<31) +#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT) + +/* Definitions for machine and pseudophysical addresses. */ +#ifdef CONFIG_X86_PAE +typedef unsigned long long paddr_t; +typedef unsigned long long maddr_t; +#else +typedef unsigned long paddr_t; +typedef unsigned long maddr_t; +#endif + +#ifdef CONFIG_XEN + +extern unsigned long *phys_to_machine_mapping; +extern unsigned long max_mapnr; + +#undef machine_to_phys_mapping +extern unsigned long *machine_to_phys_mapping; +extern unsigned int machine_to_phys_order; + +static inline unsigned long pfn_to_mfn(unsigned long pfn) +{ + if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) + return pfn; + if (likely(max_mapnr)) + BUG_ON(pfn >= max_mapnr); + return phys_to_machine_mapping[pfn] & ~FOREIGN_FRAME_BIT; +} + +static inline int phys_to_machine_mapping_valid(unsigned long pfn) +{ + if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) + return 1; + if (likely(max_mapnr)) + BUG_ON(pfn >= max_mapnr); + return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY); +} + +static inline unsigned long mfn_to_pfn(unsigned long mfn) +{ + unsigned long pfn; + + if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) + return mfn; + + if (unlikely((mfn >> machine_to_phys_order) != 0)) + return max_mapnr; + + /* The array access can fail (e.g., device space beyond end of RAM). */ + asm ( + "1: movl %1,%0\n" + "2:\n" + ".section .fixup,\"ax\"\n" + "3: movl %2,%0\n" + " jmp 2b\n" + ".previous\n" + ".section __ex_table,\"a\"\n" + " .align 4\n" + " .long 1b,3b\n" + ".previous" + : "=r" (pfn) + : "m" (machine_to_phys_mapping[mfn]), "m" (max_mapnr) ); + + return pfn; +} + +/* + * We detect special mappings in one of two ways: + * 1. If the MFN is an I/O page then Xen will set the m2p entry + * to be outside our maximum possible pseudophys range. + * 2. If the MFN belongs to a different domain then we will certainly + * not have MFN in our p2m table. Conversely, if the page is ours, + * then we'll have p2m(m2p(MFN))==MFN. + * If we detect a special mapping then it doesn't have a 'struct page'. + * We force !pfn_valid() by returning an out-of-range pointer. + * + * NB. These checks require that, for any MFN that is not in our reservation, + * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if + * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN. + * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety. + * + * NB2. When deliberately mapping foreign pages into the p2m table, you *must* + * use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we + * require. In all the cases we care about, the FOREIGN_FRAME bit is + * masked (e.g., pfn_to_mfn()) so behaviour there is correct. + */ +static inline unsigned long mfn_to_local_pfn(unsigned long mfn) +{ + unsigned long pfn = mfn_to_pfn(mfn); + if (likely(pfn < max_mapnr) + && likely(!xen_feature(XENFEAT_auto_translated_physmap)) + && unlikely(phys_to_machine_mapping[pfn] != mfn)) + return max_mapnr; /* force !pfn_valid() */ + return pfn; +} + +static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn) +{ + if (likely(max_mapnr)) + BUG_ON(pfn >= max_mapnr); + if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) { + BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY); + return; + } + phys_to_machine_mapping[pfn] = mfn; +} + +static inline maddr_t phys_to_machine(paddr_t phys) +{ + maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT); + machine = (machine << PAGE_SHIFT) | (phys & ~PAGE_MASK); + return machine; +} + +static inline paddr_t machine_to_phys(maddr_t machine) +{ + paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT); + phys = (phys << PAGE_SHIFT) | (machine & ~PAGE_MASK); + return phys; +} + +#ifdef CONFIG_X86_PAE +static inline paddr_t pte_phys_to_machine(paddr_t phys) +{ + /* + * In PAE mode, the NX bit needs to be dealt with in the value + * passed to pfn_to_mfn(). On x86_64, we need to mask it off, + * but for i386 the conversion to ulong for the argument will + * clip it off. + */ + maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT); + machine = (machine << PAGE_SHIFT) | (phys & ~PHYSICAL_PAGE_MASK); + return machine; +} + +static inline paddr_t pte_machine_to_phys(maddr_t machine) +{ + /* + * In PAE mode, the NX bit needs to be dealt with in the value + * passed to mfn_to_pfn(). On x86_64, we need to mask it off, + * but for i386 the conversion to ulong for the argument will + * clip it off. + */ + paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT); + phys = (phys << PAGE_SHIFT) | (machine & ~PHYSICAL_PAGE_MASK); + return phys; +} +#else +#define pte_phys_to_machine phys_to_machine +#define pte_machine_to_phys machine_to_phys +#endif + +#else /* !CONFIG_XEN */ + +#define pfn_to_mfn(pfn) (pfn) +#define mfn_to_pfn(mfn) (mfn) +#define mfn_to_local_pfn(mfn) (mfn) +#define set_phys_to_machine(pfn, mfn) ((void)0) +#define phys_to_machine_mapping_valid(pfn) (1) +#define phys_to_machine(phys) ((maddr_t)(phys)) +#define machine_to_phys(mach) ((paddr_t)(mach)) +#define pfn_pte_ma(pfn, prot) pfn_pte(pfn, prot) +#define __pte_ma(x) __pte(x) + +#endif /* !CONFIG_XEN */ + +/* VIRT <-> MACHINE conversion */ +#define virt_to_machine(v) (phys_to_machine(__pa(v))) +#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT)) +#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT)) + +#endif /* _I386_MADDR_H */ diff --git a/include/xenpvdrivers/asm/maddr_64.h b/include/xenpvdrivers/asm/maddr_64.h new file mode 100644 index 0000000..73c7c2c --- /dev/null +++ b/include/xenpvdrivers/asm/maddr_64.h @@ -0,0 +1,163 @@ +#ifndef _X86_64_MADDR_H +#define _X86_64_MADDR_H + +#include +#include +#include + +/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/ +#define INVALID_P2M_ENTRY (~0UL) +#define FOREIGN_FRAME_BIT (1UL<<63) +#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT) + +/* Definitions for machine and pseudophysical addresses. */ +typedef unsigned long paddr_t; +typedef unsigned long maddr_t; + +#ifdef CONFIG_XEN + +extern unsigned long *phys_to_machine_mapping; +extern unsigned long max_mapnr; + +#undef machine_to_phys_mapping +extern unsigned long *machine_to_phys_mapping; +extern unsigned int machine_to_phys_order; + +static inline unsigned long pfn_to_mfn(unsigned long pfn) +{ + if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) + return pfn; + if (likely(max_mapnr)) + BUG_ON(pfn >= max_mapnr); + return phys_to_machine_mapping[pfn] & ~FOREIGN_FRAME_BIT; +} + +static inline int phys_to_machine_mapping_valid(unsigned long pfn) +{ + if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) + return 1; + if (likely(max_mapnr)) + BUG_ON(pfn >= max_mapnr); + return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY); +} + +static inline unsigned long mfn_to_pfn(unsigned long mfn) +{ + unsigned long pfn; + + if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) + return mfn; + + if (unlikely((mfn >> machine_to_phys_order) != 0)) + return max_mapnr; + + /* The array access can fail (e.g., device space beyond end of RAM). */ + asm ( + "1: movq %1,%0\n" + "2:\n" + ".section .fixup,\"ax\"\n" + "3: movq %2,%0\n" + " jmp 2b\n" + ".previous\n" + ".section __ex_table,\"a\"\n" + " .align 8\n" + " .quad 1b,3b\n" + ".previous" + : "=r" (pfn) + : "m" (machine_to_phys_mapping[mfn]), "m" (max_mapnr) ); + + return pfn; +} + +/* + * We detect special mappings in one of two ways: + * 1. If the MFN is an I/O page then Xen will set the m2p entry + * to be outside our maximum possible pseudophys range. + * 2. If the MFN belongs to a different domain then we will certainly + * not have MFN in our p2m table. Conversely, if the page is ours, + * then we'll have p2m(m2p(MFN))==MFN. + * If we detect a special mapping then it doesn't have a 'struct page'. + * We force !pfn_valid() by returning an out-of-range pointer. + * + * NB. These checks require that, for any MFN that is not in our reservation, + * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if + * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN. + * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety. + * + * NB2. When deliberately mapping foreign pages into the p2m table, you *must* + * use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we + * require. In all the cases we care about, the FOREIGN_FRAME bit is + * masked (e.g., pfn_to_mfn()) so behaviour there is correct. + */ +static inline unsigned long mfn_to_local_pfn(unsigned long mfn) +{ + unsigned long pfn = mfn_to_pfn(mfn); + if (likely(pfn < max_mapnr) + && likely(!xen_feature(XENFEAT_auto_translated_physmap)) + && unlikely(phys_to_machine_mapping[pfn] != mfn)) + return max_mapnr; /* force !pfn_valid() */ + return pfn; +} + +static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn) +{ + if (likely(max_mapnr)) + BUG_ON(pfn >= max_mapnr); + if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) { + BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY); + return; + } + phys_to_machine_mapping[pfn] = mfn; +} + +static inline maddr_t phys_to_machine(paddr_t phys) +{ + maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT); + machine = (machine << PAGE_SHIFT) | (phys & ~PAGE_MASK); + return machine; +} + +static inline paddr_t machine_to_phys(maddr_t machine) +{ + paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT); + phys = (phys << PAGE_SHIFT) | (machine & ~PAGE_MASK); + return phys; +} + +static inline paddr_t pte_phys_to_machine(paddr_t phys) +{ + maddr_t machine; + machine = pfn_to_mfn((phys & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT); + machine = (machine << PAGE_SHIFT) | (phys & ~PHYSICAL_PAGE_MASK); + return machine; +} + +static inline paddr_t pte_machine_to_phys(maddr_t machine) +{ + paddr_t phys; + phys = mfn_to_pfn((machine & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT); + phys = (phys << PAGE_SHIFT) | (machine & ~PHYSICAL_PAGE_MASK); + return phys; +} + +#else /* !CONFIG_XEN */ + +#define pfn_to_mfn(pfn) (pfn) +#define mfn_to_pfn(mfn) (mfn) +#define mfn_to_local_pfn(mfn) (mfn) +#define set_phys_to_machine(pfn, mfn) ((void)0) +#define phys_to_machine_mapping_valid(pfn) (1) +#define phys_to_machine(phys) ((maddr_t)(phys)) +#define machine_to_phys(mach) ((paddr_t)(mach)) +#define pfn_pte_ma(pfn, prot) pfn_pte(pfn, prot) +#define __pte_ma(x) __pte(x) + +#endif /* !CONFIG_XEN */ + +/* VIRT <-> MACHINE conversion */ +#define virt_to_machine(v) (phys_to_machine(__pa(v))) +#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT)) +#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT)) + +#endif /* _X86_64_MADDR_H */ + diff --git a/include/xenpvdrivers/asm/synch_bitops.h b/include/xenpvdrivers/asm/synch_bitops.h new file mode 100644 index 0000000..b46eb9f --- /dev/null +++ b/include/xenpvdrivers/asm/synch_bitops.h @@ -0,0 +1,126 @@ +#ifndef __XEN_SYNCH_BITOPS_H__ +#define __XEN_SYNCH_BITOPS_H__ + +/* + * Copyright 1992, Linus Torvalds. + * Heavily modified to provide guaranteed strong synchronisation + * when communicating with Xen or other guest OSes running on other CPUs. + */ + +#ifdef HAVE_XEN_PLATFORM_COMPAT_H +#include +#endif + +#define ADDR (*(volatile long *) addr) + +static __inline__ void synch_set_bit(int nr, volatile void * addr) +{ + __asm__ __volatile__ ( + "lock btsl %1,%0" + : "+m" (ADDR) : "Ir" (nr) : "memory" ); +} + +static __inline__ void synch_clear_bit(int nr, volatile void * addr) +{ + __asm__ __volatile__ ( + "lock btrl %1,%0" + : "+m" (ADDR) : "Ir" (nr) : "memory" ); +} + +static __inline__ void synch_change_bit(int nr, volatile void * addr) +{ + __asm__ __volatile__ ( + "lock btcl %1,%0" + : "+m" (ADDR) : "Ir" (nr) : "memory" ); +} + +static __inline__ int synch_test_and_set_bit(int nr, volatile void * addr) +{ + int oldbit; + __asm__ __volatile__ ( + "lock btsl %2,%1\n\tsbbl %0,%0" + : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory"); + return oldbit; +} + +static __inline__ int synch_test_and_clear_bit(int nr, volatile void * addr) +{ + int oldbit; + __asm__ __volatile__ ( + "lock btrl %2,%1\n\tsbbl %0,%0" + : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory"); + return oldbit; +} + +static __inline__ int synch_test_and_change_bit(int nr, volatile void * addr) +{ + int oldbit; + + __asm__ __volatile__ ( + "lock btcl %2,%1\n\tsbbl %0,%0" + : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory"); + return oldbit; +} + +struct __synch_xchg_dummy { unsigned long a[100]; }; +#define __synch_xg(x) ((struct __synch_xchg_dummy *)(x)) + +#define synch_cmpxchg(ptr, old, new) \ +((__typeof__(*(ptr)))__synch_cmpxchg((ptr),\ + (unsigned long)(old), \ + (unsigned long)(new), \ + sizeof(*(ptr)))) + +static inline unsigned long __synch_cmpxchg(volatile void *ptr, + unsigned long old, + unsigned long new, int size) +{ + unsigned long prev; + switch (size) { + case 1: + __asm__ __volatile__("lock; cmpxchgb %b1,%2" + : "=a"(prev) + : "q"(new), "m"(*__synch_xg(ptr)), + "0"(old) + : "memory"); + return prev; + case 2: + __asm__ __volatile__("lock; cmpxchgw %w1,%2" + : "=a"(prev) + : "r"(new), "m"(*__synch_xg(ptr)), + "0"(old) + : "memory"); + return prev; +#ifdef CONFIG_X86_64 + case 4: + __asm__ __volatile__("lock; cmpxchgl %k1,%2" + : "=a"(prev) + : "r"(new), "m"(*__synch_xg(ptr)), + "0"(old) + : "memory"); + return prev; + case 8: + __asm__ __volatile__("lock; cmpxchgq %1,%2" + : "=a"(prev) + : "r"(new), "m"(*__synch_xg(ptr)), + "0"(old) + : "memory"); + return prev; +#else + case 4: + __asm__ __volatile__("lock; cmpxchgl %1,%2" + : "=a"(prev) + : "r"(new), "m"(*__synch_xg(ptr)), + "0"(old) + : "memory"); + return prev; +#endif + } + return old; +} + +#define synch_test_bit test_bit + +#define synch_cmpxchg_subword synch_cmpxchg + +#endif /* __XEN_SYNCH_BITOPS_H__ */ diff --git a/include/xenpvdrivers/balloon.h b/include/xenpvdrivers/balloon.h new file mode 100644 index 0000000..ae162b4 --- /dev/null +++ b/include/xenpvdrivers/balloon.h @@ -0,0 +1,65 @@ +/****************************************************************************** + * balloon.h + * + * Xen balloon driver - enables returning/claiming memory to/from Xen. + * + * Copyright (c) 2003, B Dragovic + * Copyright (c) 2003-2004, M Williamson, K Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __XEN_BALLOON_H__ +#define __XEN_BALLOON_H__ + +#include + +#if !defined(CONFIG_PARAVIRT_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H) +/* + * Inform the balloon driver that it should allow some slop for device-driver + * memory activities. + */ +void balloon_update_driver_allowance(long delta); + +/* Allocate/free a set of empty pages in low memory (i.e., no RAM mapped). */ +struct page **alloc_empty_pages_and_pagevec(int nr_pages); +void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages); + +/* Free an empty page range (not allocated through + alloc_empty_pages_and_pagevec), adding to the balloon. */ +void free_empty_pages(struct page **pagevec, int nr_pages); + +void balloon_release_driver_page(struct page *page); + +/* + * Prevent the balloon driver from changing the memory reservation during + * a driver critical region. + */ +extern spinlock_t balloon_lock; +#define balloon_lock(__flags) spin_lock_irqsave(&balloon_lock, __flags) +#define balloon_unlock(__flags) spin_unlock_irqrestore(&balloon_lock, __flags) +#endif + +#endif /* __XEN_BALLOON_H__ */ diff --git a/include/xenpvdrivers/blkif.h b/include/xenpvdrivers/blkif.h new file mode 100644 index 0000000..7284e92 --- /dev/null +++ b/include/xenpvdrivers/blkif.h @@ -0,0 +1,123 @@ +/* + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_BLKIF_H__ +#define __XEN_BLKIF_H__ + +#include +#include +#include + +/* Not a real protocol. Used to generate ring structs which contain + * the elements common to all protocols only. This way we get a + * compiler-checkable way to use common struct elements, so we can + * avoid using switch(protocol) in a number of places. */ +struct blkif_common_request { + char dummy; +}; +struct blkif_common_response { + char dummy; +}; + +/* i386 protocol version */ +#pragma pack(push, 4) +struct blkif_x86_32_request { + uint8_t operation; /* BLKIF_OP_??? */ + uint8_t nr_segments; /* number of segments */ + blkif_vdev_t handle; /* only for read/write requests */ + uint64_t id; /* private guest value, echoed in resp */ + blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */ + struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST]; +}; +struct blkif_x86_32_response { + uint64_t id; /* copied from request */ + uint8_t operation; /* copied from request */ + int16_t status; /* BLKIF_RSP_??? */ +}; +typedef struct blkif_x86_32_request blkif_x86_32_request_t; +typedef struct blkif_x86_32_response blkif_x86_32_response_t; +#pragma pack(pop) + +/* x86_64 protocol version */ +struct blkif_x86_64_request { + uint8_t operation; /* BLKIF_OP_??? */ + uint8_t nr_segments; /* number of segments */ + blkif_vdev_t handle; /* only for read/write requests */ + uint64_t __attribute__((__aligned__(8))) id; + blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */ + struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST]; +}; +struct blkif_x86_64_response { + uint64_t __attribute__((__aligned__(8))) id; + uint8_t operation; /* copied from request */ + int16_t status; /* BLKIF_RSP_??? */ +}; +typedef struct blkif_x86_64_request blkif_x86_64_request_t; +typedef struct blkif_x86_64_response blkif_x86_64_response_t; + +DEFINE_RING_TYPES(blkif_common, struct blkif_common_request, struct blkif_common_response); +DEFINE_RING_TYPES(blkif_x86_32, struct blkif_x86_32_request, struct blkif_x86_32_response); +DEFINE_RING_TYPES(blkif_x86_64, struct blkif_x86_64_request, struct blkif_x86_64_response); + +union blkif_back_rings { + blkif_back_ring_t native; + blkif_common_back_ring_t common; + blkif_x86_32_back_ring_t x86_32; + blkif_x86_64_back_ring_t x86_64; +}; +typedef union blkif_back_rings blkif_back_rings_t; + +enum blkif_protocol { + BLKIF_PROTOCOL_NATIVE = 1, + BLKIF_PROTOCOL_X86_32 = 2, + BLKIF_PROTOCOL_X86_64 = 3, +}; + +static void inline blkif_get_x86_32_req(blkif_request_t *dst, blkif_x86_32_request_t *src) +{ + int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST; + dst->operation = src->operation; + dst->nr_segments = src->nr_segments; + dst->handle = src->handle; + dst->id = src->id; + dst->sector_number = src->sector_number; + barrier(); + if (n > dst->nr_segments) + n = dst->nr_segments; + for (i = 0; i < n; i++) + dst->seg[i] = src->seg[i]; +} + +static void inline blkif_get_x86_64_req(blkif_request_t *dst, blkif_x86_64_request_t *src) +{ + int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST; + dst->operation = src->operation; + dst->nr_segments = src->nr_segments; + dst->handle = src->handle; + dst->id = src->id; + dst->sector_number = src->sector_number; + barrier(); + if (n > dst->nr_segments) + n = dst->nr_segments; + for (i = 0; i < n; i++) + dst->seg[i] = src->seg[i]; +} + +#endif /* __XEN_BLKIF_H__ */ diff --git a/include/xenpvdrivers/compat_ioctl.h b/include/xenpvdrivers/compat_ioctl.h new file mode 100644 index 0000000..366e3b6 --- /dev/null +++ b/include/xenpvdrivers/compat_ioctl.h @@ -0,0 +1,75 @@ +/* + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation; either version 2 of the + * License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * + * Copyright IBM Corp. 2007 + * + * Authors: Jimi Xenidis + * Hollis Blanchard + */ + +#ifndef __LINUX_XEN_COMPAT_H__ +#define __LINUX_XEN_COMPAT_H__ + +#include +#include + +#if defined(CONFIG_X86) || defined(CONFIG_IA64) +#define xen_pfn32_t __u32 +#endif + +extern int privcmd_ioctl_32(int fd, unsigned int cmd, unsigned long arg); +struct privcmd_mmap_32 { + int num; + domid_t dom; + compat_uptr_t entry; +}; + +struct privcmd_mmapbatch_32 { + int num; /* number of pages to populate */ + domid_t dom; /* target domain */ +#if defined(CONFIG_X86) || defined(CONFIG_IA64) + union { /* virtual address */ + __u64 addr __attribute__((packed)); + __u32 va; /* ensures union is 4-byte aligned */ + }; +#else + __u64 addr; /* virtual address */ +#endif + compat_uptr_t arr; /* array of mfns - top nibble set on err */ +}; + +struct privcmd_mmapbatch_v2_32 { + unsigned int num; /* number of pages to populate */ + domid_t dom; /* target domain */ +#if defined(CONFIG_X86) || defined(CONFIG_IA64) + union { /* virtual address */ + __u64 addr __attribute__((packed)); + __u32 va; /* ensures union is 4-byte aligned */ + }; +#else + __u64 addr; /* virtual address */ +#endif + compat_uptr_t arr; /* array of mfns */ + compat_uptr_t err; /* array of error codes */ +}; + +#define IOCTL_PRIVCMD_MMAP_32 \ + _IOC(_IOC_NONE, 'P', 2, sizeof(struct privcmd_mmap_32)) +#define IOCTL_PRIVCMD_MMAPBATCH_32 \ + _IOC(_IOC_NONE, 'P', 3, sizeof(struct privcmd_mmapbatch_32)) +#define IOCTL_PRIVCMD_MMAPBATCH_V2_32 \ + _IOC(_IOC_NONE, 'P', 4, sizeof(struct privcmd_mmapbatch_v2_32)) + +#endif /* __LINUX_XEN_COMPAT_H__ */ diff --git a/include/xenpvdrivers/cpu_hotplug.h b/include/xenpvdrivers/cpu_hotplug.h new file mode 100644 index 0000000..429cf64 --- /dev/null +++ b/include/xenpvdrivers/cpu_hotplug.h @@ -0,0 +1,39 @@ +#ifndef __XEN_CPU_HOTPLUG_H__ +#define __XEN_CPU_HOTPLUG_H__ + +#include +#include + +#if defined(CONFIG_X86) && defined(CONFIG_SMP) +extern cpumask_var_t vcpu_initialized_mask; +#endif + +#if defined(CONFIG_HOTPLUG_CPU) + +int cpu_up_check(unsigned int cpu); +void init_xenbus_allowed_cpumask(void); +int smp_suspend(void); +void smp_resume(void); + +#else /* !defined(CONFIG_HOTPLUG_CPU) */ + +#define cpu_up_check(cpu) (0) +#define init_xenbus_allowed_cpumask() ((void)0) + +static inline int smp_suspend(void) +{ + if (num_online_cpus() > 1) { + printk(KERN_WARNING "Can't suspend SMP guests " + "without CONFIG_HOTPLUG_CPU\n"); + return -EOPNOTSUPP; + } + return 0; +} + +static inline void smp_resume(void) +{ +} + +#endif /* !defined(CONFIG_HOTPLUG_CPU) */ + +#endif /* __XEN_CPU_HOTPLUG_H__ */ diff --git a/include/xenpvdrivers/driver_util.h b/include/xenpvdrivers/driver_util.h new file mode 100644 index 0000000..30a6f0e --- /dev/null +++ b/include/xenpvdrivers/driver_util.h @@ -0,0 +1,10 @@ + +#ifndef __ASM_XEN_DRIVER_UTIL_H__ +#define __ASM_XEN_DRIVER_UTIL_H__ + +#include +#include + +extern struct class *get_xen_class(void); + +#endif /* __ASM_XEN_DRIVER_UTIL_H__ */ diff --git a/include/xenpvdrivers/events.h b/include/xenpvdrivers/events.h new file mode 100644 index 0000000..2ad4599 --- /dev/null +++ b/include/xenpvdrivers/events.h @@ -0,0 +1,59 @@ +#ifndef _XEN_EVENTS_H +#define _XEN_EVENTS_H + +#include + +#include +#include +#include + +int bind_evtchn_to_irq(unsigned int evtchn); +int bind_evtchn_to_irqhandler(unsigned int evtchn, + irq_handler_t handler, + unsigned long irqflags, const char *devname, + void *dev_id); +int bind_virq_to_irqhandler(unsigned int virq, unsigned int cpu, + irq_handler_t handler, + unsigned long irqflags, const char *devname, + void *dev_id); +int bind_ipi_to_irqhandler(enum ipi_vector ipi, + unsigned int cpu, + irq_handler_t handler, + unsigned long irqflags, + const char *devname, + void *dev_id); + +/* + * Common unbind function for all event sources. Takes IRQ to unbind from. + * Automatically closes the underlying event channel (even for bindings + * made with bind_evtchn_to_irqhandler()). + */ +void unbind_from_irqhandler(unsigned int irq, void *dev_id); + +void xen_send_IPI_one(unsigned int cpu, enum ipi_vector vector); +int resend_irq_on_evtchn(unsigned int irq); +void rebind_evtchn_irq(int evtchn, int irq); + +static inline void notify_remote_via_evtchn(int port) +{ + struct evtchn_send send = { .port = port }; + (void)HYPERVISOR_event_channel_op(EVTCHNOP_send, &send); +} + +extern void notify_remote_via_irq(int irq); + +extern void xen_irq_resume(void); + +/* Clear an irq's pending state, in preparation for polling on it */ +void xen_clear_irq_pending(int irq); +void xen_set_irq_pending(int irq); +bool xen_test_irq_pending(int irq); + +/* Poll waiting for an irq to become pending. In the usual case, the + irq will be disabled so it won't deliver an interrupt. */ +void xen_poll_irq(int irq); + +/* Determine the IRQ which is bound to an event channel */ +unsigned irq_from_evtchn(unsigned int evtchn); + +#endif /* _XEN_EVENTS_H */ diff --git a/include/xenpvdrivers/evtchn.h b/include/xenpvdrivers/evtchn.h new file mode 100644 index 0000000..4545e5f --- /dev/null +++ b/include/xenpvdrivers/evtchn.h @@ -0,0 +1,234 @@ +#if defined(CONFIG_PARAVIRT_XEN) || !defined(__KERNEL__) +#include "public/evtchn.h" +#else +/****************************************************************************** + * evtchn.h + * + * Communication via Xen event channels. + * Also definitions for the device that demuxes notifications to userspace. + * + * Copyright (c) 2003-2005, K A Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __ASM_EVTCHN_H__ +#define __ASM_EVTCHN_H__ + +#include +#include +#include +#include +#include +#include + +/* + * LOW-LEVEL DEFINITIONS + */ +struct irq_cfg { + u32 info; + union { + int bindcount; /* for dynamic IRQs */ +#ifdef CONFIG_X86_IO_APIC + u8 vector; /* for physical IRQs */ +#endif + }; +}; + +int assign_irq_vector(int irq, struct irq_cfg *, const struct cpumask *); + +/* + * Dynamically bind an event source to an IRQ-like callback handler. + * On some platforms this may not be implemented via the Linux IRQ subsystem. + * The IRQ argument passed to the callback handler is the same as returned + * from the bind call. It may not correspond to a Linux IRQ number. + * Returns IRQ or negative errno. + */ +int bind_caller_port_to_irqhandler( + unsigned int caller_port, + irq_handler_t handler, + unsigned long irqflags, + const char *devname, + void *dev_id); +int bind_listening_port_to_irqhandler( + unsigned int remote_domain, + irq_handler_t handler, + unsigned long irqflags, + const char *devname, + void *dev_id); +int bind_interdomain_evtchn_to_irqhandler( + unsigned int remote_domain, + unsigned int remote_port, + irq_handler_t handler, + unsigned long irqflags, + const char *devname, + void *dev_id); +int bind_virq_to_irqhandler( + unsigned int virq, + unsigned int cpu, + irq_handler_t handler, + unsigned long irqflags, + const char *devname, + void *dev_id); +#if defined(CONFIG_SMP) && defined(CONFIG_XEN) && defined(CONFIG_X86) +int bind_virq_to_irqaction( + unsigned int virq, + unsigned int cpu, + struct irqaction *action); +#else +#define bind_virq_to_irqaction(virq, cpu, action) \ + bind_virq_to_irqhandler(virq, cpu, (action)->handler, \ + (action)->flags | IRQF_NOBALANCING, \ + (action)->name, action) +#endif +#if defined(CONFIG_SMP) && !defined(XENPV_COMPAT_MODULE) +#ifndef CONFIG_X86 +int bind_ipi_to_irqhandler( + unsigned int ipi, + unsigned int cpu, + irq_handler_t handler, + unsigned long irqflags, + const char *devname, + void *dev_id); +#else +int bind_ipi_to_irqaction( + unsigned int ipi, + unsigned int cpu, + struct irqaction *action); +#endif +#endif + +/* + * Common unbind function for all event sources. Takes IRQ to unbind from. + * Automatically closes the underlying event channel (except for bindings + * made with bind_caller_port_to_irqhandler()). + */ +void unbind_from_irqhandler(unsigned int irq, void *dev_id); + +#if defined(CONFIG_SMP) && defined(CONFIG_XEN) && defined(CONFIG_X86) +/* Specialized unbind function for per-CPU IRQs. */ +void unbind_from_per_cpu_irq(unsigned int irq, unsigned int cpu, + struct irqaction *); +#else +#define unbind_from_per_cpu_irq(irq, cpu, action) \ + unbind_from_irqhandler(irq, action) +#endif + +#ifndef CONFIG_XEN +void irq_resume(void); +#endif + +/* Entry point for notifications into Linux subsystems. */ +asmlinkage void evtchn_do_upcall(struct pt_regs *regs); + +/* Entry point for notifications into the userland character device. */ +void evtchn_device_upcall(int port); + +/* Mark a PIRQ as unavailable for dynamic allocation. */ +void evtchn_register_pirq(int irq); +/* Map a Xen-supplied PIRQ to a dynamically allocated one. */ +int evtchn_map_pirq(int irq, int xen_pirq); +/* Look up a Xen-supplied PIRQ for a dynamically allocated one. */ +int evtchn_get_xen_pirq(int irq); + +void mask_evtchn(int port); +void disable_all_local_evtchn(void); +void unmask_evtchn(int port); + +#ifdef CONFIG_SMP +void rebind_evtchn_to_cpu(int port, unsigned int cpu); +#else +#define rebind_evtchn_to_cpu(port, cpu) ((void)0) +#endif + +static inline int test_and_set_evtchn_mask(int port) +{ + shared_info_t *s = HYPERVISOR_shared_info; + return synch_test_and_set_bit(port, s->evtchn_mask); +} + +static inline void clear_evtchn(int port) +{ + shared_info_t *s = HYPERVISOR_shared_info; + synch_clear_bit(port, s->evtchn_pending); +} + +static inline void set_evtchn(int port) +{ + shared_info_t *s = HYPERVISOR_shared_info; + synch_set_bit(port, s->evtchn_pending); +} + +static inline int test_evtchn(int port) +{ + shared_info_t *s = HYPERVISOR_shared_info; + return synch_test_bit(port, s->evtchn_pending); +} + +static inline void notify_remote_via_evtchn(int port) +{ + struct evtchn_send send = { .port = port }; + VOID(HYPERVISOR_event_channel_op(EVTCHNOP_send, &send)); +} + +static inline void +multi_notify_remote_via_evtchn(multicall_entry_t *mcl, int port) +{ + struct evtchn_send *send = (void *)(mcl->args + 2); + + BUILD_BUG_ON(sizeof(*send) > sizeof(mcl->args) - 2 * sizeof(*mcl->args)); + send->port = port; + mcl->op = __HYPERVISOR_event_channel_op; + mcl->args[0] = EVTCHNOP_send; + mcl->args[1] = (unsigned long)send; +} + +/* Clear an irq's pending state, in preparation for polling on it. */ +void xen_clear_irq_pending(int irq); + +/* Set an irq's pending state, to avoid blocking on it. */ +void xen_set_irq_pending(int irq); + +/* Test an irq's pending state. */ +int xen_test_irq_pending(int irq); + +/* Poll waiting for an irq to become pending. In the usual case, the + irq will be disabled so it won't deliver an interrupt. */ +void xen_poll_irq(int irq); + +/* + * Use these to access the event channel underlying the IRQ handle returned + * by bind_*_to_irqhandler(). + */ +void notify_remote_via_irq(int irq); +int multi_notify_remote_via_irq(multicall_entry_t *, int irq); +int irq_to_evtchn_port(int irq); + +#if defined(CONFIG_SMP) && !defined(XENPV_COMPAT_MODULE) && defined(CONFIG_X86) +void notify_remote_via_ipi(unsigned int ipi, unsigned int cpu); +#endif + +#endif /* __ASM_EVTCHN_H__ */ +#endif /* CONFIG_PARAVIRT_XEN */ diff --git a/include/xenpvdrivers/features.h b/include/xenpvdrivers/features.h new file mode 100644 index 0000000..b92f474 --- /dev/null +++ b/include/xenpvdrivers/features.h @@ -0,0 +1,24 @@ +/****************************************************************************** + * features.h + * + * Query the features reported by Xen. + * + * Copyright (c) 2006, Ian Campbell + */ + +#ifndef __XEN_FEATURES_H__ +#define __XEN_FEATURES_H__ + +#include +#include + +void xen_setup_features(void); + +extern u8 xen_features[XENFEAT_NR_SUBMAPS * 32]; + +static inline int xen_feature(int flag) +{ + return xen_features[flag]; +} + +#endif /* __XEN_FEATURES_H__ */ diff --git a/include/xenpvdrivers/firmware.h b/include/xenpvdrivers/firmware.h new file mode 100644 index 0000000..3be378c --- /dev/null +++ b/include/xenpvdrivers/firmware.h @@ -0,0 +1,14 @@ +#ifndef __XEN_FIRMWARE_H__ +#define __XEN_FIRMWARE_H__ + +#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE) +void copy_edd(void); +#endif + +#ifdef CONFIG_XEN_PRIVILEGED_GUEST +void copy_edid(void); +#else +static inline void copy_edid(void) {} +#endif + +#endif /* __XEN_FIRMWARE_H__ */ diff --git a/include/xenpvdrivers/gnttab.h b/include/xenpvdrivers/gnttab.h new file mode 100644 index 0000000..26e60e6 --- /dev/null +++ b/include/xenpvdrivers/gnttab.h @@ -0,0 +1,169 @@ +/****************************************************************************** + * gnttab.h + * + * Two sets of functionality: + * 1. Granting foreign access to our memory reservation. + * 2. Accessing others' memory reservations via grant references. + * (i.e., mechanisms for both sender and recipient of grant references) + * + * Copyright (c) 2004-2005, K A Fraser + * Copyright (c) 2005, Christopher Clark + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __ASM_GNTTAB_H__ +#define __ASM_GNTTAB_H__ + +#include +#include /* maddr_t */ +#include +#include +#include + +struct gnttab_free_callback { + struct gnttab_free_callback *next; + void (*fn)(void *); + void *arg; + u16 count; + u8 queued; +}; + +int gnttab_grant_foreign_access(domid_t domid, unsigned long frame, + int flags); + +/* + * End access through the given grant reference, iff the grant entry is no + * longer in use. Return 1 if the grant entry was freed, 0 if it is still in + * use. + */ +int gnttab_end_foreign_access_ref(grant_ref_t ref); + +/* + * Eventually end access through the given grant reference, and once that + * access has been ended, free the given page too. Access will be ended + * immediately iff the grant entry is not in use, otherwise it will happen + * some time later. page may be 0, in which case no freeing will occur. + */ +void gnttab_end_foreign_access(grant_ref_t ref, unsigned long page); + +int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn); + +unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref); +unsigned long gnttab_end_foreign_transfer(grant_ref_t ref); + +int gnttab_query_foreign_access(grant_ref_t ref); + +/* + * operations on reserved batches of grant references + */ +int gnttab_alloc_grant_references(u16 count, grant_ref_t *pprivate_head); + +void gnttab_free_grant_reference(grant_ref_t ref); + +void gnttab_free_grant_references(grant_ref_t head); + +int gnttab_empty_grant_references(const grant_ref_t *pprivate_head); + +int gnttab_claim_grant_reference(grant_ref_t *pprivate_head); + +void gnttab_release_grant_reference(grant_ref_t *private_head, + grant_ref_t release); + +void gnttab_request_free_callback(struct gnttab_free_callback *callback, + void (*fn)(void *), void *arg, u16 count); +void gnttab_cancel_free_callback(struct gnttab_free_callback *callback); + +void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid, + unsigned long frame, int flags); + +void gnttab_grant_foreign_transfer_ref(grant_ref_t, domid_t domid, + unsigned long pfn); + +int gnttab_copy_grant_page(grant_ref_t ref, struct page **pagep); +#if defined(CONFIG_XEN_BACKEND) || defined(CONFIG_XEN_BACKEND_MODULE) +void __gnttab_dma_map_page(struct page *page); +#else +#define __gnttab_dma_map_page __gnttab_dma_unmap_page +#endif +static inline void __gnttab_dma_unmap_page(struct page *page) +{ +} + +void gnttab_reset_grant_page(struct page *page); + +#ifndef CONFIG_XEN +int gnttab_resume(void); +#endif + +void *arch_gnttab_alloc_shared(unsigned long *frames); + +static inline void +gnttab_set_map_op(struct gnttab_map_grant_ref *map, maddr_t addr, + uint32_t flags, grant_ref_t ref, domid_t domid) +{ + if (flags & GNTMAP_contains_pte) + map->host_addr = addr; + else if (xen_feature(XENFEAT_auto_translated_physmap)) + map->host_addr = __pa(addr); + else + map->host_addr = addr; + + map->flags = flags; + map->ref = ref; + map->dom = domid; +} + +static inline void +gnttab_set_unmap_op(struct gnttab_unmap_grant_ref *unmap, maddr_t addr, + uint32_t flags, grant_handle_t handle) +{ + if (flags & GNTMAP_contains_pte) + unmap->host_addr = addr; + else if (xen_feature(XENFEAT_auto_translated_physmap)) + unmap->host_addr = __pa(addr); + else + unmap->host_addr = addr; + + unmap->handle = handle; + unmap->dev_bus_addr = 0; +} + +static inline void +gnttab_set_replace_op(struct gnttab_unmap_and_replace *unmap, maddr_t addr, + maddr_t new_addr, grant_handle_t handle) +{ + if (xen_feature(XENFEAT_auto_translated_physmap)) { + unmap->host_addr = __pa(addr); + unmap->new_addr = __pa(new_addr); + } else { + unmap->host_addr = addr; + unmap->new_addr = new_addr; + } + + unmap->handle = handle; +} + +#endif /* __ASM_GNTTAB_H__ */ diff --git a/include/xenpvdrivers/grant_table.h b/include/xenpvdrivers/grant_table.h new file mode 100644 index 0000000..0dea613 --- /dev/null +++ b/include/xenpvdrivers/grant_table.h @@ -0,0 +1,117 @@ +/****************************************************************************** + * grant_table.h + * + * Two sets of functionality: + * 1. Granting foreign access to our memory reservation. + * 2. Accessing others' memory reservations via grant references. + * (i.e., mechanisms for both sender and recipient of grant references) + * + * Copyright (c) 2004-2005, K A Fraser + * Copyright (c) 2005, Christopher Clark + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __ASM_GNTTAB_H__ +#define __ASM_GNTTAB_H__ + +#include +#include +#include + +/* NR_GRANT_FRAMES must be less than or equal to that configured in Xen */ +#define NR_GRANT_FRAMES 4 + +struct gnttab_free_callback { + struct gnttab_free_callback *next; + void (*fn)(void *); + void *arg; + u16 count; +}; + +int gnttab_suspend(void); +int gnttab_resume(void); + +int gnttab_grant_foreign_access(domid_t domid, unsigned long frame, + int readonly); + +/* + * End access through the given grant reference, iff the grant entry is no + * longer in use. Return 1 if the grant entry was freed, 0 if it is still in + * use. + */ +int gnttab_end_foreign_access_ref(grant_ref_t ref, int readonly); + +/* + * Eventually end access through the given grant reference, and once that + * access has been ended, free the given page too. Access will be ended + * immediately iff the grant entry is not in use, otherwise it will happen + * some time later. page may be 0, in which case no freeing will occur. + */ +void gnttab_end_foreign_access(grant_ref_t ref, int readonly, + unsigned long page); + +int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn); + +unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref); +unsigned long gnttab_end_foreign_transfer(grant_ref_t ref); + +int gnttab_query_foreign_access(grant_ref_t ref); + +/* + * operations on reserved batches of grant references + */ +int gnttab_alloc_grant_references(u16 count, grant_ref_t *pprivate_head); + +void gnttab_free_grant_reference(grant_ref_t ref); + +void gnttab_free_grant_references(grant_ref_t head); + +int gnttab_empty_grant_references(const grant_ref_t *pprivate_head); + +int gnttab_claim_grant_reference(grant_ref_t *pprivate_head); + +void gnttab_release_grant_reference(grant_ref_t *private_head, + grant_ref_t release); + +void gnttab_request_free_callback(struct gnttab_free_callback *callback, + void (*fn)(void *), void *arg, u16 count); +void gnttab_cancel_free_callback(struct gnttab_free_callback *callback); + +void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid, + unsigned long frame, int readonly); + +void gnttab_grant_foreign_transfer_ref(grant_ref_t, domid_t domid, + unsigned long pfn); + +int arch_gnttab_map_shared(unsigned long *frames, unsigned long nr_gframes, + unsigned long max_nr_gframes, + struct grant_entry **__shared); +void arch_gnttab_unmap_shared(struct grant_entry *shared, + unsigned long nr_gframes); + +#define gnttab_map_vaddr(map) ((void *)(map.host_virt_addr)) + +#endif /* __ASM_GNTTAB_H__ */ diff --git a/include/xenpvdrivers/hvc-console.h b/include/xenpvdrivers/hvc-console.h new file mode 100644 index 0000000..c3adde3 --- /dev/null +++ b/include/xenpvdrivers/hvc-console.h @@ -0,0 +1,16 @@ +#ifndef XEN_HVC_CONSOLE_H +#define XEN_HVC_CONSOLE_H + +extern struct console xenboot_console; + +#ifdef CONFIG_HVC_XEN +void xen_console_resume(void); +void xen_raw_console_write(const char *str); +void xen_raw_printk(const char *fmt, ...); +#else +static inline void xen_console_resume(void) { } +static inline void xen_raw_console_write(const char *str) { } +static inline void xen_raw_printk(const char *fmt, ...) { } +#endif + +#endif /* XEN_HVC_CONSOLE_H */ diff --git a/include/xenpvdrivers/hvm.h b/include/xenpvdrivers/hvm.h new file mode 100644 index 0000000..d984576 --- /dev/null +++ b/include/xenpvdrivers/hvm.h @@ -0,0 +1,23 @@ +/* Simple wrappers around HVM functions */ +#ifndef XEN_HVM_H__ +#define XEN_HVM_H__ + +#include + +static inline unsigned long hvm_get_parameter(int idx) +{ + struct xen_hvm_param xhv; + int r; + + xhv.domid = DOMID_SELF; + xhv.index = idx; + r = HYPERVISOR_hvm_op(HVMOP_get_param, &xhv); + if (r < 0) { + printk(KERN_ERR "cannot get hvm parameter %d: %d.\n", + idx, r); + return 0; + } + return xhv.value; +} + +#endif /* XEN_HVM_H__ */ diff --git a/include/xenpvdrivers/hypercall.h b/include/xenpvdrivers/hypercall.h new file mode 100644 index 0000000..c27476e --- /dev/null +++ b/include/xenpvdrivers/hypercall.h @@ -0,0 +1,30 @@ +#ifndef __XEN_HYPERCALL_H__ +#define __XEN_HYPERCALL_H__ + +#include + +static inline int __must_check +HYPERVISOR_multicall_check( + multicall_entry_t *call_list, unsigned int nr_calls, + const unsigned long *rc_list) +{ + int rc = HYPERVISOR_multicall(call_list, nr_calls); + + if (unlikely(rc < 0)) + return rc; + BUG_ON(rc); + BUG_ON((int)nr_calls < 0); + + for ( ; nr_calls > 0; --nr_calls, ++call_list) + if (unlikely(call_list->result != (rc_list ? *rc_list++ : 0))) + return nr_calls; + + return 0; +} + +/* A construct to ignore the return value of hypercall wrappers in a few + * exceptional cases (simply casting the function result to void doesn't + * avoid the compiler warning): */ +#define VOID(expr) ((void)((expr)?:0)) + +#endif /* __XEN_HYPERCALL_H__ */ diff --git a/include/xenpvdrivers/hypervisor_sysfs.h b/include/xenpvdrivers/hypervisor_sysfs.h new file mode 100644 index 0000000..fed48a1 --- /dev/null +++ b/include/xenpvdrivers/hypervisor_sysfs.h @@ -0,0 +1,30 @@ +/* + * copyright (c) 2006 IBM Corporation + * Authored by: Mike D. Day + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef _HYP_SYSFS_H_ +#define _HYP_SYSFS_H_ + +#include +#include + +#define HYPERVISOR_ATTR_RO(_name) \ +static struct hyp_sysfs_attr _name##_attr = __ATTR_RO(_name) + +#define HYPERVISOR_ATTR_RW(_name) \ +static struct hyp_sysfs_attr _name##_attr = \ + __ATTR(_name, 0644, _name##_show, _name##_store) + +struct hyp_sysfs_attr { + struct attribute attr; + ssize_t (*show)(struct hyp_sysfs_attr *, char *); + ssize_t (*store)(struct hyp_sysfs_attr *, const char *, size_t); + void *hyp_attr_data; +}; + +#endif /* _HYP_SYSFS_H_ */ diff --git a/include/xenpvdrivers/interface/arch-x86/cpuid.h b/include/xenpvdrivers/interface/arch-x86/cpuid.h new file mode 100644 index 0000000..34fd72d --- /dev/null +++ b/include/xenpvdrivers/interface/arch-x86/cpuid.h @@ -0,0 +1,68 @@ +/****************************************************************************** + * arch-x86/cpuid.h + * + * CPUID interface to Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2007 Citrix Systems, Inc. + * + * Authors: + * Keir Fraser + */ + +#ifndef __XEN_PUBLIC_ARCH_X86_CPUID_H__ +#define __XEN_PUBLIC_ARCH_X86_CPUID_H__ + +/* Xen identification leaves start at 0x40000000. */ +#define XEN_CPUID_FIRST_LEAF 0x40000000 +#define XEN_CPUID_LEAF(i) (XEN_CPUID_FIRST_LEAF + (i)) + +/* + * Leaf 1 (0x40000000) + * EAX: Largest Xen-information leaf. All leaves up to an including @EAX + * are supported by the Xen host. + * EBX-EDX: "XenVMMXenVMM" signature, allowing positive identification + * of a Xen host. + */ +#define XEN_CPUID_SIGNATURE_EBX 0x566e6558 /* "XenV" */ +#define XEN_CPUID_SIGNATURE_ECX 0x65584d4d /* "MMXe" */ +#define XEN_CPUID_SIGNATURE_EDX 0x4d4d566e /* "nVMM" */ + +/* + * Leaf 2 (0x40000001) + * EAX[31:16]: Xen major version. + * EAX[15: 0]: Xen minor version. + * EBX-EDX: Reserved (currently all zeroes). + */ + +/* + * Leaf 3 (0x40000002) + * EAX: Number of hypercall transfer pages. This register is always guaranteed + * to specify one hypercall page. + * EBX: Base address of Xen-specific MSRs. + * ECX: Features 1. Unused bits are set to zero. + * EDX: Features 2. Unused bits are set to zero. + */ + +/* Does the host support MMU_PT_UPDATE_PRESERVE_AD for this guest? */ +#define _XEN_CPUID_FEAT1_MMU_PT_UPDATE_PRESERVE_AD 0 +#define XEN_CPUID_FEAT1_MMU_PT_UPDATE_PRESERVE_AD (1u<<0) + +#endif /* __XEN_PUBLIC_ARCH_X86_CPUID_H__ */ diff --git a/include/xenpvdrivers/interface/arch-x86/hvm/save.h b/include/xenpvdrivers/interface/arch-x86/hvm/save.h new file mode 100644 index 0000000..5996cbc --- /dev/null +++ b/include/xenpvdrivers/interface/arch-x86/hvm/save.h @@ -0,0 +1,439 @@ +/* + * Structure definitions for HVM state that is held by Xen and must + * be saved along with the domain's memory and device-model state. + * + * Copyright (c) 2007 XenSource Ltd. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_HVM_SAVE_X86_H__ +#define __XEN_PUBLIC_HVM_SAVE_X86_H__ + +/* + * Save/restore header: general info about the save file. + */ + +#define HVM_FILE_MAGIC 0x54381286 +#define HVM_FILE_VERSION 0x00000001 + +struct hvm_save_header { + uint32_t magic; /* Must be HVM_FILE_MAGIC */ + uint32_t version; /* File format version */ + uint64_t changeset; /* Version of Xen that saved this file */ + uint32_t cpuid; /* CPUID[0x01][%eax] on the saving machine */ + uint32_t gtsc_khz; /* Guest's TSC frequency in kHz */ +}; + +DECLARE_HVM_SAVE_TYPE(HEADER, 1, struct hvm_save_header); + + +/* + * Processor + */ + +struct hvm_hw_cpu { + uint8_t fpu_regs[512]; + + uint64_t rax; + uint64_t rbx; + uint64_t rcx; + uint64_t rdx; + uint64_t rbp; + uint64_t rsi; + uint64_t rdi; + uint64_t rsp; + uint64_t r8; + uint64_t r9; + uint64_t r10; + uint64_t r11; + uint64_t r12; + uint64_t r13; + uint64_t r14; + uint64_t r15; + + uint64_t rip; + uint64_t rflags; + + uint64_t cr0; + uint64_t cr2; + uint64_t cr3; + uint64_t cr4; + + uint64_t dr0; + uint64_t dr1; + uint64_t dr2; + uint64_t dr3; + uint64_t dr6; + uint64_t dr7; + + uint32_t cs_sel; + uint32_t ds_sel; + uint32_t es_sel; + uint32_t fs_sel; + uint32_t gs_sel; + uint32_t ss_sel; + uint32_t tr_sel; + uint32_t ldtr_sel; + + uint32_t cs_limit; + uint32_t ds_limit; + uint32_t es_limit; + uint32_t fs_limit; + uint32_t gs_limit; + uint32_t ss_limit; + uint32_t tr_limit; + uint32_t ldtr_limit; + uint32_t idtr_limit; + uint32_t gdtr_limit; + + uint64_t cs_base; + uint64_t ds_base; + uint64_t es_base; + uint64_t fs_base; + uint64_t gs_base; + uint64_t ss_base; + uint64_t tr_base; + uint64_t ldtr_base; + uint64_t idtr_base; + uint64_t gdtr_base; + + uint32_t cs_arbytes; + uint32_t ds_arbytes; + uint32_t es_arbytes; + uint32_t fs_arbytes; + uint32_t gs_arbytes; + uint32_t ss_arbytes; + uint32_t tr_arbytes; + uint32_t ldtr_arbytes; + + uint64_t sysenter_cs; + uint64_t sysenter_esp; + uint64_t sysenter_eip; + + /* msr for em64t */ + uint64_t shadow_gs; + + /* msr content saved/restored. */ + uint64_t msr_flags; + uint64_t msr_lstar; + uint64_t msr_star; + uint64_t msr_cstar; + uint64_t msr_syscall_mask; + uint64_t msr_efer; + uint64_t msr_tsc_aux; + + /* guest's idea of what rdtsc() would return */ + uint64_t tsc; + + /* pending event, if any */ + union { + uint32_t pending_event; + struct { + uint8_t pending_vector:8; + uint8_t pending_type:3; + uint8_t pending_error_valid:1; + uint32_t pending_reserved:19; + uint8_t pending_valid:1; + }; + }; + /* error code for pending event */ + uint32_t error_code; +}; + +DECLARE_HVM_SAVE_TYPE(CPU, 2, struct hvm_hw_cpu); + + +/* + * PIC + */ + +struct hvm_hw_vpic { + /* IR line bitmasks. */ + uint8_t irr; + uint8_t imr; + uint8_t isr; + + /* Line IRx maps to IRQ irq_base+x */ + uint8_t irq_base; + + /* + * Where are we in ICW2-4 initialisation (0 means no init in progress)? + * Bits 0-1 (=x): Next write at A=1 sets ICW(x+1). + * Bit 2: ICW1.IC4 (1 == ICW4 included in init sequence) + * Bit 3: ICW1.SNGL (0 == ICW3 included in init sequence) + */ + uint8_t init_state:4; + + /* IR line with highest priority. */ + uint8_t priority_add:4; + + /* Reads from A=0 obtain ISR or IRR? */ + uint8_t readsel_isr:1; + + /* Reads perform a polling read? */ + uint8_t poll:1; + + /* Automatically clear IRQs from the ISR during INTA? */ + uint8_t auto_eoi:1; + + /* Automatically rotate IRQ priorities during AEOI? */ + uint8_t rotate_on_auto_eoi:1; + + /* Exclude slave inputs when considering in-service IRQs? */ + uint8_t special_fully_nested_mode:1; + + /* Special mask mode excludes masked IRs from AEOI and priority checks. */ + uint8_t special_mask_mode:1; + + /* Is this a master PIC or slave PIC? (NB. This is not programmable.) */ + uint8_t is_master:1; + + /* Edge/trigger selection. */ + uint8_t elcr; + + /* Virtual INT output. */ + uint8_t int_output; +}; + +DECLARE_HVM_SAVE_TYPE(PIC, 3, struct hvm_hw_vpic); + + +/* + * IO-APIC + */ + +#ifdef __ia64__ +#define VIOAPIC_IS_IOSAPIC 1 +#define VIOAPIC_NUM_PINS 24 +#else +#define VIOAPIC_NUM_PINS 48 /* 16 ISA IRQs, 32 non-legacy PCI IRQS. */ +#endif + +struct hvm_hw_vioapic { + uint64_t base_address; + uint32_t ioregsel; + uint32_t id; + union vioapic_redir_entry + { + uint64_t bits; + struct { + uint8_t vector; + uint8_t delivery_mode:3; + uint8_t dest_mode:1; + uint8_t delivery_status:1; + uint8_t polarity:1; + uint8_t remote_irr:1; + uint8_t trig_mode:1; + uint8_t mask:1; + uint8_t reserve:7; +#if !VIOAPIC_IS_IOSAPIC + uint8_t reserved[4]; + uint8_t dest_id; +#else + uint8_t reserved[3]; + uint16_t dest_id; +#endif + } fields; + } redirtbl[VIOAPIC_NUM_PINS]; +}; + +DECLARE_HVM_SAVE_TYPE(IOAPIC, 4, struct hvm_hw_vioapic); + + +/* + * LAPIC + */ + +struct hvm_hw_lapic { + uint64_t apic_base_msr; + uint32_t disabled; /* VLAPIC_xx_DISABLED */ + uint32_t timer_divisor; +}; + +DECLARE_HVM_SAVE_TYPE(LAPIC, 5, struct hvm_hw_lapic); + +struct hvm_hw_lapic_regs { + uint8_t data[1024]; +}; + +DECLARE_HVM_SAVE_TYPE(LAPIC_REGS, 6, struct hvm_hw_lapic_regs); + + +/* + * IRQs + */ + +struct hvm_hw_pci_irqs { + /* + * Virtual interrupt wires for a single PCI bus. + * Indexed by: device*4 + INTx#. + */ + union { + unsigned long i[16 / sizeof (unsigned long)]; /* DECLARE_BITMAP(i, 32*4); */ + uint64_t pad[2]; + }; +}; + +DECLARE_HVM_SAVE_TYPE(PCI_IRQ, 7, struct hvm_hw_pci_irqs); + +struct hvm_hw_isa_irqs { + /* + * Virtual interrupt wires for ISA devices. + * Indexed by ISA IRQ (assumes no ISA-device IRQ sharing). + */ + union { + unsigned long i[1]; /* DECLARE_BITMAP(i, 16); */ + uint64_t pad[1]; + }; +}; + +DECLARE_HVM_SAVE_TYPE(ISA_IRQ, 8, struct hvm_hw_isa_irqs); + +struct hvm_hw_pci_link { + /* + * PCI-ISA interrupt router. + * Each PCI is 'wire-ORed' into one of four links using + * the traditional 'barber's pole' mapping ((device + INTx#) & 3). + * The router provides a programmable mapping from each link to a GSI. + */ + uint8_t route[4]; + uint8_t pad0[4]; +}; + +DECLARE_HVM_SAVE_TYPE(PCI_LINK, 9, struct hvm_hw_pci_link); + +/* + * PIT + */ + +struct hvm_hw_pit { + struct hvm_hw_pit_channel { + uint32_t count; /* can be 65536 */ + uint16_t latched_count; + uint8_t count_latched; + uint8_t status_latched; + uint8_t status; + uint8_t read_state; + uint8_t write_state; + uint8_t write_latch; + uint8_t rw_mode; + uint8_t mode; + uint8_t bcd; /* not supported */ + uint8_t gate; /* timer start */ + } channels[3]; /* 3 x 16 bytes */ + uint32_t speaker_data_on; + uint32_t pad0; +}; + +DECLARE_HVM_SAVE_TYPE(PIT, 10, struct hvm_hw_pit); + + +/* + * RTC + */ + +#define RTC_CMOS_SIZE 14 +struct hvm_hw_rtc { + /* CMOS bytes */ + uint8_t cmos_data[RTC_CMOS_SIZE]; + /* Index register for 2-part operations */ + uint8_t cmos_index; + uint8_t pad0; +}; + +DECLARE_HVM_SAVE_TYPE(RTC, 11, struct hvm_hw_rtc); + + +/* + * HPET + */ + +#define HPET_TIMER_NUM 3 /* 3 timers supported now */ +struct hvm_hw_hpet { + /* Memory-mapped, software visible registers */ + uint64_t capability; /* capabilities */ + uint64_t res0; /* reserved */ + uint64_t config; /* configuration */ + uint64_t res1; /* reserved */ + uint64_t isr; /* interrupt status reg */ + uint64_t res2[25]; /* reserved */ + uint64_t mc64; /* main counter */ + uint64_t res3; /* reserved */ + struct { /* timers */ + uint64_t config; /* configuration/cap */ + uint64_t cmp; /* comparator */ + uint64_t fsb; /* FSB route, not supported now */ + uint64_t res4; /* reserved */ + } timers[HPET_TIMER_NUM]; + uint64_t res5[4*(24-HPET_TIMER_NUM)]; /* reserved, up to 0x3ff */ + + /* Hidden register state */ + uint64_t period[HPET_TIMER_NUM]; /* Last value written to comparator */ +}; + +DECLARE_HVM_SAVE_TYPE(HPET, 12, struct hvm_hw_hpet); + + +/* + * PM timer + */ + +struct hvm_hw_pmtimer { + uint32_t tmr_val; /* PM_TMR_BLK.TMR_VAL: 32bit free-running counter */ + uint16_t pm1a_sts; /* PM1a_EVT_BLK.PM1a_STS: status register */ + uint16_t pm1a_en; /* PM1a_EVT_BLK.PM1a_EN: enable register */ +}; + +DECLARE_HVM_SAVE_TYPE(PMTIMER, 13, struct hvm_hw_pmtimer); + +/* + * MTRR MSRs + */ + +struct hvm_hw_mtrr { +#define MTRR_VCNT 8 +#define NUM_FIXED_MSR 11 + uint64_t msr_pat_cr; + /* mtrr physbase & physmask msr pair*/ + uint64_t msr_mtrr_var[MTRR_VCNT*2]; + uint64_t msr_mtrr_fixed[NUM_FIXED_MSR]; + uint64_t msr_mtrr_cap; + uint64_t msr_mtrr_def_type; +}; + +DECLARE_HVM_SAVE_TYPE(MTRR, 14, struct hvm_hw_mtrr); + +/* + * Viridian hypervisor context. + */ + +struct hvm_viridian_context { + uint64_t hypercall_gpa; + uint64_t guest_os_id; +}; + +DECLARE_HVM_SAVE_TYPE(VIRIDIAN, 15, struct hvm_viridian_context); + +/* + * Largest type-code in use + */ +#define HVM_SAVE_CODE_MAX 15 + +#endif /* __XEN_PUBLIC_HVM_SAVE_X86_H__ */ diff --git a/include/xenpvdrivers/interface/arch-x86/xen-mca.h b/include/xenpvdrivers/interface/arch-x86/xen-mca.h new file mode 100644 index 0000000..13fc5b1 --- /dev/null +++ b/include/xenpvdrivers/interface/arch-x86/xen-mca.h @@ -0,0 +1,422 @@ +/****************************************************************************** + * arch-x86/mca.h + * + * Contributed by Advanced Micro Devices, Inc. + * Author: Christoph Egger + * + * Guest OS machine check interface to x86 Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +/* Full MCA functionality has the following Usecases from the guest side: + * + * Must have's: + * 1. Dom0 and DomU register machine check trap callback handlers + * (already done via "set_trap_table" hypercall) + * 2. Dom0 registers machine check event callback handler + * (doable via EVTCHNOP_bind_virq) + * 3. Dom0 and DomU fetches machine check data + * 4. Dom0 wants Xen to notify a DomU + * 5. Dom0 gets DomU ID from physical address + * 6. Dom0 wants Xen to kill DomU (already done for "xm destroy") + * + * Nice to have's: + * 7. Dom0 wants Xen to deactivate a physical CPU + * This is better done as separate task, physical CPU hotplugging, + * and hypercall(s) should be sysctl's + * 8. Page migration proposed from Xen NUMA work, where Dom0 can tell Xen to + * move a DomU (or Dom0 itself) away from a malicious page + * producing correctable errors. + * 9. offlining physical page: + * Xen free's and never re-uses a certain physical page. + * 10. Testfacility: Allow Dom0 to write values into machine check MSR's + * and tell Xen to trigger a machine check + */ + +#ifndef __XEN_PUBLIC_ARCH_X86_MCA_H__ +#define __XEN_PUBLIC_ARCH_X86_MCA_H__ + +/* Hypercall */ +#define __HYPERVISOR_mca __HYPERVISOR_arch_0 + +/* + * The xen-unstable repo has interface version 0x03000001; out interface + * is incompatible with that and any future minor revisions, so we + * choose a different version number range that is numerically less + * than that used in xen-unstable. + */ +#define XEN_MCA_INTERFACE_VERSION 0x01ecc003 + +/* IN: Dom0 calls hypercall to retrieve nonurgent telemetry */ +#define XEN_MC_NONURGENT 0x0001 +/* IN: Dom0/DomU calls hypercall to retrieve urgent telemetry */ +#define XEN_MC_URGENT 0x0002 +/* IN: Dom0 acknowledges previosly-fetched telemetry */ +#define XEN_MC_ACK 0x0004 + +/* OUT: All is ok */ +#define XEN_MC_OK 0x0 +/* OUT: Domain could not fetch data. */ +#define XEN_MC_FETCHFAILED 0x1 +/* OUT: There was no machine check data to fetch. */ +#define XEN_MC_NODATA 0x2 +/* OUT: Between notification time and this hypercall an other + * (most likely) correctable error happened. The fetched data, + * does not match the original machine check data. */ +#define XEN_MC_NOMATCH 0x4 + +/* OUT: DomU did not register MC NMI handler. Try something else. */ +#define XEN_MC_CANNOTHANDLE 0x8 +/* OUT: Notifying DomU failed. Retry later or try something else. */ +#define XEN_MC_NOTDELIVERED 0x10 +/* Note, XEN_MC_CANNOTHANDLE and XEN_MC_NOTDELIVERED are mutually exclusive. */ + + +#ifndef __ASSEMBLY__ + +#define VIRQ_MCA VIRQ_ARCH_0 /* G. (DOM0) Machine Check Architecture */ + +/* + * Machine Check Architecure: + * structs are read-only and used to report all kinds of + * correctable and uncorrectable errors detected by the HW. + * Dom0 and DomU: register a handler to get notified. + * Dom0 only: Correctable errors are reported via VIRQ_MCA + * Dom0 and DomU: Uncorrectable errors are reported via nmi handlers + */ +#define MC_TYPE_GLOBAL 0 +#define MC_TYPE_BANK 1 +#define MC_TYPE_EXTENDED 2 +#define MC_TYPE_RECOVERY 3 + +struct mcinfo_common { + uint16_t type; /* structure type */ + uint16_t size; /* size of this struct in bytes */ +}; + + +#define MC_FLAG_CORRECTABLE (1 << 0) +#define MC_FLAG_UNCORRECTABLE (1 << 1) +#define MC_FLAG_RECOVERABLE (1 << 2) +#define MC_FLAG_POLLED (1 << 3) +#define MC_FLAG_RESET (1 << 4) +#define MC_FLAG_CMCI (1 << 5) +#define MC_FLAG_MCE (1 << 6) +/* contains global x86 mc information */ +struct mcinfo_global { + struct mcinfo_common common; + + /* running domain at the time in error (most likely the impacted one) */ + uint16_t mc_domid; + uint16_t mc_vcpuid; /* virtual cpu scheduled for mc_domid */ + uint32_t mc_socketid; /* physical socket of the physical core */ + uint16_t mc_coreid; /* physical impacted core */ + uint16_t mc_core_threadid; /* core thread of physical core */ + uint32_t mc_apicid; + uint32_t mc_flags; + uint64_t mc_gstatus; /* global status */ +}; + +/* contains bank local x86 mc information */ +struct mcinfo_bank { + struct mcinfo_common common; + + uint16_t mc_bank; /* bank nr */ + uint16_t mc_domid; /* Usecase 5: domain referenced by mc_addr on dom0 + * and if mc_addr is valid. Never valid on DomU. */ + uint64_t mc_status; /* bank status */ + uint64_t mc_addr; /* bank address, only valid + * if addr bit is set in mc_status */ + uint64_t mc_misc; + uint64_t mc_ctrl2; + uint64_t mc_tsc; +}; + + +struct mcinfo_msr { + uint64_t reg; /* MSR */ + uint64_t value; /* MSR value */ +}; + +/* contains mc information from other + * or additional mc MSRs */ +struct mcinfo_extended { + struct mcinfo_common common; + + /* You can fill up to five registers. + * If you need more, then use this structure + * multiple times. */ + + uint32_t mc_msrs; /* Number of msr with valid values. */ + /* + * Currently Intel extended MSR (32/64) include all gp registers + * and E(R)FLAGS, E(R)IP, E(R)MISC, up to 11/19 of them might be + * useful at present. So expand this array to 16/32 to leave room. + */ + struct mcinfo_msr mc_msr[sizeof(void *) * 4]; +}; + +/* Recovery Action flags. Giving recovery result information to DOM0 */ + +/* Xen takes successful recovery action, the error is recovered */ +#define REC_ACTION_RECOVERED (0x1 << 0) +/* No action is performed by XEN */ +#define REC_ACTION_NONE (0x1 << 1) +/* It's possible DOM0 might take action ownership in some case */ +#define REC_ACTION_NEED_RESET (0x1 << 2) + +/* Different Recovery Action types, if the action is performed successfully, + * REC_ACTION_RECOVERED flag will be returned. + */ + +/* Page Offline Action */ +#define MC_ACTION_PAGE_OFFLINE (0x1 << 0) +/* CPU offline Action */ +#define MC_ACTION_CPU_OFFLINE (0x1 << 1) +/* L3 cache disable Action */ +#define MC_ACTION_CACHE_SHRINK (0x1 << 2) + +/* Below interface used between XEN/DOM0 for passing XEN's recovery action + * information to DOM0. + * usage Senario: After offlining broken page, XEN might pass its page offline + * recovery action result to DOM0. DOM0 will save the information in + * non-volatile memory for further proactive actions, such as offlining the + * easy broken page earlier when doing next reboot. +*/ +struct page_offline_action +{ + /* Params for passing the offlined page number to DOM0 */ + uint64_t mfn; + uint64_t status; +}; + +struct cpu_offline_action +{ + /* Params for passing the identity of the offlined CPU to DOM0 */ + uint32_t mc_socketid; + uint16_t mc_coreid; + uint16_t mc_core_threadid; +}; + +#define MAX_UNION_SIZE 16 +struct mcinfo_recovery +{ + struct mcinfo_common common; + uint16_t mc_bank; /* bank nr */ + uint8_t action_flags; + uint8_t action_types; + union { + struct page_offline_action page_retire; + struct cpu_offline_action cpu_offline; + uint8_t pad[MAX_UNION_SIZE]; + } action_info; +}; + + +#define MCINFO_HYPERCALLSIZE 1024 +#define MCINFO_MAXSIZE 768 + +struct mc_info { + /* Number of mcinfo_* entries in mi_data */ + uint32_t mi_nentries; + uint32_t _pad0; + uint64_t mi_data[(MCINFO_MAXSIZE - 1) / 8]; +}; +typedef struct mc_info mc_info_t; +DEFINE_XEN_GUEST_HANDLE(mc_info_t); + +#define __MC_MSR_ARRAYSIZE 8 +#define __MC_NMSRS 1 +#define MC_NCAPS 7 /* 7 CPU feature flag words */ +#define MC_CAPS_STD_EDX 0 /* cpuid level 0x00000001 (%edx) */ +#define MC_CAPS_AMD_EDX 1 /* cpuid level 0x80000001 (%edx) */ +#define MC_CAPS_TM 2 /* cpuid level 0x80860001 (TransMeta) */ +#define MC_CAPS_LINUX 3 /* Linux-defined */ +#define MC_CAPS_STD_ECX 4 /* cpuid level 0x00000001 (%ecx) */ +#define MC_CAPS_VIA 5 /* cpuid level 0xc0000001 */ +#define MC_CAPS_AMD_ECX 6 /* cpuid level 0x80000001 (%ecx) */ + +struct mcinfo_logical_cpu { + uint32_t mc_cpunr; + uint32_t mc_chipid; + uint16_t mc_coreid; + uint16_t mc_threadid; + uint32_t mc_apicid; + uint32_t mc_clusterid; + uint32_t mc_ncores; + uint32_t mc_ncores_active; + uint32_t mc_nthreads; + int32_t mc_cpuid_level; + uint32_t mc_family; + uint32_t mc_vendor; + uint32_t mc_model; + uint32_t mc_step; + char mc_vendorid[16]; + char mc_brandid[64]; + uint32_t mc_cpu_caps[MC_NCAPS]; + uint32_t mc_cache_size; + uint32_t mc_cache_alignment; + int32_t mc_nmsrvals; + struct mcinfo_msr mc_msrvalues[__MC_MSR_ARRAYSIZE]; +}; +typedef struct mcinfo_logical_cpu xen_mc_logical_cpu_t; +DEFINE_XEN_GUEST_HANDLE(xen_mc_logical_cpu_t); + + +/* + * OS's should use these instead of writing their own lookup function + * each with its own bugs and drawbacks. + * We use macros instead of static inline functions to allow guests + * to include this header in assembly files (*.S). + */ +/* Prototype: + * uint32_t x86_mcinfo_nentries(struct mc_info *mi); + */ +#define x86_mcinfo_nentries(_mi) \ + (_mi)->mi_nentries +/* Prototype: + * struct mcinfo_common *x86_mcinfo_first(struct mc_info *mi); + */ +#define x86_mcinfo_first(_mi) \ + ((struct mcinfo_common *)(_mi)->mi_data) +/* Prototype: + * struct mcinfo_common *x86_mcinfo_next(struct mcinfo_common *mic); + */ +#define x86_mcinfo_next(_mic) \ + ((struct mcinfo_common *)((uint8_t *)(_mic) + (_mic)->size)) + +/* Prototype: + * void x86_mcinfo_lookup(void *ret, struct mc_info *mi, uint16_t type); + */ +#define x86_mcinfo_lookup(_ret, _mi, _type) \ + do { \ + uint32_t found, i; \ + struct mcinfo_common *_mic; \ + \ + found = 0; \ + (_ret) = NULL; \ + if (_mi == NULL) break; \ + _mic = x86_mcinfo_first(_mi); \ + for (i = 0; i < x86_mcinfo_nentries(_mi); i++) { \ + if (_mic->type == (_type)) { \ + found = 1; \ + break; \ + } \ + _mic = x86_mcinfo_next(_mic); \ + } \ + (_ret) = found ? _mic : NULL; \ + } while (0) + + +/* Usecase 1 + * Register machine check trap callback handler + * (already done via "set_trap_table" hypercall) + */ + +/* Usecase 2 + * Dom0 registers machine check event callback handler + * done by EVTCHNOP_bind_virq + */ + +/* Usecase 3 + * Fetch machine check data from hypervisor. + * Note, this hypercall is special, because both Dom0 and DomU must use this. + */ +#define XEN_MC_fetch 1 +struct xen_mc_fetch { + /* IN/OUT variables. */ + uint32_t flags; /* IN: XEN_MC_NONURGENT, XEN_MC_URGENT, + XEN_MC_ACK if ack'ing an earlier fetch */ + /* OUT: XEN_MC_OK, XEN_MC_FETCHFAILED, + XEN_MC_NODATA, XEN_MC_NOMATCH */ + uint32_t _pad0; + uint64_t fetch_id; /* OUT: id for ack, IN: id we are ack'ing */ + + /* OUT variables. */ + XEN_GUEST_HANDLE(mc_info_t) data; +}; +typedef struct xen_mc_fetch xen_mc_fetch_t; +DEFINE_XEN_GUEST_HANDLE(xen_mc_fetch_t); + + +/* Usecase 4 + * This tells the hypervisor to notify a DomU about the machine check error + */ +#define XEN_MC_notifydomain 2 +struct xen_mc_notifydomain { + /* IN variables. */ + uint16_t mc_domid; /* The unprivileged domain to notify. */ + uint16_t mc_vcpuid; /* The vcpu in mc_domid to notify. + * Usually echo'd value from the fetch hypercall. */ + + /* IN/OUT variables. */ + uint32_t flags; + +/* IN: XEN_MC_CORRECTABLE, XEN_MC_TRAP */ +/* OUT: XEN_MC_OK, XEN_MC_CANNOTHANDLE, XEN_MC_NOTDELIVERED, XEN_MC_NOMATCH */ +}; +typedef struct xen_mc_notifydomain xen_mc_notifydomain_t; +DEFINE_XEN_GUEST_HANDLE(xen_mc_notifydomain_t); + +#define XEN_MC_physcpuinfo 3 +struct xen_mc_physcpuinfo { + /* IN/OUT */ + uint32_t ncpus; + uint32_t _pad0; + /* OUT */ + XEN_GUEST_HANDLE(xen_mc_logical_cpu_t) info; +}; + +#define XEN_MC_msrinject 4 +#define MC_MSRINJ_MAXMSRS 8 +struct xen_mc_msrinject { + /* IN */ + uint32_t mcinj_cpunr; /* target processor id */ + uint32_t mcinj_flags; /* see MC_MSRINJ_F_* below */ + uint32_t mcinj_count; /* 0 .. count-1 in array are valid */ + uint32_t _pad0; + struct mcinfo_msr mcinj_msr[MC_MSRINJ_MAXMSRS]; +}; + +/* Flags for mcinj_flags above; bits 16-31 are reserved */ +#define MC_MSRINJ_F_INTERPOSE 0x1 + +#define XEN_MC_mceinject 5 +struct xen_mc_mceinject { + unsigned int mceinj_cpunr; /* target processor id */ +}; + +struct xen_mc { + uint32_t cmd; + uint32_t interface_version; /* XEN_MCA_INTERFACE_VERSION */ + union { + struct xen_mc_fetch mc_fetch; + struct xen_mc_notifydomain mc_notifydomain; + struct xen_mc_physcpuinfo mc_physcpuinfo; + struct xen_mc_msrinject mc_msrinject; + struct xen_mc_mceinject mc_mceinject; + } u; +}; +typedef struct xen_mc xen_mc_t; +DEFINE_XEN_GUEST_HANDLE(xen_mc_t); + +#endif /* __ASSEMBLY__ */ + +#endif /* __XEN_PUBLIC_ARCH_X86_MCA_H__ */ diff --git a/include/xenpvdrivers/interface/arch-x86/xen-x86_32.h b/include/xenpvdrivers/interface/arch-x86/xen-x86_32.h new file mode 100644 index 0000000..7cb6a01 --- /dev/null +++ b/include/xenpvdrivers/interface/arch-x86/xen-x86_32.h @@ -0,0 +1,180 @@ +/****************************************************************************** + * xen-x86_32.h + * + * Guest OS interface to x86 32-bit Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2004-2007, K A Fraser + */ + +#ifndef __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__ +#define __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__ + +/* + * Hypercall interface: + * Input: %ebx, %ecx, %edx, %esi, %edi (arguments 1-5) + * Output: %eax + * Access is via hypercall page (set up by guest loader or via a Xen MSR): + * call hypercall_page + hypercall-number * 32 + * Clobbered: Argument registers (e.g., 2-arg hypercall clobbers %ebx,%ecx) + */ + +#if __XEN_INTERFACE_VERSION__ < 0x00030203 +/* + * Legacy hypercall interface: + * As above, except the entry sequence to the hypervisor is: + * mov $hypercall-number*32,%eax ; int $0x82 + */ +#define TRAP_INSTR "int $0x82" +#endif + +/* + * These flat segments are in the Xen-private section of every GDT. Since these + * are also present in the initial GDT, many OSes will be able to avoid + * installing their own GDT. + */ +#define FLAT_RING1_CS 0xe019 /* GDT index 259 */ +#define FLAT_RING1_DS 0xe021 /* GDT index 260 */ +#define FLAT_RING1_SS 0xe021 /* GDT index 260 */ +#define FLAT_RING3_CS 0xe02b /* GDT index 261 */ +#define FLAT_RING3_DS 0xe033 /* GDT index 262 */ +#define FLAT_RING3_SS 0xe033 /* GDT index 262 */ + +#define FLAT_KERNEL_CS FLAT_RING1_CS +#define FLAT_KERNEL_DS FLAT_RING1_DS +#define FLAT_KERNEL_SS FLAT_RING1_SS +#define FLAT_USER_CS FLAT_RING3_CS +#define FLAT_USER_DS FLAT_RING3_DS +#define FLAT_USER_SS FLAT_RING3_SS + +#define __HYPERVISOR_VIRT_START_PAE 0xF5800000 +#define __MACH2PHYS_VIRT_START_PAE 0xF5800000 +#define __MACH2PHYS_VIRT_END_PAE 0xF6800000 +#define HYPERVISOR_VIRT_START_PAE \ + mk_unsigned_long(__HYPERVISOR_VIRT_START_PAE) +#define MACH2PHYS_VIRT_START_PAE \ + mk_unsigned_long(__MACH2PHYS_VIRT_START_PAE) +#define MACH2PHYS_VIRT_END_PAE \ + mk_unsigned_long(__MACH2PHYS_VIRT_END_PAE) + +/* Non-PAE bounds are obsolete. */ +#define __HYPERVISOR_VIRT_START_NONPAE 0xFC000000 +#define __MACH2PHYS_VIRT_START_NONPAE 0xFC000000 +#define __MACH2PHYS_VIRT_END_NONPAE 0xFC400000 +#define HYPERVISOR_VIRT_START_NONPAE \ + mk_unsigned_long(__HYPERVISOR_VIRT_START_NONPAE) +#define MACH2PHYS_VIRT_START_NONPAE \ + mk_unsigned_long(__MACH2PHYS_VIRT_START_NONPAE) +#define MACH2PHYS_VIRT_END_NONPAE \ + mk_unsigned_long(__MACH2PHYS_VIRT_END_NONPAE) + +#define __HYPERVISOR_VIRT_START __HYPERVISOR_VIRT_START_PAE +#define __MACH2PHYS_VIRT_START __MACH2PHYS_VIRT_START_PAE +#define __MACH2PHYS_VIRT_END __MACH2PHYS_VIRT_END_PAE + +#ifndef HYPERVISOR_VIRT_START +#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START) +#endif + +#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START) +#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END) +#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>2) +#ifndef machine_to_phys_mapping +#define machine_to_phys_mapping ((unsigned long *)MACH2PHYS_VIRT_START) +#endif + +/* 32-/64-bit invariability for control interfaces (domctl/sysctl). */ +#if defined(__XEN__) || defined(__XEN_TOOLS__) +#undef ___DEFINE_XEN_GUEST_HANDLE +#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \ + typedef struct { type *p; } \ + __guest_handle_ ## name; \ + typedef struct { union { type *p; uint64_aligned_t q; }; } \ + __guest_handle_64_ ## name +#undef set_xen_guest_handle +#define set_xen_guest_handle(hnd, val) \ + do { if ( sizeof(hnd) == 8 ) *(uint64_t *)&(hnd) = 0; \ + (hnd).p = val; \ + } while ( 0 ) +#define uint64_aligned_t uint64_t __attribute__((aligned(8))) +#define __XEN_GUEST_HANDLE_64(name) __guest_handle_64_ ## name +#define XEN_GUEST_HANDLE_64(name) __XEN_GUEST_HANDLE_64(name) +#endif + +#ifndef __ASSEMBLY__ + +struct cpu_user_regs { + uint32_t ebx; + uint32_t ecx; + uint32_t edx; + uint32_t esi; + uint32_t edi; + uint32_t ebp; + uint32_t eax; + uint16_t error_code; /* private */ + uint16_t entry_vector; /* private */ + uint32_t eip; + uint16_t cs; + uint8_t saved_upcall_mask; + uint8_t _pad0; + uint32_t eflags; /* eflags.IF == !saved_upcall_mask */ + uint32_t esp; + uint16_t ss, _pad1; + uint16_t es, _pad2; + uint16_t ds, _pad3; + uint16_t fs, _pad4; + uint16_t gs, _pad5; +}; +typedef struct cpu_user_regs cpu_user_regs_t; +DEFINE_XEN_GUEST_HANDLE(cpu_user_regs_t); + +/* + * Page-directory addresses above 4GB do not fit into architectural %cr3. + * When accessing %cr3, or equivalent field in vcpu_guest_context, guests + * must use the following accessor macros to pack/unpack valid MFNs. + */ +#define xen_pfn_to_cr3(pfn) (((unsigned)(pfn) << 12) | ((unsigned)(pfn) >> 20)) +#define xen_cr3_to_pfn(cr3) (((unsigned)(cr3) >> 12) | ((unsigned)(cr3) << 20)) + +struct arch_vcpu_info { + unsigned long cr2; + unsigned long pad[5]; /* sizeof(vcpu_info_t) == 64 */ +}; +typedef struct arch_vcpu_info arch_vcpu_info_t; + +struct xen_callback { + unsigned long cs; + unsigned long eip; +}; +typedef struct xen_callback xen_callback_t; + +#endif /* !__ASSEMBLY__ */ + +#endif /* __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/arch-x86/xen-x86_64.h b/include/xenpvdrivers/interface/arch-x86/xen-x86_64.h new file mode 100644 index 0000000..1e54cf9 --- /dev/null +++ b/include/xenpvdrivers/interface/arch-x86/xen-x86_64.h @@ -0,0 +1,212 @@ +/****************************************************************************** + * xen-x86_64.h + * + * Guest OS interface to x86 64-bit Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2004-2006, K A Fraser + */ + +#ifndef __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__ +#define __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__ + +/* + * Hypercall interface: + * Input: %rdi, %rsi, %rdx, %r10, %r8 (arguments 1-5) + * Output: %rax + * Access is via hypercall page (set up by guest loader or via a Xen MSR): + * call hypercall_page + hypercall-number * 32 + * Clobbered: argument registers (e.g., 2-arg hypercall clobbers %rdi,%rsi) + */ + +#if __XEN_INTERFACE_VERSION__ < 0x00030203 +/* + * Legacy hypercall interface: + * As above, except the entry sequence to the hypervisor is: + * mov $hypercall-number*32,%eax ; syscall + * Clobbered: %rcx, %r11, argument registers (as above) + */ +#define TRAP_INSTR "syscall" +#endif + +/* + * 64-bit segment selectors + * These flat segments are in the Xen-private section of every GDT. Since these + * are also present in the initial GDT, many OSes will be able to avoid + * installing their own GDT. + */ + +#define FLAT_RING3_CS32 0xe023 /* GDT index 260 */ +#define FLAT_RING3_CS64 0xe033 /* GDT index 261 */ +#define FLAT_RING3_DS32 0xe02b /* GDT index 262 */ +#define FLAT_RING3_DS64 0x0000 /* NULL selector */ +#define FLAT_RING3_SS32 0xe02b /* GDT index 262 */ +#define FLAT_RING3_SS64 0xe02b /* GDT index 262 */ + +#define FLAT_KERNEL_DS64 FLAT_RING3_DS64 +#define FLAT_KERNEL_DS32 FLAT_RING3_DS32 +#define FLAT_KERNEL_DS FLAT_KERNEL_DS64 +#define FLAT_KERNEL_CS64 FLAT_RING3_CS64 +#define FLAT_KERNEL_CS32 FLAT_RING3_CS32 +#define FLAT_KERNEL_CS FLAT_KERNEL_CS64 +#define FLAT_KERNEL_SS64 FLAT_RING3_SS64 +#define FLAT_KERNEL_SS32 FLAT_RING3_SS32 +#define FLAT_KERNEL_SS FLAT_KERNEL_SS64 + +#define FLAT_USER_DS64 FLAT_RING3_DS64 +#define FLAT_USER_DS32 FLAT_RING3_DS32 +#define FLAT_USER_DS FLAT_USER_DS64 +#define FLAT_USER_CS64 FLAT_RING3_CS64 +#define FLAT_USER_CS32 FLAT_RING3_CS32 +#define FLAT_USER_CS FLAT_USER_CS64 +#define FLAT_USER_SS64 FLAT_RING3_SS64 +#define FLAT_USER_SS32 FLAT_RING3_SS32 +#define FLAT_USER_SS FLAT_USER_SS64 + +#define __HYPERVISOR_VIRT_START 0xFFFF800000000000 +#define __HYPERVISOR_VIRT_END 0xFFFF880000000000 +#define __MACH2PHYS_VIRT_START 0xFFFF800000000000 +#define __MACH2PHYS_VIRT_END 0xFFFF804000000000 + +#ifndef HYPERVISOR_VIRT_START +#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START) +#define HYPERVISOR_VIRT_END mk_unsigned_long(__HYPERVISOR_VIRT_END) +#endif + +#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START) +#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END) +#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>3) +#ifndef machine_to_phys_mapping +#define machine_to_phys_mapping ((unsigned long *)HYPERVISOR_VIRT_START) +#endif + +/* + * int HYPERVISOR_set_segment_base(unsigned int which, unsigned long base) + * @which == SEGBASE_* ; @base == 64-bit base address + * Returns 0 on success. + */ +#define SEGBASE_FS 0 +#define SEGBASE_GS_USER 1 +#define SEGBASE_GS_KERNEL 2 +#define SEGBASE_GS_USER_SEL 3 /* Set user %gs specified in base[15:0] */ + +/* + * int HYPERVISOR_iret(void) + * All arguments are on the kernel stack, in the following format. + * Never returns if successful. Current kernel context is lost. + * The saved CS is mapped as follows: + * RING0 -> RING3 kernel mode. + * RING1 -> RING3 kernel mode. + * RING2 -> RING3 kernel mode. + * RING3 -> RING3 user mode. + * However RING0 indicates that the guest kernel should return to iteself + * directly with + * orb $3,1*8(%rsp) + * iretq + * If flags contains VGCF_in_syscall: + * Restore RAX, RIP, RFLAGS, RSP. + * Discard R11, RCX, CS, SS. + * Otherwise: + * Restore RAX, R11, RCX, CS:RIP, RFLAGS, SS:RSP. + * All other registers are saved on hypercall entry and restored to user. + */ +/* Guest exited in SYSCALL context? Return to guest with SYSRET? */ +#define _VGCF_in_syscall 8 +#define VGCF_in_syscall (1<<_VGCF_in_syscall) +#define VGCF_IN_SYSCALL VGCF_in_syscall + +#ifndef __ASSEMBLY__ + +struct iret_context { + /* Top of stack (%rsp at point of hypercall). */ + uint64_t rax, r11, rcx, flags, rip, cs, rflags, rsp, ss; + /* Bottom of iret stack frame. */ +}; + +#if defined(__GNUC__) && !defined(__STRICT_ANSI__) +/* Anonymous union includes both 32- and 64-bit names (e.g., eax/rax). */ +#define __DECL_REG(name) union { \ + uint64_t r ## name, e ## name; \ + uint32_t _e ## name; \ +} +#else +/* Non-gcc sources must always use the proper 64-bit name (e.g., rax). */ +#define __DECL_REG(name) uint64_t r ## name +#endif + +struct cpu_user_regs { + uint64_t r15; + uint64_t r14; + uint64_t r13; + uint64_t r12; + __DECL_REG(bp); + __DECL_REG(bx); + uint64_t r11; + uint64_t r10; + uint64_t r9; + uint64_t r8; + __DECL_REG(ax); + __DECL_REG(cx); + __DECL_REG(dx); + __DECL_REG(si); + __DECL_REG(di); + uint32_t error_code; /* private */ + uint32_t entry_vector; /* private */ + __DECL_REG(ip); + uint16_t cs, _pad0[1]; + uint8_t saved_upcall_mask; + uint8_t _pad1[3]; + __DECL_REG(flags); /* rflags.IF == !saved_upcall_mask */ + __DECL_REG(sp); + uint16_t ss, _pad2[3]; + uint16_t es, _pad3[3]; + uint16_t ds, _pad4[3]; + uint16_t fs, _pad5[3]; /* Non-zero => takes precedence over fs_base. */ + uint16_t gs, _pad6[3]; /* Non-zero => takes precedence over gs_base_usr. */ +}; +typedef struct cpu_user_regs cpu_user_regs_t; +DEFINE_XEN_GUEST_HANDLE(cpu_user_regs_t); + +#undef __DECL_REG + +#define xen_pfn_to_cr3(pfn) ((unsigned long)(pfn) << 12) +#define xen_cr3_to_pfn(cr3) ((unsigned long)(cr3) >> 12) + +struct arch_vcpu_info { + unsigned long cr2; + unsigned long pad; /* sizeof(vcpu_info_t) == 64 */ +}; +typedef struct arch_vcpu_info arch_vcpu_info_t; + +typedef unsigned long xen_callback_t; + +#endif /* !__ASSEMBLY__ */ + +#endif /* __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/arch-x86/xen.h b/include/xenpvdrivers/interface/arch-x86/xen.h new file mode 100644 index 0000000..eb0320f --- /dev/null +++ b/include/xenpvdrivers/interface/arch-x86/xen.h @@ -0,0 +1,203 @@ +/****************************************************************************** + * arch-x86/xen.h + * + * Guest OS interface to x86 Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2004-2006, K A Fraser + */ + +#include "../xen.h" + +#ifndef __XEN_PUBLIC_ARCH_X86_XEN_H__ +#define __XEN_PUBLIC_ARCH_X86_XEN_H__ + +/* Structural guest handles introduced in 0x00030201. */ +#if __XEN_INTERFACE_VERSION__ >= 0x00030201 +#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \ + typedef struct { type *p; } __guest_handle_ ## name +#else +#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \ + typedef type * __guest_handle_ ## name +#endif + +#define __DEFINE_XEN_GUEST_HANDLE(name, type) \ + ___DEFINE_XEN_GUEST_HANDLE(name, type); \ + ___DEFINE_XEN_GUEST_HANDLE(const_##name, const type) +#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name) +#define __XEN_GUEST_HANDLE(name) __guest_handle_ ## name +#define XEN_GUEST_HANDLE(name) __XEN_GUEST_HANDLE(name) +#define set_xen_guest_handle(hnd, val) do { (hnd).p = val; } while (0) +#ifdef __XEN_TOOLS__ +#define get_xen_guest_handle(val, hnd) do { val = (hnd).p; } while (0) +#endif + +/* Allow co-existing Linux 2.6.23+ Xen interface definitions. */ +#define DEFINE_GUEST_HANDLE_STRUCT(name) struct name + +#if defined(__i386__) +#include "xen-x86_32.h" +#elif defined(__x86_64__) +#include "xen-x86_64.h" +#endif + +#ifndef __ASSEMBLY__ +typedef unsigned long xen_pfn_t; +#define PRI_xen_pfn "lx" +#endif + +/* + * SEGMENT DESCRIPTOR TABLES + */ +/* + * A number of GDT entries are reserved by Xen. These are not situated at the + * start of the GDT because some stupid OSes export hard-coded selector values + * in their ABI. These hard-coded values are always near the start of the GDT, + * so Xen places itself out of the way, at the far end of the GDT. + */ +#define FIRST_RESERVED_GDT_PAGE 14 +#define FIRST_RESERVED_GDT_BYTE (FIRST_RESERVED_GDT_PAGE * 4096) +#define FIRST_RESERVED_GDT_ENTRY (FIRST_RESERVED_GDT_BYTE / 8) + +/* Maximum number of virtual CPUs in legacy multi-processor guests. */ +#define XEN_LEGACY_MAX_VCPUS 32 + +#ifndef __ASSEMBLY__ + +typedef unsigned long xen_ulong_t; + +/* + * Send an array of these to HYPERVISOR_set_trap_table(). + * The privilege level specifies which modes may enter a trap via a software + * interrupt. On x86/64, since rings 1 and 2 are unavailable, we allocate + * privilege levels as follows: + * Level == 0: Noone may enter + * Level == 1: Kernel may enter + * Level == 2: Kernel may enter + * Level == 3: Everyone may enter + */ +#define TI_GET_DPL(_ti) ((_ti)->flags & 3) +#define TI_GET_IF(_ti) ((_ti)->flags & 4) +#define TI_SET_DPL(_ti,_dpl) ((_ti)->flags |= (_dpl)) +#define TI_SET_IF(_ti,_if) ((_ti)->flags |= ((!!(_if))<<2)) +struct trap_info { + uint8_t vector; /* exception vector */ + uint8_t flags; /* 0-3: privilege level; 4: clear event enable? */ + uint16_t cs; /* code selector */ + unsigned long address; /* code offset */ +}; +typedef struct trap_info trap_info_t; +DEFINE_XEN_GUEST_HANDLE(trap_info_t); + +typedef uint64_t tsc_timestamp_t; /* RDTSC timestamp */ + +/* + * The following is all CPU context. Note that the fpu_ctxt block is filled + * in by FXSAVE if the CPU has feature FXSR; otherwise FSAVE is used. + */ +struct vcpu_guest_context { + /* FPU registers come first so they can be aligned for FXSAVE/FXRSTOR. */ + struct { char x[512]; } fpu_ctxt; /* User-level FPU registers */ +#define VGCF_I387_VALID (1<<0) +#define VGCF_IN_KERNEL (1<<2) +#define _VGCF_i387_valid 0 +#define VGCF_i387_valid (1<<_VGCF_i387_valid) +#define _VGCF_in_kernel 2 +#define VGCF_in_kernel (1<<_VGCF_in_kernel) +#define _VGCF_failsafe_disables_events 3 +#define VGCF_failsafe_disables_events (1<<_VGCF_failsafe_disables_events) +#define _VGCF_syscall_disables_events 4 +#define VGCF_syscall_disables_events (1<<_VGCF_syscall_disables_events) +#define _VGCF_online 5 +#define VGCF_online (1<<_VGCF_online) + unsigned long flags; /* VGCF_* flags */ + struct cpu_user_regs user_regs; /* User-level CPU registers */ + struct trap_info trap_ctxt[256]; /* Virtual IDT */ + unsigned long ldt_base, ldt_ents; /* LDT (linear address, # ents) */ + unsigned long gdt_frames[16], gdt_ents; /* GDT (machine frames, # ents) */ + unsigned long kernel_ss, kernel_sp; /* Virtual TSS (only SS1/SP1) */ + /* NB. User pagetable on x86/64 is placed in ctrlreg[1]. */ + unsigned long ctrlreg[8]; /* CR0-CR7 (control registers) */ + unsigned long debugreg[8]; /* DB0-DB7 (debug registers) */ +#ifdef __i386__ + unsigned long event_callback_cs; /* CS:EIP of event callback */ + unsigned long event_callback_eip; + unsigned long failsafe_callback_cs; /* CS:EIP of failsafe callback */ + unsigned long failsafe_callback_eip; +#else + unsigned long event_callback_eip; + unsigned long failsafe_callback_eip; +#ifdef __XEN__ + union { + unsigned long syscall_callback_eip; + struct { + unsigned int event_callback_cs; /* compat CS of event cb */ + unsigned int failsafe_callback_cs; /* compat CS of failsafe cb */ + }; + }; +#else + unsigned long syscall_callback_eip; +#endif +#endif + unsigned long vm_assist; /* VMASST_TYPE_* bitmap */ +#ifdef __x86_64__ + /* Segment base addresses. */ + uint64_t fs_base; + uint64_t gs_base_kernel; + uint64_t gs_base_user; +#endif +}; +typedef struct vcpu_guest_context vcpu_guest_context_t; +DEFINE_XEN_GUEST_HANDLE(vcpu_guest_context_t); + +struct arch_shared_info { + unsigned long max_pfn; /* max pfn that appears in table */ + /* Frame containing list of mfns containing list of mfns containing p2m. */ + xen_pfn_t pfn_to_mfn_frame_list_list; + unsigned long nmi_reason; + uint64_t pad[32]; +}; +typedef struct arch_shared_info arch_shared_info_t; + +#endif /* !__ASSEMBLY__ */ + +/* + * Prefix forces emulation of some non-trapping instructions. + * Currently only CPUID. + */ +#ifdef __ASSEMBLY__ +#define XEN_EMULATE_PREFIX .byte 0x0f,0x0b,0x78,0x65,0x6e ; +#define XEN_CPUID XEN_EMULATE_PREFIX cpuid +#else +#define XEN_EMULATE_PREFIX ".byte 0x0f,0x0b,0x78,0x65,0x6e ; " +#define XEN_CPUID XEN_EMULATE_PREFIX "cpuid" +#endif + +#endif /* __XEN_PUBLIC_ARCH_X86_XEN_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/arch-x86_32.h b/include/xenpvdrivers/interface/arch-x86_32.h new file mode 100644 index 0000000..45842b2 --- /dev/null +++ b/include/xenpvdrivers/interface/arch-x86_32.h @@ -0,0 +1,27 @@ +/****************************************************************************** + * arch-x86_32.h + * + * Guest OS interface to x86 32-bit Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2004-2006, K A Fraser + */ + +#include "arch-x86/xen.h" diff --git a/include/xenpvdrivers/interface/arch-x86_64.h b/include/xenpvdrivers/interface/arch-x86_64.h new file mode 100644 index 0000000..fbb2639 --- /dev/null +++ b/include/xenpvdrivers/interface/arch-x86_64.h @@ -0,0 +1,27 @@ +/****************************************************************************** + * arch-x86_64.h + * + * Guest OS interface to x86 64-bit Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2004-2006, K A Fraser + */ + +#include "arch-x86/xen.h" diff --git a/include/xenpvdrivers/interface/callback.h b/include/xenpvdrivers/interface/callback.h new file mode 100644 index 0000000..0323e51 --- /dev/null +++ b/include/xenpvdrivers/interface/callback.h @@ -0,0 +1,111 @@ +/****************************************************************************** + * callback.h + * + * Register guest OS callbacks with Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2006, Ian Campbell + */ + +#ifndef __XEN_PUBLIC_CALLBACK_H__ +#define __XEN_PUBLIC_CALLBACK_H__ + +#include "xen.h" + +/* + * Prototype for this hypercall is: + * long callback_op(int cmd, void *extra_args) + * @cmd == CALLBACKOP_??? (callback operation). + * @extra_args == Operation-specific extra arguments (NULL if none). + */ + +/* ia64, x86: Callback for event delivery. */ +#define CALLBACKTYPE_event 0 + +/* x86: Failsafe callback when guest state cannot be restored by Xen. */ +#define CALLBACKTYPE_failsafe 1 + +/* x86/64 hypervisor: Syscall by 64-bit guest app ('64-on-64-on-64'). */ +#define CALLBACKTYPE_syscall 2 + +/* + * x86/32 hypervisor: Only available on x86/32 when supervisor_mode_kernel + * feature is enabled. Do not use this callback type in new code. + */ +#define CALLBACKTYPE_sysenter_deprecated 3 + +/* x86: Callback for NMI delivery. */ +#define CALLBACKTYPE_nmi 4 + +/* + * x86: sysenter is only available as follows: + * - 32-bit hypervisor: with the supervisor_mode_kernel feature enabled + * - 64-bit hypervisor: 32-bit guest applications on Intel CPUs + * ('32-on-32-on-64', '32-on-64-on-64') + * [nb. also 64-bit guest applications on Intel CPUs + * ('64-on-64-on-64'), but syscall is preferred] + */ +#define CALLBACKTYPE_sysenter 5 + +/* + * x86/64 hypervisor: Syscall by 32-bit guest app on AMD CPUs + * ('32-on-32-on-64', '32-on-64-on-64') + */ +#define CALLBACKTYPE_syscall32 7 + +/* + * Disable event deliver during callback? This flag is ignored for event and + * NMI callbacks: event delivery is unconditionally disabled. + */ +#define _CALLBACKF_mask_events 0 +#define CALLBACKF_mask_events (1U << _CALLBACKF_mask_events) + +/* + * Register a callback. + */ +#define CALLBACKOP_register 0 +struct callback_register { + uint16_t type; + uint16_t flags; + xen_callback_t address; +}; +typedef struct callback_register callback_register_t; +DEFINE_XEN_GUEST_HANDLE(callback_register_t); + +/* + * Unregister a callback. + * + * Not all callbacks can be unregistered. -EINVAL will be returned if + * you attempt to unregister such a callback. + */ +#define CALLBACKOP_unregister 1 +struct callback_unregister { + uint16_t type; + uint16_t _unused; +}; +typedef struct callback_unregister callback_unregister_t; +DEFINE_XEN_GUEST_HANDLE(callback_unregister_t); + +#if __XEN_INTERFACE_VERSION__ < 0x00030207 +#undef CALLBACKTYPE_sysenter +#define CALLBACKTYPE_sysenter CALLBACKTYPE_sysenter_deprecated +#endif + +#endif /* __XEN_PUBLIC_CALLBACK_H__ */ diff --git a/include/xenpvdrivers/interface/dom0_ops.h b/include/xenpvdrivers/interface/dom0_ops.h new file mode 100644 index 0000000..5d2b324 --- /dev/null +++ b/include/xenpvdrivers/interface/dom0_ops.h @@ -0,0 +1,120 @@ +/****************************************************************************** + * dom0_ops.h + * + * Process command requests from domain-0 guest OS. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2002-2003, B Dragovic + * Copyright (c) 2002-2006, K Fraser + */ + +#ifndef __XEN_PUBLIC_DOM0_OPS_H__ +#define __XEN_PUBLIC_DOM0_OPS_H__ + +#include "xen.h" +#include "platform.h" + +#if __XEN_INTERFACE_VERSION__ >= 0x00030204 +#error "dom0_ops.h is a compatibility interface only" +#endif + +#define DOM0_INTERFACE_VERSION XENPF_INTERFACE_VERSION + +#define DOM0_SETTIME XENPF_settime +#define dom0_settime xenpf_settime +#define dom0_settime_t xenpf_settime_t + +#define DOM0_ADD_MEMTYPE XENPF_add_memtype +#define dom0_add_memtype xenpf_add_memtype +#define dom0_add_memtype_t xenpf_add_memtype_t + +#define DOM0_DEL_MEMTYPE XENPF_del_memtype +#define dom0_del_memtype xenpf_del_memtype +#define dom0_del_memtype_t xenpf_del_memtype_t + +#define DOM0_READ_MEMTYPE XENPF_read_memtype +#define dom0_read_memtype xenpf_read_memtype +#define dom0_read_memtype_t xenpf_read_memtype_t + +#define DOM0_MICROCODE XENPF_microcode_update +#define dom0_microcode xenpf_microcode_update +#define dom0_microcode_t xenpf_microcode_update_t + +#define DOM0_PLATFORM_QUIRK XENPF_platform_quirk +#define dom0_platform_quirk xenpf_platform_quirk +#define dom0_platform_quirk_t xenpf_platform_quirk_t + +typedef uint64_t cpumap_t; + +/* Unsupported legacy operation -- defined for API compatibility. */ +#define DOM0_MSR 15 +struct dom0_msr { + /* IN variables. */ + uint32_t write; + cpumap_t cpu_mask; + uint32_t msr; + uint32_t in1; + uint32_t in2; + /* OUT variables. */ + uint32_t out1; + uint32_t out2; +}; +typedef struct dom0_msr dom0_msr_t; +DEFINE_XEN_GUEST_HANDLE(dom0_msr_t); + +/* Unsupported legacy operation -- defined for API compatibility. */ +#define DOM0_PHYSICAL_MEMORY_MAP 40 +struct dom0_memory_map_entry { + uint64_t start, end; + uint32_t flags; /* reserved */ + uint8_t is_ram; +}; +typedef struct dom0_memory_map_entry dom0_memory_map_entry_t; +DEFINE_XEN_GUEST_HANDLE(dom0_memory_map_entry_t); + +struct dom0_op { + uint32_t cmd; + uint32_t interface_version; /* DOM0_INTERFACE_VERSION */ + union { + struct dom0_msr msr; + struct dom0_settime settime; + struct dom0_add_memtype add_memtype; + struct dom0_del_memtype del_memtype; + struct dom0_read_memtype read_memtype; + struct dom0_microcode microcode; + struct dom0_platform_quirk platform_quirk; + struct dom0_memory_map_entry physical_memory_map; + uint8_t pad[128]; + } u; +}; +typedef struct dom0_op dom0_op_t; +DEFINE_XEN_GUEST_HANDLE(dom0_op_t); + +#endif /* __XEN_PUBLIC_DOM0_OPS_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/domctl.h b/include/xenpvdrivers/interface/domctl.h new file mode 100644 index 0000000..af949f4 --- /dev/null +++ b/include/xenpvdrivers/interface/domctl.h @@ -0,0 +1,904 @@ +/****************************************************************************** + * domctl.h + * + * Domain management operations. For use by node control stack. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2002-2003, B Dragovic + * Copyright (c) 2002-2006, K Fraser + */ + +#ifndef __XEN_PUBLIC_DOMCTL_H__ +#define __XEN_PUBLIC_DOMCTL_H__ + +#if !defined(__XEN__) && !defined(__XEN_TOOLS__) +#error "domctl operations are intended for use by node control tools only" +#endif + +#include "xen.h" +#include "grant_table.h" + +#define XEN_DOMCTL_INTERFACE_VERSION 0x00000006 + +struct xenctl_cpumap { + XEN_GUEST_HANDLE_64(uint8) bitmap; + uint32_t nr_cpus; +}; + +/* + * NB. xen_domctl.domain is an IN/OUT parameter for this operation. + * If it is specified as zero, an id is auto-allocated and returned. + */ +/* XEN_DOMCTL_createdomain */ +struct xen_domctl_createdomain { + /* IN parameters */ + uint32_t ssidref; + xen_domain_handle_t handle; + /* Is this an HVM guest (as opposed to a PV guest)? */ +#define _XEN_DOMCTL_CDF_hvm_guest 0 +#define XEN_DOMCTL_CDF_hvm_guest (1U<<_XEN_DOMCTL_CDF_hvm_guest) + /* Use hardware-assisted paging if available? */ +#define _XEN_DOMCTL_CDF_hap 1 +#define XEN_DOMCTL_CDF_hap (1U<<_XEN_DOMCTL_CDF_hap) + /* Should domain memory integrity be verifed by tboot during Sx? */ +#define _XEN_DOMCTL_CDF_s3_integrity 2 +#define XEN_DOMCTL_CDF_s3_integrity (1U<<_XEN_DOMCTL_CDF_s3_integrity) + uint32_t flags; + /* Disable out-of-sync shadow page tables? */ +#define _XEN_DOMCTL_CDF_oos_off 3 +#define XEN_DOMCTL_CDF_oos_off (1U<<_XEN_DOMCTL_CDF_oos_off) +}; +typedef struct xen_domctl_createdomain xen_domctl_createdomain_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_createdomain_t); + +/* XEN_DOMCTL_getdomaininfo */ +struct xen_domctl_getdomaininfo { + /* OUT variables. */ + domid_t domain; /* Also echoed in domctl.domain */ + /* Domain is scheduled to die. */ +#define _XEN_DOMINF_dying 0 +#define XEN_DOMINF_dying (1U<<_XEN_DOMINF_dying) + /* Domain is an HVM guest (as opposed to a PV guest). */ +#define _XEN_DOMINF_hvm_guest 1 +#define XEN_DOMINF_hvm_guest (1U<<_XEN_DOMINF_hvm_guest) + /* The guest OS has shut down. */ +#define _XEN_DOMINF_shutdown 2 +#define XEN_DOMINF_shutdown (1U<<_XEN_DOMINF_shutdown) + /* Currently paused by control software. */ +#define _XEN_DOMINF_paused 3 +#define XEN_DOMINF_paused (1U<<_XEN_DOMINF_paused) + /* Currently blocked pending an event. */ +#define _XEN_DOMINF_blocked 4 +#define XEN_DOMINF_blocked (1U<<_XEN_DOMINF_blocked) + /* Domain is currently running. */ +#define _XEN_DOMINF_running 5 +#define XEN_DOMINF_running (1U<<_XEN_DOMINF_running) + /* Being debugged. */ +#define _XEN_DOMINF_debugged 6 +#define XEN_DOMINF_debugged (1U<<_XEN_DOMINF_debugged) + /* XEN_DOMINF_shutdown guest-supplied code. */ +#define XEN_DOMINF_shutdownmask 255 +#define XEN_DOMINF_shutdownshift 16 + uint32_t flags; /* XEN_DOMINF_* */ + uint64_aligned_t tot_pages; + uint64_aligned_t max_pages; + uint64_aligned_t shr_pages; + uint64_aligned_t shared_info_frame; /* GMFN of shared_info struct */ + uint64_aligned_t cpu_time; + uint32_t nr_online_vcpus; /* Number of VCPUs currently online. */ + uint32_t max_vcpu_id; /* Maximum VCPUID in use by this domain. */ + uint32_t ssidref; + xen_domain_handle_t handle; +}; +typedef struct xen_domctl_getdomaininfo xen_domctl_getdomaininfo_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_getdomaininfo_t); + + +/* XEN_DOMCTL_getmemlist */ +struct xen_domctl_getmemlist { + /* IN variables. */ + /* Max entries to write to output buffer. */ + uint64_aligned_t max_pfns; + /* Start index in guest's page list. */ + uint64_aligned_t start_pfn; + XEN_GUEST_HANDLE_64(uint64) buffer; + /* OUT variables. */ + uint64_aligned_t num_pfns; +}; +typedef struct xen_domctl_getmemlist xen_domctl_getmemlist_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_getmemlist_t); + + +/* XEN_DOMCTL_getpageframeinfo */ + +#define XEN_DOMCTL_PFINFO_LTAB_SHIFT 28 +#define XEN_DOMCTL_PFINFO_NOTAB (0x0U<<28) +#define XEN_DOMCTL_PFINFO_L1TAB (0x1U<<28) +#define XEN_DOMCTL_PFINFO_L2TAB (0x2U<<28) +#define XEN_DOMCTL_PFINFO_L3TAB (0x3U<<28) +#define XEN_DOMCTL_PFINFO_L4TAB (0x4U<<28) +#define XEN_DOMCTL_PFINFO_LTABTYPE_MASK (0x7U<<28) +#define XEN_DOMCTL_PFINFO_LPINTAB (0x1U<<31) +#define XEN_DOMCTL_PFINFO_XTAB (0xfU<<28) /* invalid page */ +#define XEN_DOMCTL_PFINFO_PAGEDTAB (0x8U<<28) +#define XEN_DOMCTL_PFINFO_LTAB_MASK (0xfU<<28) + +struct xen_domctl_getpageframeinfo { + /* IN variables. */ + uint64_aligned_t gmfn; /* GMFN to query */ + /* OUT variables. */ + /* Is the page PINNED to a type? */ + uint32_t type; /* see above type defs */ +}; +typedef struct xen_domctl_getpageframeinfo xen_domctl_getpageframeinfo_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_getpageframeinfo_t); + + +/* XEN_DOMCTL_getpageframeinfo2 */ +struct xen_domctl_getpageframeinfo2 { + /* IN variables. */ + uint64_aligned_t num; + /* IN/OUT variables. */ + XEN_GUEST_HANDLE_64(uint32) array; +}; +typedef struct xen_domctl_getpageframeinfo2 xen_domctl_getpageframeinfo2_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_getpageframeinfo2_t); + + +/* + * Control shadow pagetables operation + */ +/* XEN_DOMCTL_shadow_op */ + +/* Disable shadow mode. */ +#define XEN_DOMCTL_SHADOW_OP_OFF 0 + +/* Enable shadow mode (mode contains ORed XEN_DOMCTL_SHADOW_ENABLE_* flags). */ +#define XEN_DOMCTL_SHADOW_OP_ENABLE 32 + +/* Log-dirty bitmap operations. */ + /* Return the bitmap and clean internal copy for next round. */ +#define XEN_DOMCTL_SHADOW_OP_CLEAN 11 + /* Return the bitmap but do not modify internal copy. */ +#define XEN_DOMCTL_SHADOW_OP_PEEK 12 + +/* Memory allocation accessors. */ +#define XEN_DOMCTL_SHADOW_OP_GET_ALLOCATION 30 +#define XEN_DOMCTL_SHADOW_OP_SET_ALLOCATION 31 + +/* Legacy enable operations. */ + /* Equiv. to ENABLE with no mode flags. */ +#define XEN_DOMCTL_SHADOW_OP_ENABLE_TEST 1 + /* Equiv. to ENABLE with mode flag ENABLE_LOG_DIRTY. */ +#define XEN_DOMCTL_SHADOW_OP_ENABLE_LOGDIRTY 2 + /* Equiv. to ENABLE with mode flags ENABLE_REFCOUNT and ENABLE_TRANSLATE. */ +#define XEN_DOMCTL_SHADOW_OP_ENABLE_TRANSLATE 3 + +/* Mode flags for XEN_DOMCTL_SHADOW_OP_ENABLE. */ + /* + * Shadow pagetables are refcounted: guest does not use explicit mmu + * operations nor write-protect its pagetables. + */ +#define XEN_DOMCTL_SHADOW_ENABLE_REFCOUNT (1 << 1) + /* + * Log pages in a bitmap as they are dirtied. + * Used for live relocation to determine which pages must be re-sent. + */ +#define XEN_DOMCTL_SHADOW_ENABLE_LOG_DIRTY (1 << 2) + /* + * Automatically translate GPFNs into MFNs. + */ +#define XEN_DOMCTL_SHADOW_ENABLE_TRANSLATE (1 << 3) + /* + * Xen does not steal virtual address space from the guest. + * Requires HVM support. + */ +#define XEN_DOMCTL_SHADOW_ENABLE_EXTERNAL (1 << 4) + +struct xen_domctl_shadow_op_stats { + uint32_t fault_count; + uint32_t dirty_count; +}; +typedef struct xen_domctl_shadow_op_stats xen_domctl_shadow_op_stats_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_shadow_op_stats_t); + +struct xen_domctl_shadow_op { + /* IN variables. */ + uint32_t op; /* XEN_DOMCTL_SHADOW_OP_* */ + + /* OP_ENABLE */ + uint32_t mode; /* XEN_DOMCTL_SHADOW_ENABLE_* */ + + /* OP_GET_ALLOCATION / OP_SET_ALLOCATION */ + uint32_t mb; /* Shadow memory allocation in MB */ + + /* OP_PEEK / OP_CLEAN */ + XEN_GUEST_HANDLE_64(uint8) dirty_bitmap; + uint64_aligned_t pages; /* Size of buffer. Updated with actual size. */ + struct xen_domctl_shadow_op_stats stats; +}; +typedef struct xen_domctl_shadow_op xen_domctl_shadow_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_shadow_op_t); + + +/* XEN_DOMCTL_max_mem */ +struct xen_domctl_max_mem { + /* IN variables. */ + uint64_aligned_t max_memkb; +}; +typedef struct xen_domctl_max_mem xen_domctl_max_mem_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_max_mem_t); + + +/* XEN_DOMCTL_setvcpucontext */ +/* XEN_DOMCTL_getvcpucontext */ +struct xen_domctl_vcpucontext { + uint32_t vcpu; /* IN */ + XEN_GUEST_HANDLE_64(vcpu_guest_context_t) ctxt; /* IN/OUT */ +}; +typedef struct xen_domctl_vcpucontext xen_domctl_vcpucontext_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_vcpucontext_t); + + +/* XEN_DOMCTL_getvcpuinfo */ +struct xen_domctl_getvcpuinfo { + /* IN variables. */ + uint32_t vcpu; + /* OUT variables. */ + uint8_t online; /* currently online (not hotplugged)? */ + uint8_t blocked; /* blocked waiting for an event? */ + uint8_t running; /* currently scheduled on its CPU? */ + uint64_aligned_t cpu_time; /* total cpu time consumed (ns) */ + uint32_t cpu; /* current mapping */ +}; +typedef struct xen_domctl_getvcpuinfo xen_domctl_getvcpuinfo_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_getvcpuinfo_t); + + +/* Get/set which physical cpus a vcpu can execute on. */ +/* XEN_DOMCTL_setvcpuaffinity */ +/* XEN_DOMCTL_getvcpuaffinity */ +struct xen_domctl_vcpuaffinity { + uint32_t vcpu; /* IN */ + struct xenctl_cpumap cpumap; /* IN/OUT */ +}; +typedef struct xen_domctl_vcpuaffinity xen_domctl_vcpuaffinity_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_vcpuaffinity_t); + + +/* XEN_DOMCTL_max_vcpus */ +struct xen_domctl_max_vcpus { + uint32_t max; /* maximum number of vcpus */ +}; +typedef struct xen_domctl_max_vcpus xen_domctl_max_vcpus_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_max_vcpus_t); + + +/* XEN_DOMCTL_scheduler_op */ +/* Scheduler types. */ +#define XEN_SCHEDULER_SEDF 4 +#define XEN_SCHEDULER_CREDIT 5 +/* Set or get info? */ +#define XEN_DOMCTL_SCHEDOP_putinfo 0 +#define XEN_DOMCTL_SCHEDOP_getinfo 1 +struct xen_domctl_scheduler_op { + uint32_t sched_id; /* XEN_SCHEDULER_* */ + uint32_t cmd; /* XEN_DOMCTL_SCHEDOP_* */ + union { + struct xen_domctl_sched_sedf { + uint64_aligned_t period; + uint64_aligned_t slice; + uint64_aligned_t latency; + uint32_t extratime; + uint32_t weight; + } sedf; + struct xen_domctl_sched_credit { + uint16_t weight; + uint16_t cap; + } credit; + } u; +}; +typedef struct xen_domctl_scheduler_op xen_domctl_scheduler_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_scheduler_op_t); + + +/* XEN_DOMCTL_setdomainhandle */ +struct xen_domctl_setdomainhandle { + xen_domain_handle_t handle; +}; +typedef struct xen_domctl_setdomainhandle xen_domctl_setdomainhandle_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_setdomainhandle_t); + + +/* XEN_DOMCTL_setdebugging */ +struct xen_domctl_setdebugging { + uint8_t enable; +}; +typedef struct xen_domctl_setdebugging xen_domctl_setdebugging_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_setdebugging_t); + + +/* XEN_DOMCTL_irq_permission */ +struct xen_domctl_irq_permission { + uint8_t pirq; + uint8_t allow_access; /* flag to specify enable/disable of IRQ access */ +}; +typedef struct xen_domctl_irq_permission xen_domctl_irq_permission_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_irq_permission_t); + + +/* XEN_DOMCTL_iomem_permission */ +struct xen_domctl_iomem_permission { + uint64_aligned_t first_mfn;/* first page (physical page number) in range */ + uint64_aligned_t nr_mfns; /* number of pages in range (>0) */ + uint8_t allow_access; /* allow (!0) or deny (0) access to range? */ +}; +typedef struct xen_domctl_iomem_permission xen_domctl_iomem_permission_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_iomem_permission_t); + + +/* XEN_DOMCTL_ioport_permission */ +struct xen_domctl_ioport_permission { + uint32_t first_port; /* first port int range */ + uint32_t nr_ports; /* size of port range */ + uint8_t allow_access; /* allow or deny access to range? */ +}; +typedef struct xen_domctl_ioport_permission xen_domctl_ioport_permission_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_ioport_permission_t); + + +/* XEN_DOMCTL_hypercall_init */ +struct xen_domctl_hypercall_init { + uint64_aligned_t gmfn; /* GMFN to be initialised */ +}; +typedef struct xen_domctl_hypercall_init xen_domctl_hypercall_init_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_hypercall_init_t); + + +/* XEN_DOMCTL_arch_setup */ +#define _XEN_DOMAINSETUP_hvm_guest 0 +#define XEN_DOMAINSETUP_hvm_guest (1UL<<_XEN_DOMAINSETUP_hvm_guest) +#define _XEN_DOMAINSETUP_query 1 /* Get parameters (for save) */ +#define XEN_DOMAINSETUP_query (1UL<<_XEN_DOMAINSETUP_query) +#define _XEN_DOMAINSETUP_sioemu_guest 2 +#define XEN_DOMAINSETUP_sioemu_guest (1UL<<_XEN_DOMAINSETUP_sioemu_guest) +typedef struct xen_domctl_arch_setup { + uint64_aligned_t flags; /* XEN_DOMAINSETUP_* */ +#ifdef __ia64__ + uint64_aligned_t bp; /* mpaddr of boot param area */ + uint64_aligned_t maxmem; /* Highest memory address for MDT. */ + uint64_aligned_t xsi_va; /* Xen shared_info area virtual address. */ + uint32_t hypercall_imm; /* Break imm for Xen hypercalls. */ + int8_t vhpt_size_log2; /* Log2 of VHPT size. */ +#endif +} xen_domctl_arch_setup_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_arch_setup_t); + + +/* XEN_DOMCTL_settimeoffset */ +struct xen_domctl_settimeoffset { + int32_t time_offset_seconds; /* applied to domain wallclock time */ +}; +typedef struct xen_domctl_settimeoffset xen_domctl_settimeoffset_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_settimeoffset_t); + +/* XEN_DOMCTL_gethvmcontext */ +/* XEN_DOMCTL_sethvmcontext */ +typedef struct xen_domctl_hvmcontext { + uint32_t size; /* IN/OUT: size of buffer / bytes filled */ + XEN_GUEST_HANDLE_64(uint8) buffer; /* IN/OUT: data, or call + * gethvmcontext with NULL + * buffer to get size req'd */ +} xen_domctl_hvmcontext_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_hvmcontext_t); + + +/* XEN_DOMCTL_set_address_size */ +/* XEN_DOMCTL_get_address_size */ +typedef struct xen_domctl_address_size { + uint32_t size; +} xen_domctl_address_size_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_address_size_t); + + +/* XEN_DOMCTL_real_mode_area */ +struct xen_domctl_real_mode_area { + uint32_t log; /* log2 of Real Mode Area size */ +}; +typedef struct xen_domctl_real_mode_area xen_domctl_real_mode_area_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_real_mode_area_t); + + +/* XEN_DOMCTL_sendtrigger */ +#define XEN_DOMCTL_SENDTRIGGER_NMI 0 +#define XEN_DOMCTL_SENDTRIGGER_RESET 1 +#define XEN_DOMCTL_SENDTRIGGER_INIT 2 +#define XEN_DOMCTL_SENDTRIGGER_POWER 3 +struct xen_domctl_sendtrigger { + uint32_t trigger; /* IN */ + uint32_t vcpu; /* IN */ +}; +typedef struct xen_domctl_sendtrigger xen_domctl_sendtrigger_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_sendtrigger_t); + + +/* Assign PCI device to HVM guest. Sets up IOMMU structures. */ +/* XEN_DOMCTL_assign_device */ +/* XEN_DOMCTL_test_assign_device */ +/* XEN_DOMCTL_deassign_device */ +struct xen_domctl_assign_device { + uint32_t machine_bdf; /* machine PCI ID of assigned device */ +}; +typedef struct xen_domctl_assign_device xen_domctl_assign_device_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_assign_device_t); + +/* Retrieve sibling devices infomation of machine_bdf */ +/* XEN_DOMCTL_get_device_group */ +struct xen_domctl_get_device_group { + uint32_t machine_bdf; /* IN */ + uint32_t max_sdevs; /* IN */ + uint32_t num_sdevs; /* OUT */ + XEN_GUEST_HANDLE_64(uint32) sdev_array; /* OUT */ +}; +typedef struct xen_domctl_get_device_group xen_domctl_get_device_group_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_get_device_group_t); + +/* Pass-through interrupts: bind real irq -> hvm devfn. */ +/* XEN_DOMCTL_bind_pt_irq */ +/* XEN_DOMCTL_unbind_pt_irq */ +typedef enum pt_irq_type_e { + PT_IRQ_TYPE_PCI, + PT_IRQ_TYPE_ISA, + PT_IRQ_TYPE_MSI, + PT_IRQ_TYPE_MSI_TRANSLATE, +} pt_irq_type_t; +struct xen_domctl_bind_pt_irq { + uint32_t machine_irq; + pt_irq_type_t irq_type; + uint32_t hvm_domid; + + union { + struct { + uint8_t isa_irq; + } isa; + struct { + uint8_t bus; + uint8_t device; + uint8_t intx; + } pci; + struct { + uint8_t gvec; + uint32_t gflags; + uint64_aligned_t gtable; + } msi; + } u; +}; +typedef struct xen_domctl_bind_pt_irq xen_domctl_bind_pt_irq_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_bind_pt_irq_t); + + +/* Bind machine I/O address range -> HVM address range. */ +/* XEN_DOMCTL_memory_mapping */ +#define DPCI_ADD_MAPPING 1 +#define DPCI_REMOVE_MAPPING 0 +struct xen_domctl_memory_mapping { + uint64_aligned_t first_gfn; /* first page (hvm guest phys page) in range */ + uint64_aligned_t first_mfn; /* first page (machine page) in range */ + uint64_aligned_t nr_mfns; /* number of pages in range (>0) */ + uint32_t add_mapping; /* add or remove mapping */ + uint32_t padding; /* padding for 64-bit aligned structure */ +}; +typedef struct xen_domctl_memory_mapping xen_domctl_memory_mapping_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_memory_mapping_t); + + +/* Bind machine I/O port range -> HVM I/O port range. */ +/* XEN_DOMCTL_ioport_mapping */ +struct xen_domctl_ioport_mapping { + uint32_t first_gport; /* first guest IO port*/ + uint32_t first_mport; /* first machine IO port */ + uint32_t nr_ports; /* size of port range */ + uint32_t add_mapping; /* add or remove mapping */ +}; +typedef struct xen_domctl_ioport_mapping xen_domctl_ioport_mapping_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_ioport_mapping_t); + + +/* + * Pin caching type of RAM space for x86 HVM domU. + */ +/* XEN_DOMCTL_pin_mem_cacheattr */ +/* Caching types: these happen to be the same as x86 MTRR/PAT type codes. */ +#define XEN_DOMCTL_MEM_CACHEATTR_UC 0 +#define XEN_DOMCTL_MEM_CACHEATTR_WC 1 +#define XEN_DOMCTL_MEM_CACHEATTR_WT 4 +#define XEN_DOMCTL_MEM_CACHEATTR_WP 5 +#define XEN_DOMCTL_MEM_CACHEATTR_WB 6 +#define XEN_DOMCTL_MEM_CACHEATTR_UCM 7 +struct xen_domctl_pin_mem_cacheattr { + uint64_aligned_t start, end; + uint32_t type; /* XEN_DOMCTL_MEM_CACHEATTR_* */ +}; +typedef struct xen_domctl_pin_mem_cacheattr xen_domctl_pin_mem_cacheattr_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_pin_mem_cacheattr_t); + + +/* XEN_DOMCTL_set_ext_vcpucontext */ +/* XEN_DOMCTL_get_ext_vcpucontext */ +struct xen_domctl_ext_vcpucontext { + /* IN: VCPU that this call applies to. */ + uint32_t vcpu; + /* + * SET: Size of struct (IN) + * GET: Size of struct (OUT) + */ + uint32_t size; +#if defined(__i386__) || defined(__x86_64__) + /* SYSCALL from 32-bit mode and SYSENTER callback information. */ + /* NB. SYSCALL from 64-bit mode is contained in vcpu_guest_context_t */ + uint64_aligned_t syscall32_callback_eip; + uint64_aligned_t sysenter_callback_eip; + uint16_t syscall32_callback_cs; + uint16_t sysenter_callback_cs; + uint8_t syscall32_disables_events; + uint8_t sysenter_disables_events; +#endif +}; +typedef struct xen_domctl_ext_vcpucontext xen_domctl_ext_vcpucontext_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_ext_vcpucontext_t); + +/* + * Set optimizaton features for a domain + */ +/* XEN_DOMCTL_set_opt_feature */ +struct xen_domctl_set_opt_feature { +#if defined(__ia64__) + struct xen_ia64_opt_feature optf; +#else + /* Make struct non-empty: do not depend on this field name! */ + uint64_t dummy; +#endif +}; +typedef struct xen_domctl_set_opt_feature xen_domctl_set_opt_feature_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_set_opt_feature_t); + +/* + * Set the target domain for a domain + */ +/* XEN_DOMCTL_set_target */ +struct xen_domctl_set_target { + domid_t target; +}; +typedef struct xen_domctl_set_target xen_domctl_set_target_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_set_target_t); + +#if defined(__i386__) || defined(__x86_64__) +# define XEN_CPUID_INPUT_UNUSED 0xFFFFFFFF +/* XEN_DOMCTL_set_cpuid */ +struct xen_domctl_cpuid { + uint32_t input[2]; + uint32_t eax; + uint32_t ebx; + uint32_t ecx; + uint32_t edx; +}; +typedef struct xen_domctl_cpuid xen_domctl_cpuid_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_cpuid_t); +#endif + +/* XEN_DOMCTL_subscribe */ +struct xen_domctl_subscribe { + uint32_t port; /* IN */ +}; +typedef struct xen_domctl_subscribe xen_domctl_subscribe_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_subscribe_t); + +/* + * Define the maximum machine address size which should be allocated + * to a guest. + */ +/* XEN_DOMCTL_set_machine_address_size */ +/* XEN_DOMCTL_get_machine_address_size */ + +/* + * Do not inject spurious page faults into this domain. + */ +/* XEN_DOMCTL_suppress_spurious_page_faults */ + +/* XEN_DOMCTL_debug_op */ +#define XEN_DOMCTL_DEBUG_OP_SINGLE_STEP_OFF 0 +#define XEN_DOMCTL_DEBUG_OP_SINGLE_STEP_ON 1 +struct xen_domctl_debug_op { + uint32_t op; /* IN */ + uint32_t vcpu; /* IN */ +}; +typedef struct xen_domctl_debug_op xen_domctl_debug_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_debug_op_t); + +/* + * Request a particular record from the HVM context + */ +/* XEN_DOMCTL_gethvmcontext_partial */ +typedef struct xen_domctl_hvmcontext_partial { + uint32_t type; /* IN: Type of record required */ + uint32_t instance; /* IN: Instance of that type */ + XEN_GUEST_HANDLE_64(uint8) buffer; /* OUT: buffer to write record into */ +} xen_domctl_hvmcontext_partial_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_hvmcontext_partial_t); + +/* XEN_DOMCTL_disable_migrate */ +typedef struct xen_domctl_disable_migrate { + uint32_t disable; /* IN: 1: disable migration and restore */ +} xen_domctl_disable_migrate_t; + + +/* XEN_DOMCTL_gettscinfo */ +/* XEN_DOMCTL_settscinfo */ +struct xen_guest_tsc_info { + uint32_t tsc_mode; + uint32_t gtsc_khz; + uint32_t incarnation; + uint32_t pad; + uint64_aligned_t elapsed_nsec; +}; +typedef struct xen_guest_tsc_info xen_guest_tsc_info_t; +DEFINE_XEN_GUEST_HANDLE(xen_guest_tsc_info_t); +typedef struct xen_domctl_tsc_info { + XEN_GUEST_HANDLE_64(xen_guest_tsc_info_t) out_info; /* OUT */ + xen_guest_tsc_info_t info; /* IN */ +} xen_domctl_tsc_info_t; + +/* XEN_DOMCTL_gdbsx_guestmemio guest mem io */ +struct xen_domctl_gdbsx_memio { + /* IN */ + uint64_aligned_t pgd3val;/* optional: init_mm.pgd[3] value */ + uint64_aligned_t gva; /* guest virtual address */ + uint64_aligned_t uva; /* user buffer virtual address */ + uint32_t len; /* number of bytes to read/write */ + uint8_t gwr; /* 0 = read from guest. 1 = write to guest */ + /* OUT */ + uint32_t remain; /* bytes remaining to be copied */ +}; + +/* XEN_DOMCTL_gdbsx_pausevcpu */ +/* XEN_DOMCTL_gdbsx_unpausevcpu */ +struct xen_domctl_gdbsx_pauseunp_vcpu { /* pause/unpause a vcpu */ + uint32_t vcpu; /* which vcpu */ +}; + +/* XEN_DOMCTL_gdbsx_domstatus */ +struct xen_domctl_gdbsx_domstatus { + /* OUT */ + uint8_t paused; /* is the domain paused */ + uint32_t vcpu_id; /* any vcpu in an event? */ + uint32_t vcpu_ev; /* if yes, what event? */ +}; + +/* + * Memory event operations + */ + +/* XEN_DOMCTL_mem_event_op */ + +/* Add and remove memory handlers */ +#define XEN_DOMCTL_MEM_EVENT_OP_ENABLE 0 +#define XEN_DOMCTL_MEM_EVENT_OP_DISABLE 1 + +/* + * Page memory in and out. + */ +#define XEN_DOMCTL_MEM_EVENT_OP_PAGING (1 << 0) + +/* Domain memory paging */ +#define XEN_DOMCTL_MEM_EVENT_OP_PAGING_NOMINATE 0 +#define XEN_DOMCTL_MEM_EVENT_OP_PAGING_EVICT 1 +#define XEN_DOMCTL_MEM_EVENT_OP_PAGING_PREP 2 +#define XEN_DOMCTL_MEM_EVENT_OP_PAGING_RESUME 3 + +struct xen_domctl_mem_event_op { + uint32_t op; /* XEN_DOMCTL_MEM_EVENT_OP_* */ + uint32_t mode; /* XEN_DOMCTL_MEM_EVENT_ENABLE_* */ + + /* OP_ENABLE */ + uint64_aligned_t shared_addr; /* IN: Virtual address of shared page */ + uint64_aligned_t ring_addr; /* IN: Virtual address of ring page */ + + /* Other OPs */ + uint64_aligned_t gfn; /* IN: gfn of page being operated on */ +}; +typedef struct xen_domctl_mem_event_op xen_domctl_mem_event_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_mem_event_op_t); + +/* + * Memory sharing operations + */ +/* XEN_DOMCTL_mem_sharing_op */ + +#define XEN_DOMCTL_MEM_SHARING_OP_CONTROL 0 +#define XEN_DOMCTL_MEM_SHARING_OP_NOMINATE_GFN 1 +#define XEN_DOMCTL_MEM_SHARING_OP_NOMINATE_GREF 2 +#define XEN_DOMCTL_MEM_SHARING_OP_SHARE 3 +#define XEN_DOMCTL_MEM_SHARING_OP_RESUME 4 +#define XEN_DOMCTL_MEM_SHARING_OP_DEBUG_GFN 5 +#define XEN_DOMCTL_MEM_SHARING_OP_DEBUG_MFN 6 +#define XEN_DOMCTL_MEM_SHARING_OP_DEBUG_GREF 7 + +#define XEN_DOMCTL_MEM_SHARING_S_HANDLE_INVALID (-10) +#define XEN_DOMCTL_MEM_SHARING_C_HANDLE_INVALID (-9) + +struct xen_domctl_mem_sharing_op { + uint8_t op; /* XEN_DOMCTL_MEM_EVENT_OP_* */ + + union { + uint8_t enable; /* OP_CONTROL */ + + struct mem_sharing_op_nominate { /* OP_NOMINATE_xxx */ + union { + uint64_aligned_t gfn; /* IN: gfn to nominate */ + uint32_t grant_ref; /* IN: grant ref to nominate */ + } u; + uint64_aligned_t handle; /* OUT: the handle */ + } nominate; + struct mem_sharing_op_share { /* OP_SHARE */ + uint64_aligned_t source_handle; /* IN: handle to the source page */ + uint64_aligned_t client_handle; /* IN: handle to the client page */ + } share; + struct mem_sharing_op_debug { /* OP_DEBUG_xxx */ + union { + uint64_aligned_t gfn; /* IN: gfn to debug */ + uint64_aligned_t mfn; /* IN: mfn to debug */ + grant_ref_t gref; /* IN: gref to debug */ + } u; + } debug; + } u; +}; +typedef struct xen_domctl_mem_sharing_op xen_domctl_mem_sharing_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_mem_sharing_op_t); + + +struct xen_domctl { + uint32_t cmd; +#define XEN_DOMCTL_createdomain 1 +#define XEN_DOMCTL_destroydomain 2 +#define XEN_DOMCTL_pausedomain 3 +#define XEN_DOMCTL_unpausedomain 4 +#define XEN_DOMCTL_getdomaininfo 5 +#define XEN_DOMCTL_getmemlist 6 +#define XEN_DOMCTL_getpageframeinfo 7 +#define XEN_DOMCTL_getpageframeinfo2 8 +#define XEN_DOMCTL_setvcpuaffinity 9 +#define XEN_DOMCTL_shadow_op 10 +#define XEN_DOMCTL_max_mem 11 +#define XEN_DOMCTL_setvcpucontext 12 +#define XEN_DOMCTL_getvcpucontext 13 +#define XEN_DOMCTL_getvcpuinfo 14 +#define XEN_DOMCTL_max_vcpus 15 +#define XEN_DOMCTL_scheduler_op 16 +#define XEN_DOMCTL_setdomainhandle 17 +#define XEN_DOMCTL_setdebugging 18 +#define XEN_DOMCTL_irq_permission 19 +#define XEN_DOMCTL_iomem_permission 20 +#define XEN_DOMCTL_ioport_permission 21 +#define XEN_DOMCTL_hypercall_init 22 +#define XEN_DOMCTL_arch_setup 23 +#define XEN_DOMCTL_settimeoffset 24 +#define XEN_DOMCTL_getvcpuaffinity 25 +#define XEN_DOMCTL_real_mode_area 26 +#define XEN_DOMCTL_resumedomain 27 +#define XEN_DOMCTL_sendtrigger 28 +#define XEN_DOMCTL_subscribe 29 +#define XEN_DOMCTL_gethvmcontext 33 +#define XEN_DOMCTL_sethvmcontext 34 +#define XEN_DOMCTL_set_address_size 35 +#define XEN_DOMCTL_get_address_size 36 +#define XEN_DOMCTL_assign_device 37 +#define XEN_DOMCTL_bind_pt_irq 38 +#define XEN_DOMCTL_memory_mapping 39 +#define XEN_DOMCTL_ioport_mapping 40 +#define XEN_DOMCTL_pin_mem_cacheattr 41 +#define XEN_DOMCTL_set_ext_vcpucontext 42 +#define XEN_DOMCTL_get_ext_vcpucontext 43 +#define XEN_DOMCTL_set_opt_feature 44 +#define XEN_DOMCTL_test_assign_device 45 +#define XEN_DOMCTL_set_target 46 +#define XEN_DOMCTL_deassign_device 47 +#define XEN_DOMCTL_unbind_pt_irq 48 +#define XEN_DOMCTL_set_cpuid 49 +#define XEN_DOMCTL_get_device_group 50 +#define XEN_DOMCTL_set_machine_address_size 51 +#define XEN_DOMCTL_get_machine_address_size 52 +#define XEN_DOMCTL_suppress_spurious_page_faults 53 +#define XEN_DOMCTL_debug_op 54 +#define XEN_DOMCTL_gethvmcontext_partial 55 +#define XEN_DOMCTL_mem_event_op 56 +#define XEN_DOMCTL_mem_sharing_op 57 +#define XEN_DOMCTL_disable_migrate 58 +#define XEN_DOMCTL_gettscinfo 59 +#define XEN_DOMCTL_settscinfo 60 +#define XEN_DOMCTL_gdbsx_guestmemio 1000 +#define XEN_DOMCTL_gdbsx_pausevcpu 1001 +#define XEN_DOMCTL_gdbsx_unpausevcpu 1002 +#define XEN_DOMCTL_gdbsx_domstatus 1003 + uint32_t interface_version; /* XEN_DOMCTL_INTERFACE_VERSION */ + domid_t domain; + union { + struct xen_domctl_createdomain createdomain; + struct xen_domctl_getdomaininfo getdomaininfo; + struct xen_domctl_getmemlist getmemlist; + struct xen_domctl_getpageframeinfo getpageframeinfo; + struct xen_domctl_getpageframeinfo2 getpageframeinfo2; + struct xen_domctl_vcpuaffinity vcpuaffinity; + struct xen_domctl_shadow_op shadow_op; + struct xen_domctl_max_mem max_mem; + struct xen_domctl_vcpucontext vcpucontext; + struct xen_domctl_getvcpuinfo getvcpuinfo; + struct xen_domctl_max_vcpus max_vcpus; + struct xen_domctl_scheduler_op scheduler_op; + struct xen_domctl_setdomainhandle setdomainhandle; + struct xen_domctl_setdebugging setdebugging; + struct xen_domctl_irq_permission irq_permission; + struct xen_domctl_iomem_permission iomem_permission; + struct xen_domctl_ioport_permission ioport_permission; + struct xen_domctl_hypercall_init hypercall_init; + struct xen_domctl_arch_setup arch_setup; + struct xen_domctl_settimeoffset settimeoffset; + struct xen_domctl_disable_migrate disable_migrate; + struct xen_domctl_tsc_info tsc_info; + struct xen_domctl_real_mode_area real_mode_area; + struct xen_domctl_hvmcontext hvmcontext; + struct xen_domctl_hvmcontext_partial hvmcontext_partial; + struct xen_domctl_address_size address_size; + struct xen_domctl_sendtrigger sendtrigger; + struct xen_domctl_get_device_group get_device_group; + struct xen_domctl_assign_device assign_device; + struct xen_domctl_bind_pt_irq bind_pt_irq; + struct xen_domctl_memory_mapping memory_mapping; + struct xen_domctl_ioport_mapping ioport_mapping; + struct xen_domctl_pin_mem_cacheattr pin_mem_cacheattr; + struct xen_domctl_ext_vcpucontext ext_vcpucontext; + struct xen_domctl_set_opt_feature set_opt_feature; + struct xen_domctl_set_target set_target; + struct xen_domctl_subscribe subscribe; + struct xen_domctl_debug_op debug_op; + struct xen_domctl_mem_event_op mem_event_op; + struct xen_domctl_mem_sharing_op mem_sharing_op; +#if defined(__i386__) || defined(__x86_64__) + struct xen_domctl_cpuid cpuid; +#endif + struct xen_domctl_gdbsx_memio gdbsx_guest_memio; + struct xen_domctl_gdbsx_pauseunp_vcpu gdbsx_pauseunp_vcpu; + struct xen_domctl_gdbsx_domstatus gdbsx_domstatus; + uint8_t pad[128]; + } u; +}; +typedef struct xen_domctl xen_domctl_t; +DEFINE_XEN_GUEST_HANDLE(xen_domctl_t); + +#endif /* __XEN_PUBLIC_DOMCTL_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/elfnote.h b/include/xenpvdrivers/interface/elfnote.h new file mode 100644 index 0000000..ff29847 --- /dev/null +++ b/include/xenpvdrivers/interface/elfnote.h @@ -0,0 +1,244 @@ +/****************************************************************************** + * elfnote.h + * + * Definitions used for the Xen ELF notes. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2006, Ian Campbell, XenSource Ltd. + */ + +#ifndef __XEN_PUBLIC_ELFNOTE_H__ +#define __XEN_PUBLIC_ELFNOTE_H__ + +/* + * The notes should live in a PT_NOTE segment and have "Xen" in the + * name field. + * + * Numeric types are either 4 or 8 bytes depending on the content of + * the desc field. + * + * LEGACY indicated the fields in the legacy __xen_guest string which + * this a note type replaces. + */ + +/* + * NAME=VALUE pair (string). + */ +#define XEN_ELFNOTE_INFO 0 + +/* + * The virtual address of the entry point (numeric). + * + * LEGACY: VIRT_ENTRY + */ +#define XEN_ELFNOTE_ENTRY 1 + +/* The virtual address of the hypercall transfer page (numeric). + * + * LEGACY: HYPERCALL_PAGE. (n.b. legacy value is a physical page + * number not a virtual address) + */ +#define XEN_ELFNOTE_HYPERCALL_PAGE 2 + +/* The virtual address where the kernel image should be mapped (numeric). + * + * Defaults to 0. + * + * LEGACY: VIRT_BASE + */ +#define XEN_ELFNOTE_VIRT_BASE 3 + +/* + * The offset of the ELF paddr field from the acutal required + * psuedo-physical address (numeric). + * + * This is used to maintain backwards compatibility with older kernels + * which wrote __PAGE_OFFSET into that field. This field defaults to 0 + * if not present. + * + * LEGACY: ELF_PADDR_OFFSET. (n.b. legacy default is VIRT_BASE) + */ +#define XEN_ELFNOTE_PADDR_OFFSET 4 + +/* + * The version of Xen that we work with (string). + * + * LEGACY: XEN_VER + */ +#define XEN_ELFNOTE_XEN_VERSION 5 + +/* + * The name of the guest operating system (string). + * + * LEGACY: GUEST_OS + */ +#define XEN_ELFNOTE_GUEST_OS 6 + +/* + * The version of the guest operating system (string). + * + * LEGACY: GUEST_VER + */ +#define XEN_ELFNOTE_GUEST_VERSION 7 + +/* + * The loader type (string). + * + * LEGACY: LOADER + */ +#define XEN_ELFNOTE_LOADER 8 + +/* + * The kernel supports PAE (x86/32 only, string = "yes", "no" or + * "bimodal"). + * + * For compatibility with Xen 3.0.3 and earlier the "bimodal" setting + * may be given as "yes,bimodal" which will cause older Xen to treat + * this kernel as PAE. + * + * LEGACY: PAE (n.b. The legacy interface included a provision to + * indicate 'extended-cr3' support allowing L3 page tables to be + * placed above 4G. It is assumed that any kernel new enough to use + * these ELF notes will include this and therefore "yes" here is + * equivalent to "yes[entended-cr3]" in the __xen_guest interface. + */ +#define XEN_ELFNOTE_PAE_MODE 9 + +/* + * The features supported/required by this kernel (string). + * + * The string must consist of a list of feature names (as given in + * features.h, without the "XENFEAT_" prefix) separated by '|' + * characters. If a feature is required for the kernel to function + * then the feature name must be preceded by a '!' character. + * + * LEGACY: FEATURES + */ +#define XEN_ELFNOTE_FEATURES 10 + +/* + * The kernel requires the symbol table to be loaded (string = "yes" or "no") + * LEGACY: BSD_SYMTAB (n.b. The legacy treated the presence or absence + * of this string as a boolean flag rather than requiring "yes" or + * "no". + */ +#define XEN_ELFNOTE_BSD_SYMTAB 11 + +/* + * The lowest address the hypervisor hole can begin at (numeric). + * + * This must not be set higher than HYPERVISOR_VIRT_START. Its presence + * also indicates to the hypervisor that the kernel can deal with the + * hole starting at a higher address. + */ +#define XEN_ELFNOTE_HV_START_LOW 12 + +/* + * List of maddr_t-sized mask/value pairs describing how to recognize + * (non-present) L1 page table entries carrying valid MFNs (numeric). + */ +#define XEN_ELFNOTE_L1_MFN_VALID 13 + +/* + * Whether or not the guest supports cooperative suspend cancellation. + */ +#define XEN_ELFNOTE_SUSPEND_CANCEL 14 + +/* + * The (non-default) location the initial phys-to-machine map should be + * placed at by the hypervisor (Dom0) or the tools (DomU). + * The kernel must be prepared for this mapping to be established using + * large pages, despite such otherwise not being available to guests. + * The kernel must also be able to handle the page table pages used for + * this mapping not being accessible through the initial mapping. + * (Only x86-64 supports this at present.) + */ +#define XEN_ELFNOTE_INIT_P2M 15 + +/* + * The number of the highest elfnote defined. + */ +#define XEN_ELFNOTE_MAX XEN_ELFNOTE_INIT_P2M + +/* + * System information exported through crash notes. + * + * The kexec / kdump code will create one XEN_ELFNOTE_CRASH_INFO + * note in case of a system crash. This note will contain various + * information about the system, see xen/include/xen/elfcore.h. + */ +#define XEN_ELFNOTE_CRASH_INFO 0x1000001 + +/* + * System registers exported through crash notes. + * + * The kexec / kdump code will create one XEN_ELFNOTE_CRASH_REGS + * note per cpu in case of a system crash. This note is architecture + * specific and will contain registers not saved in the "CORE" note. + * See xen/include/xen/elfcore.h for more information. + */ +#define XEN_ELFNOTE_CRASH_REGS 0x1000002 + + +/* + * xen dump-core none note. + * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_NONE + * in its dump file to indicate that the file is xen dump-core + * file. This note doesn't have any other information. + * See tools/libxc/xc_core.h for more information. + */ +#define XEN_ELFNOTE_DUMPCORE_NONE 0x2000000 + +/* + * xen dump-core header note. + * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_HEADER + * in its dump file. + * See tools/libxc/xc_core.h for more information. + */ +#define XEN_ELFNOTE_DUMPCORE_HEADER 0x2000001 + +/* + * xen dump-core xen version note. + * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_XEN_VERSION + * in its dump file. It contains the xen version obtained via the + * XENVER hypercall. + * See tools/libxc/xc_core.h for more information. + */ +#define XEN_ELFNOTE_DUMPCORE_XEN_VERSION 0x2000002 + +/* + * xen dump-core format version note. + * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION + * in its dump file. It contains a format version identifier. + * See tools/libxc/xc_core.h for more information. + */ +#define XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION 0x2000003 + +#endif /* __XEN_PUBLIC_ELFNOTE_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/event_channel.h b/include/xenpvdrivers/interface/event_channel.h new file mode 100644 index 0000000..f2970e3 --- /dev/null +++ b/include/xenpvdrivers/interface/event_channel.h @@ -0,0 +1,257 @@ +/****************************************************************************** + * event_channel.h + * + * Event channels between domains. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2003-2004, K A Fraser. + */ + +#ifndef __XEN_PUBLIC_EVENT_CHANNEL_H__ +#define __XEN_PUBLIC_EVENT_CHANNEL_H__ + +#include + +/* + * Prototype for this hypercall is: + * int event_channel_op(int cmd, void *args) + * @cmd == EVTCHNOP_??? (event-channel operation). + * @args == Operation-specific extra arguments (NULL if none). + */ + +typedef uint32_t evtchn_port_t; +DEFINE_XEN_GUEST_HANDLE(evtchn_port_t); + +/* + * EVTCHNOP_alloc_unbound: Allocate a port in domain and mark as + * accepting interdomain bindings from domain . A fresh port + * is allocated in and returned as . + * NOTES: + * 1. If the caller is unprivileged then must be DOMID_SELF. + * 2. may be DOMID_SELF, allowing loopback connections. + */ +#define EVTCHNOP_alloc_unbound 6 +struct evtchn_alloc_unbound { + /* IN parameters */ + domid_t dom, remote_dom; + /* OUT parameters */ + evtchn_port_t port; +}; +typedef struct evtchn_alloc_unbound evtchn_alloc_unbound_t; + +/* + * EVTCHNOP_bind_interdomain: Construct an interdomain event channel between + * the calling domain and . must identify + * a port that is unbound and marked as accepting bindings from the calling + * domain. A fresh port is allocated in the calling domain and returned as + * . + * NOTES: + * 2. may be DOMID_SELF, allowing loopback connections. + */ +#define EVTCHNOP_bind_interdomain 0 +struct evtchn_bind_interdomain { + /* IN parameters. */ + domid_t remote_dom; + evtchn_port_t remote_port; + /* OUT parameters. */ + evtchn_port_t local_port; +}; +typedef struct evtchn_bind_interdomain evtchn_bind_interdomain_t; + +/* + * EVTCHNOP_bind_virq: Bind a local event channel to VIRQ on specified + * vcpu. + * NOTES: + * 1. Virtual IRQs are classified as per-vcpu or global. See the VIRQ list + * in xen.h for the classification of each VIRQ. + * 2. Global VIRQs must be allocated on VCPU0 but can subsequently be + * re-bound via EVTCHNOP_bind_vcpu. + * 3. Per-vcpu VIRQs may be bound to at most one event channel per vcpu. + * The allocated event channel is bound to the specified vcpu and the + * binding cannot be changed. + */ +#define EVTCHNOP_bind_virq 1 +struct evtchn_bind_virq { + /* IN parameters. */ + uint32_t virq; + uint32_t vcpu; + /* OUT parameters. */ + evtchn_port_t port; +}; +typedef struct evtchn_bind_virq evtchn_bind_virq_t; + +/* + * EVTCHNOP_bind_pirq: Bind a local event channel to PIRQ . + * NOTES: + * 1. A physical IRQ may be bound to at most one event channel per domain. + * 2. Only a sufficiently-privileged domain may bind to a physical IRQ. + */ +#define EVTCHNOP_bind_pirq 2 +struct evtchn_bind_pirq { + /* IN parameters. */ + uint32_t pirq; +#define BIND_PIRQ__WILL_SHARE 1 + uint32_t flags; /* BIND_PIRQ__* */ + /* OUT parameters. */ + evtchn_port_t port; +}; +typedef struct evtchn_bind_pirq evtchn_bind_pirq_t; + +/* + * EVTCHNOP_bind_ipi: Bind a local event channel to receive events. + * NOTES: + * 1. The allocated event channel is bound to the specified vcpu. The binding + * may not be changed. + */ +#define EVTCHNOP_bind_ipi 7 +struct evtchn_bind_ipi { + uint32_t vcpu; + /* OUT parameters. */ + evtchn_port_t port; +}; +typedef struct evtchn_bind_ipi evtchn_bind_ipi_t; + +/* + * EVTCHNOP_close: Close a local event channel . If the channel is + * interdomain then the remote end is placed in the unbound state + * (EVTCHNSTAT_unbound), awaiting a new connection. + */ +#define EVTCHNOP_close 3 +struct evtchn_close { + /* IN parameters. */ + evtchn_port_t port; +}; +typedef struct evtchn_close evtchn_close_t; + +/* + * EVTCHNOP_send: Send an event to the remote end of the channel whose local + * endpoint is . + */ +#define EVTCHNOP_send 4 +struct evtchn_send { + /* IN parameters. */ + evtchn_port_t port; +}; +typedef struct evtchn_send evtchn_send_t; + +/* + * EVTCHNOP_status: Get the current status of the communication channel which + * has an endpoint at . + * NOTES: + * 1. may be specified as DOMID_SELF. + * 2. Only a sufficiently-privileged domain may obtain the status of an event + * channel for which is not DOMID_SELF. + */ +#define EVTCHNOP_status 5 +struct evtchn_status { + /* IN parameters */ + domid_t dom; + evtchn_port_t port; + /* OUT parameters */ +#define EVTCHNSTAT_closed 0 /* Channel is not in use. */ +#define EVTCHNSTAT_unbound 1 /* Channel is waiting interdom connection.*/ +#define EVTCHNSTAT_interdomain 2 /* Channel is connected to remote domain. */ +#define EVTCHNSTAT_pirq 3 /* Channel is bound to a phys IRQ line. */ +#define EVTCHNSTAT_virq 4 /* Channel is bound to a virtual IRQ line */ +#define EVTCHNSTAT_ipi 5 /* Channel is bound to a virtual IPI line */ + uint32_t status; + uint32_t vcpu; /* VCPU to which this channel is bound. */ + union { + struct { + domid_t dom; + } unbound; /* EVTCHNSTAT_unbound */ + struct { + domid_t dom; + evtchn_port_t port; + } interdomain; /* EVTCHNSTAT_interdomain */ + uint32_t pirq; /* EVTCHNSTAT_pirq */ + uint32_t virq; /* EVTCHNSTAT_virq */ + } u; +}; +typedef struct evtchn_status evtchn_status_t; + +/* + * EVTCHNOP_bind_vcpu: Specify which vcpu a channel should notify when an + * event is pending. + * NOTES: + * 1. IPI-bound channels always notify the vcpu specified at bind time. + * This binding cannot be changed. + * 2. Per-VCPU VIRQ channels always notify the vcpu specified at bind time. + * This binding cannot be changed. + * 3. All other channels notify vcpu0 by default. This default is set when + * the channel is allocated (a port that is freed and subsequently reused + * has its binding reset to vcpu0). + */ +#define EVTCHNOP_bind_vcpu 8 +struct evtchn_bind_vcpu { + /* IN parameters. */ + evtchn_port_t port; + uint32_t vcpu; +}; +typedef struct evtchn_bind_vcpu evtchn_bind_vcpu_t; + +/* + * EVTCHNOP_unmask: Unmask the specified local event-channel port and deliver + * a notification to the appropriate VCPU if an event is pending. + */ +#define EVTCHNOP_unmask 9 +struct evtchn_unmask { + /* IN parameters. */ + evtchn_port_t port; +}; +typedef struct evtchn_unmask evtchn_unmask_t; + +/* + * EVTCHNOP_reset: Close all event channels associated with specified domain. + * NOTES: + * 1. may be specified as DOMID_SELF. + * 2. Only a sufficiently-privileged domain may specify other than DOMID_SELF. + */ +#define EVTCHNOP_reset 10 +struct evtchn_reset { + /* IN parameters. */ + domid_t dom; +}; +typedef struct evtchn_reset evtchn_reset_t; + +/* + * Argument to event_channel_op_compat() hypercall. Superceded by new + * event_channel_op() hypercall since 0x00030202. + */ +struct evtchn_op { + uint32_t cmd; /* EVTCHNOP_* */ + union { + struct evtchn_alloc_unbound alloc_unbound; + struct evtchn_bind_interdomain bind_interdomain; + struct evtchn_bind_virq bind_virq; + struct evtchn_bind_pirq bind_pirq; + struct evtchn_bind_ipi bind_ipi; + struct evtchn_close close; + struct evtchn_send send; + struct evtchn_status status; + struct evtchn_bind_vcpu bind_vcpu; + struct evtchn_unmask unmask; + } u; +}; +DEFINE_GUEST_HANDLE_STRUCT(evtchn_op); +typedef struct evtchn_op evtchn_op_t; +DEFINE_XEN_GUEST_HANDLE(evtchn_op_t); + +#endif /* __XEN_PUBLIC_EVENT_CHANNEL_H__ */ diff --git a/include/xenpvdrivers/interface/features.h b/include/xenpvdrivers/interface/features.h new file mode 100644 index 0000000..bf1a9a7 --- /dev/null +++ b/include/xenpvdrivers/interface/features.h @@ -0,0 +1,73 @@ +/****************************************************************************** + * features.h + * + * Feature flags, reported by XENVER_get_features. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2006, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_FEATURES_H__ +#define __XEN_PUBLIC_FEATURES_H__ + +/* + * If set, the guest does not need to write-protect its pagetables, and can + * update them via direct writes. + */ +#define XENFEAT_writable_page_tables 0 + +/* + * If set, the guest does not need to write-protect its segment descriptor + * tables, and can update them via direct writes. + */ +#define XENFEAT_writable_descriptor_tables 1 + +/* + * If set, translation between the guest's 'pseudo-physical' address space + * and the host's machine address space are handled by the hypervisor. In this + * mode the guest does not need to perform phys-to/from-machine translations + * when performing page table operations. + */ +#define XENFEAT_auto_translated_physmap 2 + +/* If set, the guest is running in supervisor mode (e.g., x86 ring 0). */ +#define XENFEAT_supervisor_mode_kernel 3 + +/* + * If set, the guest does not need to allocate x86 PAE page directories + * below 4GB. This flag is usually implied by auto_translated_physmap. + */ +#define XENFEAT_pae_pgdir_above_4gb 4 + +/* x86: Does this Xen host support the MMU_PT_UPDATE_PRESERVE_AD hypercall? */ +#define XENFEAT_mmu_pt_update_preserve_ad 5 + +/* x86: Does this Xen host support the MMU_{CLEAR,COPY}_PAGE hypercall? */ +#define XENFEAT_highmem_assist 6 + +/* + * If set, GNTTABOP_map_grant_ref honors flags to be placed into guest kernel + * available pte bits. + */ +#define XENFEAT_gnttab_map_avail_bits 7 + +#define XENFEAT_NR_SUBMAPS 1 + +#endif /* __XEN_PUBLIC_FEATURES_H__ */ diff --git a/include/xenpvdrivers/interface/grant_table.h b/include/xenpvdrivers/interface/grant_table.h new file mode 100644 index 0000000..9d37ee4 --- /dev/null +++ b/include/xenpvdrivers/interface/grant_table.h @@ -0,0 +1,594 @@ +/****************************************************************************** + * grant_table.h + * + * Interface for granting foreign access to page frames, and receiving + * page-ownership transfers. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2004, K A Fraser + */ + +#ifndef __XEN_PUBLIC_GRANT_TABLE_H__ +#define __XEN_PUBLIC_GRANT_TABLE_H__ + +#include "xen.h" + +/*********************************** + * GRANT TABLE REPRESENTATION + */ + +/* Some rough guidelines on accessing and updating grant-table entries + * in a concurrency-safe manner. For more information, Linux contains a + * reference implementation for guest OSes (arch/xen/kernel/grant_table.c). + * + * NB. WMB is a no-op on current-generation x86 processors. However, a + * compiler barrier will still be required. + * + * Introducing a valid entry into the grant table: + * 1. Write ent->domid. + * 2. Write ent->frame: + * GTF_permit_access: Frame to which access is permitted. + * GTF_accept_transfer: Pseudo-phys frame slot being filled by new + * frame, or zero if none. + * 3. Write memory barrier (WMB). + * 4. Write ent->flags, inc. valid type. + * + * Invalidating an unused GTF_permit_access entry: + * 1. flags = ent->flags. + * 2. Observe that !(flags & (GTF_reading|GTF_writing)). + * 3. Check result of SMP-safe CMPXCHG(&ent->flags, flags, 0). + * NB. No need for WMB as reuse of entry is control-dependent on success of + * step 3, and all architectures guarantee ordering of ctrl-dep writes. + * + * Invalidating an in-use GTF_permit_access entry: + * This cannot be done directly. Request assistance from the domain controller + * which can set a timeout on the use of a grant entry and take necessary + * action. (NB. This is not yet implemented!). + * + * Invalidating an unused GTF_accept_transfer entry: + * 1. flags = ent->flags. + * 2. Observe that !(flags & GTF_transfer_committed). [*] + * 3. Check result of SMP-safe CMPXCHG(&ent->flags, flags, 0). + * NB. No need for WMB as reuse of entry is control-dependent on success of + * step 3, and all architectures guarantee ordering of ctrl-dep writes. + * [*] If GTF_transfer_committed is set then the grant entry is 'committed'. + * The guest must /not/ modify the grant entry until the address of the + * transferred frame is written. It is safe for the guest to spin waiting + * for this to occur (detect by observing GTF_transfer_completed in + * ent->flags). + * + * Invalidating a committed GTF_accept_transfer entry: + * 1. Wait for (ent->flags & GTF_transfer_completed). + * + * Changing a GTF_permit_access from writable to read-only: + * Use SMP-safe CMPXCHG to set GTF_readonly, while checking !GTF_writing. + * + * Changing a GTF_permit_access from read-only to writable: + * Use SMP-safe bit-setting instruction. + */ + +/* + * Reference to a grant entry in a specified domain's grant table. + */ +typedef uint32_t grant_ref_t; + +/* + * A grant table comprises a packed array of grant entries in one or more + * page frames shared between Xen and a guest. + * [XEN]: This field is written by Xen and read by the sharing guest. + * [GST]: This field is written by the guest and read by Xen. + */ + +/* + * Version 1 of the grant table entry structure is maintained purely + * for backwards compatibility. New guests should use version 2. + */ +#if __XEN_INTERFACE_VERSION__ < 0x0003020a +#define grant_entry_v1 grant_entry +#define grant_entry_v1_t grant_entry_t +#endif +struct grant_entry_v1 { + /* GTF_xxx: various type and flag information. [XEN,GST] */ + uint16_t flags; + /* The domain being granted foreign privileges. [GST] */ + domid_t domid; + /* + * GTF_permit_access: Frame that @domid is allowed to map and access. [GST] + * GTF_accept_transfer: Frame whose ownership transferred by @domid. [XEN] + */ + uint32_t frame; +}; +typedef struct grant_entry_v1 grant_entry_v1_t; + +/* + * Type of grant entry. + * GTF_invalid: This grant entry grants no privileges. + * GTF_permit_access: Allow @domid to map/access @frame. + * GTF_accept_transfer: Allow @domid to transfer ownership of one page frame + * to this guest. Xen writes the page number to @frame. + * GTF_transitive: Allow @domid to transitively access a subrange of + * @trans_grant in @trans_domid. No mappings are allowed. + */ +#define GTF_invalid (0U<<0) +#define GTF_permit_access (1U<<0) +#define GTF_accept_transfer (2U<<0) +#define GTF_transitive (3U<<0) +#define GTF_type_mask (3U<<0) + +/* + * Subflags for GTF_permit_access. + * GTF_readonly: Restrict @domid to read-only mappings and accesses. [GST] + * GTF_reading: Grant entry is currently mapped for reading by @domid. [XEN] + * GTF_writing: Grant entry is currently mapped for writing by @domid. [XEN] + * GTF_PAT, GTF_PWT, GTF_PCD: (x86) cache attribute flags for the grant [GST] + * GTF_sub_page: Grant access to only a subrange of the page. @domid + * will only be allowed to copy from the grant, and not + * map it. [GST] + */ +#define _GTF_readonly (2) +#define GTF_readonly (1U<<_GTF_readonly) +#define _GTF_reading (3) +#define GTF_reading (1U<<_GTF_reading) +#define _GTF_writing (4) +#define GTF_writing (1U<<_GTF_writing) +#define _GTF_PWT (5) +#define GTF_PWT (1U<<_GTF_PWT) +#define _GTF_PCD (6) +#define GTF_PCD (1U<<_GTF_PCD) +#define _GTF_PAT (7) +#define GTF_PAT (1U<<_GTF_PAT) +#define _GTF_sub_page (8) +#define GTF_sub_page (1U<<_GTF_sub_page) + +/* + * Subflags for GTF_accept_transfer: + * GTF_transfer_committed: Xen sets this flag to indicate that it is committed + * to transferring ownership of a page frame. When a guest sees this flag + * it must /not/ modify the grant entry until GTF_transfer_completed is + * set by Xen. + * GTF_transfer_completed: It is safe for the guest to spin-wait on this flag + * after reading GTF_transfer_committed. Xen will always write the frame + * address, followed by ORing this flag, in a timely manner. + */ +#define _GTF_transfer_committed (2) +#define GTF_transfer_committed (1U<<_GTF_transfer_committed) +#define _GTF_transfer_completed (3) +#define GTF_transfer_completed (1U<<_GTF_transfer_completed) + +/* + * Version 2 grant table entries. These fulfil the same role as + * version 1 entries, but can represent more complicated operations. + * Any given domain will have either a version 1 or a version 2 table, + * and every entry in the table will be the same version. + * + * The interface by which domains use grant references does not depend + * on the grant table version in use by the other domain. + */ +#if __XEN_INTERFACE_VERSION__ >= 0x0003020a +/* + * Version 1 and version 2 grant entries share a common prefix. The + * fields of the prefix are documented as part of struct + * grant_entry_v1. + */ +struct grant_entry_header { + uint16_t flags; + domid_t domid; +}; +typedef struct grant_entry_header grant_entry_header_t; + +/* + * Version 2 of the grant entry structure. + */ +union grant_entry_v2 { + grant_entry_header_t hdr; + + /* + * This member is used for V1-style full page grants, where either: + * + * -- hdr.type is GTF_accept_transfer, or + * -- hdr.type is GTF_permit_access and GTF_sub_page is not set. + * + * In that case, the frame field has the same semantics as the + * field of the same name in the V1 entry structure. + */ + struct { + grant_entry_header_t hdr; + uint32_t pad0; + uint64_t frame; + } full_page; + + /* + * If the grant type is GTF_grant_access and GTF_sub_page is set, + * @domid is allowed to access bytes [@page_off,@page_off+@length) + * in frame @frame. + */ + struct { + grant_entry_header_t hdr; + uint16_t page_off; + uint16_t length; + uint64_t frame; + } sub_page; + + /* + * If the grant is GTF_transitive, @domid is allowed to use the + * grant @gref in domain @trans_domid, as if it was the local + * domain. Obviously, the transitive access must be compatible + * with the original grant. + * + * The current version of Xen does not allow transitive grants + * to be mapped. + */ + struct { + grant_entry_header_t hdr; + domid_t trans_domid; + uint16_t pad0; + grant_ref_t gref; + } transitive; + + uint32_t __spacer[4]; /* Pad to a power of two */ +}; +typedef union grant_entry_v2 grant_entry_v2_t; + +typedef uint16_t grant_status_t; + +#endif /* __XEN_INTERFACE_VERSION__ */ + +/*********************************** + * GRANT TABLE QUERIES AND USES + */ + +/* + * Handle to track a mapping created via a grant reference. + */ +typedef uint32_t grant_handle_t; + +/* + * GNTTABOP_map_grant_ref: Map the grant entry (,) for access + * by devices and/or host CPUs. If successful, is a tracking number + * that must be presented later to destroy the mapping(s). On error, + * is a negative status code. + * NOTES: + * 1. If GNTMAP_device_map is specified then is the address + * via which I/O devices may access the granted frame. + * 2. If GNTMAP_host_map is specified then a mapping will be added at + * either a host virtual address in the current address space, or at + * a PTE at the specified machine address. The type of mapping to + * perform is selected through the GNTMAP_contains_pte flag, and the + * address is specified in . + * 3. Mappings should only be destroyed via GNTTABOP_unmap_grant_ref. If a + * host mapping is destroyed by other means then it is *NOT* guaranteed + * to be accounted to the correct grant reference! + */ +#define GNTTABOP_map_grant_ref 0 +struct gnttab_map_grant_ref { + /* IN parameters. */ + uint64_t host_addr; + uint32_t flags; /* GNTMAP_* */ + grant_ref_t ref; + domid_t dom; + /* OUT parameters. */ + int16_t status; /* GNTST_* */ + grant_handle_t handle; + uint64_t dev_bus_addr; +}; +DEFINE_GUEST_HANDLE_STRUCT(gnttab_map_grant_ref); +typedef struct gnttab_map_grant_ref gnttab_map_grant_ref_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_map_grant_ref_t); + +/* + * GNTTABOP_unmap_grant_ref: Destroy one or more grant-reference mappings + * tracked by . If or is zero, that + * field is ignored. If non-zero, they must refer to a device/host mapping + * that is tracked by + * NOTES: + * 1. The call may fail in an undefined manner if either mapping is not + * tracked by . + * 3. After executing a batch of unmaps, it is guaranteed that no stale + * mappings will remain in the device or host TLBs. + */ +#define GNTTABOP_unmap_grant_ref 1 +struct gnttab_unmap_grant_ref { + /* IN parameters. */ + uint64_t host_addr; + uint64_t dev_bus_addr; + grant_handle_t handle; + /* OUT parameters. */ + int16_t status; /* GNTST_* */ +}; +DEFINE_GUEST_HANDLE_STRUCT(gnttab_unmap_grant_ref); +typedef struct gnttab_unmap_grant_ref gnttab_unmap_grant_ref_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_unmap_grant_ref_t); + +/* + * GNTTABOP_setup_table: Set up a grant table for comprising at least + * pages. The frame addresses are written to the . + * Only addresses are written, even if the table is larger. + * NOTES: + * 1. may be specified as DOMID_SELF. + * 2. Only a sufficiently-privileged domain may specify != DOMID_SELF. + * 3. Xen may not support more than a single grant-table page per domain. + */ +#define GNTTABOP_setup_table 2 +struct gnttab_setup_table { + /* IN parameters. */ + domid_t dom; + uint32_t nr_frames; + /* OUT parameters. */ + int16_t status; /* GNTST_* */ + XEN_GUEST_HANDLE(ulong) frame_list; +}; +DEFINE_GUEST_HANDLE_STRUCT(gnttab_setup_table); +typedef struct gnttab_setup_table gnttab_setup_table_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_setup_table_t); + +/* + * GNTTABOP_dump_table: Dump the contents of the grant table to the + * xen console. Debugging use only. + */ +#define GNTTABOP_dump_table 3 +struct gnttab_dump_table { + /* IN parameters. */ + domid_t dom; + /* OUT parameters. */ + int16_t status; /* GNTST_* */ +}; +DEFINE_GUEST_HANDLE_STRUCT(gnttab_dump_table); +typedef struct gnttab_dump_table gnttab_dump_table_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_dump_table_t); + +/* + * GNTTABOP_transfer_grant_ref: Transfer to a foreign domain. The + * foreign domain has previously registered its interest in the transfer via + * . + * + * Note that, even if the transfer fails, the specified page no longer belongs + * to the calling domain *unless* the error is GNTST_bad_page. + */ +#define GNTTABOP_transfer 4 +struct gnttab_transfer { + /* IN parameters. */ + xen_pfn_t mfn; + domid_t domid; + grant_ref_t ref; + /* OUT parameters. */ + int16_t status; +}; +DEFINE_GUEST_HANDLE_STRUCT(gnttab_transfer); +typedef struct gnttab_transfer gnttab_transfer_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_transfer_t); + + +/* + * GNTTABOP_copy: Hypervisor based copy + * source and destinations can be eithers MFNs or, for foreign domains, + * grant references. the foreign domain has to grant read/write access + * in its grant table. + * + * The flags specify what type source and destinations are (either MFN + * or grant reference). + * + * Note that this can also be used to copy data between two domains + * via a third party if the source and destination domains had previously + * grant appropriate access to their pages to the third party. + * + * source_offset specifies an offset in the source frame, dest_offset + * the offset in the target frame and len specifies the number of + * bytes to be copied. + */ + +#define _GNTCOPY_source_gref (0) +#define GNTCOPY_source_gref (1<<_GNTCOPY_source_gref) +#define _GNTCOPY_dest_gref (1) +#define GNTCOPY_dest_gref (1<<_GNTCOPY_dest_gref) +#define _GNTCOPY_can_fail (2) +#define GNTCOPY_can_fail (1<<_GNTCOPY_can_fail) + +#define GNTTABOP_copy 5 +typedef struct gnttab_copy { + /* IN parameters. */ + struct { + union { + grant_ref_t ref; + xen_pfn_t gmfn; + } u; + domid_t domid; + uint16_t offset; + } source, dest; + uint16_t len; + uint16_t flags; /* GNTCOPY_* */ + /* OUT parameters. */ + int16_t status; +} gnttab_copy_t; +DEFINE_GUEST_HANDLE_STRUCT(gnttab_copy); +DEFINE_XEN_GUEST_HANDLE(gnttab_copy_t); + +/* + * GNTTABOP_query_size: Query the current and maximum sizes of the shared + * grant table. + * NOTES: + * 1. may be specified as DOMID_SELF. + * 2. Only a sufficiently-privileged domain may specify != DOMID_SELF. + */ +#define GNTTABOP_query_size 6 +struct gnttab_query_size { + /* IN parameters. */ + domid_t dom; + /* OUT parameters. */ + uint32_t nr_frames; + uint32_t max_nr_frames; + int16_t status; /* GNTST_* */ +}; +DEFINE_GUEST_HANDLE_STRUCT(gnttab_query_size); +typedef struct gnttab_query_size gnttab_query_size_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_query_size_t); + +/* + * GNTTABOP_unmap_and_replace: Destroy one or more grant-reference mappings + * tracked by but atomically replace the page table entry with one + * pointing to the machine address under . will be + * redirected to the null entry. + * NOTES: + * 1. The call may fail in an undefined manner if either mapping is not + * tracked by . + * 2. After executing a batch of unmaps, it is guaranteed that no stale + * mappings will remain in the device or host TLBs. + */ +#define GNTTABOP_unmap_and_replace 7 +struct gnttab_unmap_and_replace { + /* IN parameters. */ + uint64_t host_addr; + uint64_t new_addr; + grant_handle_t handle; + /* OUT parameters. */ + int16_t status; /* GNTST_* */ +}; +typedef struct gnttab_unmap_and_replace gnttab_unmap_and_replace_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_unmap_and_replace_t); + +#if __XEN_INTERFACE_VERSION__ >= 0x0003020a +/* + * GNTTABOP_set_version: Request a particular version of the grant + * table shared table structure. This operation can only be performed + * once in any given domain. It must be performed before any grants + * are activated; otherwise, the domain will be stuck with version 1. + * The only defined versions are 1 and 2. + */ +#define GNTTABOP_set_version 8 +struct gnttab_set_version { + /* IN/OUT parameters */ + uint32_t version; +}; +typedef struct gnttab_set_version gnttab_set_version_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_set_version_t); + + +/* + * GNTTABOP_get_status_frames: Get the list of frames used to store grant + * status for . In grant format version 2, the status is separated + * from the other shared grant fields to allow more efficient synchronization + * using barriers instead of atomic cmpexch operations. + * specify the size of vector . + * The frame addresses are returned in the . + * Only addresses are returned, even if the table is larger. + * NOTES: + * 1. may be specified as DOMID_SELF. + * 2. Only a sufficiently-privileged domain may specify != DOMID_SELF. + */ +#define GNTTABOP_get_status_frames 9 +struct gnttab_get_status_frames { + /* IN parameters. */ + uint32_t nr_frames; + domid_t dom; + /* OUT parameters. */ + int16_t status; /* GNTST_* */ + XEN_GUEST_HANDLE(uint64_t) frame_list; +}; +typedef struct gnttab_get_status_frames gnttab_get_status_frames_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_get_status_frames_t); + +/* + * GNTTABOP_get_version: Get the grant table version which is in + * effect for domain . + */ +#define GNTTABOP_get_version 10 +struct gnttab_get_version { + /* IN parameters */ + domid_t dom; + uint16_t pad; + /* OUT parameters */ + uint32_t version; +}; +typedef struct gnttab_get_version gnttab_get_version_t; +DEFINE_XEN_GUEST_HANDLE(gnttab_get_version_t); + +#endif /* __XEN_INTERFACE_VERSION__ */ + +/* + * Bitfield values for gnttab_map_grant_ref.flags. + */ + /* Map the grant entry for access by I/O devices. */ +#define _GNTMAP_device_map (0) +#define GNTMAP_device_map (1<<_GNTMAP_device_map) + /* Map the grant entry for access by host CPUs. */ +#define _GNTMAP_host_map (1) +#define GNTMAP_host_map (1<<_GNTMAP_host_map) + /* Accesses to the granted frame will be restricted to read-only access. */ +#define _GNTMAP_readonly (2) +#define GNTMAP_readonly (1<<_GNTMAP_readonly) + /* + * GNTMAP_host_map subflag: + * 0 => The host mapping is usable only by the guest OS. + * 1 => The host mapping is usable by guest OS + current application. + */ +#define _GNTMAP_application_map (3) +#define GNTMAP_application_map (1<<_GNTMAP_application_map) + + /* + * GNTMAP_contains_pte subflag: + * 0 => This map request contains a host virtual address. + * 1 => This map request contains the machine addess of the PTE to update. + */ +#define _GNTMAP_contains_pte (4) +#define GNTMAP_contains_pte (1<<_GNTMAP_contains_pte) + +#define _GNTMAP_can_fail (5) +#define GNTMAP_can_fail (1<<_GNTMAP_can_fail) + +/* + * Bits to be placed in guest kernel available PTE bits (architecture + * dependent; only supported when XENFEAT_gnttab_map_avail_bits is set). + */ +#define _GNTMAP_guest_avail0 (16) +#define GNTMAP_guest_avail_mask ((uint32_t)~0 << _GNTMAP_guest_avail0) + +/* + * Values for error status returns. All errors are -ve. + */ +#define GNTST_okay (0) /* Normal return. */ +#define GNTST_general_error (-1) /* General undefined error. */ +#define GNTST_bad_domain (-2) /* Unrecognsed domain id. */ +#define GNTST_bad_gntref (-3) /* Unrecognised or inappropriate gntref. */ +#define GNTST_bad_handle (-4) /* Unrecognised or inappropriate handle. */ +#define GNTST_bad_virt_addr (-5) /* Inappropriate virtual address to map. */ +#define GNTST_bad_dev_addr (-6) /* Inappropriate device address to unmap.*/ +#define GNTST_no_device_space (-7) /* Out of space in I/O MMU. */ +#define GNTST_permission_denied (-8) /* Not enough privilege for operation. */ +#define GNTST_bad_page (-9) /* Specified page was invalid for op. */ +#define GNTST_bad_copy_arg (-10) /* copy arguments cross page boundary. */ +#define GNTST_address_too_big (-11) /* transfer page address too large. */ +#define GNTST_eagain (-12) /* Could not map at the moment. Retry. */ + +#define GNTTABOP_error_msgs { \ + "okay", \ + "undefined error", \ + "unrecognised domain id", \ + "invalid grant reference", \ + "invalid mapping handle", \ + "invalid virtual address", \ + "invalid device address", \ + "no spare translation slot in the I/O MMU", \ + "permission denied", \ + "bad page", \ + "copy arguments cross page boundary", \ + "page address size too large", \ + "could not map at the moment, retry" \ +} + +#endif /* __XEN_PUBLIC_GRANT_TABLE_H__ */ diff --git a/include/xenpvdrivers/interface/hvm/e820.h b/include/xenpvdrivers/interface/hvm/e820.h new file mode 100644 index 0000000..5bdc227 --- /dev/null +++ b/include/xenpvdrivers/interface/hvm/e820.h @@ -0,0 +1,34 @@ + +/* + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_HVM_E820_H__ +#define __XEN_PUBLIC_HVM_E820_H__ + +/* E820 location in HVM virtual address space. */ +#define HVM_E820_PAGE 0x00090000 +#define HVM_E820_NR_OFFSET 0x000001E8 +#define HVM_E820_OFFSET 0x000002D0 + +#define HVM_BELOW_4G_RAM_END 0xF0000000 +#define HVM_BELOW_4G_MMIO_START HVM_BELOW_4G_RAM_END +#define HVM_BELOW_4G_MMIO_LENGTH ((1ULL << 32) - HVM_BELOW_4G_MMIO_START) + +#endif /* __XEN_PUBLIC_HVM_E820_H__ */ diff --git a/include/xenpvdrivers/interface/hvm/hvm_info_table.h b/include/xenpvdrivers/interface/hvm/hvm_info_table.h new file mode 100644 index 0000000..adb3fb9 --- /dev/null +++ b/include/xenpvdrivers/interface/hvm/hvm_info_table.h @@ -0,0 +1,75 @@ +/****************************************************************************** + * hvm/hvm_info_table.h + * + * HVM parameter and information table, written into guest memory map. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__ +#define __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__ + +#define HVM_INFO_PFN 0x09F +#define HVM_INFO_OFFSET 0x800 +#define HVM_INFO_PADDR ((HVM_INFO_PFN << 12) + HVM_INFO_OFFSET) + +/* Maximum we can support with current vLAPIC ID mapping. */ +#define HVM_MAX_VCPUS 128 + +struct hvm_info_table { + char signature[8]; /* "HVM INFO" */ + uint32_t length; + uint8_t checksum; + + /* Should firmware build ACPI tables? */ + uint8_t acpi_enabled; + + /* Should firmware build APIC descriptors (APIC MADT / MP BIOS)? */ + uint8_t apic_mode; + + /* How many CPUs does this domain have? */ + uint32_t nr_vcpus; + + /* + * MEMORY MAP provided by HVM domain builder. + * Notes: + * 1. page_to_phys(x) = x << 12 + * 2. If a field is zero, the corresponding range does not exist. + */ + /* + * 0x0 to page_to_phys(low_mem_pgend)-1: + * RAM below 4GB (except for VGA hole 0xA0000-0xBFFFF) + */ + uint32_t low_mem_pgend; + /* + * page_to_phys(reserved_mem_pgstart) to 0xFFFFFFFF: + * Reserved for special memory mappings + */ + uint32_t reserved_mem_pgstart; + /* + * 0x100000000 to page_to_phys(high_mem_pgend)-1: + * RAM above 4GB + */ + uint32_t high_mem_pgend; + + /* Bitmap of which CPUs are online at boot time. */ + uint8_t vcpu_online[HVM_MAX_VCPUS/8]; +}; + +#endif /* __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__ */ diff --git a/include/xenpvdrivers/interface/hvm/hvm_op.h b/include/xenpvdrivers/interface/hvm/hvm_op.h new file mode 100644 index 0000000..c5b2598 --- /dev/null +++ b/include/xenpvdrivers/interface/hvm/hvm_op.h @@ -0,0 +1,133 @@ +/* + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_HVM_HVM_OP_H__ +#define __XEN_PUBLIC_HVM_HVM_OP_H__ + +#include "../xen.h" + +/* Get/set subcommands: extra argument == pointer to xen_hvm_param struct. */ +#define HVMOP_set_param 0 +#define HVMOP_get_param 1 +struct xen_hvm_param { + domid_t domid; /* IN */ + uint32_t index; /* IN */ + uint64_t value; /* IN/OUT */ +}; +typedef struct xen_hvm_param xen_hvm_param_t; +DEFINE_XEN_GUEST_HANDLE(xen_hvm_param_t); + +/* Set the logical level of one of a domain's PCI INTx wires. */ +#define HVMOP_set_pci_intx_level 2 +struct xen_hvm_set_pci_intx_level { + /* Domain to be updated. */ + domid_t domid; + /* PCI INTx identification in PCI topology (domain:bus:device:intx). */ + uint8_t domain, bus, device, intx; + /* Assertion level (0 = unasserted, 1 = asserted). */ + uint8_t level; +}; +typedef struct xen_hvm_set_pci_intx_level xen_hvm_set_pci_intx_level_t; +DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_pci_intx_level_t); + +/* Set the logical level of one of a domain's ISA IRQ wires. */ +#define HVMOP_set_isa_irq_level 3 +struct xen_hvm_set_isa_irq_level { + /* Domain to be updated. */ + domid_t domid; + /* ISA device identification, by ISA IRQ (0-15). */ + uint8_t isa_irq; + /* Assertion level (0 = unasserted, 1 = asserted). */ + uint8_t level; +}; +typedef struct xen_hvm_set_isa_irq_level xen_hvm_set_isa_irq_level_t; +DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_isa_irq_level_t); + +#define HVMOP_set_pci_link_route 4 +struct xen_hvm_set_pci_link_route { + /* Domain to be updated. */ + domid_t domid; + /* PCI link identifier (0-3). */ + uint8_t link; + /* ISA IRQ (1-15), or 0 (disable link). */ + uint8_t isa_irq; +}; +typedef struct xen_hvm_set_pci_link_route xen_hvm_set_pci_link_route_t; +DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_pci_link_route_t); + +/* Flushes all VCPU TLBs: @arg must be NULL. */ +#define HVMOP_flush_tlbs 5 + +/* Following tools-only interfaces may change in future. */ +#if defined(__XEN__) || defined(__XEN_TOOLS__) + +/* Track dirty VRAM. */ +#define HVMOP_track_dirty_vram 6 +struct xen_hvm_track_dirty_vram { + /* Domain to be tracked. */ + domid_t domid; + /* First pfn to track. */ + uint64_aligned_t first_pfn; + /* Number of pages to track. */ + uint64_aligned_t nr; + /* OUT variable. */ + /* Dirty bitmap buffer. */ + XEN_GUEST_HANDLE_64(uint8) dirty_bitmap; +}; +typedef struct xen_hvm_track_dirty_vram xen_hvm_track_dirty_vram_t; +DEFINE_XEN_GUEST_HANDLE(xen_hvm_track_dirty_vram_t); + +/* Notify that some pages got modified by the Device Model. */ +#define HVMOP_modified_memory 7 +struct xen_hvm_modified_memory { + /* Domain to be updated. */ + domid_t domid; + /* First pfn. */ + uint64_aligned_t first_pfn; + /* Number of pages. */ + uint64_aligned_t nr; +}; +typedef struct xen_hvm_modified_memory xen_hvm_modified_memory_t; +DEFINE_XEN_GUEST_HANDLE(xen_hvm_modified_memory_t); + +#define HVMOP_set_mem_type 8 +typedef enum { + HVMMEM_ram_rw, /* Normal read/write guest RAM */ + HVMMEM_ram_ro, /* Read-only; writes are discarded */ + HVMMEM_mmio_dm, /* Reads and write go to the device model */ +} hvmmem_type_t; +/* Notify that a region of memory is to be treated in a specific way. */ +struct xen_hvm_set_mem_type { + /* Domain to be updated. */ + domid_t domid; + /* Memory type */ + hvmmem_type_t hvmmem_type; + /* First pfn. */ + uint64_aligned_t first_pfn; + /* Number of pages. */ + uint64_aligned_t nr; +}; +typedef struct xen_hvm_set_mem_type xen_hvm_set_mem_type_t; +DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_mem_type_t); + + +#endif /* defined(__XEN__) || defined(__XEN_TOOLS__) */ + +#endif /* __XEN_PUBLIC_HVM_HVM_OP_H__ */ diff --git a/include/xenpvdrivers/interface/hvm/ioreq.h b/include/xenpvdrivers/interface/hvm/ioreq.h new file mode 100644 index 0000000..0c10c08 --- /dev/null +++ b/include/xenpvdrivers/interface/hvm/ioreq.h @@ -0,0 +1,119 @@ +/* + * ioreq.h: I/O request definitions for device models + * Copyright (c) 2004, Intel Corporation. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef _IOREQ_H_ +#define _IOREQ_H_ + +#define IOREQ_READ 1 +#define IOREQ_WRITE 0 + +#define STATE_IOREQ_NONE 0 +#define STATE_IOREQ_READY 1 +#define STATE_IOREQ_INPROCESS 2 +#define STATE_IORESP_READY 3 + +#define IOREQ_TYPE_PIO 0 /* pio */ +#define IOREQ_TYPE_COPY 1 /* mmio ops */ +#define IOREQ_TYPE_TIMEOFFSET 7 +#define IOREQ_TYPE_INVALIDATE 8 /* mapcache */ + +/* + * VMExit dispatcher should cooperate with instruction decoder to + * prepare this structure and notify service OS and DM by sending + * virq + */ +struct ioreq { + uint64_t addr; /* physical address */ + uint64_t data; /* data (or paddr of data) */ + uint32_t count; /* for rep prefixes */ + uint32_t size; /* size in bytes */ + uint32_t vp_eport; /* evtchn for notifications to/from device model */ + uint16_t _pad0; + uint8_t state:4; + uint8_t data_is_ptr:1; /* if 1, data above is the guest paddr + * of the real data to use. */ + uint8_t dir:1; /* 1=read, 0=write */ + uint8_t df:1; + uint8_t _pad1:1; + uint8_t type; /* I/O type */ +}; +typedef struct ioreq ioreq_t; + +struct shared_iopage { + struct ioreq vcpu_ioreq[1]; +}; +typedef struct shared_iopage shared_iopage_t; + +struct buf_ioreq { + uint8_t type; /* I/O type */ + uint8_t pad:1; + uint8_t dir:1; /* 1=read, 0=write */ + uint8_t size:2; /* 0=>1, 1=>2, 2=>4, 3=>8. If 8, use two buf_ioreqs */ + uint32_t addr:20;/* physical address */ + uint32_t data; /* data */ +}; +typedef struct buf_ioreq buf_ioreq_t; + +#define IOREQ_BUFFER_SLOT_NUM 511 /* 8 bytes each, plus 2 4-byte indexes */ +struct buffered_iopage { + unsigned int read_pointer; + unsigned int write_pointer; + buf_ioreq_t buf_ioreq[IOREQ_BUFFER_SLOT_NUM]; +}; /* NB. Size of this structure must be no greater than one page. */ +typedef struct buffered_iopage buffered_iopage_t; + +#if defined(__ia64__) +struct pio_buffer { + uint32_t page_offset; + uint32_t pointer; + uint32_t data_end; + uint32_t buf_size; + void *opaque; +}; + +#define PIO_BUFFER_IDE_PRIMARY 0 /* I/O port = 0x1F0 */ +#define PIO_BUFFER_IDE_SECONDARY 1 /* I/O port = 0x170 */ +#define PIO_BUFFER_ENTRY_NUM 2 +struct buffered_piopage { + struct pio_buffer pio[PIO_BUFFER_ENTRY_NUM]; + uint8_t buffer[1]; +}; +#endif /* defined(__ia64__) */ + +#define ACPI_PM1A_EVT_BLK_ADDRESS 0x0000000000001f40 +#define ACPI_PM1A_CNT_BLK_ADDRESS (ACPI_PM1A_EVT_BLK_ADDRESS + 0x04) +#define ACPI_PM_TMR_BLK_ADDRESS (ACPI_PM1A_EVT_BLK_ADDRESS + 0x08) +#define ACPI_GPE0_BLK_ADDRESS (ACPI_PM_TMR_BLK_ADDRESS + 0x20) +#define ACPI_GPE0_BLK_LEN 0x08 + +#endif /* _IOREQ_H_ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/hvm/params.h b/include/xenpvdrivers/interface/hvm/params.h new file mode 100644 index 0000000..15d828f --- /dev/null +++ b/include/xenpvdrivers/interface/hvm/params.h @@ -0,0 +1,111 @@ +/* + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_HVM_PARAMS_H__ +#define __XEN_PUBLIC_HVM_PARAMS_H__ + +#include "hvm_op.h" + +/* + * Parameter space for HVMOP_{set,get}_param. + */ + +/* + * How should CPU0 event-channel notifications be delivered? + * val[63:56] == 0: val[55:0] is a delivery GSI (Global System Interrupt). + * val[63:56] == 1: val[55:0] is a delivery PCI INTx line, as follows: + * Domain = val[47:32], Bus = val[31:16], + * DevFn = val[15: 8], IntX = val[ 1: 0] + * If val == 0 then CPU0 event-channel notifications are not delivered. + */ +#define HVM_PARAM_CALLBACK_IRQ 0 + +/* + * These are not used by Xen. They are here for convenience of HVM-guest + * xenbus implementations. + */ +#define HVM_PARAM_STORE_PFN 1 +#define HVM_PARAM_STORE_EVTCHN 2 + +#define HVM_PARAM_PAE_ENABLED 4 + +#define HVM_PARAM_IOREQ_PFN 5 + +#define HVM_PARAM_BUFIOREQ_PFN 6 + +#ifdef __ia64__ + +#define HVM_PARAM_NVRAM_FD 7 +#define HVM_PARAM_VHPT_SIZE 8 +#define HVM_PARAM_BUFPIOREQ_PFN 9 + +#elif defined(__i386__) || defined(__x86_64__) + +/* Expose Viridian interfaces to this HVM guest? */ +#define HVM_PARAM_VIRIDIAN 9 + +#endif + +/* + * Set mode for virtual timers (currently x86 only): + * delay_for_missed_ticks (default): + * Do not advance a vcpu's time beyond the correct delivery time for + * interrupts that have been missed due to preemption. Deliver missed + * interrupts when the vcpu is rescheduled and advance the vcpu's virtual + * time stepwise for each one. + * no_delay_for_missed_ticks: + * As above, missed interrupts are delivered, but guest time always tracks + * wallclock (i.e., real) time while doing so. + * no_missed_ticks_pending: + * No missed interrupts are held pending. Instead, to ensure ticks are + * delivered at some non-zero rate, if we detect missed ticks then the + * internal tick alarm is not disabled if the VCPU is preempted during the + * next tick period. + * one_missed_tick_pending: + * Missed interrupts are collapsed together and delivered as one 'late tick'. + * Guest time always tracks wallclock (i.e., real) time. + */ +#define HVM_PARAM_TIMER_MODE 10 +#define HVMPTM_delay_for_missed_ticks 0 +#define HVMPTM_no_delay_for_missed_ticks 1 +#define HVMPTM_no_missed_ticks_pending 2 +#define HVMPTM_one_missed_tick_pending 3 + +/* Boolean: Enable virtual HPET (high-precision event timer)? (x86-only) */ +#define HVM_PARAM_HPET_ENABLED 11 + +/* Identity-map page directory used by Intel EPT when CR0.PG=0. */ +#define HVM_PARAM_IDENT_PT 12 + +/* Device Model domain, defaults to 0. */ +#define HVM_PARAM_DM_DOMAIN 13 + +/* ACPI S state: currently support S0 and S3 on x86. */ +#define HVM_PARAM_ACPI_S_STATE 14 + +/* TSS used on Intel when CR0.PE=0. */ +#define HVM_PARAM_VM86_TSS 15 + +/* Boolean: Enable aligning all periodic vpts to reduce interrupts */ +#define HVM_PARAM_VPT_ALIGN 16 + +#define HVM_NR_PARAMS 17 + +#endif /* __XEN_PUBLIC_HVM_PARAMS_H__ */ diff --git a/include/xenpvdrivers/interface/hvm/save.h b/include/xenpvdrivers/interface/hvm/save.h new file mode 100644 index 0000000..d45f0c1 --- /dev/null +++ b/include/xenpvdrivers/interface/hvm/save.h @@ -0,0 +1,88 @@ +/* + * hvm/save.h + * + * Structure definitions for HVM state that is held by Xen and must + * be saved along with the domain's memory and device-model state. + * + * Copyright (c) 2007 XenSource Ltd. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_HVM_SAVE_H__ +#define __XEN_PUBLIC_HVM_SAVE_H__ + +/* + * Structures in this header *must* have the same layout in 32bit + * and 64bit environments: this means that all fields must be explicitly + * sized types and aligned to their sizes, and the structs must be + * a multiple of eight bytes long. + * + * Only the state necessary for saving and restoring (i.e. fields + * that are analogous to actual hardware state) should go in this file. + * Internal mechanisms should be kept in Xen-private headers. + */ + +#if !defined(__GNUC__) || defined(__STRICT_ANSI__) +#error "Anonymous structs/unions are a GNU extension." +#endif + +/* + * Each entry is preceded by a descriptor giving its type and length + */ +struct hvm_save_descriptor { + uint16_t typecode; /* Used to demux the various types below */ + uint16_t instance; /* Further demux within a type */ + uint32_t length; /* In bytes, *not* including this descriptor */ +}; + + +/* + * Each entry has a datatype associated with it: for example, the CPU state + * is saved as a HVM_SAVE_TYPE(CPU), which has HVM_SAVE_LENGTH(CPU), + * and is identified by a descriptor with typecode HVM_SAVE_CODE(CPU). + * DECLARE_HVM_SAVE_TYPE binds these things together with some type-system + * ugliness. + */ + +#define DECLARE_HVM_SAVE_TYPE(_x, _code, _type) \ + struct __HVM_SAVE_TYPE_##_x { _type t; char c[_code]; } + +#define HVM_SAVE_TYPE(_x) typeof (((struct __HVM_SAVE_TYPE_##_x *)(0))->t) +#define HVM_SAVE_LENGTH(_x) (sizeof (HVM_SAVE_TYPE(_x))) +#define HVM_SAVE_CODE(_x) (sizeof (((struct __HVM_SAVE_TYPE_##_x *)(0))->c)) + + +/* + * The series of save records is teminated by a zero-type, zero-length + * descriptor. + */ + +struct hvm_save_end {}; +DECLARE_HVM_SAVE_TYPE(END, 0, struct hvm_save_end); + +#if defined(__i386__) || defined(__x86_64__) +#include "../arch-x86/hvm/save.h" +#elif defined(__ia64__) +#include "../arch-ia64/hvm/save.h" +#else +#error "unsupported architecture" +#endif + +#endif /* __XEN_PUBLIC_HVM_SAVE_H__ */ diff --git a/include/xenpvdrivers/interface/io/blkif.h b/include/xenpvdrivers/interface/io/blkif.h new file mode 100644 index 0000000..a06c00a --- /dev/null +++ b/include/xenpvdrivers/interface/io/blkif.h @@ -0,0 +1,142 @@ +/****************************************************************************** + * blkif.h + * + * Unified block-device I/O interface for Xen guest OSes. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2003-2004, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_IO_BLKIF_H__ +#define __XEN_PUBLIC_IO_BLKIF_H__ + +#include "ring.h" +#include "../grant_table.h" + +/* + * Front->back notifications: When enqueuing a new request, sending a + * notification can be made conditional on req_event (i.e., the generic + * hold-off mechanism provided by the ring macros). Backends must set + * req_event appropriately (e.g., using RING_FINAL_CHECK_FOR_REQUESTS()). + * + * Back->front notifications: When enqueuing a new response, sending a + * notification can be made conditional on rsp_event (i.e., the generic + * hold-off mechanism provided by the ring macros). Frontends must set + * rsp_event appropriately (e.g., using RING_FINAL_CHECK_FOR_RESPONSES()). + */ + +#ifndef blkif_vdev_t +#define blkif_vdev_t uint16_t +#endif +#define blkif_sector_t uint64_t + +/* + * REQUEST CODES. + */ +#define BLKIF_OP_READ 0 +#define BLKIF_OP_WRITE 1 +/* + * Recognised only if "feature-barrier" is present in backend xenbus info. + * The "feature-barrier" node contains a boolean indicating whether barrier + * requests are likely to succeed or fail. Either way, a barrier request + * may fail at any time with BLKIF_RSP_EOPNOTSUPP if it is unsupported by + * the underlying block-device hardware. The boolean simply indicates whether + * or not it is worthwhile for the frontend to attempt barrier requests. + * If a backend does not recognise BLKIF_OP_WRITE_BARRIER, it should *not* + * create the "feature-barrier" node! + */ +#define BLKIF_OP_WRITE_BARRIER 2 +/* + * Recognised if "feature-flush-cache" is present in backend xenbus + * info. A flush will ask the underlying storage hardware to flush its + * non-volatile caches as appropriate. The "feature-flush-cache" node + * contains a boolean indicating whether flush requests are likely to + * succeed or fail. Either way, a flush request may fail at any time + * with BLKIF_RSP_EOPNOTSUPP if it is unsupported by the underlying + * block-device hardware. The boolean simply indicates whether or not it + * is worthwhile for the frontend to attempt flushes. If a backend does + * not recognise BLKIF_OP_WRITE_FLUSH_CACHE, it should *not* create the + * "feature-flush-cache" node! + */ +#define BLKIF_OP_FLUSH_DISKCACHE 3 +/* + * Device specific command packet contained within the request + */ +#define BLKIF_OP_PACKET 4 + +/* + * Maximum scatter/gather segments per request. + * This is carefully chosen so that sizeof(blkif_ring_t) <= PAGE_SIZE. + * NB. This could be 12 if the ring indexes weren't stored in the same page. + */ +#define BLKIF_MAX_SEGMENTS_PER_REQUEST 11 + +/* + * NB. first_sect and last_sect in blkif_request_segment, as well as + * sector_number in blkif_request, are always expressed in 512-byte units. + * However they must be properly aligned to the real sector size of the + * physical disk, which is reported in the "sector-size" node in the backend + * xenbus info. Also the xenbus "sectors" node is expressed in 512-byte units. + */ +struct blkif_request_segment { + grant_ref_t gref; /* reference to I/O buffer frame */ + /* @first_sect: first sector in frame to transfer (inclusive). */ + /* @last_sect: last sector in frame to transfer (inclusive). */ + uint8_t first_sect, last_sect; +}; + +struct blkif_request { + uint8_t operation; /* BLKIF_OP_??? */ + uint8_t nr_segments; /* number of segments */ + blkif_vdev_t handle; /* only for read/write requests */ + uint64_t id; /* private guest value, echoed in resp */ + blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */ + struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST]; +}; +typedef struct blkif_request blkif_request_t; + +struct blkif_response { + uint64_t id; /* copied from request */ + uint8_t operation; /* copied from request */ + int16_t status; /* BLKIF_RSP_??? */ +}; +typedef struct blkif_response blkif_response_t; + +/* + * STATUS RETURN CODES. + */ + /* Operation not supported (only happens on barrier writes). */ +#define BLKIF_RSP_EOPNOTSUPP -2 + /* Operation failed for some unspecified reason (-EIO). */ +#define BLKIF_RSP_ERROR -1 + /* Operation completed successfully. */ +#define BLKIF_RSP_OKAY 0 + +/* + * Generate blkif ring structures and types. + */ + +DEFINE_RING_TYPES(blkif, struct blkif_request, struct blkif_response); + +#define VDISK_CDROM 0x1 +#define VDISK_REMOVABLE 0x2 +#define VDISK_READONLY 0x4 + +#endif /* __XEN_PUBLIC_IO_BLKIF_H__ */ diff --git a/include/xenpvdrivers/interface/io/cdromif.h b/include/xenpvdrivers/interface/io/cdromif.h new file mode 100644 index 0000000..b691056 --- /dev/null +++ b/include/xenpvdrivers/interface/io/cdromif.h @@ -0,0 +1,120 @@ +/****************************************************************************** + * cdromif.h + * + * Shared definitions between backend driver and Xen guest Virtual CDROM + * block device. + * + * Copyright (c) 2008, Pat Campell plc@novell.com + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_IO_CDROMIF_H__ +#define __XEN_PUBLIC_IO_CDROMIF_H__ + +/* + * Queries backend for CDROM support + */ +#define XEN_TYPE_CDROM_SUPPORT _IO('c', 1) + +struct xen_cdrom_support +{ + uint32_t type; + int8_t ret; /* returned, 0 succeded, -1 error */ + int8_t err; /* returned, backend errno */ + int8_t supported; /* returned, 1 supported */ +}; + +/* + * Opens backend device, returns drive geometry or + * any encountered errors + */ +#define XEN_TYPE_CDROM_OPEN _IO('c', 2) + +struct xen_cdrom_open +{ + uint32_t type; + int8_t ret; + int8_t err; + int8_t pad; + int8_t media_present; /* returned */ + uint32_t sectors; /* returned */ + uint32_t sector_size; /* returned */ + int32_t payload_offset; /* offset to backend node name payload */ +}; + +/* + * Queries backend for media changed status + */ +#define XEN_TYPE_CDROM_MEDIA_CHANGED _IO('c', 3) + +struct xen_cdrom_media_changed +{ + uint32_t type; + int8_t ret; + int8_t err; + int8_t media_changed; /* returned */ +}; + +/* + * Sends vcd generic CDROM packet to backend, followed + * immediately by the vcd_generic_command payload + */ +#define XEN_TYPE_CDROM_PACKET _IO('c', 4) + +struct xen_cdrom_packet +{ + uint32_t type; + int8_t ret; + int8_t err; + int8_t pad[2]; + int32_t payload_offset; /* offset to vcd_generic_command payload */ +}; + +/* CDROM_PACKET_COMMAND, payload for XEN_TYPE_CDROM_PACKET */ +struct vcd_generic_command +{ + uint8_t cmd[CDROM_PACKET_SIZE]; + uint8_t pad[4]; + uint32_t buffer_offset; + uint32_t buflen; + int32_t stat; + uint32_t sense_offset; + uint8_t data_direction; + uint8_t pad1[3]; + int32_t quiet; + int32_t timeout; +}; + +union xen_block_packet +{ + uint32_t type; + struct xen_cdrom_support xcs; + struct xen_cdrom_open xco; + struct xen_cdrom_media_changed xcmc; + struct xen_cdrom_packet xcp; +}; + +#define PACKET_PAYLOAD_OFFSET (sizeof(struct xen_cdrom_packet)) +#define PACKET_SENSE_OFFSET (PACKET_PAYLOAD_OFFSET + sizeof(struct vcd_generic_command)) +#define PACKET_BUFFER_OFFSET (PACKET_SENSE_OFFSET + sizeof(struct request_sense)) +#define MAX_PACKET_DATA (PAGE_SIZE - sizeof(struct xen_cdrom_packet) - \ + sizeof(struct vcd_generic_command) - sizeof(struct request_sense)) + +#endif diff --git a/include/xenpvdrivers/interface/io/console.h b/include/xenpvdrivers/interface/io/console.h new file mode 100644 index 0000000..70906df --- /dev/null +++ b/include/xenpvdrivers/interface/io/console.h @@ -0,0 +1,41 @@ +/****************************************************************************** + * console.h + * + * Console I/O interface for Xen guest OSes. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2005, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_IO_CONSOLE_H__ +#define __XEN_PUBLIC_IO_CONSOLE_H__ + +typedef uint32_t XENCONS_RING_IDX; + +#define MASK_XENCONS_IDX(idx, ring) ((idx) & (sizeof(ring)-1)) + +struct xencons_interface { + char in[1024]; + char out[2048]; + XENCONS_RING_IDX in_cons, in_prod; + XENCONS_RING_IDX out_cons, out_prod; +}; + +#endif /* __XEN_PUBLIC_IO_CONSOLE_H__ */ diff --git a/include/xenpvdrivers/interface/io/fbif.h b/include/xenpvdrivers/interface/io/fbif.h new file mode 100644 index 0000000..7071586 --- /dev/null +++ b/include/xenpvdrivers/interface/io/fbif.h @@ -0,0 +1,171 @@ +/* + * fbif.h -- Xen virtual frame buffer device + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (C) 2005 Anthony Liguori + * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster + */ + +#ifndef __XEN_PUBLIC_IO_FBIF_H__ +#define __XEN_PUBLIC_IO_FBIF_H__ + +/* Out events (frontend -> backend) */ + +/* + * Out events may be sent only when requested by backend, and receipt + * of an unknown out event is an error. + */ + +/* Event type 1 currently not used */ +/* + * Framebuffer update notification event + * Capable frontend sets feature-update in xenstore. + * Backend requests it by setting request-update in xenstore. + */ +#define XENFB_TYPE_UPDATE 2 + +struct xenfb_update +{ + uint8_t type; /* XENFB_TYPE_UPDATE */ + int32_t x; /* source x */ + int32_t y; /* source y */ + int32_t width; /* rect width */ + int32_t height; /* rect height */ +}; + +/* + * Framebuffer resize notification event + * Capable backend sets feature-resize in xenstore. + */ +#define XENFB_TYPE_RESIZE 3 + +struct xenfb_resize +{ + uint8_t type; /* XENFB_TYPE_RESIZE */ + int32_t width; /* width in pixels */ + int32_t height; /* height in pixels */ + int32_t stride; /* stride in bytes */ + int32_t depth; /* depth in bits */ + int32_t offset; /* offset of the framebuffer in bytes */ +}; + +#define XENFB_OUT_EVENT_SIZE 40 + +union xenfb_out_event +{ + uint8_t type; + struct xenfb_update update; + struct xenfb_resize resize; + char pad[XENFB_OUT_EVENT_SIZE]; +}; + +/* In events (backend -> frontend) */ + +/* + * Frontends should ignore unknown in events. + */ + +/* + * Framebuffer refresh period advice + * Backend sends it to advise the frontend their preferred period of + * refresh. Frontends that keep the framebuffer constantly up-to-date + * just ignore it. Frontends that use the advice should immediately + * refresh the framebuffer (and send an update notification event if + * those have been requested), then use the update frequency to guide + * their periodical refreshs. + */ +#define XENFB_TYPE_REFRESH_PERIOD 1 +#define XENFB_NO_REFRESH 0 + +struct xenfb_refresh_period +{ + uint8_t type; /* XENFB_TYPE_UPDATE_PERIOD */ + uint32_t period; /* period of refresh, in ms, + * XENFB_NO_REFRESH if no refresh is needed */ +}; + +#define XENFB_IN_EVENT_SIZE 40 + +union xenfb_in_event +{ + uint8_t type; + struct xenfb_refresh_period refresh_period; + char pad[XENFB_IN_EVENT_SIZE]; +}; + +/* shared page */ + +#define XENFB_IN_RING_SIZE 1024 +#define XENFB_IN_RING_LEN (XENFB_IN_RING_SIZE / XENFB_IN_EVENT_SIZE) +#define XENFB_IN_RING_OFFS 1024 +#define XENFB_IN_RING(page) \ + ((union xenfb_in_event *)((char *)(page) + XENFB_IN_RING_OFFS)) +#define XENFB_IN_RING_REF(page, idx) \ + (XENFB_IN_RING((page))[(idx) % XENFB_IN_RING_LEN]) + +#define XENFB_OUT_RING_SIZE 2048 +#define XENFB_OUT_RING_LEN (XENFB_OUT_RING_SIZE / XENFB_OUT_EVENT_SIZE) +#define XENFB_OUT_RING_OFFS (XENFB_IN_RING_OFFS + XENFB_IN_RING_SIZE) +#define XENFB_OUT_RING(page) \ + ((union xenfb_out_event *)((char *)(page) + XENFB_OUT_RING_OFFS)) +#define XENFB_OUT_RING_REF(page, idx) \ + (XENFB_OUT_RING((page))[(idx) % XENFB_OUT_RING_LEN]) + +struct xenfb_page +{ + uint32_t in_cons, in_prod; + uint32_t out_cons, out_prod; + + int32_t width; /* the width of the framebuffer (in pixels) */ + int32_t height; /* the height of the framebuffer (in pixels) */ + uint32_t line_length; /* the length of a row of pixels (in bytes) */ + uint32_t mem_length; /* the length of the framebuffer (in bytes) */ + uint8_t depth; /* the depth of a pixel (in bits) */ + + /* + * Framebuffer page directory + * + * Each directory page holds PAGE_SIZE / sizeof(*pd) + * framebuffer pages, and can thus map up to PAGE_SIZE * + * PAGE_SIZE / sizeof(*pd) bytes. With PAGE_SIZE == 4096 and + * sizeof(unsigned long) == 4/8, that's 4 Megs 32 bit and 2 Megs + * 64 bit. 256 directories give enough room for a 512 Meg + * framebuffer with a max resolution of 12,800x10,240. Should + * be enough for a while with room leftover for expansion. + */ +#ifndef CONFIG_PARAVIRT_XEN + unsigned long pd[256]; +#else + /* Two directory pages should be enough for a while. */ + unsigned long pd[2]; +#endif +}; + +/* + * Wart: xenkbd needs to know default resolution. Put it here until a + * better solution is found, but don't leak it to the backend. + */ +#ifdef __KERNEL__ +#define XENFB_WIDTH 800 +#define XENFB_HEIGHT 600 +#define XENFB_DEPTH 32 +#endif + +#endif diff --git a/include/xenpvdrivers/interface/io/fsif.h b/include/xenpvdrivers/interface/io/fsif.h new file mode 100644 index 0000000..8fc2174 --- /dev/null +++ b/include/xenpvdrivers/interface/io/fsif.h @@ -0,0 +1,192 @@ +/****************************************************************************** + * fsif.h + * + * Interface to FS level split device drivers. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2007, Grzegorz Milos, . + */ + +#ifndef __XEN_PUBLIC_IO_FSIF_H__ +#define __XEN_PUBLIC_IO_FSIF_H__ + +#include "ring.h" +#include "../grant_table.h" + +#define REQ_FILE_OPEN 1 +#define REQ_FILE_CLOSE 2 +#define REQ_FILE_READ 3 +#define REQ_FILE_WRITE 4 +#define REQ_STAT 5 +#define REQ_FILE_TRUNCATE 6 +#define REQ_REMOVE 7 +#define REQ_RENAME 8 +#define REQ_CREATE 9 +#define REQ_DIR_LIST 10 +#define REQ_CHMOD 11 +#define REQ_FS_SPACE 12 +#define REQ_FILE_SYNC 13 + +struct fsif_open_request { + grant_ref_t gref; +}; + +struct fsif_close_request { + uint32_t fd; +}; + +struct fsif_read_request { + uint32_t fd; + int32_t pad; + uint64_t len; + uint64_t offset; + grant_ref_t grefs[1]; /* Variable length */ +}; + +struct fsif_write_request { + uint32_t fd; + int32_t pad; + uint64_t len; + uint64_t offset; + grant_ref_t grefs[1]; /* Variable length */ +}; + +struct fsif_stat_request { + uint32_t fd; +}; + +/* This structure is a copy of some fields from stat structure, returned + * via the ring. */ +struct fsif_stat_response { + int32_t stat_mode; + uint32_t stat_uid; + uint32_t stat_gid; + int32_t stat_ret; + int64_t stat_size; + int64_t stat_atime; + int64_t stat_mtime; + int64_t stat_ctime; +}; + +struct fsif_truncate_request { + uint32_t fd; + int32_t pad; + int64_t length; +}; + +struct fsif_remove_request { + grant_ref_t gref; +}; + +struct fsif_rename_request { + uint16_t old_name_offset; + uint16_t new_name_offset; + grant_ref_t gref; +}; + +struct fsif_create_request { + int8_t directory; + int8_t pad; + int16_t pad2; + int32_t mode; + grant_ref_t gref; +}; + +struct fsif_list_request { + uint32_t offset; + grant_ref_t gref; +}; + +#define NR_FILES_SHIFT 0 +#define NR_FILES_SIZE 16 /* 16 bits for the number of files mask */ +#define NR_FILES_MASK (((1ULL << NR_FILES_SIZE) - 1) << NR_FILES_SHIFT) +#define ERROR_SIZE 32 /* 32 bits for the error mask */ +#define ERROR_SHIFT (NR_FILES_SIZE + NR_FILES_SHIFT) +#define ERROR_MASK (((1ULL << ERROR_SIZE) - 1) << ERROR_SHIFT) +#define HAS_MORE_SHIFT (ERROR_SHIFT + ERROR_SIZE) +#define HAS_MORE_FLAG (1ULL << HAS_MORE_SHIFT) + +struct fsif_chmod_request { + uint32_t fd; + int32_t mode; +}; + +struct fsif_space_request { + grant_ref_t gref; +}; + +struct fsif_sync_request { + uint32_t fd; +}; + + +/* FS operation request */ +struct fsif_request { + uint8_t type; /* Type of the request */ + uint8_t pad; + uint16_t id; /* Request ID, copied to the response */ + uint32_t pad2; + union { + struct fsif_open_request fopen; + struct fsif_close_request fclose; + struct fsif_read_request fread; + struct fsif_write_request fwrite; + struct fsif_stat_request fstat; + struct fsif_truncate_request ftruncate; + struct fsif_remove_request fremove; + struct fsif_rename_request frename; + struct fsif_create_request fcreate; + struct fsif_list_request flist; + struct fsif_chmod_request fchmod; + struct fsif_space_request fspace; + struct fsif_sync_request fsync; + } u; +}; +typedef struct fsif_request fsif_request_t; + +/* FS operation response */ +struct fsif_response { + uint16_t id; + uint16_t pad1; + uint32_t pad2; + union { + uint64_t ret_val; + struct fsif_stat_response fstat; + } u; +}; + +typedef struct fsif_response fsif_response_t; + +#define FSIF_RING_ENTRY_SIZE 64 + +#define FSIF_NR_READ_GNTS ((FSIF_RING_ENTRY_SIZE - sizeof(struct fsif_read_request)) / \ + sizeof(grant_ref_t) + 1) +#define FSIF_NR_WRITE_GNTS ((FSIF_RING_ENTRY_SIZE - sizeof(struct fsif_write_request)) / \ + sizeof(grant_ref_t) + 1) + +DEFINE_RING_TYPES(fsif, struct fsif_request, struct fsif_response); + +#define STATE_INITIALISED "init" +#define STATE_READY "ready" +#define STATE_CLOSING "closing" +#define STATE_CLOSED "closed" + + +#endif diff --git a/include/xenpvdrivers/interface/io/kbdif.h b/include/xenpvdrivers/interface/io/kbdif.h new file mode 100644 index 0000000..39bae65 --- /dev/null +++ b/include/xenpvdrivers/interface/io/kbdif.h @@ -0,0 +1,122 @@ +/* + * kbdif.h -- Xen virtual keyboard/mouse + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (C) 2005 Anthony Liguori + * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster + */ + +#ifndef __XEN_PUBLIC_IO_KBDIF_H__ +#define __XEN_PUBLIC_IO_KBDIF_H__ + +/* In events (backend -> frontend) */ + +/* + * Frontends should ignore unknown in events. + */ + +/* Pointer movement event */ +#define XENKBD_TYPE_MOTION 1 +/* Event type 2 currently not used */ +/* Key event (includes pointer buttons) */ +#define XENKBD_TYPE_KEY 3 +/* + * Pointer position event + * Capable backend sets feature-abs-pointer in xenstore. + * Frontend requests ot instead of XENKBD_TYPE_MOTION by setting + * request-abs-update in xenstore. + */ +#define XENKBD_TYPE_POS 4 + +struct xenkbd_motion +{ + uint8_t type; /* XENKBD_TYPE_MOTION */ + int32_t rel_x; /* relative X motion */ + int32_t rel_y; /* relative Y motion */ + int32_t rel_z; /* relative Z motion (wheel) */ +}; + +struct xenkbd_key +{ + uint8_t type; /* XENKBD_TYPE_KEY */ + uint8_t pressed; /* 1 if pressed; 0 otherwise */ + uint32_t keycode; /* KEY_* from linux/input.h */ +}; + +struct xenkbd_position +{ + uint8_t type; /* XENKBD_TYPE_POS */ + int32_t abs_x; /* absolute X position (in FB pixels) */ + int32_t abs_y; /* absolute Y position (in FB pixels) */ + int32_t rel_z; /* relative Z motion (wheel) */ +}; + +#define XENKBD_IN_EVENT_SIZE 40 + +union xenkbd_in_event +{ + uint8_t type; + struct xenkbd_motion motion; + struct xenkbd_key key; + struct xenkbd_position pos; + char pad[XENKBD_IN_EVENT_SIZE]; +}; + +/* Out events (frontend -> backend) */ + +/* + * Out events may be sent only when requested by backend, and receipt + * of an unknown out event is an error. + * No out events currently defined. + */ + +#define XENKBD_OUT_EVENT_SIZE 40 + +union xenkbd_out_event +{ + uint8_t type; + char pad[XENKBD_OUT_EVENT_SIZE]; +}; + +/* shared page */ + +#define XENKBD_IN_RING_SIZE 2048 +#define XENKBD_IN_RING_LEN (XENKBD_IN_RING_SIZE / XENKBD_IN_EVENT_SIZE) +#define XENKBD_IN_RING_OFFS 1024 +#define XENKBD_IN_RING(page) \ + ((union xenkbd_in_event *)((char *)(page) + XENKBD_IN_RING_OFFS)) +#define XENKBD_IN_RING_REF(page, idx) \ + (XENKBD_IN_RING((page))[(idx) % XENKBD_IN_RING_LEN]) + +#define XENKBD_OUT_RING_SIZE 1024 +#define XENKBD_OUT_RING_LEN (XENKBD_OUT_RING_SIZE / XENKBD_OUT_EVENT_SIZE) +#define XENKBD_OUT_RING_OFFS (XENKBD_IN_RING_OFFS + XENKBD_IN_RING_SIZE) +#define XENKBD_OUT_RING(page) \ + ((union xenkbd_out_event *)((char *)(page) + XENKBD_OUT_RING_OFFS)) +#define XENKBD_OUT_RING_REF(page, idx) \ + (XENKBD_OUT_RING((page))[(idx) % XENKBD_OUT_RING_LEN]) + +struct xenkbd_page +{ + uint32_t in_cons, in_prod; + uint32_t out_cons, out_prod; +}; + +#endif diff --git a/include/xenpvdrivers/interface/io/netif.h b/include/xenpvdrivers/interface/io/netif.h new file mode 100644 index 0000000..58a4be0 --- /dev/null +++ b/include/xenpvdrivers/interface/io/netif.h @@ -0,0 +1,209 @@ +/****************************************************************************** + * netif.h + * + * Unified network-device I/O interface for Xen guest OSes. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2003-2004, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_IO_NETIF_H__ +#define __XEN_PUBLIC_IO_NETIF_H__ + +#include "ring.h" +#include "../grant_table.h" + +/* + * Notifications after enqueuing any type of message should be conditional on + * the appropriate req_event or rsp_event field in the shared ring. + * If the client sends notification for rx requests then it should specify + * feature 'feature-rx-notify' via xenbus. Otherwise the backend will assume + * that it cannot safely queue packets (as it may not be kicked to send them). + */ + +/* + * This is the 'wire' format for packets: + * Request 1: netif_tx_request -- NETTXF_* (any flags) + * [Request 2: netif_tx_extra] (only if request 1 has NETTXF_extra_info) + * [Request 3: netif_tx_extra] (only if request 2 has XEN_NETIF_EXTRA_MORE) + * Request 4: netif_tx_request -- NETTXF_more_data + * Request 5: netif_tx_request -- NETTXF_more_data + * ... + * Request N: netif_tx_request -- 0 + */ + +/* Protocol checksum field is blank in the packet (hardware offload)? */ +#define _NETTXF_csum_blank (0) +#define NETTXF_csum_blank (1U<<_NETTXF_csum_blank) + +/* Packet data has been validated against protocol checksum. */ +#define _NETTXF_data_validated (1) +#define NETTXF_data_validated (1U<<_NETTXF_data_validated) + +/* Packet continues in the next request descriptor. */ +#define _NETTXF_more_data (2) +#define NETTXF_more_data (1U<<_NETTXF_more_data) + +/* Packet to be followed by extra descriptor(s). */ +#define _NETTXF_extra_info (3) +#define NETTXF_extra_info (1U<<_NETTXF_extra_info) + +struct netif_tx_request { + grant_ref_t gref; /* Reference to buffer page */ + uint16_t offset; /* Offset within buffer page */ + uint16_t flags; /* NETTXF_* */ + uint16_t id; /* Echoed in response message. */ + uint16_t size; /* Packet size in bytes. */ +}; +typedef struct netif_tx_request netif_tx_request_t; + +/* Types of netif_extra_info descriptors. */ +#define XEN_NETIF_EXTRA_TYPE_NONE (0) /* Never used - invalid */ +#define XEN_NETIF_EXTRA_TYPE_GSO (1) /* u.gso */ +#define XEN_NETIF_EXTRA_TYPE_MCAST_ADD (2) /* u.mcast */ +#define XEN_NETIF_EXTRA_TYPE_MCAST_DEL (3) /* u.mcast */ +#define XEN_NETIF_EXTRA_TYPE_MAX (4) + +/* netif_extra_info flags. */ +#define _XEN_NETIF_EXTRA_FLAG_MORE (0) +#define XEN_NETIF_EXTRA_FLAG_MORE (1U<<_XEN_NETIF_EXTRA_FLAG_MORE) + +/* GSO types - only TCPv4 currently supported. */ +#define XEN_NETIF_GSO_TYPE_TCPV4 (1) + +/* + * This structure needs to fit within both netif_tx_request and + * netif_rx_response for compatibility. + */ +struct netif_extra_info { + uint8_t type; /* XEN_NETIF_EXTRA_TYPE_* */ + uint8_t flags; /* XEN_NETIF_EXTRA_FLAG_* */ + + union { + /* + * XEN_NETIF_EXTRA_TYPE_GSO: + */ + struct { + /* + * Maximum payload size of each segment. For example, for TCP this + * is just the path MSS. + */ + uint16_t size; + + /* + * GSO type. This determines the protocol of the packet and any + * extra features required to segment the packet properly. + */ + uint8_t type; /* XEN_NETIF_GSO_TYPE_* */ + + /* Future expansion. */ + uint8_t pad; + + /* + * GSO features. This specifies any extra GSO features required + * to process this packet, such as ECN support for TCPv4. + */ + uint16_t features; /* XEN_NETIF_GSO_FEAT_* */ + } gso; + + /* + * XEN_NETIF_EXTRA_TYPE_MCAST_{ADD,DEL}: + * Backend advertises availability via 'feature-multicast-control' + * xenbus node containing value '1'. + * Frontend requests this feature by advertising + * 'request-multicast-control' xenbus node containing value '1'. + * If multicast control is requested then multicast flooding is + * disabled and the frontend must explicitly register its interest + * in multicast groups using dummy transmit requests containing + * MCAST_{ADD,DEL} extra-info fragments. + */ + struct { + uint8_t addr[6]; /* Address to add/remove. */ + } mcast; + + uint16_t pad[3]; + } u; +}; +typedef struct netif_extra_info netif_extra_info_t; + +struct netif_tx_response { + uint16_t id; + int16_t status; /* NETIF_RSP_* */ +}; +typedef struct netif_tx_response netif_tx_response_t; + +struct netif_rx_request { + uint16_t id; /* Echoed in response message. */ + grant_ref_t gref; /* Reference to incoming granted frame */ +}; +typedef struct netif_rx_request netif_rx_request_t; + +/* Packet data has been validated against protocol checksum. */ +#define _NETRXF_data_validated (0) +#define NETRXF_data_validated (1U<<_NETRXF_data_validated) + +/* Protocol checksum field is blank in the packet (hardware offload)? */ +#define _NETRXF_csum_blank (1) +#define NETRXF_csum_blank (1U<<_NETRXF_csum_blank) + +/* Packet continues in the next request descriptor. */ +#define _NETRXF_more_data (2) +#define NETRXF_more_data (1U<<_NETRXF_more_data) + +/* Packet to be followed by extra descriptor(s). */ +#define _NETRXF_extra_info (3) +#define NETRXF_extra_info (1U<<_NETRXF_extra_info) + +struct netif_rx_response { + uint16_t id; + uint16_t offset; /* Offset in page of start of received packet */ + uint16_t flags; /* NETRXF_* */ + int16_t status; /* -ve: BLKIF_RSP_* ; +ve: Rx'ed pkt size. */ +}; +typedef struct netif_rx_response netif_rx_response_t; + +/* + * Generate netif ring structures and types. + */ + +#if defined(CONFIG_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H) +DEFINE_RING_TYPES(netif_tx, struct netif_tx_request, struct netif_tx_response); +DEFINE_RING_TYPES(netif_rx, struct netif_rx_request, struct netif_rx_response); +#else +#define xen_netif_tx_request netif_tx_request +#define xen_netif_rx_request netif_rx_request +#define xen_netif_tx_response netif_tx_response +#define xen_netif_rx_response netif_rx_response +DEFINE_RING_TYPES(xen_netif_tx, + struct xen_netif_tx_request, + struct xen_netif_tx_response); +DEFINE_RING_TYPES(xen_netif_rx, + struct xen_netif_rx_request, + struct xen_netif_rx_response); +#define xen_netif_extra_info netif_extra_info +#endif + +#define NETIF_RSP_DROPPED -2 +#define NETIF_RSP_ERROR -1 +#define NETIF_RSP_OKAY 0 +/* No response: used for auxiliary requests (e.g., netif_tx_extra). */ +#define NETIF_RSP_NULL 1 + +#endif diff --git a/include/xenpvdrivers/interface/io/pciif.h b/include/xenpvdrivers/interface/io/pciif.h new file mode 100644 index 0000000..c4177f3 --- /dev/null +++ b/include/xenpvdrivers/interface/io/pciif.h @@ -0,0 +1,124 @@ +/* + * PCI Backend/Frontend Common Data Structures & Macros + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Author: Ryan Wilson + */ +#ifndef __XEN_PCI_COMMON_H__ +#define __XEN_PCI_COMMON_H__ + +/* Be sure to bump this number if you change this file */ +#define XEN_PCI_MAGIC "7" + +/* xen_pci_sharedinfo flags */ +#define _XEN_PCIF_active (0) +#define XEN_PCIF_active (1<<_XEN_PCIF_active) +#define _XEN_PCIB_AERHANDLER (1) +#define XEN_PCIB_AERHANDLER (1<<_XEN_PCIB_AERHANDLER) +#define _XEN_PCIB_active (2) +#define XEN_PCIB_active (1<<_XEN_PCIB_active) + +/* xen_pci_op commands */ +#define XEN_PCI_OP_conf_read (0) +#define XEN_PCI_OP_conf_write (1) +#define XEN_PCI_OP_enable_msi (2) +#define XEN_PCI_OP_disable_msi (3) +#define XEN_PCI_OP_enable_msix (4) +#define XEN_PCI_OP_disable_msix (5) +#define XEN_PCI_OP_aer_detected (6) +#define XEN_PCI_OP_aer_resume (7) +#define XEN_PCI_OP_aer_mmio (8) +#define XEN_PCI_OP_aer_slotreset (9) + +/* xen_pci_op error numbers */ +#define XEN_PCI_ERR_success (0) +#define XEN_PCI_ERR_dev_not_found (-1) +#define XEN_PCI_ERR_invalid_offset (-2) +#define XEN_PCI_ERR_access_denied (-3) +#define XEN_PCI_ERR_not_implemented (-4) +/* XEN_PCI_ERR_op_failed - backend failed to complete the operation */ +#define XEN_PCI_ERR_op_failed (-5) + +/* + * it should be PAGE_SIZE-sizeof(struct xen_pci_op))/sizeof(struct msix_entry)) + * Should not exceed 128 + */ +#define SH_INFO_MAX_VEC 128 + +struct xen_msix_entry { + uint16_t vector; + uint16_t entry; +}; +struct xen_pci_op { + /* IN: what action to perform: XEN_PCI_OP_* */ + uint32_t cmd; + + /* OUT: will contain an error number (if any) from errno.h */ + int32_t err; + + /* IN: which device to touch */ + uint32_t domain; /* PCI Domain/Segment */ + uint32_t bus; + uint32_t devfn; + + /* IN: which configuration registers to touch */ + int32_t offset; + int32_t size; + + /* IN/OUT: Contains the result after a READ or the value to WRITE */ + uint32_t value; + /* IN: Contains extra infor for this operation */ + uint32_t info; + /*IN: param for msi-x */ + struct xen_msix_entry msix_entries[SH_INFO_MAX_VEC]; +}; + +/*used for pcie aer handling*/ +struct xen_pcie_aer_op +{ + + /* IN: what action to perform: XEN_PCI_OP_* */ + uint32_t cmd; + /*IN/OUT: return aer_op result or carry error_detected state as input*/ + int32_t err; + + /* IN: which device to touch */ + uint32_t domain; /* PCI Domain/Segment*/ + uint32_t bus; + uint32_t devfn; +}; +struct xen_pci_sharedinfo { + /* flags - XEN_PCIF_* */ + uint32_t flags; + struct xen_pci_op op; + struct xen_pcie_aer_op aer_op; +}; + +#endif /* __XEN_PCI_COMMON_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/io/protocols.h b/include/xenpvdrivers/interface/io/protocols.h new file mode 100644 index 0000000..5fbf10c --- /dev/null +++ b/include/xenpvdrivers/interface/io/protocols.h @@ -0,0 +1,40 @@ +/****************************************************************************** + * protocols.h + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PROTOCOLS_H__ +#define __XEN_PROTOCOLS_H__ + +#define XEN_IO_PROTO_ABI_X86_32 "x86_32-abi" +#define XEN_IO_PROTO_ABI_X86_64 "x86_64-abi" +#define XEN_IO_PROTO_ABI_IA64 "ia64-abi" + +#if defined(__i386__) +# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_32 +#elif defined(__x86_64__) +# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_64 +#elif defined(__ia64__) +# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_IA64 +#else +# error arch fixup needed here +#endif + +#endif diff --git a/include/xenpvdrivers/interface/io/ring.h b/include/xenpvdrivers/interface/io/ring.h new file mode 100644 index 0000000..9403f4b --- /dev/null +++ b/include/xenpvdrivers/interface/io/ring.h @@ -0,0 +1,304 @@ +/****************************************************************************** + * ring.h + * + * Shared producer-consumer ring macros. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Tim Deegan and Andrew Warfield November 2004. + */ + +#ifndef __XEN_PUBLIC_IO_RING_H__ +#define __XEN_PUBLIC_IO_RING_H__ + +#include "../xen-compat.h" + +#if __XEN_INTERFACE_VERSION__ < 0x00030208 +#define xen_mb() mb() +#define xen_rmb() rmb() +#define xen_wmb() wmb() +#endif + +typedef unsigned int RING_IDX; + +/* Round a 32-bit unsigned constant down to the nearest power of two. */ +#define __RD2(_x) (((_x) & 0x00000002) ? 0x2 : ((_x) & 0x1)) +#define __RD4(_x) (((_x) & 0x0000000c) ? __RD2((_x)>>2)<<2 : __RD2(_x)) +#define __RD8(_x) (((_x) & 0x000000f0) ? __RD4((_x)>>4)<<4 : __RD4(_x)) +#define __RD16(_x) (((_x) & 0x0000ff00) ? __RD8((_x)>>8)<<8 : __RD8(_x)) +#define __RD32(_x) (((_x) & 0xffff0000) ? __RD16((_x)>>16)<<16 : __RD16(_x)) + +/* + * Calculate size of a shared ring, given the total available space for the + * ring and indexes (_sz), and the name tag of the request/response structure. + * A ring contains as many entries as will fit, rounded down to the nearest + * power of two (so we can mask with (size-1) to loop around). + */ +#define __CONST_RING_SIZE(_s, _sz) \ + (__RD32(((_sz) - offsetof(struct _s##_sring, ring)) / \ + sizeof(((struct _s##_sring *)0)->ring[0]))) +/* + * The same for passing in an actual pointer instead of a name tag. + */ +#define __RING_SIZE(_s, _sz) \ + (__RD32(((_sz) - (long)(_s)->ring + (long)(_s)) / sizeof((_s)->ring[0]))) + +/* + * Macros to make the correct C datatypes for a new kind of ring. + * + * To make a new ring datatype, you need to have two message structures, + * let's say request_t, and response_t already defined. + * + * In a header where you want the ring datatype declared, you then do: + * + * DEFINE_RING_TYPES(mytag, request_t, response_t); + * + * These expand out to give you a set of types, as you can see below. + * The most important of these are: + * + * mytag_sring_t - The shared ring. + * mytag_front_ring_t - The 'front' half of the ring. + * mytag_back_ring_t - The 'back' half of the ring. + * + * To initialize a ring in your code you need to know the location and size + * of the shared memory area (PAGE_SIZE, for instance). To initialise + * the front half: + * + * mytag_front_ring_t front_ring; + * SHARED_RING_INIT((mytag_sring_t *)shared_page); + * FRONT_RING_INIT(&front_ring, (mytag_sring_t *)shared_page, PAGE_SIZE); + * + * Initializing the back follows similarly (note that only the front + * initializes the shared ring): + * + * mytag_back_ring_t back_ring; + * BACK_RING_INIT(&back_ring, (mytag_sring_t *)shared_page, PAGE_SIZE); + */ + +#define DEFINE_RING_TYPES(__name, __req_t, __rsp_t) \ + \ +/* Shared ring entry */ \ +union __name##_sring_entry { \ + __req_t req; \ + __rsp_t rsp; \ +}; \ + \ +/* Shared ring page */ \ +struct __name##_sring { \ + RING_IDX req_prod, req_event; \ + RING_IDX rsp_prod, rsp_event; \ + uint8_t netfront_smartpoll_active; \ + uint8_t pad[47]; \ + union __name##_sring_entry ring[1]; /* variable-length */ \ +}; \ + \ +/* "Front" end's private variables */ \ +struct __name##_front_ring { \ + RING_IDX req_prod_pvt; \ + RING_IDX rsp_cons; \ + unsigned int nr_ents; \ + struct __name##_sring *sring; \ +}; \ + \ +/* "Back" end's private variables */ \ +struct __name##_back_ring { \ + RING_IDX rsp_prod_pvt; \ + RING_IDX req_cons; \ + unsigned int nr_ents; \ + struct __name##_sring *sring; \ +}; \ + \ +/* Syntactic sugar */ \ +typedef struct __name##_sring __name##_sring_t; \ +typedef struct __name##_front_ring __name##_front_ring_t; \ +typedef struct __name##_back_ring __name##_back_ring_t + +/* + * Macros for manipulating rings. + * + * FRONT_RING_whatever works on the "front end" of a ring: here + * requests are pushed on to the ring and responses taken off it. + * + * BACK_RING_whatever works on the "back end" of a ring: here + * requests are taken off the ring and responses put on. + * + * N.B. these macros do NO INTERLOCKS OR FLOW CONTROL. + * This is OK in 1-for-1 request-response situations where the + * requestor (front end) never has more than RING_SIZE()-1 + * outstanding requests. + */ + +/* Initialising empty rings */ +#define SHARED_RING_INIT(_s) do { \ + (_s)->req_prod = (_s)->rsp_prod = 0; \ + (_s)->req_event = (_s)->rsp_event = 1; \ + (void)memset((_s)->pad, 0, sizeof((_s)->pad)); \ +} while(0) + +#define FRONT_RING_INIT(_r, _s, __size) do { \ + (_r)->req_prod_pvt = 0; \ + (_r)->rsp_cons = 0; \ + (_r)->nr_ents = __RING_SIZE(_s, __size); \ + (_r)->sring = (_s); \ +} while (0) + +#define BACK_RING_INIT(_r, _s, __size) do { \ + (_r)->rsp_prod_pvt = 0; \ + (_r)->req_cons = 0; \ + (_r)->nr_ents = __RING_SIZE(_s, __size); \ + (_r)->sring = (_s); \ +} while (0) + +/* Initialize to existing shared indexes -- for recovery */ +#define FRONT_RING_ATTACH(_r, _s, __size) do { \ + (_r)->sring = (_s); \ + (_r)->req_prod_pvt = (_s)->req_prod; \ + (_r)->rsp_cons = (_s)->rsp_prod; \ + (_r)->nr_ents = __RING_SIZE(_s, __size); \ +} while (0) + +#define BACK_RING_ATTACH(_r, _s, __size) do { \ + (_r)->sring = (_s); \ + (_r)->rsp_prod_pvt = (_s)->rsp_prod; \ + (_r)->req_cons = (_s)->req_prod; \ + (_r)->nr_ents = __RING_SIZE(_s, __size); \ +} while (0) + +/* How big is this ring? */ +#define RING_SIZE(_r) \ + ((_r)->nr_ents) + +/* Number of free requests (for use on front side only). */ +#define RING_FREE_REQUESTS(_r) \ + (RING_SIZE(_r) - ((_r)->req_prod_pvt - (_r)->rsp_cons)) + +/* Test if there is an empty slot available on the front ring. + * (This is only meaningful from the front. ) + */ +#define RING_FULL(_r) \ + (RING_FREE_REQUESTS(_r) == 0) + +/* Test if there are outstanding messages to be processed on a ring. */ +#define RING_HAS_UNCONSUMED_RESPONSES(_r) \ + ((_r)->sring->rsp_prod - (_r)->rsp_cons) + +#ifdef __GNUC__ +#define RING_HAS_UNCONSUMED_REQUESTS(_r) ({ \ + unsigned int req = (_r)->sring->req_prod - (_r)->req_cons; \ + unsigned int rsp = RING_SIZE(_r) - \ + ((_r)->req_cons - (_r)->rsp_prod_pvt); \ + req < rsp ? req : rsp; \ +}) +#else +/* Same as above, but without the nice GCC ({ ... }) syntax. */ +#define RING_HAS_UNCONSUMED_REQUESTS(_r) \ + ((((_r)->sring->req_prod - (_r)->req_cons) < \ + (RING_SIZE(_r) - ((_r)->req_cons - (_r)->rsp_prod_pvt))) ? \ + ((_r)->sring->req_prod - (_r)->req_cons) : \ + (RING_SIZE(_r) - ((_r)->req_cons - (_r)->rsp_prod_pvt))) +#endif + +/* Direct access to individual ring elements, by index. */ +#define RING_GET_REQUEST(_r, _idx) \ + (&((_r)->sring->ring[((_idx) & (RING_SIZE(_r) - 1))].req)) + +#define RING_GET_RESPONSE(_r, _idx) \ + (&((_r)->sring->ring[((_idx) & (RING_SIZE(_r) - 1))].rsp)) + +/* Loop termination condition: Would the specified index overflow the ring? */ +#define RING_REQUEST_CONS_OVERFLOW(_r, _cons) \ + (((_cons) - (_r)->rsp_prod_pvt) >= RING_SIZE(_r)) + +#define RING_PUSH_REQUESTS(_r) do { \ + xen_wmb(); /* back sees requests /before/ updated producer index */ \ + (_r)->sring->req_prod = (_r)->req_prod_pvt; \ +} while (0) + +#define RING_PUSH_RESPONSES(_r) do { \ + xen_wmb(); /* front sees resps /before/ updated producer index */ \ + (_r)->sring->rsp_prod = (_r)->rsp_prod_pvt; \ +} while (0) + +/* + * Notification hold-off (req_event and rsp_event): + * + * When queueing requests or responses on a shared ring, it may not always be + * necessary to notify the remote end. For example, if requests are in flight + * in a backend, the front may be able to queue further requests without + * notifying the back (if the back checks for new requests when it queues + * responses). + * + * When enqueuing requests or responses: + * + * Use RING_PUSH_{REQUESTS,RESPONSES}_AND_CHECK_NOTIFY(). The second argument + * is a boolean return value. True indicates that the receiver requires an + * asynchronous notification. + * + * After dequeuing requests or responses (before sleeping the connection): + * + * Use RING_FINAL_CHECK_FOR_REQUESTS() or RING_FINAL_CHECK_FOR_RESPONSES(). + * The second argument is a boolean return value. True indicates that there + * are pending messages on the ring (i.e., the connection should not be put + * to sleep). + * + * These macros will set the req_event/rsp_event field to trigger a + * notification on the very next message that is enqueued. If you want to + * create batches of work (i.e., only receive a notification after several + * messages have been enqueued) then you will need to create a customised + * version of the FINAL_CHECK macro in your own code, which sets the event + * field appropriately. + */ + +#define RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(_r, _notify) do { \ + RING_IDX __old = (_r)->sring->req_prod; \ + RING_IDX __new = (_r)->req_prod_pvt; \ + xen_wmb(); /* back sees requests /before/ updated producer index */ \ + (_r)->sring->req_prod = __new; \ + xen_mb(); /* back sees new requests /before/ we check req_event */ \ + (_notify) = ((RING_IDX)(__new - (_r)->sring->req_event) < \ + (RING_IDX)(__new - __old)); \ +} while (0) + +#define RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(_r, _notify) do { \ + RING_IDX __old = (_r)->sring->rsp_prod; \ + RING_IDX __new = (_r)->rsp_prod_pvt; \ + xen_wmb(); /* front sees resps /before/ updated producer index */ \ + (_r)->sring->rsp_prod = __new; \ + xen_mb(); /* front sees new resps /before/ we check rsp_event */ \ + (_notify) = ((RING_IDX)(__new - (_r)->sring->rsp_event) < \ + (RING_IDX)(__new - __old)); \ +} while (0) + +#define RING_FINAL_CHECK_FOR_REQUESTS(_r, _work_to_do) do { \ + (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \ + if (_work_to_do) break; \ + (_r)->sring->req_event = (_r)->req_cons + 1; \ + xen_mb(); \ + (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \ +} while (0) + +#define RING_FINAL_CHECK_FOR_RESPONSES(_r, _work_to_do) do { \ + (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \ + if (_work_to_do) break; \ + (_r)->sring->rsp_event = (_r)->rsp_cons + 1; \ + xen_mb(); \ + (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \ +} while (0) + +#endif /* __XEN_PUBLIC_IO_RING_H__ */ diff --git a/include/xenpvdrivers/interface/io/tpmif.h b/include/xenpvdrivers/interface/io/tpmif.h new file mode 100644 index 0000000..02ccdab --- /dev/null +++ b/include/xenpvdrivers/interface/io/tpmif.h @@ -0,0 +1,77 @@ +/****************************************************************************** + * tpmif.h + * + * TPM I/O interface for Xen guest OSes. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2005, IBM Corporation + * + * Author: Stefan Berger, stefanb@us.ibm.com + * Grant table support: Mahadevan Gomathisankaran + * + * This code has been derived from tools/libxc/xen/io/netif.h + * + * Copyright (c) 2003-2004, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_IO_TPMIF_H__ +#define __XEN_PUBLIC_IO_TPMIF_H__ + +#include "../grant_table.h" + +struct tpmif_tx_request { + unsigned long addr; /* Machine address of packet. */ + grant_ref_t ref; /* grant table access reference */ + uint16_t unused; + uint16_t size; /* Packet size in bytes. */ +}; +typedef struct tpmif_tx_request tpmif_tx_request_t; + +/* + * The TPMIF_TX_RING_SIZE defines the number of pages the + * front-end and backend can exchange (= size of array). + */ +typedef uint32_t TPMIF_RING_IDX; + +#define TPMIF_TX_RING_SIZE 1 + +/* This structure must fit in a memory page. */ + +struct tpmif_ring { + struct tpmif_tx_request req; +}; +typedef struct tpmif_ring tpmif_ring_t; + +struct tpmif_tx_interface { + struct tpmif_ring ring[TPMIF_TX_RING_SIZE]; +}; +typedef struct tpmif_tx_interface tpmif_tx_interface_t; + +#endif + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/io/usbif.h b/include/xenpvdrivers/interface/io/usbif.h new file mode 100644 index 0000000..6099c29 --- /dev/null +++ b/include/xenpvdrivers/interface/io/usbif.h @@ -0,0 +1,151 @@ +/* + * usbif.h + * + * USB I/O interface for Xen guest OSes. + * + * Copyright (C) 2009, FUJITSU LABORATORIES LTD. + * Author: Noboru Iwamatsu + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_IO_USBIF_H__ +#define __XEN_PUBLIC_IO_USBIF_H__ + +#include "ring.h" +#include "../grant_table.h" + +enum usb_spec_version { + USB_VER_UNKNOWN = 0, + USB_VER_USB11, + USB_VER_USB20, + USB_VER_USB30, /* not supported yet */ +}; + +/* + * USB pipe in usbif_request + * + * bits 0-5 are specific bits for virtual USB driver. + * bits 7-31 are standard urb pipe. + * + * - port number(NEW): bits 0-4 + * (USB_MAXCHILDREN is 31) + * + * - operation flag(NEW): bit 5 + * (0 = submit urb, + * 1 = unlink urb) + * + * - direction: bit 7 + * (0 = Host-to-Device [Out] + * 1 = Device-to-Host [In]) + * + * - device address: bits 8-14 + * + * - endpoint: bits 15-18 + * + * - pipe type: bits 30-31 + * (00 = isochronous, 01 = interrupt, + * 10 = control, 11 = bulk) + */ +#define usbif_pipeportnum(pipe) ((pipe) & 0x1f) +#define usbif_setportnum_pipe(pipe, portnum) \ + ((pipe)|(portnum)) + +#define usbif_pipeunlink(pipe) ((pipe) & 0x20) +#define usbif_pipesubmit(pipe) (!usbif_pipeunlink(pipe)) +#define usbif_setunlink_pipe(pipe) ((pipe)|(0x20)) + +#define USBIF_BACK_MAX_PENDING_REQS (128) +#define USBIF_MAX_SEGMENTS_PER_REQUEST (16) + +/* + * RING for transferring urbs. + */ +struct usbif_request_segment { + grant_ref_t gref; + uint16_t offset; + uint16_t length; +}; + +struct usbif_urb_request { + uint16_t id; /* request id */ + uint16_t nr_buffer_segs; /* number of urb->transfer_buffer segments */ + + /* basic urb parameter */ + uint32_t pipe; + uint16_t transfer_flags; + uint16_t buffer_length; + union { + uint8_t ctrl[8]; /* setup_packet (Ctrl) */ + + struct { + uint16_t interval; /* maximum (1024*8) in usb core */ + uint16_t start_frame; /* start frame */ + uint16_t number_of_packets; /* number of ISO packet */ + uint16_t nr_frame_desc_segs; /* number of iso_frame_desc segments */ + } isoc; + + struct { + uint16_t interval; /* maximum (1024*8) in usb core */ + uint16_t pad[3]; + } intr; + + struct { + uint16_t unlink_id; /* unlink request id */ + uint16_t pad[3]; + } unlink; + + } u; + + /* urb data segments */ + struct usbif_request_segment seg[USBIF_MAX_SEGMENTS_PER_REQUEST]; +}; +typedef struct usbif_urb_request usbif_urb_request_t; + +struct usbif_urb_response { + uint16_t id; /* request id */ + uint16_t start_frame; /* start frame (ISO) */ + int32_t status; /* status (non-ISO) */ + int32_t actual_length; /* actual transfer length */ + int32_t error_count; /* number of ISO errors */ +}; +typedef struct usbif_urb_response usbif_urb_response_t; + +DEFINE_RING_TYPES(usbif_urb, struct usbif_urb_request, struct usbif_urb_response); +#define USB_URB_RING_SIZE __CONST_RING_SIZE(usbif_urb, PAGE_SIZE) + +/* + * RING for notifying connect/disconnect events to frontend + */ +struct usbif_conn_request { + uint16_t id; +}; +typedef struct usbif_conn_request usbif_conn_request_t; + +struct usbif_conn_response { + uint16_t id; /* request id */ + uint8_t portnum; /* port number */ + uint8_t speed; /* usb_device_speed */ +}; +typedef struct usbif_conn_response usbif_conn_response_t; + +DEFINE_RING_TYPES(usbif_conn, struct usbif_conn_request, struct usbif_conn_response); +#define USB_CONN_RING_SIZE __CONST_RING_SIZE(usbif_conn, PAGE_SIZE) + +#endif /* __XEN_PUBLIC_IO_USBIF_H__ */ diff --git a/include/xenpvdrivers/interface/io/vscsiif.h b/include/xenpvdrivers/interface/io/vscsiif.h new file mode 100644 index 0000000..3ce2914 --- /dev/null +++ b/include/xenpvdrivers/interface/io/vscsiif.h @@ -0,0 +1,105 @@ +/****************************************************************************** + * vscsiif.h + * + * Based on the blkif.h code. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright(c) FUJITSU Limited 2008. + */ + +#ifndef __XEN__PUBLIC_IO_SCSI_H__ +#define __XEN__PUBLIC_IO_SCSI_H__ + +#include "ring.h" +#include "../grant_table.h" + +/* command between backend and frontend */ +#define VSCSIIF_ACT_SCSI_CDB 1 /* SCSI CDB command */ +#define VSCSIIF_ACT_SCSI_ABORT 2 /* SCSI Device(Lun) Abort*/ +#define VSCSIIF_ACT_SCSI_RESET 3 /* SCSI Device(Lun) Reset*/ + + +#define VSCSIIF_BACK_MAX_PENDING_REQS 128 + +/* + * Maximum scatter/gather segments per request. + * + * Considering balance between allocating al least 16 "vscsiif_request" + * structures on one page (4096bytes) and number of scatter gather + * needed, we decided to use 26 as a magic number. + */ +#define VSCSIIF_SG_TABLESIZE 26 + +/* + * base on linux kernel 2.6.18 + */ +#define VSCSIIF_MAX_COMMAND_SIZE 16 +#define VSCSIIF_SENSE_BUFFERSIZE 96 + + +struct vscsiif_request { + uint16_t rqid; /* private guest value, echoed in resp */ + uint8_t act; /* command between backend and frontend */ + uint8_t cmd_len; + + uint8_t cmnd[VSCSIIF_MAX_COMMAND_SIZE]; + uint16_t timeout_per_command; /* The command is issued by twice + the value in Backend. */ + uint16_t channel, id, lun; + uint16_t padding; + uint8_t sc_data_direction; /* for DMA_TO_DEVICE(1) + DMA_FROM_DEVICE(2) + DMA_NONE(3) requests */ + uint8_t nr_segments; /* Number of pieces of scatter-gather */ + + struct scsiif_request_segment { + grant_ref_t gref; + uint16_t offset; + uint16_t length; + } seg[VSCSIIF_SG_TABLESIZE]; + uint32_t reserved[3]; +}; +typedef struct vscsiif_request vscsiif_request_t; + +struct vscsiif_response { + uint16_t rqid; + uint8_t padding; + uint8_t sense_len; + uint8_t sense_buffer[VSCSIIF_SENSE_BUFFERSIZE]; + int32_t rslt; + uint32_t residual_len; /* request bufflen - + return the value from physical device */ + uint32_t reserved[36]; +}; +typedef struct vscsiif_response vscsiif_response_t; + +DEFINE_RING_TYPES(vscsiif, struct vscsiif_request, struct vscsiif_response); + + +#endif /*__XEN__PUBLIC_IO_SCSI_H__*/ +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/io/xenbus.h b/include/xenpvdrivers/interface/io/xenbus.h new file mode 100644 index 0000000..2fbf2a7 --- /dev/null +++ b/include/xenpvdrivers/interface/io/xenbus.h @@ -0,0 +1,70 @@ +/***************************************************************************** + * xenbus.h + * + * Xenbus protocol details. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (C) 2005 XenSource Ltd. + */ + +#ifndef _XEN_PUBLIC_IO_XENBUS_H +#define _XEN_PUBLIC_IO_XENBUS_H + +/* + * The state of either end of the Xenbus, i.e. the current communication + * status of initialisation across the bus. States here imply nothing about + * the state of the connection between the driver and the kernel's device + * layers. + */ +enum xenbus_state { + XenbusStateUnknown = 0, + + XenbusStateInitialising = 1, + + /* + * InitWait: Finished early initialisation but waiting for information + * from the peer or hotplug scripts. + */ + XenbusStateInitWait = 2, + + /* + * Initialised: Waiting for a connection from the peer. + */ + XenbusStateInitialised = 3, + + XenbusStateConnected = 4, + + /* + * Closing: The device is being closed due to an error or an unplug event. + */ + XenbusStateClosing = 5, + + XenbusStateClosed = 6, + + /* + * Reconfiguring: The device is being reconfigured. + */ + XenbusStateReconfiguring = 7, + + XenbusStateReconfigured = 8 +}; +typedef enum xenbus_state XenbusState; + +#endif /* _XEN_PUBLIC_IO_XENBUS_H */ diff --git a/include/xenpvdrivers/interface/io/xs_wire.h b/include/xenpvdrivers/interface/io/xs_wire.h new file mode 100644 index 0000000..b0be9d4 --- /dev/null +++ b/include/xenpvdrivers/interface/io/xs_wire.h @@ -0,0 +1,122 @@ +/* + * Details of the "wire" protocol between Xen Store Daemon and client + * library or guest kernel. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (C) 2005 Rusty Russell IBM Corporation + */ + +#ifndef _XS_WIRE_H +#define _XS_WIRE_H + +enum xsd_sockmsg_type +{ + XS_DEBUG, + XS_DIRECTORY, + XS_READ, + XS_GET_PERMS, + XS_WATCH, + XS_UNWATCH, + XS_TRANSACTION_START, + XS_TRANSACTION_END, + XS_INTRODUCE, + XS_RELEASE, + XS_GET_DOMAIN_PATH, + XS_WRITE, + XS_MKDIR, + XS_RM, + XS_SET_PERMS, + XS_WATCH_EVENT, + XS_ERROR, + XS_IS_DOMAIN_INTRODUCED, + XS_RESUME, + XS_SET_TARGET +}; + +#define XS_WRITE_NONE "NONE" +#define XS_WRITE_CREATE "CREATE" +#define XS_WRITE_CREATE_EXCL "CREATE|EXCL" + +/* We hand errors as strings, for portability. */ +struct xsd_errors +{ + int errnum; + const char *errstring; +}; +#ifdef EINVAL +#define XSD_ERROR(x) { x, #x } +/* LINTED: static unused */ +static struct xsd_errors xsd_errors[] +#if defined(__GNUC__) +__attribute__((unused)) +#endif + = { + XSD_ERROR(EINVAL), + XSD_ERROR(EACCES), + XSD_ERROR(EEXIST), + XSD_ERROR(EISDIR), + XSD_ERROR(ENOENT), + XSD_ERROR(ENOMEM), + XSD_ERROR(ENOSPC), + XSD_ERROR(EIO), + XSD_ERROR(ENOTEMPTY), + XSD_ERROR(ENOSYS), + XSD_ERROR(EROFS), + XSD_ERROR(EBUSY), + XSD_ERROR(EAGAIN), + XSD_ERROR(EISCONN) +}; +#endif + +struct xsd_sockmsg +{ + uint32_t type; /* XS_??? */ + uint32_t req_id;/* Request identifier, echoed in daemon's response. */ + uint32_t tx_id; /* Transaction id (0 if not related to a transaction). */ + uint32_t len; /* Length of data following this. */ + + /* Generally followed by nul-terminated string(s). */ +}; + +enum xs_watch_type +{ + XS_WATCH_PATH = 0, + XS_WATCH_TOKEN +}; + +/* Inter-domain shared memory communications. */ +#define XENSTORE_RING_SIZE 1024 +typedef uint32_t XENSTORE_RING_IDX; +#define MASK_XENSTORE_IDX(idx) ((idx) & (XENSTORE_RING_SIZE-1)) +struct xenstore_domain_interface { + char req[XENSTORE_RING_SIZE]; /* Requests to xenstore daemon. */ + char rsp[XENSTORE_RING_SIZE]; /* Replies and async watch events. */ + XENSTORE_RING_IDX req_cons, req_prod; + XENSTORE_RING_IDX rsp_cons, rsp_prod; +}; + +/* Violating this is very bad. See docs/misc/xenstore.txt. */ +#define XENSTORE_PAYLOAD_MAX 4096 + +/* Violating these just gets you an error back */ +#define XENSTORE_ABS_PATH_MAX 3072 +#define XENSTORE_REL_PATH_MAX 2048 + +#endif /* _XS_WIRE_H */ diff --git a/include/xenpvdrivers/interface/kexec.h b/include/xenpvdrivers/interface/kexec.h new file mode 100644 index 0000000..0425222 --- /dev/null +++ b/include/xenpvdrivers/interface/kexec.h @@ -0,0 +1,168 @@ +/****************************************************************************** + * kexec.h - Public portion + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Xen port written by: + * - Simon 'Horms' Horman + * - Magnus Damm + */ + +#ifndef _XEN_PUBLIC_KEXEC_H +#define _XEN_PUBLIC_KEXEC_H + + +/* This file describes the Kexec / Kdump hypercall interface for Xen. + * + * Kexec under vanilla Linux allows a user to reboot the physical machine + * into a new user-specified kernel. The Xen port extends this idea + * to allow rebooting of the machine from dom0. When kexec for dom0 + * is used to reboot, both the hypervisor and the domains get replaced + * with some other kernel. It is possible to kexec between vanilla + * Linux and Xen and back again. Xen to Xen works well too. + * + * The hypercall interface for kexec can be divided into three main + * types of hypercall operations: + * + * 1) Range information: + * This is used by the dom0 kernel to ask the hypervisor about various + * address information. This information is needed to allow kexec-tools + * to fill in the ELF headers for /proc/vmcore properly. + * + * 2) Load and unload of images: + * There are no big surprises here, the kexec binary from kexec-tools + * runs in userspace in dom0. The tool loads/unloads data into the + * dom0 kernel such as new kernel, initramfs and hypervisor. When + * loaded the dom0 kernel performs a load hypercall operation, and + * before releasing all page references the dom0 kernel calls unload. + * + * 3) Kexec operation: + * This is used to start a previously loaded kernel. + */ + +#include "xen.h" + +#if defined(__i386__) || defined(__x86_64__) +#define KEXEC_XEN_NO_PAGES 17 +#endif + +/* + * Prototype for this hypercall is: + * int kexec_op(int cmd, void *args) + * @cmd == KEXEC_CMD_... + * KEXEC operation to perform + * @args == Operation-specific extra arguments (NULL if none). + */ + +/* + * Kexec supports two types of operation: + * - kexec into a regular kernel, very similar to a standard reboot + * - KEXEC_TYPE_DEFAULT is used to specify this type + * - kexec into a special "crash kernel", aka kexec-on-panic + * - KEXEC_TYPE_CRASH is used to specify this type + * - parts of our system may be broken at kexec-on-panic time + * - the code should be kept as simple and self-contained as possible + */ + +#define KEXEC_TYPE_DEFAULT 0 +#define KEXEC_TYPE_CRASH 1 + + +/* The kexec implementation for Xen allows the user to load two + * types of kernels, KEXEC_TYPE_DEFAULT and KEXEC_TYPE_CRASH. + * All data needed for a kexec reboot is kept in one xen_kexec_image_t + * per "instance". The data mainly consists of machine address lists to pages + * together with destination addresses. The data in xen_kexec_image_t + * is passed to the "code page" which is one page of code that performs + * the final relocations before jumping to the new kernel. + */ + +typedef struct xen_kexec_image { +#if defined(__i386__) || defined(__x86_64__) + unsigned long page_list[KEXEC_XEN_NO_PAGES]; +#endif +#if defined(__ia64__) + unsigned long reboot_code_buffer; +#endif + unsigned long indirection_page; + unsigned long start_address; +} xen_kexec_image_t; + +/* + * Perform kexec having previously loaded a kexec or kdump kernel + * as appropriate. + * type == KEXEC_TYPE_DEFAULT or KEXEC_TYPE_CRASH [in] + */ +#define KEXEC_CMD_kexec 0 +typedef struct xen_kexec_exec { + int type; +} xen_kexec_exec_t; + +/* + * Load/Unload kernel image for kexec or kdump. + * type == KEXEC_TYPE_DEFAULT or KEXEC_TYPE_CRASH [in] + * image == relocation information for kexec (ignored for unload) [in] + */ +#define KEXEC_CMD_kexec_load 1 +#define KEXEC_CMD_kexec_unload 2 +typedef struct xen_kexec_load { + int type; + xen_kexec_image_t image; +} xen_kexec_load_t; + +#define KEXEC_RANGE_MA_CRASH 0 /* machine address and size of crash area */ +#define KEXEC_RANGE_MA_XEN 1 /* machine address and size of Xen itself */ +#define KEXEC_RANGE_MA_CPU 2 /* machine address and size of a CPU note */ +#define KEXEC_RANGE_MA_XENHEAP 3 /* machine address and size of xenheap + * Note that although this is adjacent + * to Xen it exists in a separate EFI + * region on ia64, and thus needs to be + * inserted into iomem_machine separately */ +#define KEXEC_RANGE_MA_BOOT_PARAM 4 /* machine address and size of + * the ia64_boot_param */ +#define KEXEC_RANGE_MA_EFI_MEMMAP 5 /* machine address and size of + * of the EFI Memory Map */ +#define KEXEC_RANGE_MA_VMCOREINFO 6 /* machine address and size of vmcoreinfo */ + +/* + * Find the address and size of certain memory areas + * range == KEXEC_RANGE_... [in] + * nr == physical CPU number (starting from 0) if KEXEC_RANGE_MA_CPU [in] + * size == number of bytes reserved in window [out] + * start == address of the first byte in the window [out] + */ +#define KEXEC_CMD_kexec_get_range 3 +typedef struct xen_kexec_range { + int range; + int nr; + unsigned long size; + unsigned long start; +} xen_kexec_range_t; + +#endif /* _XEN_PUBLIC_KEXEC_H */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/mem_event.h b/include/xenpvdrivers/interface/mem_event.h new file mode 100644 index 0000000..db0acf5 --- /dev/null +++ b/include/xenpvdrivers/interface/mem_event.h @@ -0,0 +1,69 @@ +/****************************************************************************** + * mem_event.h + * + * Memory event common structures. + * + * Copyright (c) 2009 by Citrix Systems, Inc. (Patrick Colp) + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef _XEN_PUBLIC_MEM_EVENT_H +#define _XEN_PUBLIC_MEM_EVENT_H + + +#include "xen.h" +#include "io/ring.h" + + +/* Memory event notification modes */ +#define MEM_EVENT_MODE_ASYNC 0 +#define MEM_EVENT_MODE_SYNC (1 << 0) +#define MEM_EVENT_MODE_SYNC_ALL (1 << 1) + +/* Memory event flags */ +#define MEM_EVENT_FLAG_VCPU_PAUSED (1 << 0) +#define MEM_EVENT_FLAG_DOM_PAUSED (1 << 1) +#define MEM_EVENT_FLAG_OUT_OF_MEM (1 << 2) + + +typedef struct mem_event_shared_page { + int port; +} mem_event_shared_page_t; + +typedef struct mem_event_st { + unsigned long gfn; + unsigned long offset; + unsigned long p2mt; + int vcpu_id; + uint64_t flags; +} mem_event_request_t, mem_event_response_t; + + +DEFINE_RING_TYPES(mem_event, mem_event_request_t, mem_event_response_t); + + +#endif + + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/memory.h b/include/xenpvdrivers/interface/memory.h new file mode 100644 index 0000000..ce5c17e --- /dev/null +++ b/include/xenpvdrivers/interface/memory.h @@ -0,0 +1,294 @@ +/****************************************************************************** + * memory.h + * + * Memory reservation and information. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2005, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_MEMORY_H__ +#define __XEN_PUBLIC_MEMORY_H__ + +#include "xen.h" + +/* + * Increase or decrease the specified domain's memory reservation. Returns the + * number of extents successfully allocated or freed. + * arg == addr of struct xen_memory_reservation. + */ +#define XENMEM_increase_reservation 0 +#define XENMEM_decrease_reservation 1 +#define XENMEM_populate_physmap 6 + +#if __XEN_INTERFACE_VERSION__ >= 0x00030209 +/* + * Maximum # bits addressable by the user of the allocated region (e.g., I/O + * devices often have a 32-bit limitation even in 64-bit systems). If zero + * then the user has no addressing restriction. This field is not used by + * XENMEM_decrease_reservation. + */ +#define XENMEMF_address_bits(x) (x) +#define XENMEMF_get_address_bits(x) ((x) & 0xffu) +/* NUMA node to allocate from. */ +#define XENMEMF_node(x) (((x) + 1) << 8) +#define XENMEMF_get_node(x) ((((x) >> 8) - 1) & 0xffu) +/* Flag to populate physmap with populate-on-demand entries */ +#define XENMEMF_populate_on_demand (1<<16) +#endif + +struct xen_memory_reservation { + + /* + * XENMEM_increase_reservation: + * OUT: MFN (*not* GMFN) bases of extents that were allocated + * XENMEM_decrease_reservation: + * IN: GMFN bases of extents to free + * XENMEM_populate_physmap: + * IN: GPFN bases of extents to populate with memory + * OUT: GMFN bases of extents that were allocated + * (NB. This command also updates the mach_to_phys translation table) + */ + XEN_GUEST_HANDLE(xen_pfn_t) extent_start; + + /* Number of extents, and size/alignment of each (2^extent_order pages). */ + xen_ulong_t nr_extents; + unsigned int extent_order; + +#if __XEN_INTERFACE_VERSION__ >= 0x00030209 + /* XENMEMF flags. */ + unsigned int mem_flags; +#else + unsigned int address_bits; +#endif + + /* + * Domain whose reservation is being changed. + * Unprivileged domains can specify only DOMID_SELF. + */ + domid_t domid; +}; +DEFINE_GUEST_HANDLE_STRUCT(xen_memory_reservation); +typedef struct xen_memory_reservation xen_memory_reservation_t; +DEFINE_XEN_GUEST_HANDLE(xen_memory_reservation_t); + +/* + * An atomic exchange of memory pages. If return code is zero then + * @out.extent_list provides GMFNs of the newly-allocated memory. + * Returns zero on complete success, otherwise a negative error code. + * On complete success then always @nr_exchanged == @in.nr_extents. + * On partial success @nr_exchanged indicates how much work was done. + */ +#define XENMEM_exchange 11 +struct xen_memory_exchange { + /* + * [IN] Details of memory extents to be exchanged (GMFN bases). + * Note that @in.address_bits is ignored and unused. + */ + struct xen_memory_reservation in; + + /* + * [IN/OUT] Details of new memory extents. + * We require that: + * 1. @in.domid == @out.domid + * 2. @in.nr_extents << @in.extent_order == + * @out.nr_extents << @out.extent_order + * 3. @in.extent_start and @out.extent_start lists must not overlap + * 4. @out.extent_start lists GPFN bases to be populated + * 5. @out.extent_start is overwritten with allocated GMFN bases + */ + struct xen_memory_reservation out; + + /* + * [OUT] Number of input extents that were successfully exchanged: + * 1. The first @nr_exchanged input extents were successfully + * deallocated. + * 2. The corresponding first entries in the output extent list correctly + * indicate the GMFNs that were successfully exchanged. + * 3. All other input and output extents are untouched. + * 4. If not all input exents are exchanged then the return code of this + * command will be non-zero. + * 5. THIS FIELD MUST BE INITIALISED TO ZERO BY THE CALLER! + */ + xen_ulong_t nr_exchanged; +}; +typedef struct xen_memory_exchange xen_memory_exchange_t; +DEFINE_XEN_GUEST_HANDLE(xen_memory_exchange_t); + +/* + * Returns the maximum machine frame number of mapped RAM in this system. + * This command always succeeds (it never returns an error code). + * arg == NULL. + */ +#define XENMEM_maximum_ram_page 2 + +/* + * Returns the current or maximum memory reservation, in pages, of the + * specified domain (may be DOMID_SELF). Returns -ve errcode on failure. + * arg == addr of domid_t. + */ +#define XENMEM_current_reservation 3 +#define XENMEM_maximum_reservation 4 + +/* + * Returns the maximum GPFN in use by the guest, or -ve errcode on failure. + */ +#define XENMEM_maximum_gpfn 14 + +/* + * Returns a list of MFN bases of 2MB extents comprising the machine_to_phys + * mapping table. Architectures which do not have a m2p table do not implement + * this command. + * arg == addr of xen_machphys_mfn_list_t. + */ +#define XENMEM_machphys_mfn_list 5 +struct xen_machphys_mfn_list { + /* + * Size of the 'extent_start' array. Fewer entries will be filled if the + * machphys table is smaller than max_extents * 2MB. + */ + unsigned int max_extents; + + /* + * Pointer to buffer to fill with list of extent starts. If there are + * any large discontiguities in the machine address space, 2MB gaps in + * the machphys table will be represented by an MFN base of zero. + */ + XEN_GUEST_HANDLE(xen_pfn_t) extent_start; + + /* + * Number of extents written to the above array. This will be smaller + * than 'max_extents' if the machphys table is smaller than max_e * 2MB. + */ + unsigned int nr_extents; +}; +DEFINE_GUEST_HANDLE_STRUCT(xen_machphys_mfn_list); +typedef struct xen_machphys_mfn_list xen_machphys_mfn_list_t; +DEFINE_XEN_GUEST_HANDLE(xen_machphys_mfn_list_t); + +/* + * Returns the location in virtual address space of the machine_to_phys + * mapping table. Architectures which do not have a m2p table, or which do not + * map it by default into guest address space, do not implement this command. + * arg == addr of xen_machphys_mapping_t. + */ +#define XENMEM_machphys_mapping 12 +struct xen_machphys_mapping { + xen_ulong_t v_start, v_end; /* Start and end virtual addresses. */ + xen_ulong_t max_mfn; /* Maximum MFN that can be looked up. */ +}; +typedef struct xen_machphys_mapping xen_machphys_mapping_t; +DEFINE_XEN_GUEST_HANDLE(xen_machphys_mapping_t); + +/* + * Sets the GPFN at which a particular page appears in the specified guest's + * pseudophysical address space. + * arg == addr of xen_add_to_physmap_t. + */ +#define XENMEM_add_to_physmap 7 +struct xen_add_to_physmap { + /* Which domain to change the mapping for. */ + domid_t domid; + + /* Source mapping space. */ +#define XENMAPSPACE_shared_info 0 /* shared info page */ +#define XENMAPSPACE_grant_table 1 /* grant table page */ +#define XENMAPSPACE_gmfn 2 /* GMFN */ + unsigned int space; + +#define XENMAPIDX_grant_table_status 0x80000000 + + /* Index into source mapping space. */ + xen_ulong_t idx; + + /* GPFN where the source mapping page should appear. */ + xen_pfn_t gpfn; +}; +DEFINE_GUEST_HANDLE_STRUCT(xen_add_to_physmap); +typedef struct xen_add_to_physmap xen_add_to_physmap_t; +DEFINE_XEN_GUEST_HANDLE(xen_add_to_physmap_t); + +/*** REMOVED ***/ +/*#define XENMEM_translate_gpfn_list 8*/ + +/* + * Returns the pseudo-physical memory map as it was when the domain + * was started (specified by XENMEM_set_memory_map). + * arg == addr of xen_memory_map_t. + */ +#define XENMEM_memory_map 9 +struct xen_memory_map { + /* + * On call the number of entries which can be stored in buffer. On + * return the number of entries which have been stored in + * buffer. + */ + unsigned int nr_entries; + + /* + * Entries in the buffer are in the same format as returned by the + * BIOS INT 0x15 EAX=0xE820 call. + */ + XEN_GUEST_HANDLE(void) buffer; +}; +typedef struct xen_memory_map xen_memory_map_t; +DEFINE_XEN_GUEST_HANDLE(xen_memory_map_t); + +/* + * Returns the real physical memory map. Passes the same structure as + * XENMEM_memory_map. + * arg == addr of xen_memory_map_t. + */ +#define XENMEM_machine_memory_map 10 + +/* + * Set the pseudo-physical memory map of a domain, as returned by + * XENMEM_memory_map. + * arg == addr of xen_foreign_memory_map_t. + */ +#define XENMEM_set_memory_map 13 +struct xen_foreign_memory_map { + domid_t domid; + struct xen_memory_map map; +}; +typedef struct xen_foreign_memory_map xen_foreign_memory_map_t; +DEFINE_XEN_GUEST_HANDLE(xen_foreign_memory_map_t); + +#define XENMEM_set_pod_target 16 +#define XENMEM_get_pod_target 17 +struct xen_pod_target { + /* IN */ + uint64_t target_pages; + /* OUT */ + uint64_t tot_pages; + uint64_t pod_cache_pages; + uint64_t pod_entries; + /* IN */ + domid_t domid; +}; +typedef struct xen_pod_target xen_pod_target_t; + +/* + * Get the number of MFNs saved through memory sharing. + * The call never fails. + */ +#define XENMEM_get_sharing_freed_pages 18 + +#endif /* __XEN_PUBLIC_MEMORY_H__ */ diff --git a/include/xenpvdrivers/interface/nmi.h b/include/xenpvdrivers/interface/nmi.h new file mode 100644 index 0000000..2fd21d2 --- /dev/null +++ b/include/xenpvdrivers/interface/nmi.h @@ -0,0 +1,80 @@ +/****************************************************************************** + * nmi.h + * + * NMI callback registration and reason codes. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2005, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_NMI_H__ +#define __XEN_PUBLIC_NMI_H__ + +#include "xen.h" + +/* + * NMI reason codes: + * Currently these are x86-specific, stored in arch_shared_info.nmi_reason. + */ + /* I/O-check error reported via ISA port 0x61, bit 6. */ +#define _XEN_NMIREASON_io_error 0 +#define XEN_NMIREASON_io_error (1UL << _XEN_NMIREASON_io_error) + /* Parity error reported via ISA port 0x61, bit 7. */ +#define _XEN_NMIREASON_parity_error 1 +#define XEN_NMIREASON_parity_error (1UL << _XEN_NMIREASON_parity_error) + /* Unknown hardware-generated NMI. */ +#define _XEN_NMIREASON_unknown 2 +#define XEN_NMIREASON_unknown (1UL << _XEN_NMIREASON_unknown) + +/* + * long nmi_op(unsigned int cmd, void *arg) + * NB. All ops return zero on success, else a negative error code. + */ + +/* + * Register NMI callback for this (calling) VCPU. Currently this only makes + * sense for domain 0, vcpu 0. All other callers will be returned EINVAL. + * arg == pointer to xennmi_callback structure. + */ +#define XENNMI_register_callback 0 +struct xennmi_callback { + unsigned long handler_address; + unsigned long pad; +}; +typedef struct xennmi_callback xennmi_callback_t; +DEFINE_XEN_GUEST_HANDLE(xennmi_callback_t); + +/* + * Deregister NMI callback for this (calling) VCPU. + * arg == NULL. + */ +#define XENNMI_unregister_callback 1 + +#endif /* __XEN_PUBLIC_NMI_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/physdev.h b/include/xenpvdrivers/interface/physdev.h new file mode 100644 index 0000000..07e7284 --- /dev/null +++ b/include/xenpvdrivers/interface/physdev.h @@ -0,0 +1,264 @@ +/* + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __XEN_PUBLIC_PHYSDEV_H__ +#define __XEN_PUBLIC_PHYSDEV_H__ + +#include "xen.h" + +/* + * Prototype for this hypercall is: + * int physdev_op(int cmd, void *args) + * @cmd == PHYSDEVOP_??? (physdev operation). + * @args == Operation-specific extra arguments (NULL if none). + */ + +/* + * Notify end-of-interrupt (EOI) for the specified IRQ. + * @arg == pointer to physdev_eoi structure. + */ +#define PHYSDEVOP_eoi 12 +struct physdev_eoi { + /* IN */ + uint32_t irq; +}; +typedef struct physdev_eoi physdev_eoi_t; +DEFINE_XEN_GUEST_HANDLE(physdev_eoi_t); + +/* + * Register a shared page for the hypervisor to indicate whether the guest + * must issue PHYSDEVOP_eoi. The semantics of PHYSDEVOP_eoi change slightly + * once the guest used this function in that the associated event channel + * will automatically get unmasked. The page registered is used as a bit + * array indexed by Xen's PIRQ value. + */ +#define PHYSDEVOP_pirq_eoi_gmfn 17 +struct physdev_pirq_eoi_gmfn { + /* IN */ + xen_pfn_t gmfn; +}; +typedef struct physdev_pirq_eoi_gmfn physdev_pirq_eoi_gmfn_t; +DEFINE_XEN_GUEST_HANDLE(physdev_pirq_eoi_gmfn_t); + +/* + * Query the status of an IRQ line. + * @arg == pointer to physdev_irq_status_query structure. + */ +#define PHYSDEVOP_irq_status_query 5 +struct physdev_irq_status_query { + /* IN */ + uint32_t irq; + /* OUT */ + uint32_t flags; /* XENIRQSTAT_* */ +}; +typedef struct physdev_irq_status_query physdev_irq_status_query_t; +DEFINE_XEN_GUEST_HANDLE(physdev_irq_status_query_t); + +/* Need to call PHYSDEVOP_eoi when the IRQ has been serviced? */ +#define _XENIRQSTAT_needs_eoi (0) +#define XENIRQSTAT_needs_eoi (1U<<_XENIRQSTAT_needs_eoi) + +/* IRQ shared by multiple guests? */ +#define _XENIRQSTAT_shared (1) +#define XENIRQSTAT_shared (1U<<_XENIRQSTAT_shared) + +/* + * Set the current VCPU's I/O privilege level. + * @arg == pointer to physdev_set_iopl structure. + */ +#define PHYSDEVOP_set_iopl 6 +struct physdev_set_iopl { + /* IN */ + uint32_t iopl; +}; +typedef struct physdev_set_iopl physdev_set_iopl_t; +DEFINE_XEN_GUEST_HANDLE(physdev_set_iopl_t); + +/* + * Set the current VCPU's I/O-port permissions bitmap. + * @arg == pointer to physdev_set_iobitmap structure. + */ +#define PHYSDEVOP_set_iobitmap 7 +struct physdev_set_iobitmap { + /* IN */ +#if __XEN_INTERFACE_VERSION__ >= 0x00030205 + XEN_GUEST_HANDLE(uint8) bitmap; +#else + uint8_t *bitmap; +#endif + uint32_t nr_ports; +}; +typedef struct physdev_set_iobitmap physdev_set_iobitmap_t; +DEFINE_XEN_GUEST_HANDLE(physdev_set_iobitmap_t); + +/* + * Read or write an IO-APIC register. + * @arg == pointer to physdev_apic structure. + */ +#define PHYSDEVOP_apic_read 8 +#define PHYSDEVOP_apic_write 9 +struct physdev_apic { + /* IN */ + unsigned long apic_physbase; + uint32_t reg; + /* IN or OUT */ + uint32_t value; +}; +typedef struct physdev_apic physdev_apic_t; +DEFINE_XEN_GUEST_HANDLE(physdev_apic_t); + +/* + * Allocate or free a physical upcall vector for the specified IRQ line. + * @arg == pointer to physdev_irq structure. + */ +#define PHYSDEVOP_alloc_irq_vector 10 +#define PHYSDEVOP_free_irq_vector 11 +struct physdev_irq { + /* IN */ + uint32_t irq; + /* IN or OUT */ + uint32_t vector; +}; +typedef struct physdev_irq physdev_irq_t; +DEFINE_XEN_GUEST_HANDLE(physdev_irq_t); + +#define MAP_PIRQ_TYPE_MSI 0x0 +#define MAP_PIRQ_TYPE_GSI 0x1 +#define MAP_PIRQ_TYPE_UNKNOWN 0x2 + +#define PHYSDEVOP_map_pirq 13 +struct physdev_map_pirq { + domid_t domid; + /* IN */ + int type; + /* IN */ + int index; + /* IN or OUT */ + int pirq; + /* IN */ + int bus; + /* IN */ + int devfn; + /* IN */ + int entry_nr; + /* IN */ + uint64_t table_base; +}; +typedef struct physdev_map_pirq physdev_map_pirq_t; +DEFINE_XEN_GUEST_HANDLE(physdev_map_pirq_t); + +#define PHYSDEVOP_unmap_pirq 14 +struct physdev_unmap_pirq { + domid_t domid; + /* IN */ + int pirq; +}; + +typedef struct physdev_unmap_pirq physdev_unmap_pirq_t; +DEFINE_XEN_GUEST_HANDLE(physdev_unmap_pirq_t); + +#define PHYSDEVOP_manage_pci_add 15 +#define PHYSDEVOP_manage_pci_remove 16 +struct physdev_manage_pci { + /* IN */ + uint8_t bus; + uint8_t devfn; +}; + +typedef struct physdev_manage_pci physdev_manage_pci_t; +DEFINE_XEN_GUEST_HANDLE(physdev_manage_pci_t); + +#define PHYSDEVOP_restore_msi 19 +struct physdev_restore_msi { + /* IN */ + uint8_t bus; + uint8_t devfn; +}; +typedef struct physdev_restore_msi physdev_restore_msi_t; +DEFINE_XEN_GUEST_HANDLE(physdev_restore_msi_t); + +#define PHYSDEVOP_manage_pci_add_ext 20 +struct physdev_manage_pci_ext { + /* IN */ + uint8_t bus; + uint8_t devfn; + unsigned is_extfn; + unsigned is_virtfn; + struct { + uint8_t bus; + uint8_t devfn; + } physfn; +}; + +typedef struct physdev_manage_pci_ext physdev_manage_pci_ext_t; +DEFINE_XEN_GUEST_HANDLE(physdev_manage_pci_ext_t); + +/* + * Argument to physdev_op_compat() hypercall. Superceded by new physdev_op() + * hypercall since 0x00030202. + */ +struct physdev_op { + uint32_t cmd; + union { + struct physdev_irq_status_query irq_status_query; + struct physdev_set_iopl set_iopl; + struct physdev_set_iobitmap set_iobitmap; + struct physdev_apic apic_op; + struct physdev_irq irq_op; + } u; +}; +typedef struct physdev_op physdev_op_t; +DEFINE_XEN_GUEST_HANDLE(physdev_op_t); + +#define PHYSDEVOP_setup_gsi 21 +struct physdev_setup_gsi { + int gsi; + /* IN */ + uint8_t triggering; + /* IN */ + uint8_t polarity; + /* IN */ +}; + +typedef struct physdev_setup_gsi physdev_setup_gsi_t; +DEFINE_XEN_GUEST_HANDLE(physdev_setup_gsi_t); + +/* + * Notify that some PIRQ-bound event channels have been unmasked. + * ** This command is obsolete since interface version 0x00030202 and is ** + * ** unsupported by newer versions of Xen. ** + */ +#define PHYSDEVOP_IRQ_UNMASK_NOTIFY 4 + +/* + * These all-capitals physdev operation names are superceded by the new names + * (defined above) since interface version 0x00030202. + */ +#define PHYSDEVOP_IRQ_STATUS_QUERY PHYSDEVOP_irq_status_query +#define PHYSDEVOP_SET_IOPL PHYSDEVOP_set_iopl +#define PHYSDEVOP_SET_IOBITMAP PHYSDEVOP_set_iobitmap +#define PHYSDEVOP_APIC_READ PHYSDEVOP_apic_read +#define PHYSDEVOP_APIC_WRITE PHYSDEVOP_apic_write +#define PHYSDEVOP_ASSIGN_VECTOR PHYSDEVOP_alloc_irq_vector +#define PHYSDEVOP_FREE_VECTOR PHYSDEVOP_free_irq_vector +#define PHYSDEVOP_IRQ_NEEDS_UNMASK_NOTIFY XENIRQSTAT_needs_eoi +#define PHYSDEVOP_IRQ_SHARED XENIRQSTAT_shared + +#endif /* __XEN_PUBLIC_PHYSDEV_H__ */ diff --git a/include/xenpvdrivers/interface/platform.h b/include/xenpvdrivers/interface/platform.h new file mode 100644 index 0000000..1a519f4 --- /dev/null +++ b/include/xenpvdrivers/interface/platform.h @@ -0,0 +1,402 @@ +/****************************************************************************** + * platform.h + * + * Hardware platform operations. Intended for use by domain-0 kernel. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2002-2006, K Fraser + */ + +#ifndef __XEN_PUBLIC_PLATFORM_H__ +#define __XEN_PUBLIC_PLATFORM_H__ + +#include "xen.h" + +#define XENPF_INTERFACE_VERSION 0x03000001 + +/* + * Set clock such that it would read after 00:00:00 UTC, + * 1 January, 1970 if the current system time was . + */ +#define XENPF_settime 17 +struct xenpf_settime { + /* IN variables. */ + uint32_t secs; + uint32_t nsecs; + uint64_t system_time; +}; +typedef struct xenpf_settime xenpf_settime_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_settime_t); + +/* + * Request memory range (@mfn, @mfn+@nr_mfns-1) to have type @type. + * On x86, @type is an architecture-defined MTRR memory type. + * On success, returns the MTRR that was used (@reg) and a handle that can + * be passed to XENPF_DEL_MEMTYPE to accurately tear down the new setting. + * (x86-specific). + */ +#define XENPF_add_memtype 31 +struct xenpf_add_memtype { + /* IN variables. */ + xen_pfn_t mfn; + uint64_t nr_mfns; + uint32_t type; + /* OUT variables. */ + uint32_t handle; + uint32_t reg; +}; +typedef struct xenpf_add_memtype xenpf_add_memtype_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_add_memtype_t); + +/* + * Tear down an existing memory-range type. If @handle is remembered then it + * should be passed in to accurately tear down the correct setting (in case + * of overlapping memory regions with differing types). If it is not known + * then @handle should be set to zero. In all cases @reg must be set. + * (x86-specific). + */ +#define XENPF_del_memtype 32 +struct xenpf_del_memtype { + /* IN variables. */ + uint32_t handle; + uint32_t reg; +}; +typedef struct xenpf_del_memtype xenpf_del_memtype_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_del_memtype_t); + +/* Read current type of an MTRR (x86-specific). */ +#define XENPF_read_memtype 33 +struct xenpf_read_memtype { + /* IN variables. */ + uint32_t reg; + /* OUT variables. */ + xen_pfn_t mfn; + uint64_t nr_mfns; + uint32_t type; +}; +typedef struct xenpf_read_memtype xenpf_read_memtype_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_read_memtype_t); + +#define XENPF_microcode_update 35 +struct xenpf_microcode_update { + /* IN variables. */ + XEN_GUEST_HANDLE(const_void) data;/* Pointer to microcode data */ + uint32_t length; /* Length of microcode data. */ +}; +typedef struct xenpf_microcode_update xenpf_microcode_update_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_microcode_update_t); + +#define XENPF_platform_quirk 39 +#define QUIRK_NOIRQBALANCING 1 /* Do not restrict IO-APIC RTE targets */ +#define QUIRK_IOAPIC_BAD_REGSEL 2 /* IO-APIC REGSEL forgets its value */ +#define QUIRK_IOAPIC_GOOD_REGSEL 3 /* IO-APIC REGSEL behaves properly */ +struct xenpf_platform_quirk { + /* IN variables. */ + uint32_t quirk_id; +}; +typedef struct xenpf_platform_quirk xenpf_platform_quirk_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_platform_quirk_t); + +#define XENPF_firmware_info 50 +#define XEN_FW_DISK_INFO 1 /* from int 13 AH=08/41/48 */ +#define XEN_FW_DISK_MBR_SIGNATURE 2 /* from MBR offset 0x1b8 */ +#define XEN_FW_VBEDDC_INFO 3 /* from int 10 AX=4f15 */ +struct xenpf_firmware_info { + /* IN variables. */ + uint32_t type; + uint32_t index; + /* OUT variables. */ + union { + struct { + /* Int13, Fn48: Check Extensions Present. */ + uint8_t device; /* %dl: bios device number */ + uint8_t version; /* %ah: major version */ + uint16_t interface_support; /* %cx: support bitmap */ + /* Int13, Fn08: Legacy Get Device Parameters. */ + uint16_t legacy_max_cylinder; /* %cl[7:6]:%ch: max cyl # */ + uint8_t legacy_max_head; /* %dh: max head # */ + uint8_t legacy_sectors_per_track; /* %cl[5:0]: max sector # */ + /* Int13, Fn41: Get Device Parameters (as filled into %ds:%esi). */ + /* NB. First uint16_t of buffer must be set to buffer size. */ + XEN_GUEST_HANDLE(void) edd_params; + } disk_info; /* XEN_FW_DISK_INFO */ + struct { + uint8_t device; /* bios device number */ + uint32_t mbr_signature; /* offset 0x1b8 in mbr */ + } disk_mbr_signature; /* XEN_FW_DISK_MBR_SIGNATURE */ + struct { + /* Int10, AX=4F15: Get EDID info. */ + uint8_t capabilities; + uint8_t edid_transfer_time; + /* must refer to 128-byte buffer */ + XEN_GUEST_HANDLE(uint8) edid; + } vbeddc_info; /* XEN_FW_VBEDDC_INFO */ + } u; +}; +typedef struct xenpf_firmware_info xenpf_firmware_info_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_firmware_info_t); + +#define XENPF_enter_acpi_sleep 51 +struct xenpf_enter_acpi_sleep { + /* IN variables */ + uint16_t pm1a_cnt_val; /* PM1a control value. */ + uint16_t pm1b_cnt_val; /* PM1b control value. */ + uint32_t sleep_state; /* Which state to enter (Sn). */ + uint32_t flags; /* Must be zero. */ +}; +typedef struct xenpf_enter_acpi_sleep xenpf_enter_acpi_sleep_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_enter_acpi_sleep_t); + +#define XENPF_change_freq 52 +struct xenpf_change_freq { + /* IN variables */ + uint32_t flags; /* Must be zero. */ + uint32_t cpu; /* Physical cpu. */ + uint64_t freq; /* New frequency (Hz). */ +}; +typedef struct xenpf_change_freq xenpf_change_freq_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_change_freq_t); + +/* + * Get idle times (nanoseconds since boot) for physical CPUs specified in the + * @cpumap_bitmap with range [0..@cpumap_nr_cpus-1]. The @idletime array is + * indexed by CPU number; only entries with the corresponding @cpumap_bitmap + * bit set are written to. On return, @cpumap_bitmap is modified so that any + * non-existent CPUs are cleared. Such CPUs have their @idletime array entry + * cleared. + */ +#define XENPF_getidletime 53 +struct xenpf_getidletime { + /* IN/OUT variables */ + /* IN: CPUs to interrogate; OUT: subset of IN which are present */ + XEN_GUEST_HANDLE(uint8) cpumap_bitmap; + /* IN variables */ + /* Size of cpumap bitmap. */ + uint32_t cpumap_nr_cpus; + /* Must be indexable for every cpu in cpumap_bitmap. */ + XEN_GUEST_HANDLE(uint64) idletime; + /* OUT variables */ + /* System time when the idletime snapshots were taken. */ + uint64_t now; +}; +typedef struct xenpf_getidletime xenpf_getidletime_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_getidletime_t); + +#define XENPF_set_processor_pminfo 54 + +/* ability bits */ +#define XEN_PROCESSOR_PM_CX 1 +#define XEN_PROCESSOR_PM_PX 2 +#define XEN_PROCESSOR_PM_TX 4 + +/* cmd type */ +#define XEN_PM_CX 0 +#define XEN_PM_PX 1 +#define XEN_PM_TX 2 + +/* Px sub info type */ +#define XEN_PX_PCT 1 +#define XEN_PX_PSS 2 +#define XEN_PX_PPC 4 +#define XEN_PX_PSD 8 + +struct xen_power_register { + uint32_t space_id; + uint32_t bit_width; + uint32_t bit_offset; + uint32_t access_size; + uint64_t address; +}; + +struct xen_processor_csd { + uint32_t domain; /* domain number of one dependent group */ + uint32_t coord_type; /* coordination type */ + uint32_t num; /* number of processors in same domain */ +}; +typedef struct xen_processor_csd xen_processor_csd_t; +DEFINE_XEN_GUEST_HANDLE(xen_processor_csd_t); + +struct xen_processor_cx { + struct xen_power_register reg; /* GAS for Cx trigger register */ + uint8_t type; /* cstate value, c0: 0, c1: 1, ... */ + uint32_t latency; /* worst latency (ms) to enter/exit this cstate */ + uint32_t power; /* average power consumption(mW) */ + uint32_t dpcnt; /* number of dependency entries */ + XEN_GUEST_HANDLE(xen_processor_csd_t) dp; /* NULL if no dependency */ +}; +typedef struct xen_processor_cx xen_processor_cx_t; +DEFINE_XEN_GUEST_HANDLE(xen_processor_cx_t); + +struct xen_processor_flags { + uint32_t bm_control:1; + uint32_t bm_check:1; + uint32_t has_cst:1; + uint32_t power_setup_done:1; + uint32_t bm_rld_set:1; +}; + +struct xen_processor_power { + uint32_t count; /* number of C state entries in array below */ + struct xen_processor_flags flags; /* global flags of this processor */ + XEN_GUEST_HANDLE(xen_processor_cx_t) states; /* supported c states */ +}; + +struct xen_pct_register { + uint8_t descriptor; + uint16_t length; + uint8_t space_id; + uint8_t bit_width; + uint8_t bit_offset; + uint8_t reserved; + uint64_t address; +}; + +struct xen_processor_px { + uint64_t core_frequency; /* megahertz */ + uint64_t power; /* milliWatts */ + uint64_t transition_latency; /* microseconds */ + uint64_t bus_master_latency; /* microseconds */ + uint64_t control; /* control value */ + uint64_t status; /* success indicator */ +}; +typedef struct xen_processor_px xen_processor_px_t; +DEFINE_XEN_GUEST_HANDLE(xen_processor_px_t); + +struct xen_psd_package { + uint64_t num_entries; + uint64_t revision; + uint64_t domain; + uint64_t coord_type; + uint64_t num_processors; +}; + +struct xen_processor_performance { + uint32_t flags; /* flag for Px sub info type */ + uint32_t platform_limit; /* Platform limitation on freq usage */ + struct xen_pct_register control_register; + struct xen_pct_register status_register; + uint32_t state_count; /* total available performance states */ + XEN_GUEST_HANDLE(xen_processor_px_t) states; + struct xen_psd_package domain_info; + uint32_t shared_type; /* coordination type of this processor */ +}; +typedef struct xen_processor_performance xen_processor_performance_t; +DEFINE_XEN_GUEST_HANDLE(xen_processor_performance_t); + +struct xenpf_set_processor_pminfo { + /* IN variables */ + uint32_t id; /* ACPI CPU ID */ + uint32_t type; /* {XEN_PM_CX, XEN_PM_PX} */ + union { + struct xen_processor_power power;/* Cx: _CST/_CSD */ + struct xen_processor_performance perf; /* Px: _PPC/_PCT/_PSS/_PSD */ + } u; +}; +typedef struct xenpf_set_processor_pminfo xenpf_set_processor_pminfo_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_set_processor_pminfo_t); + +#define XENPF_get_cpuinfo 55 +struct xenpf_pcpuinfo { + /* IN */ + uint32_t xen_cpuid; + /* OUT */ + /* The maxium cpu_id that is present */ + uint32_t max_present; +#define XEN_PCPU_FLAGS_ONLINE 1 + /* Correponding xen_cpuid is not present*/ +#define XEN_PCPU_FLAGS_INVALID 2 + uint32_t flags; + uint32_t apic_id; + uint32_t acpi_id; +}; +typedef struct xenpf_pcpuinfo xenpf_pcpuinfo_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_pcpuinfo_t); + +#define XENPF_cpu_online 56 +#define XENPF_cpu_offline 57 +struct xenpf_cpu_ol +{ + uint32_t cpuid; +}; +typedef struct xenpf_cpu_ol xenpf_cpu_ol_t; +DEFINE_XEN_GUEST_HANDLE(xenpf_cpu_ol_t); + +#define XENPF_cpu_hotadd 58 +struct xenpf_cpu_hotadd +{ + uint32_t apic_id; + uint32_t acpi_id; + uint32_t pxm; +}; + +#define XENPF_mem_hotadd 59 +struct xenpf_mem_hotadd +{ + uint64_t spfn; + uint64_t epfn; + uint32_t pxm; + uint32_t flags; +}; + +#define XENPF_get_cpu_freq ('N' << 24) +struct xenpf_get_cpu_freq { + /* IN variables */ + uint32_t vcpu; + /* OUT variables */ + uint32_t freq; /* in kHz */ +}; + +struct xen_platform_op { + uint32_t cmd; + uint32_t interface_version; /* XENPF_INTERFACE_VERSION */ + union { + struct xenpf_settime settime; + struct xenpf_add_memtype add_memtype; + struct xenpf_del_memtype del_memtype; + struct xenpf_read_memtype read_memtype; + struct xenpf_microcode_update microcode; + struct xenpf_platform_quirk platform_quirk; + struct xenpf_firmware_info firmware_info; + struct xenpf_enter_acpi_sleep enter_acpi_sleep; + struct xenpf_change_freq change_freq; + struct xenpf_getidletime getidletime; + struct xenpf_set_processor_pminfo set_pminfo; + struct xenpf_pcpuinfo pcpu_info; + struct xenpf_cpu_ol cpu_ol; + struct xenpf_cpu_hotadd cpu_add; + struct xenpf_mem_hotadd mem_add; + struct xenpf_get_cpu_freq get_cpu_freq; + uint8_t pad[128]; + } u; +}; +typedef struct xen_platform_op xen_platform_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_platform_op_t); + +#endif /* __XEN_PUBLIC_PLATFORM_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/sched.h b/include/xenpvdrivers/interface/sched.h new file mode 100644 index 0000000..1974b3e --- /dev/null +++ b/include/xenpvdrivers/interface/sched.h @@ -0,0 +1,113 @@ +/****************************************************************************** + * sched.h + * + * Scheduler state interactions + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2005, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_SCHED_H__ +#define __XEN_PUBLIC_SCHED_H__ + +#include "event_channel.h" + +/* + * The prototype for this hypercall is: + * long sched_op(int cmd, void *arg) + * @cmd == SCHEDOP_??? (scheduler operation). + * @arg == Operation-specific extra argument(s), as described below. + * + * Versions of Xen prior to 3.0.2 provided only the following legacy version + * of this hypercall, supporting only the commands yield, block and shutdown: + * long sched_op(int cmd, unsigned long arg) + * @cmd == SCHEDOP_??? (scheduler operation). + * @arg == 0 (SCHEDOP_yield and SCHEDOP_block) + * == SHUTDOWN_* code (SCHEDOP_shutdown) + * This legacy version is available to new guests as sched_op_compat(). + */ + +/* + * Voluntarily yield the CPU. + * @arg == NULL. + */ +#define SCHEDOP_yield 0 + +/* + * Block execution of this VCPU until an event is received for processing. + * If called with event upcalls masked, this operation will atomically + * reenable event delivery and check for pending events before blocking the + * VCPU. This avoids a "wakeup waiting" race. + * @arg == NULL. + */ +#define SCHEDOP_block 1 + +/* + * Halt execution of this domain (all VCPUs) and notify the system controller. + * @arg == pointer to sched_shutdown structure. + */ +#define SCHEDOP_shutdown 2 +struct sched_shutdown { + unsigned int reason; /* SHUTDOWN_* */ +}; +DEFINE_GUEST_HANDLE_STRUCT(sched_shutdown); +typedef struct sched_shutdown sched_shutdown_t; +DEFINE_XEN_GUEST_HANDLE(sched_shutdown_t); + +/* + * Poll a set of event-channel ports. Return when one or more are pending. An + * optional timeout may be specified. + * @arg == pointer to sched_poll structure. + */ +#define SCHEDOP_poll 3 +struct sched_poll { + XEN_GUEST_HANDLE(evtchn_port_t) ports; + unsigned int nr_ports; + uint64_t timeout; +}; +DEFINE_GUEST_HANDLE_STRUCT(sched_poll); +typedef struct sched_poll sched_poll_t; +DEFINE_XEN_GUEST_HANDLE(sched_poll_t); + +/* + * Declare a shutdown for another domain. The main use of this function is + * in interpreting shutdown requests and reasons for fully-virtualized + * domains. A para-virtualized domain may use SCHEDOP_shutdown directly. + * @arg == pointer to sched_remote_shutdown structure. + */ +#define SCHEDOP_remote_shutdown 4 +struct sched_remote_shutdown { + domid_t domain_id; /* Remote domain ID */ + unsigned int reason; /* SHUTDOWN_xxx reason */ +}; +typedef struct sched_remote_shutdown sched_remote_shutdown_t; +DEFINE_XEN_GUEST_HANDLE(sched_remote_shutdown_t); + +/* + * Reason codes for SCHEDOP_shutdown. These may be interpreted by control + * software to determine the appropriate action. For the most part, Xen does + * not care about the shutdown code. + */ +#define SHUTDOWN_poweroff 0 /* Domain exited normally. Clean up and kill. */ +#define SHUTDOWN_reboot 1 /* Clean up, kill, and then restart. */ +#define SHUTDOWN_suspend 2 /* Clean up, save suspend info, kill. */ +#define SHUTDOWN_crash 3 /* Tell controller we've crashed. */ + +#endif /* __XEN_PUBLIC_SCHED_H__ */ diff --git a/include/xenpvdrivers/interface/sysctl.h b/include/xenpvdrivers/interface/sysctl.h new file mode 100644 index 0000000..f869f9c --- /dev/null +++ b/include/xenpvdrivers/interface/sysctl.h @@ -0,0 +1,523 @@ +/****************************************************************************** + * sysctl.h + * + * System management operations. For use by node control stack. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2002-2006, K Fraser + */ + +#ifndef __XEN_PUBLIC_SYSCTL_H__ +#define __XEN_PUBLIC_SYSCTL_H__ + +#if !defined(__XEN__) && !defined(__XEN_TOOLS__) +#error "sysctl operations are intended for use by node control tools only" +#endif + +#include "xen.h" +#include "domctl.h" + +#define XEN_SYSCTL_INTERFACE_VERSION 0x00000007 + +/* + * Read console content from Xen buffer ring. + */ +#define XEN_SYSCTL_readconsole 1 +struct xen_sysctl_readconsole { + /* IN: Non-zero -> clear after reading. */ + uint8_t clear; + /* IN: Non-zero -> start index specified by @index field. */ + uint8_t incremental; + uint8_t pad0, pad1; + /* + * IN: Start index for consuming from ring buffer (if @incremental); + * OUT: End index after consuming from ring buffer. + */ + uint32_t index; + /* IN: Virtual address to write console data. */ + XEN_GUEST_HANDLE_64(char) buffer; + /* IN: Size of buffer; OUT: Bytes written to buffer. */ + uint32_t count; +}; +typedef struct xen_sysctl_readconsole xen_sysctl_readconsole_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_readconsole_t); + +/* Get trace buffers machine base address */ +#define XEN_SYSCTL_tbuf_op 2 +struct xen_sysctl_tbuf_op { + /* IN variables */ +#define XEN_SYSCTL_TBUFOP_get_info 0 +#define XEN_SYSCTL_TBUFOP_set_cpu_mask 1 +#define XEN_SYSCTL_TBUFOP_set_evt_mask 2 +#define XEN_SYSCTL_TBUFOP_set_size 3 +#define XEN_SYSCTL_TBUFOP_enable 4 +#define XEN_SYSCTL_TBUFOP_disable 5 + uint32_t cmd; + /* IN/OUT variables */ + struct xenctl_cpumap cpu_mask; + uint32_t evt_mask; + /* OUT variables */ + uint64_aligned_t buffer_mfn; + uint32_t size; +}; +typedef struct xen_sysctl_tbuf_op xen_sysctl_tbuf_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_tbuf_op_t); + +/* + * Get physical information about the host machine + */ +#define XEN_SYSCTL_physinfo 3 + /* (x86) The platform supports HVM guests. */ +#define _XEN_SYSCTL_PHYSCAP_hvm 0 +#define XEN_SYSCTL_PHYSCAP_hvm (1u<<_XEN_SYSCTL_PHYSCAP_hvm) + /* (x86) The platform supports HVM-guest direct access to I/O devices. */ +#define _XEN_SYSCTL_PHYSCAP_hvm_directio 1 +#define XEN_SYSCTL_PHYSCAP_hvm_directio (1u<<_XEN_SYSCTL_PHYSCAP_hvm_directio) +struct xen_sysctl_physinfo { + uint32_t threads_per_core; + uint32_t cores_per_socket; + uint32_t nr_cpus; + uint32_t max_node_id; + uint32_t cpu_khz; + uint64_aligned_t total_pages; + uint64_aligned_t free_pages; + uint64_aligned_t scrub_pages; + uint32_t hw_cap[8]; + + /* + * IN: maximum addressable entry in the caller-provided cpu_to_node array. + * OUT: largest cpu identifier in the system. + * If OUT is greater than IN then the cpu_to_node array is truncated! + */ + uint32_t max_cpu_id; + /* + * If not NULL, this array is filled with node identifier for each cpu. + * If a cpu has no node information (e.g., cpu not present) then the + * sentinel value ~0u is written. + * The size of this array is specified by the caller in @max_cpu_id. + * If the actual @max_cpu_id is smaller than the array then the trailing + * elements of the array will not be written by the sysctl. + */ + XEN_GUEST_HANDLE_64(uint32) cpu_to_node; + + /* XEN_SYSCTL_PHYSCAP_??? */ + uint32_t capabilities; +}; +typedef struct xen_sysctl_physinfo xen_sysctl_physinfo_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_physinfo_t); + +/* + * Get the ID of the current scheduler. + */ +#define XEN_SYSCTL_sched_id 4 +struct xen_sysctl_sched_id { + /* OUT variable */ + uint32_t sched_id; +}; +typedef struct xen_sysctl_sched_id xen_sysctl_sched_id_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_sched_id_t); + +/* Interface for controlling Xen software performance counters. */ +#define XEN_SYSCTL_perfc_op 5 +/* Sub-operations: */ +#define XEN_SYSCTL_PERFCOP_reset 1 /* Reset all counters to zero. */ +#define XEN_SYSCTL_PERFCOP_query 2 /* Get perfctr information. */ +struct xen_sysctl_perfc_desc { + char name[80]; /* name of perf counter */ + uint32_t nr_vals; /* number of values for this counter */ +}; +typedef struct xen_sysctl_perfc_desc xen_sysctl_perfc_desc_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_desc_t); +typedef uint32_t xen_sysctl_perfc_val_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_val_t); + +struct xen_sysctl_perfc_op { + /* IN variables. */ + uint32_t cmd; /* XEN_SYSCTL_PERFCOP_??? */ + /* OUT variables. */ + uint32_t nr_counters; /* number of counters description */ + uint32_t nr_vals; /* number of values */ + /* counter information (or NULL) */ + XEN_GUEST_HANDLE_64(xen_sysctl_perfc_desc_t) desc; + /* counter values (or NULL) */ + XEN_GUEST_HANDLE_64(xen_sysctl_perfc_val_t) val; +}; +typedef struct xen_sysctl_perfc_op xen_sysctl_perfc_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_op_t); + +#define XEN_SYSCTL_getdomaininfolist 6 +struct xen_sysctl_getdomaininfolist { + /* IN variables. */ + domid_t first_domain; + uint32_t max_domains; + XEN_GUEST_HANDLE_64(xen_domctl_getdomaininfo_t) buffer; + /* OUT variables. */ + uint32_t num_domains; +}; +typedef struct xen_sysctl_getdomaininfolist xen_sysctl_getdomaininfolist_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_getdomaininfolist_t); + +/* Inject debug keys into Xen. */ +#define XEN_SYSCTL_debug_keys 7 +struct xen_sysctl_debug_keys { + /* IN variables. */ + XEN_GUEST_HANDLE_64(char) keys; + uint32_t nr_keys; +}; +typedef struct xen_sysctl_debug_keys xen_sysctl_debug_keys_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_debug_keys_t); + +/* Get physical CPU information. */ +#define XEN_SYSCTL_getcpuinfo 8 +struct xen_sysctl_cpuinfo { + uint64_aligned_t idletime; +}; +typedef struct xen_sysctl_cpuinfo xen_sysctl_cpuinfo_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_cpuinfo_t); +struct xen_sysctl_getcpuinfo { + /* IN variables. */ + uint32_t max_cpus; + XEN_GUEST_HANDLE_64(xen_sysctl_cpuinfo_t) info; + /* OUT variables. */ + uint32_t nr_cpus; +}; +typedef struct xen_sysctl_getcpuinfo xen_sysctl_getcpuinfo_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_getcpuinfo_t); + +#define XEN_SYSCTL_availheap 9 +struct xen_sysctl_availheap { + /* IN variables. */ + uint32_t min_bitwidth; /* Smallest address width (zero if don't care). */ + uint32_t max_bitwidth; /* Largest address width (zero if don't care). */ + int32_t node; /* NUMA node of interest (-1 for all nodes). */ + /* OUT variables. */ + uint64_aligned_t avail_bytes;/* Bytes available in the specified region. */ +}; +typedef struct xen_sysctl_availheap xen_sysctl_availheap_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_availheap_t); + +#define XEN_SYSCTL_get_pmstat 10 +struct pm_px_val { + uint64_aligned_t freq; /* Px core frequency */ + uint64_aligned_t residency; /* Px residency time */ + uint64_aligned_t count; /* Px transition count */ +}; +typedef struct pm_px_val pm_px_val_t; +DEFINE_XEN_GUEST_HANDLE(pm_px_val_t); + +struct pm_px_stat { + uint8_t total; /* total Px states */ + uint8_t usable; /* usable Px states */ + uint8_t last; /* last Px state */ + uint8_t cur; /* current Px state */ + XEN_GUEST_HANDLE_64(uint64) trans_pt; /* Px transition table */ + XEN_GUEST_HANDLE_64(pm_px_val_t) pt; +}; +typedef struct pm_px_stat pm_px_stat_t; +DEFINE_XEN_GUEST_HANDLE(pm_px_stat_t); + +struct pm_cx_stat { + uint32_t nr; /* entry nr in triggers & residencies, including C0 */ + uint32_t last; /* last Cx state */ + uint64_aligned_t idle_time; /* idle time from boot */ + XEN_GUEST_HANDLE_64(uint64) triggers; /* Cx trigger counts */ + XEN_GUEST_HANDLE_64(uint64) residencies; /* Cx residencies */ +}; + +struct xen_sysctl_get_pmstat { +#define PMSTAT_CATEGORY_MASK 0xf0 +#define PMSTAT_PX 0x10 +#define PMSTAT_CX 0x20 +#define PMSTAT_get_max_px (PMSTAT_PX | 0x1) +#define PMSTAT_get_pxstat (PMSTAT_PX | 0x2) +#define PMSTAT_reset_pxstat (PMSTAT_PX | 0x3) +#define PMSTAT_get_max_cx (PMSTAT_CX | 0x1) +#define PMSTAT_get_cxstat (PMSTAT_CX | 0x2) +#define PMSTAT_reset_cxstat (PMSTAT_CX | 0x3) + uint32_t type; + uint32_t cpuid; + union { + struct pm_px_stat getpx; + struct pm_cx_stat getcx; + /* other struct for tx, etc */ + } u; +}; +typedef struct xen_sysctl_get_pmstat xen_sysctl_get_pmstat_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_get_pmstat_t); + +/* + * Status codes. Must be greater than 0 to avoid confusing + * sysctl callers that see 0 as a plain successful return. + */ +#define XEN_CPU_HOTPLUG_STATUS_OFFLINE 1 +#define XEN_CPU_HOTPLUG_STATUS_ONLINE 2 +#define XEN_CPU_HOTPLUG_STATUS_NEW 3 + +#define XEN_SYSCTL_cpu_hotplug 11 +struct xen_sysctl_cpu_hotplug { + /* IN variables */ + uint32_t cpu; /* Physical cpu. */ +#define XEN_SYSCTL_CPU_HOTPLUG_ONLINE 0 +#define XEN_SYSCTL_CPU_HOTPLUG_OFFLINE 1 +#define XEN_SYSCTL_CPU_HOTPLUG_STATUS 2 + uint32_t op; /* hotplug opcode */ +}; +typedef struct xen_sysctl_cpu_hotplug xen_sysctl_cpu_hotplug_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_cpu_hotplug_t); + +/* + * Get/set xen power management, include + * 1. cpufreq governors and related parameters + */ +#define XEN_SYSCTL_pm_op 12 +struct xen_userspace { + uint32_t scaling_setspeed; +}; +typedef struct xen_userspace xen_userspace_t; + +struct xen_ondemand { + uint32_t sampling_rate_max; + uint32_t sampling_rate_min; + + uint32_t sampling_rate; + uint32_t up_threshold; +}; +typedef struct xen_ondemand xen_ondemand_t; + +/* + * cpufreq para name of this structure named + * same as sysfs file name of native linux + */ +#define CPUFREQ_NAME_LEN 16 +struct xen_get_cpufreq_para { + /* IN/OUT variable */ + uint32_t cpu_num; + uint32_t freq_num; + uint32_t gov_num; + + /* for all governors */ + /* OUT variable */ + XEN_GUEST_HANDLE_64(uint32) affected_cpus; + XEN_GUEST_HANDLE_64(uint32) scaling_available_frequencies; + XEN_GUEST_HANDLE_64(char) scaling_available_governors; + char scaling_driver[CPUFREQ_NAME_LEN]; + + uint32_t cpuinfo_cur_freq; + uint32_t cpuinfo_max_freq; + uint32_t cpuinfo_min_freq; + uint32_t scaling_cur_freq; + + char scaling_governor[CPUFREQ_NAME_LEN]; + uint32_t scaling_max_freq; + uint32_t scaling_min_freq; + + /* for specific governor */ + union { + struct xen_userspace userspace; + struct xen_ondemand ondemand; + } u; +}; + +struct xen_set_cpufreq_gov { + char scaling_governor[CPUFREQ_NAME_LEN]; +}; + +struct xen_set_cpufreq_para { + #define SCALING_MAX_FREQ 1 + #define SCALING_MIN_FREQ 2 + #define SCALING_SETSPEED 3 + #define SAMPLING_RATE 4 + #define UP_THRESHOLD 5 + + uint32_t ctrl_type; + uint32_t ctrl_value; +}; + +/* Get physical CPU topology information. */ +#define INVALID_TOPOLOGY_ID (~0U) +struct xen_get_cputopo { + /* IN: maximum addressable entry in + * the caller-provided cpu_to_core/socket. + */ + uint32_t max_cpus; + XEN_GUEST_HANDLE_64(uint32) cpu_to_core; + XEN_GUEST_HANDLE_64(uint32) cpu_to_socket; + + /* OUT: number of cpus returned + * If OUT is greater than IN then the cpu_to_core/socket is truncated! + */ + uint32_t nr_cpus; +}; + +struct xen_sysctl_pm_op { + #define PM_PARA_CATEGORY_MASK 0xf0 + #define CPUFREQ_PARA 0x10 + + /* cpufreq command type */ + #define GET_CPUFREQ_PARA (CPUFREQ_PARA | 0x01) + #define SET_CPUFREQ_GOV (CPUFREQ_PARA | 0x02) + #define SET_CPUFREQ_PARA (CPUFREQ_PARA | 0x03) + #define GET_CPUFREQ_AVGFREQ (CPUFREQ_PARA | 0x04) + + /* get CPU topology */ + #define XEN_SYSCTL_pm_op_get_cputopo 0x20 + + /* set/reset scheduler power saving option */ + #define XEN_SYSCTL_pm_op_set_sched_opt_smt 0x21 + + /* cpuidle max_cstate access command */ + #define XEN_SYSCTL_pm_op_get_max_cstate 0x22 + #define XEN_SYSCTL_pm_op_set_max_cstate 0x23 + + /* set scheduler migration cost value */ + #define XEN_SYSCTL_pm_op_set_vcpu_migration_delay 0x24 + #define XEN_SYSCTL_pm_op_get_vcpu_migration_delay 0x25 + + uint32_t cmd; + uint32_t cpuid; + union { + struct xen_get_cpufreq_para get_para; + struct xen_set_cpufreq_gov set_gov; + struct xen_set_cpufreq_para set_para; + uint64_aligned_t get_avgfreq; + struct xen_get_cputopo get_topo; + uint32_t set_sched_opt_smt; + uint32_t get_max_cstate; + uint32_t set_max_cstate; + uint32_t get_vcpu_migration_delay; + uint32_t set_vcpu_migration_delay; + } u; +}; + +#define XEN_SYSCTL_page_offline_op 14 +struct xen_sysctl_page_offline_op { + /* IN: range of page to be offlined */ +#define sysctl_page_offline 1 +#define sysctl_page_online 2 +#define sysctl_query_page_offline 3 + uint32_t cmd; + uint32_t start; + uint32_t end; + /* OUT: result of page offline request */ + /* + * bit 0~15: result flags + * bit 16~31: owner + */ + XEN_GUEST_HANDLE(uint32) status; +}; + +#define PG_OFFLINE_STATUS_MASK (0xFFUL) + +/* The result is invalid, i.e. HV does not handle it */ +#define PG_OFFLINE_INVALID (0x1UL << 0) + +#define PG_OFFLINE_OFFLINED (0x1UL << 1) +#define PG_OFFLINE_PENDING (0x1UL << 2) +#define PG_OFFLINE_FAILED (0x1UL << 3) + +#define PG_ONLINE_FAILED PG_OFFLINE_FAILED +#define PG_ONLINE_ONLINED PG_OFFLINE_OFFLINED + +#define PG_OFFLINE_STATUS_OFFLINED (0x1UL << 1) +#define PG_OFFLINE_STATUS_ONLINE (0x1UL << 2) +#define PG_OFFLINE_STATUS_OFFLINE_PENDING (0x1UL << 3) +#define PG_OFFLINE_STATUS_BROKEN (0x1UL << 4) + +#define PG_OFFLINE_MISC_MASK (0xFFUL << 4) + +/* only valid when PG_OFFLINE_FAILED */ +#define PG_OFFLINE_XENPAGE (0x1UL << 8) +#define PG_OFFLINE_DOM0PAGE (0x1UL << 9) +#define PG_OFFLINE_ANONYMOUS (0x1UL << 10) +#define PG_OFFLINE_NOT_CONV_RAM (0x1UL << 11) +#define PG_OFFLINE_OWNED (0x1UL << 12) + +#define PG_OFFLINE_BROKEN (0x1UL << 13) +#define PG_ONLINE_BROKEN PG_OFFLINE_BROKEN + +#define PG_OFFLINE_OWNER_SHIFT 16 + +#define XEN_SYSCTL_lockprof_op 15 +/* Sub-operations: */ +#define XEN_SYSCTL_LOCKPROF_reset 1 /* Reset all profile data to zero. */ +#define XEN_SYSCTL_LOCKPROF_query 2 /* Get lock profile information. */ +/* Record-type: */ +#define LOCKPROF_TYPE_GLOBAL 0 /* global lock, idx meaningless */ +#define LOCKPROF_TYPE_PERDOM 1 /* per-domain lock, idx is domid */ +#define LOCKPROF_TYPE_N 2 /* number of types */ +struct xen_sysctl_lockprof_data { + char name[40]; /* lock name (may include up to 2 %d specifiers) */ + int32_t type; /* LOCKPROF_TYPE_??? */ + int32_t idx; /* index (e.g. domain id) */ + uint64_aligned_t lock_cnt; /* # of locking succeeded */ + uint64_aligned_t block_cnt; /* # of wait for lock */ + uint64_aligned_t lock_time; /* nsecs lock held */ + uint64_aligned_t block_time; /* nsecs waited for lock */ +}; +typedef struct xen_sysctl_lockprof_data xen_sysctl_lockprof_data_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_lockprof_data_t); +struct xen_sysctl_lockprof_op { + /* IN variables. */ + uint32_t cmd; /* XEN_SYSCTL_LOCKPROF_??? */ + uint32_t max_elem; /* size of output buffer */ + /* OUT variables (query only). */ + uint32_t nr_elem; /* number of elements available */ + uint64_aligned_t time; /* nsecs of profile measurement */ + /* profile information (or NULL) */ + XEN_GUEST_HANDLE_64(xen_sysctl_lockprof_data_t) data; +}; +typedef struct xen_sysctl_lockprof_op xen_sysctl_lockprof_op_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_lockprof_op_t); + +struct xen_sysctl { + uint32_t cmd; + uint32_t interface_version; /* XEN_SYSCTL_INTERFACE_VERSION */ + union { + struct xen_sysctl_readconsole readconsole; + struct xen_sysctl_tbuf_op tbuf_op; + struct xen_sysctl_physinfo physinfo; + struct xen_sysctl_sched_id sched_id; + struct xen_sysctl_perfc_op perfc_op; + struct xen_sysctl_getdomaininfolist getdomaininfolist; + struct xen_sysctl_debug_keys debug_keys; + struct xen_sysctl_getcpuinfo getcpuinfo; + struct xen_sysctl_availheap availheap; + struct xen_sysctl_get_pmstat get_pmstat; + struct xen_sysctl_cpu_hotplug cpu_hotplug; + struct xen_sysctl_pm_op pm_op; + struct xen_sysctl_page_offline_op page_offline; + struct xen_sysctl_lockprof_op lockprof_op; + uint8_t pad[128]; + } u; +}; +typedef struct xen_sysctl xen_sysctl_t; +DEFINE_XEN_GUEST_HANDLE(xen_sysctl_t); + +#endif /* __XEN_PUBLIC_SYSCTL_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/tmem.h b/include/xenpvdrivers/interface/tmem.h new file mode 100644 index 0000000..0019085 --- /dev/null +++ b/include/xenpvdrivers/interface/tmem.h @@ -0,0 +1,144 @@ +/****************************************************************************** + * tmem.h + * + * Guest OS interface to Xen Transcendent Memory. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2004, K A Fraser + */ + +#ifndef __XEN_PUBLIC_TMEM_H__ +#define __XEN_PUBLIC_TMEM_H__ + +#include "xen.h" + +/* Commands to HYPERVISOR_tmem_op() */ +#define TMEM_CONTROL 0 +#define TMEM_NEW_POOL 1 +#define TMEM_DESTROY_POOL 2 +#define TMEM_NEW_PAGE 3 +#define TMEM_PUT_PAGE 4 +#define TMEM_GET_PAGE 5 +#define TMEM_FLUSH_PAGE 6 +#define TMEM_FLUSH_OBJECT 7 +#define TMEM_READ 8 +#define TMEM_WRITE 9 +#define TMEM_XCHG 10 + +/* Privileged commands to HYPERVISOR_tmem_op() */ +#define TMEM_AUTH 101 +#define TMEM_RESTORE_NEW 102 + +/* Subops for HYPERVISOR_tmem_op(TMEM_CONTROL) */ +#define TMEMC_THAW 0 +#define TMEMC_FREEZE 1 +#define TMEMC_FLUSH 2 +#define TMEMC_DESTROY 3 +#define TMEMC_LIST 4 +#define TMEMC_SET_WEIGHT 5 +#define TMEMC_SET_CAP 6 +#define TMEMC_SET_COMPRESS 7 +#define TMEMC_QUERY_FREEABLE_MB 8 +#define TMEMC_SAVE_BEGIN 10 +#define TMEMC_SAVE_GET_VERSION 11 +#define TMEMC_SAVE_GET_MAXPOOLS 12 +#define TMEMC_SAVE_GET_CLIENT_WEIGHT 13 +#define TMEMC_SAVE_GET_CLIENT_CAP 14 +#define TMEMC_SAVE_GET_CLIENT_FLAGS 15 +#define TMEMC_SAVE_GET_POOL_FLAGS 16 +#define TMEMC_SAVE_GET_POOL_NPAGES 17 +#define TMEMC_SAVE_GET_POOL_UUID 18 +#define TMEMC_SAVE_GET_NEXT_PAGE 19 +#define TMEMC_SAVE_GET_NEXT_INV 20 +#define TMEMC_SAVE_END 21 +#define TMEMC_RESTORE_BEGIN 30 +#define TMEMC_RESTORE_PUT_PAGE 32 +#define TMEMC_RESTORE_FLUSH_PAGE 33 + +/* Bits for HYPERVISOR_tmem_op(TMEM_NEW_POOL) */ +#define TMEM_POOL_PERSIST 1 +#define TMEM_POOL_SHARED 2 +#define TMEM_POOL_PAGESIZE_SHIFT 4 +#define TMEM_POOL_PAGESIZE_MASK 0xf +#define TMEM_POOL_VERSION_SHIFT 24 +#define TMEM_POOL_VERSION_MASK 0xff + +/* Bits for client flags (save/restore) */ +#define TMEM_CLIENT_COMPRESS 1 +#define TMEM_CLIENT_FROZEN 2 + +/* Special errno values */ +#define EFROZEN 1000 +#define EEMPTY 1001 + + +#ifndef __ASSEMBLY__ +typedef xen_pfn_t tmem_cli_mfn_t; +typedef XEN_GUEST_HANDLE(char) tmem_cli_va_t; +struct tmem_op { + uint32_t cmd; + int32_t pool_id; + union { + struct { + uint64_t uuid[2]; + uint32_t flags; + uint32_t arg1; + } new; /* for cmd == TMEM_NEW_POOL, TMEM_AUTH, TMEM_RESTORE_NEW */ + struct { + uint32_t subop; + uint32_t cli_id; + uint32_t arg1; + uint32_t arg2; + uint64_t arg3; + tmem_cli_va_t buf; + } ctrl; /* for cmd == TMEM_CONTROL */ + struct { + + uint64_t object; + uint32_t index; + uint32_t tmem_offset; + uint32_t pfn_offset; + uint32_t len; + tmem_cli_mfn_t cmfn; /* client machine page frame */ + } gen; /* for all other cmd ("generic") */ + } u; +}; +typedef struct tmem_op tmem_op_t; +DEFINE_XEN_GUEST_HANDLE(tmem_op_t); + +struct tmem_handle { + uint32_t pool_id; + uint32_t index; + uint64_t oid; +}; + +#endif + +#endif /* __XEN_PUBLIC_TMEM_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/trace.h b/include/xenpvdrivers/interface/trace.h new file mode 100644 index 0000000..6eb125a --- /dev/null +++ b/include/xenpvdrivers/interface/trace.h @@ -0,0 +1,208 @@ +/****************************************************************************** + * include/public/trace.h + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Mark Williamson, (C) 2004 Intel Research Cambridge + * Copyright (C) 2005 Bin Ren + */ + +#ifndef __XEN_PUBLIC_TRACE_H__ +#define __XEN_PUBLIC_TRACE_H__ + +#define TRACE_EXTRA_MAX 7 +#define TRACE_EXTRA_SHIFT 28 + +/* Trace classes */ +#define TRC_CLS_SHIFT 16 +#define TRC_GEN 0x0001f000 /* General trace */ +#define TRC_SCHED 0x0002f000 /* Xen Scheduler trace */ +#define TRC_DOM0OP 0x0004f000 /* Xen DOM0 operation trace */ +#define TRC_HVM 0x0008f000 /* Xen HVM trace */ +#define TRC_MEM 0x0010f000 /* Xen memory trace */ +#define TRC_PV 0x0020f000 /* Xen PV traces */ +#define TRC_SHADOW 0x0040f000 /* Xen shadow tracing */ +#define TRC_PM 0x0080f000 /* Xen power management trace */ +#define TRC_ALL 0x0ffff000 +#define TRC_HD_TO_EVENT(x) ((x)&0x0fffffff) +#define TRC_HD_CYCLE_FLAG (1UL<<31) +#define TRC_HD_INCLUDES_CYCLE_COUNT(x) ( !!( (x) & TRC_HD_CYCLE_FLAG ) ) +#define TRC_HD_EXTRA(x) (((x)>>TRACE_EXTRA_SHIFT)&TRACE_EXTRA_MAX) + +/* Trace subclasses */ +#define TRC_SUBCLS_SHIFT 12 + +/* trace subclasses for SVM */ +#define TRC_HVM_ENTRYEXIT 0x00081000 /* VMENTRY and #VMEXIT */ +#define TRC_HVM_HANDLER 0x00082000 /* various HVM handlers */ + +#define TRC_SCHED_MIN 0x00021000 /* Just runstate changes */ +#define TRC_SCHED_VERBOSE 0x00028000 /* More inclusive scheduling */ + +/* Trace events per class */ +#define TRC_LOST_RECORDS (TRC_GEN + 1) +#define TRC_TRACE_WRAP_BUFFER (TRC_GEN + 2) +#define TRC_TRACE_CPU_CHANGE (TRC_GEN + 3) +#define TRC_TRACE_IRQ (TRC_GEN + 4) + +#define TRC_SCHED_RUNSTATE_CHANGE (TRC_SCHED_MIN + 1) +#define TRC_SCHED_CONTINUE_RUNNING (TRC_SCHED_MIN + 2) +#define TRC_SCHED_DOM_ADD (TRC_SCHED_VERBOSE + 1) +#define TRC_SCHED_DOM_REM (TRC_SCHED_VERBOSE + 2) +#define TRC_SCHED_SLEEP (TRC_SCHED_VERBOSE + 3) +#define TRC_SCHED_WAKE (TRC_SCHED_VERBOSE + 4) +#define TRC_SCHED_YIELD (TRC_SCHED_VERBOSE + 5) +#define TRC_SCHED_BLOCK (TRC_SCHED_VERBOSE + 6) +#define TRC_SCHED_SHUTDOWN (TRC_SCHED_VERBOSE + 7) +#define TRC_SCHED_CTL (TRC_SCHED_VERBOSE + 8) +#define TRC_SCHED_ADJDOM (TRC_SCHED_VERBOSE + 9) +#define TRC_SCHED_SWITCH (TRC_SCHED_VERBOSE + 10) +#define TRC_SCHED_S_TIMER_FN (TRC_SCHED_VERBOSE + 11) +#define TRC_SCHED_T_TIMER_FN (TRC_SCHED_VERBOSE + 12) +#define TRC_SCHED_DOM_TIMER_FN (TRC_SCHED_VERBOSE + 13) +#define TRC_SCHED_SWITCH_INFPREV (TRC_SCHED_VERBOSE + 14) +#define TRC_SCHED_SWITCH_INFNEXT (TRC_SCHED_VERBOSE + 15) + +#define TRC_MEM_PAGE_GRANT_MAP (TRC_MEM + 1) +#define TRC_MEM_PAGE_GRANT_UNMAP (TRC_MEM + 2) +#define TRC_MEM_PAGE_GRANT_TRANSFER (TRC_MEM + 3) + +#define TRC_PV_HYPERCALL (TRC_PV + 1) +#define TRC_PV_TRAP (TRC_PV + 3) +#define TRC_PV_PAGE_FAULT (TRC_PV + 4) +#define TRC_PV_FORCED_INVALID_OP (TRC_PV + 5) +#define TRC_PV_EMULATE_PRIVOP (TRC_PV + 6) +#define TRC_PV_EMULATE_4GB (TRC_PV + 7) +#define TRC_PV_MATH_STATE_RESTORE (TRC_PV + 8) +#define TRC_PV_PAGING_FIXUP (TRC_PV + 9) +#define TRC_PV_GDT_LDT_MAPPING_FAULT (TRC_PV + 10) +#define TRC_PV_PTWR_EMULATION (TRC_PV + 11) +#define TRC_PV_PTWR_EMULATION_PAE (TRC_PV + 12) + /* Indicates that addresses in trace record are 64 bits */ +#define TRC_64_FLAG (0x100) + +#define TRC_SHADOW_NOT_SHADOW (TRC_SHADOW + 1) +#define TRC_SHADOW_FAST_PROPAGATE (TRC_SHADOW + 2) +#define TRC_SHADOW_FAST_MMIO (TRC_SHADOW + 3) +#define TRC_SHADOW_FALSE_FAST_PATH (TRC_SHADOW + 4) +#define TRC_SHADOW_MMIO (TRC_SHADOW + 5) +#define TRC_SHADOW_FIXUP (TRC_SHADOW + 6) +#define TRC_SHADOW_DOMF_DYING (TRC_SHADOW + 7) +#define TRC_SHADOW_EMULATE (TRC_SHADOW + 8) +#define TRC_SHADOW_EMULATE_UNSHADOW_USER (TRC_SHADOW + 9) +#define TRC_SHADOW_EMULATE_UNSHADOW_EVTINJ (TRC_SHADOW + 10) +#define TRC_SHADOW_EMULATE_UNSHADOW_UNHANDLED (TRC_SHADOW + 11) +#define TRC_SHADOW_WRMAP_BF (TRC_SHADOW + 12) +#define TRC_SHADOW_PREALLOC_UNPIN (TRC_SHADOW + 13) +#define TRC_SHADOW_RESYNC_FULL (TRC_SHADOW + 14) +#define TRC_SHADOW_RESYNC_ONLY (TRC_SHADOW + 15) + +/* trace events per subclass */ +#define TRC_HVM_VMENTRY (TRC_HVM_ENTRYEXIT + 0x01) +#define TRC_HVM_VMEXIT (TRC_HVM_ENTRYEXIT + 0x02) +#define TRC_HVM_VMEXIT64 (TRC_HVM_ENTRYEXIT + TRC_64_FLAG + 0x02) +#define TRC_HVM_PF_XEN (TRC_HVM_HANDLER + 0x01) +#define TRC_HVM_PF_XEN64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x01) +#define TRC_HVM_PF_INJECT (TRC_HVM_HANDLER + 0x02) +#define TRC_HVM_PF_INJECT64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x02) +#define TRC_HVM_INJ_EXC (TRC_HVM_HANDLER + 0x03) +#define TRC_HVM_INJ_VIRQ (TRC_HVM_HANDLER + 0x04) +#define TRC_HVM_REINJ_VIRQ (TRC_HVM_HANDLER + 0x05) +#define TRC_HVM_IO_READ (TRC_HVM_HANDLER + 0x06) +#define TRC_HVM_IO_WRITE (TRC_HVM_HANDLER + 0x07) +#define TRC_HVM_CR_READ (TRC_HVM_HANDLER + 0x08) +#define TRC_HVM_CR_READ64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x08) +#define TRC_HVM_CR_WRITE (TRC_HVM_HANDLER + 0x09) +#define TRC_HVM_CR_WRITE64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x09) +#define TRC_HVM_DR_READ (TRC_HVM_HANDLER + 0x0A) +#define TRC_HVM_DR_WRITE (TRC_HVM_HANDLER + 0x0B) +#define TRC_HVM_MSR_READ (TRC_HVM_HANDLER + 0x0C) +#define TRC_HVM_MSR_WRITE (TRC_HVM_HANDLER + 0x0D) +#define TRC_HVM_CPUID (TRC_HVM_HANDLER + 0x0E) +#define TRC_HVM_INTR (TRC_HVM_HANDLER + 0x0F) +#define TRC_HVM_NMI (TRC_HVM_HANDLER + 0x10) +#define TRC_HVM_SMI (TRC_HVM_HANDLER + 0x11) +#define TRC_HVM_VMMCALL (TRC_HVM_HANDLER + 0x12) +#define TRC_HVM_HLT (TRC_HVM_HANDLER + 0x13) +#define TRC_HVM_INVLPG (TRC_HVM_HANDLER + 0x14) +#define TRC_HVM_INVLPG64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x14) +#define TRC_HVM_MCE (TRC_HVM_HANDLER + 0x15) +#define TRC_HVM_IOPORT_READ (TRC_HVM_HANDLER + 0x16) +#define TRC_HVM_IOMEM_READ (TRC_HVM_HANDLER + 0x17) +#define TRC_HVM_CLTS (TRC_HVM_HANDLER + 0x18) +#define TRC_HVM_LMSW (TRC_HVM_HANDLER + 0x19) +#define TRC_HVM_LMSW64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x19) +#define TRC_HVM_INTR_WINDOW (TRC_HVM_HANDLER + 0x20) +#define TRC_HVM_IOPORT_WRITE (TRC_HVM_HANDLER + 0x216) +#define TRC_HVM_IOMEM_WRITE (TRC_HVM_HANDLER + 0x217) + +/* trace subclasses for power management */ +#define TRC_PM_FREQ 0x00801000 /* xen cpu freq events */ +#define TRC_PM_IDLE 0x00802000 /* xen cpu idle events */ + +/* trace events for per class */ +#define TRC_PM_FREQ_CHANGE (TRC_PM_FREQ + 0x01) +#define TRC_PM_IDLE_ENTRY (TRC_PM_IDLE + 0x01) +#define TRC_PM_IDLE_EXIT (TRC_PM_IDLE + 0x02) + +/* This structure represents a single trace buffer record. */ +struct t_rec { + uint32_t event:28; + uint32_t extra_u32:3; /* # entries in trailing extra_u32[] array */ + uint32_t cycles_included:1; /* u.cycles or u.no_cycles? */ + union { + struct { + uint32_t cycles_lo, cycles_hi; /* cycle counter timestamp */ + uint32_t extra_u32[7]; /* event data items */ + } cycles; + struct { + uint32_t extra_u32[7]; /* event data items */ + } nocycles; + } u; +}; + +/* + * This structure contains the metadata for a single trace buffer. The head + * field, indexes into an array of struct t_rec's. + */ +struct t_buf { + /* Assume the data buffer size is X. X is generally not a power of 2. + * CONS and PROD are incremented modulo (2*X): + * 0 <= cons < 2*X + * 0 <= prod < 2*X + * This is done because addition modulo X breaks at 2^32 when X is not a + * power of 2: + * (((2^32 - 1) % X) + 1) % X != (2^32) % X + */ + uint32_t cons; /* Offset of next item to be consumed by control tools. */ + uint32_t prod; /* Offset of next item to be produced by Xen. */ + /* Records follow immediately after the meta-data header. */ +}; + +#endif /* __XEN_PUBLIC_TRACE_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/vcpu.h b/include/xenpvdrivers/interface/vcpu.h new file mode 100644 index 0000000..131ddc7 --- /dev/null +++ b/include/xenpvdrivers/interface/vcpu.h @@ -0,0 +1,237 @@ +/****************************************************************************** + * vcpu.h + * + * VCPU initialisation, query, and hotplug. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2005, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_VCPU_H__ +#define __XEN_PUBLIC_VCPU_H__ + +#include "xen.h" + +/* + * Prototype for this hypercall is: + * int vcpu_op(int cmd, int vcpuid, void *extra_args) + * @cmd == VCPUOP_??? (VCPU operation). + * @vcpuid == VCPU to operate on. + * @extra_args == Operation-specific extra arguments (NULL if none). + */ + +/* + * Initialise a VCPU. Each VCPU can be initialised only once. A + * newly-initialised VCPU will not run until it is brought up by VCPUOP_up. + * + * @extra_arg == pointer to vcpu_guest_context structure containing initial + * state for the VCPU. + */ +#define VCPUOP_initialise 0 + +/* + * Bring up a VCPU. This makes the VCPU runnable. This operation will fail + * if the VCPU has not been initialised (VCPUOP_initialise). + */ +#define VCPUOP_up 1 + +/* + * Bring down a VCPU (i.e., make it non-runnable). + * There are a few caveats that callers should observe: + * 1. This operation may return, and VCPU_is_up may return false, before the + * VCPU stops running (i.e., the command is asynchronous). It is a good + * idea to ensure that the VCPU has entered a non-critical loop before + * bringing it down. Alternatively, this operation is guaranteed + * synchronous if invoked by the VCPU itself. + * 2. After a VCPU is initialised, there is currently no way to drop all its + * references to domain memory. Even a VCPU that is down still holds + * memory references via its pagetable base pointer and GDT. It is good + * practise to move a VCPU onto an 'idle' or default page table, LDT and + * GDT before bringing it down. + */ +#define VCPUOP_down 2 + +/* Returns 1 if the given VCPU is up. */ +#define VCPUOP_is_up 3 + +/* + * Return information about the state and running time of a VCPU. + * @extra_arg == pointer to vcpu_runstate_info structure. + */ +#define VCPUOP_get_runstate_info 4 +struct vcpu_runstate_info { + /* VCPU's current state (RUNSTATE_*). */ + int state; + /* When was current state entered (system time, ns)? */ + uint64_t state_entry_time; + /* + * Time spent in each RUNSTATE_* (ns). The sum of these times is + * guaranteed not to drift from system time. + */ + uint64_t time[4]; +}; +DEFINE_GUEST_HANDLE_STRUCT(vcpu_runstate_info); +typedef struct vcpu_runstate_info vcpu_runstate_info_t; +DEFINE_XEN_GUEST_HANDLE(vcpu_runstate_info_t); + +/* VCPU is currently running on a physical CPU. */ +#define RUNSTATE_running 0 + +/* VCPU is runnable, but not currently scheduled on any physical CPU. */ +#define RUNSTATE_runnable 1 + +/* VCPU is blocked (a.k.a. idle). It is therefore not runnable. */ +#define RUNSTATE_blocked 2 + +/* + * VCPU is not runnable, but it is not blocked. + * This is a 'catch all' state for things like hotplug and pauses by the + * system administrator (or for critical sections in the hypervisor). + * RUNSTATE_blocked dominates this state (it is the preferred state). + */ +#define RUNSTATE_offline 3 + +/* + * Register a shared memory area from which the guest may obtain its own + * runstate information without needing to execute a hypercall. + * Notes: + * 1. The registered address may be virtual or physical or guest handle, + * depending on the platform. Virtual address or guest handle should be + * registered on x86 systems. + * 2. Only one shared area may be registered per VCPU. The shared area is + * updated by the hypervisor each time the VCPU is scheduled. Thus + * runstate.state will always be RUNSTATE_running and + * runstate.state_entry_time will indicate the system time at which the + * VCPU was last scheduled to run. + * @extra_arg == pointer to vcpu_register_runstate_memory_area structure. + */ +#define VCPUOP_register_runstate_memory_area 5 +struct vcpu_register_runstate_memory_area { + union { + XEN_GUEST_HANDLE(vcpu_runstate_info_t) h; + struct vcpu_runstate_info *v; + uint64_t p; + } addr; +}; +typedef struct vcpu_register_runstate_memory_area vcpu_register_runstate_memory_area_t; +DEFINE_XEN_GUEST_HANDLE(vcpu_register_runstate_memory_area_t); + +/* + * Set or stop a VCPU's periodic timer. Every VCPU has one periodic timer + * which can be set via these commands. Periods smaller than one millisecond + * may not be supported. + */ +#define VCPUOP_set_periodic_timer 6 /* arg == vcpu_set_periodic_timer_t */ +#define VCPUOP_stop_periodic_timer 7 /* arg == NULL */ +struct vcpu_set_periodic_timer { + uint64_t period_ns; +}; +DEFINE_GUEST_HANDLE_STRUCT(vcpu_set_periodic_timer); +typedef struct vcpu_set_periodic_timer vcpu_set_periodic_timer_t; +DEFINE_XEN_GUEST_HANDLE(vcpu_set_periodic_timer_t); + +/* + * Set or stop a VCPU's single-shot timer. Every VCPU has one single-shot + * timer which can be set via these commands. + */ +#define VCPUOP_set_singleshot_timer 8 /* arg == vcpu_set_singleshot_timer_t */ +#define VCPUOP_stop_singleshot_timer 9 /* arg == NULL */ +struct vcpu_set_singleshot_timer { + uint64_t timeout_abs_ns; /* Absolute system time value in nanoseconds. */ + uint32_t flags; /* VCPU_SSHOTTMR_??? */ +}; +DEFINE_GUEST_HANDLE_STRUCT(vcpu_set_singleshot_timer); +typedef struct vcpu_set_singleshot_timer vcpu_set_singleshot_timer_t; +DEFINE_XEN_GUEST_HANDLE(vcpu_set_singleshot_timer_t); + +/* Flags to VCPUOP_set_singleshot_timer. */ + /* Require the timeout to be in the future (return -ETIME if it's passed). */ +#define _VCPU_SSHOTTMR_future (0) +#define VCPU_SSHOTTMR_future (1U << _VCPU_SSHOTTMR_future) + +/* + * Register a memory location in the guest address space for the + * vcpu_info structure. This allows the guest to place the vcpu_info + * structure in a convenient place, such as in a per-cpu data area. + * The pointer need not be page aligned, but the structure must not + * cross a page boundary. + * + * This may be called only once per vcpu. + */ +#define VCPUOP_register_vcpu_info 10 /* arg == vcpu_register_vcpu_info_t */ +struct vcpu_register_vcpu_info { + uint64_t mfn; /* mfn of page to place vcpu_info */ + uint32_t offset; /* offset within page */ + uint32_t rsvd; /* unused */ +}; +DEFINE_GUEST_HANDLE_STRUCT(vcpu_register_vcpu_info); +typedef struct vcpu_register_vcpu_info vcpu_register_vcpu_info_t; +DEFINE_XEN_GUEST_HANDLE(vcpu_register_vcpu_info_t); + +/* Send an NMI to the specified VCPU. @extra_arg == NULL. */ +#define VCPUOP_send_nmi 11 + +/* + * Get the physical ID information for a pinned vcpu's underlying physical + * processor. The physical ID informmation is architecture-specific. + * On x86: id[31:0]=apic_id, id[63:32]=acpi_id, and all values 0xff and + * greater are reserved. + * This command returns -EINVAL if it is not a valid operation for this VCPU. + */ +#define VCPUOP_get_physid 12 /* arg == vcpu_get_physid_t */ +struct vcpu_get_physid { + uint64_t phys_id; +}; +typedef struct vcpu_get_physid vcpu_get_physid_t; +DEFINE_XEN_GUEST_HANDLE(vcpu_get_physid_t); +#define xen_vcpu_physid_to_x86_apicid(physid) \ + ((((uint32_t)(physid)) >= 0xff) ? 0xff : ((uint8_t)(physid))) +#define xen_vcpu_physid_to_x86_acpiid(physid) \ + ((((uint32_t)((physid)>>32)) >= 0xff) ? 0xff : ((uint8_t)((physid)>>32))) + +/* + * Register a memory location to get a secondary copy of the vcpu time + * parameters. The master copy still exists as part of the vcpu shared + * memory area, and this secondary copy is updated whenever the master copy + * is updated (and using the same versioning scheme for synchronisation). + * + * The intent is that this copy may be mapped (RO) into userspace so + * that usermode can compute system time using the time info and the + * tsc. Usermode will see an array of vcpu_time_info structures, one + * for each vcpu, and choose the right one by an existing mechanism + * which allows it to get the current vcpu number (such as via a + * segment limit). It can then apply the normal algorithm to compute + * system time from the tsc. + * + * @extra_arg == pointer to vcpu_register_time_info_memory_area structure. + */ +#define VCPUOP_register_vcpu_time_memory_area 13 +DEFINE_XEN_GUEST_HANDLE(vcpu_time_info_t); +struct vcpu_register_time_memory_area { + union { + XEN_GUEST_HANDLE(vcpu_time_info_t) h; + struct vcpu_time_info *v; + uint64_t p; + } addr; +}; +typedef struct vcpu_register_time_memory_area vcpu_register_time_memory_area_t; +DEFINE_XEN_GUEST_HANDLE(vcpu_register_time_memory_area_t); + +#endif /* __XEN_PUBLIC_VCPU_H__ */ diff --git a/include/xenpvdrivers/interface/version.h b/include/xenpvdrivers/interface/version.h new file mode 100644 index 0000000..da54fd8 --- /dev/null +++ b/include/xenpvdrivers/interface/version.h @@ -0,0 +1,93 @@ +/****************************************************************************** + * version.h + * + * Xen version, type, and compile information. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2005, Nguyen Anh Quynh + * Copyright (c) 2005, Keir Fraser + */ + +#ifndef __XEN_PUBLIC_VERSION_H__ +#define __XEN_PUBLIC_VERSION_H__ + +/* NB. All ops return zero on success, except XENVER_{version,pagesize} */ + +/* arg == NULL; returns major:minor (16:16). */ +#define XENVER_version 0 + +/* arg == xen_extraversion_t. */ +#define XENVER_extraversion 1 +typedef char xen_extraversion_t[16]; +struct xen_extraversion { + xen_extraversion_t extraversion; +}; +#define XEN_EXTRAVERSION_LEN (sizeof(xen_extraversion_t)) + +/* arg == xen_compile_info_t. */ +#define XENVER_compile_info 2 +struct xen_compile_info { + char compiler[64]; + char compile_by[16]; + char compile_domain[32]; + char compile_date[32]; +}; +typedef struct xen_compile_info xen_compile_info_t; + +#define XENVER_capabilities 3 +typedef char xen_capabilities_info_t[1024]; +struct xen_capabilities_info { + xen_capabilities_info_t info; +}; +#define XEN_CAPABILITIES_INFO_LEN (sizeof(xen_capabilities_info_t)) + +#define XENVER_changeset 4 +typedef char xen_changeset_info_t[64]; +struct xen_changeset_info { + xen_changeset_info_t info; +}; +#define XEN_CHANGESET_INFO_LEN (sizeof(xen_changeset_info_t)) + +#define XENVER_platform_parameters 5 +struct xen_platform_parameters { + unsigned long virt_start; +}; +typedef struct xen_platform_parameters xen_platform_parameters_t; + +#define XENVER_get_features 6 +struct xen_feature_info { + unsigned int submap_idx; /* IN: which 32-bit submap to return */ + uint32_t submap; /* OUT: 32-bit submap */ +}; +typedef struct xen_feature_info xen_feature_info_t; + +/* Declares the features reported by XENVER_get_features. */ +#include "features.h" + +/* arg == NULL; returns host memory page size. */ +#define XENVER_pagesize 7 + +/* arg == xen_domain_handle_t. */ +#define XENVER_guest_handle 8 + +#define XENVER_commandline 9 +typedef char xen_commandline_t[1024]; + +#endif /* __XEN_PUBLIC_VERSION_H__ */ diff --git a/include/xenpvdrivers/interface/xen-compat.h b/include/xenpvdrivers/interface/xen-compat.h new file mode 100644 index 0000000..2e38003 --- /dev/null +++ b/include/xenpvdrivers/interface/xen-compat.h @@ -0,0 +1,44 @@ +/****************************************************************************** + * xen-compat.h + * + * Guest OS interface to Xen. Compatibility layer. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2006, Christian Limpach + */ + +#ifndef __XEN_PUBLIC_XEN_COMPAT_H__ +#define __XEN_PUBLIC_XEN_COMPAT_H__ + +#define __XEN_LATEST_INTERFACE_VERSION__ 0x0003020a + +#if defined(__XEN__) || defined(__XEN_TOOLS__) +/* Xen is built with matching headers and implements the latest interface. */ +#define __XEN_INTERFACE_VERSION__ __XEN_LATEST_INTERFACE_VERSION__ +#elif !defined(__XEN_INTERFACE_VERSION__) +/* Guests which do not specify a version get the legacy interface. */ +#define __XEN_INTERFACE_VERSION__ 0x00000000 +#endif + +#if __XEN_INTERFACE_VERSION__ > __XEN_LATEST_INTERFACE_VERSION__ +#error "These header files do not support the requested interface version." +#endif + +#endif /* __XEN_PUBLIC_XEN_COMPAT_H__ */ diff --git a/include/xenpvdrivers/interface/xen.h b/include/xenpvdrivers/interface/xen.h new file mode 100644 index 0000000..41e8f6e --- /dev/null +++ b/include/xenpvdrivers/interface/xen.h @@ -0,0 +1,714 @@ +/****************************************************************************** + * xen.h + * + * Guest OS interface to Xen. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (c) 2004, K A Fraser + */ + +#ifndef __XEN_PUBLIC_XEN_H__ +#define __XEN_PUBLIC_XEN_H__ + +#include "xen-compat.h" +#ifdef CONFIG_PARAVIRT_XEN +#include +#endif + +#if defined(CONFIG_PARAVIRT_XEN) && !defined(HAVE_XEN_PLATFORM_COMPAT_H) +#include +#elif defined(__i386__) || defined(__x86_64__) +#include "arch-x86/xen.h" +#elif defined(__ia64__) +#include "arch-ia64.h" +#else +#error "Unsupported architecture" +#endif + +#ifndef __ASSEMBLY__ +/* Guest handles for primitive C types. */ +DEFINE_XEN_GUEST_HANDLE(char); +__DEFINE_XEN_GUEST_HANDLE(uchar, unsigned char); +DEFINE_XEN_GUEST_HANDLE(int); +__DEFINE_XEN_GUEST_HANDLE(uint, unsigned int); +DEFINE_XEN_GUEST_HANDLE(long); +__DEFINE_XEN_GUEST_HANDLE(ulong, unsigned long); +DEFINE_XEN_GUEST_HANDLE(void); + +DEFINE_XEN_GUEST_HANDLE(uint64_t); +DEFINE_XEN_GUEST_HANDLE(xen_pfn_t); +#endif + +/* + * HYPERCALLS + */ + +#define __HYPERVISOR_set_trap_table 0 +#define __HYPERVISOR_mmu_update 1 +#define __HYPERVISOR_set_gdt 2 +#define __HYPERVISOR_stack_switch 3 +#define __HYPERVISOR_set_callbacks 4 +#define __HYPERVISOR_fpu_taskswitch 5 +#define __HYPERVISOR_sched_op_compat 6 /* compat since 0x00030101 */ +#define __HYPERVISOR_platform_op 7 +#define __HYPERVISOR_set_debugreg 8 +#define __HYPERVISOR_get_debugreg 9 +#define __HYPERVISOR_update_descriptor 10 +#define __HYPERVISOR_memory_op 12 +#define __HYPERVISOR_multicall 13 +#define __HYPERVISOR_update_va_mapping 14 +#define __HYPERVISOR_set_timer_op 15 +#define __HYPERVISOR_event_channel_op_compat 16 /* compat since 0x00030202 */ +#define __HYPERVISOR_xen_version 17 +#define __HYPERVISOR_console_io 18 +#define __HYPERVISOR_physdev_op_compat 19 /* compat since 0x00030202 */ +#define __HYPERVISOR_grant_table_op 20 +#define __HYPERVISOR_vm_assist 21 +#define __HYPERVISOR_update_va_mapping_otherdomain 22 +#define __HYPERVISOR_iret 23 /* x86 only */ +#define __HYPERVISOR_vcpu_op 24 +#define __HYPERVISOR_set_segment_base 25 /* x86/64 only */ +#define __HYPERVISOR_mmuext_op 26 +#define __HYPERVISOR_xsm_op 27 +#define __HYPERVISOR_nmi_op 28 +#define __HYPERVISOR_sched_op_new 29 +#define __HYPERVISOR_callback_op 30 +#define __HYPERVISOR_xenoprof_op 31 +#define __HYPERVISOR_event_channel_op 32 +#define __HYPERVISOR_physdev_op 33 +#define __HYPERVISOR_hvm_op 34 +#define __HYPERVISOR_sysctl 35 +#define __HYPERVISOR_domctl 36 +#define __HYPERVISOR_kexec_op 37 +#define __HYPERVISOR_tmem_op 38 + +/* Architecture-specific hypercall definitions. */ +#define __HYPERVISOR_arch_0 48 +#define __HYPERVISOR_arch_1 49 +#define __HYPERVISOR_arch_2 50 +#define __HYPERVISOR_arch_3 51 +#define __HYPERVISOR_arch_4 52 +#define __HYPERVISOR_arch_5 53 +#define __HYPERVISOR_arch_6 54 +#define __HYPERVISOR_arch_7 55 + +/* + * HYPERCALL COMPATIBILITY. + */ + +/* New sched_op hypercall introduced in 0x00030101. */ +#if __XEN_INTERFACE_VERSION__ < 0x00030101 || (defined(CONFIG_PARAVIRT_XEN) && !defined(HAVE_XEN_PLATFORM_COMPAT_H)) +#undef __HYPERVISOR_sched_op +#define __HYPERVISOR_sched_op __HYPERVISOR_sched_op_compat +#else +#define __HYPERVISOR_sched_op __HYPERVISOR_sched_op_new +#endif + +/* New event-channel and physdev hypercalls introduced in 0x00030202. */ +#if __XEN_INTERFACE_VERSION__ < 0x00030202 +#undef __HYPERVISOR_event_channel_op +#define __HYPERVISOR_event_channel_op __HYPERVISOR_event_channel_op_compat +#undef __HYPERVISOR_physdev_op +#define __HYPERVISOR_physdev_op __HYPERVISOR_physdev_op_compat +#endif + +/* New platform_op hypercall introduced in 0x00030204. */ +#if __XEN_INTERFACE_VERSION__ < 0x00030204 || (defined(CONFIG_PARAVIRT_XEN) && !defined(HAVE_XEN_PLATFORM_COMPAT_H)) +#define __HYPERVISOR_dom0_op __HYPERVISOR_platform_op +#endif + +/* + * VIRTUAL INTERRUPTS + * + * Virtual interrupts that a guest OS may receive from Xen. + * + * In the side comments, 'V.' denotes a per-VCPU VIRQ while 'G.' denotes a + * global VIRQ. The former can be bound once per VCPU and cannot be re-bound. + * The latter can be allocated only once per guest: they must initially be + * allocated to VCPU0 but can subsequently be re-bound. + */ +#define VIRQ_TIMER 0 /* V. Timebase update, and/or requested timeout. */ +#define VIRQ_DEBUG 1 /* V. Request guest to dump debug info. */ +#define VIRQ_CONSOLE 2 /* G. (DOM0) Bytes received on emergency console. */ +#define VIRQ_DOM_EXC 3 /* G. (DOM0) Exceptional event for some domain. */ +#define VIRQ_TBUF 4 /* G. (DOM0) Trace buffer has records available. */ +#define VIRQ_DEBUGGER 6 /* G. (DOM0) A domain has paused for debugging. */ +#define VIRQ_XENOPROF 7 /* V. XenOprofile interrupt: new sample available */ +#define VIRQ_CON_RING 8 /* G. (DOM0) Bytes received on console */ +#define VIRQ_PCPU_STATE 9 /* G. (DOM0) PCPU state changed */ +#define VIRQ_MEM_EVENT 10 /* G. (DOM0) A memory event has occured */ + +/* Architecture-specific VIRQ definitions. */ +#define VIRQ_ARCH_0 16 +#define VIRQ_ARCH_1 17 +#define VIRQ_ARCH_2 18 +#define VIRQ_ARCH_3 19 +#define VIRQ_ARCH_4 20 +#define VIRQ_ARCH_5 21 +#define VIRQ_ARCH_6 22 +#define VIRQ_ARCH_7 23 + +#define NR_VIRQS 24 + +/* + * HYPERVISOR_mmu_update(reqs, count, pdone, foreigndom) + * + * @reqs is an array of mmu_update_t structures ((ptr, val) pairs). + * @count is the length of the above array. + * @pdone is an output parameter indicating number of completed operations + * @foreigndom[15:0]: FD, the expected owner of data pages referenced in this + * hypercall invocation. Can be DOMID_SELF. + * @foreigndom[31:16]: PFD, the expected owner of pagetable pages referenced + * in this hypercall invocation. The value of this field + * (x) encodes the PFD as follows: + * x == 0 => PFD == DOMID_SELF + * x != 0 => PFD == x - 1 + * + * Sub-commands: ptr[1:0] specifies the appropriate MMU_* command. + * ------------- + * ptr[1:0] == MMU_NORMAL_PT_UPDATE: + * Updates an entry in a page table belonging to PFD. If updating an L1 table, + * and the new table entry is valid/present, the mapped frame must belong to + * FD. If attempting to map an I/O page then the caller assumes the privilege + * of the FD. + * FD == DOMID_IO: Permit /only/ I/O mappings, at the priv level of the caller. + * FD == DOMID_XEN: Map restricted areas of Xen's heap space. + * ptr[:2] -- Machine address of the page-table entry to modify. + * val -- Value to write. + * + * ptr[1:0] == MMU_MACHPHYS_UPDATE: + * Updates an entry in the machine->pseudo-physical mapping table. + * ptr[:2] -- Machine address within the frame whose mapping to modify. + * The frame must belong to the FD, if one is specified. + * val -- Value to write into the mapping entry. + * + * ptr[1:0] == MMU_PT_UPDATE_PRESERVE_AD: + * As MMU_NORMAL_PT_UPDATE above, but A/D bits currently in the PTE are ORed + * with those in @val. + */ +#define MMU_NORMAL_PT_UPDATE 0 /* checked '*ptr = val'. ptr is MA. */ +#define MMU_MACHPHYS_UPDATE 1 /* ptr = MA of frame to modify entry for */ +#define MMU_PT_UPDATE_PRESERVE_AD 2 /* atomically: *ptr = val | (*ptr&(A|D)) */ + +/* + * MMU EXTENDED OPERATIONS + * + * HYPERVISOR_mmuext_op() accepts a list of mmuext_op structures. + * A foreigndom (FD) can be specified (or DOMID_SELF for none). + * Where the FD has some effect, it is described below. + * + * cmd: MMUEXT_(UN)PIN_*_TABLE + * mfn: Machine frame number to be (un)pinned as a p.t. page. + * The frame must belong to the FD, if one is specified. + * + * cmd: MMUEXT_NEW_BASEPTR + * mfn: Machine frame number of new page-table base to install in MMU. + * + * cmd: MMUEXT_NEW_USER_BASEPTR [x86/64 only] + * mfn: Machine frame number of new page-table base to install in MMU + * when in user space. + * + * cmd: MMUEXT_TLB_FLUSH_LOCAL + * No additional arguments. Flushes local TLB. + * + * cmd: MMUEXT_INVLPG_LOCAL + * linear_addr: Linear address to be flushed from the local TLB. + * + * cmd: MMUEXT_TLB_FLUSH_MULTI + * vcpumask: Pointer to bitmap of VCPUs to be flushed. + * + * cmd: MMUEXT_INVLPG_MULTI + * linear_addr: Linear address to be flushed. + * vcpumask: Pointer to bitmap of VCPUs to be flushed. + * + * cmd: MMUEXT_TLB_FLUSH_ALL + * No additional arguments. Flushes all VCPUs' TLBs. + * + * cmd: MMUEXT_INVLPG_ALL + * linear_addr: Linear address to be flushed from all VCPUs' TLBs. + * + * cmd: MMUEXT_FLUSH_CACHE + * No additional arguments. Writes back and flushes cache contents. + * + * cmd: MMUEXT_SET_LDT + * linear_addr: Linear address of LDT base (NB. must be page-aligned). + * nr_ents: Number of entries in LDT. + * + * cmd: MMUEXT_CLEAR_PAGE + * mfn: Machine frame number to be cleared. + * + * cmd: MMUEXT_COPY_PAGE + * mfn: Machine frame number of the destination page. + * src_mfn: Machine frame number of the source page. + */ +#define MMUEXT_PIN_L1_TABLE 0 +#define MMUEXT_PIN_L2_TABLE 1 +#define MMUEXT_PIN_L3_TABLE 2 +#define MMUEXT_PIN_L4_TABLE 3 +#define MMUEXT_UNPIN_TABLE 4 +#define MMUEXT_NEW_BASEPTR 5 +#define MMUEXT_TLB_FLUSH_LOCAL 6 +#define MMUEXT_INVLPG_LOCAL 7 +#define MMUEXT_TLB_FLUSH_MULTI 8 +#define MMUEXT_INVLPG_MULTI 9 +#define MMUEXT_TLB_FLUSH_ALL 10 +#define MMUEXT_INVLPG_ALL 11 +#define MMUEXT_FLUSH_CACHE 12 +#define MMUEXT_SET_LDT 13 +#define MMUEXT_NEW_USER_BASEPTR 15 +#define MMUEXT_CLEAR_PAGE 16 +#define MMUEXT_COPY_PAGE 17 + +#ifndef __ASSEMBLY__ +struct mmuext_op { + unsigned int cmd; + union { + /* [UN]PIN_TABLE, NEW_BASEPTR, NEW_USER_BASEPTR + * CLEAR_PAGE, COPY_PAGE */ + xen_pfn_t mfn; + /* INVLPG_LOCAL, INVLPG_ALL, SET_LDT */ + unsigned long linear_addr; + } arg1; + union { + /* SET_LDT */ + unsigned int nr_ents; + /* TLB_FLUSH_MULTI, INVLPG_MULTI */ +#if __XEN_INTERFACE_VERSION__ >= 0x00030205 + XEN_GUEST_HANDLE(const_void) vcpumask; +#else + const void *vcpumask; +#endif + /* COPY_PAGE */ + xen_pfn_t src_mfn; + } arg2; +}; +DEFINE_GUEST_HANDLE_STRUCT(mmuext_op); +typedef struct mmuext_op mmuext_op_t; +DEFINE_XEN_GUEST_HANDLE(mmuext_op_t); +#endif + +/* These are passed as 'flags' to update_va_mapping. They can be ORed. */ +/* When specifying UVMF_MULTI, also OR in a pointer to a CPU bitmap. */ +/* UVMF_LOCAL is merely UVMF_MULTI with a NULL bitmap pointer. */ +#define UVMF_NONE (0UL<<0) /* No flushing at all. */ +#define UVMF_TLB_FLUSH (1UL<<0) /* Flush entire TLB(s). */ +#define UVMF_INVLPG (2UL<<0) /* Flush only one entry. */ +#define UVMF_FLUSHTYPE_MASK (3UL<<0) +#define UVMF_MULTI (0UL<<2) /* Flush subset of TLBs. */ +#define UVMF_LOCAL (0UL<<2) /* Flush local TLB. */ +#define UVMF_ALL (1UL<<2) /* Flush all TLBs. */ + +/* + * Commands to HYPERVISOR_console_io(). + */ +#define CONSOLEIO_write 0 +#define CONSOLEIO_read 1 + +/* + * Commands to HYPERVISOR_vm_assist(). + */ +#define VMASST_CMD_enable 0 +#define VMASST_CMD_disable 1 + +/* x86/32 guests: simulate full 4GB segment limits. */ +#define VMASST_TYPE_4gb_segments 0 + +/* x86/32 guests: trap (vector 15) whenever above vmassist is used. */ +#define VMASST_TYPE_4gb_segments_notify 1 + +/* + * x86 guests: support writes to bottom-level PTEs. + * NB1. Page-directory entries cannot be written. + * NB2. Guest must continue to remove all writable mappings of PTEs. + */ +#define VMASST_TYPE_writable_pagetables 2 + +/* x86/PAE guests: support PDPTs above 4GB. */ +#define VMASST_TYPE_pae_extended_cr3 3 + +#define MAX_VMASST_TYPE 3 + +#ifndef __ASSEMBLY__ + +typedef uint16_t domid_t; + +/* Domain ids >= DOMID_FIRST_RESERVED cannot be used for ordinary domains. */ +#define DOMID_FIRST_RESERVED (0x7FF0U) + +/* DOMID_SELF is used in certain contexts to refer to oneself. */ +#define DOMID_SELF (0x7FF0U) + +/* + * DOMID_IO is used to restrict page-table updates to mapping I/O memory. + * Although no Foreign Domain need be specified to map I/O pages, DOMID_IO + * is useful to ensure that no mappings to the OS's own heap are accidentally + * installed. (e.g., in Linux this could cause havoc as reference counts + * aren't adjusted on the I/O-mapping code path). + * This only makes sense in MMUEXT_SET_FOREIGNDOM, but in that context can + * be specified by any calling domain. + */ +#define DOMID_IO (0x7FF1U) + +/* + * DOMID_XEN is used to allow privileged domains to map restricted parts of + * Xen's heap space (e.g., the machine_to_phys table). + * This only makes sense in MMUEXT_SET_FOREIGNDOM, and is only permitted if + * the caller is privileged. + */ +#define DOMID_XEN (0x7FF2U) + +/* + * DOMID_COW is used as the owner of sharable pages */ +#define DOMID_COW (0x7FF3U) + +/* DOMID_INVALID is used to identity invalid domid */ +#define DOMID_INVALID (0x7FFFU) + +/* + * Send an array of these to HYPERVISOR_mmu_update(). + * NB. The fields are natural pointer/address size for this architecture. + */ +struct mmu_update { + uint64_t ptr; /* Machine address of PTE. */ + uint64_t val; /* New contents of PTE. */ +}; +DEFINE_GUEST_HANDLE_STRUCT(mmu_update); +typedef struct mmu_update mmu_update_t; +DEFINE_XEN_GUEST_HANDLE(mmu_update_t); + +/* + * Send an array of these to HYPERVISOR_multicall(). + * NB. The fields are natural register size for this architecture. + */ +struct multicall_entry { + unsigned long op; +#if !defined(CONFIG_PARAVIRT_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H) + unsigned long result; +#else + long result; +#endif + unsigned long args[6]; +}; +DEFINE_GUEST_HANDLE_STRUCT(multicall_entry); +typedef struct multicall_entry multicall_entry_t; +DEFINE_XEN_GUEST_HANDLE(multicall_entry_t); + +/* + * Event channel endpoints per domain: + * 1024 if a long is 32 bits; 4096 if a long is 64 bits. + */ +#define NR_EVENT_CHANNELS (sizeof(unsigned long) * sizeof(unsigned long) * 64) + +struct vcpu_time_info { + /* + * Updates to the following values are preceded and followed by an + * increment of 'version'. The guest can therefore detect updates by + * looking for changes to 'version'. If the least-significant bit of + * the version number is set then an update is in progress and the guest + * must wait to read a consistent set of values. + * The correct way to interact with the version number is similar to + * Linux's seqlock: see the implementations of read_seqbegin/read_seqretry. + */ + uint32_t version; + uint32_t pad0; + uint64_t tsc_timestamp; /* TSC at last update of time vals. */ + uint64_t system_time; /* Time, in nanosecs, since boot. */ + /* + * Current system time: + * system_time + + * ((((tsc - tsc_timestamp) << tsc_shift) * tsc_to_system_mul) >> 32) + * CPU frequency (Hz): + * ((10^9 << 32) / tsc_to_system_mul) >> tsc_shift + */ + uint32_t tsc_to_system_mul; + int8_t tsc_shift; + int8_t pad1[3]; +}; /* 32 bytes */ +typedef struct vcpu_time_info vcpu_time_info_t; + +struct vcpu_info { + /* + * 'evtchn_upcall_pending' is written non-zero by Xen to indicate + * a pending notification for a particular VCPU. It is then cleared + * by the guest OS /before/ checking for pending work, thus avoiding + * a set-and-check race. Note that the mask is only accessed by Xen + * on the CPU that is currently hosting the VCPU. This means that the + * pending and mask flags can be updated by the guest without special + * synchronisation (i.e., no need for the x86 LOCK prefix). + * This may seem suboptimal because if the pending flag is set by + * a different CPU then an IPI may be scheduled even when the mask + * is set. However, note: + * 1. The task of 'interrupt holdoff' is covered by the per-event- + * channel mask bits. A 'noisy' event that is continually being + * triggered can be masked at source at this very precise + * granularity. + * 2. The main purpose of the per-VCPU mask is therefore to restrict + * reentrant execution: whether for concurrency control, or to + * prevent unbounded stack usage. Whatever the purpose, we expect + * that the mask will be asserted only for short periods at a time, + * and so the likelihood of a 'spurious' IPI is suitably small. + * The mask is read before making an event upcall to the guest: a + * non-zero mask therefore guarantees that the VCPU will not receive + * an upcall activation. The mask is cleared when the VCPU requests + * to block: this avoids wakeup-waiting races. + */ + uint8_t evtchn_upcall_pending; + uint8_t evtchn_upcall_mask; + unsigned long evtchn_pending_sel; + struct arch_vcpu_info arch; +#ifdef CONFIG_PARAVIRT_XEN + struct pvclock_vcpu_time_info time; +#else + struct vcpu_time_info time; +#endif +}; /* 64 bytes (x86) */ +#ifndef __XEN__ +typedef struct vcpu_info vcpu_info_t; +#endif + +/* + * Xen/kernel shared data -- pointer provided in start_info. + * + * This structure is defined to be both smaller than a page, and the + * only data on the shared page, but may vary in actual size even within + * compatible Xen versions; guests should not rely on the size + * of this structure remaining constant. + */ +struct shared_info { + struct vcpu_info vcpu_info[XEN_LEGACY_MAX_VCPUS]; + + /* + * A domain can create "event channels" on which it can send and receive + * asynchronous event notifications. There are three classes of event that + * are delivered by this mechanism: + * 1. Bi-directional inter- and intra-domain connections. Domains must + * arrange out-of-band to set up a connection (usually by allocating + * an unbound 'listener' port and avertising that via a storage service + * such as xenstore). + * 2. Physical interrupts. A domain with suitable hardware-access + * privileges can bind an event-channel port to a physical interrupt + * source. + * 3. Virtual interrupts ('events'). A domain can bind an event-channel + * port to a virtual interrupt source, such as the virtual-timer + * device or the emergency console. + * + * Event channels are addressed by a "port index". Each channel is + * associated with two bits of information: + * 1. PENDING -- notifies the domain that there is a pending notification + * to be processed. This bit is cleared by the guest. + * 2. MASK -- if this bit is clear then a 0->1 transition of PENDING + * will cause an asynchronous upcall to be scheduled. This bit is only + * updated by the guest. It is read-only within Xen. If a channel + * becomes pending while the channel is masked then the 'edge' is lost + * (i.e., when the channel is unmasked, the guest must manually handle + * pending notifications as no upcall will be scheduled by Xen). + * + * To expedite scanning of pending notifications, any 0->1 pending + * transition on an unmasked channel causes a corresponding bit in a + * per-vcpu selector word to be set. Each bit in the selector covers a + * 'C long' in the PENDING bitfield array. + */ + unsigned long evtchn_pending[sizeof(unsigned long) * 8]; + unsigned long evtchn_mask[sizeof(unsigned long) * 8]; + + /* + * Wallclock time: updated only by control software. Guests should base + * their gettimeofday() syscall on this wallclock-base value. + */ +#ifdef CONFIG_PARAVIRT_XEN + struct pvclock_wall_clock wc; +#else + uint32_t wc_version; /* Version counter: see vcpu_time_info_t. */ + uint32_t wc_sec; /* Secs 00:00:00 UTC, Jan 1, 1970. */ + uint32_t wc_nsec; /* Nsecs 00:00:00 UTC, Jan 1, 1970. */ +#endif + + struct arch_shared_info arch; + +}; +#ifndef __XEN__ +typedef struct shared_info shared_info_t; +#endif + +/* + * Start-of-day memory layout: + * 1. The domain is started within contiguous virtual-memory region. + * 2. The contiguous region ends on an aligned 4MB boundary. + * 3. This the order of bootstrap elements in the initial virtual region: + * a. relocated kernel image + * b. initial ram disk [mod_start, mod_len] + * c. list of allocated page frames [mfn_list, nr_pages] + * (unless relocated due to XEN_ELFNOTE_INIT_P2M) + * d. start_info_t structure [register ESI (x86)] + * e. bootstrap page tables [pt_base, CR3 (x86)] + * f. bootstrap stack [register ESP (x86)] + * 4. Bootstrap elements are packed together, but each is 4kB-aligned. + * 5. The initial ram disk may be omitted. + * 6. The list of page frames forms a contiguous 'pseudo-physical' memory + * layout for the domain. In particular, the bootstrap virtual-memory + * region is a 1:1 mapping to the first section of the pseudo-physical map. + * 7. All bootstrap elements are mapped read-writable for the guest OS. The + * only exception is the bootstrap page table, which is mapped read-only. + * 8. There is guaranteed to be at least 512kB padding after the final + * bootstrap element. If necessary, the bootstrap virtual region is + * extended by an extra 4MB to ensure this. + */ + +#define MAX_GUEST_CMDLINE 1024 +struct start_info { + /* THE FOLLOWING ARE FILLED IN BOTH ON INITIAL BOOT AND ON RESUME. */ + char magic[32]; /* "xen--". */ + unsigned long nr_pages; /* Total pages allocated to this domain. */ + unsigned long shared_info; /* MACHINE address of shared info struct. */ + uint32_t flags; /* SIF_xxx flags. */ + xen_pfn_t store_mfn; /* MACHINE page number of shared page. */ + uint32_t store_evtchn; /* Event channel for store communication. */ + union { + struct { + xen_pfn_t mfn; /* MACHINE page number of console page. */ + uint32_t evtchn; /* Event channel for console page. */ + } domU; + struct { + uint32_t info_off; /* Offset of console_info struct. */ + uint32_t info_size; /* Size of console_info struct from start.*/ + } dom0; + } console; + /* THE FOLLOWING ARE ONLY FILLED IN ON INITIAL BOOT (NOT RESUME). */ + unsigned long pt_base; /* VIRTUAL address of page directory. */ + unsigned long nr_pt_frames; /* Number of bootstrap p.t. frames. */ + unsigned long mfn_list; /* VIRTUAL address of page-frame list. */ + unsigned long mod_start; /* VIRTUAL address of pre-loaded module. */ + unsigned long mod_len; /* Size (bytes) of pre-loaded module. */ + int8_t cmd_line[MAX_GUEST_CMDLINE]; + /* The pfn range here covers both page table and p->m table frames. */ + unsigned long first_p2m_pfn;/* 1st pfn forming initial P->M table. */ + unsigned long nr_p2m_frames;/* # of pfns forming initial P->M table. */ +}; +typedef struct start_info start_info_t; + +/* New console union for dom0 introduced in 0x00030203. */ +#if __XEN_INTERFACE_VERSION__ < 0x00030203 +#define console_mfn console.domU.mfn +#define console_evtchn console.domU.evtchn +#endif + +/* These flags are passed in the 'flags' field of start_info_t. */ +#define SIF_PRIVILEGED (1<<0) /* Is the domain privileged? */ +#define SIF_INITDOMAIN (1<<1) /* Is this the initial control domain? */ +#define SIF_MULTIBOOT_MOD (1<<2) /* Is mod_start a multiboot module? */ +#define SIF_PM_MASK (0xFF<<8) /* reserve 1 byte for xen-pm options */ + +/* + * A multiboot module is a package containing modules very similar to a + * multiboot module array. The only differences are: + * - the array of module descriptors is by convention simply at the beginning + * of the multiboot module, + * - addresses in the module descriptors are based on the beginning of the + * multiboot module, + * - the number of modules is determined by a termination descriptor that has + * mod_start == 0. + * + * This permits to both build it statically and reference it in a configuration + * file, and let the PV guest easily rebase the addresses to virtual addresses + * and at the same time count the number of modules. + */ +struct xen_multiboot_mod_list +{ + /* Address of first byte of the module */ + uint32_t mod_start; + /* Address of last byte of the module (inclusive) */ + uint32_t mod_end; + /* Address of zero-terminated command line */ + uint32_t cmdline; + /* Unused, must be zero */ + uint32_t pad; +}; + +typedef struct dom0_vga_console_info { + uint8_t video_type; /* DOM0_VGA_CONSOLE_??? */ +#define XEN_VGATYPE_TEXT_MODE_3 0x03 +#define XEN_VGATYPE_VESA_LFB 0x23 + + union { + struct { + /* Font height, in pixels. */ + uint16_t font_height; + /* Cursor location (column, row). */ + uint16_t cursor_x, cursor_y; + /* Number of rows and columns (dimensions in characters). */ + uint16_t rows, columns; + } text_mode_3; + + struct { + /* Width and height, in pixels. */ + uint16_t width, height; + /* Bytes per scan line. */ + uint16_t bytes_per_line; + /* Bits per pixel. */ + uint16_t bits_per_pixel; + /* LFB physical address, and size (in units of 64kB). */ + uint32_t lfb_base; + uint32_t lfb_size; + /* RGB mask offsets and sizes, as defined by VBE 1.2+ */ + uint8_t red_pos, red_size; + uint8_t green_pos, green_size; + uint8_t blue_pos, blue_size; + uint8_t rsvd_pos, rsvd_size; +#if __XEN_INTERFACE_VERSION__ >= 0x00030206 + /* VESA capabilities (offset 0xa, VESA command 0x4f00). */ + uint32_t gbl_caps; + /* Mode attributes (offset 0x0, VESA command 0x4f01). */ + uint16_t mode_attrs; +#endif + } vesa_lfb; + } u; +} dom0_vga_console_info_t; +#define xen_vga_console_info dom0_vga_console_info +#define xen_vga_console_info_t dom0_vga_console_info_t + +typedef uint8_t xen_domain_handle_t[16]; + +/* Turn a plain number into a C unsigned long constant. */ +#define __mk_unsigned_long(x) x ## UL +#define mk_unsigned_long(x) __mk_unsigned_long(x) + +__DEFINE_XEN_GUEST_HANDLE(uint8, uint8_t); +__DEFINE_XEN_GUEST_HANDLE(uint16, uint16_t); +__DEFINE_XEN_GUEST_HANDLE(uint32, uint32_t); +__DEFINE_XEN_GUEST_HANDLE(uint64, uint64_t); + +#else /* __ASSEMBLY__ */ + +/* In assembly code we cannot use C numeric constant suffixes. */ +#define mk_unsigned_long(x) x + +#endif /* !__ASSEMBLY__ */ + +/* Default definitions for macros used by domctl/sysctl. */ +#if defined(__XEN__) || defined(__XEN_TOOLS__) +#ifndef uint64_aligned_t +#define uint64_aligned_t uint64_t +#endif +#ifndef XEN_GUEST_HANDLE_64 +#define XEN_GUEST_HANDLE_64(name) XEN_GUEST_HANDLE(name) +#endif +#endif + +#endif /* __XEN_PUBLIC_XEN_H__ */ diff --git a/include/xenpvdrivers/interface/xencomm.h b/include/xenpvdrivers/interface/xencomm.h new file mode 100644 index 0000000..ac45e07 --- /dev/null +++ b/include/xenpvdrivers/interface/xencomm.h @@ -0,0 +1,41 @@ +/* + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (C) IBM Corp. 2006 + */ + +#ifndef _XEN_XENCOMM_H_ +#define _XEN_XENCOMM_H_ + +/* A xencomm descriptor is a scatter/gather list containing physical + * addresses corresponding to a virtually contiguous memory area. The + * hypervisor translates these physical addresses to machine addresses to copy + * to and from the virtually contiguous area. + */ + +#define XENCOMM_MAGIC 0x58434F4D /* 'XCOM' */ +#define XENCOMM_INVALID (~0UL) + +struct xencomm_desc { + uint32_t magic; + uint32_t nr_addrs; /* the number of entries in address[] */ + uint64_t address[0]; +}; + +#endif /* _XEN_XENCOMM_H_ */ diff --git a/include/xenpvdrivers/interface/xenoprof.h b/include/xenpvdrivers/interface/xenoprof.h new file mode 100644 index 0000000..183078d --- /dev/null +++ b/include/xenpvdrivers/interface/xenoprof.h @@ -0,0 +1,138 @@ +/****************************************************************************** + * xenoprof.h + * + * Interface for enabling system wide profiling based on hardware performance + * counters + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Copyright (C) 2005 Hewlett-Packard Co. + * Written by Aravind Menon & Jose Renato Santos + */ + +#ifndef __XEN_PUBLIC_XENOPROF_H__ +#define __XEN_PUBLIC_XENOPROF_H__ + +#include "xen.h" + +/* + * Commands to HYPERVISOR_xenoprof_op(). + */ +#define XENOPROF_init 0 +#define XENOPROF_reset_active_list 1 +#define XENOPROF_reset_passive_list 2 +#define XENOPROF_set_active 3 +#define XENOPROF_set_passive 4 +#define XENOPROF_reserve_counters 5 +#define XENOPROF_counter 6 +#define XENOPROF_setup_events 7 +#define XENOPROF_enable_virq 8 +#define XENOPROF_start 9 +#define XENOPROF_stop 10 +#define XENOPROF_disable_virq 11 +#define XENOPROF_release_counters 12 +#define XENOPROF_shutdown 13 +#define XENOPROF_get_buffer 14 +#define XENOPROF_set_backtrace 15 +#define XENOPROF_last_op 15 + +#define MAX_OPROF_EVENTS 32 +#define MAX_OPROF_DOMAINS 25 +#define XENOPROF_CPU_TYPE_SIZE 64 + +/* Xenoprof performance events (not Xen events) */ +struct event_log { + uint64_t eip; + uint8_t mode; + uint8_t event; +}; + +/* PC value that indicates a special code */ +#define XENOPROF_ESCAPE_CODE ~0UL +/* Transient events for the xenoprof->oprofile cpu buf */ +#define XENOPROF_TRACE_BEGIN 1 + +/* Xenoprof buffer shared between Xen and domain - 1 per VCPU */ +struct xenoprof_buf { + uint32_t event_head; + uint32_t event_tail; + uint32_t event_size; + uint32_t vcpu_id; + uint64_t xen_samples; + uint64_t kernel_samples; + uint64_t user_samples; + uint64_t lost_samples; + struct event_log event_log[1]; +}; +#ifndef __XEN__ +typedef struct xenoprof_buf xenoprof_buf_t; +DEFINE_XEN_GUEST_HANDLE(xenoprof_buf_t); +#endif + +struct xenoprof_init { + int32_t num_events; + int32_t is_primary; + char cpu_type[XENOPROF_CPU_TYPE_SIZE]; +}; +typedef struct xenoprof_init xenoprof_init_t; +DEFINE_XEN_GUEST_HANDLE(xenoprof_init_t); + +struct xenoprof_get_buffer { + int32_t max_samples; + int32_t nbuf; + int32_t bufsize; + uint64_t buf_gmaddr; +}; +typedef struct xenoprof_get_buffer xenoprof_get_buffer_t; +DEFINE_XEN_GUEST_HANDLE(xenoprof_get_buffer_t); + +struct xenoprof_counter { + uint32_t ind; + uint64_t count; + uint32_t enabled; + uint32_t event; + uint32_t hypervisor; + uint32_t kernel; + uint32_t user; + uint64_t unit_mask; +}; +typedef struct xenoprof_counter xenoprof_counter_t; +DEFINE_XEN_GUEST_HANDLE(xenoprof_counter_t); + +typedef struct xenoprof_passive { + uint16_t domain_id; + int32_t max_samples; + int32_t nbuf; + int32_t bufsize; + uint64_t buf_gmaddr; +} xenoprof_passive_t; +DEFINE_XEN_GUEST_HANDLE(xenoprof_passive_t); + + +#endif /* __XEN_PUBLIC_XENOPROF_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/xsm/acm.h b/include/xenpvdrivers/interface/xsm/acm.h new file mode 100644 index 0000000..b6ac8d5 --- /dev/null +++ b/include/xenpvdrivers/interface/xsm/acm.h @@ -0,0 +1,223 @@ +/* + * acm.h: Xen access control module interface defintions + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Reiner Sailer + * Copyright (c) 2005, International Business Machines Corporation. + */ + +#ifndef _XEN_PUBLIC_ACM_H +#define _XEN_PUBLIC_ACM_H + +#include "../xen.h" + +/* default ssid reference value if not supplied */ +#define ACM_DEFAULT_SSID 0x0 +#define ACM_DEFAULT_LOCAL_SSID 0x0 + +/* Internal ACM ERROR types */ +#define ACM_OK 0 +#define ACM_UNDEF -1 +#define ACM_INIT_SSID_ERROR -2 +#define ACM_INIT_SOID_ERROR -3 +#define ACM_ERROR -4 + +/* External ACCESS DECISIONS */ +#define ACM_ACCESS_PERMITTED 0 +#define ACM_ACCESS_DENIED -111 +#define ACM_NULL_POINTER_ERROR -200 + +/* + Error codes reported in when trying to test for a new policy + These error codes are reported in an array of tuples where + each error code is followed by a parameter describing the error + more closely, such as a domain id. +*/ +#define ACM_EVTCHN_SHARING_VIOLATION 0x100 +#define ACM_GNTTAB_SHARING_VIOLATION 0x101 +#define ACM_DOMAIN_LOOKUP 0x102 +#define ACM_CHWALL_CONFLICT 0x103 +#define ACM_SSIDREF_IN_USE 0x104 + + +/* primary policy in lower 4 bits */ +#define ACM_NULL_POLICY 0 +#define ACM_CHINESE_WALL_POLICY 1 +#define ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY 2 +#define ACM_POLICY_UNDEFINED 15 + +/* combinations have secondary policy component in higher 4bit */ +#define ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY \ + ((ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY << 4) | ACM_CHINESE_WALL_POLICY) + +/* policy: */ +#define ACM_POLICY_NAME(X) \ + ((X) == (ACM_NULL_POLICY)) ? "NULL" : \ + ((X) == (ACM_CHINESE_WALL_POLICY)) ? "CHINESE WALL" : \ + ((X) == (ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "SIMPLE TYPE ENFORCEMENT" : \ + ((X) == (ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "CHINESE WALL AND SIMPLE TYPE ENFORCEMENT" : \ + "UNDEFINED" + +/* the following policy versions must be increased + * whenever the interpretation of the related + * policy's data structure changes + */ +#define ACM_POLICY_VERSION 4 +#define ACM_CHWALL_VERSION 1 +#define ACM_STE_VERSION 1 + +/* defines a ssid reference used by xen */ +typedef uint32_t ssidref_t; + +/* hooks that are known to domains */ +#define ACMHOOK_none 0 +#define ACMHOOK_sharing 1 +#define ACMHOOK_authorization 2 +#define ACMHOOK_conflictset 3 + +/* -------security policy relevant type definitions-------- */ + +/* type identifier; compares to "equal" or "not equal" */ +typedef uint16_t domaintype_t; + +/* CHINESE WALL POLICY DATA STRUCTURES + * + * current accumulated conflict type set: + * When a domain is started and has a type that is in + * a conflict set, the conflicting types are incremented in + * the aggregate set. When a domain is destroyed, the + * conflicting types to its type are decremented. + * If a domain has multiple types, this procedure works over + * all those types. + * + * conflict_aggregate_set[i] holds the number of + * running domains that have a conflict with type i. + * + * running_types[i] holds the number of running domains + * that include type i in their ssidref-referenced type set + * + * conflict_sets[i][j] is "0" if type j has no conflict + * with type i and is "1" otherwise. + */ +/* high-16 = version, low-16 = check magic */ +#define ACM_MAGIC 0x0001debc + +/* size of the SHA1 hash identifying the XML policy from which the + binary policy was created */ +#define ACM_SHA1_HASH_SIZE 20 + +/* each offset in bytes from start of the struct they + * are part of */ + +/* V3 of the policy buffer aded a version structure */ +struct acm_policy_version +{ + uint32_t major; + uint32_t minor; +}; + + +/* each buffer consists of all policy information for + * the respective policy given in the policy code + * + * acm_policy_buffer, acm_chwall_policy_buffer, + * and acm_ste_policy_buffer need to stay 32-bit aligned + * because we create binary policies also with external + * tools that assume packed representations (e.g. the java tool) + */ +struct acm_policy_buffer { + uint32_t magic; + uint32_t policy_version; /* ACM_POLICY_VERSION */ + uint32_t len; + uint32_t policy_reference_offset; + uint32_t primary_policy_code; + uint32_t primary_buffer_offset; + uint32_t secondary_policy_code; + uint32_t secondary_buffer_offset; + struct acm_policy_version xml_pol_version; /* add in V3 */ + uint8_t xml_policy_hash[ACM_SHA1_HASH_SIZE]; /* added in V4 */ +}; + + +struct acm_policy_reference_buffer { + uint32_t len; +}; + +struct acm_chwall_policy_buffer { + uint32_t policy_version; /* ACM_CHWALL_VERSION */ + uint32_t policy_code; + uint32_t chwall_max_types; + uint32_t chwall_max_ssidrefs; + uint32_t chwall_max_conflictsets; + uint32_t chwall_ssid_offset; + uint32_t chwall_conflict_sets_offset; + uint32_t chwall_running_types_offset; + uint32_t chwall_conflict_aggregate_offset; +}; + +struct acm_ste_policy_buffer { + uint32_t policy_version; /* ACM_STE_VERSION */ + uint32_t policy_code; + uint32_t ste_max_types; + uint32_t ste_max_ssidrefs; + uint32_t ste_ssid_offset; +}; + +struct acm_stats_buffer { + uint32_t magic; + uint32_t len; + uint32_t primary_policy_code; + uint32_t primary_stats_offset; + uint32_t secondary_policy_code; + uint32_t secondary_stats_offset; +}; + +struct acm_ste_stats_buffer { + uint32_t ec_eval_count; + uint32_t gt_eval_count; + uint32_t ec_denied_count; + uint32_t gt_denied_count; + uint32_t ec_cachehit_count; + uint32_t gt_cachehit_count; +}; + +struct acm_ssid_buffer { + uint32_t len; + ssidref_t ssidref; + uint32_t policy_reference_offset; + uint32_t primary_policy_code; + uint32_t primary_max_types; + uint32_t primary_types_offset; + uint32_t secondary_policy_code; + uint32_t secondary_max_types; + uint32_t secondary_types_offset; +}; + +#endif + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/xsm/acm_ops.h b/include/xenpvdrivers/interface/xsm/acm_ops.h new file mode 100644 index 0000000..1fef7a0 --- /dev/null +++ b/include/xenpvdrivers/interface/xsm/acm_ops.h @@ -0,0 +1,159 @@ +/* + * acm_ops.h: Xen access control module hypervisor commands + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Reiner Sailer + * Copyright (c) 2005,2006 International Business Machines Corporation. + */ + +#ifndef __XEN_PUBLIC_ACM_OPS_H__ +#define __XEN_PUBLIC_ACM_OPS_H__ + +#include "../xen.h" +#include "acm.h" + +/* + * Make sure you increment the interface version whenever you modify this file! + * This makes sure that old versions of acm tools will stop working in a + * well-defined way (rather than crashing the machine, for instance). + */ +#define ACM_INTERFACE_VERSION 0xAAAA000A + +/************************************************************************/ + +/* + * Prototype for this hypercall is: + * int acm_op(int cmd, void *args) + * @cmd == ACMOP_??? (access control module operation). + * @args == Operation-specific extra arguments (NULL if none). + */ + + +#define ACMOP_setpolicy 1 +struct acm_setpolicy { + /* IN */ + XEN_GUEST_HANDLE_64(void) pushcache; + uint32_t pushcache_size; +}; + + +#define ACMOP_getpolicy 2 +struct acm_getpolicy { + /* IN */ + XEN_GUEST_HANDLE_64(void) pullcache; + uint32_t pullcache_size; +}; + + +#define ACMOP_dumpstats 3 +struct acm_dumpstats { + /* IN */ + XEN_GUEST_HANDLE_64(void) pullcache; + uint32_t pullcache_size; +}; + + +#define ACMOP_getssid 4 +#define ACM_GETBY_ssidref 1 +#define ACM_GETBY_domainid 2 +struct acm_getssid { + /* IN */ + uint32_t get_ssid_by; /* ACM_GETBY_* */ + union { + domaintype_t domainid; + ssidref_t ssidref; + } id; + XEN_GUEST_HANDLE_64(void) ssidbuf; + uint32_t ssidbuf_size; +}; + +#define ACMOP_getdecision 5 +struct acm_getdecision { + /* IN */ + uint32_t get_decision_by1; /* ACM_GETBY_* */ + uint32_t get_decision_by2; /* ACM_GETBY_* */ + union { + domaintype_t domainid; + ssidref_t ssidref; + } id1; + union { + domaintype_t domainid; + ssidref_t ssidref; + } id2; + uint32_t hook; + /* OUT */ + uint32_t acm_decision; +}; + + +#define ACMOP_chgpolicy 6 +struct acm_change_policy { + /* IN */ + XEN_GUEST_HANDLE_64(void) policy_pushcache; + uint32_t policy_pushcache_size; + XEN_GUEST_HANDLE_64(void) del_array; + uint32_t delarray_size; + XEN_GUEST_HANDLE_64(void) chg_array; + uint32_t chgarray_size; + /* OUT */ + /* array with error code */ + XEN_GUEST_HANDLE_64(void) err_array; + uint32_t errarray_size; +}; + +#define ACMOP_relabeldoms 7 +struct acm_relabel_doms { + /* IN */ + XEN_GUEST_HANDLE_64(void) relabel_map; + uint32_t relabel_map_size; + /* OUT */ + XEN_GUEST_HANDLE_64(void) err_array; + uint32_t errarray_size; +}; + +/* future interface to Xen */ +struct xen_acmctl { + uint32_t cmd; + uint32_t interface_version; + union { + struct acm_setpolicy setpolicy; + struct acm_getpolicy getpolicy; + struct acm_dumpstats dumpstats; + struct acm_getssid getssid; + struct acm_getdecision getdecision; + struct acm_change_policy change_policy; + struct acm_relabel_doms relabel_doms; + } u; +}; + +typedef struct xen_acmctl xen_acmctl_t; +DEFINE_XEN_GUEST_HANDLE(xen_acmctl_t); + +#endif /* __XEN_PUBLIC_ACM_OPS_H__ */ + +/* + * Local variables: + * mode: C + * c-set-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/include/xenpvdrivers/interface/xsm/flask_op.h b/include/xenpvdrivers/interface/xsm/flask_op.h new file mode 100644 index 0000000..2fc1184 --- /dev/null +++ b/include/xenpvdrivers/interface/xsm/flask_op.h @@ -0,0 +1,47 @@ +/* + * This file contains the flask_op hypercall commands and definitions. + * + * Author: George Coker, + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2, + * as published by the Free Software Foundation. + */ + +#ifndef __FLASK_OP_H__ +#define __FLASK_OP_H__ + +#define FLASK_LOAD 1 +#define FLASK_GETENFORCE 2 +#define FLASK_SETENFORCE 3 +#define FLASK_CONTEXT_TO_SID 4 +#define FLASK_SID_TO_CONTEXT 5 +#define FLASK_ACCESS 6 +#define FLASK_CREATE 7 +#define FLASK_RELABEL 8 +#define FLASK_USER 9 +#define FLASK_POLICYVERS 10 +#define FLASK_GETBOOL 11 +#define FLASK_SETBOOL 12 +#define FLASK_COMMITBOOLS 13 +#define FLASK_MLS 14 +#define FLASK_DISABLE 15 +#define FLASK_GETAVC_THRESHOLD 16 +#define FLASK_SETAVC_THRESHOLD 17 +#define FLASK_AVC_HASHSTATS 18 +#define FLASK_AVC_CACHESTATS 19 +#define FLASK_MEMBER 20 +#define FLASK_ADD_OCONTEXT 21 +#define FLASK_DEL_OCONTEXT 22 + +#define FLASK_LAST FLASK_DEL_OCONTEXT + +typedef struct flask_op { + uint32_t cmd; + uint32_t size; + char *buf; +} flask_op_t; + +DEFINE_XEN_GUEST_HANDLE(flask_op_t); + +#endif diff --git a/include/xenpvdrivers/page.h b/include/xenpvdrivers/page.h new file mode 100644 index 0000000..eaf85fa --- /dev/null +++ b/include/xenpvdrivers/page.h @@ -0,0 +1 @@ +#include diff --git a/include/xenpvdrivers/pcifront.h b/include/xenpvdrivers/pcifront.h new file mode 100644 index 0000000..bb5713f --- /dev/null +++ b/include/xenpvdrivers/pcifront.h @@ -0,0 +1,69 @@ +/* + * PCI Frontend - arch-dependendent declarations + * + * Author: Ryan Wilson + */ +#ifndef __XEN_ASM_PCIFRONT_H__ +#define __XEN_ASM_PCIFRONT_H__ + +#include + +#ifdef __KERNEL__ + +#ifndef __ia64__ + +#include + +struct pcifront_device; +struct pci_bus; +#define pcifront_sd pci_sysdata + +static inline struct pcifront_device * +pcifront_get_pdev(struct pcifront_sd *sd) +{ + return sd->pdev; +} + +static inline void pcifront_init_sd(struct pcifront_sd *sd, + unsigned int domain, unsigned int bus, + struct pcifront_device *pdev) +{ + sd->domain = domain; + sd->pdev = pdev; +} + +static inline void pcifront_setup_root_resources(struct pci_bus *bus, + struct pcifront_sd *sd) +{ +} + +#else /* __ia64__ */ + +#include +#include +#define pcifront_sd pci_controller + +extern void xen_add_resource(struct pci_controller *, unsigned int, + unsigned int, struct acpi_resource *); +extern void xen_pcibios_setup_root_windows(struct pci_bus *, + struct pci_controller *); + +static inline struct pcifront_device * +pcifront_get_pdev(struct pcifront_sd *sd) +{ + return (struct pcifront_device *)sd->platform_data; +} + +static inline void pcifront_setup_root_resources(struct pci_bus *bus, + struct pcifront_sd *sd) +{ + xen_pcibios_setup_root_windows(bus, sd); +} + +#endif /* __ia64__ */ + +extern struct rw_semaphore pci_bus_sem; + +#endif /* __KERNEL__ */ + +#endif /* __XEN_ASM_PCIFRONT_H__ */ diff --git a/include/xenpvdrivers/platform-compat.h b/include/xenpvdrivers/platform-compat.h new file mode 100644 index 0000000..381d5bb --- /dev/null +++ b/include/xenpvdrivers/platform-compat.h @@ -0,0 +1,164 @@ +#ifndef COMPAT_INCLUDE_XEN_PLATFORM_COMPAT_H +#define COMPAT_INCLUDE_XEN_PLATFORM_COMPAT_H + +#include +#include +#include + +#if defined(__LINUX_COMPILER_H) && !defined(__always_inline) +#define __always_inline inline +#endif + +#if defined(__LINUX_SPINLOCK_H) && !defined(DEFINE_SPINLOCK) +#define DEFINE_SPINLOCK(x) spinlock_t x = SPIN_LOCK_UNLOCKED +#endif + +#if defined(_LINUX_INIT_H) && !defined(__init) +#define __init +#endif + +#if defined(__LINUX_CACHE_H) && !defined(__read_mostly) +#define __read_mostly +#endif + +#if defined(_LINUX_SKBUFF_H) && !defined(NET_IP_ALIGN) +#define NET_IP_ALIGN 0 +#endif + +#if defined(_LINUX_SKBUFF_H) && !defined(CHECKSUM_HW) +#define CHECKSUM_HW CHECKSUM_PARTIAL +#endif + +#if defined(_LINUX_ERR_H) && !defined(IS_ERR_VALUE) +#define IS_ERR_VALUE(x) unlikely((x) > (unsigned long)-1000L) +#endif + +#if defined(_ASM_IA64_PGTABLE_H) && !defined(_PGTABLE_NOPUD_H) +#include +#endif + +/* Some kernels have this typedef backported so we cannot reliably + * detect based on version number, hence we forcibly #define it. + */ +#if defined(__LINUX_TYPES_H) || defined(__LINUX_GFP_H) || defined(_LINUX_KERNEL_H) +#define gfp_t unsigned +#endif + +#if defined(_LINUX_NOTIFIER_H) && !defined(ATOMIC_NOTIFIER_HEAD) +#define ATOMIC_NOTIFIER_HEAD(name) struct notifier_block *name +#define atomic_notifier_chain_register(chain,nb) notifier_chain_register(chain,nb) +#define atomic_notifier_chain_unregister(chain,nb) notifier_chain_unregister(chain,nb) +#define atomic_notifier_call_chain(chain,val,v) notifier_call_chain(chain,val,v) +#endif + +#if defined(_LINUX_NOTIFIER_H) && !defined(BLOCKING_NOTIFIER_HEAD) +#define BLOCKING_NOTIFIER_HEAD(name) struct notifier_block *name +#define blocking_notifier_chain_register(chain,nb) notifier_chain_register(chain,nb) +#define blocking_notifier_chain_unregister(chain,nb) notifier_chain_unregister(chain,nb) +#define blocking_notifier_call_chain(chain,val,v) notifier_call_chain(chain,val,v) +#endif + +#if defined(_LINUX_MM_H) && defined set_page_count +#define init_page_count(page) set_page_count(page, 1) +#endif + +#if defined(__LINUX_GFP_H) && !defined __GFP_NOMEMALLOC +#define __GFP_NOMEMALLOC 0 +#endif + +#if defined(_LINUX_FS_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,9) +#define nonseekable_open(inode, filp) /* Nothing to do */ +#endif + +#if defined(_LINUX_MM_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,10) +unsigned long vmalloc_to_pfn(void *addr); +#endif + +#if defined(__LINUX_COMPLETION_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,11) +unsigned long wait_for_completion_timeout(struct completion *x, unsigned long timeout); +#endif + +#if defined(_LINUX_SCHED_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14) +signed long schedule_timeout_interruptible(signed long timeout); +#endif + +#if defined(_LINUX_SLAB_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14) +void *kzalloc(size_t size, int flags); +#endif + +#if defined(_LINUX_BLKDEV_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16) +#define end_that_request_last(req, uptodate) end_that_request_last(req) +#endif + +#if defined(_LINUX_CAPABILITY_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16) +#define capable(cap) (1) +#endif + +#if defined(_LINUX_KERNEL_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18) +extern char *kasprintf(gfp_t gfp, const char *fmt, ...) + __attribute__ ((format (printf, 2, 3))); +#endif + +#if defined(_LINUX_SYSRQ_H) && LINUX_VERSION_CODE > KERNEL_VERSION(2,6,18) +#define handle_sysrq(x,y,z) handle_sysrq(x,y) +#endif + +#if defined(_PAGE_PRESENT) && !defined(_PAGE_NX) +#define _PAGE_NX 0 +/* + * This variable at present is referenced by netfront, but only in code that + * is dead when running in hvm guests. To detect potential active uses of it + * in the future, don't try to supply a 'valid' value here, so that any + * mappings created with it will fault when accessed. + */ +#define __supported_pte_mask ((maddr_t)0) +#endif + +/* This code duplication is not ideal, but || does not seem to properly + * short circuit in a #if condition. + **/ +#if defined(_LINUX_NETDEVICE_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18) +#if !defined(SLE_VERSION) +#define netif_tx_lock_bh(dev) spin_lock_bh(&(dev)->xmit_lock) +#define netif_tx_unlock_bh(dev) spin_unlock_bh(&(dev)->xmit_lock) +#elif SLE_VERSION_CODE < SLE_VERSION(10,1,0) +#define netif_tx_lock_bh(dev) spin_lock_bh(&(dev)->xmit_lock) +#define netif_tx_unlock_bh(dev) spin_unlock_bh(&(dev)->xmit_lock) +#endif +#endif + +#if defined(__LINUX_SEQLOCK_H) && !defined(DEFINE_SEQLOCK) +#define DEFINE_SEQLOCK(x) seqlock_t x = SEQLOCK_UNLOCKED +#endif + +/* Bug in RHEL4-U3: rw_lock_t is mistakenly defined in DEFINE_RWLOCK() macro */ +#if defined(__LINUX_SPINLOCK_H) && defined(DEFINE_RWLOCK) +#define rw_lock_t rwlock_t +#endif + +#if defined(__LINUX_SPINLOCK_H) && !defined(DEFINE_RWLOCK) +#define DEFINE_RWLOCK(x) rwlock_t x = RW_LOCK_UNLOCKED +#endif + +#if defined(_LINUX_INTERRUPT_H) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19) +/** + * RHEL4-U5 pulled back this feature into the older kernel + * Since it is a typedef, and not a macro - detect this kernel via + * RHEL_VERSION + */ +#if !defined(RHEL_VERSION) || (RHEL_VERSION == 4 && RHEL_UPDATE < 5) +#if !defined(RHEL_MAJOR) || (RHEL_MAJOR == 4 && RHEL_MINOR < 5) +typedef irqreturn_t (*irq_handler_t)(int, void *, struct pt_regs *); +#endif +#endif +#endif + +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,23) +#define setup_xen_features xen_setup_features +#endif + +#ifndef atomic_cmpxchg +#define atomic_cmpxchg(v, old, new) (cmpxchg(&((v)->counter), (old), (new))) +#endif + +#endif diff --git a/include/xenpvdrivers/public/evtchn.h b/include/xenpvdrivers/public/evtchn.h new file mode 100644 index 0000000..938d4da --- /dev/null +++ b/include/xenpvdrivers/public/evtchn.h @@ -0,0 +1,88 @@ +/****************************************************************************** + * evtchn.h + * + * Interface to /dev/xen/evtchn. + * + * Copyright (c) 2003-2005, K A Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __LINUX_PUBLIC_EVTCHN_H__ +#define __LINUX_PUBLIC_EVTCHN_H__ + +/* + * Bind a fresh port to VIRQ @virq. + * Return allocated port. + */ +#define IOCTL_EVTCHN_BIND_VIRQ \ + _IOC(_IOC_NONE, 'E', 0, sizeof(struct ioctl_evtchn_bind_virq)) +struct ioctl_evtchn_bind_virq { + unsigned int virq; +}; + +/* + * Bind a fresh port to remote <@remote_domain, @remote_port>. + * Return allocated port. + */ +#define IOCTL_EVTCHN_BIND_INTERDOMAIN \ + _IOC(_IOC_NONE, 'E', 1, sizeof(struct ioctl_evtchn_bind_interdomain)) +struct ioctl_evtchn_bind_interdomain { + unsigned int remote_domain, remote_port; +}; + +/* + * Allocate a fresh port for binding to @remote_domain. + * Return allocated port. + */ +#define IOCTL_EVTCHN_BIND_UNBOUND_PORT \ + _IOC(_IOC_NONE, 'E', 2, sizeof(struct ioctl_evtchn_bind_unbound_port)) +struct ioctl_evtchn_bind_unbound_port { + unsigned int remote_domain; +}; + +/* + * Unbind previously allocated @port. + */ +#define IOCTL_EVTCHN_UNBIND \ + _IOC(_IOC_NONE, 'E', 3, sizeof(struct ioctl_evtchn_unbind)) +struct ioctl_evtchn_unbind { + unsigned int port; +}; + +/* + * Unbind previously allocated @port. + */ +#define IOCTL_EVTCHN_NOTIFY \ + _IOC(_IOC_NONE, 'E', 4, sizeof(struct ioctl_evtchn_notify)) +struct ioctl_evtchn_notify { + unsigned int port; +}; + +/* Clear and reinitialise the event buffer. Clear error condition. */ +#define IOCTL_EVTCHN_RESET \ + _IOC(_IOC_NONE, 'E', 5, 0) + +#endif /* __LINUX_PUBLIC_EVTCHN_H__ */ diff --git a/include/xenpvdrivers/public/gntdev.h b/include/xenpvdrivers/public/gntdev.h new file mode 100644 index 0000000..8bd1467 --- /dev/null +++ b/include/xenpvdrivers/public/gntdev.h @@ -0,0 +1,119 @@ +/****************************************************************************** + * gntdev.h + * + * Interface to /dev/xen/gntdev. + * + * Copyright (c) 2007, D G Murray + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __LINUX_PUBLIC_GNTDEV_H__ +#define __LINUX_PUBLIC_GNTDEV_H__ + +struct ioctl_gntdev_grant_ref { + /* The domain ID of the grant to be mapped. */ + uint32_t domid; + /* The grant reference of the grant to be mapped. */ + uint32_t ref; +}; + +/* + * Inserts the grant references into the mapping table of an instance + * of gntdev. N.B. This does not perform the mapping, which is deferred + * until mmap() is called with @index as the offset. + */ +#define IOCTL_GNTDEV_MAP_GRANT_REF \ +_IOC(_IOC_NONE, 'G', 0, sizeof(struct ioctl_gntdev_map_grant_ref)) +struct ioctl_gntdev_map_grant_ref { + /* IN parameters */ + /* The number of grants to be mapped. */ + uint32_t count; + uint32_t pad; + /* OUT parameters */ + /* The offset to be used on a subsequent call to mmap(). */ + uint64_t index; + /* Variable IN parameter. */ + /* Array of grant references, of size @count. */ + struct ioctl_gntdev_grant_ref refs[1]; +}; + +/* + * Removes the grant references from the mapping table of an instance of + * of gntdev. N.B. munmap() must be called on the relevant virtual address(es) + * before this ioctl is called, or an error will result. + */ +#define IOCTL_GNTDEV_UNMAP_GRANT_REF \ +_IOC(_IOC_NONE, 'G', 1, sizeof(struct ioctl_gntdev_unmap_grant_ref)) +struct ioctl_gntdev_unmap_grant_ref { + /* IN parameters */ + /* The offset was returned by the corresponding map operation. */ + uint64_t index; + /* The number of pages to be unmapped. */ + uint32_t count; + uint32_t pad; +}; + +/* + * Returns the offset in the driver's address space that corresponds + * to @vaddr. This can be used to perform a munmap(), followed by an + * UNMAP_GRANT_REF ioctl, where no state about the offset is retained by + * the caller. The number of pages that were allocated at the same time as + * @vaddr is returned in @count. + * + * N.B. Where more than one page has been mapped into a contiguous range, the + * supplied @vaddr must correspond to the start of the range; otherwise + * an error will result. It is only possible to munmap() the entire + * contiguously-allocated range at once, and not any subrange thereof. + */ +#define IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR \ +_IOC(_IOC_NONE, 'G', 2, sizeof(struct ioctl_gntdev_get_offset_for_vaddr)) +struct ioctl_gntdev_get_offset_for_vaddr { + /* IN parameters */ + /* The virtual address of the first mapped page in a range. */ + uint64_t vaddr; + /* OUT parameters */ + /* The offset that was used in the initial mmap() operation. */ + uint64_t offset; + /* The number of pages mapped in the VM area that begins at @vaddr. */ + uint32_t count; + uint32_t pad; +}; + +/* + * Sets the maximum number of grants that may mapped at once by this gntdev + * instance. + * + * N.B. This must be called before any other ioctl is performed on the device. + */ +#define IOCTL_GNTDEV_SET_MAX_GRANTS \ +_IOC(_IOC_NONE, 'G', 3, sizeof(struct ioctl_gntdev_set_max_grants)) +struct ioctl_gntdev_set_max_grants { + /* IN parameter */ + /* The maximum number of grants that may be mapped at once. */ + uint32_t count; +}; + +#endif /* __LINUX_PUBLIC_GNTDEV_H__ */ diff --git a/include/xenpvdrivers/public/privcmd.h b/include/xenpvdrivers/public/privcmd.h new file mode 100644 index 0000000..aacbe19 --- /dev/null +++ b/include/xenpvdrivers/public/privcmd.h @@ -0,0 +1,85 @@ +/****************************************************************************** + * privcmd.h + * + * Interface to /proc/xen/privcmd. + * + * Copyright (c) 2003-2005, K A Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __LINUX_PUBLIC_PRIVCMD_H__ +#define __LINUX_PUBLIC_PRIVCMD_H__ + +#include + +typedef struct privcmd_hypercall +{ + __u64 op; + __u64 arg[5]; +} privcmd_hypercall_t; + +typedef struct privcmd_mmap_entry { + __u64 va; + __u64 mfn; + __u64 npages; +} privcmd_mmap_entry_t; + +typedef struct privcmd_mmap { + int num; + domid_t dom; /* target domain */ + privcmd_mmap_entry_t __user *entry; +} privcmd_mmap_t; + +typedef struct privcmd_mmapbatch { + int num; /* number of pages to populate */ + domid_t dom; /* target domain */ + __u64 addr; /* virtual address */ + xen_pfn_t __user *arr; /* array of mfns - top nibble set on err */ +} privcmd_mmapbatch_t; + +typedef struct privcmd_mmapbatch_v2 { + unsigned int num; /* number of pages to populate */ + domid_t dom; /* target domain */ + __u64 addr; /* virtual address */ + const xen_pfn_t __user *arr; /* array of mfns */ + int __user *err; /* array of error codes */ +} privcmd_mmapbatch_v2_t; + +/* + * @cmd: IOCTL_PRIVCMD_HYPERCALL + * @arg: &privcmd_hypercall_t + * Return: Value returned from execution of the specified hypercall. + */ +#define IOCTL_PRIVCMD_HYPERCALL \ + _IOC(_IOC_NONE, 'P', 0, sizeof(privcmd_hypercall_t)) +#define IOCTL_PRIVCMD_MMAP \ + _IOC(_IOC_NONE, 'P', 2, sizeof(privcmd_mmap_t)) +#define IOCTL_PRIVCMD_MMAPBATCH \ + _IOC(_IOC_NONE, 'P', 3, sizeof(privcmd_mmapbatch_t)) +#define IOCTL_PRIVCMD_MMAPBATCH_V2 \ + _IOC(_IOC_NONE, 'P', 4, sizeof(privcmd_mmapbatch_v2_t)) + +#endif /* __LINUX_PUBLIC_PRIVCMD_H__ */ diff --git a/include/xenpvdrivers/public/xenbus.h b/include/xenpvdrivers/public/xenbus.h new file mode 100644 index 0000000..fd61373 --- /dev/null +++ b/include/xenpvdrivers/public/xenbus.h @@ -0,0 +1,52 @@ +/****************************************************************************** + * xenbus.h + * + * Interface to /proc/xen/xenbus. + * + * Copyright (c) 2008, Diego Ongaro + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __LINUX_PUBLIC_XENBUS_H__ +#define __LINUX_PUBLIC_XENBUS_H__ + +#include + +typedef struct xenbus_alloc { + domid_t dom; + __u32 port; + __u32 grant_ref; +} xenbus_alloc_t; + +/* + * @cmd: IOCTL_XENBUS_ALLOC + * @arg: &xenbus_alloc_t + * Return: 0, or -1 for error + */ +#define IOCTL_XENBUS_ALLOC \ + _IOC(_IOC_NONE, 'X', 0, sizeof(xenbus_alloc_t)) + +#endif /* __LINUX_PUBLIC_XENBUS_H__ */ diff --git a/include/xenpvdrivers/sysctl.h b/include/xenpvdrivers/sysctl.h new file mode 100644 index 0000000..7fe9250 --- /dev/null +++ b/include/xenpvdrivers/sysctl.h @@ -0,0 +1,11 @@ +#ifndef _XEN_SYSCTL_H +#define _XEN_SYSCTL_H + +/* CTL_XEN names: */ +enum +{ + CTL_XEN_INDEPENDENT_WALLCLOCK=1, + CTL_XEN_PERMITTED_CLOCK_JITTER=2, +}; + +#endif /* _XEN_SYSCTL_H */ diff --git a/include/xenpvdrivers/xen-ops.h b/include/xenpvdrivers/xen-ops.h new file mode 100644 index 0000000..883a21b --- /dev/null +++ b/include/xenpvdrivers/xen-ops.h @@ -0,0 +1,17 @@ +#ifndef INCLUDE_XEN_OPS_H +#define INCLUDE_XEN_OPS_H + +#include + +DECLARE_PER_CPU(struct vcpu_info *, xen_vcpu); + +void xen_pre_suspend(void); +void xen_post_suspend(int suspend_cancelled); + +void xen_mm_pin_all(void); +void xen_mm_unpin_all(void); + +void xen_timer_resume(void); +void xen_arch_resume(void); + +#endif /* INCLUDE_XEN_OPS_H */ diff --git a/include/xenpvdrivers/xen_proc.h b/include/xenpvdrivers/xen_proc.h new file mode 100644 index 0000000..44af17c --- /dev/null +++ b/include/xenpvdrivers/xen_proc.h @@ -0,0 +1,12 @@ + +#ifndef __ASM_XEN_PROC_H__ +#define __ASM_XEN_PROC_H__ + +#include + +extern struct proc_dir_entry *create_xen_proc_entry( + const char *name, mode_t mode); +extern void remove_xen_proc_entry( + const char *name); + +#endif /* __ASM_XEN_PROC_H__ */ diff --git a/include/xenpvdrivers/xenbus.h b/include/xenpvdrivers/xenbus.h new file mode 100644 index 0000000..eb1ed5b --- /dev/null +++ b/include/xenpvdrivers/xenbus.h @@ -0,0 +1,340 @@ +/****************************************************************************** + * xenbus.h + * + * Talks to Xen Store to figure out what devices we have. + * + * Copyright (C) 2005 Rusty Russell, IBM Corporation + * Copyright (C) 2005 XenSource Ltd. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef _XEN_XENBUS_H +#define _XEN_XENBUS_H + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* Register callback to watch this node. */ +struct xenbus_watch +{ + struct list_head list; + + /* Path being watched. */ + const char *node; + + /* Callback (executed in a process context with no locks held). */ + void (*callback)(struct xenbus_watch *, + const char **vec, unsigned int len); + +#if defined(CONFIG_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H) + /* See XBWF_ definitions below. */ + unsigned long flags; +#endif +}; + +#if defined(CONFIG_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H) +/* + * Execute callback in its own kthread. Useful if the callback is long + * running or heavily serialised, to avoid taking out the main xenwatch thread + * for a long period of time (or even unwittingly causing a deadlock). + */ +#define XBWF_new_thread 1 +#endif + +/* A xenbus device. */ +struct xenbus_device { + const char *devicetype; + const char *nodename; + const char *otherend; + int otherend_id; + struct xenbus_watch otherend_watch; + struct device dev; + enum xenbus_state state; + struct completion down; +}; + +static inline struct xenbus_device *to_xenbus_device(struct device *dev) +{ + return container_of(dev, struct xenbus_device, dev); +} + +struct xenbus_device_id +{ + /* .../device// */ + char devicetype[32]; /* General class of device. */ +}; + +/* A xenbus driver. */ +struct xenbus_driver { + const char *name; + const struct xenbus_device_id *ids; + int (*probe)(struct xenbus_device *dev, + const struct xenbus_device_id *id); + void (*otherend_changed)(struct xenbus_device *dev, + enum xenbus_state backend_state); + int (*remove)(struct xenbus_device *dev); +#if !defined(CONFIG_XEN) && !defined(XENPV_COMPAT_MODULE) + int (*suspend)(struct xenbus_device *dev, pm_message_t state); +#else + int (*suspend)(struct xenbus_device *dev); + int (*suspend_cancel)(struct xenbus_device *dev); +#endif + int (*resume)(struct xenbus_device *dev); + int (*uevent)(struct xenbus_device *, struct kobj_uevent_env *); + struct device_driver driver; + int (*read_otherend_details)(struct xenbus_device *dev); + int (*is_ready)(struct xenbus_device *dev); +}; + +static inline struct xenbus_driver *to_xenbus_driver(struct device_driver *drv) +{ + return container_of(drv, struct xenbus_driver, driver); +} + +int __must_check __xenbus_register_frontend(struct xenbus_driver *drv, + struct module *owner, + const char *mod_name); + +static inline int __must_check +xenbus_register_frontend(struct xenbus_driver *drv) +{ + return __xenbus_register_frontend(drv, THIS_MODULE, KBUILD_MODNAME); +} + +int __must_check __xenbus_register_backend(struct xenbus_driver *drv, + struct module *owner, + const char *mod_name); +static inline int __must_check +xenbus_register_backend(struct xenbus_driver *drv) +{ + return __xenbus_register_backend(drv, THIS_MODULE, KBUILD_MODNAME); +} + +void xenbus_unregister_driver(struct xenbus_driver *drv); + +struct xenbus_transaction +{ + u32 id; +}; + +/* Nil transaction ID. */ +#define XBT_NIL ((struct xenbus_transaction) { 0 }) + +char **xenbus_directory(struct xenbus_transaction t, + const char *dir, const char *node, unsigned int *num); +void *xenbus_read(struct xenbus_transaction t, + const char *dir, const char *node, unsigned int *len); +int xenbus_write(struct xenbus_transaction t, + const char *dir, const char *node, const char *string); +int xenbus_mkdir(struct xenbus_transaction t, + const char *dir, const char *node); +int xenbus_exists(struct xenbus_transaction t, + const char *dir, const char *node); +int xenbus_rm(struct xenbus_transaction t, const char *dir, const char *node); +int xenbus_transaction_start(struct xenbus_transaction *t); +int xenbus_transaction_end(struct xenbus_transaction t, int abort); + +/* Single read and scanf: returns -errno or num scanned if > 0. */ +int xenbus_scanf(struct xenbus_transaction t, + const char *dir, const char *node, const char *fmt, ...) + __attribute__((format(scanf, 4, 5))); + +/* Single printf and write: returns -errno or 0. */ +int xenbus_printf(struct xenbus_transaction t, + const char *dir, const char *node, const char *fmt, ...) + __attribute__((format(printf, 4, 5))); + +/* Generic read function: NULL-terminated triples of name, + * sprintf-style type string, and pointer. Returns 0 or errno.*/ +int xenbus_gather(struct xenbus_transaction t, const char *dir, ...); + +/* notifer routines for when the xenstore comes up */ +int register_xenstore_notifier(struct notifier_block *nb); +void unregister_xenstore_notifier(struct notifier_block *nb); + +int register_xenbus_watch(struct xenbus_watch *watch); +void unregister_xenbus_watch(struct xenbus_watch *watch); +void xs_suspend(void); +void xs_resume(void); +void xs_suspend_cancel(void); + +/* Used by xenbus_dev to borrow kernel's store connection. */ +void *xenbus_dev_request_and_reply(struct xsd_sockmsg *msg); + +/* Prepare for domain suspend: then resume or cancel the suspend. */ +void xenbus_suspend(void); +void xenbus_resume(void); +void xenbus_suspend_cancel(void); + +#define XENBUS_IS_ERR_READ(str) ({ \ + if (!IS_ERR(str) && strlen(str) == 0) { \ + kfree(str); \ + str = ERR_PTR(-ERANGE); \ + } \ + IS_ERR(str); \ +}) + +#define XENBUS_EXIST_ERR(err) ((err) == -ENOENT || (err) == -ERANGE) + + +/** + * Register a watch on the given path, using the given xenbus_watch structure + * for storage, and the given callback function as the callback. Return 0 on + * success, or -errno on error. On success, the given path will be saved as + * watch->node, and remains the caller's to free. On error, watch->node will + * be NULL, the device will switch to XenbusStateClosing, and the error will + * be saved in the store. + */ +int xenbus_watch_path(struct xenbus_device *dev, const char *path, + struct xenbus_watch *watch, + void (*callback)(struct xenbus_watch *, + const char **, unsigned int)); + + +#if defined(CONFIG_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H) +/** + * Register a watch on the given path/path2, using the given xenbus_watch + * structure for storage, and the given callback function as the callback. + * Return 0 on success, or -errno on error. On success, the watched path + * (path/path2) will be saved as watch->node, and becomes the caller's to + * kfree(). On error, watch->node will be NULL, so the caller has nothing to + * free, the device will switch to XenbusStateClosing, and the error will be + * saved in the store. + */ +int xenbus_watch_path2(struct xenbus_device *dev, const char *path, + const char *path2, struct xenbus_watch *watch, + void (*callback)(struct xenbus_watch *, + const char **, unsigned int)); +#else +int xenbus_watch_pathfmt(struct xenbus_device *dev, struct xenbus_watch *watch, + void (*callback)(struct xenbus_watch *, + const char **, unsigned int), + const char *pathfmt, ...) + __attribute__ ((format (printf, 4, 5))); +#endif + +/** + * Advertise in the store a change of the given driver to the given new_state. + * Return 0 on success, or -errno on error. On error, the device will switch + * to XenbusStateClosing, and the error will be saved in the store. + */ +int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state new_state); + + +/** + * Grant access to the given ring_mfn to the peer of the given device. Return + * 0 on success, or -errno on error. On error, the device will switch to + * XenbusStateClosing, and the error will be saved in the store. + */ +int xenbus_grant_ring(struct xenbus_device *dev, unsigned long ring_mfn); + + +/** + * Map a page of memory into this domain from another domain's grant table. + * xenbus_map_ring_valloc allocates a page of virtual address space, maps the + * page to that address, and sets *vaddr to that address. + * xenbus_map_ring does not allocate the virtual address space (you must do + * this yourself!). It only maps in the page to the specified address. + * Returns 0 on success, and GNTST_* (see xen/include/interface/grant_table.h) + * or -ENOMEM on error. If an error is returned, device will switch to + * XenbusStateClosing and the error message will be saved in XenStore. + */ +struct vm_struct *xenbus_map_ring_valloc(struct xenbus_device *dev, + int gnt_ref); +int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref, + grant_handle_t *handle, void *vaddr); + + +/** + * Unmap a page of memory in this domain that was imported from another domain. + * Use xenbus_unmap_ring_vfree if you mapped in your memory with + * xenbus_map_ring_valloc (it will free the virtual address space). + * Returns 0 on success and returns GNTST_* on error + * (see xen/include/interface/grant_table.h). + */ +int xenbus_unmap_ring_vfree(struct xenbus_device *dev, struct vm_struct *); +int xenbus_unmap_ring(struct xenbus_device *dev, + grant_handle_t handle, void *vaddr); + + +/** + * Allocate an event channel for the given xenbus_device, assigning the newly + * created local port to *port. Return 0 on success, or -errno on error. On + * error, the device will switch to XenbusStateClosing, and the error will be + * saved in the store. + */ +int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port); + + +/** + * Free an existing event channel. Returns 0 on success or -errno on error. + */ +int xenbus_free_evtchn(struct xenbus_device *dev, int port); + + +/** + * Return the state of the driver rooted at the given store path, or + * XenbusStateUnknown if no state can be read. + */ +enum xenbus_state xenbus_read_driver_state(const char *path); + + +/*** + * Report the given negative errno into the store, along with the given + * formatted message. + */ +void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt, + ...); + + +/*** + * Equivalent to xenbus_dev_error(dev, err, fmt, args), followed by + * xenbus_switch_state(dev, NULL, XenbusStateClosing) to schedule an orderly + * closedown of this driver and its peer. + */ +void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt, + ...); + +#if defined(CONFIG_XEN) || defined(XENPV_COMPAT_MODULE) +int xenbus_dev_init(void); +#endif + +const char *xenbus_strstate(enum xenbus_state state); +int xenbus_dev_is_online(struct xenbus_device *dev); +int xenbus_frontend_closed(struct xenbus_device *dev); + +int xenbus_for_each_backend(void *arg, int (*fn)(struct device *, void *)); +int xenbus_for_each_frontend(void *arg, int (*fn)(struct device *, void *)); + +#endif /* _XEN_XENBUS_H */ diff --git a/include/xenpvdrivers/xencomm.h b/include/xenpvdrivers/xencomm.h new file mode 100644 index 0000000..86eebed --- /dev/null +++ b/include/xenpvdrivers/xencomm.h @@ -0,0 +1,77 @@ +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Copyright (C) IBM Corp. 2006 + * + * Authors: Hollis Blanchard + * Jerone Young + */ + +#ifndef _LINUX_XENCOMM_H_ +#define _LINUX_XENCOMM_H_ + +#include + +#define XENCOMM_MINI_ADDRS 3 +struct xencomm_mini { + struct xencomm_desc _desc; + uint64_t address[XENCOMM_MINI_ADDRS]; +}; + +/* To avoid additionnal virt to phys conversion, an opaque structure is + presented. */ +struct xencomm_handle; + +extern void xencomm_free(struct xencomm_handle *desc); +extern struct xencomm_handle *xencomm_map(void *ptr, unsigned long bytes); +extern struct xencomm_handle *__xencomm_map_no_alloc(void *ptr, + unsigned long bytes, struct xencomm_mini *xc_area); + +#if 0 +#define XENCOMM_MINI_ALIGNED(xc_desc, n) \ + struct xencomm_mini xc_desc ## _base[(n)] \ + __attribute__((__aligned__(sizeof(struct xencomm_mini)))); \ + struct xencomm_mini *xc_desc = &xc_desc ## _base[0]; +#else +/* + * gcc bug workaround: + * http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16660 + * gcc doesn't handle properly stack variable with + * __attribute__((__align__(sizeof(struct xencomm_mini)))) + */ +#define XENCOMM_MINI_ALIGNED(xc_desc, n) \ + unsigned char xc_desc ## _base[((n) + 1 ) * \ + sizeof(struct xencomm_mini)]; \ + struct xencomm_mini *xc_desc = (struct xencomm_mini *) \ + ((unsigned long)xc_desc ## _base + \ + (sizeof(struct xencomm_mini) - \ + ((unsigned long)xc_desc ## _base) % \ + sizeof(struct xencomm_mini))); +#endif +#define xencomm_map_no_alloc(ptr, bytes) \ + ({ XENCOMM_MINI_ALIGNED(xc_desc, 1); \ + __xencomm_map_no_alloc(ptr, bytes, xc_desc); }) + +/* provided by architecture code: */ +extern unsigned long xencomm_vtop(unsigned long vaddr); + +static inline void *xencomm_pa(void *ptr) +{ + return (void *)xencomm_vtop((unsigned long)ptr); +} + +#define xen_guest_handle(hnd) ((hnd).p) + +#endif /* _LINUX_XENCOMM_H_ */ diff --git a/include/xenpvdrivers/xencons.h b/include/xenpvdrivers/xencons.h new file mode 100644 index 0000000..c5a55d6 --- /dev/null +++ b/include/xenpvdrivers/xencons.h @@ -0,0 +1,17 @@ +#ifndef __ASM_XENCONS_H__ +#define __ASM_XENCONS_H__ + +struct dom0_vga_console_info; +void dom0_init_screen_info(const struct dom0_vga_console_info *, size_t); + +void xencons_force_flush(void); +void xencons_resume(void); + +/* Interrupt work hooks. Receive data, or kick data out. */ +void xencons_rx(char *buf, unsigned len); +void xencons_tx(void); + +int xencons_ring_init(void); +int xencons_ring_send(const char *data, unsigned len); + +#endif /* __ASM_XENCONS_H__ */ diff --git a/include/xenpvdrivers/xenoprof.h b/include/xenpvdrivers/xenoprof.h new file mode 100644 index 0000000..4c3ab0f --- /dev/null +++ b/include/xenpvdrivers/xenoprof.h @@ -0,0 +1,42 @@ +/****************************************************************************** + * xen/xenoprof.h + * + * Copyright (c) 2006 Isaku Yamahata + * VA Linux Systems Japan K.K. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + */ + +#ifndef __XEN_XENOPROF_H__ +#define __XEN_XENOPROF_H__ +#ifdef CONFIG_XEN + +#include + +struct oprofile_operations; +int xenoprofile_init(struct oprofile_operations * ops); +void xenoprofile_exit(void); + +struct xenoprof_shared_buffer { + char *buffer; + struct xenoprof_arch_shared_buffer arch; +}; +#else +#define xenoprofile_init(ops) (-ENOSYS) +#define xenoprofile_exit() do { } while (0) + +#endif /* CONFIG_XEN */ +#endif /* __XEN_XENOPROF_H__ */