From d407bd25a204bd66b7346dde24bd3d37ef0e0b05 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 18 Jan 2017 15:14:17 +0100 Subject: bpf: don't trigger OOM killer under pressure with map alloc This patch adds two helpers, bpf_map_area_alloc() and bpf_map_area_free(), that are to be used for map allocations. Using kmalloc() for very large allocations can cause excessive work within the page allocator, so i) fall back earlier to vmalloc() when the attempt is considered costly anyway, and even more importantly ii) don't trigger OOM killer with any of the allocators. Since this is based on a user space request, for example, when creating maps with element pre-allocation, we really want such requests to fail instead of killing other user space processes. Also, don't spam the kernel log with warnings should any of the allocations fail under pressure. Given that, we can make backend selection in bpf_map_area_alloc() generic, and convert all maps over to use this API for spots with potentially large allocation requests. Note, replacing the one kmalloc_array() is fine as overflow checks happen earlier in htab_map_alloc(), since it must also protect the multiplication for vmalloc() should kmalloc_array() fail. Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- kernel/bpf/arraymap.c | 18 +++++++----------- kernel/bpf/hashtab.c | 22 +++++++++------------- kernel/bpf/stackmap.c | 20 ++++++++------------ kernel/bpf/syscall.c | 26 ++++++++++++++++++++++++++ 4 files changed, 50 insertions(+), 36 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c index 229a5d5df977..3d55d95dcf49 100644 --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -11,7 +11,6 @@ */ #include #include -#include #include #include #include @@ -74,14 +73,10 @@ static struct bpf_map *array_map_alloc(union bpf_attr *attr) if (array_size >= U32_MAX - PAGE_SIZE) return ERR_PTR(-ENOMEM); - /* allocate all map elements and zero-initialize them */ - array = kzalloc(array_size, GFP_USER | __GFP_NOWARN); - if (!array) { - array = vzalloc(array_size); - if (!array) - return ERR_PTR(-ENOMEM); - } + array = bpf_map_area_alloc(array_size); + if (!array) + return ERR_PTR(-ENOMEM); /* copy mandatory map attributes */ array->map.map_type = attr->map_type; @@ -97,7 +92,7 @@ static struct bpf_map *array_map_alloc(union bpf_attr *attr) if (array_size >= U32_MAX - PAGE_SIZE || elem_size > PCPU_MIN_UNIT_SIZE || bpf_array_alloc_percpu(array)) { - kvfree(array); + bpf_map_area_free(array); return ERR_PTR(-ENOMEM); } out: @@ -262,7 +257,7 @@ static void array_map_free(struct bpf_map *map) if (array->map.map_type == BPF_MAP_TYPE_PERCPU_ARRAY) bpf_array_free_percpu(array); - kvfree(array); + bpf_map_area_free(array); } static const struct bpf_map_ops array_ops = { @@ -319,7 +314,8 @@ static void fd_array_map_free(struct bpf_map *map) /* make sure it's empty */ for (i = 0; i < array->map.max_entries; i++) BUG_ON(array->ptrs[i] != NULL); - kvfree(array); + + bpf_map_area_free(array); } static void *fd_array_map_lookup_elem(struct bpf_map *map, void *key) diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 3f2bb58952d8..a753bbe7df0a 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -13,7 +13,6 @@ #include #include #include -#include #include "percpu_freelist.h" #include "bpf_lru_list.h" @@ -103,7 +102,7 @@ static void htab_free_elems(struct bpf_htab *htab) free_percpu(pptr); } free_elems: - vfree(htab->elems); + bpf_map_area_free(htab->elems); } static struct htab_elem *prealloc_lru_pop(struct bpf_htab *htab, void *key, @@ -125,7 +124,8 @@ static int prealloc_init(struct bpf_htab *htab) { int err = -ENOMEM, i; - htab->elems = vzalloc(htab->elem_size * htab->map.max_entries); + htab->elems = bpf_map_area_alloc(htab->elem_size * + htab->map.max_entries); if (!htab->elems) return -ENOMEM; @@ -320,14 +320,10 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) goto free_htab; err = -ENOMEM; - htab->buckets = kmalloc_array(htab->n_buckets, sizeof(struct bucket), - GFP_USER | __GFP_NOWARN); - - if (!htab->buckets) { - htab->buckets = vmalloc(htab->n_buckets * sizeof(struct bucket)); - if (!htab->buckets) - goto free_htab; - } + htab->buckets = bpf_map_area_alloc(htab->n_buckets * + sizeof(struct bucket)); + if (!htab->buckets) + goto free_htab; for (i = 0; i < htab->n_buckets; i++) { INIT_HLIST_HEAD(&htab->buckets[i].head); @@ -354,7 +350,7 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) free_extra_elems: free_percpu(htab->extra_elems); free_buckets: - kvfree(htab->buckets); + bpf_map_area_free(htab->buckets); free_htab: kfree(htab); return ERR_PTR(err); @@ -1014,7 +1010,7 @@ static void htab_map_free(struct bpf_map *map) prealloc_destroy(htab); free_percpu(htab->extra_elems); - kvfree(htab->buckets); + bpf_map_area_free(htab->buckets); kfree(htab); } diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c index 732ae16d12b7..be8519148c25 100644 --- a/kernel/bpf/stackmap.c +++ b/kernel/bpf/stackmap.c @@ -7,7 +7,6 @@ #include #include #include -#include #include #include #include "percpu_freelist.h" @@ -32,7 +31,7 @@ static int prealloc_elems_and_freelist(struct bpf_stack_map *smap) u32 elem_size = sizeof(struct stack_map_bucket) + smap->map.value_size; int err; - smap->elems = vzalloc(elem_size * smap->map.max_entries); + smap->elems = bpf_map_area_alloc(elem_size * smap->map.max_entries); if (!smap->elems) return -ENOMEM; @@ -45,7 +44,7 @@ static int prealloc_elems_and_freelist(struct bpf_stack_map *smap) return 0; free_elems: - vfree(smap->elems); + bpf_map_area_free(smap->elems); return err; } @@ -76,12 +75,9 @@ static struct bpf_map *stack_map_alloc(union bpf_attr *attr) if (cost >= U32_MAX - PAGE_SIZE) return ERR_PTR(-E2BIG); - smap = kzalloc(cost, GFP_USER | __GFP_NOWARN); - if (!smap) { - smap = vzalloc(cost); - if (!smap) - return ERR_PTR(-ENOMEM); - } + smap = bpf_map_area_alloc(cost); + if (!smap) + return ERR_PTR(-ENOMEM); err = -E2BIG; cost += n_buckets * (value_size + sizeof(struct stack_map_bucket)); @@ -112,7 +108,7 @@ static struct bpf_map *stack_map_alloc(union bpf_attr *attr) put_buffers: put_callchain_buffers(); free_smap: - kvfree(smap); + bpf_map_area_free(smap); return ERR_PTR(err); } @@ -262,9 +258,9 @@ static void stack_map_free(struct bpf_map *map) /* wait for bpf programs to complete before freeing stack map */ synchronize_rcu(); - vfree(smap->elems); + bpf_map_area_free(smap->elems); pcpu_freelist_destroy(&smap->freelist); - kvfree(smap); + bpf_map_area_free(smap); put_callchain_buffers(); } diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 1d6b29e4e2c3..19b6129eab23 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -12,6 +12,8 @@ #include #include #include +#include +#include #include #include #include @@ -49,6 +51,30 @@ void bpf_register_map_type(struct bpf_map_type_list *tl) list_add(&tl->list_node, &bpf_map_types); } +void *bpf_map_area_alloc(size_t size) +{ + /* We definitely need __GFP_NORETRY, so OOM killer doesn't + * trigger under memory pressure as we really just want to + * fail instead. + */ + const gfp_t flags = __GFP_NOWARN | __GFP_NORETRY | __GFP_ZERO; + void *area; + + if (size <= (PAGE_SIZE << PAGE_ALLOC_COSTLY_ORDER)) { + area = kmalloc(size, GFP_USER | flags); + if (area != NULL) + return area; + } + + return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM | flags, + PAGE_KERNEL); +} + +void bpf_map_area_free(void *area) +{ + kvfree(area); +} + int bpf_map_precharge_memlock(u32 pages) { struct user_struct *user = get_current_user(); -- cgit v1.2.3 From e326ce013a8e851193eb337aafb1aa396c533a61 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Fri, 20 Jan 2017 03:25:34 +0100 Subject: Revert "PM / sleep / ACPI: Use the ACPI_FADT_LOW_POWER_S0 flag" Revert commit 08b98d329165 (PM / sleep / ACPI: Use the ACPI_FADT_LOW_POWER_S0 flag) as it caused system suspend (in the default configuration) to fail on Dell XPS13 (9360) with the Kaby Lake processor. Fixes: 08b98d329165 (PM / sleep / ACPI: Use the ACPI_FADT_LOW_POWER_S0 flag) Reported-by: Paul Menzel Signed-off-by: Rafael J. Wysocki --- Documentation/power/states.txt | 4 +--- drivers/acpi/sleep.c | 8 -------- include/linux/suspend.h | 2 -- kernel/power/suspend.c | 4 ++-- 4 files changed, 3 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/Documentation/power/states.txt b/Documentation/power/states.txt index 8a39ce45d8a0..008ecb588317 100644 --- a/Documentation/power/states.txt +++ b/Documentation/power/states.txt @@ -35,9 +35,7 @@ only one way to cause the system to go into the Suspend-To-RAM state (write The default suspend mode (ie. the one to be used without writing anything into /sys/power/mem_sleep) is either "deep" (if Suspend-To-RAM is supported) or "s2idle", but it can be overridden by the value of the "mem_sleep_default" -parameter in the kernel command line. On some ACPI-based systems, depending on -the information in the FADT, the default may be "s2idle" even if Suspend-To-RAM -is supported. +parameter in the kernel command line. The properties of all of the sleep states are described below. diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c index ce1855fd584b..deb0ff78eba8 100644 --- a/drivers/acpi/sleep.c +++ b/drivers/acpi/sleep.c @@ -691,14 +691,6 @@ static void acpi_sleep_suspend_setup(void) if (acpi_sleep_state_supported(i)) sleep_states[i] = 1; - /* - * Use suspend-to-idle by default if ACPI_FADT_LOW_POWER_S0 is set and - * the default suspend mode was not selected from the command line. - */ - if (acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0 && - mem_sleep_default > PM_SUSPEND_MEM) - mem_sleep_default = PM_SUSPEND_FREEZE; - suspend_set_ops(old_suspend_ordering ? &acpi_suspend_ops_old : &acpi_suspend_ops); freeze_set_ops(&acpi_freeze_ops); diff --git a/include/linux/suspend.h b/include/linux/suspend.h index 0c729c3c8549..d9718378a8be 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -194,8 +194,6 @@ struct platform_freeze_ops { }; #ifdef CONFIG_SUSPEND -extern suspend_state_t mem_sleep_default; - /** * suspend_set_ops - set platform dependent suspend operations * @ops: The new suspend operations to set. diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c index f67ceb7768b8..15e6baef5c73 100644 --- a/kernel/power/suspend.c +++ b/kernel/power/suspend.c @@ -46,7 +46,7 @@ static const char * const mem_sleep_labels[] = { const char *mem_sleep_states[PM_SUSPEND_MAX]; suspend_state_t mem_sleep_current = PM_SUSPEND_FREEZE; -suspend_state_t mem_sleep_default = PM_SUSPEND_MAX; +static suspend_state_t mem_sleep_default = PM_SUSPEND_MEM; unsigned int pm_suspend_global_flags; EXPORT_SYMBOL_GPL(pm_suspend_global_flags); @@ -168,7 +168,7 @@ void suspend_set_ops(const struct platform_suspend_ops *ops) } if (valid_state(PM_SUSPEND_MEM)) { mem_sleep_states[PM_SUSPEND_MEM] = mem_sleep_labels[PM_SUSPEND_MEM]; - if (mem_sleep_default >= PM_SUSPEND_MEM) + if (mem_sleep_default == PM_SUSPEND_MEM) mem_sleep_current = PM_SUSPEND_MEM; } -- cgit v1.2.3 From ff9f8a7cf935468a94d9927c68b00daae701667e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 25 Jan 2017 18:20:55 -0800 Subject: sysctl: fix proc_doulongvec_ms_jiffies_minmax() We perform the conversion between kernel jiffies and ms only when exporting kernel value to user space. We need to do the opposite operation when value is written by user. Only matters when HZ != 1000 Signed-off-by: Eric Dumazet Cc: stable@vger.kernel.org Signed-off-by: Linus Torvalds --- kernel/sysctl.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 8dbaec0e4f7f..1aea594a54db 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -2475,6 +2475,7 @@ static int __do_proc_doulongvec_minmax(void *data, struct ctl_table *table, int break; if (neg) continue; + val = convmul * val / convdiv; if ((min && val < *min) || (max && val > *max)) continue; *i = val; -- cgit v1.2.3 From 07cd12945551b63ecb1a349d50a6d69d1d6feb4a Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 26 Jan 2017 16:47:28 -0500 Subject: cgroup: don't online subsystems before cgroup_name/path() are operational While refactoring cgroup creation, a5bca2152036 ("cgroup: factor out cgroup_create() out of cgroup_mkdir()") incorrectly onlined subsystems before the new cgroup is associated with it kernfs_node. This is fine for cgroup proper but cgroup_name/path() depend on the associated kernfs_node and if a subsystem makes the new cgroup_subsys_state visible, which they're allowed to after onlining, it can lead to NULL dereference. The current code performs cgroup creation and subsystem onlining in cgroup_create() and cgroup_mkdir() makes the cgroup and subsystems visible afterwards. There's no reason to online the subsystems early and we can simply drop cgroup_apply_control_enable() call from cgroup_create() so that the subsystems are onlined and made visible at the same time. Signed-off-by: Tejun Heo Reported-by: Konstantin Khlebnikov Fixes: a5bca2152036 ("cgroup: factor out cgroup_create() out of cgroup_mkdir()") Cc: stable@vger.kernel.org # v4.6+ --- kernel/cgroup.c | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 2ee9ec3051b2..688dd02af985 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -5221,6 +5221,11 @@ err_free_css: return ERR_PTR(err); } +/* + * The returned cgroup is fully initialized including its control mask, but + * it isn't associated with its kernfs_node and doesn't have the control + * mask applied. + */ static struct cgroup *cgroup_create(struct cgroup *parent) { struct cgroup_root *root = parent->root; @@ -5288,11 +5293,6 @@ static struct cgroup *cgroup_create(struct cgroup *parent) cgroup_propagate_control(cgrp); - /* @cgrp doesn't have dir yet so the following will only create csses */ - ret = cgroup_apply_control_enable(cgrp); - if (ret) - goto out_destroy; - return cgrp; out_cancel_ref: @@ -5300,9 +5300,6 @@ out_cancel_ref: out_free_cgrp: kfree(cgrp); return ERR_PTR(ret); -out_destroy: - cgroup_destroy_locked(cgrp); - return ERR_PTR(ret); } static int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, -- cgit v1.2.3 From a76a82a3e38c8d3fb6499e3dfaeb0949241ab588 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 26 Jan 2017 16:39:55 +0100 Subject: perf/core: Fix use-after-free bug Dmitry reported a KASAN use-after-free on event->group_leader. It turns out there's a hole in perf_remove_from_context() due to event_function_call() not calling its function when the task associated with the event is already dead. In this case the event will have been detached from the task, but the grouping will have been retained, such that group operations might still work properly while there are live child events etc. This does however mean that we can miss a perf_group_detach() call when the group decomposes, this in turn can then lead to use-after-free. Fix it by explicitly doing the group detach if its still required. Reported-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Mathieu Desnoyers Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: stable@vger.kernel.org # v4.5+ Cc: syzkaller Fixes: 63b6da39bb38 ("perf: Fix perf_event_exit_task() race") Link: http://lkml.kernel.org/r/20170126153955.GD6515@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar --- kernel/events/core.c | 27 +++++++++++++++++++++++++-- 1 file changed, 25 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 110b38a58493..4e1f4c0070ce 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -1469,7 +1469,6 @@ ctx_group_list(struct perf_event *event, struct perf_event_context *ctx) static void list_add_event(struct perf_event *event, struct perf_event_context *ctx) { - lockdep_assert_held(&ctx->lock); WARN_ON_ONCE(event->attach_state & PERF_ATTACH_CONTEXT); @@ -1624,6 +1623,8 @@ static void perf_group_attach(struct perf_event *event) { struct perf_event *group_leader = event->group_leader, *pos; + lockdep_assert_held(&event->ctx->lock); + /* * We can have double attach due to group movement in perf_event_open. */ @@ -1697,6 +1698,8 @@ static void perf_group_detach(struct perf_event *event) struct perf_event *sibling, *tmp; struct list_head *list = NULL; + lockdep_assert_held(&event->ctx->lock); + /* * We can have double detach due to exit/hot-unplug + close. */ @@ -1895,9 +1898,29 @@ __perf_remove_from_context(struct perf_event *event, */ static void perf_remove_from_context(struct perf_event *event, unsigned long flags) { - lockdep_assert_held(&event->ctx->mutex); + struct perf_event_context *ctx = event->ctx; + + lockdep_assert_held(&ctx->mutex); event_function_call(event, __perf_remove_from_context, (void *)flags); + + /* + * The above event_function_call() can NO-OP when it hits + * TASK_TOMBSTONE. In that case we must already have been detached + * from the context (by perf_event_exit_event()) but the grouping + * might still be in-tact. + */ + WARN_ON_ONCE(event->attach_state & PERF_ATTACH_CONTEXT); + if ((flags & DETACH_GROUP) && + (event->attach_state & PERF_ATTACH_GROUP)) { + /* + * Since in that case we cannot possibly be scheduled, simply + * detach now. + */ + raw_spin_lock_irq(&ctx->lock); + perf_group_detach(event); + raw_spin_unlock_irq(&ctx->lock); + } } /* -- cgit v1.2.3 From 0b3589be9b98994ce3d5aeca52445d1f5627c4ba Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 26 Jan 2017 23:15:08 +0100 Subject: perf/core: Fix PERF_RECORD_MMAP2 prot/flags for anonymous memory Andres reported that MMAP2 records for anonymous memory always have their protection field 0. Turns out, someone daft put the prot/flags generation code in the file branch, leaving them unset for anonymous memory. Reported-by: Andres Freund Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Don Zickus Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Stephane Eranian Cc: Thomas Gleixner Cc: acme@kernel.org Cc: anton@ozlabs.org Cc: namhyung@kernel.org Cc: stable@vger.kernel.org # v3.16+ Fixes: f972eb63b100 ("perf: Pass protection and flags bits through mmap2 interface") Link: http://lkml.kernel.org/r/20170126221508.GF6536@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar --- kernel/events/core.c | 42 +++++++++++++++++++++--------------------- 1 file changed, 21 insertions(+), 21 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 4e1f4c0070ce..e5aaa806702d 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -6632,6 +6632,27 @@ static void perf_event_mmap_event(struct perf_mmap_event *mmap_event) char *buf = NULL; char *name; + if (vma->vm_flags & VM_READ) + prot |= PROT_READ; + if (vma->vm_flags & VM_WRITE) + prot |= PROT_WRITE; + if (vma->vm_flags & VM_EXEC) + prot |= PROT_EXEC; + + if (vma->vm_flags & VM_MAYSHARE) + flags = MAP_SHARED; + else + flags = MAP_PRIVATE; + + if (vma->vm_flags & VM_DENYWRITE) + flags |= MAP_DENYWRITE; + if (vma->vm_flags & VM_MAYEXEC) + flags |= MAP_EXECUTABLE; + if (vma->vm_flags & VM_LOCKED) + flags |= MAP_LOCKED; + if (vma->vm_flags & VM_HUGETLB) + flags |= MAP_HUGETLB; + if (file) { struct inode *inode; dev_t dev; @@ -6658,27 +6679,6 @@ static void perf_event_mmap_event(struct perf_mmap_event *mmap_event) maj = MAJOR(dev); min = MINOR(dev); - if (vma->vm_flags & VM_READ) - prot |= PROT_READ; - if (vma->vm_flags & VM_WRITE) - prot |= PROT_WRITE; - if (vma->vm_flags & VM_EXEC) - prot |= PROT_EXEC; - - if (vma->vm_flags & VM_MAYSHARE) - flags = MAP_SHARED; - else - flags = MAP_PRIVATE; - - if (vma->vm_flags & VM_DENYWRITE) - flags |= MAP_DENYWRITE; - if (vma->vm_flags & VM_MAYEXEC) - flags |= MAP_EXECUTABLE; - if (vma->vm_flags & VM_LOCKED) - flags |= MAP_LOCKED; - if (vma->vm_flags & VM_HUGETLB) - flags |= MAP_HUGETLB; - goto got_name; } else { if (vma->vm_ops && vma->vm_ops->name) { -- cgit v1.2.3 From 08d85f3ea99f1eeafc4e8507936190e86a16ee8c Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Tue, 17 Jan 2017 16:00:48 +0000 Subject: irqdomain: Avoid activating interrupts more than once Since commit f3b0946d629c ("genirq/msi: Make sure PCI MSIs are activated early"), we can end-up activating a PCI/MSI twice (once at allocation time, and once at startup time). This is normally of no consequences, except that there is some HW out there that may misbehave if activate is used more than once (the GICv3 ITS, for example, uses the activate callback to issue the MAPVI command, and the architecture spec says that "If there is an existing mapping for the EventID-DeviceID combination, behavior is UNPREDICTABLE"). While this could be worked around in each individual driver, it may make more sense to tackle the issue at the core level. In order to avoid getting in that situation, let's have a per-interrupt flag to remember if we have already activated that interrupt or not. Fixes: f3b0946d629c ("genirq/msi: Make sure PCI MSIs are activated early") Reported-and-tested-by: Andre Przywara Signed-off-by: Marc Zyngier Cc: stable@vger.kernel.org Link: http://lkml.kernel.org/r/1484668848-24361-1-git-send-email-marc.zyngier@arm.com Signed-off-by: Thomas Gleixner --- include/linux/irq.h | 17 +++++++++++++++++ kernel/irq/irqdomain.c | 44 ++++++++++++++++++++++++++++++-------------- 2 files changed, 47 insertions(+), 14 deletions(-) (limited to 'kernel') diff --git a/include/linux/irq.h b/include/linux/irq.h index e79875574b39..39e3254e5769 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -184,6 +184,7 @@ struct irq_data { * * IRQD_TRIGGER_MASK - Mask for the trigger type bits * IRQD_SETAFFINITY_PENDING - Affinity setting is pending + * IRQD_ACTIVATED - Interrupt has already been activated * IRQD_NO_BALANCING - Balancing disabled for this IRQ * IRQD_PER_CPU - Interrupt is per cpu * IRQD_AFFINITY_SET - Interrupt affinity was set @@ -202,6 +203,7 @@ struct irq_data { enum { IRQD_TRIGGER_MASK = 0xf, IRQD_SETAFFINITY_PENDING = (1 << 8), + IRQD_ACTIVATED = (1 << 9), IRQD_NO_BALANCING = (1 << 10), IRQD_PER_CPU = (1 << 11), IRQD_AFFINITY_SET = (1 << 12), @@ -312,6 +314,21 @@ static inline bool irqd_affinity_is_managed(struct irq_data *d) return __irqd_to_state(d) & IRQD_AFFINITY_MANAGED; } +static inline bool irqd_is_activated(struct irq_data *d) +{ + return __irqd_to_state(d) & IRQD_ACTIVATED; +} + +static inline void irqd_set_activated(struct irq_data *d) +{ + __irqd_to_state(d) |= IRQD_ACTIVATED; +} + +static inline void irqd_clr_activated(struct irq_data *d) +{ + __irqd_to_state(d) &= ~IRQD_ACTIVATED; +} + #undef __irqd_to_state static inline irq_hw_number_t irqd_to_hwirq(struct irq_data *d) diff --git a/kernel/irq/irqdomain.c b/kernel/irq/irqdomain.c index 8c0a0ae43521..b59e6768c5e9 100644 --- a/kernel/irq/irqdomain.c +++ b/kernel/irq/irqdomain.c @@ -1346,6 +1346,30 @@ void irq_domain_free_irqs_parent(struct irq_domain *domain, } EXPORT_SYMBOL_GPL(irq_domain_free_irqs_parent); +static void __irq_domain_activate_irq(struct irq_data *irq_data) +{ + if (irq_data && irq_data->domain) { + struct irq_domain *domain = irq_data->domain; + + if (irq_data->parent_data) + __irq_domain_activate_irq(irq_data->parent_data); + if (domain->ops->activate) + domain->ops->activate(domain, irq_data); + } +} + +static void __irq_domain_deactivate_irq(struct irq_data *irq_data) +{ + if (irq_data && irq_data->domain) { + struct irq_domain *domain = irq_data->domain; + + if (domain->ops->deactivate) + domain->ops->deactivate(domain, irq_data); + if (irq_data->parent_data) + __irq_domain_deactivate_irq(irq_data->parent_data); + } +} + /** * irq_domain_activate_irq - Call domain_ops->activate recursively to activate * interrupt @@ -1356,13 +1380,9 @@ EXPORT_SYMBOL_GPL(irq_domain_free_irqs_parent); */ void irq_domain_activate_irq(struct irq_data *irq_data) { - if (irq_data && irq_data->domain) { - struct irq_domain *domain = irq_data->domain; - - if (irq_data->parent_data) - irq_domain_activate_irq(irq_data->parent_data); - if (domain->ops->activate) - domain->ops->activate(domain, irq_data); + if (!irqd_is_activated(irq_data)) { + __irq_domain_activate_irq(irq_data); + irqd_set_activated(irq_data); } } @@ -1376,13 +1396,9 @@ void irq_domain_activate_irq(struct irq_data *irq_data) */ void irq_domain_deactivate_irq(struct irq_data *irq_data) { - if (irq_data && irq_data->domain) { - struct irq_domain *domain = irq_data->domain; - - if (domain->ops->deactivate) - domain->ops->deactivate(domain, irq_data); - if (irq_data->parent_data) - irq_domain_deactivate_irq(irq_data->parent_data); + if (irqd_is_activated(irq_data)) { + __irq_domain_deactivate_irq(irq_data); + irqd_clr_activated(irq_data); } } -- cgit v1.2.3 From 79c6f448c8b79c321e4a1f31f98194e4f6b6cae7 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Mon, 30 Jan 2017 19:27:10 -0500 Subject: tracing: Fix hwlat kthread migration The hwlat tracer creates a kernel thread at start of the tracer. It is pinned to a single CPU and will move to the next CPU after each period of running. If the user modifies the migration thread's affinity, it will not change after that happens. The original code created the thread at the first instance it was called, but later was changed to destroy the thread after the tracer was finished, and would not be created until the next instance of the tracer was established. The code that initialized the affinity was only called on the initial instantiation of the tracer. After that, it was not initialized, and the previous affinity did not match the current newly created one, making it appear that the user modified the thread's affinity when it did not, and the thread failed to migrate again. Cc: stable@vger.kernel.org Fixes: 0330f7aa8ee6 ("tracing: Have hwlat trace migrate across tracing_cpumask CPUs") Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_hwlat.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_hwlat.c b/kernel/trace/trace_hwlat.c index 775569ec50d0..af344a1bf0d0 100644 --- a/kernel/trace/trace_hwlat.c +++ b/kernel/trace/trace_hwlat.c @@ -266,7 +266,7 @@ out: static struct cpumask save_cpumask; static bool disable_migrate; -static void move_to_next_cpu(void) +static void move_to_next_cpu(bool initmask) { static struct cpumask *current_mask; int next_cpu; @@ -275,7 +275,7 @@ static void move_to_next_cpu(void) return; /* Just pick the first CPU on first iteration */ - if (!current_mask) { + if (initmask) { current_mask = &save_cpumask; get_online_cpus(); cpumask_and(current_mask, cpu_online_mask, tracing_buffer_mask); @@ -330,10 +330,12 @@ static void move_to_next_cpu(void) static int kthread_fn(void *data) { u64 interval; + bool initmask = true; while (!kthread_should_stop()) { - move_to_next_cpu(); + move_to_next_cpu(initmask); + initmask = false; local_irq_disable(); get_sample(); -- cgit v1.2.3 From 26a346f23c5291d1d9521e72763103daf2c6f0d1 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Wed, 1 Feb 2017 17:57:56 +0100 Subject: tracing/kprobes: Fix __init annotation clang complains about "__init" being attached to a struct name: kernel/trace/trace_kprobe.c:1375:15: error: '__section__' attribute only applies to functions and global variables The intention must have been to mark the function as __init instead of the type, so move the attribute there. Link: http://lkml.kernel.org/r/20170201165826.2625888-1-arnd@arndb.de Fixes: f18f97ac43d7 ("tracing/kprobes: Add a helper method to return number of probe hits") Signed-off-by: Arnd Bergmann Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_kprobe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index a133ecd741e4..7ad9e53ad174 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1372,7 +1372,7 @@ kprobe_trace_selftest_target(int a1, int a2, int a3, int a4, int a5, int a6) return a1 + a2 + a3 + a4 + a5 + a6; } -static struct __init trace_event_file * +static __init struct trace_event_file * find_trace_probe_file(struct trace_kprobe *tk, struct trace_array *tr) { struct trace_event_file *file; -- cgit v1.2.3 From 71810db27c1c853b335675bee335d893bc3d324b Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Fri, 3 Feb 2017 09:54:06 +0000 Subject: modversions: treat symbol CRCs as 32 bit quantities The modversion symbol CRCs are emitted as ELF symbols, which allows us to easily populate the kcrctab sections by relying on the linker to associate each kcrctab slot with the correct value. This has a couple of downsides: - Given that the CRCs are treated as memory addresses, we waste 4 bytes for each CRC on 64 bit architectures, - On architectures that support runtime relocation, a R__RELATIVE relocation entry is emitted for each CRC value, which identifies it as a quantity that requires fixing up based on the actual runtime load offset of the kernel. This results in corrupted CRCs unless we explicitly undo the fixup (and this is currently being handled in the core module code) - Such runtime relocation entries take up 24 bytes of __init space each, resulting in a x8 overhead in [uncompressed] kernel size for CRCs. Switching to explicit 32 bit values on 64 bit architectures fixes most of these issues, given that 32 bit values are not treated as quantities that require fixing up based on the actual runtime load offset. Note that on some ELF64 architectures [such as PPC64], these 32-bit values are still emitted as [absolute] runtime relocatable quantities, even if the value resolves to a build time constant. Since relative relocations are always resolved at build time, this patch enables MODULE_REL_CRCS on powerpc when CONFIG_RELOCATABLE=y, which turns the absolute CRC references into relative references into .rodata where the actual CRC value is stored. So redefine all CRC fields and variables as u32, and redefine the __CRC_SYMBOL() macro for 64 bit builds to emit the CRC reference using inline assembler (which is necessary since 64-bit C code cannot use 32-bit types to hold memory addresses, even if they are ultimately resolved using values that do not exceed 0xffffffff). To avoid potential problems with legacy 32-bit architectures using legacy toolchains, the equivalent C definition of the kcrctab entry is retained for 32-bit architectures. Note that this mostly reverts commit d4703aefdbc8 ("module: handle ppc64 relocating kcrctabs when CONFIG_RELOCATABLE=y") Acked-by: Rusty Russell Signed-off-by: Ard Biesheuvel Signed-off-by: Linus Torvalds --- arch/powerpc/Kconfig | 1 + arch/powerpc/include/asm/module.h | 4 --- arch/powerpc/kernel/module_64.c | 8 ------ include/asm-generic/export.h | 11 ++++---- include/linux/export.h | 14 +++++++++++ include/linux/module.h | 14 +++++------ kernel/module.c | 53 ++++++++++++++++++--------------------- 7 files changed, 53 insertions(+), 52 deletions(-) (limited to 'kernel') diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index a8ee573fe610..db8a1ef6bfaf 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -484,6 +484,7 @@ config RELOCATABLE bool "Build a relocatable kernel" depends on (PPC64 && !COMPILE_TEST) || (FLATMEM && (44x || FSL_BOOKE)) select NONSTATIC_KERNEL + select MODULE_REL_CRCS if MODVERSIONS help This builds a kernel image that is capable of running at the location the kernel is loaded at. For ppc32, there is no any diff --git a/arch/powerpc/include/asm/module.h b/arch/powerpc/include/asm/module.h index cc12c61ef315..53885512b8d3 100644 --- a/arch/powerpc/include/asm/module.h +++ b/arch/powerpc/include/asm/module.h @@ -90,9 +90,5 @@ static inline int module_finalize_ftrace(struct module *mod, const Elf_Shdr *sec } #endif -#if defined(CONFIG_MODVERSIONS) && defined(CONFIG_PPC64) -#define ARCH_RELOCATES_KCRCTAB -#define reloc_start PHYSICAL_START -#endif #endif /* __KERNEL__ */ #endif /* _ASM_POWERPC_MODULE_H */ diff --git a/arch/powerpc/kernel/module_64.c b/arch/powerpc/kernel/module_64.c index bb1807184bad..0b0f89685b67 100644 --- a/arch/powerpc/kernel/module_64.c +++ b/arch/powerpc/kernel/module_64.c @@ -286,14 +286,6 @@ static void dedotify_versions(struct modversion_info *vers, for (end = (void *)vers + size; vers < end; vers++) if (vers->name[0] == '.') { memmove(vers->name, vers->name+1, strlen(vers->name)); -#ifdef ARCH_RELOCATES_KCRCTAB - /* The TOC symbol has no CRC computed. To avoid CRC - * check failing, we must force it to the expected - * value (see CRC check in module.c). - */ - if (!strcmp(vers->name, "TOC.")) - vers->crc = -(unsigned long)reloc_start; -#endif } } diff --git a/include/asm-generic/export.h b/include/asm-generic/export.h index 63554e9f6e0c..719db1968d81 100644 --- a/include/asm-generic/export.h +++ b/include/asm-generic/export.h @@ -9,18 +9,15 @@ #ifndef KSYM_ALIGN #define KSYM_ALIGN 8 #endif -#ifndef KCRC_ALIGN -#define KCRC_ALIGN 8 -#endif #else #define __put .long #ifndef KSYM_ALIGN #define KSYM_ALIGN 4 #endif +#endif #ifndef KCRC_ALIGN #define KCRC_ALIGN 4 #endif -#endif #ifdef CONFIG_HAVE_UNDERSCORE_SYMBOL_PREFIX #define KSYM(name) _##name @@ -52,7 +49,11 @@ KSYM(__kstrtab_\name): .section ___kcrctab\sec+\name,"a" .balign KCRC_ALIGN KSYM(__kcrctab_\name): - __put KSYM(__crc_\name) +#if defined(CONFIG_MODULE_REL_CRCS) + .long KSYM(__crc_\name) - . +#else + .long KSYM(__crc_\name) +#endif .weak KSYM(__crc_\name) .previous #endif diff --git a/include/linux/export.h b/include/linux/export.h index 2a0f61fbc731..7473fba6a60c 100644 --- a/include/linux/export.h +++ b/include/linux/export.h @@ -43,6 +43,13 @@ extern struct module __this_module; #ifdef CONFIG_MODVERSIONS /* Mark the CRC weak since genksyms apparently decides not to * generate a checksums for some symbols */ +#if defined(CONFIG_MODULE_REL_CRCS) +#define __CRC_SYMBOL(sym, sec) \ + asm(" .section \"___kcrctab" sec "+" #sym "\", \"a\" \n" \ + " .weak " VMLINUX_SYMBOL_STR(__crc_##sym) " \n" \ + " .long " VMLINUX_SYMBOL_STR(__crc_##sym) " - . \n" \ + " .previous \n"); +#elif !defined(CONFIG_64BIT) #define __CRC_SYMBOL(sym, sec) \ extern __visible void *__crc_##sym __attribute__((weak)); \ static const unsigned long __kcrctab_##sym \ @@ -50,6 +57,13 @@ extern struct module __this_module; __attribute__((section("___kcrctab" sec "+" #sym), used)) \ = (unsigned long) &__crc_##sym; #else +#define __CRC_SYMBOL(sym, sec) \ + asm(" .section \"___kcrctab" sec "+" #sym "\", \"a\" \n" \ + " .weak " VMLINUX_SYMBOL_STR(__crc_##sym) " \n" \ + " .long " VMLINUX_SYMBOL_STR(__crc_##sym) " \n" \ + " .previous \n"); +#endif +#else #define __CRC_SYMBOL(sym, sec) #endif diff --git a/include/linux/module.h b/include/linux/module.h index 7c84273d60b9..cc7cba219b20 100644 --- a/include/linux/module.h +++ b/include/linux/module.h @@ -346,7 +346,7 @@ struct module { /* Exported symbols */ const struct kernel_symbol *syms; - const unsigned long *crcs; + const s32 *crcs; unsigned int num_syms; /* Kernel parameters. */ @@ -359,18 +359,18 @@ struct module { /* GPL-only exported symbols. */ unsigned int num_gpl_syms; const struct kernel_symbol *gpl_syms; - const unsigned long *gpl_crcs; + const s32 *gpl_crcs; #ifdef CONFIG_UNUSED_SYMBOLS /* unused exported symbols. */ const struct kernel_symbol *unused_syms; - const unsigned long *unused_crcs; + const s32 *unused_crcs; unsigned int num_unused_syms; /* GPL-only, unused exported symbols. */ unsigned int num_unused_gpl_syms; const struct kernel_symbol *unused_gpl_syms; - const unsigned long *unused_gpl_crcs; + const s32 *unused_gpl_crcs; #endif #ifdef CONFIG_MODULE_SIG @@ -382,7 +382,7 @@ struct module { /* symbols that will be GPL-only in the near future. */ const struct kernel_symbol *gpl_future_syms; - const unsigned long *gpl_future_crcs; + const s32 *gpl_future_crcs; unsigned int num_gpl_future_syms; /* Exception table */ @@ -523,7 +523,7 @@ struct module *find_module(const char *name); struct symsearch { const struct kernel_symbol *start, *stop; - const unsigned long *crcs; + const s32 *crcs; enum { NOT_GPL_ONLY, GPL_ONLY, @@ -539,7 +539,7 @@ struct symsearch { */ const struct kernel_symbol *find_symbol(const char *name, struct module **owner, - const unsigned long **crc, + const s32 **crc, bool gplok, bool warn); diff --git a/kernel/module.c b/kernel/module.c index 38d4270925d4..3d8f126208e3 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -389,16 +389,16 @@ extern const struct kernel_symbol __start___ksymtab_gpl[]; extern const struct kernel_symbol __stop___ksymtab_gpl[]; extern const struct kernel_symbol __start___ksymtab_gpl_future[]; extern const struct kernel_symbol __stop___ksymtab_gpl_future[]; -extern const unsigned long __start___kcrctab[]; -extern const unsigned long __start___kcrctab_gpl[]; -extern const unsigned long __start___kcrctab_gpl_future[]; +extern const s32 __start___kcrctab[]; +extern const s32 __start___kcrctab_gpl[]; +extern const s32 __start___kcrctab_gpl_future[]; #ifdef CONFIG_UNUSED_SYMBOLS extern const struct kernel_symbol __start___ksymtab_unused[]; extern const struct kernel_symbol __stop___ksymtab_unused[]; extern const struct kernel_symbol __start___ksymtab_unused_gpl[]; extern const struct kernel_symbol __stop___ksymtab_unused_gpl[]; -extern const unsigned long __start___kcrctab_unused[]; -extern const unsigned long __start___kcrctab_unused_gpl[]; +extern const s32 __start___kcrctab_unused[]; +extern const s32 __start___kcrctab_unused_gpl[]; #endif #ifndef CONFIG_MODVERSIONS @@ -497,7 +497,7 @@ struct find_symbol_arg { /* Output */ struct module *owner; - const unsigned long *crc; + const s32 *crc; const struct kernel_symbol *sym; }; @@ -563,7 +563,7 @@ static bool find_symbol_in_section(const struct symsearch *syms, * (optional) module which owns it. Needs preempt disabled or module_mutex. */ const struct kernel_symbol *find_symbol(const char *name, struct module **owner, - const unsigned long **crc, + const s32 **crc, bool gplok, bool warn) { @@ -1249,23 +1249,17 @@ static int try_to_force_load(struct module *mod, const char *reason) } #ifdef CONFIG_MODVERSIONS -/* If the arch applies (non-zero) relocations to kernel kcrctab, unapply it. */ -static unsigned long maybe_relocated(unsigned long crc, - const struct module *crc_owner) + +static u32 resolve_rel_crc(const s32 *crc) { -#ifdef ARCH_RELOCATES_KCRCTAB - if (crc_owner == NULL) - return crc - (unsigned long)reloc_start; -#endif - return crc; + return *(u32 *)((void *)crc + *crc); } static int check_version(Elf_Shdr *sechdrs, unsigned int versindex, const char *symname, struct module *mod, - const unsigned long *crc, - const struct module *crc_owner) + const s32 *crc) { unsigned int i, num_versions; struct modversion_info *versions; @@ -1283,13 +1277,19 @@ static int check_version(Elf_Shdr *sechdrs, / sizeof(struct modversion_info); for (i = 0; i < num_versions; i++) { + u32 crcval; + if (strcmp(versions[i].name, symname) != 0) continue; - if (versions[i].crc == maybe_relocated(*crc, crc_owner)) + if (IS_ENABLED(CONFIG_MODULE_REL_CRCS)) + crcval = resolve_rel_crc(crc); + else + crcval = *crc; + if (versions[i].crc == crcval) return 1; - pr_debug("Found checksum %lX vs module %lX\n", - maybe_relocated(*crc, crc_owner), versions[i].crc); + pr_debug("Found checksum %X vs module %lX\n", + crcval, versions[i].crc); goto bad_version; } @@ -1307,7 +1307,7 @@ static inline int check_modstruct_version(Elf_Shdr *sechdrs, unsigned int versindex, struct module *mod) { - const unsigned long *crc; + const s32 *crc; /* * Since this should be found in kernel (which can't be removed), no @@ -1321,8 +1321,7 @@ static inline int check_modstruct_version(Elf_Shdr *sechdrs, } preempt_enable(); return check_version(sechdrs, versindex, - VMLINUX_SYMBOL_STR(module_layout), mod, crc, - NULL); + VMLINUX_SYMBOL_STR(module_layout), mod, crc); } /* First part is kernel version, which we ignore if module has crcs. */ @@ -1340,8 +1339,7 @@ static inline int check_version(Elf_Shdr *sechdrs, unsigned int versindex, const char *symname, struct module *mod, - const unsigned long *crc, - const struct module *crc_owner) + const s32 *crc) { return 1; } @@ -1368,7 +1366,7 @@ static const struct kernel_symbol *resolve_symbol(struct module *mod, { struct module *owner; const struct kernel_symbol *sym; - const unsigned long *crc; + const s32 *crc; int err; /* @@ -1383,8 +1381,7 @@ static const struct kernel_symbol *resolve_symbol(struct module *mod, if (!sym) goto unlock; - if (!check_version(info->sechdrs, info->index.vers, name, mod, crc, - owner)) { + if (!check_version(info->sechdrs, info->index.vers, name, mod, crc)) { sym = ERR_PTR(-EINVAL); goto getname; } -- cgit v1.2.3