Re: [PATCH] mm, compaction: properly signal and act upon lock and need_sched() contention

[Vlastimil Babka <vbabka@suse.cz>]

On 05/12/2014 05:34 PM, Naoya Horiguchi wrote:
> On Mon, May 12, 2014 at 04:15:11PM +0200, Vlastimil Babka wrote:
>> Compaction uses compact_checklock_irqsave() function to periodically check for
>> lock contention and need_resched() to either abort async compaction, or to
>> free the lock, schedule and retake the lock. When aborting, cc->contended is
>> set to signal the contended state to the caller. Two problems have been
>> identified in this mechanism.
>>
>> First, compaction also calls directly cond_resched() in both scanners when no
>> lock is yet taken. This call either does not abort async compaction, or set
>> cc->contended appropriately. This patch introduces a new
>> compact_check_resched() function to achieve both.
>>
>> Second, isolate_freepages() does not check if isolate_freepages_block()
>> aborted due to contention, and advances to the next pageblock. This violates
>> the principle of aborting on contention, and might result in pageblocks not
>> being scanned completely, since the scanning cursor is advanced. This patch
>> makes isolate_freepages_block() check the cc->contended flag and abort.
>>
>> Reported-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>> Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
>> Cc: Minchan Kim <minchan@kernel.org>
>> Cc: Mel Gorman <mgorman@suse.de>
>> Cc: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
>> Cc: Michal Nazarewicz <mina86@mina86.com>
>> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
>> Cc: Christoph Lameter <cl@linux.com>
>> Cc: Rik van Riel <riel@redhat.com>
>> ---
>>   mm/compaction.c | 40 +++++++++++++++++++++++++++++++++-------
>>   1 file changed, 33 insertions(+), 7 deletions(-)
>>
>> diff --git a/mm/compaction.c b/mm/compaction.c
>> index 83ca6f9..b34ab7c 100644
>> --- a/mm/compaction.c
>> +++ b/mm/compaction.c
>> @@ -222,6 +222,27 @@ static bool compact_checklock_irqsave(spinlock_t *lock, unsigned long *flags,
>>   	return true;
>>   }
>>
>> +/*
>> + * Similar to compact_checklock_irqsave() (see its comment) for places where
>> + * a zone lock is not concerned.
>> + *
>> + * Returns false when compaction should abort.
>> + */
>> +static inline bool compact_check_resched(struct compact_control *cc)
>> +{
>> +	/* async compaction aborts if contended */
>> +	if (need_resched()) {
>> +		if (cc->mode == MIGRATE_ASYNC) {
>> +			cc->contended = true;
>
> This changes the meaning of contended in struct compact_control (not just
> indicating lock contention,) so please update the comment in mm/internal.h too.

It doesn't change it, since compact_checklock_irqsave() already has this 
semantic:
if (should_release_lock(lock) && cc->mode == MIGRATE_ASYNC)
	cc->contended = true;

and should_release_lock() is:
	need_resched() || spin_is_contended(lock)

So the comment was already outdated, I will update it in v2.

> Reviewed-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>

Thanks.

> Thanks,
> Naoya Horiguchi
>