Re: [PATCH] [RFC] mm/compaction: initialize compaction information

[Gioh Kim <gioh.kim@lge.com>]

2015-03-19 i??i?? 6:01i?? Vlastimil Babka i?'(e??) i?' e,?:
> On 03/19/2015 09:52 AM, Gioh Kim wrote:
>>
>>
>> 2015-03-19 i??i?? 5:41i?? Vlastimil Babka i?'(e??) i?' e,?:
>>> On 03/19/2015 06:30 AM, Gioh Kim wrote:
>>>
>>> The code below this comment already does the initialization if the cached values
>>> are outside zone boundaries (e.g. due to not being initialized). So if I go
>>> through what your __reset_isolation_suitable(zone) call possibly fixes:
>>>
>>> - the code below comment should take care of zone->compact_cached_migrate_pfn
>>> and zone->compact_cached_free_pfn.
>>> - the value of zone->compact_blockskip_flush shouldn't affect whether compaction
>>> is done.
>>> - the state of pageblock_skip bits shouldn't matter for compaction via
>>> /proc/sys... as that sets ignore_skip_hint = true
>>>
>>> It might be perhaps possible that the cached scanner positions are close to
>>> meeting and compaction occurs but doesn't process much. That would be also true
>>> if both were zero, but at least on my x86 system, lowest zone's start_pfn is 1
>>> so that would be detected and corrected. Maybe it is zero on yours though? (ARM?).
>>
>> YES, it is. As comment above, my platform is based on ARM.
>
> Ah, I see.
>
>> zone's start_pfn is 0.
>
> OK, good to know that's possible. In that case it's clear that the proper
> initialization doesn't happen, and __compact_finished() decides that scanners
> have already met at pfn 0.
>
>>>
>>> So in any case, the problem should be identified in more detail so we know the
>>> fix is not accidental. It could be also worthwile to always reset scanner
>>> positions when doing a /proc triggered compaction, so it's not depending on what
>>> happened before.
>>>
>>
>> Excuse my poor english.
>> I cannot catch exactly what you want.
>> Is this what you want? This resets the position if compaction is started via /proc.
>
> Yes that's right, but..
>
>> diff --git a/mm/compaction.c b/mm/compaction.c
>> index 8c0d945..827ec06 100644
>> --- a/mm/compaction.c
>> +++ b/mm/compaction.c
>> @@ -1587,8 +1587,10 @@ static void __compact_pgdat(pg_data_t *pgdat, struct compact_control *cc)
>>                   INIT_LIST_HEAD(&cc->freepages);
>>                   INIT_LIST_HEAD(&cc->migratepages);
>>
>> -               if (cc->order == -1 || !compaction_deferred(zone, cc->order))
>> +               if (cc->order == -1 || !compaction_deferred(zone, cc->order)) {
>> +                       __reset_isolation_suitable(zone);
>
> This will also trigger reset when called from kswapd through compact_pgdat() and
> !compaction_deferred() is true.
> The reset should be restricted to cc->order == -1 which only happens from /proc
> trigger.
>
>>                           compact_zone(zone, cc);
>> +               }
>>
>>                   if (cc->order > 0) {
>>                           if (zone_watermark_ok(zone, cc->order,
>>
>
>

I've not been familiar with compaction code.
I think cc->order is -1 only if __compact_pgdat is called via /proc.
This is ugly but I don't have better solution.
Do you have better idea?


diff --git a/mm/compaction.c b/mm/compaction.c
index 8c0d945..5b4e255 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -1587,6 +1587,9 @@ static void __compact_pgdat(pg_data_t *pgdat, struct compact_control *cc)
                 INIT_LIST_HEAD(&cc->freepages);
                 INIT_LIST_HEAD(&cc->migratepages);

+               if (cc->order == -1)
+                       __reset_isolation_suitable(zone);
+
                 if (cc->order == -1 || !compaction_deferred(zone, cc->order))
                         compact_zone(zone, cc);



--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>