From: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
To: Minchan Kim <minchan@kernel.org>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>,
Andrew Morton <akpm@linux-foundation.org>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Subject: Re: [RFC][PATCHv2 8/8] zsmalloc: register a shrinker to trigger auto-compaction
Date: Thu, 18 Jun 2015 11:41:07 +0900 [thread overview]
Message-ID: <20150618023906.GC3422@swordfish> (raw)
In-Reply-To: <20150618015028.GA2370@bgram>
Hi,
On (06/18/15 10:50), Minchan Kim wrote:
[..]
> > hm, what's the difference with the existing implementation?
> > The 'new one' aborts when (a) !zs_can_compact() and (b) !migrate_zspage().
> > It holds the class lock less time than current compaction.
>
> At old, it unlocks periodically(ie, per-zspage migration) so other who
> want to allocate a zspage in the class can have a chance but your patch
> increases lock holding time until all of zspages in the class is done
> so other will be blocked until all of zspage migration in the class is
> done.
ah, I see.
it doesn't hold the lock `until all the pages are done`. it holds it
as long as zs_can_compact() returns > 0. hm, I'm not entirely sure that
this patch set has increased the locking time (in average).
> >
> > > I will review remain parts tomorrow(I hope) but what I want to say
> > > before going sleep is:
> > >
> > > I like the idea but still have a concern to lack of fragmented zspages
> > > during memory pressure because auto-compaction will prevent fragment
> > > most of time. Surely, using fragment space as buffer in heavy memory
> > > pressure is not intened design so it could be fragile but I'm afraid
> > > this feature might accelrate it and it ends up having a problem and
> > > change current behavior in zram as swap.
> >
> > Well, it's nearly impossible to prove anything with the numbers obtained
> > during some particular case. I agree that fragmentation can be both
> > 'good' (depending on IO pattern) and 'bad'.
>
> Yes, it's not easy and I believe a few artificial testing are not enough
> to prove no regression but we don't have any choice.
> Actually, I think this patchset does make sense. Although it might have
> a problem on situation heavy memory pressure by lacking of fragment space,
I tested exactly this scenario yesterday (and sent an email). We leave `no holes'
in classes only in ~1.35% of cases. so, no, this argument is not valid. we preserve
fragmentation.
-ss
> I think we should go with this patchset and fix the problem with another way
> (e,g. memory pooling rather than relying on the luck of fragment).
> But I need something to take the risk. That's why I ask the number
> although it's not complete. It can cover a case at least, it is better than
> none. :)
>
> >
> >
> > Auto-compaction of IDLE zram devices certainly makes sense, when system
> > is getting low on memory. zram devices are not always 'busy', serving
> > heavy IO. There may be N idle zram devices simply sitting and wasting
> > memory; or being 'moderately' busy; so compaction will not cause any
> > significant slow down there.
> >
> > Auto-compaction of BUSY zram devices is less `desired', of course;
> > but not entirely terrible I think (zs_can_compact() can help here a
> > lot).
>
> My concern is not a compacion overhead but higher memory footprint
> consumed by zram in reserved memory.
> It might hang system if zram used up reserved memory of system with
> ALLOC_NO_WATERMARKS. With auto-compaction, userspace has a higher chance
> to use more memory with uncompressible pages or file-backed pages
> so zram-swap can use more reserved memory. We need to evaluate it, I think.
>
WARNING: multiple messages have this Message-ID (diff)
From: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
To: Minchan Kim <minchan@kernel.org>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>,
Andrew Morton <akpm@linux-foundation.org>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Subject: Re: [RFC][PATCHv2 8/8] zsmalloc: register a shrinker to trigger auto-compaction
Date: Thu, 18 Jun 2015 11:41:07 +0900 [thread overview]
Message-ID: <20150618023906.GC3422@swordfish> (raw)
In-Reply-To: <20150618015028.GA2370@bgram>
Hi,
On (06/18/15 10:50), Minchan Kim wrote:
[..]
> > hm, what's the difference with the existing implementation?
> > The 'new one' aborts when (a) !zs_can_compact() and (b) !migrate_zspage().
> > It holds the class lock less time than current compaction.
>
> At old, it unlocks periodically(ie, per-zspage migration) so other who
> want to allocate a zspage in the class can have a chance but your patch
> increases lock holding time until all of zspages in the class is done
> so other will be blocked until all of zspage migration in the class is
> done.
ah, I see.
it doesn't hold the lock `until all the pages are done`. it holds it
as long as zs_can_compact() returns > 0. hm, I'm not entirely sure that
this patch set has increased the locking time (in average).
> >
> > > I will review remain parts tomorrow(I hope) but what I want to say
> > > before going sleep is:
> > >
> > > I like the idea but still have a concern to lack of fragmented zspages
> > > during memory pressure because auto-compaction will prevent fragment
> > > most of time. Surely, using fragment space as buffer in heavy memory
> > > pressure is not intened design so it could be fragile but I'm afraid
> > > this feature might accelrate it and it ends up having a problem and
> > > change current behavior in zram as swap.
> >
> > Well, it's nearly impossible to prove anything with the numbers obtained
> > during some particular case. I agree that fragmentation can be both
> > 'good' (depending on IO pattern) and 'bad'.
>
> Yes, it's not easy and I believe a few artificial testing are not enough
> to prove no regression but we don't have any choice.
> Actually, I think this patchset does make sense. Although it might have
> a problem on situation heavy memory pressure by lacking of fragment space,
I tested exactly this scenario yesterday (and sent an email). We leave `no holes'
in classes only in ~1.35% of cases. so, no, this argument is not valid. we preserve
fragmentation.
-ss
> I think we should go with this patchset and fix the problem with another way
> (e,g. memory pooling rather than relying on the luck of fragment).
> But I need something to take the risk. That's why I ask the number
> although it's not complete. It can cover a case at least, it is better than
> none. :)
>
> >
> >
> > Auto-compaction of IDLE zram devices certainly makes sense, when system
> > is getting low on memory. zram devices are not always 'busy', serving
> > heavy IO. There may be N idle zram devices simply sitting and wasting
> > memory; or being 'moderately' busy; so compaction will not cause any
> > significant slow down there.
> >
> > Auto-compaction of BUSY zram devices is less `desired', of course;
> > but not entirely terrible I think (zs_can_compact() can help here a
> > lot).
>
> My concern is not a compacion overhead but higher memory footprint
> consumed by zram in reserved memory.
> It might hang system if zram used up reserved memory of system with
> ALLOC_NO_WATERMARKS. With auto-compaction, userspace has a higher chance
> to use more memory with uncompressible pages or file-backed pages
> so zram-swap can use more reserved memory. We need to evaluate it, I think.
>
--
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>
next prev parent reply other threads:[~2015-06-18 2:41 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-06-05 12:03 [RFC][PATCHv2 0/8] introduce automatic pool compaction Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-05 12:03 ` [RFC][PATCHv2 1/8] zsmalloc: drop unused variable `nr_to_migrate' Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-05 12:03 ` [RFC][PATCHv2 2/8] zsmalloc: partial page ordering within a fullness_list Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-16 13:19 ` Minchan Kim
2015-06-16 13:19 ` Minchan Kim
2015-06-16 14:30 ` Sergey Senozhatsky
2015-06-16 14:30 ` Sergey Senozhatsky
2015-06-05 12:03 ` [RFC][PATCHv2 3/8] zsmalloc: lower ZS_ALMOST_FULL waterline Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-16 13:37 ` Minchan Kim
2015-06-16 13:37 ` Minchan Kim
2015-06-16 14:35 ` Sergey Senozhatsky
2015-06-16 14:35 ` Sergey Senozhatsky
2015-06-05 12:03 ` [RFC][PATCHv2 4/8] zsmalloc: always keep per-class stats Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-05 12:03 ` [RFC][PATCHv2 5/8] zsmalloc: introduce zs_can_compact() function Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-16 14:19 ` Minchan Kim
2015-06-16 14:19 ` Minchan Kim
2015-06-16 14:41 ` Sergey Senozhatsky
2015-06-16 14:41 ` Sergey Senozhatsky
2015-06-05 12:03 ` [RFC][PATCHv2 6/8] zsmalloc: cosmetic compaction code adjustments Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-05 12:03 ` [RFC][PATCHv2 7/8] zsmalloc/zram: move `num_migrated' to zs_pool Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-05 12:03 ` [RFC][PATCHv2 8/8] zsmalloc: register a shrinker to trigger auto-compaction Sergey Senozhatsky
2015-06-05 12:03 ` Sergey Senozhatsky
2015-06-16 14:47 ` Minchan Kim
2015-06-16 14:47 ` Minchan Kim
2015-06-16 15:45 ` Sergey Senozhatsky
2015-06-16 15:45 ` Sergey Senozhatsky
2015-06-18 1:50 ` Minchan Kim
2015-06-18 1:50 ` Minchan Kim
2015-06-18 2:41 ` Sergey Senozhatsky [this message]
2015-06-18 2:41 ` Sergey Senozhatsky
2015-06-18 3:01 ` Sergey Senozhatsky
2015-06-18 3:01 ` Sergey Senozhatsky
2015-06-18 3:46 ` Minchan Kim
2015-06-18 3:46 ` Minchan Kim
2015-06-18 3:39 ` Minchan Kim
2015-06-18 3:39 ` Minchan Kim
2015-06-18 3:58 ` Sergey Senozhatsky
2015-06-18 3:58 ` Sergey Senozhatsky
2015-06-17 7:11 ` Sergey Senozhatsky
2015-06-17 7:11 ` Sergey Senozhatsky
2015-06-10 0:04 ` [RFC][PATCHv2 0/8] introduce automatic pool compaction Minchan Kim
2015-06-10 0:04 ` Minchan Kim
2015-06-10 0:07 ` Sergey Senozhatsky
2015-06-10 0:07 ` Sergey Senozhatsky
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20150618023906.GC3422@swordfish \
--to=sergey.senozhatsky.work@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=minchan@kernel.org \
--cc=sergey.senozhatsky@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.