Message-ID: <52148730.4000709@oracle.com>
Date:	Wed, 21 Aug 2013 17:24:00 +0800
From:	Bob Liu <bob.liu@...cle.com>
To:	Minchan Kim <minchan@...nel.org>
CC:	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Jens Axboe <axboe@...nel.dk>,
	Seth Jennings <sjenning@...ux.vnet.ibm.com>,
	Nitin Gupta <ngupta@...are.org>,
	Konrad Rzeszutek Wilk <konrad@...nok.org>,
	Luigi Semenzato <semenzato@...gle.com>,
	linux-kernel@...r.kernel.org, linux-mm@...ck.org,
	Pekka Enberg <penberg@...helsinki.fi>,
	Mel Gorman <mgorman@...e.de>, lliubbo@...il.com
Subject: Re: [PATCH v7 0/5] zram/zsmalloc promotion

Hi Minchan,

On 08/21/2013 02:16 PM, Minchan Kim wrote:
> It's 7th trial of zram/zsmalloc promotion.
> I rewrote cover-letter totally based on previous discussion.
> 
> The main reason to prevent zram promotion was no review of
> zsmalloc part while Jens, block maintainer, already acked
> zram part.
> 
> At that time, zsmalloc was used for zram, zcache and zswap so
> everybody wanted to make it general and at last, Mel reviewed it
> when zswap was submitted to merge mainline a few month ago.
> Most of review was related to zswap writeback mechanism which
> can pageout compressed page in memory into real swap storage
> in runtime and the conclusion was that zsmalloc isn't good for
> zswap writeback so zswap borrowed zbud allocator from zcache to
> replace zsmalloc. The zbud is bad for memory compression ratio(2)
> but it's very predictable behavior because we can expect a zpage
> includes just two pages as maximum. Other reviews were not major. 
> http://lkml.indiana.edu/hypermail/linux/kernel/1304.1/04334.html
> 
> Zcache doesn't use zsmalloc either so zsmalloc's user is only
> zram now so this patchset moves it into zsmalloc directory.
> Recently, Bob tried to move zsmalloc under mm directory to unify
> zram and zswap with adding pseudo block device in zswap(It's
> very weired to me) but he was simple ignoring zram's block device
> (a.k.a zram-blk) feature and considered only swap usecase of zram,
> in turn, it lose zram's good concept.
> 

Yes, I didn't notice the feature that zram can be used as a normal block
device.


> Mel raised an another issue in v6, "maintainance headache".
> He claimed zswap and zram has a similar goal that is to compresss
> swap pages so if we promote zram, maintainance headache happens
> sometime by diverging implementaion between zswap and zram
> so that he want to unify zram and zswap. For it, he want zswap
> to implement pseudo block device like Bob did to emulate zram so
> zswap can have an advantage of writeback as well as zram's benefit.

If consider zram as a swap device only, I still think it's better to add
a pseudo block device to zswap and just disable the writeback of zswap.

But I have no idea of zram's block device feature.

> But I wonder frontswap-based zswap's writeback is really good
> approach for writeback POV. I think that problem isn't only
> specific for zswap. If we want to configure multiple swap hierarchy
> with various speed device such as RAM, NVRAM, SSD, eMMC, NAS etc,
> it would be a general problem. So we should think of more general
> approach. At a glance, I can see two approach.
> 
> First, VM could be aware of heterogeneous swap configuration
> so it could aim for being able to configure cache hierarchy
> among swap devices. It may need indirction layer on swap, which
> was already talked about that way so VM can migrate a block from 
> A to B easily. It will support various configuration with VM's
> hints, maybe, in future.
> http://lkml.indiana.edu/hypermail/linux/kernel/1203.3/03812.html
> 
> Second, as more practical solution, we could use device mapper like
> dm-cache(https://lwn.net/Articles/540996/), which makes it very
> flexible. Now, it supports various configruation and cache policy
> (block size, writeback/writethrough, LRU, MFU although MQ is merged
> now) so it would be good fit for our purpose. Even, it can make zram
> support writeback. I tested it following as following scenario
> in KVM 4 CPU, 1G DRAM with background 800M memory hogger, which is
> allocates random data up to 800M.
> 
> 1) zram swap disk 1G, untar kernel.tgz to tmpfs, build -j 4
>    Fail to untar due to shortage of memory space by tmpfs default size limit
> 
> 2) zram swap disk 1G, untar kernel.tgz to ext2 on zram-blk, build -j 4
>    OOM happens while building the kernel but it untar successfully
>    on ext2 based on zram-blk. The reason OOM happend is zram can not find
>    free pages from main memory to store swap out pages although empty
>    swap space is still enough.
> 
> 3) dm-cache swap disk 1G, untar kernel.tgz to ext2 on zram-blk, build -j 4
>    dmcache consists of zram-meta 10M, zram-cache 1G and real swap storage 1G
>    No OOM happens and successfully building done.
> 
> Above tests proves zram can support writeback into real swap storage
> so that zram-cache can always have a free space. If necessary, we could
> add new plugin in dm-cache. I see It's really flexible and well-layered
> architecure so zram-blk's concept is good for us and it has lots of
> potential to be enhanced by MM/FS/Block developers. 
> 

That's an exciting direction!

-- 
Regards,
-Bob
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists