From b5a9b340789b2b24c6896bcf7a065c31a4db671c Mon Sep 17 00:00:00 2001 From: Vincent Guittot Date: Wed, 19 Oct 2016 14:45:23 +0200 Subject: sched/fair: Fix incorrect task group ->load_avg A scheduler performance regression has been reported by Joseph Salisbury, which he bisected back to: 3d30544f0212 ("sched/fair: Apply more PELT fixes) The regression triggers when several levels of task groups are involved (read: SystemD) and cpu_possible_mask != cpu_present_mask. The root cause is that group entity's load (tg_child->se[i]->avg.load_avg) is initialized to scale_load_down(se->load.weight). During the creation of a child task group, its group entities on possible CPUs are attached to parent's cfs_rq (tg_parent) and their loads are added to the parent's load (tg_parent->load_avg) with update_tg_load_avg(). But only the load on online CPUs will then be updated to reflect real load, whereas load on other CPUs will stay at the initial value. The result is a tg_parent->load_avg that is higher than the real load, the weight of group entities (tg_parent->se[i]->load.weight) on online CPUs is smaller than it should be, and the task group gets a less running time than what it could expect. ( This situation can be detected with /proc/sched_debug. The ".tg_load_avg" of the task group will be much higher than sum of ".tg_load_avg_contrib" of online cfs_rqs of the task group. ) The load of group entities don't have to be intialized to something else than 0 because their load will increase when an entity is attached. Reported-by: Joseph Salisbury Tested-by: Dietmar Eggemann Signed-off-by: Vincent Guittot Acked-by: Peter Zijlstra Cc: # 4.8.x Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: joonwoop@codeaurora.org Fixes: 3d30544f0212 ("sched/fair: Apply more PELT fixes) Link: http://lkml.kernel.org/r/1476881123-10159-1-git-send-email-vincent.guittot@linaro.org Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 76ee7de1859d..d941c97dfbc3 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -690,7 +690,14 @@ void init_entity_runnable_average(struct sched_entity *se) * will definitely be update (after enqueue). */ sa->period_contrib = 1023; - sa->load_avg = scale_load_down(se->load.weight); + /* + * Tasks are intialized with full load to be seen as heavy tasks until + * they get a chance to stabilize to their real load level. + * Group entities are intialized with zero load to reflect the fact that + * nothing has been attached to the task group yet. + */ + if (entity_is_task(se)) + sa->load_avg = scale_load_down(se->load.weight); sa->load_sum = sa->load_avg * LOAD_AVG_MAX; /* * At this point, util_avg won't be used in select_task_rq_fair anyway -- cgit From 9beae1ea89305a9667ceaab6d0bf46a045ad71e7 Mon Sep 17 00:00:00 2001 From: Lorenzo Stoakes Date: Thu, 13 Oct 2016 01:20:17 +0100 Subject: mm: replace get_user_pages_remote() write/force parameters with gup_flags This removes the 'write' and 'force' from get_user_pages_remote() and replaces them with 'gup_flags' to make the use of FOLL_FORCE explicit in callers as use of this flag can result in surprising behaviour (and hence bugs) within the mm subsystem. Signed-off-by: Lorenzo Stoakes Acked-by: Michal Hocko Reviewed-by: Jan Kara Signed-off-by: Linus Torvalds --- kernel/events/uprobes.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c index d4129bb05e5d..f9ec9add2164 100644 --- a/kernel/events/uprobes.c +++ b/kernel/events/uprobes.c @@ -300,7 +300,8 @@ int uprobe_write_opcode(struct mm_struct *mm, unsigned long vaddr, retry: /* Read the page with vaddr into memory */ - ret = get_user_pages_remote(NULL, mm, vaddr, 1, 0, 1, &old_page, &vma); + ret = get_user_pages_remote(NULL, mm, vaddr, 1, FOLL_FORCE, &old_page, + &vma); if (ret <= 0) return ret; @@ -1710,7 +1711,8 @@ static int is_trap_at_addr(struct mm_struct *mm, unsigned long vaddr) * but we treat this as a 'remote' access since it is * essentially a kernel access to the memory. */ - result = get_user_pages_remote(NULL, mm, vaddr, 1, 0, 1, &page, NULL); + result = get_user_pages_remote(NULL, mm, vaddr, 1, FOLL_FORCE, &page, + NULL); if (result < 0) return result; -- cgit From f307ab6dcea03f9d8e4d70508fd7d1ca57cfa7f9 Mon Sep 17 00:00:00 2001 From: Lorenzo Stoakes Date: Thu, 13 Oct 2016 01:20:20 +0100 Subject: mm: replace access_process_vm() write parameter with gup_flags This removes the 'write' argument from access_process_vm() and replaces it with 'gup_flags' as use of this function previously silently implied FOLL_FORCE, whereas after this patch callers explicitly pass this flag. We make this explicit as use of FOLL_FORCE can result in surprising behaviour (and hence bugs) within the mm subsystem. Signed-off-by: Lorenzo Stoakes Acked-by: Jesper Nilsson Acked-by: Michal Hocko Acked-by: Michael Ellerman Signed-off-by: Linus Torvalds --- kernel/ptrace.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 2a99027312a6..e6474f7272ec 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -537,7 +537,7 @@ int ptrace_readdata(struct task_struct *tsk, unsigned long src, char __user *dst int this_len, retval; this_len = (len > sizeof(buf)) ? sizeof(buf) : len; - retval = access_process_vm(tsk, src, buf, this_len, 0); + retval = access_process_vm(tsk, src, buf, this_len, FOLL_FORCE); if (!retval) { if (copied) break; @@ -564,7 +564,8 @@ int ptrace_writedata(struct task_struct *tsk, char __user *src, unsigned long ds this_len = (len > sizeof(buf)) ? sizeof(buf) : len; if (copy_from_user(buf, src, this_len)) return -EFAULT; - retval = access_process_vm(tsk, dst, buf, this_len, 1); + retval = access_process_vm(tsk, dst, buf, this_len, + FOLL_FORCE | FOLL_WRITE); if (!retval) { if (copied) break; @@ -1127,7 +1128,7 @@ int generic_ptrace_peekdata(struct task_struct *tsk, unsigned long addr, unsigned long tmp; int copied; - copied = access_process_vm(tsk, addr, &tmp, sizeof(tmp), 0); + copied = access_process_vm(tsk, addr, &tmp, sizeof(tmp), FOLL_FORCE); if (copied != sizeof(tmp)) return -EIO; return put_user(tmp, (unsigned long __user *)data); @@ -1138,7 +1139,8 @@ int generic_ptrace_pokedata(struct task_struct *tsk, unsigned long addr, { int copied; - copied = access_process_vm(tsk, addr, &data, sizeof(data), 1); + copied = access_process_vm(tsk, addr, &data, sizeof(data), + FOLL_FORCE | FOLL_WRITE); return (copied == sizeof(data)) ? 0 : -EIO; } @@ -1155,7 +1157,8 @@ int compat_ptrace_request(struct task_struct *child, compat_long_t request, switch (request) { case PTRACE_PEEKTEXT: case PTRACE_PEEKDATA: - ret = access_process_vm(child, addr, &word, sizeof(word), 0); + ret = access_process_vm(child, addr, &word, sizeof(word), + FOLL_FORCE); if (ret != sizeof(word)) ret = -EIO; else @@ -1164,7 +1167,8 @@ int compat_ptrace_request(struct task_struct *child, compat_long_t request, case PTRACE_POKETEXT: case PTRACE_POKEDATA: - ret = access_process_vm(child, addr, &data, sizeof(data), 1); + ret = access_process_vm(child, addr, &data, sizeof(data), + FOLL_FORCE | FOLL_WRITE); ret = (ret != sizeof(data) ? -EIO : 0); break; -- cgit From 8835ca59dac2bc1e0136791abf3ccd51588803ce Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 19 Oct 2016 09:11:24 -0700 Subject: printk: suppress empty continuation lines We have a fairly common pattern where you print several things as continuations on one single line in a loop, and then at the end you do printk(KERN_CONT "\n"); to flush the buffered output. But if the output was flushed by something else (concurrent printk activity, or just system logging), we don't want that final flushing to just print an empty line. So just suppress empty continuation lines when they couldn't be merged into the line they are a continuation of. Signed-off-by: Linus Torvalds --- kernel/printk/printk.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'kernel') diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index d5e397315473..de08fc90baaf 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -1769,6 +1769,10 @@ static size_t log_output(int facility, int level, enum log_flags lflags, const c cont_flush(); } + /* Skip empty continuation lines that couldn't be added - they just flush */ + if (!text_len && (lflags & LOG_CONT)) + return 0; + /* If it doesn't end in a newline, try to buffer the current line */ if (!(lflags & LOG_NEWLINE)) { if (cont_add(facility, level, lflags, text, text_len)) -- cgit From 3118dac501bc0317de099db81618d589503351e1 Mon Sep 17 00:00:00 2001 From: Sudip Mukherjee Date: Thu, 6 Oct 2016 23:06:43 +0530 Subject: kernel/irq: Export irq_set_parent() The TPS65217 driver grew interrupt support which uses irq_set_parent(). While it's not yet clear why this is used in the first place, building the driver as a module fails with: ERROR: ".irq_set_parent" [drivers/mfd/tps65217.ko] undefined! The correctness of the driver change is still investigated, but for now it's less trouble to export irq_set_parent() than dealing with the build wreckage. [ tglx: Rewrote changelog and made the export GPL ] Fixes: 6556bdacf646 ("mfd: tps65217: Add support for IRQs") Signed-off-by: Sudip Mukherjee Cc: Sudip Mukherjee Cc: Marcin Niestroj Cc: Grygorii Strashko Cc: Tony Lindgren Cc: Lee Jones Link: http://lkml.kernel.org/r/1475775403-27207-1-git-send-email-sudipm.mukherjee@gmail.com Signed-off-by: Thomas Gleixner --- kernel/irq/manage.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index 0c5f1a5db654..9c4d30483264 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -721,6 +721,7 @@ int irq_set_parent(int irq, int parent_irq) irq_put_desc_unlock(desc, flags); return 0; } +EXPORT_SYMBOL_GPL(irq_set_parent); #endif /* -- cgit From f660f6066716b700148f60dba3461e65efff3123 Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Mon, 10 Oct 2016 15:10:51 +0300 Subject: softirq: Display IRQ_POLL for irq-poll statistics This library was moved to the generic area and was renamed to irq-poll. Hence, update proc/softirqs output accordingly. Signed-off-by: Sagi Grimberg Reviewed-by: Johannes Thumshirn Reviewed-by: Christoph Hellwig Signed-off-by: Jens Axboe --- kernel/softirq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/softirq.c b/kernel/softirq.c index 1bf81ef91375..744fa611cae0 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -58,7 +58,7 @@ static struct softirq_action softirq_vec[NR_SOFTIRQS] __cacheline_aligned_in_smp DEFINE_PER_CPU(struct task_struct *, ksoftirqd); const char * const softirq_to_name[NR_SOFTIRQS] = { - "HI", "TIMER", "NET_TX", "NET_RX", "BLOCK", "BLOCK_IOPOLL", + "HI", "TIMER", "NET_TX", "NET_RX", "BLOCK", "IRQ_POLL", "TASKLET", "SCHED", "HRTIMER", "RCU" }; -- cgit From 1adb469b9b76276d7e5ea36a20a24c47d6618a0b Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Fri, 21 Oct 2016 16:24:09 +0100 Subject: PM / suspend: Fix missing KERN_CONT for suspend message Commit 4bcc595ccd80 (printk: reinstate KERN_CONT for printing continuation lines) exposed a missing KERN_CONT from one of the messages shown on entering suspend. With v4.9-rc1, the 'done.' shown after syncing the filesystems no longer appears as a continuation but a new message with its own timestamp. [ 9.259566] PM: Syncing filesystems ... [ 9.264119] done. Fix this by adding the KERN_CONT log level for the 'done.' part of the message seen after syncing filesystems. While we are at it, convert these suspend printks to pr_info and pr_cont, respectively. Signed-off-by: Jon Hunter Signed-off-by: Rafael J. Wysocki --- kernel/power/suspend.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c index 1e7f5da648d9..6ccb08f57fcb 100644 --- a/kernel/power/suspend.c +++ b/kernel/power/suspend.c @@ -498,9 +498,9 @@ static int enter_state(suspend_state_t state) #ifndef CONFIG_SUSPEND_SKIP_SYNC trace_suspend_resume(TPS("sync_filesystems"), 0, true); - printk(KERN_INFO "PM: Syncing filesystems ... "); + pr_info("PM: Syncing filesystems ... "); sys_sync(); - printk("done.\n"); + pr_cont("done.\n"); trace_suspend_resume(TPS("sync_filesystems"), 0, false); #endif -- cgit From b831275a3553c32091222ac619cfddd73a5553fb Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Mon, 24 Oct 2016 11:41:56 +0200 Subject: timers: Plug locking race vs. timer migration Linus noticed that lock_timer_base() lacks a READ_ONCE() for accessing the timer flags. As a consequence the compiler is allowed to reload the flags between the initial check for TIMER_MIGRATION and the following timer base computation and the spin lock of the base. While this has not been observed (yet), we need to make sure that it never happens. Fixes: 0eeda71bc30d ("timer: Replace timer base by a cpu index") Reported-by: Linus Torvalds Signed-off-by: Thomas Gleixner Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1610241711220.4983@nanos Cc: stable@vger.kernel.org Cc: Andrew Morton Cc: Peter Zijlstra --- kernel/time/timer.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/time/timer.c b/kernel/time/timer.c index 2d47980a1bc4..0d4b91c5a374 100644 --- a/kernel/time/timer.c +++ b/kernel/time/timer.c @@ -943,7 +943,14 @@ static struct timer_base *lock_timer_base(struct timer_list *timer, { for (;;) { struct timer_base *base; - u32 tf = timer->flags; + u32 tf; + + /* + * We need to use READ_ONCE() here, otherwise the compiler + * might re-read @tf between the check for TIMER_MIGRATING + * and spin_lock(). + */ + tf = READ_ONCE(timer->flags); if (!(tf & TIMER_MIGRATING)) { base = get_timer_base(tf); -- cgit From 4da9152a4308dcbf611cde399c695c359fc9145f Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Mon, 24 Oct 2016 11:55:10 +0200 Subject: timers: Lock base for same bucket optimization Linus stumbled over the unlocked modification of the timer expiry value in mod_timer() which is an optimization for timers which stay in the same bucket - due to the bucket granularity - despite their expiry time getting updated. The optimization itself still makes sense even if we take the lock, because in case that the bucket stays the same, we avoid the pointless queue/enqueue dance. Make the check and the modification of timer->expires protected by the base lock and shuffle the remaining code around so we can keep the lock held when we actually have to requeue the timer to a different bucket. Fixes: f00c0afdfa62 ("timers: Implement optimization for same expiry time in mod_timer()") Reported-by: Linus Torvalds Signed-off-by: Thomas Gleixner Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1610241711220.4983@nanos Cc: stable@vger.kernel.org Cc: Andrew Morton Cc: Peter Zijlstra --- kernel/time/timer.c | 28 +++++++++++++++++----------- 1 file changed, 17 insertions(+), 11 deletions(-) (limited to 'kernel') diff --git a/kernel/time/timer.c b/kernel/time/timer.c index 0d4b91c5a374..ccf913038f9f 100644 --- a/kernel/time/timer.c +++ b/kernel/time/timer.c @@ -971,6 +971,8 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) unsigned long clk = 0, flags; int ret = 0; + BUG_ON(!timer->function); + /* * This is a common optimization triggered by the networking code - if * the timer is re-modified to have the same timeout or ends up in the @@ -979,13 +981,16 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) if (timer_pending(timer)) { if (timer->expires == expires) return 1; + /* - * Take the current timer_jiffies of base, but without holding - * the lock! + * We lock timer base and calculate the bucket index right + * here. If the timer ends up in the same bucket, then we + * just update the expiry time and avoid the whole + * dequeue/enqueue dance. */ - base = get_timer_base(timer->flags); - clk = base->clk; + base = lock_timer_base(timer, &flags); + clk = base->clk; idx = calc_wheel_index(expires, clk); /* @@ -995,14 +1000,14 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) */ if (idx == timer_get_idx(timer)) { timer->expires = expires; - return 1; + ret = 1; + goto out_unlock; } + } else { + base = lock_timer_base(timer, &flags); } timer_stats_timer_set_start_info(timer); - BUG_ON(!timer->function); - - base = lock_timer_base(timer, &flags); ret = detach_if_pending(timer, base, false); if (!ret && pending_only) @@ -1035,9 +1040,10 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) timer->expires = expires; /* * If 'idx' was calculated above and the base time did not advance - * between calculating 'idx' and taking the lock, only enqueue_timer() - * and trigger_dyntick_cpu() is required. Otherwise we need to - * (re)calculate the wheel index via internal_add_timer(). + * between calculating 'idx' and possibly switching the base, only + * enqueue_timer() and trigger_dyntick_cpu() is required. Otherwise + * we need to (re)calculate the wheel index via + * internal_add_timer(). */ if (idx != UINT_MAX && clk == base->clk) { enqueue_timer(base, timer, idx); -- cgit From 041ad7bc758db259bb960ef795197dd14aab19a6 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 22 Oct 2016 11:07:35 +0000 Subject: timers: Prevent base clock rewind when forwarding clock Ashton and Michael reported, that kernel versions 4.8 and later suffer from USB timeouts which are caused by the timer wheel rework. This is caused by a bug in the base clock forwarding mechanism, which leads to timers expiring early. The scenario which leads to this is: run_timers() while (jiffies >= base->clk) { collect_expired_timers(); base->clk++; expire_timers(); } So base->clk = jiffies + 1. Now the cpu goes idle: idle() get_next_timer_interrupt() nextevt = __next_time_interrupt(); if (time_after(nextevt, base->clk)) base->clk = jiffies; jiffies has not advanced since run_timers(), so this assignment effectively decrements base->clk by one. base->clk is the index into the timer wheel arrays. So let's assume the following state after the base->clk increment in run_timers(): jiffies = 0 base->clk = 1 A timer gets enqueued with an expiry delta of 63 ticks (which is the case with the USB timeout and HZ=250) so the resulting bucket index is: base->clk + delta = 1 + 63 = 64 The timer goes into the first wheel level. The array size is 64 so it ends up in bucket 0, which is correct as it takes 63 ticks to advance base->clk to index into bucket 0 again. If the cpu goes idle before jiffies advance, then the bug in the forwarding mechanism sets base->clk back to 0, so the next invocation of run_timers() at the next tick will index into bucket 0 and therefore expire the timer 62 ticks too early. Instead of blindly setting base->clk to jiffies we must make the forwarding conditional on jiffies > base->clk, but we cannot use jiffies for this as we might run into the following issue: if (time_after(jiffies, base->clk) { if (time_after(nextevt, base->clk)) base->clk = jiffies; jiffies can increment between the check and the assigment far enough to advance beyond nextevt. So we need to use a stable value for checking. get_next_timer_interrupt() has the basej argument which is the jiffies value snapshot taken in the calling code. So we can just that. Thanks to Ashton for bisecting and providing trace data! Fixes: a683f390b93f ("timers: Forward the wheel clock whenever possible") Reported-by: Ashton Holmes Reported-by: Michael Thayer Signed-off-by: Thomas Gleixner Cc: Michal Necasek Cc: Peter Zijlstra Cc: knut.osmundsen@oracle.com Cc: stable@vger.kernel.org Cc: stern@rowland.harvard.edu Cc: rt@linutronix.de Link: http://lkml.kernel.org/r/20161022110552.175308322@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/timer.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/time/timer.c b/kernel/time/timer.c index ccf913038f9f..7c446fb5163a 100644 --- a/kernel/time/timer.c +++ b/kernel/time/timer.c @@ -1523,12 +1523,16 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem) is_max_delta = (nextevt == base->clk + NEXT_TIMER_MAX_DELTA); base->next_expiry = nextevt; /* - * We have a fresh next event. Check whether we can forward the base: + * We have a fresh next event. Check whether we can forward the + * base. We can only do that when @basej is past base->clk + * otherwise we might rewind base->clk. */ - if (time_after(nextevt, jiffies)) - base->clk = jiffies; - else if (time_after(nextevt, base->clk)) - base->clk = nextevt; + if (time_after(basej, base->clk)) { + if (time_after(nextevt, basej)) + base->clk = basej; + else if (time_after(nextevt, base->clk)) + base->clk = nextevt; + } if (time_before_eq(nextevt, basej)) { expires = basem; -- cgit From 6bad6bccf2d717f652d37e63cf261eaa23466009 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 22 Oct 2016 11:07:37 +0000 Subject: timers: Prevent base clock corruption when forwarding When a timer is enqueued we try to forward the timer base clock. This mechanism has two issues: 1) Forwarding a remote base unlocked The forwarding function is called from get_target_base() with the current timer base lock held. But if the new target base is a different base than the current base (can happen with NOHZ, sigh!) then the forwarding is done on an unlocked base. This can lead to corruption of base->clk. Solution is simple: Invoke the forwarding after the target base is locked. 2) Possible corruption due to jiffies advancing This is similar to the issue in get_net_timer_interrupt() which was fixed in the previous patch. jiffies can advance between check and assignement and therefore advancing base->clk beyond the next expiry value. So we need to read jiffies into a local variable once and do the checks and assignment with the local copy. Fixes: a683f390b93f("timers: Forward the wheel clock whenever possible") Reported-by: Ashton Holmes Reported-by: Michael Thayer Signed-off-by: Thomas Gleixner Cc: Michal Necasek Cc: Peter Zijlstra Cc: knut.osmundsen@oracle.com Cc: stable@vger.kernel.org Cc: stern@rowland.harvard.edu Cc: rt@linutronix.de Link: http://lkml.kernel.org/r/20161022110552.253640125@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/timer.c | 23 ++++++++++------------- 1 file changed, 10 insertions(+), 13 deletions(-) (limited to 'kernel') diff --git a/kernel/time/timer.c b/kernel/time/timer.c index 7c446fb5163a..c611c47de884 100644 --- a/kernel/time/timer.c +++ b/kernel/time/timer.c @@ -878,7 +878,7 @@ static inline struct timer_base *get_timer_base(u32 tflags) #ifdef CONFIG_NO_HZ_COMMON static inline struct timer_base * -__get_target_base(struct timer_base *base, unsigned tflags) +get_target_base(struct timer_base *base, unsigned tflags) { #ifdef CONFIG_SMP if ((tflags & TIMER_PINNED) || !base->migration_enabled) @@ -891,25 +891,27 @@ __get_target_base(struct timer_base *base, unsigned tflags) static inline void forward_timer_base(struct timer_base *base) { + unsigned long jnow = READ_ONCE(jiffies); + /* * We only forward the base when it's idle and we have a delta between * base clock and jiffies. */ - if (!base->is_idle || (long) (jiffies - base->clk) < 2) + if (!base->is_idle || (long) (jnow - base->clk) < 2) return; /* * If the next expiry value is > jiffies, then we fast forward to * jiffies otherwise we forward to the next expiry value. */ - if (time_after(base->next_expiry, jiffies)) - base->clk = jiffies; + if (time_after(base->next_expiry, jnow)) + base->clk = jnow; else base->clk = base->next_expiry; } #else static inline struct timer_base * -__get_target_base(struct timer_base *base, unsigned tflags) +get_target_base(struct timer_base *base, unsigned tflags) { return get_timer_this_cpu_base(tflags); } @@ -917,14 +919,6 @@ __get_target_base(struct timer_base *base, unsigned tflags) static inline void forward_timer_base(struct timer_base *base) { } #endif -static inline struct timer_base * -get_target_base(struct timer_base *base, unsigned tflags) -{ - struct timer_base *target = __get_target_base(base, tflags); - - forward_timer_base(target); - return target; -} /* * We are using hashed locking: Holding per_cpu(timer_bases[x]).lock means @@ -1037,6 +1031,9 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) } } + /* Try to forward a stale timer base clock */ + forward_timer_base(base); + timer->expires = expires; /* * If 'idx' was calculated above and the base time did not advance -- cgit From f5d6d2da0d9098a4aa0ebcc187aa0fc167045d6b Mon Sep 17 00:00:00 2001 From: Tobias Klauser Date: Wed, 26 Oct 2016 13:37:04 +0200 Subject: sched/fair: Remove unused but set variable 'rq' MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Since commit: 8663e24d56dc ("sched/fair: Reorder cgroup creation code") ... the variable 'rq' in alloc_fair_sched_group() is set but no longer used. Remove it to fix the following GCC warning when building with 'W=1': kernel/sched/fair.c:8842:13: warning: variable ‘rq’ set but not used [-Wunused-but-set-variable] Signed-off-by: Tobias Klauser Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20161026113704.8981-1-tklauser@distanz.ch Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index d941c97dfbc3..c242944f5cbd 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -8839,7 +8839,6 @@ int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent) { struct sched_entity *se; struct cfs_rq *cfs_rq; - struct rq *rq; int i; tg->cfs_rq = kzalloc(sizeof(cfs_rq) * nr_cpu_ids, GFP_KERNEL); @@ -8854,8 +8853,6 @@ int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent) init_cfs_bandwidth(tg_cfs_bandwidth(tg)); for_each_possible_cpu(i) { - rq = cpu_rq(i); - cfs_rq = kzalloc_node(sizeof(struct cfs_rq), GFP_KERNEL, cpu_to_node(i)); if (!cfs_rq) -- cgit From 9dcb8b685fc30813b35ab4b4bf39244430753190 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 26 Oct 2016 10:15:30 -0700 Subject: mm: remove per-zone hashtable of bitlock waitqueues The per-zone waitqueues exist because of a scalability issue with the page waitqueues on some NUMA machines, but it turns out that they hurt normal loads, and now with the vmalloced stacks they also end up breaking gfs2 that uses a bit_wait on a stack object: wait_on_bit(&gh->gh_iflags, HIF_WAIT, TASK_UNINTERRUPTIBLE) where 'gh' can be a reference to the local variable 'mount_gh' on the stack of fill_super(). The reason the per-zone hash table breaks for this case is that there is no "zone" for virtual allocations, and trying to look up the physical page to get at it will fail (with a BUG_ON()). It turns out that I actually complained to the mm people about the per-zone hash table for another reason just a month ago: the zone lookup also hurts the regular use of "unlock_page()" a lot, because the zone lookup ends up forcing several unnecessary cache misses and generates horrible code. As part of that earlier discussion, we had a much better solution for the NUMA scalability issue - by just making the page lock have a separate contention bit, the waitqueue doesn't even have to be looked at for the normal case. Peter Zijlstra already has a patch for that, but let's see if anybody even notices. In the meantime, let's fix the actual gfs2 breakage by simplifying the bitlock waitqueues and removing the per-zone issue. Reported-by: Andreas Gruenbacher Tested-by: Bob Peterson Acked-by: Mel Gorman Cc: Peter Zijlstra Cc: Andy Lutomirski Cc: Steven Whitehouse Signed-off-by: Linus Torvalds --- kernel/sched/core.c | 16 ++++++++++++++++ kernel/sched/wait.c | 10 ---------- 2 files changed, 16 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 94732d1ab00a..42d4027f9e26 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -7515,11 +7515,27 @@ static struct kmem_cache *task_group_cache __read_mostly; DECLARE_PER_CPU(cpumask_var_t, load_balance_mask); DECLARE_PER_CPU(cpumask_var_t, select_idle_mask); +#define WAIT_TABLE_BITS 8 +#define WAIT_TABLE_SIZE (1 << WAIT_TABLE_BITS) +static wait_queue_head_t bit_wait_table[WAIT_TABLE_SIZE] __cacheline_aligned; + +wait_queue_head_t *bit_waitqueue(void *word, int bit) +{ + const int shift = BITS_PER_LONG == 32 ? 5 : 6; + unsigned long val = (unsigned long)word << shift | bit; + + return bit_wait_table + hash_long(val, WAIT_TABLE_BITS); +} +EXPORT_SYMBOL(bit_waitqueue); + void __init sched_init(void) { int i, j; unsigned long alloc_size = 0, ptr; + for (i = 0; i < WAIT_TABLE_SIZE; i++) + init_waitqueue_head(bit_wait_table + i); + #ifdef CONFIG_FAIR_GROUP_SCHED alloc_size += 2 * nr_cpu_ids * sizeof(void **); #endif diff --git a/kernel/sched/wait.c b/kernel/sched/wait.c index 4f7053579fe3..9453efe9b25a 100644 --- a/kernel/sched/wait.c +++ b/kernel/sched/wait.c @@ -480,16 +480,6 @@ void wake_up_bit(void *word, int bit) } EXPORT_SYMBOL(wake_up_bit); -wait_queue_head_t *bit_waitqueue(void *word, int bit) -{ - const int shift = BITS_PER_LONG == 32 ? 5 : 6; - const struct zone *zone = page_zone(virt_to_page(word)); - unsigned long val = (unsigned long)word << shift | bit; - - return &zone->wait_table[hash_long(val, zone->wait_table_bits)]; -} -EXPORT_SYMBOL(bit_waitqueue); - /* * Manipulate the atomic_t address to produce a better bit waitqueue table hash * index (we're keying off bit -1, but that would produce a horrible hash -- cgit From b274c0bb394c6a69ac12feac7c2db81f5aff5a55 Mon Sep 17 00:00:00 2001 From: Andrey Konovalov Date: Thu, 27 Oct 2016 17:46:21 -0700 Subject: kcov: properly check if we are in an interrupt in_interrupt() returns a nonzero value when we are either in an interrupt or have bh disabled via local_bh_disable(). Since we are interested in only ignoring coverage from actual interrupts, do a proper check instead of just calling in_interrupt(). As a result of this change, kcov will start to collect coverage from within local_bh_disable()/local_bh_enable() sections. Link: http://lkml.kernel.org/r/1476115803-20712-1-git-send-email-andreyknvl@google.com Signed-off-by: Andrey Konovalov Acked-by: Dmitry Vyukov Cc: Nicolai Stange Cc: Andrey Ryabinin Cc: Kees Cook Cc: James Morse Cc: Vegard Nossum Cc: Quentin Casasnovas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/kcov.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/kcov.c b/kernel/kcov.c index 8d44b3fea9d0..30e6d05aa5a9 100644 --- a/kernel/kcov.c +++ b/kernel/kcov.c @@ -53,8 +53,15 @@ void notrace __sanitizer_cov_trace_pc(void) /* * We are interested in code coverage as a function of a syscall inputs, * so we ignore code executed in interrupts. + * The checks for whether we are in an interrupt are open-coded, because + * 1. We can't use in_interrupt() here, since it also returns true + * when we are inside local_bh_disable() section. + * 2. We don't want to use (in_irq() | in_serving_softirq() | in_nmi()), + * since that leads to slower generated code (three separate tests, + * one for each of the flags). */ - if (!t || in_interrupt()) + if (!t || (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_OFFSET + | NMI_MASK))) return; mode = READ_ONCE(t->kcov_mode); if (mode == KCOV_MODE_TRACE) { -- cgit From 0933840acf7b65d6d30a5b6089d882afea57aca3 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Thu, 20 Oct 2016 13:10:11 +0200 Subject: perf/core: Protect PMU device removal with a 'pmu_bus_running' check, to fix CONFIG_DEBUG_TEST_DRIVER_REMOVE=y kernel panic CAI Qian reported a crash in the PMU uncore device removal code, enabled by the CONFIG_DEBUG_TEST_DRIVER_REMOVE=y option: https://marc.info/?l=linux-kernel&m=147688837328451 The reason for the crash is that perf_pmu_unregister() tries to remove a PMU device which is not added at this point. We add PMU devices only after pmu_bus is registered, which happens in the perf_event_sysfs_init() call and sets the 'pmu_bus_running' flag. The fix is to get the 'pmu_bus_running' flag state at the point the PMU is taken out of the PMU list and remove the device later only if it's set. Reported-by: CAI Qian Tested-by: CAI Qian Signed-off-by: Jiri Olsa Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Greg Kroah-Hartman Cc: Jiri Olsa Cc: Kan Liang Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Rob Herring Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20161020111011.GA13361@krava Signed-off-by: Ingo Molnar --- kernel/events/core.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index c6e47e97b33f..a5d2e62faf7e 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -8855,7 +8855,10 @@ EXPORT_SYMBOL_GPL(perf_pmu_register); void perf_pmu_unregister(struct pmu *pmu) { + int remove_device; + mutex_lock(&pmus_lock); + remove_device = pmu_bus_running; list_del_rcu(&pmu->entry); mutex_unlock(&pmus_lock); @@ -8869,10 +8872,12 @@ void perf_pmu_unregister(struct pmu *pmu) free_percpu(pmu->pmu_disable_count); if (pmu->type >= PERF_TYPE_MAX) idr_remove(&pmu_idr, pmu->type); - if (pmu->nr_addr_filters) - device_remove_file(pmu->dev, &dev_attr_nr_addr_filters); - device_del(pmu->dev); - put_device(pmu->dev); + if (remove_device) { + if (pmu->nr_addr_filters) + device_remove_file(pmu->dev, &dev_attr_nr_addr_filters); + device_del(pmu->dev); + put_device(pmu->dev); + } free_pmu_context(pmu); } EXPORT_SYMBOL_GPL(perf_pmu_unregister); -- cgit From 5aab90ce1ec449912a2ebc4d45e0c85dac29e9dd Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 26 Oct 2016 11:48:24 +0200 Subject: perf/powerpc: Don't call perf_event_disable() from atomic context The trinity syscall fuzzer triggered following WARN() on powerpc: WARNING: CPU: 9 PID: 2998 at arch/powerpc/kernel/hw_breakpoint.c:278 ... NIP [c00000000093aedc] .hw_breakpoint_handler+0x28c/0x2b0 LR [c00000000093aed8] .hw_breakpoint_handler+0x288/0x2b0 Call Trace: [c0000002f7933580] [c00000000093aed8] .hw_breakpoint_handler+0x288/0x2b0 (unreliable) [c0000002f7933630] [c0000000000f671c] .notifier_call_chain+0x7c/0xf0 [c0000002f79336d0] [c0000000000f6abc] .__atomic_notifier_call_chain+0xbc/0x1c0 [c0000002f7933780] [c0000000000f6c40] .notify_die+0x70/0xd0 [c0000002f7933820] [c00000000001a74c] .do_break+0x4c/0x100 [c0000002f7933920] [c0000000000089fc] handle_dabr_fault+0x14/0x48 Followed by a lockdep warning: =============================== [ INFO: suspicious RCU usage. ] 4.8.0-rc5+ #7 Tainted: G W ------------------------------- ./include/linux/rcupdate.h:556 Illegal context switch in RCU read-side critical section! other info that might help us debug this: rcu_scheduler_active = 1, debug_locks = 0 2 locks held by ls/2998: #0: (rcu_read_lock){......}, at: [] .__atomic_notifier_call_chain+0x0/0x1c0 #1: (rcu_read_lock){......}, at: [] .hw_breakpoint_handler+0x0/0x2b0 stack backtrace: CPU: 9 PID: 2998 Comm: ls Tainted: G W 4.8.0-rc5+ #7 Call Trace: [c0000002f7933150] [c00000000094b1f8] .dump_stack+0xe0/0x14c (unreliable) [c0000002f79331e0] [c00000000013c468] .lockdep_rcu_suspicious+0x138/0x180 [c0000002f7933270] [c0000000001005d8] .___might_sleep+0x278/0x2e0 [c0000002f7933300] [c000000000935584] .mutex_lock_nested+0x64/0x5a0 [c0000002f7933410] [c00000000023084c] .perf_event_ctx_lock_nested+0x16c/0x380 [c0000002f7933500] [c000000000230a80] .perf_event_disable+0x20/0x60 [c0000002f7933580] [c00000000093aeec] .hw_breakpoint_handler+0x29c/0x2b0 [c0000002f7933630] [c0000000000f671c] .notifier_call_chain+0x7c/0xf0 [c0000002f79336d0] [c0000000000f6abc] .__atomic_notifier_call_chain+0xbc/0x1c0 [c0000002f7933780] [c0000000000f6c40] .notify_die+0x70/0xd0 [c0000002f7933820] [c00000000001a74c] .do_break+0x4c/0x100 [c0000002f7933920] [c0000000000089fc] handle_dabr_fault+0x14/0x48 While it looks like the first WARN() is probably valid, the other one is triggered by disabling event via perf_event_disable() from atomic context. The event is disabled here in case we were not able to emulate the instruction that hit the breakpoint. By disabling the event we unschedule the event and make sure it's not scheduled back. But we can't call perf_event_disable() from atomic context, instead we need to use the event's pending_disable irq_work method to disable it. Reported-by: Jan Stancek Signed-off-by: Jiri Olsa Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Huang Ying Cc: Jiri Olsa Cc: Linus Torvalds Cc: Michael Neuling Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20161026094824.GA21397@krava Signed-off-by: Ingo Molnar --- kernel/events/core.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index a5d2e62faf7e..0e292132efac 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -1960,6 +1960,12 @@ void perf_event_disable(struct perf_event *event) } EXPORT_SYMBOL_GPL(perf_event_disable); +void perf_event_disable_inatomic(struct perf_event *event) +{ + event->pending_disable = 1; + irq_work_queue(&event->pending); +} + static void perf_set_shadow_time(struct perf_event *event, struct perf_event_context *ctx, u64 tstamp) @@ -7075,8 +7081,8 @@ static int __perf_event_overflow(struct perf_event *event, if (events && atomic_dec_and_test(&event->event_limit)) { ret = 1; event->pending_kill = POLL_HUP; - event->pending_disable = 1; - irq_work_queue(&event->pending); + + perf_event_disable_inatomic(event); } READ_ONCE(event->overflow_handler)(event, data, regs); -- cgit From 405c0759712f57b680f66aee9c55cd06ad1cbdef Mon Sep 17 00:00:00 2001 From: Andy Lutomirski Date: Mon, 31 Oct 2016 08:11:43 -0700 Subject: fork: Add task stack refcounting sanity check and prevent premature task stack freeing If something goes wrong with task stack refcounting and a stack refcount hits zero too early, warn and leak it rather than potentially freeing it early (and silently). Signed-off-by: Andy Lutomirski Cc: Borislav Petkov Cc: Brian Gerst Cc: Denys Vlasenko Cc: H. Peter Anvin Cc: Josh Poimboeuf Cc: Linus Torvalds Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/f29119c783a9680a4b4656e751b6123917ace94b.1477926663.git.luto@kernel.org Signed-off-by: Ingo Molnar --- kernel/fork.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 623259fc794d..997ac1d584f7 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -315,6 +315,9 @@ static void account_kernel_stack(struct task_struct *tsk, int account) static void release_task_stack(struct task_struct *tsk) { + if (WARN_ON(tsk->state != TASK_DEAD)) + return; /* Better to leak the stack than to free prematurely */ + account_kernel_stack(tsk, -1); arch_release_thread_stack(tsk->stack); free_thread_stack(tsk); @@ -1862,6 +1865,7 @@ bad_fork_cleanup_count: atomic_dec(&p->cred->user->processes); exit_creds(p); bad_fork_free: + p->state = TASK_DEAD; put_task_stack(p); free_task(p); fork_out: -- cgit From ceb75787bc75d0a7b88519ab8a68067ac690f55a Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 1 Nov 2016 11:49:56 +0100 Subject: PM / sleep: fix device reference leak in test_suspend Make sure to drop the reference taken by class_find_device() after opening the RTC device. Fixes: 77437fd4e61f (pm: boot time suspend selftest) Signed-off-by: Johan Hovold Signed-off-by: Rafael J. Wysocki --- kernel/power/suspend_test.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/power/suspend_test.c b/kernel/power/suspend_test.c index 084452e34a12..bdff5ed57f10 100644 --- a/kernel/power/suspend_test.c +++ b/kernel/power/suspend_test.c @@ -203,8 +203,10 @@ static int __init test_suspend(void) /* RTCs have initialized by now too ... can we use one? */ dev = class_find_device(rtc_class, NULL, NULL, has_wakealarm); - if (dev) + if (dev) { rtc = rtc_class_open(dev_name(dev)); + put_device(dev); + } if (!rtc) { printk(warn_no_rtc); return 0; -- cgit From 382005027fedc50b28d40ae64ef1461cca38953e Mon Sep 17 00:00:00 2001 From: Tetsuo Handa Date: Wed, 2 Nov 2016 19:50:29 +0900 Subject: sched/core: Fix oops in sched_show_task() When CONFIG_THREAD_INFO_IN_TASK=y, it is possible that an exited thread remains in the task list after its stack pointer was already set to NULL. Therefore, thread_saved_pc() and stack_not_used() in sched_show_task() will trigger NULL pointer dereference if an attempt to dump such thread's traces (e.g. SysRq-t, khungtaskd) is made. Since show_stack() in sched_show_task() calls try_get_task_stack() and sched_show_task() is called from interrupt context, calling try_get_task_stack() from sched_show_task() will be safe as well. Signed-off-by: Tetsuo Handa Acked-by: Andy Lutomirski Acked-by: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: bp@alien8.de Cc: brgerst@gmail.com Cc: jann@thejh.net Cc: keescook@chromium.org Cc: linux-api@vger.kernel.org Cc: tycho.andersen@canonical.com Link: http://lkml.kernel.org/r/201611021950.FEJ34368.HFFJOOMLtQOVSF@I-love.SAKURA.ne.jp Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 42d4027f9e26..9abf66b6c271 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -5192,6 +5192,8 @@ void sched_show_task(struct task_struct *p) int ppid; unsigned long state = p->state; + if (!try_get_task_stack(p)) + return; if (state) state = __ffs(state) + 1; printk(KERN_INFO "%-15.15s %c", p->comm, @@ -5221,6 +5223,7 @@ void sched_show_task(struct task_struct *p) print_worker_info(KERN_INFO, p); show_stack(p, NULL); + put_task_stack(p); } void show_state_filter(unsigned long state_filter) -- cgit From 8243d5597793b5e85143c9a935e1b971c59740a9 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 1 Nov 2016 17:47:18 -0600 Subject: sched/core: Remove pointless printout in sched_show_task() In sched_show_task() we print out a useless hex number, not even a symbol, and there's a big question mark whether this even makes sense anyway, I suspect we should just remove it all. Signed-off-by: Linus Torvalds Acked-by: Andy Lutomirski Cc: Peter Zijlstra Cc: Tetsuo Handa Cc: Thomas Gleixner Cc: bp@alien8.de Cc: brgerst@gmail.com Cc: jann@thejh.net Cc: keescook@chromium.org Cc: linux-api@vger.kernel.org Cc: tycho.andersen@canonical.com Link: http://lkml.kernel.org/r/CA+55aFzphURPFzAvU4z6Moy7ZmimcwPuUdYU8bj9z0J+S8X1rw@mail.gmail.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 9 --------- 1 file changed, 9 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 9abf66b6c271..154fd689fe02 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -5198,17 +5198,8 @@ void sched_show_task(struct task_struct *p) state = __ffs(state) + 1; printk(KERN_INFO "%-15.15s %c", p->comm, state < sizeof(stat_nam) - 1 ? stat_nam[state] : '?'); -#if BITS_PER_LONG == 32 - if (state == TASK_RUNNING) - printk(KERN_CONT " running "); - else - printk(KERN_CONT " %08lx ", thread_saved_pc(p)); -#else if (state == TASK_RUNNING) printk(KERN_CONT " running task "); - else - printk(KERN_CONT " %016lx ", thread_saved_pc(p)); -#endif #ifdef CONFIG_DEBUG_STACK_USAGE free = stack_not_used(p); #endif -- cgit From 243d52126184b072a18fe2130ce0008f8aa3a340 Mon Sep 17 00:00:00 2001 From: WANG Cong Date: Thu, 3 Nov 2016 09:42:36 -0700 Subject: taskstats: fix the length of cgroupstats_cmd_get_policy cgroupstats_cmd_get_policy is [CGROUPSTATS_CMD_ATTR_MAX+1], taskstats_cmd_get_policy[TASKSTATS_CMD_ATTR_MAX+1], but their family.maxattr is TASKSTATS_CMD_ATTR_MAX. CGROUPSTATS_CMD_ATTR_MAX is less than TASKSTATS_CMD_ATTR_MAX, so we could end up accessing out-of-bound. Change cgroupstats_cmd_get_policy to TASKSTATS_CMD_ATTR_MAX+1, this is safe because the rest are initialized to 0's. Reported-by: Andrey Konovalov Tested-by: Andrey Konovalov Signed-off-by: Cong Wang Signed-off-by: David S. Miller --- kernel/taskstats.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/taskstats.c b/kernel/taskstats.c index b3f05ee20d18..cbb387a265db 100644 --- a/kernel/taskstats.c +++ b/kernel/taskstats.c @@ -54,7 +54,11 @@ static const struct nla_policy taskstats_cmd_get_policy[TASKSTATS_CMD_ATTR_MAX+1 [TASKSTATS_CMD_ATTR_REGISTER_CPUMASK] = { .type = NLA_STRING }, [TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK] = { .type = NLA_STRING },}; -static const struct nla_policy cgroupstats_cmd_get_policy[CGROUPSTATS_CMD_ATTR_MAX+1] = { +/* + * We have to use TASKSTATS_CMD_ATTR_MAX here, it is the maxattr in the family. + * Make sure they are always aligned. + */ +static const struct nla_policy cgroupstats_cmd_get_policy[TASKSTATS_CMD_ATTR_MAX+1] = { [CGROUPSTATS_CMD_ATTR_FD] = { .type = NLA_U32 }, }; -- cgit From 483bed2b0ddd12ec33fc9407e0c6e1088e77a97c Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Fri, 4 Nov 2016 00:01:19 +0100 Subject: bpf: fix htab map destruction when extra reserve is in use Commit a6ed3ea65d98 ("bpf: restore behavior of bpf_map_update_elem") added an extra per-cpu reserve to the hash table map to restore old behaviour from pre prealloc times. When non-prealloc is in use for a map, then problem is that once a hash table extra element has been linked into the hash-table, and the hash table is destroyed due to refcount dropping to zero, then htab_map_free() -> delete_all_elements() will walk the whole hash table and drop all elements via htab_elem_free(). The problem is that the element from the extra reserve is first fed to the wrong backend allocator and eventually freed twice. Fixes: a6ed3ea65d98 ("bpf: restore behavior of bpf_map_update_elem") Reported-by: Dmitry Vyukov Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- kernel/bpf/hashtab.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 570eeca7bdfa..ad1bc67aff1b 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -687,7 +687,8 @@ static void delete_all_elements(struct bpf_htab *htab) hlist_for_each_entry_safe(l, n, head, hash_node) { hlist_del_rcu(&l->hash_node); - htab_elem_free(htab, l); + if (l->state != HTAB_EXTRA_ELEM_USED) + htab_elem_free(htab, l); } } } -- cgit From 20b2b24f91f70e7d3f0918c077546cb21bd73a87 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Fri, 4 Nov 2016 00:56:31 +0100 Subject: bpf: fix map not being uncharged during map creation failure In map_create(), we first find and create the map, then once that suceeded, we charge it to the user's RLIMIT_MEMLOCK, and then fetch a new anon fd through anon_inode_getfd(). The problem is, once the latter fails f.e. due to RLIMIT_NOFILE limit, then we only destruct the map via map->ops->map_free(), but without uncharging the previously locked memory first. That means that the user_struct allocation is leaked as well as the accounted RLIMIT_MEMLOCK memory not released. Make the label names in the fix consistent with bpf_prog_load(). Fixes: aaac3ba95e4c ("bpf: charge user for creation of BPF maps and programs") Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- kernel/bpf/syscall.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 228f962447a5..237f3d6a7ddc 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -194,7 +194,7 @@ static int map_create(union bpf_attr *attr) err = bpf_map_charge_memlock(map); if (err) - goto free_map; + goto free_map_nouncharge; err = bpf_map_new_fd(map); if (err < 0) @@ -204,6 +204,8 @@ static int map_create(union bpf_attr *attr) return err; free_map: + bpf_map_uncharge_memlock(map); +free_map_nouncharge: map->ops->map_free(map); return err; } -- cgit From 7ee7e87dfb158e79019ea1d5ea1b0e6f2bc93ee4 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Mon, 7 Nov 2016 19:57:00 +0100 Subject: genirq: Use irq type from irqdata instead of irqdesc The type flags in the irq descriptor are there for historical reasons and only updated via irq_modify_status() or irq_set_type(). Both functions also update the type flags in irqdata. __setup_irq() is the only left over user of the type flags in the irq descriptor. If __setup_irq() is called with empty irq type flags, then the type flags are retrieved from irqdata. If an interrupt is shared, then the type flags are compared with the type flags stored in the irq descriptor. On x86 the ioapic does not have a irq_set_type() callback because the type is defined in the BIOS tables and cannot be changed. The type is stored in irqdata at setup time without updating the type data in the irq descriptor. As a result the comparison described above fails. There is no point in updating the irq descriptor flags because the only relevant storage is irqdata. Use the type flags from irqdata for both retrieval and comparison in __setup_irq() instead. Aside of that the print out in case of non matching type flags has the old and new type flags arguments flipped. Fix that as well. For correctness sake the flags stored in the irq descriptor should be removed, but this is beyond the scope of this bugfix and will be done in a later patch. Fixes: 4b357daed698 ("genirq: Look-up trigger type if not specified by caller") Reported-and-tested-by: Mika Westerberg Signed-off-by: Thomas Gleixner Cc: Marc Zyngier Cc: Jon Hunter Cc: stable@vger.kernel.org Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1611072020360.3501@nanos Signed-off-by: Thomas Gleixner --- kernel/irq/manage.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index 9c4d30483264..6b669593e7eb 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -1341,12 +1341,12 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) } else if (new->flags & IRQF_TRIGGER_MASK) { unsigned int nmsk = new->flags & IRQF_TRIGGER_MASK; - unsigned int omsk = irq_settings_get_trigger_mask(desc); + unsigned int omsk = irqd_get_trigger_type(&desc->irq_data); if (nmsk != omsk) /* hope the handler works with current trigger mode */ pr_warn("irq %d uses trigger mode %u; requested %u\n", - irq, nmsk, omsk); + irq, omsk, nmsk); } *old_ptr = new; -- cgit From c6c7d83b9c9e6a8b3e6d84c820ac61fbffc9e396 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Thu, 10 Nov 2016 10:46:26 -0800 Subject: Revert "console: don't prefer first registered if DT specifies stdout-path" This reverts commit 05fd007e4629 ("console: don't prefer first registered if DT specifies stdout-path"). The reverted commit changes existing behavior on which many ARM boards rely. Many ARM small-board-computers, like e.g. the Raspberry Pi have both a video output and a serial console. Depending on whether the user is using the device as a more regular computer; or as a headless device we need to have the console on either one or the other. Many users rely on the kernel behavior of the console being present on both outputs, before the reverted commit the console setup with no console= kernel arguments on an ARM board which sets stdout-path in dt would look like this: [root@localhost ~]# cat /proc/consoles ttyS0 -W- (EC p a) 4:64 tty0 -WU (E p ) 4:1 Where as after the reverted commit, it looks like this: [root@localhost ~]# cat /proc/consoles ttyS0 -W- (EC p a) 4:64 This commit reverts commit 05fd007e4629 ("console: don't prefer first registered if DT specifies stdout-path") restoring the original behavior. Fixes: 05fd007e4629 ("console: don't prefer first registered if DT specifies stdout-path") Link: http://lkml.kernel.org/r/20161104121135.4780-2-hdegoede@redhat.com Signed-off-by: Hans de Goede Cc: Paul Burton Cc: Rob Herring Cc: Frank Rowand Cc: Thorsten Leemhuis Cc: Greg Kroah-Hartman Cc: Tejun Heo Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/printk/printk.c | 13 +------------ 1 file changed, 1 insertion(+), 12 deletions(-) (limited to 'kernel') diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index de08fc90baaf..5028f4fd504a 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -253,17 +253,6 @@ static int preferred_console = -1; int console_set_on_cmdline; EXPORT_SYMBOL(console_set_on_cmdline); -#ifdef CONFIG_OF -static bool of_specified_console; - -void console_set_by_of(void) -{ - of_specified_console = true; -} -#else -# define of_specified_console false -#endif - /* Flag: console code may call schedule() */ static int console_may_schedule; @@ -2657,7 +2646,7 @@ void register_console(struct console *newcon) * didn't select a console we take the first one * that registers here. */ - if (preferred_console < 0 && !of_specified_console) { + if (preferred_console < 0) { if (newcon->index < 0) newcon->index = 0; if (newcon->setup == NULL || -- cgit From f5c9f9c72395c3291c2e35c905dedae2b98475a4 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 14 Nov 2016 09:31:52 -0800 Subject: Revert "printk: make reading the kernel log flush pending lines" This reverts commit bfd8d3f23b51018388be0411ccbc2d56277fe294. It turns out that this flushes things much too aggressiverly, and causes lines to break up when the system logger races with new continuation lines being printed. There's a pending patch to make printk() flushing much more straightforward, but it's too invasive for 4.9, so in the meantime let's just not make the system message logging flush continuation lines. They'll be flushed by the final newline anyway. Suggested-by: Petr Mladek Signed-off-by: Linus Torvalds --- kernel/printk/printk.c | 11 ----------- 1 file changed, 11 deletions(-) (limited to 'kernel') diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 5028f4fd504a..f7a55e9ff2f7 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -783,8 +783,6 @@ static ssize_t devkmsg_write(struct kiocb *iocb, struct iov_iter *from) return ret; } -static void cont_flush(void); - static ssize_t devkmsg_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { @@ -800,7 +798,6 @@ static ssize_t devkmsg_read(struct file *file, char __user *buf, if (ret) return ret; raw_spin_lock_irq(&logbuf_lock); - cont_flush(); while (user->seq == log_next_seq) { if (file->f_flags & O_NONBLOCK) { ret = -EAGAIN; @@ -863,7 +860,6 @@ static loff_t devkmsg_llseek(struct file *file, loff_t offset, int whence) return -ESPIPE; raw_spin_lock_irq(&logbuf_lock); - cont_flush(); switch (whence) { case SEEK_SET: /* the first record */ @@ -902,7 +898,6 @@ static unsigned int devkmsg_poll(struct file *file, poll_table *wait) poll_wait(file, &log_wait, wait); raw_spin_lock_irq(&logbuf_lock); - cont_flush(); if (user->seq < log_next_seq) { /* return error when data has vanished underneath us */ if (user->seq < log_first_seq) @@ -1289,7 +1284,6 @@ static int syslog_print(char __user *buf, int size) size_t skip; raw_spin_lock_irq(&logbuf_lock); - cont_flush(); if (syslog_seq < log_first_seq) { /* messages are gone, move to first one */ syslog_seq = log_first_seq; @@ -1349,7 +1343,6 @@ static int syslog_print_all(char __user *buf, int size, bool clear) return -ENOMEM; raw_spin_lock_irq(&logbuf_lock); - cont_flush(); if (buf) { u64 next_seq; u64 seq; @@ -1511,7 +1504,6 @@ int do_syslog(int type, char __user *buf, int len, int source) /* Number of chars in the log buffer */ case SYSLOG_ACTION_SIZE_UNREAD: raw_spin_lock_irq(&logbuf_lock); - cont_flush(); if (syslog_seq < log_first_seq) { /* messages are gone, move to first one */ syslog_seq = log_first_seq; @@ -3028,7 +3020,6 @@ void kmsg_dump(enum kmsg_dump_reason reason) dumper->active = true; raw_spin_lock_irqsave(&logbuf_lock, flags); - cont_flush(); dumper->cur_seq = clear_seq; dumper->cur_idx = clear_idx; dumper->next_seq = log_next_seq; @@ -3119,7 +3110,6 @@ bool kmsg_dump_get_line(struct kmsg_dumper *dumper, bool syslog, bool ret; raw_spin_lock_irqsave(&logbuf_lock, flags); - cont_flush(); ret = kmsg_dump_get_line_nolock(dumper, syslog, line, size, len); raw_spin_unlock_irqrestore(&logbuf_lock, flags); @@ -3162,7 +3152,6 @@ bool kmsg_dump_get_buffer(struct kmsg_dumper *dumper, bool syslog, goto out; raw_spin_lock_irqsave(&logbuf_lock, flags); - cont_flush(); if (dumper->cur_seq < log_first_seq) { /* messages are gone, move to first available one */ dumper->cur_seq = log_first_seq; -- cgit From 977c1f9c8c022d0173181766b34a0db3705265a4 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Mon, 7 Nov 2016 15:14:20 -0800 Subject: ftrace: Ignore FTRACE_FL_DISABLED while walking dyn_ftrace records ftrace_shutdown() checks for sanity of ftrace records and if dyn_ftrace->flags is not zero, it will warn. It can happen that 'flags' are set to FTRACE_FL_DISABLED at this point, since some module was loaded, but before ftrace_module_enable() cleared the flags for this module. In other words the module.c is doing: ftrace_module_init(mod); // calls ftrace_update_code() that sets flags=FTRACE_FL_DISABLED ... // here ftrace_shutdown() is called that warns, since err = prepare_coming_module(mod); // didn't have a chance to clear FTRACE_FL_DISABLED Fix it by ignoring disabled records. It's similar to what __ftrace_hash_rec_update() is already doing. Link: http://lkml.kernel.org/r/1478560460-3818619-1-git-send-email-ast@fb.com Cc: stable@vger.kernel.org Fixes: b7ffffbb46f2 "ftrace: Add infrastructure for delayed enabling of module functions" Signed-off-by: Alexei Starovoitov Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 2050a7652a86..326498baab83 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -2763,7 +2763,7 @@ static int ftrace_shutdown(struct ftrace_ops *ops, int command) struct dyn_ftrace *rec; do_for_each_ftrace_rec(pg, rec) { - if (FTRACE_WARN_ON_ONCE(rec->flags)) + if (FTRACE_WARN_ON_ONCE(rec->flags & ~FTRACE_FL_DISABLED)) pr_warn(" %pS flags:%lx\n", (void *)rec->ip, rec->flags); } while_for_each_ftrace_rec(); -- cgit From 546fece4eae871f033925ccf0ff2b740725ae915 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (Red Hat)" Date: Mon, 14 Nov 2016 16:31:49 -0500 Subject: ftrace: Add more checks for FTRACE_FL_DISABLED in processing ip records When a module is first loaded and its function ip records are added to the ftrace list of functions to modify, they are set to DISABLED, as their text is still in a read only state. When the module is fully loaded, and can be updated, the flag is cleared, and if their's any functions that should be tracing them, it is updated at that moment. But there's several locations that do record accounting and should ignore records that are marked as disabled, or they can cause issues. Alexei already fixed one location, but others need to be addressed. Cc: stable@vger.kernel.org Fixes: b7ffffbb46f2 "ftrace: Add infrastructure for delayed enabling of module functions" Reported-by: Alexei Starovoitov Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 326498baab83..da87b3cba5b3 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1862,6 +1862,10 @@ static int __ftrace_hash_update_ipmodify(struct ftrace_ops *ops, /* Update rec->flags */ do_for_each_ftrace_rec(pg, rec) { + + if (rec->flags & FTRACE_FL_DISABLED) + continue; + /* We need to update only differences of filter_hash */ in_old = !!ftrace_lookup_ip(old_hash, rec->ip); in_new = !!ftrace_lookup_ip(new_hash, rec->ip); @@ -1884,6 +1888,10 @@ rollback: /* Roll back what we did above */ do_for_each_ftrace_rec(pg, rec) { + + if (rec->flags & FTRACE_FL_DISABLED) + continue; + if (rec == end) goto err_out; @@ -2397,6 +2405,10 @@ void __weak ftrace_replace_code(int enable) return; do_for_each_ftrace_rec(pg, rec) { + + if (rec->flags & FTRACE_FL_DISABLED) + continue; + failed = __ftrace_replace_code(rec, enable); if (failed) { ftrace_bug(failed, rec); @@ -3598,6 +3610,10 @@ match_records(struct ftrace_hash *hash, char *func, int len, char *mod) goto out_unlock; do_for_each_ftrace_rec(pg, rec) { + + if (rec->flags & FTRACE_FL_DISABLED) + continue; + if (ftrace_match_record(rec, &func_g, mod_match, exclude_mod)) { ret = enter_record(hash, rec, clear_filter); if (ret < 0) { @@ -3793,6 +3809,9 @@ register_ftrace_function_probe(char *glob, struct ftrace_probe_ops *ops, do_for_each_ftrace_rec(pg, rec) { + if (rec->flags & FTRACE_FL_DISABLED) + continue; + if (!ftrace_match_record(rec, &func_g, NULL, 0)) continue; @@ -4685,6 +4704,9 @@ ftrace_set_func(unsigned long *array, int *idx, int size, char *buffer) do_for_each_ftrace_rec(pg, rec) { + if (rec->flags & FTRACE_FL_DISABLED) + continue; + if (ftrace_match_record(rec, &func_g, NULL, 0)) { /* if it is in the array */ exists = false; -- cgit From 864c2357ca898c6171fe5284f5ecc795c8ce27a8 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Tue, 1 Nov 2016 11:52:58 -0700 Subject: perf/core: Do not set cpuctx->cgrp for unscheduled cgroups Commit: db4a835601b7 ("perf/core: Set cgroup in CPU contexts for new cgroup events") failed to verify that event->cgrp is actually the scheduled cgroup in a CPU before setting cpuctx->cgrp. This patch fixes that. Now that there is a different path for scheduled and unscheduled cgroup, add a warning to catch when cpuctx->cgrp is still set after the last cgroup event has been unsheduled. To verify the bug: # Create 2 cgroups. mkdir /dev/cgroups/devices/g1 mkdir /dev/cgroups/devices/g2 # launch a task, bind it to a cpu and move it to g1 CPU=2 while :; do : ; done & P=$! taskset -pc $CPU $P echo $P > /dev/cgroups/devices/g1/tasks # monitor g2 (it runs no tasks) and observe output perf stat -e cycles -I 1000 -C $CPU -G g2 # time counts unit events 1.000091408 7,579,527 cycles g2 2.000350111 cycles g2 3.000589181 cycles g2 4.000771428 cycles g2 # note first line that displays that a task run in g2, despite # g2 having no tasks. This is because cpuctx->cgrp was wrongly # set when context of new event was installed. # After applying the fix we obtain the right output: perf stat -e cycles -I 1000 -C $CPU -G g2 # time counts unit events 1.000119615 cycles g2 2.000389430 cycles g2 3.000590962 cycles g2 Signed-off-by: David Carrillo-Cisneros Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Kan Liang Cc: Linus Torvalds Cc: Nilay Vaish Cc: Paul Turner Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vegard Nossum Link: http://lkml.kernel.org/r/1478026378-86083-1-git-send-email-davidcc@google.com Signed-off-by: Ingo Molnar --- kernel/events/core.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 0e292132efac..ff230bb4a02e 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -902,6 +902,17 @@ list_update_cgroup_event(struct perf_event *event, * this will always be called from the right CPU. */ cpuctx = __get_cpu_context(ctx); + + /* Only set/clear cpuctx->cgrp if current task uses event->cgrp. */ + if (perf_cgroup_from_task(current, ctx) != event->cgrp) { + /* + * We are removing the last cpu event in this context. + * If that event is not active in this cpu, cpuctx->cgrp + * should've been cleared by perf_cgroup_switch. + */ + WARN_ON_ONCE(!add && cpuctx->cgrp); + return; + } cpuctx->cgrp = add ? event->cgrp : NULL; } -- cgit From f23cc643f9baec7f71f2b74692da3cf03abbbfda Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 14 Nov 2016 15:45:36 -0500 Subject: bpf: fix range arithmetic for bpf map access I made some invalid assumptions with BPF_AND and BPF_MOD that could result in invalid accesses to bpf map entries. Fix this up by doing a few things 1) Kill BPF_MOD support. This doesn't actually get used by the compiler in real life and just adds extra complexity. 2) Fix the logic for BPF_AND, don't allow AND of negative numbers and set the minimum value to 0 for positive AND's. 3) Don't do operations on the ranges if they are set to the limits, as they are by definition undefined, and allowing arithmetic operations on those values could make them appear valid when they really aren't. This fixes the testcase provided by Jann as well as a few other theoretical problems. Reported-by: Jann Horn Signed-off-by: Josef Bacik Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- kernel/bpf/verifier.c | 70 ++++++++++++++++++++++++++++++++++----------------- 1 file changed, 47 insertions(+), 23 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 99a7e5b388f2..6a936159c6e0 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -216,8 +216,8 @@ static void print_verifier_state(struct bpf_verifier_state *state) reg->map_ptr->key_size, reg->map_ptr->value_size); if (reg->min_value != BPF_REGISTER_MIN_RANGE) - verbose(",min_value=%llu", - (unsigned long long)reg->min_value); + verbose(",min_value=%lld", + (long long)reg->min_value); if (reg->max_value != BPF_REGISTER_MAX_RANGE) verbose(",max_value=%llu", (unsigned long long)reg->max_value); @@ -758,7 +758,7 @@ static int check_mem_access(struct bpf_verifier_env *env, u32 regno, int off, * index'es we need to make sure that whatever we use * will have a set floor within our range. */ - if ((s64)reg->min_value < 0) { + if (reg->min_value < 0) { verbose("R%d min value is negative, either use unsigned index or do a if (index >=0) check.\n", regno); return -EACCES; @@ -1468,7 +1468,8 @@ static void check_reg_overflow(struct bpf_reg_state *reg) { if (reg->max_value > BPF_REGISTER_MAX_RANGE) reg->max_value = BPF_REGISTER_MAX_RANGE; - if ((s64)reg->min_value < BPF_REGISTER_MIN_RANGE) + if (reg->min_value < BPF_REGISTER_MIN_RANGE || + reg->min_value > BPF_REGISTER_MAX_RANGE) reg->min_value = BPF_REGISTER_MIN_RANGE; } @@ -1476,7 +1477,8 @@ static void adjust_reg_min_max_vals(struct bpf_verifier_env *env, struct bpf_insn *insn) { struct bpf_reg_state *regs = env->cur_state.regs, *dst_reg; - u64 min_val = BPF_REGISTER_MIN_RANGE, max_val = BPF_REGISTER_MAX_RANGE; + s64 min_val = BPF_REGISTER_MIN_RANGE; + u64 max_val = BPF_REGISTER_MAX_RANGE; bool min_set = false, max_set = false; u8 opcode = BPF_OP(insn->code); @@ -1512,22 +1514,43 @@ static void adjust_reg_min_max_vals(struct bpf_verifier_env *env, return; } + /* If one of our values was at the end of our ranges then we can't just + * do our normal operations to the register, we need to set the values + * to the min/max since they are undefined. + */ + if (min_val == BPF_REGISTER_MIN_RANGE) + dst_reg->min_value = BPF_REGISTER_MIN_RANGE; + if (max_val == BPF_REGISTER_MAX_RANGE) + dst_reg->max_value = BPF_REGISTER_MAX_RANGE; + switch (opcode) { case BPF_ADD: - dst_reg->min_value += min_val; - dst_reg->max_value += max_val; + if (dst_reg->min_value != BPF_REGISTER_MIN_RANGE) + dst_reg->min_value += min_val; + if (dst_reg->max_value != BPF_REGISTER_MAX_RANGE) + dst_reg->max_value += max_val; break; case BPF_SUB: - dst_reg->min_value -= min_val; - dst_reg->max_value -= max_val; + if (dst_reg->min_value != BPF_REGISTER_MIN_RANGE) + dst_reg->min_value -= min_val; + if (dst_reg->max_value != BPF_REGISTER_MAX_RANGE) + dst_reg->max_value -= max_val; break; case BPF_MUL: - dst_reg->min_value *= min_val; - dst_reg->max_value *= max_val; + if (dst_reg->min_value != BPF_REGISTER_MIN_RANGE) + dst_reg->min_value *= min_val; + if (dst_reg->max_value != BPF_REGISTER_MAX_RANGE) + dst_reg->max_value *= max_val; break; case BPF_AND: - /* & is special since it could end up with 0 bits set. */ - dst_reg->min_value &= min_val; + /* Disallow AND'ing of negative numbers, ain't nobody got time + * for that. Otherwise the minimum is 0 and the max is the max + * value we could AND against. + */ + if (min_val < 0) + dst_reg->min_value = BPF_REGISTER_MIN_RANGE; + else + dst_reg->min_value = 0; dst_reg->max_value = max_val; break; case BPF_LSH: @@ -1537,24 +1560,25 @@ static void adjust_reg_min_max_vals(struct bpf_verifier_env *env, */ if (min_val > ilog2(BPF_REGISTER_MAX_RANGE)) dst_reg->min_value = BPF_REGISTER_MIN_RANGE; - else + else if (dst_reg->min_value != BPF_REGISTER_MIN_RANGE) dst_reg->min_value <<= min_val; if (max_val > ilog2(BPF_REGISTER_MAX_RANGE)) dst_reg->max_value = BPF_REGISTER_MAX_RANGE; - else + else if (dst_reg->max_value != BPF_REGISTER_MAX_RANGE) dst_reg->max_value <<= max_val; break; case BPF_RSH: - dst_reg->min_value >>= min_val; - dst_reg->max_value >>= max_val; - break; - case BPF_MOD: - /* % is special since it is an unsigned modulus, so the floor - * will always be 0. + /* RSH by a negative number is undefined, and the BPF_RSH is an + * unsigned shift, so make the appropriate casts. */ - dst_reg->min_value = 0; - dst_reg->max_value = max_val - 1; + if (min_val < 0 || dst_reg->min_value < 0) + dst_reg->min_value = BPF_REGISTER_MIN_RANGE; + else + dst_reg->min_value = + (u64)(dst_reg->min_value) >> min_val; + if (dst_reg->max_value != BPF_REGISTER_MAX_RANGE) + dst_reg->max_value >>= max_val; break; default: reset_reg_range_values(regs, insn->dst_reg); -- cgit From e245d99e6cc4a0b904b87b46b4f60d46fb405987 Mon Sep 17 00:00:00 2001 From: Babu Moger Date: Wed, 2 Nov 2016 09:36:33 -0700 Subject: lockdep: Limit static allocations if PROVE_LOCKING_SMALL is defined Reduce the size of data structure for lockdep entries by half if PROVE_LOCKING_SMALL if defined. This is used only for sparc. Signed-off-by: Babu Moger Acked-by: Sam Ravnborg Signed-off-by: David S. Miller --- kernel/locking/lockdep_internals.h | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/lockdep_internals.h b/kernel/locking/lockdep_internals.h index 51c4b24b6328..c2b88490d857 100644 --- a/kernel/locking/lockdep_internals.h +++ b/kernel/locking/lockdep_internals.h @@ -45,6 +45,14 @@ enum { #define LOCKF_USED_IN_IRQ_READ \ (LOCKF_USED_IN_HARDIRQ_READ | LOCKF_USED_IN_SOFTIRQ_READ) +/* + * CONFIG_PROVE_LOCKING_SMALL is defined for sparc. Sparc requires .text, + * .data and .bss to fit in required 32MB limit for the kernel. With + * PROVE_LOCKING we could go over this limit and cause system boot-up problems. + * So, reduce the static allocations for lockdeps related structures so that + * everything fits in current required size limit. + */ +#ifdef CONFIG_PROVE_LOCKING_SMALL /* * MAX_LOCKDEP_ENTRIES is the maximum number of lock dependencies * we track. @@ -54,18 +62,24 @@ enum { * table (if it's not there yet), and we check it for lock order * conflicts and deadlocks. */ +#define MAX_LOCKDEP_ENTRIES 16384UL +#define MAX_LOCKDEP_CHAINS_BITS 15 +#define MAX_STACK_TRACE_ENTRIES 262144UL +#else #define MAX_LOCKDEP_ENTRIES 32768UL #define MAX_LOCKDEP_CHAINS_BITS 16 -#define MAX_LOCKDEP_CHAINS (1UL << MAX_LOCKDEP_CHAINS_BITS) - -#define MAX_LOCKDEP_CHAIN_HLOCKS (MAX_LOCKDEP_CHAINS*5) /* * Stack-trace: tightly packed array of stack backtrace * addresses. Protected by the hash_lock. */ #define MAX_STACK_TRACE_ENTRIES 524288UL +#endif + +#define MAX_LOCKDEP_CHAINS (1UL << MAX_LOCKDEP_CHAINS_BITS) + +#define MAX_LOCKDEP_CHAIN_HLOCKS (MAX_LOCKDEP_CHAINS*5) extern struct list_head all_lock_classes; extern struct lock_chain lock_chains[]; -- cgit From e96271f3ed7e702fa36dd0605c0c5b5f065af816 Mon Sep 17 00:00:00 2001 From: Alexander Shishkin Date: Fri, 18 Nov 2016 13:38:43 +0200 Subject: perf/core: Fix address filter parser The token table passed into match_token() must be null-terminated, which it currently is not in the perf's address filter string parser, as caught by Vince's perf_fuzzer and KASAN. It doesn't blow up otherwise because of the alignment padding of the table to the next element in the .rodata, which is luck. Fixing by adding a null-terminator to the token table. Reported-by: Vince Weaver Tested-by: Vince Weaver Signed-off-by: Alexander Shishkin Acked-by: Peter Zijlstra (Intel) Cc: Arnaldo Carvalho de Melo Cc: Linus Torvalds Cc: Thomas Gleixner Cc: dvyukov@google.com Cc: stable@vger.kernel.org # v4.7+ Fixes: 375637bc524 ("perf/core: Introduce address range filtering") Link: http://lkml.kernel.org/r/877f81f264.fsf@ashishki-desk.ger.corp.intel.com Signed-off-by: Ingo Molnar --- kernel/events/core.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index ff230bb4a02e..6ee1febdf6ff 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -8029,6 +8029,7 @@ restart: * if is not specified, the range is treated as a single address. */ enum { + IF_ACT_NONE = -1, IF_ACT_FILTER, IF_ACT_START, IF_ACT_STOP, @@ -8052,6 +8053,7 @@ static const match_table_t if_tokens = { { IF_SRC_KERNEL, "%u/%u" }, { IF_SRC_FILEADDR, "%u@%s" }, { IF_SRC_KERNELADDR, "%u" }, + { IF_ACT_NONE, NULL }, }; /* -- cgit From 18f649ef344127ef6de23a5a4272dbe2fdb73dde Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 14 Nov 2016 19:46:09 +0100 Subject: sched/autogroup: Fix autogroup_move_group() to never skip sched_move_task() The PF_EXITING check in task_wants_autogroup() is no longer needed. Remove it, but see the next patch. However the comment is correct in that autogroup_move_group() must always change task_group() for every thread so the sysctl_ check is very wrong; we can race with cgroups and even sys_setsid() is not safe because a task running with task_group() == ag->tg must participate in refcounting: int main(void) { int sctl = open("/proc/sys/kernel/sched_autogroup_enabled", O_WRONLY); assert(sctl > 0); if (fork()) { wait(NULL); // destroy the child's ag/tg pause(); } assert(pwrite(sctl, "1\n", 2, 0) == 2); assert(setsid() > 0); if (fork()) pause(); kill(getppid(), SIGKILL); sleep(1); // The child has gone, the grandchild runs with kref == 1 assert(pwrite(sctl, "0\n", 2, 0) == 2); assert(setsid() > 0); // runs with the freed ag/tg for (;;) sleep(1); return 0; } crashes the kernel. It doesn't really need sleep(1), it doesn't matter if autogroup_move_group() actually frees the task_group or this happens later. Reported-by: Vern Lovejoy Signed-off-by: Oleg Nesterov Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: hartsjc@redhat.com Cc: vbendel@redhat.com Link: http://lkml.kernel.org/r/20161114184609.GA15965@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/auto_group.c | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/auto_group.c b/kernel/sched/auto_group.c index a5d966cb8891..ad2b19ad6ca0 100644 --- a/kernel/sched/auto_group.c +++ b/kernel/sched/auto_group.c @@ -111,14 +111,11 @@ bool task_wants_autogroup(struct task_struct *p, struct task_group *tg) { if (tg != &root_task_group) return false; - /* - * We can only assume the task group can't go away on us if - * autogroup_move_group() can see us on ->thread_group list. + * If we race with autogroup_move_group() the caller can use the old + * value of signal->autogroup but in this case sched_move_task() will + * be called again before autogroup_kref_put(). */ - if (p->flags & PF_EXITING) - return false; - return true; } @@ -138,13 +135,17 @@ autogroup_move_group(struct task_struct *p, struct autogroup *ag) } p->signal->autogroup = autogroup_kref_get(ag); - - if (!READ_ONCE(sysctl_sched_autogroup_enabled)) - goto out; - + /* + * We can't avoid sched_move_task() after we changed signal->autogroup, + * this process can already run with task_group() == prev->tg or we can + * race with cgroup code which can read autogroup = prev under rq->lock. + * In the latter case for_each_thread() can not miss a migrating thread, + * cpu_cgroup_attach() must not be possible after cgroup_exit() and it + * can't be removed from thread list, we hold ->siglock. + */ for_each_thread(p, t) sched_move_task(t); -out: + unlock_task_sighand(p, &flags); autogroup_kref_put(prev); } -- cgit From 8e5bfa8c1f8471aa4a2d30be631ef2b50e10abaf Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 14 Nov 2016 19:46:12 +0100 Subject: sched/autogroup: Do not use autogroup->tg in zombie threads Exactly because for_each_thread() in autogroup_move_group() can't see it and update its ->sched_task_group before _put() and possibly free(). So the exiting task needs another sched_move_task() before exit_notify() and we need to re-introduce the PF_EXITING (or similar) check removed by the previous change for another reason. Signed-off-by: Oleg Nesterov Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: hartsjc@redhat.com Cc: vbendel@redhat.com Cc: vlovejoy@redhat.com Link: http://lkml.kernel.org/r/20161114184612.GA15968@redhat.com Signed-off-by: Ingo Molnar --- kernel/exit.c | 1 + kernel/sched/auto_group.c | 19 +++++++++++++++++++ 2 files changed, 20 insertions(+) (limited to 'kernel') diff --git a/kernel/exit.c b/kernel/exit.c index 9d68c45ebbe3..3076f3089919 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -836,6 +836,7 @@ void __noreturn do_exit(long code) */ perf_event_exit_task(tsk); + sched_autogroup_exit_task(tsk); cgroup_exit(tsk); /* diff --git a/kernel/sched/auto_group.c b/kernel/sched/auto_group.c index ad2b19ad6ca0..f1c8fd566246 100644 --- a/kernel/sched/auto_group.c +++ b/kernel/sched/auto_group.c @@ -115,10 +115,26 @@ bool task_wants_autogroup(struct task_struct *p, struct task_group *tg) * If we race with autogroup_move_group() the caller can use the old * value of signal->autogroup but in this case sched_move_task() will * be called again before autogroup_kref_put(). + * + * However, there is no way sched_autogroup_exit_task() could tell us + * to avoid autogroup->tg, so we abuse PF_EXITING flag for this case. */ + if (p->flags & PF_EXITING) + return false; + return true; } +void sched_autogroup_exit_task(struct task_struct *p) +{ + /* + * We are going to call exit_notify() and autogroup_move_group() can't + * see this thread after that: we can no longer use signal->autogroup. + * See the PF_EXITING check in task_wants_autogroup(). + */ + sched_move_task(p); +} + static void autogroup_move_group(struct task_struct *p, struct autogroup *ag) { @@ -142,6 +158,9 @@ autogroup_move_group(struct task_struct *p, struct autogroup *ag) * In the latter case for_each_thread() can not miss a migrating thread, * cpu_cgroup_attach() must not be possible after cgroup_exit() and it * can't be removed from thread list, we hold ->siglock. + * + * If an exiting thread was already removed from thread list we rely on + * sched_autogroup_exit_task(). */ for_each_thread(p, t) sched_move_task(t); -- cgit