On Tue, Apr 11, 2017 at 08:27:15PM +0900, Tetsuo Handa wrote:
> We are using warn_alloc() for reporting both allocation failures and
> allocation stalls. If we add debug_guardpage_minorder=1 parameter,
> all allocation failure and allocation stall reports become pointless
> like below. (Below output would be an OOM livelock were all __GFP_FS
> allocations got stuck at too_many_isolated() in shrink_inactive_list()
> waiting for kswapd, kswapd is waiting for !__GFP_FS allocations, and
> all !__GFP_FS allocations did not get stuck at too_many_isolated() in
> shrink_inactive_list() but are unable to invoke the OOM killer.)
>
> ----------
> [ 0.000000] Linux version 4.11.0-rc6-next-20170410 (root@ccsecurity) (gcc version 4.8.5 20150623 (Red Hat 4.8.5-11) (GCC) ) #578 SMP Mon Apr 10 23:08:53 JST 2017
> [ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-4.11.0-rc6-next-20170410 (...snipped...) debug_guardpage_minorder=1
> (...snipped...)
> [ 0.000000] Setting debug_guardpage_minorder to 1
> (...snipped...)
> [ 99.064207] Out of memory: Kill process 3097 (a.out) score 999 or sacrifice child
> [ 99.066488] Killed process 3097 (a.out) total-vm:14408kB, anon-rss:84kB, file-rss:36kB, shmem-rss:0kB
> [ 99.180378] oom_reaper: reaped process 3097 (a.out), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
> [ 128.310487] warn_alloc: 266 callbacks suppressed
> [ 133.445395] warn_alloc: 74 callbacks suppressed
> [ 138.517471] warn_alloc: 300 callbacks suppressed
> [ 143.537630] warn_alloc: 34 callbacks suppressed
> [ 148.610773] warn_alloc: 277 callbacks suppressed
> [ 153.630652] warn_alloc: 70 callbacks suppressed
> [ 158.639891] warn_alloc: 217 callbacks suppressed
> [ 163.687727] warn_alloc: 120 callbacks suppressed
> [ 168.709610] warn_alloc: 252 callbacks suppressed
> [ 173.714659] warn_alloc: 103 callbacks suppressed
> [ 178.730858] warn_alloc: 248 callbacks suppressed
> [ 183.797587] warn_alloc: 82 callbacks suppressed
> [ 188.825250] warn_alloc: 238 callbacks suppressed
> [ 193.832834] warn_alloc: 102 callbacks suppressed
> [ 198.876409] warn_alloc: 259 callbacks suppressed
> [ 203.940073] warn_alloc: 102 callbacks suppressed
> [ 207.620979] sysrq: SysRq : Resetting
> ----------
>
> Commit c0a32fc5a2e470d0 ("mm: more intensive memory corruption debugging")
> changed to check debug_guardpage_minorder() > 0 when reporting allocation
> failures. But the patch description seems to lack why we want to check it.
When we use guard page to debug memory corruption, it shrinks available
pages to 1/2, 1/4, 1/8 and so on, depending on parameter value.
In such case memory allocation failures can be common and printing
errors can flood dmesg. If sombody debug corruption, allocation
failures are not the things he/she is interested about.
> Let's remove that check so that administrators can get some clue by
> allowing warn_alloc() to report e.g. GFP_NOFS | __GFP_NOWARN allocations
> are stalling.
This is ok for me, but perhaps move debug_guardpage_minorder() > 0
check before calling warn_alloc() in buddy allocator when it fails,
or move it before __ratelimit(), will be better option.
Thanks
Stanislaw
> Signed-off-by: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx>
> Cc: Stanislaw Gruszka <sgruszka@xxxxxxxxxx>
> Cc: Mel Gorman <mgorman@xxxxxxx>
> Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
> Cc: "Rafael J. Wysocki" <rjw@xxxxxxx>
> Cc: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
> Cc: Pekka Enberg <penberg@xxxxxxxxxxxxxx>
> ---
> mm/page_alloc.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 32b31d6..5c8cf2a 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -3154,8 +3154,7 @@ void warn_alloc(gfp_t gfp_mask, nodemask_t *nodemask, const char *fmt, ...)
> static DEFINE_RATELIMIT_STATE(nopage_rs, DEFAULT_RATELIMIT_INTERVAL,
> DEFAULT_RATELIMIT_BURST);
>
> - if ((gfp_mask & __GFP_NOWARN) || !__ratelimit(&nopage_rs) ||
> - debug_guardpage_minorder() > 0)
> + if ((gfp_mask & __GFP_NOWARN) || !__ratelimit(&nopage_rs))
> return;
>
> pr_warn("%s: ", current->comm);
> --
> 1.8.3.1
>
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>
