lkml.org 
[lkml]   [2013]   [Sep]   [26]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [BUG REPORT] ZSWAP: theoretical race condition issues
From
On Thu, Sep 26, 2013 at 4:48 PM, Weijie Yang <weijie.yang.kh@gmail.com> wrote:
> On Thu, Sep 26, 2013 at 3:57 PM, Minchan Kim <minchan@kernel.org> wrote:
>> On Thu, Sep 26, 2013 at 03:26:33PM +0800, Weijie Yang wrote:
>>> On Thu, Sep 26, 2013 at 1:58 PM, Minchan Kim <minchan@kernel.org> wrote:
>>> > Hello Weigie,
>>> >
>>> > On Wed, Sep 25, 2013 at 05:33:43PM +0800, Weijie Yang wrote:
>>> >> On Wed, Sep 25, 2013 at 4:31 PM, Bob Liu <lliubbo@gmail.com> wrote:
>>> >> > On Wed, Sep 25, 2013 at 4:09 PM, Weijie Yang <weijie.yang.kh@gmail.com> wrote:
>>> >> >> I think I find a new issue, for integrity of this mail thread, I reply
>>> >> >> to this mail.
>>> >> >>
>>> >> >> It is a concurrence issue either, when duplicate store and reclaim
>>> >> >> concurrentlly.
>>> >> >>
>>> >> >> zswap entry x with offset A is already stored in zswap backend.
>>> >> >> Consider the following scenario:
>>> >> >>
>>> >> >> thread 0: reclaim entry x (get refcount, but not call zswap_get_swap_cache_page)
>>> >> >>
>>> >> >> thread 1: store new page with the same offset A, alloc a new zswap entry y.
>>> >> >> store finished. shrink_page_list() call __remove_mapping(), and now
>>> >> >> it is not in swap_cache
>>> >> >>
>>> >> >
>>> >> > But I don't think swap layer will call zswap with the same offset A.
>>> >>
>>> >> 1. store page of offset A in zswap
>>> >> 2. some time later, pagefault occur, load page data from zswap.
>>> >> But notice that zswap entry x is still in zswap because it is not
>>> >> frontswap_tmem_exclusive_gets_enabled.
>>> >
>>> > frontswap_tmem_exclusive_gets_enabled is just option to see tradeoff
>>> > between CPU burining by frequent swapout and memory footprint by duplicate
>>> > copy in swap cache and frontswap backend so it shouldn't affect the stability.
>>>
>>> Thanks for explain this.
>>> I don't mean to say this option affects the stability, but that zswap
>>> only realize
>>> one option. Maybe it's better to realize both options for different workloads.
>>
>> "zswap only relize one option"
>> What does it mena? Sorry. I couldn't parse your intention. :)
>> You mean zswap should do something special to support frontswap_tmem_exclusive_gets?
>
> Yes. But I am not sure whether it is worth.
>
>>>
>>> >> this page is with PageSwapCache(page) and page_private(page) = entry.val
>>> >> 3. change this page data, and it become dirty
>>> >
>>> > If non-shared swapin page become redirty, it should remove the page from
>>> > swapcache. If shared swapin page become redirty, it should do CoW so it's a
>>> > new page so that it doesn't live in swap cache. It means it should have new
>>> > offset which is different with old's one for swap out.
>>> >
>>> > What's wrong with that?
>>>
>>> It is really not a right scene for duplicate store. And I can not think out one.
>>> If duplicate store is impossible, How about delete the handle code in zswap?
>>> If it does exist, I think there is a potential issue as I described.
>>
>> You mean "zswap_duplicate_entry"?
>> AFAIR, I already had a question to Seth when zswap was born but AFAIRC,
>> he said that he didn't know exact reason but he saw that case during
>> experiement so copy the code peice from zcache.
>>
>> Do you see the case, too?
>
> Yes, I mean duplicate store.
> I check the /Documentation/vm/frontswap.txt, it mentions "duplicate stores",
> but I am still confused.
>
> I wrote a zcache varietas which swap out compressed page to swapfile.
> I did see that case when I test it on andorid smartphone(arm v7),

Why not test it based on zswap directly?
My suggestion is do more and more stress testing against current
zswap, if there is really any bug triggered or unusual performance
issue then we dig it out and fix it.

--
Regards,
--Bob


\
 
 \ /
  Last update: 2013-09-26 12:21    [W:0.119 / U:1.504 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site