On Tue 16-12-14 21:13:02, Johannes Weiner wrote:
[...]
> From 45362d1920340716ef58bf1024d9674b5dfa809e Mon Sep 17 00:00:00 2001
> From: Johannes Weiner <hannes@xxxxxxxxxxx>
> Date: Tue, 16 Dec 2014 21:04:24 -0500
> Subject: [patch] mm: page_alloc: embed OOM killing naturally into allocation
> slowpath fix
>
> When retrying the allocation after potentially invoking OOM, make sure
> the alloc flags are recalculated, as they have to consider TIF_MEMDIE.
>
> Restore the original restart label, but rename it to 'retry' to match
> the should_alloc_retry() it depends on.
>
> Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx>
Missed that too. Well spotted!
Acked-by: Michal Hocko <mhocko@xxxxxxx>
> ---
> mm/page_alloc.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 83ec725aec36..e8f5997c557c 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -2673,6 +2673,7 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
> (gfp_mask & GFP_THISNODE) == GFP_THISNODE)
> goto nopage;
>
> +retry:
> if (!(gfp_mask & __GFP_NO_KSWAPD))
> wake_all_kswapds(order, zonelist, high_zoneidx,
> preferred_zone, nodemask);
> @@ -2695,7 +2696,6 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
> classzone_idx = zonelist_zone_idx(preferred_zoneref);
> }
>
> -rebalance:
> /* This is the last chance, in general, before the goto nopage. */
> page = get_page_from_freelist(gfp_mask, nodemask, order, zonelist,
> high_zoneidx, alloc_flags & ~ALLOC_NO_WATERMARKS,
> @@ -2823,7 +2823,7 @@ rebalance:
> }
> /* Wait for some write requests to complete then retry */
> wait_iff_congested(preferred_zone, BLK_RW_ASYNC, HZ/50);
> - goto rebalance;
> + goto retry;
> } else {
> /*
> * High-order allocations do not necessarily loop after
> --
> 2.1.3
>
> --
> 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>
--
Michal Hocko
SUSE Labs
--
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>