]> git.ipfire.org Git - thirdparty/kernel/stable.git/commitdiff
perf/ring_buffer: Add ordering to rb->nest increment
authorPeter Zijlstra <peterz@infradead.org>
Fri, 17 May 2019 11:52:32 +0000 (13:52 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sat, 22 Jun 2019 06:15:16 +0000 (08:15 +0200)
[ Upstream commit 3f9fbe9bd86c534eba2faf5d840fd44c6049f50e ]

Similar to how decrementing rb->next too early can cause data_head to
(temporarily) be observed to go backward, so too can this happen when
we increment too late.

This barrier() ensures the rb->head load happens after the increment,
both the one in the 'goto again' path, as the one from
perf_output_get_handle() -- albeit very unlikely to matter for the
latter.

Suggested-by: Yabin Cui <yabinc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: acme@kernel.org
Cc: mark.rutland@arm.com
Cc: namhyung@kernel.org
Fixes: ef60777c9abd ("perf: Optimize the perf_output() path by removing IRQ-disables")
Link: http://lkml.kernel.org/r/20190517115418.309516009@infradead.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
kernel/events/ring_buffer.c

index 31edf1f39cca6cb4fe27096193e8a8471a2075f6..d32b9375ec0e85f5e78b9946326f1fb723f8ae06 100644 (file)
@@ -49,6 +49,15 @@ static void perf_output_put_handle(struct perf_output_handle *handle)
        unsigned long head;
 
 again:
+       /*
+        * In order to avoid publishing a head value that goes backwards,
+        * we must ensure the load of @rb->head happens after we've
+        * incremented @rb->nest.
+        *
+        * Otherwise we can observe a @rb->head value before one published
+        * by an IRQ/NMI happening between the load and the increment.
+        */
+       barrier();
        head = local_read(&rb->head);
 
        /*