From 325d73bf8fea8af2227240b7305253fb052d3a68 Mon Sep 17 00:00:00 2001 From: Bob Liu Date: Fri, 3 Apr 2015 14:42:58 +0800 Subject: [PATCH 01/10] xen/blkback: safely unmap purge persistent grants MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit c43cf3ea8385 ("xen-blkback: safely unmap grants in case they are still in use") use gnttab_unmap_refs_async() to wait until the mapped pages are no longer in use before unmapping them, but that commit missed the persistent case. Purge persistent pages can't be unmapped either unless no longer in use. Signed-off-by: Bob Liu Acked-by: Roger Pau Monné Signed-off-by: David Vrabel --- drivers/block/xen-blkback/blkback.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c index bd2b3bbbb22c..48e98f2712b5 100644 --- a/drivers/block/xen-blkback/blkback.c +++ b/drivers/block/xen-blkback/blkback.c @@ -329,8 +329,18 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST]; struct page *pages[BLKIF_MAX_SEGMENTS_PER_REQUEST]; struct persistent_gnt *persistent_gnt; - int ret, segs_to_unmap = 0; + int segs_to_unmap = 0; struct xen_blkif *blkif = container_of(work, typeof(*blkif), persistent_purge_work); + struct gntab_unmap_queue_data unmap_data; + struct completion unmap_completion; + + init_completion(&unmap_completion); + + unmap_data.data = &unmap_completion; + unmap_data.done = &free_persistent_gnts_unmap_callback; + unmap_data.pages = pages; + unmap_data.unmap_ops = unmap; + unmap_data.kunmap_ops = NULL; while(!list_empty(&blkif->persistent_purge_list)) { persistent_gnt = list_first_entry(&blkif->persistent_purge_list, @@ -346,17 +356,19 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) pages[segs_to_unmap] = persistent_gnt->page; if (++segs_to_unmap == BLKIF_MAX_SEGMENTS_PER_REQUEST) { - ret = gnttab_unmap_refs(unmap, NULL, pages, - segs_to_unmap); - BUG_ON(ret); + unmap_data.count = segs_to_unmap; + gnttab_unmap_refs_async(&unmap_data); + wait_for_completion(&unmap_completion); + put_free_pages(blkif, pages, segs_to_unmap); segs_to_unmap = 0; } kfree(persistent_gnt); } if (segs_to_unmap > 0) { - ret = gnttab_unmap_refs(unmap, NULL, pages, segs_to_unmap); - BUG_ON(ret); + unmap_data.count = segs_to_unmap; + gnttab_unmap_refs_async(&unmap_data); + wait_for_completion(&unmap_completion); put_free_pages(blkif, pages, segs_to_unmap); } } From b44166cd46e28dd608d5baa5873047a40f32919c Mon Sep 17 00:00:00 2001 From: Bob Liu Date: Fri, 3 Apr 2015 14:42:59 +0800 Subject: [PATCH 02/10] xen/grant: introduce func gnttab_unmap_refs_sync() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There are several place using gnttab async unmap and wait for completion, so move the common code to a function gnttab_unmap_refs_sync(). Signed-off-by: Bob Liu Acked-by: Roger Pau Monné Acked-by: Konrad Rzeszutek Wilk Signed-off-by: David Vrabel --- drivers/block/xen-blkback/blkback.c | 31 +++-------------------------- drivers/xen/gntdev.c | 28 +++----------------------- drivers/xen/grant-table.c | 28 ++++++++++++++++++++++++++ include/xen/grant_table.h | 1 + 4 files changed, 35 insertions(+), 53 deletions(-) diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c index 48e98f2712b5..713fc9ff1149 100644 --- a/drivers/block/xen-blkback/blkback.c +++ b/drivers/block/xen-blkback/blkback.c @@ -265,17 +265,6 @@ static void put_persistent_gnt(struct xen_blkif *blkif, atomic_dec(&blkif->persistent_gnt_in_use); } -static void free_persistent_gnts_unmap_callback(int result, - struct gntab_unmap_queue_data *data) -{ - struct completion *c = data->data; - - /* BUG_ON used to reproduce existing behaviour, - but is this the best way to deal with this? */ - BUG_ON(result); - complete(c); -} - static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, unsigned int num) { @@ -285,12 +274,7 @@ static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, struct rb_node *n; int segs_to_unmap = 0; struct gntab_unmap_queue_data unmap_data; - struct completion unmap_completion; - init_completion(&unmap_completion); - - unmap_data.data = &unmap_completion; - unmap_data.done = &free_persistent_gnts_unmap_callback; unmap_data.pages = pages; unmap_data.unmap_ops = unmap; unmap_data.kunmap_ops = NULL; @@ -310,8 +294,7 @@ static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, !rb_next(&persistent_gnt->node)) { unmap_data.count = segs_to_unmap; - gnttab_unmap_refs_async(&unmap_data); - wait_for_completion(&unmap_completion); + BUG_ON(gnttab_unmap_refs_sync(&unmap_data)); put_free_pages(blkif, pages, segs_to_unmap); segs_to_unmap = 0; @@ -332,12 +315,7 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) int segs_to_unmap = 0; struct xen_blkif *blkif = container_of(work, typeof(*blkif), persistent_purge_work); struct gntab_unmap_queue_data unmap_data; - struct completion unmap_completion; - init_completion(&unmap_completion); - - unmap_data.data = &unmap_completion; - unmap_data.done = &free_persistent_gnts_unmap_callback; unmap_data.pages = pages; unmap_data.unmap_ops = unmap; unmap_data.kunmap_ops = NULL; @@ -357,9 +335,7 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) if (++segs_to_unmap == BLKIF_MAX_SEGMENTS_PER_REQUEST) { unmap_data.count = segs_to_unmap; - gnttab_unmap_refs_async(&unmap_data); - wait_for_completion(&unmap_completion); - + BUG_ON(gnttab_unmap_refs_sync(&unmap_data)); put_free_pages(blkif, pages, segs_to_unmap); segs_to_unmap = 0; } @@ -367,8 +343,7 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) } if (segs_to_unmap > 0) { unmap_data.count = segs_to_unmap; - gnttab_unmap_refs_async(&unmap_data); - wait_for_completion(&unmap_completion); + BUG_ON(gnttab_unmap_refs_sync(&unmap_data)); put_free_pages(blkif, pages, segs_to_unmap); } } diff --git a/drivers/xen/gntdev.c b/drivers/xen/gntdev.c index d5bb1a33d0a3..89274850741b 100644 --- a/drivers/xen/gntdev.c +++ b/drivers/xen/gntdev.c @@ -327,30 +327,10 @@ static int map_grant_pages(struct grant_map *map) return err; } -struct unmap_grant_pages_callback_data -{ - struct completion completion; - int result; -}; - -static void unmap_grant_callback(int result, - struct gntab_unmap_queue_data *data) -{ - struct unmap_grant_pages_callback_data* d = data->data; - - d->result = result; - complete(&d->completion); -} - static int __unmap_grant_pages(struct grant_map *map, int offset, int pages) { int i, err = 0; struct gntab_unmap_queue_data unmap_data; - struct unmap_grant_pages_callback_data data; - - init_completion(&data.completion); - unmap_data.data = &data; - unmap_data.done= &unmap_grant_callback; if (map->notify.flags & UNMAP_NOTIFY_CLEAR_BYTE) { int pgno = (map->notify.addr >> PAGE_SHIFT); @@ -367,11 +347,9 @@ static int __unmap_grant_pages(struct grant_map *map, int offset, int pages) unmap_data.pages = map->pages + offset; unmap_data.count = pages; - gnttab_unmap_refs_async(&unmap_data); - - wait_for_completion(&data.completion); - if (data.result) - return data.result; + err = gnttab_unmap_refs_sync(&unmap_data); + if (err) + return err; for (i = 0; i < pages; i++) { if (map->unmap_ops[offset+i].status) diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c index 17972fbacddc..b1c7170e5c9e 100644 --- a/drivers/xen/grant-table.c +++ b/drivers/xen/grant-table.c @@ -123,6 +123,11 @@ struct gnttab_ops { int (*query_foreign_access)(grant_ref_t ref); }; +struct unmap_refs_callback_data { + struct completion completion; + int result; +}; + static struct gnttab_ops *gnttab_interface; static int grant_table_version; @@ -863,6 +868,29 @@ void gnttab_unmap_refs_async(struct gntab_unmap_queue_data* item) } EXPORT_SYMBOL_GPL(gnttab_unmap_refs_async); +static void unmap_refs_callback(int result, + struct gntab_unmap_queue_data *data) +{ + struct unmap_refs_callback_data *d = data->data; + + d->result = result; + complete(&d->completion); +} + +int gnttab_unmap_refs_sync(struct gntab_unmap_queue_data *item) +{ + struct unmap_refs_callback_data data; + + init_completion(&data.completion); + item->data = &data; + item->done = &unmap_refs_callback; + gnttab_unmap_refs_async(item); + wait_for_completion(&data.completion); + + return data.result; +} +EXPORT_SYMBOL_GPL(gnttab_unmap_refs_sync); + static int gnttab_map_frames_v1(xen_pfn_t *frames, unsigned int nr_gframes) { int rc; diff --git a/include/xen/grant_table.h b/include/xen/grant_table.h index 143ca5ffab7a..4478f4b4aae2 100644 --- a/include/xen/grant_table.h +++ b/include/xen/grant_table.h @@ -191,6 +191,7 @@ int gnttab_unmap_refs(struct gnttab_unmap_grant_ref *unmap_ops, struct gnttab_unmap_grant_ref *kunmap_ops, struct page **pages, unsigned int count); void gnttab_unmap_refs_async(struct gntab_unmap_queue_data* item); +int gnttab_unmap_refs_sync(struct gntab_unmap_queue_data *item); /* Perform a batch of grant map/copy operations. Retry every batch slot From 2b953a5e994ce279904ec70220f7d4f31d380a0a Mon Sep 17 00:00:00 2001 From: Boris Ostrovsky Date: Tue, 28 Apr 2015 18:46:20 -0400 Subject: [PATCH 03/10] xen: Suspend ticks on all CPUs during suspend Commit 77e32c89a711 ("clockevents: Manage device's state separately for the core") decouples clockevent device's modes from states. With this change when a Xen guest tries to resume, it won't be calling its set_mode op which needs to be done on each VCPU in order to make the hypervisor aware that we are in oneshot mode. This happens because clockevents_tick_resume() (which is an intermediate step of resuming ticks on a processor) doesn't call clockevents_set_state() anymore and because during suspend clockevent devices on all VCPUs (except for the one doing the suspend) are left in ONESHOT state. As result, during resume the clockevents state machine will assume that device is already where it should be and doesn't need to be updated. To avoid this problem we should suspend ticks on all VCPUs during suspend. Signed-off-by: Boris Ostrovsky Signed-off-by: David Vrabel --- arch/x86/xen/suspend.c | 10 ++++++++++ drivers/xen/manage.c | 9 ++++++--- include/xen/xen-ops.h | 1 + 3 files changed, 17 insertions(+), 3 deletions(-) diff --git a/arch/x86/xen/suspend.c b/arch/x86/xen/suspend.c index d9497698645a..53b4c0811f4f 100644 --- a/arch/x86/xen/suspend.c +++ b/arch/x86/xen/suspend.c @@ -88,7 +88,17 @@ static void xen_vcpu_notify_restore(void *data) tick_resume_local(); } +static void xen_vcpu_notify_suspend(void *data) +{ + tick_suspend_local(); +} + void xen_arch_resume(void) { on_each_cpu(xen_vcpu_notify_restore, NULL, 1); } + +void xen_arch_suspend(void) +{ + on_each_cpu(xen_vcpu_notify_suspend, NULL, 1); +} diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c index bf1940706422..9e6a85104a20 100644 --- a/drivers/xen/manage.c +++ b/drivers/xen/manage.c @@ -131,6 +131,8 @@ static void do_suspend(void) goto out_resume; } + xen_arch_suspend(); + si.cancelled = 1; err = stop_machine(xen_suspend, &si, cpumask_of(0)); @@ -148,11 +150,12 @@ static void do_suspend(void) si.cancelled = 1; } + xen_arch_resume(); + out_resume: - if (!si.cancelled) { - xen_arch_resume(); + if (!si.cancelled) xs_resume(); - } else + else xs_suspend_cancel(); dpm_resume_end(si.cancelled ? PMSG_THAW : PMSG_RESTORE); diff --git a/include/xen/xen-ops.h b/include/xen/xen-ops.h index c643e6a94c9a..0ce4f32017ea 100644 --- a/include/xen/xen-ops.h +++ b/include/xen/xen-ops.h @@ -13,6 +13,7 @@ void xen_arch_post_suspend(int suspend_cancelled); void xen_timer_resume(void); void xen_arch_resume(void); +void xen_arch_suspend(void); void xen_resume_notifier_register(struct notifier_block *nb); void xen_resume_notifier_unregister(struct notifier_block *nb); From 8014bcc86ef112eab9ee1db312dba4e6b608cf89 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Mon, 13 Apr 2015 00:26:35 +0100 Subject: [PATCH 04/10] xen-pciback: Add name prefix to global 'permissive' variable The variable for the 'permissive' module parameter used to be static but was recently changed to be extern. This puts it in the kernel global namespace if the driver is built-in, so its name should begin with a prefix identifying the driver. Signed-off-by: Ben Hutchings Fixes: af6fc858a35b ("xen-pciback: limit guest control of command register") Signed-off-by: David Vrabel --- drivers/xen/xen-pciback/conf_space.c | 6 +++--- drivers/xen/xen-pciback/conf_space.h | 2 +- drivers/xen/xen-pciback/conf_space_header.c | 2 +- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/xen/xen-pciback/conf_space.c b/drivers/xen/xen-pciback/conf_space.c index 75fe3d466515..9c234209d8b5 100644 --- a/drivers/xen/xen-pciback/conf_space.c +++ b/drivers/xen/xen-pciback/conf_space.c @@ -16,8 +16,8 @@ #include "conf_space.h" #include "conf_space_quirks.h" -bool permissive; -module_param(permissive, bool, 0644); +bool xen_pcibk_permissive; +module_param_named(permissive, xen_pcibk_permissive, bool, 0644); /* This is where xen_pcibk_read_config_byte, xen_pcibk_read_config_word, * xen_pcibk_write_config_word, and xen_pcibk_write_config_byte are created. */ @@ -262,7 +262,7 @@ int xen_pcibk_config_write(struct pci_dev *dev, int offset, int size, u32 value) * This means that some fields may still be read-only because * they have entries in the config_field list that intercept * the write and do nothing. */ - if (dev_data->permissive || permissive) { + if (dev_data->permissive || xen_pcibk_permissive) { switch (size) { case 1: err = pci_write_config_byte(dev, offset, diff --git a/drivers/xen/xen-pciback/conf_space.h b/drivers/xen/xen-pciback/conf_space.h index 2e1d73d1d5d0..62461a8ba1d6 100644 --- a/drivers/xen/xen-pciback/conf_space.h +++ b/drivers/xen/xen-pciback/conf_space.h @@ -64,7 +64,7 @@ struct config_field_entry { void *data; }; -extern bool permissive; +extern bool xen_pcibk_permissive; #define OFFSET(cfg_entry) ((cfg_entry)->base_offset+(cfg_entry)->field->offset) diff --git a/drivers/xen/xen-pciback/conf_space_header.c b/drivers/xen/xen-pciback/conf_space_header.c index c2260a0456c9..ad3d17d29c81 100644 --- a/drivers/xen/xen-pciback/conf_space_header.c +++ b/drivers/xen/xen-pciback/conf_space_header.c @@ -118,7 +118,7 @@ static int command_write(struct pci_dev *dev, int offset, u16 value, void *data) cmd->val = value; - if (!permissive && (!dev_data || !dev_data->permissive)) + if (!xen_pcibk_permissive && (!dev_data || !dev_data->permissive)) return 0; /* Only allow the guest to control certain bits. */ From 5cec98834989a014a9560b1841649eaca95cf00e Mon Sep 17 00:00:00 2001 From: Boris Ostrovsky Date: Wed, 29 Apr 2015 17:10:12 -0400 Subject: [PATCH 05/10] xen/events: Clear cpu_evtchn_mask before resuming When a guest is resumed, the hypervisor may change event channel assignments. If this happens and the guest uses 2-level events it is possible for the interrupt to be claimed by wrong VCPU since cpu_evtchn_mask bits may be stale. This can happen even though evtchn_2l_bind_to_cpu() attempts to clear old bits: irq_info that is passed in is not necessarily the original one (from pre-migration times) but instead is freshly allocated during resume and so any information about which CPU the channel was bound to is lost. Thus we should clear the mask during resume. We also need to make sure that bits for xenstore and console channels are set when these two subsystems are resumed. While rebind_evtchn_irq() (which is invoked for both of them on a resume) calls irq_set_affinity(), the latter will in fact postpone setting affinity until handling the interrupt. But because cpu_evtchn_mask will have bits for these two cleared we won't be able to take the interrupt. With that in mind, we need to bind those two channels explicitly in rebind_evtchn_irq(). We will keep irq_set_affinity() so that we have a pass through generic irq affinity code later, in case something needs to be updated there as well. (Also replace cpumask_of(0) with cpumask_of(info->cpu) in rebind_evtchn_irq(): it should be set to zero in preceding xen_irq_info_evtchn_setup().) Signed-off-by: Boris Ostrovsky Reported-by: Annie Li Cc: # 3.14+ Signed-off-by: David Vrabel --- drivers/xen/events/events_2l.c | 10 ++++++++++ drivers/xen/events/events_base.c | 5 +++-- 2 files changed, 13 insertions(+), 2 deletions(-) diff --git a/drivers/xen/events/events_2l.c b/drivers/xen/events/events_2l.c index 5db43fc100a4..7dd46312c180 100644 --- a/drivers/xen/events/events_2l.c +++ b/drivers/xen/events/events_2l.c @@ -345,6 +345,15 @@ irqreturn_t xen_debug_interrupt(int irq, void *dev_id) return IRQ_HANDLED; } +static void evtchn_2l_resume(void) +{ + int i; + + for_each_online_cpu(i) + memset(per_cpu(cpu_evtchn_mask, i), 0, sizeof(xen_ulong_t) * + EVTCHN_2L_NR_CHANNELS/BITS_PER_EVTCHN_WORD); +} + static const struct evtchn_ops evtchn_ops_2l = { .max_channels = evtchn_2l_max_channels, .nr_channels = evtchn_2l_max_channels, @@ -356,6 +365,7 @@ static const struct evtchn_ops evtchn_ops_2l = { .mask = evtchn_2l_mask, .unmask = evtchn_2l_unmask, .handle_events = evtchn_2l_handle_events, + .resume = evtchn_2l_resume, }; void __init xen_evtchn_2l_init(void) diff --git a/drivers/xen/events/events_base.c b/drivers/xen/events/events_base.c index 70fba973a107..a1ec564d791c 100644 --- a/drivers/xen/events/events_base.c +++ b/drivers/xen/events/events_base.c @@ -1279,8 +1279,9 @@ void rebind_evtchn_irq(int evtchn, int irq) mutex_unlock(&irq_mapping_update_lock); - /* new event channels are always bound to cpu 0 */ - irq_set_affinity(irq, cpumask_of(0)); + bind_evtchn_to_cpu(evtchn, info->cpu); + /* This will be deferred until interrupt is processed */ + irq_set_affinity(irq, cpumask_of(info->cpu)); /* Unmask the event channel. */ enable_irq(irq); From 16f1cf3ba7303228372d3756677bf7d10e79cf9f Mon Sep 17 00:00:00 2001 From: Boris Ostrovsky Date: Wed, 29 Apr 2015 17:10:13 -0400 Subject: [PATCH 06/10] xen/xenbus: Update xenbus event channel on resume After a resume the hypervisor/tools may change xenbus event channel number. We should re-query it. Signed-off-by: Boris Ostrovsky Cc: Signed-off-by: David Vrabel --- drivers/xen/xenbus/xenbus_probe.c | 29 +++++++++++++++++++++++++++++ 1 file changed, 29 insertions(+) diff --git a/drivers/xen/xenbus/xenbus_probe.c b/drivers/xen/xenbus/xenbus_probe.c index 564b31584860..5390a674b5e3 100644 --- a/drivers/xen/xenbus/xenbus_probe.c +++ b/drivers/xen/xenbus/xenbus_probe.c @@ -57,6 +57,7 @@ #include #include #include +#include #include #include @@ -735,6 +736,30 @@ static int __init xenstored_local_init(void) return err; } +static int xenbus_resume_cb(struct notifier_block *nb, + unsigned long action, void *data) +{ + int err = 0; + + if (xen_hvm_domain()) { + uint64_t v; + + err = hvm_get_parameter(HVM_PARAM_STORE_EVTCHN, &v); + if (!err && v) + xen_store_evtchn = v; + else + pr_warn("Cannot update xenstore event channel: %d\n", + err); + } else + xen_store_evtchn = xen_start_info->store_evtchn; + + return err; +} + +static struct notifier_block xenbus_resume_nb = { + .notifier_call = xenbus_resume_cb, +}; + static int __init xenbus_init(void) { int err = 0; @@ -793,6 +818,10 @@ static int __init xenbus_init(void) goto out_error; } + if ((xen_store_domain_type != XS_LOCAL) && + (xen_store_domain_type != XS_UNKNOWN)) + xen_resume_notifier_register(&xenbus_resume_nb); + #ifdef CONFIG_XEN_COMPAT_XENFS /* * Create xenfs mountpoint in /proc for compatibility with From b9d934f27c91b878c4b2e64299d6e419a4022f8d Mon Sep 17 00:00:00 2001 From: Boris Ostrovsky Date: Wed, 29 Apr 2015 17:10:14 -0400 Subject: [PATCH 07/10] xen/console: Update console event channel on resume After a resume the hypervisor/tools may change console event channel number. We should re-query it. Signed-off-by: Boris Ostrovsky Cc: Signed-off-by: David Vrabel --- drivers/tty/hvc/hvc_xen.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/drivers/tty/hvc/hvc_xen.c b/drivers/tty/hvc/hvc_xen.c index f1e57425e39f..5bab1c684bb1 100644 --- a/drivers/tty/hvc/hvc_xen.c +++ b/drivers/tty/hvc/hvc_xen.c @@ -299,11 +299,27 @@ static int xen_initial_domain_console_init(void) return 0; } +static void xen_console_update_evtchn(struct xencons_info *info) +{ + if (xen_hvm_domain()) { + uint64_t v; + int err; + + err = hvm_get_parameter(HVM_PARAM_CONSOLE_EVTCHN, &v); + if (!err && v) + info->evtchn = v; + } else + info->evtchn = xen_start_info->console.domU.evtchn; +} + void xen_console_resume(void) { struct xencons_info *info = vtermno_to_xencons(HVC_COOKIE); - if (info != NULL && info->irq) + if (info != NULL && info->irq) { + if (!xen_initial_domain()) + xen_console_update_evtchn(info); rebind_evtchn_irq(info->evtchn, info->irq); + } } static void xencons_disconnect_backend(struct xencons_info *info) From 16e6bd5970c88a2ac018b84a5f1dd5c2ff1fdf2c Mon Sep 17 00:00:00 2001 From: Boris Ostrovsky Date: Wed, 29 Apr 2015 17:10:15 -0400 Subject: [PATCH 08/10] xen/events: Set irq_info->evtchn before binding the channel to CPU in __startup_pirq() .. because bind_evtchn_to_cpu(evtchn, cpu) will map evtchn to 'info' and pass 'info' down to xen_evtchn_port_bind_to_cpu(). Signed-off-by: Boris Ostrovsky Tested-by: Annie Li Cc: Signed-off-by: David Vrabel --- drivers/xen/events/events_base.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/xen/events/events_base.c b/drivers/xen/events/events_base.c index a1ec564d791c..2b8553bd8715 100644 --- a/drivers/xen/events/events_base.c +++ b/drivers/xen/events/events_base.c @@ -529,8 +529,8 @@ static unsigned int __startup_pirq(unsigned int irq) if (rc) goto err; - bind_evtchn_to_cpu(evtchn, 0); info->evtchn = evtchn; + bind_evtchn_to_cpu(evtchn, 0); rc = xen_evtchn_port_setup(info); if (rc) From a71dbdaa8ca2933391b08e0ae5567083e3af0892 Mon Sep 17 00:00:00 2001 From: Boris Ostrovsky Date: Mon, 4 May 2015 11:02:15 -0400 Subject: [PATCH 09/10] hypervisor/x86/xen: Unset X86_BUG_SYSRET_SS_ATTRS on Xen PV guests Commit 61f01dd941ba ("x86_64, asm: Work around AMD SYSRET SS descriptor attribute issue") makes AMD processors set SS to __KERNEL_DS in __switch_to() to deal with cases when SS is NULL. This breaks Xen PV guests who do not want to load SS with__KERNEL_DS. Since the problem that the commit is trying to address would have to be fixed in the hypervisor (if it in fact exists under Xen) there is no reason to set X86_BUG_SYSRET_SS_ATTRS flag for PV VPCUs here. This can be easily achieved by adding x86_hyper_xen_hvm.set_cpu_features op which will clear this flag. (And since this structure is no longer HVM-specific we should do some renaming). Signed-off-by: Boris Ostrovsky Reported-by: Sander Eikelenboom Signed-off-by: David Vrabel --- arch/x86/include/asm/hypervisor.h | 2 +- arch/x86/kernel/cpu/hypervisor.c | 4 ++-- arch/x86/xen/enlighten.c | 29 +++++++++++++++++++---------- 3 files changed, 22 insertions(+), 13 deletions(-) diff --git a/arch/x86/include/asm/hypervisor.h b/arch/x86/include/asm/hypervisor.h index e42f758a0fbd..055ea9941dd5 100644 --- a/arch/x86/include/asm/hypervisor.h +++ b/arch/x86/include/asm/hypervisor.h @@ -50,7 +50,7 @@ extern const struct hypervisor_x86 *x86_hyper; /* Recognized hypervisors */ extern const struct hypervisor_x86 x86_hyper_vmware; extern const struct hypervisor_x86 x86_hyper_ms_hyperv; -extern const struct hypervisor_x86 x86_hyper_xen_hvm; +extern const struct hypervisor_x86 x86_hyper_xen; extern const struct hypervisor_x86 x86_hyper_kvm; extern void init_hypervisor(struct cpuinfo_x86 *c); diff --git a/arch/x86/kernel/cpu/hypervisor.c b/arch/x86/kernel/cpu/hypervisor.c index 36ce402a3fa5..d820d8eae96b 100644 --- a/arch/x86/kernel/cpu/hypervisor.c +++ b/arch/x86/kernel/cpu/hypervisor.c @@ -27,8 +27,8 @@ static const __initconst struct hypervisor_x86 * const hypervisors[] = { -#ifdef CONFIG_XEN_PVHVM - &x86_hyper_xen_hvm, +#ifdef CONFIG_XEN + &x86_hyper_xen, #endif &x86_hyper_vmware, &x86_hyper_ms_hyperv, diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index 94578efd3067..46957ead3060 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -1760,6 +1760,9 @@ static struct notifier_block xen_hvm_cpu_notifier = { static void __init xen_hvm_guest_init(void) { + if (xen_pv_domain()) + return; + init_hvm_pv_info(); xen_hvm_init_shared_info(); @@ -1775,6 +1778,7 @@ static void __init xen_hvm_guest_init(void) xen_hvm_init_time_ops(); xen_hvm_init_mmu_ops(); } +#endif static bool xen_nopv = false; static __init int xen_parse_nopv(char *arg) @@ -1784,14 +1788,11 @@ static __init int xen_parse_nopv(char *arg) } early_param("xen_nopv", xen_parse_nopv); -static uint32_t __init xen_hvm_platform(void) +static uint32_t __init xen_platform(void) { if (xen_nopv) return 0; - if (xen_pv_domain()) - return 0; - return xen_cpuid_base(); } @@ -1809,11 +1810,19 @@ bool xen_hvm_need_lapic(void) } EXPORT_SYMBOL_GPL(xen_hvm_need_lapic); -const struct hypervisor_x86 x86_hyper_xen_hvm __refconst = { - .name = "Xen HVM", - .detect = xen_hvm_platform, +static void xen_set_cpu_features(struct cpuinfo_x86 *c) +{ + if (xen_pv_domain()) + clear_cpu_bug(c, X86_BUG_SYSRET_SS_ATTRS); +} + +const struct hypervisor_x86 x86_hyper_xen = { + .name = "Xen", + .detect = xen_platform, +#ifdef CONFIG_XEN_PVHVM .init_platform = xen_hvm_guest_init, - .x2apic_available = xen_x2apic_para_available, -}; -EXPORT_SYMBOL(x86_hyper_xen_hvm); #endif + .x2apic_available = xen_x2apic_para_available, + .set_cpu_features = xen_set_cpu_features, +}; +EXPORT_SYMBOL(x86_hyper_xen); From 8746515d7f04c9ea94cf43e2db1fd2cfca93276d Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Fri, 24 Apr 2015 10:16:40 +0100 Subject: [PATCH 10/10] xen: Add __GFP_DMA flag when xen_swiotlb_init gets free pages on ARM Make sure that xen_swiotlb_init allocates buffers that are DMA capable when at least one memblock is available below 4G. Otherwise we assume that all devices on the SoC can cope with >4G addresses. We do this on ARM and ARM64, where dom0 is mapped 1:1, so pfn == mfn in this case. No functional changes on x86. From: Chen Baozi Signed-off-by: Chen Baozi Signed-off-by: Stefano Stabellini Tested-by: Chen Baozi Acked-by: Konrad Rzeszutek Wilk Signed-off-by: David Vrabel --- arch/arm/include/asm/xen/page.h | 1 + arch/arm/xen/mm.c | 15 +++++++++++++++ arch/x86/include/asm/xen/page.h | 5 +++++ drivers/xen/swiotlb-xen.c | 2 +- 4 files changed, 22 insertions(+), 1 deletion(-) diff --git a/arch/arm/include/asm/xen/page.h b/arch/arm/include/asm/xen/page.h index 2f7e6ff67d51..0b579b2f4e0e 100644 --- a/arch/arm/include/asm/xen/page.h +++ b/arch/arm/include/asm/xen/page.h @@ -110,5 +110,6 @@ static inline bool set_phys_to_machine(unsigned long pfn, unsigned long mfn) bool xen_arch_need_swiotlb(struct device *dev, unsigned long pfn, unsigned long mfn); +unsigned long xen_get_swiotlb_free_pages(unsigned int order); #endif /* _ASM_ARM_XEN_PAGE_H */ diff --git a/arch/arm/xen/mm.c b/arch/arm/xen/mm.c index 793551d15f1d..498325074a06 100644 --- a/arch/arm/xen/mm.c +++ b/arch/arm/xen/mm.c @@ -4,6 +4,7 @@ #include #include #include +#include #include #include #include @@ -21,6 +22,20 @@ #include #include +unsigned long xen_get_swiotlb_free_pages(unsigned int order) +{ + struct memblock_region *reg; + gfp_t flags = __GFP_NOWARN; + + for_each_memblock(memory, reg) { + if (reg->base < (phys_addr_t)0xffffffff) { + flags |= __GFP_DMA; + break; + } + } + return __get_free_pages(flags, order); +} + enum dma_cache_op { DMA_UNMAP, DMA_MAP, diff --git a/arch/x86/include/asm/xen/page.h b/arch/x86/include/asm/xen/page.h index 358dcd338915..c44a5d53e464 100644 --- a/arch/x86/include/asm/xen/page.h +++ b/arch/x86/include/asm/xen/page.h @@ -269,4 +269,9 @@ static inline bool xen_arch_need_swiotlb(struct device *dev, return false; } +static inline unsigned long xen_get_swiotlb_free_pages(unsigned int order) +{ + return __get_free_pages(__GFP_NOWARN, order); +} + #endif /* _ASM_X86_XEN_PAGE_H */ diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c index 810ad419e34c..4c549323c605 100644 --- a/drivers/xen/swiotlb-xen.c +++ b/drivers/xen/swiotlb-xen.c @@ -235,7 +235,7 @@ retry: #define SLABS_PER_PAGE (1 << (PAGE_SHIFT - IO_TLB_SHIFT)) #define IO_TLB_MIN_SLABS ((1<<20) >> IO_TLB_SHIFT) while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) { - xen_io_tlb_start = (void *)__get_free_pages(__GFP_NOWARN, order); + xen_io_tlb_start = (void *)xen_get_swiotlb_free_pages(order); if (xen_io_tlb_start) break; order--;