Re: [PATCH] mm/page_alloc: fix deferred compaction accounting

[Vlastimil Babka (SUSE)]

On 5/26/26 11:12, fujunjie wrote:
> COMPACT_DEFERRED means compaction did not start because past failures
> caused the zone to be deferred. try_to_compact_pages() returns the
> maximum result seen while walking the zonelist, so a final
> COMPACT_DEFERRED result means no later zone reported that compaction
> actually ran.
> 
> __alloc_pages_direct_compact() skips COMPACTSTALL and COMPACTFAIL
> accounting when try_to_compact_pages() returns COMPACT_SKIPPED, but not
> when it returns COMPACT_DEFERRED. A deferred-only direct compaction
> attempt can therefore look like a stall, and then a failure if the
> allocation still cannot be satisfied.
> 
> Treat COMPACT_DEFERRED like COMPACT_SKIPPED in this accounting path. If
> a later zone runs compaction and returns a result above COMPACT_DEFERRED,
> or compact_zone_order() reports COMPACT_SUCCESS for a captured page, the
> final result is not COMPACT_DEFERRED and the existing accounting still
> runs.
> 
> Signed-off-by: fujunjie <fujunjie1@xxxxxx>

Hm I wonder why didn't we do that right away in 06dac2f467fe ("mm:
compaction: update the COMPACT[STALL|FAIL] events properly"), seems like
even back then there was nothing that would e.g. convert _DEFERED to
_SKIPPED from try_to_compact_pages(), and the changelog did mention deferred.

So I guess for posterity we could do (but no Cc: stable, it's just stats).

Fixes: 06dac2f467fe ("mm: compaction: update the COMPACT[STALL|FAIL] events
properly")

Reviewed-by: Vlastimil Babka (SUSE) <vbabka@xxxxxxxxxx>

Thanks!

> ---
>  mm/page_alloc.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 23c7298d3be2..ddeb79fa12db 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -4161,7 +4161,8 @@ __alloc_pages_direct_compact(gfp_t gfp_mask, unsigned int order,
>  	psi_memstall_leave(&pflags);
>  	delayacct_compact_end();
>  
> -	if (*compact_result == COMPACT_SKIPPED)
> +	if (*compact_result == COMPACT_SKIPPED ||
> +	    *compact_result == COMPACT_DEFERRED)
>  		return NULL;
>  	/*
>  	 * At least in one zone compaction wasn't deferred or skipped, so let's
> 
> base-commit: e8c2f9fdadee7cbc75134dc463c1e0d856d6e5c7