From 4cd4262034849da01eb88659af677b69f8169f06 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 26 Nov 2008 21:04:24 -0500 Subject: sched: prevent divide by zero error in cpu_avg_load_per_task Impact: fix divide by zero crash in scheduler rebalance irq While testing the branch profiler, I hit this crash: divide error: 0000 [#1] PREEMPT SMP [...] RIP: 0010:[] [] cpu_avg_load_per_task+0x50/0x7f [...] Call Trace: <0> [] find_busiest_group+0x3e5/0xcaa [] rebalance_domains+0x2da/0xa21 [] ? find_next_bit+0x1b2/0x1e6 [] run_rebalance_domains+0x112/0x19f [] __do_softirq+0xa8/0x232 [] call_softirq+0x1c/0x3e [] do_softirq+0x94/0x1cd [] irq_exit+0x6b/0x10e [] smp_apic_timer_interrupt+0xd3/0xff [] apic_timer_interrupt+0x13/0x20 The code for cpu_avg_load_per_task has: if (rq->nr_running) rq->avg_load_per_task = rq->load.weight / rq->nr_running; The runqueue lock is not held here, and there is nothing that prevents the rq->nr_running from going to zero after it passes the if condition. The branch profiler simply made the race window bigger. This patch saves off the rq->nr_running to a local variable and uses that for both the condition and the division. Signed-off-by: Steven Rostedt Peter Zijlstra Signed-off-by: Ingo Molnar --- kernel/sched.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel/sched.c') diff --git a/kernel/sched.c b/kernel/sched.c index 9b1e79371c20..700aa9a1413f 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -1453,9 +1453,10 @@ static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd); static unsigned long cpu_avg_load_per_task(int cpu) { struct rq *rq = cpu_rq(cpu); + unsigned long nr_running = rq->nr_running; - if (rq->nr_running) - rq->avg_load_per_task = rq->load.weight / rq->nr_running; + if (nr_running) + rq->avg_load_per_task = rq->load.weight / nr_running; else rq->avg_load_per_task = 0; -- cgit v1.2.3 From af6d596fd603219b054c1c90fb16672a9fd441bd Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 29 Nov 2008 20:45:15 +0100 Subject: sched: prevent divide by zero error in cpu_avg_load_per_task, update Regarding the bug addressed in: 4cd4262: sched: prevent divide by zero error in cpu_avg_load_per_task Linus points out that the fix is not complete: > There's nothing that keeps gcc from deciding not to reload > rq->nr_running. > > Of course, in _practice_, I don't think gcc ever will (if it decides > that it will spill, gcc is likely going to decide that it will > literally spill the local variable to the stack rather than decide to > reload off the pointer), but it's a valid compiler optimization, and > it even has a name (rematerialization). > > So I suspect that your patch does fix the bug, but it still leaves the > fairly unlikely _potential_ for it to re-appear at some point. > > We have ACCESS_ONCE() as a macro to guarantee that the compiler > doesn't rematerialize a pointer access. That also would clarify > the fact that we access something unsafe outside a lock. So make sure our nr_running value is immutable and cannot change after we check it for nonzero. Signed-off-by: Ingo Molnar --- kernel/sched.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/sched.c') diff --git a/kernel/sched.c b/kernel/sched.c index 700aa9a1413f..b7480fb5c3dc 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -1453,7 +1453,7 @@ static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd); static unsigned long cpu_avg_load_per_task(int cpu) { struct rq *rq = cpu_rq(cpu); - unsigned long nr_running = rq->nr_running; + unsigned long nr_running = ACCESS_ONCE(rq->nr_running); if (nr_running) rq->avg_load_per_task = rq->load.weight / nr_running; -- cgit v1.2.3 From 9a2bd244e18ffbb96c8b783210fda4eded7c7e6f Mon Sep 17 00:00:00 2001 From: Brian King Date: Tue, 9 Dec 2008 08:47:00 -0600 Subject: sched: CPU remove deadlock fix Impact: fix possible deadlock in CPU hot-remove path This patch fixes a possible deadlock scenario in the CPU remove path. migration_call grabs rq->lock, then wakes up everything on rq->migration_queue with the lock held. Then one of the tasks on the migration queue ends up calling tg_shares_up which then also tries to acquire the same rq->lock. [c000000058eab2e0] c000000000502078 ._spin_lock_irqsave+0x98/0xf0 [c000000058eab370] c00000000008011c .tg_shares_up+0x10c/0x20c [c000000058eab430] c00000000007867c .walk_tg_tree+0xc4/0xfc [c000000058eab4d0] c0000000000840c8 .try_to_wake_up+0xb0/0x3c4 [c000000058eab590] c0000000000799a0 .__wake_up_common+0x6c/0xe0 [c000000058eab640] c00000000007ada4 .complete+0x54/0x80 [c000000058eab6e0] c000000000509fa8 .migration_call+0x5fc/0x6f8 [c000000058eab7c0] c000000000504074 .notifier_call_chain+0x68/0xe0 [c000000058eab860] c000000000506568 ._cpu_down+0x2b0/0x3f4 [c000000058eaba60] c000000000506750 .cpu_down+0xa4/0x108 [c000000058eabb10] c000000000507e54 .store_online+0x44/0xa8 [c000000058eabba0] c000000000396260 .sysdev_store+0x3c/0x50 [c000000058eabc10] c0000000001a39b8 .sysfs_write_file+0x124/0x18c [c000000058eabcd0] c00000000013061c .vfs_write+0xd0/0x1bc [c000000058eabd70] c0000000001308a4 .sys_write+0x68/0x114 [c000000058eabe30] c0000000000086b4 syscall_exit+0x0/0x40 Signed-off-by: Brian King Acked-by: Peter Zijlstra Signed-off-by: Ingo Molnar --- kernel/sched.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel/sched.c') diff --git a/kernel/sched.c b/kernel/sched.c index b7480fb5c3dc..e4bb1dd7b308 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -6587,7 +6587,9 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) req = list_entry(rq->migration_queue.next, struct migration_req, list); list_del_init(&req->list); + spin_unlock_irq(&rq->lock); complete(&req->done); + spin_lock_irq(&rq->lock); } spin_unlock_irq(&rq->lock); break; -- cgit v1.2.3