diff options
Diffstat (limited to 'drivers/xen')
| -rw-r--r-- | drivers/xen/Kconfig | 30 | ||||
| -rw-r--r-- | drivers/xen/Makefile | 11 | ||||
| -rw-r--r-- | drivers/xen/balloon.c | 375 | ||||
| -rw-r--r-- | drivers/xen/events.c | 784 | ||||
| -rw-r--r-- | drivers/xen/gntalloc.c | 545 | ||||
| -rw-r--r-- | drivers/xen/gntdev.c | 751 | ||||
| -rw-r--r-- | drivers/xen/grant-table.c | 56 | ||||
| -rw-r--r-- | drivers/xen/manage.c | 155 | ||||
| -rw-r--r-- | drivers/xen/platform-pci.c | 24 | ||||
| -rw-r--r-- | drivers/xen/xen-balloon.c | 256 | ||||
| -rw-r--r-- | drivers/xen/xenbus/Makefile | 5 | ||||
| -rw-r--r-- | drivers/xen/xenbus/xenbus_probe.c | 361 | ||||
| -rw-r--r-- | drivers/xen/xenbus/xenbus_probe.h | 32 | ||||
| -rw-r--r-- | drivers/xen/xenbus/xenbus_probe_backend.c | 276 | ||||
| -rw-r--r-- | drivers/xen/xenbus/xenbus_probe_frontend.c | 301 | ||||
| -rw-r--r-- | drivers/xen/xenfs/xenbus.c | 31 |
16 files changed, 2968 insertions, 1025 deletions
diff --git a/drivers/xen/Kconfig b/drivers/xen/Kconfig index 6e6180ccd72..a59638b37c1 100644 --- a/drivers/xen/Kconfig +++ b/drivers/xen/Kconfig @@ -29,6 +29,14 @@ config XEN_DEV_EVTCHN firing. If in doubt, say yes. +config XEN_BACKEND + bool "Backend driver support" + depends on XEN_DOM0 + default y + help + Support for backend device drivers that provide I/O services + to other virtual machines. + config XENFS tristate "Xen filesystem" default y @@ -62,9 +70,29 @@ config XEN_SYS_HYPERVISOR virtual environment, /sys/hypervisor will still be present, but will have no xen contents. +config XEN_XENBUS_FRONTEND + tristate + +config XEN_GNTDEV + tristate "userspace grant access device driver" + depends on XEN + default m + select MMU_NOTIFIER + help + Allows userspace processes to use grants. + +config XEN_GRANT_DEV_ALLOC + tristate "User-space grant reference allocator driver" + depends on XEN + default m + help + Allows userspace processes to create pages with access granted + to other domains. This can be used to implement frontend drivers + or as part of an inter-domain shared memory channel. + config XEN_PLATFORM_PCI tristate "xen platform pci device driver" - depends on XEN_PVHVM + depends on XEN_PVHVM && PCI default m help Driver for the Xen PCI Platform device: it is responsible for diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile index 533a199e7a3..f420f1ff7f1 100644 --- a/drivers/xen/Makefile +++ b/drivers/xen/Makefile @@ -1,4 +1,4 @@ -obj-y += grant-table.o features.o events.o manage.o +obj-y += grant-table.o features.o events.o manage.o balloon.o obj-y += xenbus/ nostackp := $(call cc-option, -fno-stack-protector) @@ -7,13 +7,18 @@ CFLAGS_features.o := $(nostackp) obj-$(CONFIG_BLOCK) += biomerge.o obj-$(CONFIG_HOTPLUG_CPU) += cpu_hotplug.o obj-$(CONFIG_XEN_XENCOMM) += xencomm.o -obj-$(CONFIG_XEN_BALLOON) += balloon.o +obj-$(CONFIG_XEN_BALLOON) += xen-balloon.o obj-$(CONFIG_XEN_DEV_EVTCHN) += xen-evtchn.o +obj-$(CONFIG_XEN_GNTDEV) += xen-gntdev.o +obj-$(CONFIG_XEN_GRANT_DEV_ALLOC) += xen-gntalloc.o obj-$(CONFIG_XENFS) += xenfs/ obj-$(CONFIG_XEN_SYS_HYPERVISOR) += sys-hypervisor.o -obj-$(CONFIG_XEN_PLATFORM_PCI) += platform-pci.o +obj-$(CONFIG_XEN_PLATFORM_PCI) += xen-platform-pci.o obj-$(CONFIG_SWIOTLB_XEN) += swiotlb-xen.o obj-$(CONFIG_XEN_DOM0) += pci.o xen-evtchn-y := evtchn.o +xen-gntdev-y := gntdev.o +xen-gntalloc-y := gntalloc.o +xen-platform-pci-y := platform-pci.o diff --git a/drivers/xen/balloon.c b/drivers/xen/balloon.c index 43f9f02c7db..043af8ad6b6 100644 --- a/drivers/xen/balloon.c +++ b/drivers/xen/balloon.c @@ -1,6 +1,4 @@ /****************************************************************************** - * balloon.c - * * Xen balloon driver - enables returning/claiming memory to/from Xen. * * Copyright (c) 2003, B Dragovic @@ -33,7 +31,6 @@ */ #include <linux/kernel.h> -#include <linux/module.h> #include <linux/sched.h> #include <linux/errno.h> #include <linux/mm.h> @@ -42,13 +39,11 @@ #include <linux/highmem.h> #include <linux/mutex.h> #include <linux/list.h> -#include <linux/sysdev.h> #include <linux/gfp.h> #include <asm/page.h> #include <asm/pgalloc.h> #include <asm/pgtable.h> -#include <asm/uaccess.h> #include <asm/tlb.h> #include <asm/e820.h> @@ -58,35 +53,29 @@ #include <xen/xen.h> #include <xen/interface/xen.h> #include <xen/interface/memory.h> -#include <xen/xenbus.h> +#include <xen/balloon.h> #include <xen/features.h> #include <xen/page.h> -#define PAGES2KB(_p) ((_p)<<(PAGE_SHIFT-10)) - -#define BALLOON_CLASS_NAME "xen_memory" +/* + * balloon_process() state: + * + * BP_DONE: done or nothing to do, + * BP_EAGAIN: error, go to sleep, + * BP_ECANCELED: error, balloon operation canceled. + */ -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; +enum bp_state { + BP_DONE, + BP_EAGAIN, + BP_ECANCELED }; -static DEFINE_MUTEX(balloon_mutex); - -static struct sys_device balloon_sysdev; -static int register_balloon(struct sys_device *sysdev); +static DEFINE_MUTEX(balloon_mutex); -static struct balloon_stats balloon_stats; +struct balloon_stats balloon_stats; +EXPORT_SYMBOL_GPL(balloon_stats); /* We increase/decrease in batches which fit in a page */ static unsigned long frame_list[PAGE_SIZE / sizeof(unsigned long)]; @@ -104,8 +93,7 @@ static LIST_HEAD(ballooned_pages); /* Main work function, always executed in process context. */ static void balloon_process(struct work_struct *work); -static DECLARE_WORK(balloon_worker, balloon_process); -static struct timer_list balloon_timer; +static DECLARE_DELAYED_WORK(balloon_worker, balloon_process); /* 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. */ @@ -140,14 +128,17 @@ static void balloon_append(struct page *page) } /* balloon_retrieve: rescue a page from the balloon, if it is not empty. */ -static struct page *balloon_retrieve(void) +static struct page *balloon_retrieve(bool prefer_highmem) { struct page *page; if (list_empty(&ballooned_pages)) return NULL; - page = list_entry(ballooned_pages.next, struct page, lru); + if (prefer_highmem) + page = list_entry(ballooned_pages.prev, struct page, lru); + else + page = list_entry(ballooned_pages.next, struct page, lru); list_del(&page->lru); if (PageHighMem(page)) { @@ -177,9 +168,29 @@ static struct page *balloon_next_page(struct page *page) return list_entry(next, struct page, lru); } -static void balloon_alarm(unsigned long unused) +static enum bp_state update_schedule(enum bp_state state) { - schedule_work(&balloon_worker); + if (state == BP_DONE) { + balloon_stats.schedule_delay = 1; + balloon_stats.retry_count = 1; + return BP_DONE; + } + + ++balloon_stats.retry_count; + + if (balloon_stats.max_retry_count != RETRY_UNLIMITED && + balloon_stats.retry_count > balloon_stats.max_retry_count) { + balloon_stats.schedule_delay = 1; + balloon_stats.retry_count = 1; + return BP_ECANCELED; + } + + balloon_stats.schedule_delay <<= 1; + + if (balloon_stats.schedule_delay > balloon_stats.max_schedule_delay) + balloon_stats.schedule_delay = balloon_stats.max_schedule_delay; + + return BP_EAGAIN; } static unsigned long current_target(void) @@ -194,11 +205,11 @@ static unsigned long current_target(void) return target; } -static int increase_reservation(unsigned long nr_pages) +static enum bp_state increase_reservation(unsigned long nr_pages) { + int rc; unsigned long pfn, i; struct page *page; - long rc; struct xen_memory_reservation reservation = { .address_bits = 0, .extent_order = 0, @@ -210,7 +221,10 @@ static int increase_reservation(unsigned long nr_pages) page = balloon_first_page(); for (i = 0; i < nr_pages; i++) { - BUG_ON(page == NULL); + if (!page) { + nr_pages = i; + break; + } frame_list[i] = page_to_pfn(page); page = balloon_next_page(page); } @@ -218,11 +232,11 @@ static int increase_reservation(unsigned long nr_pages) 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; + if (rc <= 0) + return BP_EAGAIN; for (i = 0; i < rc; i++) { - page = balloon_retrieve(); + page = balloon_retrieve(false); BUG_ON(page == NULL); pfn = page_to_pfn(page); @@ -232,7 +246,7 @@ static int increase_reservation(unsigned long nr_pages) set_phys_to_machine(pfn, frame_list[i]); /* Link back into the page tables if not highmem. */ - if (pfn < max_low_pfn) { + if (!xen_hvm_domain() && pfn < max_low_pfn) { int ret; ret = HYPERVISOR_update_va_mapping( (unsigned long)__va(pfn << PAGE_SHIFT), @@ -249,15 +263,14 @@ static int increase_reservation(unsigned long nr_pages) balloon_stats.current_pages += rc; - out: - return rc < 0 ? rc : rc != nr_pages; + return BP_DONE; } -static int decrease_reservation(unsigned long nr_pages) +static enum bp_state decrease_reservation(unsigned long nr_pages, gfp_t gfp) { + enum bp_state state = BP_DONE; unsigned long pfn, i; struct page *page; - int need_sleep = 0; int ret; struct xen_memory_reservation reservation = { .address_bits = 0, @@ -269,9 +282,9 @@ static int decrease_reservation(unsigned long nr_pages) nr_pages = ARRAY_SIZE(frame_list); for (i = 0; i < nr_pages; i++) { - if ((page = alloc_page(GFP_BALLOON)) == NULL) { + if ((page = alloc_page(gfp)) == NULL) { nr_pages = i; - need_sleep = 1; + state = BP_EAGAIN; break; } @@ -280,7 +293,7 @@ static int decrease_reservation(unsigned long nr_pages) scrub_page(page); - if (!PageHighMem(page)) { + if (!xen_hvm_domain() && !PageHighMem(page)) { ret = HYPERVISOR_update_va_mapping( (unsigned long)__va(pfn << PAGE_SHIFT), __pte_ma(0), 0); @@ -296,7 +309,7 @@ static int decrease_reservation(unsigned long nr_pages) /* No more mappings: invalidate P2M and add to balloon. */ for (i = 0; i < nr_pages; i++) { pfn = mfn_to_pfn(frame_list[i]); - set_phys_to_machine(pfn, INVALID_P2M_ENTRY); + __set_phys_to_machine(pfn, INVALID_P2M_ENTRY); balloon_append(pfn_to_page(pfn)); } @@ -307,7 +320,7 @@ static int decrease_reservation(unsigned long nr_pages) balloon_stats.current_pages -= nr_pages; - return need_sleep; + return state; } /* @@ -318,99 +331,125 @@ static int decrease_reservation(unsigned long nr_pages) */ static void balloon_process(struct work_struct *work) { - int need_sleep = 0; + enum bp_state state = BP_DONE; long credit; mutex_lock(&balloon_mutex); do { credit = current_target() - balloon_stats.current_pages; + if (credit > 0) - need_sleep = (increase_reservation(credit) != 0); + state = increase_reservation(credit); + if (credit < 0) - need_sleep = (decrease_reservation(-credit) != 0); + state = decrease_reservation(-credit, GFP_BALLOON); + + state = update_schedule(state); #ifndef CONFIG_PREEMPT if (need_resched()) schedule(); #endif - } while ((credit != 0) && !need_sleep); + } while (credit && state == BP_DONE); /* Schedule more work if there is some still to be done. */ - if (current_target() != balloon_stats.current_pages) - mod_timer(&balloon_timer, jiffies + HZ); + if (state == BP_EAGAIN) + schedule_delayed_work(&balloon_worker, balloon_stats.schedule_delay * HZ); mutex_unlock(&balloon_mutex); } /* Resets the Xen limit, sets new target, and kicks off processing. */ -static void balloon_set_new_target(unsigned long target) +void balloon_set_new_target(unsigned long target) { /* No need for lock. Not read-modify-write updates. */ balloon_stats.target_pages = target; - schedule_work(&balloon_worker); + schedule_delayed_work(&balloon_worker, 0); } +EXPORT_SYMBOL_GPL(balloon_set_new_target); -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) +/** + * alloc_xenballooned_pages - get pages that have been ballooned out + * @nr_pages: Number of pages to get + * @pages: pages returned + * @return 0 on success, error otherwise + */ +int alloc_xenballooned_pages(int nr_pages, struct page** pages) { - 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; + int pgno = 0; + struct page* page; + mutex_lock(&balloon_mutex); + while (pgno < nr_pages) { + page = balloon_retrieve(true); + if (page) { + pages[pgno++] = page; + } else { + enum bp_state st; + st = decrease_reservation(nr_pages - pgno, GFP_HIGHUSER); + if (st != BP_DONE) + goto out_undo; + } } - - /* 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)); + mutex_unlock(&balloon_mutex); + return 0; + out_undo: + while (pgno) + balloon_append(pages[--pgno]); + /* Free the memory back to the kernel soon */ + schedule_delayed_work(&balloon_worker, 0); + mutex_unlock(&balloon_mutex); + return -ENOMEM; } +EXPORT_SYMBOL(alloc_xenballooned_pages); -static int balloon_init_watcher(struct notifier_block *notifier, - unsigned long event, - void *data) +/** + * free_xenballooned_pages - return pages retrieved with get_ballooned_pages + * @nr_pages: Number of pages + * @pages: pages to return + */ +void free_xenballooned_pages(int nr_pages, struct page** pages) { - int err; + int i; - err = register_xenbus_watch(&target_watch); - if (err) - printk(KERN_ERR "Failed to set balloon watcher\n"); + mutex_lock(&balloon_mutex); - return NOTIFY_DONE; -} + for (i = 0; i < nr_pages; i++) { + if (pages[i]) + balloon_append(pages[i]); + } + + /* The balloon may be too large now. Shrink it if needed. */ + if (current_target() != balloon_stats.current_pages) + schedule_delayed_work(&balloon_worker, 0); -static struct notifier_block xenstore_notifier; + mutex_unlock(&balloon_mutex); +} +EXPORT_SYMBOL(free_xenballooned_pages); static int __init balloon_init(void) { - unsigned long pfn, extra_pfn_end; + unsigned long pfn, nr_pages, extra_pfn_end; struct page *page; - if (!xen_pv_domain()) + if (!xen_domain()) return -ENODEV; - pr_info("xen_balloon: Initialising balloon driver.\n"); + pr_info("xen/balloon: Initialising balloon driver.\n"); - balloon_stats.current_pages = min(xen_start_info->nr_pages, max_pfn); + if (xen_pv_domain()) + nr_pages = xen_start_info->nr_pages; + else + nr_pages = max_pfn; + balloon_stats.current_pages = min(nr_pages, max_pfn); balloon_stats.target_pages = balloon_stats.current_pages; balloon_stats.balloon_low = 0; balloon_stats.balloon_high = 0; - balloon_stats.driver_pages = 0UL; - - init_timer(&balloon_timer); - balloon_timer.data = 0; - balloon_timer.function = balloon_alarm; - register_balloon(&balloon_sysdev); + balloon_stats.schedule_delay = 1; + balloon_stats.max_schedule_delay = 32; + balloon_stats.retry_count = 1; + balloon_stats.max_retry_count = RETRY_UNLIMITED; /* * Initialise the balloon with excess memory space. We need @@ -432,153 +471,9 @@ static int __init balloon_init(void) __balloon_append(page); } - 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 balloon_exit(void) -{ - /* XXX - release balloon here */ - return; -} - -module_exit(balloon_exit); - -#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(balloon_stats.current_pages)); -BALLOON_SHOW(low_kb, "%lu\n", PAGES2KB(balloon_stats.balloon_low)); -BALLOON_SHOW(high_kb, "%lu\n", PAGES2KB(balloon_stats.balloon_high)); -BALLOON_SHOW(driver_kb, "%lu\n", PAGES2KB(balloon_stats.driver_pages)); - -static ssize_t show_target_kb(struct sys_device *dev, struct sysdev_attribute *attr, - char *buf) -{ - return sprintf(buf, "%lu\n", PAGES2KB(balloon_stats.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; - - target_bytes = simple_strtoull(buf, &endchar, 0) * 1024; - - 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; - - 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_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 int 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; -} - MODULE_LICENSE("GPL"); diff --git a/drivers/xen/events.c b/drivers/xen/events.c index 31af0ac31a9..02b5a9c05cf 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -56,6 +56,8 @@ */ static DEFINE_SPINLOCK(irq_mapping_update_lock); +static LIST_HEAD(xen_irq_list_head); + /* IRQ <-> VIRQ mapping. */ static DEFINE_PER_CPU(int [NR_VIRQS], virq_to_irq) = {[0 ... NR_VIRQS-1] = -1}; @@ -85,7 +87,9 @@ enum xen_irq_type { */ struct irq_info { + struct list_head list; enum xen_irq_type type; /* type */ + unsigned irq; unsigned short evtchn; /* event channel */ unsigned short cpu; /* cpu bound */ @@ -103,23 +107,10 @@ struct irq_info #define PIRQ_NEEDS_EOI (1 << 0) #define PIRQ_SHAREABLE (1 << 1) -static struct irq_info *irq_info; -static int *pirq_to_irq; - static int *evtchn_to_irq; -struct cpu_evtchn_s { - unsigned long bits[NR_EVENT_CHANNELS/BITS_PER_LONG]; -}; - -static __initdata struct cpu_evtchn_s init_evtchn_mask = { - .bits[0 ... (NR_EVENT_CHANNELS/BITS_PER_LONG)-1] = ~0ul, -}; -static struct cpu_evtchn_s *cpu_evtchn_mask_p = &init_evtchn_mask; -static inline unsigned long *cpu_evtchn_mask(int cpu) -{ - return cpu_evtchn_mask_p[cpu].bits; -} +static DEFINE_PER_CPU(unsigned long [NR_EVENT_CHANNELS/BITS_PER_LONG], + cpu_evtchn_mask); /* Xen will never allocate port zero for any purpose. */ #define VALID_EVTCHN(chn) ((chn) != 0) @@ -128,48 +119,91 @@ static struct irq_chip xen_dynamic_chip; static struct irq_chip xen_percpu_chip; static struct irq_chip xen_pirq_chip; -/* Constructor for packed IRQ information. */ -static struct irq_info mk_unbound_info(void) +/* Get info for IRQ */ +static struct irq_info *info_for_irq(unsigned irq) { - return (struct irq_info) { .type = IRQT_UNBOUND }; + return get_irq_data(irq); } -static struct irq_info mk_evtchn_info(unsigned short evtchn) +/* Constructors for packed IRQ information. */ +static void xen_irq_info_common_init(struct irq_info *info, + unsigned irq, + enum xen_irq_type type, + unsigned short evtchn, + unsigned short cpu) { - return (struct irq_info) { .type = IRQT_EVTCHN, .evtchn = evtchn, - .cpu = 0 }; + + BUG_ON(info->type != IRQT_UNBOUND && info->type != type); + + info->type = type; + info->irq = irq; + info->evtchn = evtchn; + info->cpu = cpu; + + evtchn_to_irq[evtchn] = irq; } -static struct irq_info mk_ipi_info(unsigned short evtchn, enum ipi_vector ipi) +static void xen_irq_info_evtchn_init(unsigned irq, + unsigned short evtchn) { - return (struct irq_info) { .type = IRQT_IPI, .evtchn = evtchn, - .cpu = 0, .u.ipi = ipi }; + struct irq_info *info = info_for_irq(irq); + + xen_irq_info_common_init(info, irq, IRQT_EVTCHN, evtchn, 0); } -static struct irq_info mk_virq_info(unsigned short evtchn, unsigned short virq) +static void xen_irq_info_ipi_init(unsigned cpu, + unsigned irq, + unsigned short evtchn, + enum ipi_vector ipi) { - return (struct irq_info) { .type = IRQT_VIRQ, .evtchn = evtchn, - .cpu = 0, .u.virq = virq }; + struct irq_info *info = info_for_irq(irq); + + xen_irq_info_common_init(info, irq, IRQT_IPI, evtchn, 0); + + info->u.ipi = ipi; + + per_cpu(ipi_to_irq, cpu)[ipi] = irq; } -static struct irq_info mk_pirq_info(unsigned short evtchn, unsigned short pirq, - unsigned short gsi, unsigned short vector) +static void xen_irq_info_virq_init(unsigned cpu, + unsigned irq, + unsigned short evtchn, + unsigned short virq) { - return (struct irq_info) { .type = IRQT_PIRQ, .evtchn = evtchn, - .cpu = 0, - .u.pirq = { .pirq = pirq, .gsi = gsi, .vector = vector } }; + struct irq_info *info = info_for_irq(irq); + + xen_irq_info_common_init(info, irq, IRQT_VIRQ, evtchn, 0); + + info->u.virq = virq; + + per_cpu(virq_to_irq, cpu)[virq] = irq; } -/* - * Accessors for packed IRQ information. - */ -static struct irq_info *info_for_irq(unsigned irq) +static void xen_irq_info_pirq_init(unsigned irq, + unsigned short evtchn, + unsigned short pirq, + unsigned short gsi, + unsigned short vector, + unsigned char flags) { - return &irq_info[irq]; + struct irq_info *info = info_for_irq(irq); + + xen_irq_info_common_init(info, irq, IRQT_PIRQ, evtchn, 0); + + info->u.pirq.pirq = pirq; + info->u.pirq.gsi = gsi; + info->u.pirq.vector = vector; + info->u.pirq.flags = flags; } +/* + * Accessors for packed IRQ information. + */ static unsigned int evtchn_from_irq(unsigned irq) { + if (unlikely(WARN(irq < 0 || irq >= nr_irqs, "Invalid irq %d!\n", irq))) + return 0; + return info_for_irq(irq)->evtchn; } @@ -209,26 +243,6 @@ static unsigned pirq_from_irq(unsigned irq) return info->u.pirq.pirq; } -static unsigned gsi_from_irq(unsigned irq) -{ - struct irq_info *info = info_for_irq(irq); - - BUG_ON(info == NULL); - BUG_ON(info->type != IRQT_PIRQ); - - return info->u.pirq.gsi; -} - -static unsigned vector_from_irq(unsigned irq) -{ - struct irq_info *info = info_for_irq(irq); - - BUG_ON(info == NULL); - BUG_ON(info->type != IRQT_PIRQ); - - return info->u.pirq.vector; -} - static enum xen_irq_type type_from_irq(unsigned irq) { return info_for_irq(irq)->type; @@ -264,7 +278,7 @@ static inline unsigned long active_evtchns(unsigned int cpu, unsigned int idx) { return (sh->evtchn_pending[idx] & - cpu_evtchn_mask(cpu)[idx] & + per_cpu(cpu_evtchn_mask, cpu)[idx] & ~sh->evtchn_mask[idx]); } @@ -274,31 +288,31 @@ static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu) BUG_ON(irq == -1); #ifdef CONFIG_SMP - cpumask_copy(irq_to_desc(irq)->affinity, cpumask_of(cpu)); + cpumask_copy(irq_to_desc(irq)->irq_data.affinity, cpumask_of(cpu)); #endif - clear_bit(chn, cpu_evtchn_mask(cpu_from_irq(irq))); - set_bit(chn, cpu_evtchn_mask(cpu)); + clear_bit(chn, per_cpu(cpu_evtchn_mask, cpu_from_irq(irq))); + set_bit(chn, per_cpu(cpu_evtchn_mask, cpu)); - irq_info[irq].cpu = cpu; + info_for_irq(irq)->cpu = cpu; } static void init_evtchn_cpu_bindings(void) { int i; #ifdef CONFIG_SMP - struct irq_desc *desc; + struct irq_info *info; /* By default all event channels notify CPU#0. */ - for_each_irq_desc(i, desc) { - cpumask_copy(desc->affinity, cpumask_of(0)); + list_for_each_entry(info, &xen_irq_list_head, list) { + struct irq_desc *desc = irq_to_desc(info->irq); + cpumask_copy(desc->irq_data.affinity, cpumask_of(0)); } #endif for_each_possible_cpu(i) - memset(cpu_evtchn_mask(i), - (i == 0) ? ~0 : 0, sizeof(struct cpu_evtchn_s)); - + memset(per_cpu(cpu_evtchn_mask, i), + (i == 0) ? ~0 : 0, sizeof(*per_cpu(cpu_evtchn_mask, i))); } static inline void clear_evtchn(int port) @@ -355,7 +369,7 @@ static void unmask_evtchn(int port) struct evtchn_unmask unmask = { .port = port }; (void)HYPERVISOR_event_channel_op(EVTCHNOP_unmask, &unmask); } else { - struct vcpu_info *vcpu_info = __get_cpu_var(xen_vcpu); + struct vcpu_info *vcpu_info = __this_cpu_read(xen_vcpu); sync_clear_bit(port, &s->evtchn_mask[0]); @@ -373,75 +387,90 @@ static void unmask_evtchn(int port) put_cpu(); } -static int get_nr_hw_irqs(void) +static void xen_irq_init(unsigned irq) { - int ret = 1; + struct irq_info *info; + struct irq_desc *desc = irq_to_desc(irq); -#ifdef CONFIG_X86_IO_APIC - ret = get_nr_irqs_gsi(); +#ifdef CONFIG_SMP + /* By default all event channels notify CPU#0. */ + cpumask_copy(desc->irq_data.affinity, cpumask_of(0)); #endif - return ret; -} + info = kzalloc(sizeof(*info), GFP_KERNEL); + if (info == NULL) + panic("Unable to allocate metadata for IRQ%d\n", irq); -static int find_unbound_pirq(int type) -{ - int rc, i; - struct physdev_get_free_pirq op_get_free_pirq; - op_get_free_pirq.type = type; + info->type = IRQT_UNBOUND; - rc = HYPERVISOR_physdev_op(PHYSDEVOP_get_free_pirq, &op_get_free_pirq); - if (!rc) - return op_get_free_pirq.pirq; + set_irq_data(irq, info); - for (i = 0; i < nr_irqs; i++) { - if (pirq_to_irq[i] < 0) - return i; - } - return -1; + list_add_tail(&info->list, &xen_irq_list_head); } -static int find_unbound_irq(void) +static int __must_check xen_allocate_irq_dynamic(void) { - struct irq_data *data; - int irq, res; - int start = get_nr_hw_irqs(); + int first = 0; + int irq; - if (start == nr_irqs) - goto no_irqs; +#ifdef CONFIG_X86_IO_APIC + /* + * For an HVM guest or domain 0 which see "real" (emulated or + * actual repectively) GSIs we allocate dynamic IRQs + * e.g. those corresponding to event channels or MSIs + * etc. from the range above those "real" GSIs to avoid + * collisions. + */ + if (xen_initial_domain() || xen_hvm_domain()) + first = get_nr_irqs_gsi(); +#endif - /* nr_irqs is a magic value. Must not use it.*/ - for (irq = nr_irqs-1; irq > start; irq--) { - data = irq_get_irq_data(irq); - /* only 0->15 have init'd desc; handle irq > 16 */ - if (!data) - break; - if (data->chip == &no_irq_chip) - break; - if (data->chip != &xen_dynamic_chip) - continue; - if (irq_info[irq].type == IRQT_UNBOUND) - return irq; - } + irq = irq_alloc_desc_from(first, -1); - if (irq == start) - goto no_irqs; + xen_irq_init(irq); - res = irq_alloc_desc_at(irq, -1); + return irq; +} - if (WARN_ON(res != irq)) - return -1; +static int __must_check xen_allocate_irq_gsi(unsigned gsi) +{ + int irq; - return irq; + /* + * A PV guest has no concept of a GSI (since it has no ACPI + * nor access to/knowledge of the physical APICs). Therefore + * all IRQs are dynamically allocated from the entire IRQ + * space. + */ + if (xen_pv_domain() && !xen_initial_domain()) + return xen_allocate_irq_dynamic(); + + /* Legacy IRQ descriptors are already allocated by the arch. */ + if (gsi < NR_IRQS_LEGACY) + irq = gsi; + else + irq = irq_alloc_desc_at(gsi, -1); + + xen_irq_init(irq); -no_irqs: - panic("No available IRQ to bind to: increase nr_irqs!\n"); + return irq; } -static bool identity_mapped_irq(unsigned irq) +static void xen_free_irq(unsigned irq) { - /* identity map all the hardware irqs */ - return irq < get_nr_hw_irqs(); + struct irq_info *info = get_irq_data(irq); + + list_del(&info->list); + + set_irq_data(irq, NULL); + + kfree(info); + + /* Legacy IRQ descriptors are managed by the arch. */ + if (irq < NR_IRQS_LEGACY) + return; + + irq_free_desc(irq); } static void pirq_unmask_notify(int irq) @@ -477,7 +506,7 @@ static bool probing_irq(int irq) return desc && desc->action == NULL; } -static unsigned int startup_pirq(unsigned int irq) +static unsigned int __startup_pirq(unsigned int irq) { struct evtchn_bind_pirq bind_pirq; struct irq_info *info = info_for_irq(irq); @@ -515,9 +544,15 @@ out: return 0; } -static void shutdown_pirq(unsigned int irq) +static unsigned int startup_pirq(struct irq_data *data) +{ + return __startup_pirq(data->irq); +} + +static void shutdown_pirq(struct irq_data *data) { struct evtchn_close close; + unsigned int irq = data->irq; struct irq_info *info = info_for_irq(irq); int evtchn = evtchn_from_irq(irq); @@ -537,20 +572,20 @@ static void shutdown_pirq(unsigned int irq) info->evtchn = 0; } -static void enable_pirq(unsigned int irq) +static void enable_pirq(struct irq_data *data) { - startup_pirq(irq); + startup_pirq(data); } -static void disable_pirq(unsigned int irq) +static void disable_pirq(struct irq_data *data) { } -static void ack_pirq(unsigned int irq) +static void ack_pirq(struct irq_data *data) { - int evtchn = evtchn_from_irq(irq); + int evtchn = evtchn_from_irq(data->irq); - move_native_irq(irq); + move_native_irq(data->irq); if (VALID_EVTCHN(evtchn)) { mask_evtchn(evtchn); @@ -558,70 +593,41 @@ static void ack_pirq(unsigned int irq) } } -static void end_pirq(unsigned int irq) -{ - int evtchn = evtchn_from_irq(irq); - struct irq_desc *desc = irq_to_desc(irq); - - if (WARN_ON(!desc)) - return; - - if ((desc->status & (IRQ_DISABLED|IRQ_PENDING)) == - (IRQ_DISABLED|IRQ_PENDING)) { - shutdown_pirq(irq); - } else if (VALID_EVTCHN(evtchn)) { - unmask_evtchn(evtchn); - pirq_unmask_notify(irq); - } -} - static int find_irq_by_gsi(unsigned gsi) { - int irq; - - for (irq = 0; irq < nr_irqs; irq++) { - struct irq_info *info = info_for_irq(irq); + struct irq_info *info; - if (info == NULL || info->type != IRQT_PIRQ) + list_for_each_entry(info, &xen_irq_list_head, list) { + if (info->type != IRQT_PIRQ) continue; - if (gsi_from_irq(irq) == gsi) - return irq; + if (info->u.pirq.gsi == gsi) + return info->irq; } return -1; } -int xen_allocate_pirq(unsigned gsi, int shareable, char *name) +int xen_allocate_pirq_gsi(unsigned gsi) { - return xen_map_pirq_gsi(gsi, gsi, shareable, name); + return gsi; } -/* xen_map_pirq_gsi might allocate irqs from the top down, as a - * consequence don't assume that the irq number returned has a low value - * or can be used as a pirq number unless you know otherwise. - * - * One notable exception is when xen_map_pirq_gsi is called passing an - * hardware gsi as argument, in that case the irq number returned - * matches the gsi number passed as second argument. +/* + * Do not make any assumptions regarding the relationship between the + * IRQ number returned here and the Xen pirq argument. * * Note: We don't assign an event channel until the irq actually started * up. Return an existing irq if we've already got one for the gsi. */ -int xen_map_pirq_gsi(unsigned pirq, unsigned gsi, int shareable, char *name) +int xen_bind_pirq_gsi_to_irq(unsigned gsi, + unsigned pirq, int shareable, char *name) { - int irq = 0; + int irq = -1; struct physdev_irq irq_op; spin_lock(&irq_mapping_update_lock); - if ((pirq > nr_irqs) || (gsi > nr_irqs)) { - printk(KERN_WARNING "xen_map_pirq_gsi: %s %s is incorrect!\n", - pirq > nr_irqs ? "pirq" :"", - gsi > nr_irqs ? "gsi" : ""); - goto out; - } - irq = find_irq_by_gsi(gsi); if (irq != -1) { printk(KERN_INFO "xen_map_pirq_gsi: returning irq %d for gsi %u\n", @@ -629,14 +635,9 @@ int xen_map_pirq_gsi(unsigned pirq, unsigned gsi, int shareable, char *name) goto out; /* XXX need refcount? */ } - /* If we are a PV guest, we don't have GSIs (no ACPI passed). Therefore - * we are using the !xen_initial_domain() to drop in the function.*/ - if (identity_mapped_irq(gsi) || (!xen_initial_domain() && - xen_pv_domain())) { - irq = gsi; - irq_alloc_desc_at(irq, -1); - } else - irq = find_unbound_irq(); + irq = xen_allocate_irq_gsi(gsi); + if (irq < 0) + goto out; set_irq_chip_and_handler_name(irq, &xen_pirq_chip, handle_level_irq, name); @@ -649,14 +650,13 @@ int xen_map_pirq_gsi(unsigned pirq, unsigned gsi, int shareable, char *name) * this in the priv domain. */ if (xen_initial_domain() && HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op)) { - irq_free_desc(irq); + xen_free_irq(irq); irq = -ENOSPC; goto out; } - irq_info[irq] = mk_pirq_info(0, pirq, gsi, irq_op.vector); - irq_info[irq].u.pirq.flags |= shareable ? PIRQ_SHAREABLE : 0; - pirq_to_irq[pirq] = irq; + xen_irq_info_pirq_init(irq, 0, pirq, gsi, irq_op.vector, + shareable ? PIRQ_SHAREABLE : 0); out: spin_unlock(&irq_mapping_update_lock); @@ -665,87 +665,45 @@ out: } #ifdef CONFIG_PCI_MSI -#include <linux/msi.h> -#include "../pci/msi.h" - -void xen_allocate_pirq_msi(char *name, int *irq, int *pirq, int alloc) +int xen_allocate_pirq_msi(struct pci_dev *dev, struct msi_desc *msidesc) { - spin_lock(&irq_mapping_update_lock); - - if (alloc & XEN_ALLOC_IRQ) { - *irq = find_unbound_irq(); - if (*irq == -1) - goto out; - } - - if (alloc & XEN_ALLOC_PIRQ) { - *pirq = find_unbound_pirq(MAP_PIRQ_TYPE_MSI); - if (*pirq == -1) - goto out; - } + int rc; + struct physdev_get_free_pirq op_get_free_pirq; - set_irq_chip_and_handler_name(*irq, &xen_pirq_chip, - handle_level_irq, name); + op_get_free_pirq.type = MAP_PIRQ_TYPE_MSI; + rc = HYPERVISOR_physdev_op(PHYSDEVOP_get_free_pirq, &op_get_free_pirq); - irq_info[*irq] = mk_pirq_info(0, *pirq, 0, 0); - pirq_to_irq[*pirq] = *irq; + WARN_ONCE(rc == -ENOSYS, + "hypervisor does not support the PHYSDEVOP_get_free_pirq interface\n"); -out: - spin_unlock(&irq_mapping_update_lock); + return rc ? -1 : op_get_free_pirq.pirq; } -int xen_create_msi_irq(struct pci_dev *dev, struct msi_desc *msidesc, int type) +int xen_bind_pirq_msi_to_irq(struct pci_dev *dev, struct msi_desc *msidesc, + int pirq, int vector, const char *name) { - int irq = -1; - struct physdev_map_pirq map_irq; - int rc; - int pos; - u32 table_offset, bir; - - memset(&map_irq, 0, sizeof(map_irq)); - map_irq.domid = DOMID_SELF; - map_irq.type = MAP_PIRQ_TYPE_MSI; - map_irq.index = -1; - map_irq.pirq = -1; - map_irq.bus = dev->bus->number; - map_irq.devfn = dev->devfn; - - if (type == PCI_CAP_ID_MSIX) { - pos = pci_find_capability(dev, PCI_CAP_ID_MSIX); - - pci_read_config_dword(dev, msix_table_offset_reg(pos), - &table_offset); - bir = (u8)(table_offset & PCI_MSIX_FLAGS_BIRMASK); - - map_irq.table_base = pci_resource_start(dev, bir); - map_irq.entry_nr = msidesc->msi_attrib.entry_nr; - } + int irq, ret; spin_lock(&irq_mapping_update_lock); - irq = find_unbound_irq(); - + irq = xen_allocate_irq_dynamic(); if (irq == -1) goto out; - rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq); - if (rc) { - printk(KERN_WARNING "xen map irq failed %d\n", rc); - - irq_free_desc(irq); - - irq = -1; - goto out; - } - irq_info[irq] = mk_pirq_info(0, map_irq.pirq, 0, map_irq.index); - set_irq_chip_and_handler_name(irq, &xen_pirq_chip, - handle_level_irq, - (type == PCI_CAP_ID_MSIX) ? "msi-x":"msi"); + handle_level_irq, name); + xen_irq_info_pirq_init(irq, 0, pirq, 0, vector, 0); + ret = irq_set_msi_desc(irq, msidesc); + if (ret < 0) + goto error_irq; out: spin_unlock(&irq_mapping_update_lock); return irq; +error_irq: + spin_unlock(&irq_mapping_update_lock); + xen_free_irq(irq); + return -1; } #endif @@ -770,30 +728,35 @@ int xen_destroy_irq(int irq) printk(KERN_WARNING "unmap irq failed %d\n", rc); goto out; } - pirq_to_irq[info->u.pirq.pirq] = -1; } - irq_info[irq] = mk_unbound_info(); - irq_free_desc(irq); + xen_free_irq(irq); out: spin_unlock(&irq_mapping_update_lock); return rc; } -int xen_vector_from_irq(unsigned irq) +int xen_irq_from_pirq(unsigned pirq) { - return vector_from_irq(irq); -} + int irq; -int xen_gsi_from_irq(unsigned irq) -{ - return gsi_from_irq(irq); -} + struct irq_info *info; -int xen_irq_from_pirq(unsigned pirq) -{ - return pirq_to_irq[pirq]; + spin_lock(&irq_mapping_update_lock); + + list_for_each_entry(info, &xen_irq_list_head, list) { + if (info == NULL || info->type != IRQT_PIRQ) + continue; + irq = info->irq; + if (info->u.pirq.pirq == pirq) + goto out; + } + irq = -1; +out: + spin_unlock(&irq_mapping_update_lock); + + return irq; } int bind_evtchn_to_irq(unsigned int evtchn) @@ -805,15 +768,17 @@ int bind_evtchn_to_irq(unsigned int evtchn) irq = evtchn_to_irq[evtchn]; if (irq == -1) { - irq = find_unbound_irq(); + irq = xen_allocate_irq_dynamic(); + if (irq == -1) + goto out; set_irq_chip_and_handler_name(irq, &xen_dynamic_chip, handle_fasteoi_irq, "event"); - evtchn_to_irq[evtchn] = irq; - irq_info[irq] = mk_evtchn_info(evtchn); + xen_irq_info_evtchn_init(irq, evtchn); } +out: spin_unlock(&irq_mapping_update_lock); return irq; @@ -830,7 +795,7 @@ static int bind_ipi_to_irq(unsigned int ipi, unsigned int cpu) irq = per_cpu(ipi_to_irq, cpu)[ipi]; if (irq == -1) { - irq = find_unbound_irq(); + irq = xen_allocate_irq_dynamic(); if (irq < 0) goto out; @@ -843,9 +808,7 @@ static int bind_ipi_to_irq(unsigned int ipi, unsigned int cpu) BUG(); evtchn = bind_ipi.port; - evtchn_to_irq[evtchn] = irq; - irq_info[irq] = mk_ipi_info(evtchn, ipi); - per_cpu(ipi_to_irq, cpu)[ipi] = irq; + xen_irq_info_ipi_init(cpu, irq, evtchn, ipi); bind_evtchn_to_cpu(evtchn, cpu); } @@ -855,6 +818,21 @@ static int bind_ipi_to_irq(unsigned int ipi, unsigned int cpu) return irq; } +static int bind_interdomain_evtchn_to_irq(unsigned int remote_domain, + unsigned int remote_port) +{ + struct evtchn_bind_interdomain bind_interdomain; + int err; + + bind_interdomain.remote_dom = remote_domain; + bind_interdomain.remote_port = remote_port; + + err = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain, + &bind_interdomain); + + return err ? : bind_evtchn_to_irq(bind_interdomain.local_port); +} + int bind_virq_to_irq(unsigned int virq, unsigned int cpu) { @@ -866,7 +844,9 @@ int bind_virq_to_irq(unsigned int virq, unsigned int cpu) irq = per_cpu(virq_to_irq, cpu)[virq]; if (irq == -1) { - irq = find_unbound_irq(); + irq = xen_allocate_irq_dynamic(); + if (irq == -1) + goto out; set_irq_chip_and_handler_name(irq, &xen_percpu_chip, handle_percpu_irq, "virq"); @@ -878,14 +858,12 @@ int bind_virq_to_irq(unsigned int virq, unsigned int cpu) BUG(); evtchn = bind_virq.port; - evtchn_to_irq[evtchn] = irq; - irq_info[irq] = mk_virq_info(evtchn, virq); - - per_cpu(virq_to_irq, cpu)[virq] = irq; + xen_irq_info_virq_init(cpu, irq, evtchn, virq); bind_evtchn_to_cpu(evtchn, cpu); } +out: spin_unlock(&irq_mapping_update_lock); return irq; @@ -922,11 +900,9 @@ static void unbind_from_irq(unsigned int irq) evtchn_to_irq[evtchn] = -1; } - if (irq_info[irq].type != IRQT_UNBOUND) { - irq_info[irq] = mk_unbound_info(); + BUG_ON(info_for_irq(irq)->type == IRQT_UNBOUND); - irq_free_desc(irq); - } + xen_free_irq(irq); spin_unlock(&irq_mapping_update_lock); } @@ -940,6 +916,8 @@ int bind_evtchn_to_irqhandler(unsigned int evtchn, int retval; irq = bind_evtchn_to_irq(evtchn); + if (irq < 0) + return irq; retval = request_irq(irq, handler, irqflags, devname, dev_id); if (retval != 0) { unbind_from_irq(irq); @@ -950,6 +928,29 @@ int bind_evtchn_to_irqhandler(unsigned int evtchn, } EXPORT_SYMBOL_GPL(bind_evtchn_to_irqhandler); +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 irq, retval; + + irq = bind_interdomain_evtchn_to_irq(remote_domain, remote_port); + if (irq < 0) + return irq; + + retval = request_irq(irq, handler, irqflags, devname, dev_id); + if (retval != 0) { + unbind_from_irq(irq); + return retval; + } + + return irq; +} +EXPORT_SYMBOL_GPL(bind_interdomain_evtchn_to_irqhandler); + int bind_virq_to_irqhandler(unsigned int virq, unsigned int cpu, irq_handler_t handler, unsigned long irqflags, const char *devname, void *dev_id) @@ -958,6 +959,8 @@ int bind_virq_to_irqhandler(unsigned int virq, unsigned int cpu, int retval; irq = bind_virq_to_irq(virq, cpu); + if (irq < 0) + return irq; retval = request_irq(irq, handler, irqflags, devname, dev_id); if (retval != 0) { unbind_from_irq(irq); @@ -981,7 +984,7 @@ int bind_ipi_to_irqhandler(enum ipi_vector ipi, if (irq < 0) return irq; - irqflags |= IRQF_NO_SUSPEND; + irqflags |= IRQF_NO_SUSPEND | IRQF_FORCE_RESUME; retval = request_irq(irq, handler, irqflags, devname, dev_id); if (retval != 0) { unbind_from_irq(irq); @@ -1009,7 +1012,7 @@ irqreturn_t xen_debug_interrupt(int irq, void *dev_id) { struct shared_info *sh = HYPERVISOR_shared_info; int cpu = smp_processor_id(); - unsigned long *cpu_evtchn = cpu_evtchn_mask(cpu); + unsigned long *cpu_evtchn = per_cpu(cpu_evtchn_mask, cpu); int i; unsigned long flags; static DEFINE_SPINLOCK(debug_lock); @@ -1087,6 +1090,13 @@ irqreturn_t xen_debug_interrupt(int irq, void *dev_id) } static DEFINE_PER_CPU(unsigned, xed_nesting_count); +static DEFINE_PER_CPU(unsigned int, current_word_idx); +static DEFINE_PER_CPU(unsigned int, current_bit_idx); + +/* + * Mask out the i least significant bits of w + */ +#define MASK_LSBS(w, i) (w & ((~0UL) << i)) /* * Search the CPUs pending events bitmasks. For each one found, map @@ -1099,9 +1109,12 @@ static DEFINE_PER_CPU(unsigned, xed_nesting_count); */ static void __xen_evtchn_do_upcall(void) { + int start_word_idx, start_bit_idx; + int word_idx, bit_idx; + int i; int cpu = get_cpu(); struct shared_info *s = HYPERVISOR_shared_info; - struct vcpu_info *vcpu_info = __get_cpu_var(xen_vcpu); + struct vcpu_info *vcpu_info = __this_cpu_read(xen_vcpu); unsigned count; do { @@ -1109,7 +1122,7 @@ static void __xen_evtchn_do_upcall(void) vcpu_info->evtchn_upcall_pending = 0; - if (__get_cpu_var(xed_nesting_count)++) + if (__this_cpu_inc_return(xed_nesting_count) - 1) goto out; #ifndef CONFIG_X86 /* No need for a barrier -- XCHG is a barrier on x86. */ @@ -1117,17 +1130,57 @@ static void __xen_evtchn_do_upcall(void) wmb(); #endif pending_words = xchg(&vcpu_info->evtchn_pending_sel, 0); - while (pending_words != 0) { + + start_word_idx = __this_cpu_read(current_word_idx); + start_bit_idx = __this_cpu_read(current_bit_idx); + + word_idx = start_word_idx; + + for (i = 0; pending_words != 0; i++) { unsigned long pending_bits; - int word_idx = __ffs(pending_words); - pending_words &= ~(1UL << word_idx); + unsigned long words; + + words = MASK_LSBS(pending_words, word_idx); - while ((pending_bits = active_evtchns(cpu, s, word_idx)) != 0) { - int bit_idx = __ffs(pending_bits); - int port = (word_idx * BITS_PER_LONG) + bit_idx; - int irq = evtchn_to_irq[port]; + /* + * If we masked out all events, wrap to beginning. + */ + if (words == 0) { + word_idx = 0; + bit_idx = 0; + continue; + } + word_idx = __ffs(words); + + pending_bits = active_evtchns(cpu, s, word_idx); + bit_idx = 0; /* usually scan entire word from start */ + if (word_idx == start_word_idx) { + /* We scan the starting word in two parts */ + if (i == 0) + /* 1st time: start in the middle */ + bit_idx = start_bit_idx; + else + /* 2nd time: mask bits done already */ + bit_idx &= (1UL << start_bit_idx) - 1; + } + + do { + unsigned long bits; + int port, irq; struct irq_desc *desc; + bits = MASK_LSBS(pending_bits, bit_idx); + + /* If we masked out all events, move on. */ + if (bits == 0) + break; + + bit_idx = __ffs(bits); + + /* Process port. */ + port = (word_idx * BITS_PER_LONG) + bit_idx; + irq = evtchn_to_irq[port]; + mask_evtchn(port); clear_evtchn(port); @@ -1136,13 +1189,27 @@ static void __xen_evtchn_do_upcall(void) if (desc) generic_handle_irq_desc(irq, desc); } - } + + bit_idx = (bit_idx + 1) % BITS_PER_LONG; + + /* Next caller starts at last processed + 1 */ + __this_cpu_write(current_word_idx, + bit_idx ? word_idx : + (word_idx+1) % BITS_PER_LONG); + __this_cpu_write(current_bit_idx, bit_idx); + } while (bit_idx != 0); + + /* Scan start_l1i twice; all others once. */ + if ((word_idx != start_word_idx) || (i != 0)) + pending_words &= ~(1UL << word_idx); + + word_idx = (word_idx + 1) % BITS_PER_LONG; } BUG_ON(!irqs_disabled()); - count = __get_cpu_var(xed_nesting_count); - __get_cpu_var(xed_nesting_count) = 0; + count = __this_cpu_read(xed_nesting_count); + __this_cpu_write(xed_nesting_count, 0); } while (count != 1 || vcpu_info->evtchn_upcall_pending); out: @@ -1186,8 +1253,7 @@ void rebind_evtchn_irq(int evtchn, int irq) so there should be a proper type */ BUG_ON(info->type == IRQT_UNBOUND); - evtchn_to_irq[evtchn] = irq; - irq_info[irq] = mk_evtchn_info(evtchn); + xen_irq_info_evtchn_init(irq, evtchn); spin_unlock(&irq_mapping_update_lock); @@ -1204,10 +1270,14 @@ static int rebind_irq_to_cpu(unsigned irq, unsigned tcpu) struct evtchn_bind_vcpu bind_vcpu; int evtchn = evtchn_from_irq(irq); - /* events delivered via platform PCI interrupts are always - * routed to vcpu 0 */ - if (!VALID_EVTCHN(evtchn) || - (xen_hvm_domain() && !xen_have_vector_callback)) + if (!VALID_EVTCHN(evtchn)) + return -1; + + /* + * Events delivered via platform PCI interrupts are always + * routed to vcpu 0 and hence cannot be rebound. + */ + if (xen_hvm_domain() && !xen_have_vector_callback) return -1; /* Send future instances of this interrupt to other vcpu. */ @@ -1225,11 +1295,12 @@ static int rebind_irq_to_cpu(unsigned irq, unsigned tcpu) return 0; } -static int set_affinity_irq(unsigned irq, const struct cpumask *dest) +static int set_affinity_irq(struct irq_data *data, const struct cpumask *dest, + bool force) { unsigned tcpu = cpumask_first(dest); - return rebind_irq_to_cpu(irq, tcpu); + return rebind_irq_to_cpu(data->irq, tcpu); } int resend_irq_on_evtchn(unsigned int irq) @@ -1248,35 +1319,35 @@ int resend_irq_on_evtchn(unsigned int irq) return 1; } -static void enable_dynirq(unsigned int irq) +static void enable_dynirq(struct irq_data *data) { - int evtchn = evtchn_from_irq(irq); + int evtchn = evtchn_from_irq(data->irq); if (VALID_EVTCHN(evtchn)) unmask_evtchn(evtchn); } -static void disable_dynirq(unsigned int irq) +static void disable_dynirq(struct irq_data *data) { - int evtchn = evtchn_from_irq(irq); + int evtchn = evtchn_from_irq(data->irq); if (VALID_EVTCHN(evtchn)) mask_evtchn(evtchn); } -static void ack_dynirq(unsigned int irq) +static void ack_dynirq(struct irq_data *data) { - int evtchn = evtchn_from_irq(irq); + int evtchn = evtchn_from_irq(data->irq); - move_masked_irq(irq); + move_masked_irq(data->irq); if (VALID_EVTCHN(evtchn)) unmask_evtchn(evtchn); } -static int retrigger_dynirq(unsigned int irq) +static int retrigger_dynirq(struct irq_data *data) { - int evtchn = evtchn_from_irq(irq); + int evtchn = evtchn_from_irq(data->irq); struct shared_info *sh = HYPERVISOR_shared_info; int ret = 0; @@ -1293,19 +1364,22 @@ static int retrigger_dynirq(unsigned int irq) return ret; } -static void restore_cpu_pirqs(void) +static void restore_pirqs(void) { int pirq, rc, irq, gsi; struct physdev_map_pirq map_irq; + struct irq_info *info; - for (pirq = 0; pirq < nr_irqs; pirq++) { - irq = pirq_to_irq[pirq]; - if (irq == -1) + list_for_each_entry(info, &xen_irq_list_head, list) { + if (info->type != IRQT_PIRQ) continue; + pirq = info->u.pirq.pirq; + gsi = info->u.pirq.gsi; + irq = info->irq; + /* save/restore of PT devices doesn't work, so at this point the * only devices present are GSI based emulated devices */ - gsi = gsi_from_irq(irq); if (!gsi) continue; @@ -1318,14 +1392,13 @@ static void restore_cpu_pirqs(void) if (rc) { printk(KERN_WARNING "xen map irq failed gsi=%d irq=%d pirq=%d rc=%d\n", gsi, irq, pirq, rc); - irq_info[irq] = mk_unbound_info(); - pirq_to_irq[pirq] = -1; + xen_free_irq(irq); continue; } printk(KERN_DEBUG "xen: --> irq=%d, pirq=%d\n", irq, map_irq.pirq); - startup_pirq(irq); + __startup_pirq(irq); } } @@ -1349,8 +1422,7 @@ static void restore_cpu_virqs(unsigned int cpu) evtchn = bind_virq.port; /* Record the new mapping. */ - evtchn_to_irq[evtchn] = irq; - irq_info[irq] = mk_virq_info(evtchn, virq); + xen_irq_info_virq_init(cpu, irq, evtchn, virq); bind_evtchn_to_cpu(evtchn, cpu); } } @@ -1374,8 +1446,7 @@ static void restore_cpu_ipis(unsigned int cpu) evtchn = bind_ipi.port; /* Record the new mapping. */ - evtchn_to_irq[evtchn] = irq; - irq_info[irq] = mk_ipi_info(evtchn, ipi); + xen_irq_info_ipi_init(cpu, irq, evtchn, ipi); bind_evtchn_to_cpu(evtchn, cpu); } } @@ -1435,8 +1506,8 @@ void xen_poll_irq(int irq) void xen_irq_resume(void) { - unsigned int cpu, irq, evtchn; - struct irq_desc *desc; + unsigned int cpu, evtchn; + struct irq_info *info; init_evtchn_cpu_bindings(); @@ -1445,8 +1516,8 @@ void xen_irq_resume(void) mask_evtchn(evtchn); /* No IRQ <-> event-channel mappings. */ - for (irq = 0; irq < nr_irqs; irq++) - irq_info[irq].evtchn = 0; /* zap event-channel binding */ + list_for_each_entry(info, &xen_irq_list_head, list) + info->evtchn = 0; /* zap event-channel binding */ for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++) evtchn_to_irq[evtchn] = -1; @@ -1456,66 +1527,48 @@ void xen_irq_resume(void) restore_cpu_ipis(cpu); } - /* - * Unmask any IRQF_NO_SUSPEND IRQs which are enabled. These - * are not handled by the IRQ core. - */ - for_each_irq_desc(irq, desc) { - if (!desc->action || !(desc->action->flags & IRQF_NO_SUSPEND)) - continue; - if (desc->status & IRQ_DISABLED) - continue; - - evtchn = evtchn_from_irq(irq); - if (evtchn == -1) - continue; - - unmask_evtchn(evtchn); - } - - restore_cpu_pirqs(); + restore_pirqs(); } static struct irq_chip xen_dynamic_chip __read_mostly = { - .name = "xen-dyn", + .name = "xen-dyn", - .disable = disable_dynirq, - .mask = disable_dynirq, - .unmask = enable_dynirq, + .irq_disable = disable_dynirq, + .irq_mask = disable_dynirq, + .irq_unmask = enable_dynirq, - .eoi = ack_dynirq, - .set_affinity = set_affinity_irq, - .retrigger = retrigger_dynirq, + .irq_eoi = ack_dynirq, + .irq_set_affinity = set_affinity_irq, + .irq_retrigger = retrigger_dynirq, }; static struct irq_chip xen_pirq_chip __read_mostly = { - .name = "xen-pirq", + .name = "xen-pirq", - .startup = startup_pirq, - .shutdown = shutdown_pirq, + .irq_startup = startup_pirq, + .irq_shutdown = shutdown_pirq, - .enable = enable_pirq, - .unmask = enable_pirq, + .irq_enable = enable_pirq, + .irq_unmask = enable_pirq, - .disable = disable_pirq, - .mask = disable_pirq, + .irq_disable = disable_pirq, + .irq_mask = disable_pirq, - .ack = ack_pirq, - .end = end_pirq, + .irq_ack = ack_pirq, - .set_affinity = set_affinity_irq, + .irq_set_affinity = set_affinity_irq, - .retrigger = retrigger_dynirq, + .irq_retrigger = retrigger_dynirq, }; static struct irq_chip xen_percpu_chip __read_mostly = { - .name = "xen-percpu", + .name = "xen-percpu", - .disable = disable_dynirq, - .mask = disable_dynirq, - .unmask = enable_dynirq, + .irq_disable = disable_dynirq, + .irq_mask = disable_dynirq, + .irq_unmask = enable_dynirq, - .ack = ack_dynirq, + .irq_ack = ack_dynirq, }; int xen_set_callback_via(uint64_t via) @@ -1560,17 +1613,6 @@ void __init xen_init_IRQ(void) { int i; - cpu_evtchn_mask_p = kcalloc(nr_cpu_ids, sizeof(struct cpu_evtchn_s), - GFP_KERNEL); - irq_info = kcalloc(nr_irqs, sizeof(*irq_info), GFP_KERNEL); - - /* We are using nr_irqs as the maximum number of pirq available but - * that number is actually chosen by Xen and we don't know exactly - * what it is. Be careful choosing high pirq numbers. */ - pirq_to_irq = kcalloc(nr_irqs, sizeof(*pirq_to_irq), GFP_KERNEL); - for (i = 0; i < nr_irqs; i++) - pirq_to_irq[i] = -1; - evtchn_to_irq = kcalloc(NR_EVENT_CHANNELS, sizeof(*evtchn_to_irq), GFP_KERNEL); for (i = 0; i < NR_EVENT_CHANNELS; i++) diff --git a/drivers/xen/gntalloc.c b/drivers/xen/gntalloc.c new file mode 100644 index 00000000000..a7ffdfe19fc --- /dev/null +++ b/drivers/xen/gntalloc.c @@ -0,0 +1,545 @@ +/****************************************************************************** + * gntalloc.c + * + * Device for creating grant references (in user-space) that may be shared + * with other domains. + * + * 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 + */ + +/* + * This driver exists to allow userspace programs in Linux to allocate kernel + * memory that will later be shared with another domain. Without this device, + * Linux userspace programs cannot create grant references. + * + * How this stuff works: + * X -> granting a page to Y + * Y -> mapping the grant from X + * + * 1. X uses the gntalloc device to allocate a page of kernel memory, P. + * 2. X creates an entry in the grant table that says domid(Y) can access P. + * This is done without a hypercall unless the grant table needs expansion. + * 3. X gives the grant reference identifier, GREF, to Y. + * 4. Y maps the page, either directly into kernel memory for use in a backend + * driver, or via a the gntdev device to map into the address space of an + * application running in Y. This is the first point at which Xen does any + * tracking of the page. + * 5. A program in X mmap()s a segment of the gntalloc device that corresponds + * to the shared page, and can now communicate with Y over the shared page. + * + * + * NOTE TO USERSPACE LIBRARIES: + * The grant allocation and mmap()ing are, naturally, two separate operations. + * You set up the sharing by calling the create ioctl() and then the mmap(). + * Teardown requires munmap() and either close() or ioctl(). + * + * WARNING: Since Xen does not allow a guest to forcibly end the use of a grant + * reference, this device can be used to consume kernel memory by leaving grant + * references mapped by another domain when an application exits. Therefore, + * there is a global limit on the number of pages that can be allocated. When + * all references to the page are unmapped, it will be freed during the next + * grant operation. + */ + +#include <linux/atomic.h> +#include <linux/module.h> +#include <linux/miscdevice.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/slab.h> +#include <linux/fs.h> +#include <linux/device.h> +#include <linux/mm.h> +#include <linux/uaccess.h> +#include <linux/types.h> +#include <linux/list.h> +#include <linux/highmem.h> + +#include <xen/xen.h> +#include <xen/page.h> +#include <xen/grant_table.h> +#include <xen/gntalloc.h> +#include <xen/events.h> + +static int limit = 1024; +module_param(limit, int, 0644); +MODULE_PARM_DESC(limit, "Maximum number of grants that may be allocated by " + "the gntalloc device"); + +static LIST_HEAD(gref_list); +static DEFINE_SPINLOCK(gref_lock); +static int gref_size; + +struct notify_info { + uint16_t pgoff:12; /* Bits 0-11: Offset of the byte to clear */ + uint16_t flags:2; /* Bits 12-13: Unmap notification flags */ + int event; /* Port (event channel) to notify */ +}; + +/* Metadata on a grant reference. */ +struct gntalloc_gref { + struct list_head next_gref; /* list entry gref_list */ + struct list_head next_file; /* list entry file->list, if open */ + struct page *page; /* The shared page */ + uint64_t file_index; /* File offset for mmap() */ + unsigned int users; /* Use count - when zero, waiting on Xen */ + grant_ref_t gref_id; /* The grant reference number */ + struct notify_info notify; /* Unmap notification */ +}; + +struct gntalloc_file_private_data { + struct list_head list; + uint64_t index; +}; + +static void __del_gref(struct gntalloc_gref *gref); + +static void do_cleanup(void) +{ + struct gntalloc_gref *gref, *n; + list_for_each_entry_safe(gref, n, &gref_list, next_gref) { + if (!gref->users) + __del_gref(gref); + } +} + +static int add_grefs(struct ioctl_gntalloc_alloc_gref *op, + uint32_t *gref_ids, struct gntalloc_file_private_data *priv) +{ + int i, rc, readonly; + LIST_HEAD(queue_gref); + LIST_HEAD(queue_file); + struct gntalloc_gref *gref; + + readonly = !(op->flags & GNTALLOC_FLAG_WRITABLE); + rc = -ENOMEM; + for (i = 0; i < op->count; i++) { + gref = kzalloc(sizeof(*gref), GFP_KERNEL); + if (!gref) + goto undo; + list_add_tail(&gref->next_gref, &queue_gref); + list_add_tail(&gref->next_file, &queue_file); + gref->users = 1; + gref->file_index = op->index + i * PAGE_SIZE; + gref->page = alloc_page(GFP_KERNEL|__GFP_ZERO); + if (!gref->page) + goto undo; + + /* Grant foreign access to the page. */ + gref->gref_id = gnttab_grant_foreign_access(op->domid, + pfn_to_mfn(page_to_pfn(gref->page)), readonly); + if (gref->gref_id < 0) { + rc = gref->gref_id; + goto undo; + } + gref_ids[i] = gref->gref_id; + } + + /* Add to gref lists. */ + spin_lock(&gref_lock); + list_splice_tail(&queue_gref, &gref_list); + list_splice_tail(&queue_file, &priv->list); + spin_unlock(&gref_lock); + + return 0; + +undo: + spin_lock(&gref_lock); + gref_size -= (op->count - i); + + list_for_each_entry(gref, &queue_file, next_file) { + /* __del_gref does not remove from queue_file */ + __del_gref(gref); + } + + /* It's possible for the target domain to map the just-allocated grant + * references by blindly guessing their IDs; if this is done, then + * __del_gref will leave them in the queue_gref list. They need to be + * added to the global list so that we can free them when they are no + * longer referenced. + */ + if (unlikely(!list_empty(&queue_gref))) + list_splice_tail(&queue_gref, &gref_list); + spin_unlock(&gref_lock); + return rc; +} + +static void __del_gref(struct gntalloc_gref *gref) +{ + if (gref->notify.flags & UNMAP_NOTIFY_CLEAR_BYTE) { + uint8_t *tmp = kmap(gref->page); + tmp[gref->notify.pgoff] = 0; + kunmap(gref->page); + } + if (gref->notify.flags & UNMAP_NOTIFY_SEND_EVENT) + notify_remote_via_evtchn(gref->notify.event); + + gref->notify.flags = 0; + + if (gref->gref_id > 0) { + if (gnttab_query_foreign_access(gref->gref_id)) + return; + + if (!gnttab_end_foreign_access_ref(gref->gref_id, 0)) + return; + } + + gref_size--; + list_del(&gref->next_gref); + + if (gref->page) + __free_page(gref->page); + + kfree(gref); +} + +/* finds contiguous grant references in a file, returns the first */ +static struct gntalloc_gref *find_grefs(struct gntalloc_file_private_data *priv, + uint64_t index, uint32_t count) +{ + struct gntalloc_gref *rv = NULL, *gref; + list_for_each_entry(gref, &priv->list, next_file) { + if (gref->file_index == index && !rv) + rv = gref; + if (rv) { + if (gref->file_index != index) + return NULL; + index += PAGE_SIZE; + count--; + if (count == 0) + return rv; + } + } + return NULL; +} + +/* + * ------------------------------------- + * File operations. + * ------------------------------------- + */ +static int gntalloc_open(struct inode *inode, struct file *filp) +{ + struct gntalloc_file_private_data *priv; + + priv = kzalloc(sizeof(*priv), GFP_KERNEL); + if (!priv) + goto out_nomem; + INIT_LIST_HEAD(&priv->list); + + filp->private_data = priv; + + pr_debug("%s: priv %p\n", __func__, priv); + + return 0; + +out_nomem: + return -ENOMEM; +} + +static int gntalloc_release(struct inode *inode, struct file *filp) +{ + struct gntalloc_file_private_data *priv = filp->private_data; + struct gntalloc_gref *gref; + + pr_debug("%s: priv %p\n", __func__, priv); + + spin_lock(&gref_lock); + while (!list_empty(&priv->list)) { + gref = list_entry(priv->list.next, + struct gntalloc_gref, next_file); + list_del(&gref->next_file); + gref->users--; + if (gref->users == 0) + __del_gref(gref); + } + kfree(priv); + spin_unlock(&gref_lock); + + return 0; +} + +static long gntalloc_ioctl_alloc(struct gntalloc_file_private_data *priv, + struct ioctl_gntalloc_alloc_gref __user *arg) +{ + int rc = 0; + struct ioctl_gntalloc_alloc_gref op; + uint32_t *gref_ids; + + pr_debug("%s: priv %p\n", __func__, priv); + + if (copy_from_user(&op, arg, sizeof(op))) { + rc = -EFAULT; + goto out; + } + + gref_ids = kzalloc(sizeof(gref_ids[0]) * op.count, GFP_TEMPORARY); + if (!gref_ids) { + rc = -ENOMEM; + goto out; + } + + spin_lock(&gref_lock); + /* Clean up pages that were at zero (local) users but were still mapped + * by remote domains. Since those pages count towards the limit that we + * are about to enforce, removing them here is a good idea. + */ + do_cleanup(); + if (gref_size + op.count > limit) { + spin_unlock(&gref_lock); + rc = -ENOSPC; + goto out_free; + } + gref_size += op.count; + op.index = priv->index; + priv->index += op.count * PAGE_SIZE; + spin_unlock(&gref_lock); + + rc = add_grefs(&op, gref_ids, priv); + if (rc < 0) + goto out_free; + + /* Once we finish add_grefs, it is unsafe to touch the new reference, + * since it is possible for a concurrent ioctl to remove it (by guessing + * its index). If the userspace application doesn't provide valid memory + * to write the IDs to, then it will need to close the file in order to + * release - which it will do by segfaulting when it tries to access the + * IDs to close them. + */ + if (copy_to_user(arg, &op, sizeof(op))) { + rc = -EFAULT; + goto out_free; + } + if (copy_to_user(arg->gref_ids, gref_ids, + sizeof(gref_ids[0]) * op.count)) { + rc = -EFAULT; + goto out_free; + } + +out_free: + kfree(gref_ids); +out: + return rc; +} + +static long gntalloc_ioctl_dealloc(struct gntalloc_file_private_data *priv, + void __user *arg) +{ + int i, rc = 0; + struct ioctl_gntalloc_dealloc_gref op; + struct gntalloc_gref *gref, *n; + + pr_debug("%s: priv %p\n", __func__, priv); + + if (copy_from_user(&op, arg, sizeof(op))) { + rc = -EFAULT; + goto dealloc_grant_out; + } + + spin_lock(&gref_lock); + gref = find_grefs(priv, op.index, op.count); + if (gref) { + /* Remove from the file list only, and decrease reference count. + * The later call to do_cleanup() will remove from gref_list and + * free the memory if the pages aren't mapped anywhere. + */ + for (i = 0; i < op.count; i++) { + n = list_entry(gref->next_file.next, + struct gntalloc_gref, next_file); + list_del(&gref->next_file); + gref->users--; + gref = n; + } + } else { + rc = -EINVAL; + } + + do_cleanup(); + + spin_unlock(&gref_lock); +dealloc_grant_out: + return rc; +} + +static long gntalloc_ioctl_unmap_notify(struct gntalloc_file_private_data *priv, + void __user *arg) +{ + struct ioctl_gntalloc_unmap_notify op; + struct gntalloc_gref *gref; + uint64_t index; + int pgoff; + int rc; + + if (copy_from_user(&op, arg, sizeof(op))) + return -EFAULT; + + index = op.index & ~(PAGE_SIZE - 1); + pgoff = op.index & (PAGE_SIZE - 1); + + spin_lock(&gref_lock); + + gref = find_grefs(priv, index, 1); + if (!gref) { + rc = -ENOENT; + goto unlock_out; + } + + if (op.action & ~(UNMAP_NOTIFY_CLEAR_BYTE|UNMAP_NOTIFY_SEND_EVENT)) { + rc = -EINVAL; + goto unlock_out; + } + + gref->notify.flags = op.action; + gref->notify.pgoff = pgoff; + gref->notify.event = op.event_channel_port; + rc = 0; + unlock_out: + spin_unlock(&gref_lock); + return rc; +} + +static long gntalloc_ioctl(struct file *filp, unsigned int cmd, + unsigned long arg) +{ + struct gntalloc_file_private_data *priv = filp->private_data; + + switch (cmd) { + case IOCTL_GNTALLOC_ALLOC_GREF: + return gntalloc_ioctl_alloc(priv, (void __user *)arg); + + case IOCTL_GNTALLOC_DEALLOC_GREF: + return gntalloc_ioctl_dealloc(priv, (void __user *)arg); + + case IOCTL_GNTALLOC_SET_UNMAP_NOTIFY: + return gntalloc_ioctl_unmap_notify(priv, (void __user *)arg); + + default: + return -ENOIOCTLCMD; + } + + return 0; +} + +static void gntalloc_vma_close(struct vm_area_struct *vma) +{ + struct gntalloc_gref *gref = vma->vm_private_data; + if (!gref) + return; + + spin_lock(&gref_lock); + gref->users--; + if (gref->users == 0) + __del_gref(gref); + spin_unlock(&gref_lock); +} + +static struct vm_operations_struct gntalloc_vmops = { + .close = gntalloc_vma_close, +}; + +static int gntalloc_mmap(struct file *filp, struct vm_area_struct *vma) +{ + struct gntalloc_file_private_data *priv = filp->private_data; + struct gntalloc_gref *gref; + int count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT; + int rv, i; + + pr_debug("%s: priv %p, page %lu+%d\n", __func__, + priv, vma->vm_pgoff, count); + + if (!(vma->vm_flags & VM_SHARED)) { + printk(KERN_ERR "%s: Mapping must be shared.\n", __func__); + return -EINVAL; + } + + spin_lock(&gref_lock); + gref = find_grefs(priv, vma->vm_pgoff << PAGE_SHIFT, count); + if (gref == NULL) { + rv = -ENOENT; + pr_debug("%s: Could not find grant reference", + __func__); + goto out_unlock; + } + + vma->vm_private_data = gref; + + vma->vm_flags |= VM_RESERVED; + vma->vm_flags |= VM_DONTCOPY; + vma->vm_flags |= VM_PFNMAP | VM_PFN_AT_MMAP; + + vma->vm_ops = &gntalloc_vmops; + + for (i = 0; i < count; i++) { + gref->users++; + rv = vm_insert_page(vma, vma->vm_start + i * PAGE_SIZE, + gref->page); + if (rv) + goto out_unlock; + + gref = list_entry(gref->next_file.next, + struct gntalloc_gref, next_file); + } + rv = 0; + +out_unlock: + spin_unlock(&gref_lock); + return rv; +} + +static const struct file_operations gntalloc_fops = { + .owner = THIS_MODULE, + .open = gntalloc_open, + .release = gntalloc_release, + .unlocked_ioctl = gntalloc_ioctl, + .mmap = gntalloc_mmap +}; + +/* + * ------------------------------------- + * Module creation/destruction. + * ------------------------------------- + */ +static struct miscdevice gntalloc_miscdev = { + .minor = MISC_DYNAMIC_MINOR, + .name = "xen/gntalloc", + .fops = &gntalloc_fops, +}; + +static int __init gntalloc_init(void) +{ + int err; + + if (!xen_domain()) + return -ENODEV; + + err = misc_register(&gntalloc_miscdev); + if (err != 0) { + printk(KERN_ERR "Could not register misc gntalloc device\n"); + return err; + } + + pr_debug("Created grant allocation device at %d,%d\n", + MISC_MAJOR, gntalloc_miscdev.minor); + + return 0; +} + +static void __exit gntalloc_exit(void) +{ + misc_deregister(&gntalloc_miscdev); +} + +module_init(gntalloc_init); +module_exit(gntalloc_exit); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Carter Weatherly <carter.weatherly@jhuapl.edu>, " + "Daniel De Graaf <dgdegra@tycho.nsa.gov>"); +MODULE_DESCRIPTION("User-space grant reference allocator driver"); diff --git a/drivers/xen/gntdev.c b/drivers/xen/gntdev.c new file mode 100644 index 00000000000..017ce600fbc --- /dev/null +++ b/drivers/xen/gntdev.c @@ -0,0 +1,751 @@ +/****************************************************************************** + * gntdev.c + * + * Device for accessing (in user-space) pages that have been granted by other + * domains. + * + * Copyright (c) 2006-2007, D G Murray. + * (c) 2009 Gerd Hoffmann <kraxel@redhat.com> + * + * 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 + */ + +#undef DEBUG + +#include <linux/module.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/miscdevice.h> +#include <linux/fs.h> +#include <linux/mm.h> +#include <linux/mman.h> +#include <linux/mmu_notifier.h> +#include <linux/types.h> +#include <linux/uaccess.h> +#include <linux/sched.h> +#include <linux/spinlock.h> +#include <linux/slab.h> +#include <linux/highmem.h> + +#include <xen/xen.h> +#include <xen/grant_table.h> +#include <xen/balloon.h> +#include <xen/gntdev.h> +#include <xen/events.h> +#include <asm/xen/hypervisor.h> +#include <asm/xen/hypercall.h> +#include <asm/xen/page.h> + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Derek G. Murray <Derek.Murray@cl.cam.ac.uk>, " + "Gerd Hoffmann <kraxel@redhat.com>"); +MODULE_DESCRIPTION("User-space granted page access driver"); + +static int limit = 1024*1024; +module_param(limit, int, 0644); +MODULE_PARM_DESC(limit, "Maximum number of grants that may be mapped by " + "the gntdev device"); + +static atomic_t pages_mapped = ATOMIC_INIT(0); + +static int use_ptemod; + +struct gntdev_priv { + struct list_head maps; + /* lock protects maps from concurrent changes */ + spinlock_t lock; + struct mm_struct *mm; + struct mmu_notifier mn; +}; + +struct unmap_notify { + int flags; + /* Address relative to the start of the grant_map */ + int addr; + int event; +}; + +struct grant_map { + struct list_head next; + struct vm_area_struct *vma; + int index; + int count; + int flags; + atomic_t users; + struct unmap_notify notify; + struct ioctl_gntdev_grant_ref *grants; + struct gnttab_map_grant_ref *map_ops; + struct gnttab_unmap_grant_ref *unmap_ops; + struct page **pages; +}; + +static int unmap_grant_pages(struct grant_map *map, int offset, int pages); + +/* ------------------------------------------------------------------ */ + +static void gntdev_print_maps(struct gntdev_priv *priv, + char *text, int text_index) +{ +#ifdef DEBUG + struct grant_map *map; + + pr_debug("%s: maps list (priv %p)\n", __func__, priv); + list_for_each_entry(map, &priv->maps, next) + pr_debug(" index %2d, count %2d %s\n", + map->index, map->count, + map->index == text_index && text ? text : ""); +#endif +} + +static struct grant_map *gntdev_alloc_map(struct gntdev_priv *priv, int count) +{ + struct grant_map *add; + int i; + + add = kzalloc(sizeof(struct grant_map), GFP_KERNEL); + if (NULL == add) + return NULL; + + add->grants = kzalloc(sizeof(add->grants[0]) * count, GFP_KERNEL); + add->map_ops = kzalloc(sizeof(add->map_ops[0]) * count, GFP_KERNEL); + add->unmap_ops = kzalloc(sizeof(add->unmap_ops[0]) * count, GFP_KERNEL); + add->pages = kzalloc(sizeof(add->pages[0]) * count, GFP_KERNEL); + if (NULL == add->grants || + NULL == add->map_ops || + NULL == add->unmap_ops || + NULL == add->pages) + goto err; + + if (alloc_xenballooned_pages(count, add->pages)) + goto err; + + for (i = 0; i < count; i++) { + add->map_ops[i].handle = -1; + add->unmap_ops[i].handle = -1; + } + + add->index = 0; + add->count = count; + atomic_set(&add->users, 1); + + return add; + +err: + kfree(add->pages); + kfree(add->grants); + kfree(add->map_ops); + kfree(add->unmap_ops); + kfree(add); + return NULL; +} + +static void gntdev_add_map(struct gntdev_priv *priv, struct grant_map *add) +{ + struct grant_map *map; + + list_for_each_entry(map, &priv->maps, next) { + if (add->index + add->count < map->index) { + list_add_tail(&add->next, &map->next); + goto done; + } + add->index = map->index + map->count; + } + list_add_tail(&add->next, &priv->maps); + +done: + gntdev_print_maps(priv, "[new]", add->index); +} + +static struct grant_map *gntdev_find_map_index(struct gntdev_priv *priv, + int index, int count) +{ + struct grant_map *map; + + list_for_each_entry(map, &priv->maps, next) { + if (map->index != index) + continue; + if (count && map->count != count) + continue; + return map; + } + return NULL; +} + +static void gntdev_put_map(struct grant_map *map) +{ + if (!map) + return; + + if (!atomic_dec_and_test(&map->users)) + return; + + atomic_sub(map->count, &pages_mapped); + + if (map->notify.flags & UNMAP_NOTIFY_SEND_EVENT) { + notify_remote_via_evtchn(map->notify.event); + } + + if (map->pages) { + if (!use_ptemod) + unmap_grant_pages(map, 0, map->count); + + free_xenballooned_pages(map->count, map->pages); + } + kfree(map->pages); + kfree(map->grants); + kfree(map->map_ops); + kfree(map->unmap_ops); + kfree(map); +} + +/* ------------------------------------------------------------------ */ + +static int find_grant_ptes(pte_t *pte, pgtable_t token, + unsigned long addr, void *data) +{ + struct grant_map *map = data; + unsigned int pgnr = (addr - map->vma->vm_start) >> PAGE_SHIFT; + int flags = map->flags | GNTMAP_application_map | GNTMAP_contains_pte; + u64 pte_maddr; + + BUG_ON(pgnr >= map->count); + pte_maddr = arbitrary_virt_to_machine(pte).maddr; + + gnttab_set_map_op(&map->map_ops[pgnr], pte_maddr, flags, + map->grants[pgnr].ref, + map->grants[pgnr].domid); + gnttab_set_unmap_op(&map->unmap_ops[pgnr], pte_maddr, flags, + -1 /* handle */); + return 0; +} + +static int map_grant_pages(struct grant_map *map) +{ + int i, err = 0; + + if (!use_ptemod) { + /* Note: it could already be mapped */ + if (map->map_ops[0].handle != -1) + return 0; + for (i = 0; i < map->count; i++) { + unsigned long addr = (unsigned long) + pfn_to_kaddr(page_to_pfn(map->pages[i])); + gnttab_set_map_op(&map->map_ops[i], addr, map->flags, + map->grants[i].ref, + map->grants[i].domid); + gnttab_set_unmap_op(&map->unmap_ops[i], addr, + map->flags, -1 /* handle */); + } + } + + pr_debug("map %d+%d\n", map->index, map->count); + err = gnttab_map_refs(map->map_ops, map->pages, map->count); + if (err) + return err; + + for (i = 0; i < map->count; i++) { + if (map->map_ops[i].status) + err = -EINVAL; + else { + BUG_ON(map->map_ops[i].handle == -1); + map->unmap_ops[i].handle = map->map_ops[i].handle; + pr_debug("map handle=%d\n", map->map_ops[i].handle); + } + } + return err; +} + +static int __unmap_grant_pages(struct grant_map *map, int offset, int pages) +{ + int i, err = 0; + + if (map->notify.flags & UNMAP_NOTIFY_CLEAR_BYTE) { + int pgno = (map->notify.addr >> PAGE_SHIFT); + if (pgno >= offset && pgno < offset + pages && use_ptemod) { + void __user *tmp = (void __user *) + map->vma->vm_start + map->notify.addr; + err = copy_to_user(tmp, &err, 1); + if (err) + return err; + map->notify.flags &= ~UNMAP_NOTIFY_CLEAR_BYTE; + } else if (pgno >= offset && pgno < offset + pages) { + uint8_t *tmp = kmap(map->pages[pgno]); + tmp[map->notify.addr & (PAGE_SIZE-1)] = 0; + kunmap(map->pages[pgno]); + map->notify.flags &= ~UNMAP_NOTIFY_CLEAR_BYTE; + } + } + + err = gnttab_unmap_refs(map->unmap_ops + offset, map->pages + offset, pages); + if (err) + return err; + + for (i = 0; i < pages; i++) { + if (map->unmap_ops[offset+i].status) + err = -EINVAL; + pr_debug("unmap handle=%d st=%d\n", + map->unmap_ops[offset+i].handle, + map->unmap_ops[offset+i].status); + map->unmap_ops[offset+i].handle = -1; + } + return err; +} + +static int unmap_grant_pages(struct grant_map *map, int offset, int pages) +{ + int range, err = 0; + + pr_debug("unmap %d+%d [%d+%d]\n", map->index, map->count, offset, pages); + + /* It is possible the requested range will have a "hole" where we + * already unmapped some of the grants. Only unmap valid ranges. + */ + while (pages && !err) { + while (pages && map->unmap_ops[offset].handle == -1) { + offset++; + pages--; + } + range = 0; + while (range < pages) { + if (map->unmap_ops[offset+range].handle == -1) { + range--; + break; + } + range++; + } + err = __unmap_grant_pages(map, offset, range); + offset += range; + pages -= range; + } + + return err; +} + +/* ------------------------------------------------------------------ */ + +static void gntdev_vma_close(struct vm_area_struct *vma) +{ + struct grant_map *map = vma->vm_private_data; + + pr_debug("close %p\n", vma); + map->vma = NULL; + vma->vm_private_data = NULL; + gntdev_put_map(map); +} + +static struct vm_operations_struct gntdev_vmops = { + .close = gntdev_vma_close, +}; + +/* ------------------------------------------------------------------ */ + +static void mn_invl_range_start(struct mmu_notifier *mn, + struct mm_struct *mm, + unsigned long start, unsigned long end) +{ + struct gntdev_priv *priv = container_of(mn, struct gntdev_priv, mn); + struct grant_map *map; + unsigned long mstart, mend; + int err; + + spin_lock(&priv->lock); + list_for_each_entry(map, &priv->maps, next) { + if (!map->vma) + continue; + if (map->vma->vm_start >= end) + continue; + if (map->vma->vm_end <= start) + continue; + mstart = max(start, map->vma->vm_start); + mend = min(end, map->vma->vm_end); + pr_debug("map %d+%d (%lx %lx), range %lx %lx, mrange %lx %lx\n", + map->index, map->count, + map->vma->vm_start, map->vma->vm_end, + start, end, mstart, mend); + err = unmap_grant_pages(map, + (mstart - map->vma->vm_start) >> PAGE_SHIFT, + (mend - mstart) >> PAGE_SHIFT); + WARN_ON(err); + } + spin_unlock(&priv->lock); +} + +static void mn_invl_page(struct mmu_notifier *mn, + struct mm_struct *mm, + unsigned long address) +{ + mn_invl_range_start(mn, mm, address, address + PAGE_SIZE); +} + +static void mn_release(struct mmu_notifier *mn, + struct mm_struct *mm) +{ + struct gntdev_priv *priv = container_of(mn, struct gntdev_priv, mn); + struct grant_map *map; + int err; + + spin_lock(&priv->lock); + list_for_each_entry(map, &priv->maps, next) { + if (!map->vma) + continue; + pr_debug("map %d+%d (%lx %lx)\n", + map->index, map->count, + map->vma->vm_start, map->vma->vm_end); + err = unmap_grant_pages(map, /* offset */ 0, map->count); + WARN_ON(err); + } + spin_unlock(&priv->lock); +} + +struct mmu_notifier_ops gntdev_mmu_ops = { + .release = mn_release, + .invalidate_page = mn_invl_page, + .invalidate_range_start = mn_invl_range_start, +}; + +/* ------------------------------------------------------------------ */ + +static int gntdev_open(struct inode *inode, struct file *flip) +{ + struct gntdev_priv *priv; + int ret = 0; + + priv = kzalloc(sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + INIT_LIST_HEAD(&priv->maps); + spin_lock_init(&priv->lock); + + if (use_ptemod) { + priv->mm = get_task_mm(current); + if (!priv->mm) { + kfree(priv); + return -ENOMEM; + } + priv->mn.ops = &gntdev_mmu_ops; + ret = mmu_notifier_register(&priv->mn, priv->mm); + mmput(priv->mm); + } + + if (ret) { + kfree(priv); + return ret; + } + + flip->private_data = priv; + pr_debug("priv %p\n", priv); + + return 0; +} + +static int gntdev_release(struct inode *inode, struct file *flip) +{ + struct gntdev_priv *priv = flip->private_data; + struct grant_map *map; + + pr_debug("priv %p\n", priv); + + spin_lock(&priv->lock); + while (!list_empty(&priv->maps)) { + map = list_entry(priv->maps.next, struct grant_map, next); + list_del(&map->next); + gntdev_put_map(map); + } + spin_unlock(&priv->lock); + + if (use_ptemod) + mmu_notifier_unregister(&priv->mn, priv->mm); + kfree(priv); + return 0; +} + +static long gntdev_ioctl_map_grant_ref(struct gntdev_priv *priv, + struct ioctl_gntdev_map_grant_ref __user *u) +{ + struct ioctl_gntdev_map_grant_ref op; + struct grant_map *map; + int err; + + if (copy_from_user(&op, u, sizeof(op)) != 0) + return -EFAULT; + pr_debug("priv %p, add %d\n", priv, op.count); + if (unlikely(op.count <= 0)) + return -EINVAL; + + err = -ENOMEM; + map = gntdev_alloc_map(priv, op.count); + if (!map) + return err; + + if (unlikely(atomic_add_return(op.count, &pages_mapped) > limit)) { + pr_debug("can't map: over limit\n"); + gntdev_put_map(map); + return err; + } + + if (copy_from_user(map->grants, &u->refs, + sizeof(map->grants[0]) * op.count) != 0) { + gntdev_put_map(map); + return err; + } + + spin_lock(&priv->lock); + gntdev_add_map(priv, map); + op.index = map->index << PAGE_SHIFT; + spin_unlock(&priv->lock); + + if (copy_to_user(u, &op, sizeof(op)) != 0) + return -EFAULT; + + return 0; +} + +static long gntdev_ioctl_unmap_grant_ref(struct gntdev_priv *priv, + struct ioctl_gntdev_unmap_grant_ref __user *u) +{ + struct ioctl_gntdev_unmap_grant_ref op; + struct grant_map *map; + int err = -ENOENT; + + if (copy_from_user(&op, u, sizeof(op)) != 0) + return -EFAULT; + pr_debug("priv %p, del %d+%d\n", priv, (int)op.index, (int)op.count); + + spin_lock(&priv->lock); + map = gntdev_find_map_index(priv, op.index >> PAGE_SHIFT, op.count); + if (map) { + list_del(&map->next); + gntdev_put_map(map); + err = 0; + } + spin_unlock(&priv->lock); + return err; +} + +static long gntdev_ioctl_get_offset_for_vaddr(struct gntdev_priv *priv, + struct ioctl_gntdev_get_offset_for_vaddr __user *u) +{ + struct ioctl_gntdev_get_offset_for_vaddr op; + struct vm_area_struct *vma; + struct grant_map *map; + + if (copy_from_user(&op, u, sizeof(op)) != 0) + return -EFAULT; + pr_debug("priv %p, offset for vaddr %lx\n", priv, (unsigned long)op.vaddr); + + vma = find_vma(current->mm, op.vaddr); + if (!vma || vma->vm_ops != &gntdev_vmops) + return -EINVAL; + + map = vma->vm_private_data; + if (!map) + return -EINVAL; + + op.offset = map->index << PAGE_SHIFT; + op.count = map->count; + + if (copy_to_user(u, &op, sizeof(op)) != 0) + return -EFAULT; + return 0; +} + +static long gntdev_ioctl_notify(struct gntdev_priv *priv, void __user *u) +{ + struct ioctl_gntdev_unmap_notify op; + struct grant_map *map; + int rc; + + if (copy_from_user(&op, u, sizeof(op))) + return -EFAULT; + + if (op.action & ~(UNMAP_NOTIFY_CLEAR_BYTE|UNMAP_NOTIFY_SEND_EVENT)) + return -EINVAL; + + spin_lock(&priv->lock); + + list_for_each_entry(map, &priv->maps, next) { + uint64_t begin = map->index << PAGE_SHIFT; + uint64_t end = (map->index + map->count) << PAGE_SHIFT; + if (op.index >= begin && op.index < end) + goto found; + } + rc = -ENOENT; + goto unlock_out; + + found: + if ((op.action & UNMAP_NOTIFY_CLEAR_BYTE) && + (map->flags & GNTMAP_readonly)) { + rc = -EINVAL; + goto unlock_out; + } + + map->notify.flags = op.action; + map->notify.addr = op.index - (map->index << PAGE_SHIFT); + map->notify.event = op.event_channel_port; + rc = 0; + unlock_out: + spin_unlock(&priv->lock); + return rc; +} + +static long gntdev_ioctl(struct file *flip, + unsigned int cmd, unsigned long arg) +{ + struct gntdev_priv *priv = flip->private_data; + void __user *ptr = (void __user *)arg; + + switch (cmd) { + case IOCTL_GNTDEV_MAP_GRANT_REF: + return gntdev_ioctl_map_grant_ref(priv, ptr); + + case IOCTL_GNTDEV_UNMAP_GRANT_REF: + return gntdev_ioctl_unmap_grant_ref(priv, ptr); + + case IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR: + return gntdev_ioctl_get_offset_for_vaddr(priv, ptr); + + case IOCTL_GNTDEV_SET_UNMAP_NOTIFY: + return gntdev_ioctl_notify(priv, ptr); + + default: + pr_debug("priv %p, unknown cmd %x\n", priv, cmd); + return -ENOIOCTLCMD; + } + + return 0; +} + +static int gntdev_mmap(struct file *flip, struct vm_area_struct *vma) +{ + struct gntdev_priv *priv = flip->private_data; + int index = vma->vm_pgoff; + int count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT; + struct grant_map *map; + int i, err = -EINVAL; + + if ((vma->vm_flags & VM_WRITE) && !(vma->vm_flags & VM_SHARED)) + return -EINVAL; + + pr_debug("map %d+%d at %lx (pgoff %lx)\n", + index, count, vma->vm_start, vma->vm_pgoff); + + spin_lock(&priv->lock); + map = gntdev_find_map_index(priv, index, count); + if (!map) + goto unlock_out; + if (use_ptemod && map->vma) + goto unlock_out; + if (use_ptemod && priv->mm != vma->vm_mm) { + printk(KERN_WARNING "Huh? Other mm?\n"); + goto unlock_out; + } + + atomic_inc(&map->users); + + vma->vm_ops = &gntdev_vmops; + + vma->vm_flags |= VM_RESERVED|VM_DONTCOPY|VM_DONTEXPAND|VM_PFNMAP; + + vma->vm_private_data = map; + + if (use_ptemod) + map->vma = vma; + + if (map->flags) { + if ((vma->vm_flags & VM_WRITE) && + (map->flags & GNTMAP_readonly)) + return -EINVAL; + } else { + map->flags = GNTMAP_host_map; + if (!(vma->vm_flags & VM_WRITE)) + map->flags |= GNTMAP_readonly; + } + + spin_unlock(&priv->lock); + + if (use_ptemod) { + err = apply_to_page_range(vma->vm_mm, vma->vm_start, + vma->vm_end - vma->vm_start, + find_grant_ptes, map); + if (err) { + printk(KERN_WARNING "find_grant_ptes() failure.\n"); + goto out_put_map; + } + } + + err = map_grant_pages(map); + if (err) + goto out_put_map; + + if (!use_ptemod) { + for (i = 0; i < count; i++) { + err = vm_insert_page(vma, vma->vm_start + i*PAGE_SIZE, + map->pages[i]); + if (err) + goto out_put_map; + } + } + + return 0; + +unlock_out: + spin_unlock(&priv->lock); + return err; + +out_put_map: + if (use_ptemod) + map->vma = NULL; + gntdev_put_map(map); + return err; +} + +static const struct file_operations gntdev_fops = { + .owner = THIS_MODULE, + .open = gntdev_open, + .release = gntdev_release, + .mmap = gntdev_mmap, + .unlocked_ioctl = gntdev_ioctl +}; + +static struct miscdevice gntdev_miscdev = { + .minor = MISC_DYNAMIC_MINOR, + .name = "xen/gntdev", + .fops = &gntdev_fops, +}; + +/* ------------------------------------------------------------------ */ + +static int __init gntdev_init(void) +{ + int err; + + if (!xen_domain()) + return -ENODEV; + + use_ptemod = xen_pv_domain(); + + err = misc_register(&gntdev_miscdev); + if (err != 0) { + printk(KERN_ERR "Could not register gntdev device\n"); + return err; + } + return 0; +} + +static void __exit gntdev_exit(void) +{ + misc_deregister(&gntdev_miscdev); +} + +module_init(gntdev_init); +module_exit(gntdev_exit); + +/* ------------------------------------------------------------------ */ diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c index 6c453181649..3745a318def 100644 --- a/drivers/xen/grant-table.c +++ b/drivers/xen/grant-table.c @@ -447,6 +447,62 @@ unsigned int gnttab_max_grant_frames(void) } EXPORT_SYMBOL_GPL(gnttab_max_grant_frames); +int gnttab_map_refs(struct gnttab_map_grant_ref *map_ops, + struct page **pages, unsigned int count) +{ + int i, ret; + pte_t *pte; + unsigned long mfn; + + ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map_ops, count); + if (ret) + return ret; + + if (xen_feature(XENFEAT_auto_translated_physmap)) + return ret; + + for (i = 0; i < count; i++) { + /* Do not add to override if the map failed. */ + if (map_ops[i].status) + continue; + + /* m2p override only supported for GNTMAP_contains_pte mappings */ + if (!(map_ops[i].flags & GNTMAP_contains_pte)) + continue; + pte = (pte_t *) (mfn_to_virt(PFN_DOWN(map_ops[i].host_addr)) + + (map_ops[i].host_addr & ~PAGE_MASK)); + mfn = pte_mfn(*pte); + ret = m2p_add_override(mfn, pages[i]); + if (ret) + return ret; + } + + return ret; +} +EXPORT_SYMBOL_GPL(gnttab_map_refs); + +int gnttab_unmap_refs(struct gnttab_unmap_grant_ref *unmap_ops, + struct page **pages, unsigned int count) +{ + int i, ret; + + ret = HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, unmap_ops, count); + if (ret) + return ret; + + if (xen_feature(XENFEAT_auto_translated_physmap)) + return ret; + + for (i = 0; i < count; i++) { + ret = m2p_remove_override(pages[i]); + if (ret) + return ret; + } + + return ret; +} +EXPORT_SYMBOL_GPL(gnttab_unmap_refs); + static int gnttab_map(unsigned int start_idx, unsigned int end_idx) { struct gnttab_setup_table setup; diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c index db8c4c4ac88..95143dd6904 100644 --- a/drivers/xen/manage.c +++ b/drivers/xen/manage.c @@ -34,58 +34,62 @@ enum shutdown_state { /* Ignore multiple shutdown requests. */ static enum shutdown_state shutting_down = SHUTDOWN_INVALID; -#ifdef CONFIG_PM_SLEEP -static int xen_hvm_suspend(void *data) -{ - struct sched_shutdown r = { .reason = SHUTDOWN_suspend }; - int *cancelled = data; - - BUG_ON(!irqs_disabled()); - - *cancelled = HYPERVISOR_sched_op(SCHEDOP_shutdown, &r); +struct suspend_info { + int cancelled; + unsigned long arg; /* extra hypercall argument */ + void (*pre)(void); + void (*post)(int cancelled); +}; - xen_hvm_post_suspend(*cancelled); +static void xen_hvm_post_suspend(int cancelled) +{ + xen_arch_hvm_post_suspend(cancelled); gnttab_resume(); +} - if (!*cancelled) { - xen_irq_resume(); - xen_console_resume(); - xen_timer_resume(); - } +static void xen_pre_suspend(void) +{ + xen_mm_pin_all(); + gnttab_suspend(); + xen_arch_pre_suspend(); +} - return 0; +static void xen_post_suspend(int cancelled) +{ + xen_arch_post_suspend(cancelled); + gnttab_resume(); + xen_mm_unpin_all(); } +#ifdef CONFIG_HIBERNATION static int xen_suspend(void *data) { + struct suspend_info *si = data; int err; - int *cancelled = data; BUG_ON(!irqs_disabled()); - err = sysdev_suspend(PMSG_SUSPEND); + err = sysdev_suspend(PMSG_FREEZE); if (err) { printk(KERN_ERR "xen_suspend: sysdev_suspend failed: %d\n", err); return err; } - xen_mm_pin_all(); - gnttab_suspend(); - xen_pre_suspend(); + if (si->pre) + si->pre(); /* * This hypercall returns 1 if suspend was cancelled * or the domain was merely checkpointed, and 0 if it * is resuming in a new domain. */ - *cancelled = HYPERVISOR_suspend(virt_to_mfn(xen_start_info)); + si->cancelled = HYPERVISOR_suspend(si->arg); - xen_post_suspend(*cancelled); - gnttab_resume(); - xen_mm_unpin_all(); + if (si->post) + si->post(si->cancelled); - if (!*cancelled) { + if (!si->cancelled) { xen_irq_resume(); xen_console_resume(); xen_timer_resume(); @@ -99,7 +103,7 @@ static int xen_suspend(void *data) static void do_suspend(void) { int err; - int cancelled = 1; + struct suspend_info si; shutting_down = SHUTDOWN_SUSPEND; @@ -114,7 +118,7 @@ static void do_suspend(void) } #endif - err = dpm_suspend_start(PMSG_SUSPEND); + err = dpm_suspend_start(PMSG_FREEZE); if (err) { printk(KERN_ERR "xen suspend: dpm_suspend_start %d\n", err); goto out_thaw; @@ -123,32 +127,41 @@ static void do_suspend(void) printk(KERN_DEBUG "suspending xenstore...\n"); xs_suspend(); - err = dpm_suspend_noirq(PMSG_SUSPEND); + err = dpm_suspend_noirq(PMSG_FREEZE); if (err) { printk(KERN_ERR "dpm_suspend_noirq failed: %d\n", err); goto out_resume; } - if (xen_hvm_domain()) - err = stop_machine(xen_hvm_suspend, &cancelled, cpumask_of(0)); - else - err = stop_machine(xen_suspend, &cancelled, cpumask_of(0)); + si.cancelled = 1; + + if (xen_hvm_domain()) { + si.arg = 0UL; + si.pre = NULL; + si.post = &xen_hvm_post_suspend; + } else { + si.arg = virt_to_mfn(xen_start_info); + si.pre = &xen_pre_suspend; + si.post = &xen_post_suspend; + } + + err = stop_machine(xen_suspend, &si, cpumask_of(0)); - dpm_resume_noirq(PMSG_RESUME); + dpm_resume_noirq(si.cancelled ? PMSG_THAW : PMSG_RESTORE); if (err) { printk(KERN_ERR "failed to start xen_suspend: %d\n", err); - cancelled = 1; + si.cancelled = 1; } out_resume: - if (!cancelled) { + if (!si.cancelled) { xen_arch_resume(); xs_resume(); } else xs_suspend_cancel(); - dpm_resume_end(PMSG_RESUME); + dpm_resume_end(si.cancelled ? PMSG_THAW : PMSG_RESTORE); /* Make sure timer events get retriggered on all CPUs */ clock_was_set(); @@ -160,7 +173,24 @@ out: #endif shutting_down = SHUTDOWN_INVALID; } -#endif /* CONFIG_PM_SLEEP */ +#endif /* CONFIG_HIBERNATION */ + +struct shutdown_handler { + const char *command; + void (*cb)(void); +}; + +static void do_poweroff(void) +{ + shutting_down = SHUTDOWN_POWEROFF; + orderly_poweroff(false); +} + +static void do_reboot(void) +{ + shutting_down = SHUTDOWN_POWEROFF; /* ? */ + ctrl_alt_del(); +} static void shutdown_handler(struct xenbus_watch *watch, const char **vec, unsigned int len) @@ -168,6 +198,16 @@ static void shutdown_handler(struct xenbus_watch *watch, char *str; struct xenbus_transaction xbt; int err; + static struct shutdown_handler handlers[] = { + { "poweroff", do_poweroff }, + { "halt", do_poweroff }, + { "reboot", do_reboot }, +#ifdef CONFIG_HIBERNATION + { "suspend", do_suspend }, +#endif + {NULL, NULL}, + }; + static struct shutdown_handler *handler; if (shutting_down != SHUTDOWN_INVALID) return; @@ -184,7 +224,14 @@ static void shutdown_handler(struct xenbus_watch *watch, return; } - xenbus_write(xbt, "control", "shutdown", ""); + for (handler = &handlers[0]; handler->command; handler++) { + if (strcmp(str, handler->command) == 0) + break; + } + + /* Only acknowledge commands which we are prepared to handle. */ + if (handler->cb) + xenbus_write(xbt, "control", "shutdown", ""); err = xenbus_transaction_end(xbt, 0); if (err == -EAGAIN) { @@ -192,17 +239,8 @@ static void shutdown_handler(struct xenbus_watch *watch, goto again; } - if (strcmp(str, "poweroff") == 0 || - strcmp(str, "halt") == 0) { - shutting_down = SHUTDOWN_POWEROFF; - orderly_poweroff(false); - } else if (strcmp(str, "reboot") == 0) { - shutting_down = SHUTDOWN_POWEROFF; /* ? */ - ctrl_alt_del(); -#ifdef CONFIG_PM_SLEEP - } else if (strcmp(str, "suspend") == 0) { - do_suspend(); -#endif + if (handler->cb) { + handler->cb(); } else { printk(KERN_INFO "Ignoring shutdown request: %s\n", str); shutting_down = SHUTDOWN_INVALID; @@ -281,27 +319,18 @@ static int shutdown_event(struct notifier_block *notifier, return NOTIFY_DONE; } -static int __init __setup_shutdown_event(void) -{ - /* Delay initialization in the PV on HVM case */ - if (xen_hvm_domain()) - return 0; - - if (!xen_pv_domain()) - return -ENODEV; - - return xen_setup_shutdown_event(); -} - int xen_setup_shutdown_event(void) { static struct notifier_block xenstore_notifier = { .notifier_call = shutdown_event }; + + if (!xen_domain()) + return -ENODEV; register_xenstore_notifier(&xenstore_notifier); return 0; } EXPORT_SYMBOL_GPL(xen_setup_shutdown_event); -subsys_initcall(__setup_shutdown_event); +subsys_initcall(xen_setup_shutdown_event); diff --git a/drivers/xen/platform-pci.c b/drivers/xen/platform-pci.c index c01b5ddce52..319dd0a94d5 100644 --- a/drivers/xen/platform-pci.c +++ b/drivers/xen/platform-pci.c @@ -105,7 +105,7 @@ static int __devinit platform_pci_init(struct pci_dev *pdev, const struct pci_device_id *ent) { int i, ret; - long ioaddr, iolen; + long ioaddr; long mmio_addr, mmio_len; unsigned int max_nr_gframes; @@ -114,7 +114,6 @@ static int __devinit platform_pci_init(struct pci_dev *pdev, 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); @@ -125,19 +124,13 @@ static int __devinit platform_pci_init(struct pci_dev *pdev, goto pci_out; } - if (request_mem_region(mmio_addr, mmio_len, DRV_NAME) == NULL) { - dev_err(&pdev->dev, "MEM I/O resource 0x%lx @ 0x%lx busy\n", - mmio_addr, mmio_len); - ret = -EBUSY; + ret = pci_request_region(pdev, 1, DRV_NAME); + if (ret < 0) goto pci_out; - } - if (request_region(ioaddr, iolen, DRV_NAME) == NULL) { - dev_err(&pdev->dev, "I/O resource 0x%lx @ 0x%lx busy\n", - iolen, ioaddr); - ret = -EBUSY; + ret = pci_request_region(pdev, 0, DRV_NAME); + if (ret < 0) goto mem_out; - } platform_mmio = mmio_addr; platform_mmiolen = mmio_len; @@ -163,15 +156,12 @@ static int __devinit platform_pci_init(struct pci_dev *pdev, if (ret) goto out; xenbus_probe(NULL); - ret = xen_setup_shutdown_event(); - if (ret) - goto out; return 0; out: - release_region(ioaddr, iolen); + pci_release_region(pdev, 0); mem_out: - release_mem_region(mmio_addr, mmio_len); + pci_release_region(pdev, 1); pci_out: pci_disable_device(pdev); return ret; diff --git a/drivers/xen/xen-balloon.c b/drivers/xen/xen-balloon.c new file mode 100644 index 00000000000..a4ff225ee86 --- /dev/null +++ b/drivers/xen/xen-balloon.c @@ -0,0 +1,256 @@ +/****************************************************************************** + * 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 <linux/kernel.h> +#include <linux/module.h> +#include <linux/sysdev.h> +#include <linux/capability.h> + +#include <xen/xen.h> +#include <xen/interface/xen.h> +#include <xen/balloon.h> +#include <xen/xenbus.h> +#include <xen/features.h> +#include <xen/page.h> + +#define PAGES2KB(_p) ((_p)<<(PAGE_SHIFT-10)) + +#define BALLOON_CLASS_NAME "xen_memory" + +static struct sys_device balloon_sysdev; + +static int register_balloon(struct sys_device *sysdev); + +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; +} + +static struct notifier_block xenstore_notifier; + +static int __init balloon_init(void) +{ + if (!xen_domain()) + return -ENODEV; + + pr_info("xen-balloon: Initialising balloon driver.\n"); + + register_balloon(&balloon_sysdev); + + 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 balloon_exit(void) +{ + /* XXX - release balloon here */ + return; +} + +module_exit(balloon_exit); + +#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(balloon_stats.current_pages)); +BALLOON_SHOW(low_kb, "%lu\n", PAGES2KB(balloon_stats.balloon_low)); +BALLOON_SHOW(high_kb, "%lu\n", PAGES2KB(balloon_stats.balloon_high)); + +static SYSDEV_ULONG_ATTR(schedule_delay, 0444, balloon_stats.schedule_delay); +static SYSDEV_ULONG_ATTR(max_schedule_delay, 0644, balloon_stats.max_schedule_delay); +static SYSDEV_ULONG_ATTR(retry_count, 0444, balloon_stats.retry_count); +static SYSDEV_ULONG_ATTR(max_retry_count, 0644, balloon_stats.max_retry_count); + +static ssize_t show_target_kb(struct sys_device *dev, struct sysdev_attribute *attr, + char *buf) +{ + return sprintf(buf, "%lu\n", PAGES2KB(balloon_stats.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; + + target_bytes = simple_strtoull(buf, &endchar, 0) * 1024; + + 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; + + 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, + &attr_schedule_delay.attr, + &attr_max_schedule_delay.attr, + &attr_retry_count.attr, + &attr_max_retry_count.attr +}; + +static struct attribute *balloon_info_attrs[] = { + &attr_current_kb.attr, + &attr_low_kb.attr, + &attr_high_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 int 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; +} + +MODULE_LICENSE("GPL"); diff --git a/drivers/xen/xenbus/Makefile b/drivers/xen/xenbus/Makefile index 5571f5b8422..8dca685358b 100644 --- a/drivers/xen/xenbus/Makefile +++ b/drivers/xen/xenbus/Makefile @@ -5,3 +5,8 @@ xenbus-objs += xenbus_client.o xenbus-objs += xenbus_comms.o xenbus-objs += xenbus_xs.o xenbus-objs += xenbus_probe.o + +xenbus-be-objs-$(CONFIG_XEN_BACKEND) += xenbus_probe_backend.o +xenbus-objs += $(xenbus-be-objs-y) + +obj-$(CONFIG_XEN_XENBUS_FRONTEND) += xenbus_probe_frontend.o diff --git a/drivers/xen/xenbus/xenbus_probe.c b/drivers/xen/xenbus/xenbus_probe.c index deb9c4ba3a9..739769551e3 100644 --- a/drivers/xen/xenbus/xenbus_probe.c +++ b/drivers/xen/xenbus/xenbus_probe.c @@ -56,7 +56,6 @@ #include <xen/events.h> #include <xen/page.h> -#include <xen/platform_pci.h> #include <xen/hvm.h> #include "xenbus_comms.h" @@ -73,15 +72,6 @@ static unsigned long xen_store_mfn; 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); - -static int xenbus_dev_suspend(struct device *dev, pm_message_t state); -static int xenbus_dev_resume(struct device *dev); - /* 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) @@ -102,34 +92,7 @@ int xenbus_match(struct device *_dev, struct device_driver *_drv) return match_device(drv->ids, to_xenbus_device(_dev)) != NULL; } - -static int xenbus_uevent(struct device *_dev, struct kobj_uevent_env *env) -{ - struct xenbus_device *dev = to_xenbus_device(_dev); - - if (add_uevent_var(env, "MODALIAS=xen:%s", dev->devicetype)) - return -ENOMEM; - - return 0; -} - -/* device/<type>/<id> => <type>-<id> */ -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; -} +EXPORT_SYMBOL_GPL(xenbus_match); static void free_otherend_details(struct xenbus_device *dev) @@ -149,7 +112,30 @@ static void free_otherend_watch(struct xenbus_device *dev) } -int read_otherend_details(struct xenbus_device *xendev, +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) +{ + struct xen_bus_type *bus = + container_of(dev->dev.bus, struct xen_bus_type, bus); + + return xenbus_watch_pathfmt(dev, &dev->otherend_watch, + bus->otherend_changed, + "%s/%s", dev->otherend, "state"); +} + + +int xenbus_read_otherend_details(struct xenbus_device *xendev, char *id_node, char *path_node) { int err = xenbus_gather(XBT_NIL, xendev->nodename, @@ -174,39 +160,11 @@ int read_otherend_details(struct xenbus_device *xendev, return 0; } +EXPORT_SYMBOL_GPL(xenbus_read_otherend_details); - -static int read_backend_details(struct xenbus_device *xendev) -{ - return read_otherend_details(xendev, "backend-id", "backend"); -} - -static struct device_attribute xenbus_dev_attrs[] = { - __ATTR_NULL -}; - -/* Bus type for frontend drivers. */ -static struct xen_bus_type xenbus_frontend = { - .root = "device", - .levels = 2, /* device/type/<id> */ - .get_bus_id = frontend_bus_id, - .probe = xenbus_probe_frontend, - .bus = { - .name = "xen", - .match = xenbus_match, - .uevent = xenbus_uevent, - .probe = xenbus_dev_probe, - .remove = xenbus_dev_remove, - .shutdown = xenbus_dev_shutdown, - .dev_attrs = xenbus_dev_attrs, - - .suspend = xenbus_dev_suspend, - .resume = xenbus_dev_resume, - }, -}; - -static void otherend_changed(struct xenbus_watch *watch, - const char **vec, unsigned int len) +void xenbus_otherend_changed(struct xenbus_watch *watch, + const char **vec, unsigned int len, + int ignore_on_shutdown) { struct xenbus_device *dev = container_of(watch, struct xenbus_device, otherend_watch); @@ -234,11 +192,7 @@ static void otherend_changed(struct xenbus_watch *watch, * 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)) + if (ignore_on_shutdown && (state == XenbusStateClosing)) xenbus_frontend_closed(dev); return; } @@ -246,25 +200,7 @@ static void otherend_changed(struct xenbus_watch *watch, 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) -{ - return xenbus_watch_pathfmt(dev, &dev->otherend_watch, otherend_changed, - "%s/%s", dev->otherend, "state"); -} - +EXPORT_SYMBOL_GPL(xenbus_otherend_changed); int xenbus_dev_probe(struct device *_dev) { @@ -308,8 +244,9 @@ int xenbus_dev_probe(struct device *_dev) fail: xenbus_dev_error(dev, err, "xenbus_dev_probe on %s", dev->nodename); xenbus_switch_state(dev, XenbusStateClosed); - return -ENODEV; + return err; } +EXPORT_SYMBOL_GPL(xenbus_dev_probe); int xenbus_dev_remove(struct device *_dev) { @@ -327,8 +264,9 @@ int xenbus_dev_remove(struct device *_dev) xenbus_switch_state(dev, XenbusStateClosed); return 0; } +EXPORT_SYMBOL_GPL(xenbus_dev_remove); -static void xenbus_dev_shutdown(struct device *_dev) +void xenbus_dev_shutdown(struct device *_dev) { struct xenbus_device *dev = to_xenbus_device(_dev); unsigned long timeout = 5*HZ; @@ -349,6 +287,7 @@ static void xenbus_dev_shutdown(struct device *_dev) out: put_device(&dev->dev); } +EXPORT_SYMBOL_GPL(xenbus_dev_shutdown); int xenbus_register_driver_common(struct xenbus_driver *drv, struct xen_bus_type *bus, @@ -362,25 +301,7 @@ int xenbus_register_driver_common(struct xenbus_driver *drv, return driver_register(&drv->driver); } - -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); +EXPORT_SYMBOL_GPL(xenbus_register_driver_common); void xenbus_unregister_driver(struct xenbus_driver *drv) { @@ -551,24 +472,7 @@ fail: kfree(xendev); return err; } - -/* device/<typename>/<name> */ -static int xenbus_probe_frontend(const char *type, const char *name) -{ - char *nodename; - int err; - - 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; -} +EXPORT_SYMBOL_GPL(xenbus_probe_node); static int xenbus_probe_device_type(struct xen_bus_type *bus, const char *type) { @@ -582,10 +486,11 @@ static int xenbus_probe_device_type(struct xen_bus_type *bus, const char *type) return PTR_ERR(dir); for (i = 0; i < dir_n; i++) { - err = bus->probe(type, dir[i]); + err = bus->probe(bus, type, dir[i]); if (err) break; } + kfree(dir); return err; } @@ -605,9 +510,11 @@ int xenbus_probe_devices(struct xen_bus_type *bus) if (err) break; } + kfree(dir); return err; } +EXPORT_SYMBOL_GPL(xenbus_probe_devices); static unsigned int char_count(const char *str, char c) { @@ -670,54 +577,39 @@ void xenbus_dev_changed(const char *node, struct xen_bus_type *bus) } EXPORT_SYMBOL_GPL(xenbus_dev_changed); -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, -}; - -static int xenbus_dev_suspend(struct device *dev, pm_message_t state) +int xenbus_dev_suspend(struct device *dev) { int err = 0; struct xenbus_driver *drv; - struct xenbus_device *xdev; + struct xenbus_device *xdev + = container_of(dev, struct xenbus_device, dev); - DPRINTK(""); + DPRINTK("%s", xdev->nodename); if (dev->driver == NULL) return 0; drv = to_xenbus_driver(dev->driver); - xdev = container_of(dev, struct xenbus_device, dev); if (drv->suspend) - err = drv->suspend(xdev, state); + err = drv->suspend(xdev); if (err) printk(KERN_WARNING "xenbus: suspend %s failed: %i\n", dev_name(dev), err); return 0; } +EXPORT_SYMBOL_GPL(xenbus_dev_suspend); -static int xenbus_dev_resume(struct device *dev) +int xenbus_dev_resume(struct device *dev) { int err; struct xenbus_driver *drv; - struct xenbus_device *xdev; + struct xenbus_device *xdev + = container_of(dev, struct xenbus_device, dev); - DPRINTK(""); + DPRINTK("%s", xdev->nodename); 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 @@ -748,6 +640,15 @@ static int xenbus_dev_resume(struct device *dev) return 0; } +EXPORT_SYMBOL_GPL(xenbus_dev_resume); + +int xenbus_dev_cancel(struct device *dev) +{ + /* Do nothing */ + DPRINTK("cancel"); + return 0; +} +EXPORT_SYMBOL_GPL(xenbus_dev_cancel); /* A flag to determine if xenstored is 'ready' (i.e. has started) */ int xenstored_ready = 0; @@ -776,11 +677,6 @@ void xenbus_probe(struct work_struct *unused) { xenstored_ready = 1; - /* 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); } @@ -809,16 +705,7 @@ static int __init xenbus_init(void) err = -ENODEV; if (!xen_domain()) - goto out_error; - - /* Register ourselves with the kernel bus subsystem */ - err = bus_register(&xenbus_frontend.bus); - if (err) - goto out_error; - - err = xenbus_backend_bus_register(); - if (err) - goto out_unreg_front; + return err; /* * Domain0 doesn't have a store_evtchn or store_mfn yet. @@ -874,7 +761,7 @@ static int __init xenbus_init(void) if (err) { printk(KERN_WARNING "XENBUS: Error initializing xenstore comms: %i\n", err); - goto out_unreg_back; + goto out_error; } #ifdef CONFIG_XEN_COMPAT_XENFS @@ -887,133 +774,13 @@ static int __init xenbus_init(void) return 0; - out_unreg_back: - xenbus_backend_bus_unregister(); - - out_unreg_front: - bus_unregister(&xenbus_frontend.bus); - out_error: if (page != 0) free_page(page); + return err; } postcore_initcall(xenbus_init); MODULE_LICENSE("GPL"); - -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) -{ - 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; - - /* 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); - } else 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); - } - - 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 || !xen_domain()) - 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 MODULE -static int __init boot_wait_for_devices(void) -{ - if (xen_hvm_domain() && !xen_platform_pci_unplug) - return -ENODEV; - - ready_to_wait_for_devices = 1; - wait_for_devices(NULL); - return 0; -} - -late_initcall(boot_wait_for_devices); -#endif diff --git a/drivers/xen/xenbus/xenbus_probe.h b/drivers/xen/xenbus/xenbus_probe.h index 6c5e3185a6a..888b9900ca0 100644 --- a/drivers/xen/xenbus/xenbus_probe.h +++ b/drivers/xen/xenbus/xenbus_probe.h @@ -36,26 +36,15 @@ #define XEN_BUS_ID_SIZE 20 -#ifdef CONFIG_XEN_BACKEND -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 int xenbus_backend_bus_register(void); -extern void xenbus_backend_bus_unregister(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 int xenbus_backend_bus_register(void) { return 0; } -static inline void xenbus_backend_bus_unregister(void) {} -#endif - struct xen_bus_type { char *root; 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); + int (*probe)(struct xen_bus_type *bus, const char *type, + const char *dir); + void (*otherend_changed)(struct xenbus_watch *watch, const char **vec, + unsigned int len); struct bus_type bus; }; @@ -73,4 +62,17 @@ extern int xenbus_probe_devices(struct xen_bus_type *bus); extern void xenbus_dev_changed(const char *node, struct xen_bus_type *bus); +extern void xenbus_dev_shutdown(struct device *_dev); + +extern int xenbus_dev_suspend(struct device *dev); +extern int xenbus_dev_resume(struct device *dev); +extern int xenbus_dev_cancel(struct device *dev); + +extern void xenbus_otherend_changed(struct xenbus_watch *watch, + const char **vec, unsigned int len, + int ignore_on_shutdown); + +extern int xenbus_read_otherend_details(struct xenbus_device *xendev, + char *id_node, char *path_node); + #endif diff --git a/drivers/xen/xenbus/xenbus_probe_backend.c b/drivers/xen/xenbus/xenbus_probe_backend.c new file mode 100644 index 00000000000..6cf467bf63e --- /dev/null +++ b/drivers/xen/xenbus/xenbus_probe_backend.c @@ -0,0 +1,276 @@ +/****************************************************************************** + * 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", \ + __func__, __LINE__, ##args) + +#include <linux/kernel.h> +#include <linux/err.h> +#include <linux/string.h> +#include <linux/ctype.h> +#include <linux/fcntl.h> +#include <linux/mm.h> +#include <linux/notifier.h> + +#include <asm/page.h> +#include <asm/pgtable.h> +#include <asm/xen/hypervisor.h> +#include <asm/hypervisor.h> +#include <xen/xenbus.h> +#include <xen/features.h> + +#include "xenbus_comms.h" +#include "xenbus_probe.h" + +/* backend/<type>/<fe-uuid>/<id> => <type>-<fe-domid>-<id> */ +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 int xenbus_uevent_backend(struct device *dev, + struct kobj_uevent_env *env) +{ + struct xenbus_device *xdev; + struct xenbus_driver *drv; + struct xen_bus_type *bus; + + DPRINTK(""); + + if (dev == NULL) + return -ENODEV; + + xdev = to_xenbus_device(dev); + bus = container_of(xdev->dev.bus, struct xen_bus_type, bus); + if (xdev == NULL) + return -ENODEV; + + /* stuff we want to pass to /sbin/hotplug */ + if (add_uevent_var(env, "XENBUS_TYPE=%s", xdev->devicetype)) + return -ENOMEM; + + if (add_uevent_var(env, "XENBUS_PATH=%s", xdev->nodename)) + return -ENOMEM; + + if (add_uevent_var(env, "XENBUS_BASE_PATH=%s", bus->root)) + return -ENOMEM; + + if (dev->driver) { + drv = to_xenbus_driver(dev->driver); + if (drv && drv->uevent) + return drv->uevent(xdev, env); + } + + return 0; +} + +/* backend/<typename>/<frontend-uuid>/<name> */ +static int xenbus_probe_backend_unit(struct xen_bus_type *bus, + 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(bus, type, nodename); + kfree(nodename); + return err; +} + +/* backend/<typename>/<frontend-domid> */ +static int xenbus_probe_backend(struct xen_bus_type *bus, 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", bus->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(bus, nodename, type, dir[i]); + if (err) + break; + } + kfree(dir); + kfree(nodename); + return err; +} + +static void frontend_changed(struct xenbus_watch *watch, + const char **vec, unsigned int len) +{ + xenbus_otherend_changed(watch, vec, len, 0); +} + +static struct device_attribute xenbus_backend_dev_attrs[] = { + __ATTR_NULL +}; + +static struct xen_bus_type xenbus_backend = { + .root = "backend", + .levels = 3, /* backend/type/<frontend>/<id> */ + .get_bus_id = backend_bus_id, + .probe = xenbus_probe_backend, + .otherend_changed = frontend_changed, + .bus = { + .name = "xen-backend", + .match = xenbus_match, + .uevent = xenbus_uevent_backend, + .probe = xenbus_dev_probe, + .remove = xenbus_dev_remove, + .shutdown = xenbus_dev_shutdown, + .dev_attrs = xenbus_backend_dev_attrs, + }, +}; + +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, +}; + +static int read_frontend_details(struct xenbus_device *xendev) +{ + return xenbus_read_otherend_details(xendev, "frontend-id", "frontend"); +} + +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); + +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); + +static int backend_probe_and_watch(struct notifier_block *notifier, + unsigned long event, + void *data) +{ + /* Enumerate devices in xenstore and watch for changes. */ + xenbus_probe_devices(&xenbus_backend); + register_xenbus_watch(&be_watch); + + return NOTIFY_DONE; +} + +static int __init xenbus_probe_backend_init(void) +{ + static struct notifier_block xenstore_notifier = { + .notifier_call = backend_probe_and_watch + }; + int err; + + DPRINTK(""); + + /* Register ourselves with the kernel bus subsystem */ + err = bus_register(&xenbus_backend.bus); + if (err) + return err; + + register_xenstore_notifier(&xenstore_notifier); + + return 0; +} +subsys_initcall(xenbus_probe_backend_init); diff --git a/drivers/xen/xenbus/xenbus_probe_frontend.c b/drivers/xen/xenbus/xenbus_probe_frontend.c new file mode 100644 index 00000000000..b6a2690c9d4 --- /dev/null +++ b/drivers/xen/xenbus/xenbus_probe_frontend.c @@ -0,0 +1,301 @@ +#define DPRINTK(fmt, args...) \ + pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \ + __func__, __LINE__, ##args) + +#include <linux/kernel.h> +#include <linux/err.h> +#include <linux/string.h> +#include <linux/ctype.h> +#include <linux/fcntl.h> +#include <linux/mm.h> +#include <linux/proc_fs.h> +#include <linux/notifier.h> +#include <linux/kthread.h> +#include <linux/mutex.h> +#include <linux/io.h> + +#include <asm/page.h> +#include <asm/pgtable.h> +#include <asm/xen/hypervisor.h> +#include <xen/xenbus.h> +#include <xen/events.h> +#include <xen/page.h> + +#include <xen/platform_pci.h> + +#include "xenbus_comms.h" +#include "xenbus_probe.h" + + +/* device/<type>/<id> => <type>-<id> */ +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; +} + +/* device/<typename>/<name> */ +static int xenbus_probe_frontend(struct xen_bus_type *bus, const char *type, + const char *name) +{ + char *nodename; + int err; + + nodename = kasprintf(GFP_KERNEL, "%s/%s/%s", bus->root, type, name); + if (!nodename) + return -ENOMEM; + + DPRINTK("%s", nodename); + + err = xenbus_probe_node(bus, type, nodename); + kfree(nodename); + return err; +} + +static int xenbus_uevent_frontend(struct device *_dev, + struct kobj_uevent_env *env) +{ + struct xenbus_device *dev = to_xenbus_device(_dev); + + if (add_uevent_var(env, "MODALIAS=xen:%s", dev->devicetype)) + return -ENOMEM; + + return 0; +} + + +static void backend_changed(struct xenbus_watch *watch, + const char **vec, unsigned int len) +{ + xenbus_otherend_changed(watch, vec, len, 1); +} + +static struct device_attribute xenbus_frontend_dev_attrs[] = { + __ATTR_NULL +}; + +static const struct dev_pm_ops xenbus_pm_ops = { + .suspend = xenbus_dev_suspend, + .resume = xenbus_dev_resume, + .freeze = xenbus_dev_suspend, + .thaw = xenbus_dev_cancel, + .restore = xenbus_dev_resume, +}; + +static struct xen_bus_type xenbus_frontend = { + .root = "device", + .levels = 2, /* device/type/<id> */ + .get_bus_id = frontend_bus_id, + .probe = xenbus_probe_frontend, + .otherend_changed = backend_changed, + .bus = { + .name = "xen", + .match = xenbus_match, + .uevent = xenbus_uevent_frontend, + .probe = xenbus_dev_probe, + .remove = xenbus_dev_remove, + .shutdown = xenbus_dev_shutdown, + .dev_attrs = xenbus_frontend_dev_attrs, + + .pm = &xenbus_pm_ops, + }, +}; + +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, +}; + +static int read_backend_details(struct xenbus_device *xendev) +{ + return xenbus_read_otherend_details(xendev, "backend-id", "backend"); +} + +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) +{ + 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; + + /* 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); + } else 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); + } + + 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 || !xen_domain()) + 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); +} + +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); + +static int frontend_probe_and_watch(struct notifier_block *notifier, + unsigned long event, + void *data) +{ + /* Enumerate devices in xenstore and watch for changes. */ + xenbus_probe_devices(&xenbus_frontend); + register_xenbus_watch(&fe_watch); + + return NOTIFY_DONE; +} + + +static int __init xenbus_probe_frontend_init(void) +{ + static struct notifier_block xenstore_notifier = { + .notifier_call = frontend_probe_and_watch + }; + int err; + + DPRINTK(""); + + /* Register ourselves with the kernel bus subsystem */ + err = bus_register(&xenbus_frontend.bus); + if (err) + return err; + + register_xenstore_notifier(&xenstore_notifier); + + return 0; +} +subsys_initcall(xenbus_probe_frontend_init); + +#ifndef MODULE +static int __init boot_wait_for_devices(void) +{ + if (xen_hvm_domain() && !xen_platform_pci_unplug) + return -ENODEV; + + ready_to_wait_for_devices = 1; + wait_for_devices(NULL); + return 0; +} + +late_initcall(boot_wait_for_devices); +#endif + +MODULE_LICENSE("GPL"); diff --git a/drivers/xen/xenfs/xenbus.c b/drivers/xen/xenfs/xenbus.c index 1c1236087f7..bbd000f88af 100644 --- a/drivers/xen/xenfs/xenbus.c +++ b/drivers/xen/xenfs/xenbus.c @@ -122,6 +122,7 @@ static ssize_t xenbus_file_read(struct file *filp, int ret; mutex_lock(&u->reply_mutex); +again: while (list_empty(&u->read_buffers)) { mutex_unlock(&u->reply_mutex); if (filp->f_flags & O_NONBLOCK) @@ -144,7 +145,7 @@ static ssize_t xenbus_file_read(struct file *filp, i += sz - ret; rb->cons += sz - ret; - if (ret != sz) { + if (ret != 0) { if (i == 0) i = -EFAULT; goto out; @@ -160,6 +161,8 @@ static ssize_t xenbus_file_read(struct file *filp, struct read_buffer, list); } } + if (i == 0) + goto again; out: mutex_unlock(&u->reply_mutex); @@ -407,6 +410,7 @@ static int xenbus_write_watch(unsigned msg_type, struct xenbus_file_priv *u) mutex_lock(&u->reply_mutex); rc = queue_reply(&u->read_buffers, &reply, sizeof(reply)); + wake_up(&u->read_waitq); mutex_unlock(&u->reply_mutex); } @@ -455,7 +459,7 @@ static ssize_t xenbus_file_write(struct file *filp, ret = copy_from_user(u->u.buffer + u->len, ubuf, len); - if (ret == len) { + if (ret != 0) { rc = -EFAULT; goto out; } @@ -488,21 +492,6 @@ static ssize_t xenbus_file_write(struct file *filp, msg_type = u->u.msg.type; switch (msg_type) { - case XS_TRANSACTION_START: - case XS_TRANSACTION_END: - case XS_DIRECTORY: - case XS_READ: - case XS_GET_PERMS: - case XS_RELEASE: - case XS_GET_DOMAIN_PATH: - case XS_WRITE: - case XS_MKDIR: - case XS_RM: - case XS_SET_PERMS: - /* Send out a transaction */ - ret = xenbus_write_transaction(msg_type, u); - break; - case XS_WATCH: case XS_UNWATCH: /* (Un)Ask for some path to be watched for changes */ @@ -510,7 +499,8 @@ static ssize_t xenbus_file_write(struct file *filp, break; default: - ret = -EINVAL; + /* Send out a transaction */ + ret = xenbus_write_transaction(msg_type, u); break; } if (ret != 0) @@ -555,6 +545,7 @@ static int xenbus_file_release(struct inode *inode, struct file *filp) struct xenbus_file_priv *u = filp->private_data; struct xenbus_transaction_holder *trans, *tmp; struct watch_adapter *watch, *tmp_watch; + struct read_buffer *rb, *tmp_rb; /* * No need for locking here because there are no other users, @@ -573,6 +564,10 @@ static int xenbus_file_release(struct inode *inode, struct file *filp) free_watch_adapter(watch); } + list_for_each_entry_safe(rb, tmp_rb, &u->read_buffers, list) { + list_del(&rb->list); + kfree(rb); + } kfree(u); return 0; |