From 58763148758057ffc447bf990321d3ea86d199a0 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 30 Aug 2016 10:15:03 +0200 Subject: perf/core: Remove WARN from perf_event_read() This effectively reverts commit: 71e7bc2bab77 ("perf/core: Check return value of the perf_event_read() IPI") ... and puts in a comment explaining why we ignore the return value. Reported-by: Vegard Nossum Signed-off-by: Peter Zijlstra (Intel) Cc: David Carrillo-Cisneros Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Fixes: 71e7bc2bab77 ("perf/core: Check return value of the perf_event_read() IPI") Signed-off-by: Ingo Molnar --- kernel/events/core.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 3cfabdf7b942..07ac8596a728 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -3549,10 +3549,18 @@ static int perf_event_read(struct perf_event *event, bool group) .group = group, .ret = 0, }; - ret = smp_call_function_single(event->oncpu, __perf_event_read, &data, 1); - /* The event must have been read from an online CPU: */ - WARN_ON_ONCE(ret); - ret = ret ? : data.ret; + /* + * Purposely ignore the smp_call_function_single() return + * value. + * + * If event->oncpu isn't a valid CPU it means the event got + * scheduled out and that will have updated the event count. + * + * Therefore, either way, we'll have an up-to-date event count + * after this. + */ + (void)smp_call_function_single(event->oncpu, __perf_event_read, &data, 1); + ret = data.ret; } else if (event->state == PERF_EVENT_STATE_INACTIVE) { struct perf_event_context *ctx = event->ctx; unsigned long flags; -- cgit v1.2.3-71-gd317 From 135e8c9250dd5c8c9aae5984fde6f230d0cbfeaf Mon Sep 17 00:00:00 2001 From: Balbir Singh Date: Mon, 5 Sep 2016 13:16:40 +1000 Subject: sched/core: Fix a race between try_to_wake_up() and a woken up task The origin of the issue I've seen is related to a missing memory barrier between check for task->state and the check for task->on_rq. The task being woken up is already awake from a schedule() and is doing the following: do { schedule() set_current_state(TASK_(UN)INTERRUPTIBLE); } while (!cond); The waker, actually gets stuck doing the following in try_to_wake_up(): while (p->on_cpu) cpu_relax(); Analysis: The instance I've seen involves the following race: CPU1 CPU2 while () { if (cond) break; do { schedule(); set_current_state(TASK_UN..) } while (!cond); wakeup_routine() spin_lock_irqsave(wait_lock) raw_spin_lock_irqsave(wait_lock) wake_up_process() } try_to_wake_up() set_current_state(TASK_RUNNING); .. list_del(&waiter.list); CPU2 wakes up CPU1, but before it can get the wait_lock and set current state to TASK_RUNNING the following occurs: CPU3 wakeup_routine() raw_spin_lock_irqsave(wait_lock) if (!list_empty) wake_up_process() try_to_wake_up() raw_spin_lock_irqsave(p->pi_lock) .. if (p->on_rq && ttwu_wakeup()) .. while (p->on_cpu) cpu_relax() .. CPU3 tries to wake up the task on CPU1 again since it finds it on the wait_queue, CPU1 is spinning on wait_lock, but immediately after CPU2, CPU3 got it. CPU3 checks the state of p on CPU1, it is TASK_UNINTERRUPTIBLE and the task is spinning on the wait_lock. Interestingly since p->on_rq is checked under pi_lock, I've noticed that try_to_wake_up() finds p->on_rq to be 0. This was the most confusing bit of the analysis, but p->on_rq is changed under runqueue lock, rq_lock, the p->on_rq check is not reliable without this fix IMHO. The race is visible (based on the analysis) only when ttwu_queue() does a remote wakeup via ttwu_queue_remote. In which case the p->on_rq change is not done uder the pi_lock. The result is that after a while the entire system locks up on the raw_spin_irqlock_save(wait_lock) and the holder spins infintely Reproduction of the issue: The issue can be reproduced after a long run on my system with 80 threads and having to tweak available memory to very low and running memory stress-ng mmapfork test. It usually takes a long time to reproduce. I am trying to work on a test case that can reproduce the issue faster, but thats work in progress. I am still testing the changes on my still in a loop and the tests seem OK thus far. Big thanks to Benjamin and Nick for helping debug this as well. Ben helped catch the missing barrier, Nick caught every missing bit in my theory. Signed-off-by: Balbir Singh [ Updated comment to clarify matching barriers. Many architectures do not have a full barrier in switch_to() so that cannot be relied upon. ] Signed-off-by: Peter Zijlstra (Intel) Acked-by: Benjamin Herrenschmidt Cc: Alexey Kardashevskiy Cc: Linus Torvalds Cc: Nicholas Piggin Cc: Nicholas Piggin Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Link: http://lkml.kernel.org/r/e02cce7b-d9ca-1ad0-7a61-ea97c7582b37@gmail.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 2a906f20fba7..44817c640e99 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -2016,6 +2016,28 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) success = 1; /* we're going to change ->state */ cpu = task_cpu(p); + /* + * Ensure we load p->on_rq _after_ p->state, otherwise it would + * be possible to, falsely, observe p->on_rq == 0 and get stuck + * in smp_cond_load_acquire() below. + * + * sched_ttwu_pending() try_to_wake_up() + * [S] p->on_rq = 1; [L] P->state + * UNLOCK rq->lock -----. + * \ + * +--- RMB + * schedule() / + * LOCK rq->lock -----' + * UNLOCK rq->lock + * + * [task p] + * [S] p->state = UNINTERRUPTIBLE [L] p->on_rq + * + * Pairs with the UNLOCK+LOCK on rq->lock from the + * last wakeup of our task and the schedule that got our task + * current. + */ + smp_rmb(); if (p->on_rq && ttwu_remote(p, wake_flags)) goto stat; -- cgit v1.2.3-71-gd317 From 767ae08678c2c796bcd7f582ee457aee20a28a1e Mon Sep 17 00:00:00 2001 From: Alexander Shishkin Date: Tue, 6 Sep 2016 16:23:49 +0300 Subject: perf/core: Fix a race between mmap_close() and set_output() of AUX events In the mmap_close() path we need to stop all the AUX events that are writing data to the AUX area that we are unmapping, before we can safely free the pages. To determine if an event needs to be stopped, we're comparing its ->rb against the one that's getting unmapped. However, a SET_OUTPUT ioctl may turn up inside an AUX transaction and swizzle event::rb to some other ring buffer, but the transaction will keep writing data to the old ring buffer until the event gets scheduled out. At this point, mmap_close() will skip over such an event and will proceed to free the AUX area, while it's still being used by this event, which will set off a warning in the mmap_close() path and cause a memory corruption. To avoid this, always stop an AUX event before its ->rb is updated; this will release the (potentially) last reference on the AUX area of the buffer. If the event gets restarted, its new ring buffer will be used. If another SET_OUTPUT comes and switches it back to the old ring buffer that's getting unmapped, it's also fine: this ring buffer's aux_mmap_count will be zero and AUX transactions won't start any more. Reported-by: Vince Weaver Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Arnaldo Carvalho de Melo Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160906132353.19887-2-alexander.shishkin@linux.intel.com Signed-off-by: Ingo Molnar --- kernel/events/core.c | 31 +++++++++++++++++++++++++------ 1 file changed, 25 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 07ac8596a728..a54f2c2cdb20 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -2496,11 +2496,11 @@ static int __perf_event_stop(void *info) return 0; } -static int perf_event_restart(struct perf_event *event) +static int perf_event_stop(struct perf_event *event, int restart) { struct stop_event_data sd = { .event = event, - .restart = 1, + .restart = restart, }; int ret = 0; @@ -4845,6 +4845,19 @@ static void ring_buffer_attach(struct perf_event *event, spin_unlock_irqrestore(&rb->event_lock, flags); } + /* + * Avoid racing with perf_mmap_close(AUX): stop the event + * before swizzling the event::rb pointer; if it's getting + * unmapped, its aux_mmap_count will be 0 and it won't + * restart. See the comment in __perf_pmu_output_stop(). + * + * Data will inevitably be lost when set_output is done in + * mid-air, but then again, whoever does it like this is + * not in for the data anyway. + */ + if (has_aux(event)) + perf_event_stop(event, 0); + rcu_assign_pointer(event->rb, rb); if (old_rb) { @@ -6120,7 +6133,7 @@ static void perf_event_addr_filters_exec(struct perf_event *event, void *data) raw_spin_unlock_irqrestore(&ifh->lock, flags); if (restart) - perf_event_restart(event); + perf_event_stop(event, 1); } void perf_event_exec(void) @@ -6164,7 +6177,13 @@ static void __perf_event_output_stop(struct perf_event *event, void *data) /* * In case of inheritance, it will be the parent that links to the - * ring-buffer, but it will be the child that's actually using it: + * ring-buffer, but it will be the child that's actually using it. + * + * We are using event::rb to determine if the event should be stopped, + * however this may race with ring_buffer_attach() (through set_output), + * which will make us skip the event that actually needs to be stopped. + * So ring_buffer_attach() has to stop an aux event before re-assigning + * its rb pointer. */ if (rcu_dereference(parent->rb) == rb) ro->err = __perf_event_stop(&sd); @@ -6678,7 +6697,7 @@ static void __perf_addr_filters_adjust(struct perf_event *event, void *data) raw_spin_unlock_irqrestore(&ifh->lock, flags); if (restart) - perf_event_restart(event); + perf_event_stop(event, 1); } /* @@ -7867,7 +7886,7 @@ static void perf_event_addr_filters_apply(struct perf_event *event) mmput(mm); restart: - perf_event_restart(event); + perf_event_stop(event, 1); } /* -- cgit v1.2.3-71-gd317 From b79ccadd6bb10e72cf784a298ca6dc1398eb9a24 Mon Sep 17 00:00:00 2001 From: Alexander Shishkin Date: Tue, 6 Sep 2016 16:23:50 +0300 Subject: perf/core: Fix aux_mmap_count vs aux_refcount order The order of accesses to ring buffer's aux_mmap_count and aux_refcount has to be preserved across the users, namely perf_mmap_close() and perf_aux_output_begin(), otherwise the inversion can result in the latter holding the last reference to the aux buffer and subsequently free'ing it in atomic context, triggering a warning. > ------------[ cut here ]------------ > WARNING: CPU: 0 PID: 257 at kernel/events/ring_buffer.c:541 __rb_free_aux+0x11a/0x130 > CPU: 0 PID: 257 Comm: stopbug Not tainted 4.8.0-rc1+ #2596 > Call Trace: > [] __warn+0xcb/0xf0 > [] warn_slowpath_null+0x1d/0x20 > [] __rb_free_aux+0x11a/0x130 > [] rb_free_aux+0x18/0x20 > [] perf_aux_output_begin+0x163/0x1e0 > [] bts_event_start+0x3a/0xd0 > [] bts_event_add+0x5d/0x80 > [] event_sched_in.isra.104+0xf6/0x2f0 > [] group_sched_in+0x6e/0x190 > [] ctx_sched_in+0x2fe/0x5f0 > [] perf_event_sched_in+0x60/0x80 > [] ctx_resched+0x5b/0x90 > [] __perf_event_enable+0x1e1/0x240 > [] event_function+0xa9/0x180 > [] ? perf_cgroup_attach+0x70/0x70 > [] remote_function+0x3f/0x50 > [] flush_smp_call_function_queue+0x83/0x150 > [] generic_smp_call_function_single_interrupt+0x13/0x60 > [] smp_call_function_single_interrupt+0x27/0x40 > [] call_function_single_interrupt+0x89/0x90 > [] finish_task_switch+0xa6/0x210 > [] ? finish_task_switch+0x67/0x210 > [] __schedule+0x3dd/0xb50 > [] schedule+0x35/0x80 > [] sys_sched_yield+0x61/0x70 > [] entry_SYSCALL_64_fastpath+0x18/0xa8 > ---[ end trace 6235f556f5ea83a9 ]--- This patch puts the checks in perf_aux_output_begin() in the same order as that of perf_mmap_close(). Reported-by: Vince Weaver Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Arnaldo Carvalho de Melo Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: vince@deater.net Link: http://lkml.kernel.org/r/20160906132353.19887-3-alexander.shishkin@linux.intel.com Signed-off-by: Ingo Molnar --- kernel/events/ring_buffer.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c index ae9b90dc9a5a..257fa460b846 100644 --- a/kernel/events/ring_buffer.c +++ b/kernel/events/ring_buffer.c @@ -330,15 +330,22 @@ void *perf_aux_output_begin(struct perf_output_handle *handle, if (!rb) return NULL; - if (!rb_has_aux(rb) || !atomic_inc_not_zero(&rb->aux_refcount)) + if (!rb_has_aux(rb)) goto err; /* - * If rb::aux_mmap_count is zero (and rb_has_aux() above went through), - * the aux buffer is in perf_mmap_close(), about to get freed. + * If aux_mmap_count is zero, the aux buffer is in perf_mmap_close(), + * about to get freed, so we leave immediately. + * + * Checking rb::aux_mmap_count and rb::refcount has to be done in + * the same order, see perf_mmap_close. Otherwise we end up freeing + * aux pages in this path, which is a bug, because in_atomic(). */ if (!atomic_read(&rb->aux_mmap_count)) - goto err_put; + goto err; + + if (!atomic_inc_not_zero(&rb->aux_refcount)) + goto err; /* * Nesting is not supported for AUX area, make sure nested -- cgit v1.2.3-71-gd317 From d979a39d7242e0601bf9b60e89628fb8ac577179 Mon Sep 17 00:00:00 2001 From: Johannes Weiner Date: Mon, 19 Sep 2016 14:44:38 -0700 Subject: cgroup: duplicate cgroup reference when cloning sockets When a socket is cloned, the associated sock_cgroup_data is duplicated but not its reference on the cgroup. As a result, the cgroup reference count will underflow when both sockets are destroyed later on. Fixes: bd1060a1d671 ("sock, cgroup: add sock->sk_cgroup") Link: http://lkml.kernel.org/r/20160914194846.11153-2-hannes@cmpxchg.org Signed-off-by: Johannes Weiner Acked-by: Tejun Heo Cc: Michal Hocko Cc: Vladimir Davydov Cc: [4.5+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/cgroup.c | 6 ++++++ net/core/sock.c | 5 ++++- 2 files changed, 10 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/cgroup.c b/kernel/cgroup.c index d1c51b7f5221..5e8dab5bf9ad 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -6270,6 +6270,12 @@ void cgroup_sk_alloc(struct sock_cgroup_data *skcd) if (cgroup_sk_alloc_disabled) return; + /* Socket clone path */ + if (skcd->val) { + cgroup_get(sock_cgroup_ptr(skcd)); + return; + } + rcu_read_lock(); while (true) { diff --git a/net/core/sock.c b/net/core/sock.c index 25dab8b60223..fd7b41edf1ce 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1362,7 +1362,6 @@ static struct sock *sk_prot_alloc(struct proto *prot, gfp_t priority, if (!try_module_get(prot->owner)) goto out_free_sec; sk_tx_queue_clear(sk); - cgroup_sk_alloc(&sk->sk_cgrp_data); } return sk; @@ -1422,6 +1421,7 @@ struct sock *sk_alloc(struct net *net, int family, gfp_t priority, sock_net_set(sk, net); atomic_set(&sk->sk_wmem_alloc, 1); + cgroup_sk_alloc(&sk->sk_cgrp_data); sock_update_classid(&sk->sk_cgrp_data); sock_update_netprioidx(&sk->sk_cgrp_data); } @@ -1566,6 +1566,9 @@ struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) newsk->sk_priority = 0; newsk->sk_incoming_cpu = raw_smp_processor_id(); atomic64_set(&newsk->sk_cookie, 0); + + cgroup_sk_alloc(&newsk->sk_cgrp_data); + /* * Before updating sk_refcnt, we must commit prior changes to memory * (Documentation/RCU/rculist_nulls.txt for details) -- cgit v1.2.3-71-gd317