From: Kevin Hao <haokexin@xxxxxxxxx> commit 5cbf2fff3bba8d3c6a4d47c1754de1cf57e2b01f upstream. In the current code, we use the atomic_cmpxchg() to serialize the output of the dump_stack(), but this implementation suffers the thundering herd problem. We have observed such kind of livelock on a Marvell cn96xx board(24 cpus) when heavily using the dump_stack() in a kprobe handler. Actually we can let the competitors to wait for the releasing of the lock before jumping to atomic_cmpxchg(). This will definitely mitigate the thundering herd problem. Thanks Linus for the suggestion. [akpm@xxxxxxxxxxxxxxxxxxxx: fix comment] Link: http://lkml.kernel.org/r/20191030031637.6025-1-haokexin@xxxxxxxxx Fixes: b58d977432c8 ("dump_stack: serialize the output from dump_stack()") Signed-off-by: Kevin Hao <haokexin@xxxxxxxxx> Suggested-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> Cc: <stable@xxxxxxxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- lib/dump_stack.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) --- a/lib/dump_stack.c +++ b/lib/dump_stack.c @@ -106,7 +106,12 @@ retry: was_locked = 1; } else { local_irq_restore(flags); - cpu_relax(); + /* + * Wait for the lock to release before jumping to + * atomic_cmpxchg() in order to mitigate the thundering herd + * problem. + */ + do { cpu_relax(); } while (atomic_read(&dump_lock) != -1); goto retry; }
