Message-ID: <20160323044245.GA4624@js1304-P5Q-DELUXE>
Date:	Wed, 23 Mar 2016 13:42:46 +0900
From:	Joonsoo Kim <iamjoonsoo.kim@....com>
To:	Lucas Stach <l.stach@...gutronix.de>
Cc:	Vlastimil Babka <vbabka@...e.cz>,
	Laura Abbott <lauraa@...eaurora.org>,
	Arnd Bergmann <arnd@...db.de>,
	Catalin Marinas <Catalin.Marinas@....com>,
	Hanjun Guo <guohanjun@...wei.com>,
	Will Deacon <will.deacon@....com>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	qiuxishi <qiuxishi@...wei.com>,
	"linux-mm@...ck.org" <linux-mm@...ck.org>,
	dingtinahong <dingtianhong@...wei.com>,
	"Leizhen (ThunderTown)" <thunder.leizhen@...wei.com>,
	Sasha Levin <sasha.levin@...cle.com>,
	Laura Abbott <labbott@...hat.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	"linux-arm-kernel@...ts.infradead.org" 
	<linux-arm-kernel@...ts.infradead.org>, chenjie6@...wei.com
Subject: Re: Suspicious error for CMA stress test

On Tue, Mar 22, 2016 at 03:56:46PM +0100, Lucas Stach wrote:
> Am Montag, den 21.03.2016, 13:42 +0900 schrieb Joonsoo Kim:
> > On Fri, Mar 18, 2016 at 02:32:35PM +0100, Lucas Stach wrote:
> > > Hi Vlastimil, Joonsoo,
> > > 
> > > Am Freitag, den 18.03.2016, 00:52 +0900 schrieb Joonsoo Kim:
> > > > 2016-03-18 0:43 GMT+09:00 Vlastimil Babka <vbabka@...e.cz>:
> > > > > On 03/17/2016 10:24 AM, Hanjun Guo wrote:
> > > > >>
> > > > >> On 2016/3/17 14:54, Joonsoo Kim wrote:
> > > > >>>
> > > > >>> On Wed, Mar 16, 2016 at 05:44:28PM +0800, Hanjun Guo wrote:
> > > > >>>>
> > > > >>>> On 2016/3/14 15:18, Joonsoo Kim wrote:
> > > > >>>>>
> > > > >>>>> On Mon, Mar 14, 2016 at 08:06:16AM +0100, Vlastimil Babka wrote:
> > > > >>>>>>
> > > > >>>>>> On 03/14/2016 07:49 AM, Joonsoo Kim wrote:
> > > > >>>>>>>
> > > > >>>>>>> On Fri, Mar 11, 2016 at 06:07:40PM +0100, Vlastimil Babka wrote:
> > > > >>>>>>>>
> > > > >>>>>>>> On 03/11/2016 04:00 PM, Joonsoo Kim wrote:
> > > > >>>>>>>>
> > > > >>>>>>>> How about something like this? Just and idea, probably buggy
> > > > >>>>>>>> (off-by-one etc.).
> > > > >>>>>>>> Should keep away cost from <pageblock_order iterations at the
> > > > >>>>>>>> expense of the
> > > > >>>>>>>> relatively fewer >pageblock_order iterations.
> > > > >>>>>>>
> > > > >>>>>>> Hmm... I tested this and found that it's code size is a little bit
> > > > >>>>>>> larger than mine. I'm not sure why this happens exactly but I guess
> > > > >>>>>>> it would be
> > > > >>>>>>> related to compiler optimization. In this case, I'm in favor of my
> > > > >>>>>>> implementation because it looks like well abstraction. It adds one
> > > > >>>>>>> unlikely branch to the merge loop but compiler would optimize it to
> > > > >>>>>>> check it once.
> > > > >>>>>>
> > > > >>>>>> I would be surprised if compiler optimized that to check it once, as
> > > > >>>>>> order increases with each loop iteration. But maybe it's smart
> > > > >>>>>> enough to do something like I did by hand? Guess I'll check the
> > > > >>>>>> disassembly.
> > > > >>>>>
> > > > >>>>> Okay. I used following slightly optimized version and I need to
> > > > >>>>> add 'max_order = min_t(unsigned int, MAX_ORDER, pageblock_order + 1)'
> > > > >>>>> to yours. Please consider it, too.
> > > > >>>>
> > > > >>>> Hmm, this one is not work, I still can see the bug is there after
> > > > >>>> applying
> > > > >>>> this patch, did I miss something?
> > > > >>>
> > > > >>> I may find that there is a bug which was introduced by me some time
> > > > >>> ago. Could you test following change in __free_one_page() on top of
> > > > >>> Vlastimil's patch?
> > > > >>>
> > > > >>> -page_idx = pfn & ((1 << max_order) - 1);
> > > > >>> +page_idx = pfn & ((1 << MAX_ORDER) - 1);
> > > > >>
> > > > >>
> > > > >> I tested Vlastimil's patch + your change with stress for more than half
> > > > >> hour, the bug
> > > > >> I reported is gone :)
> > > > >
> > > > >
> > > > > Oh, ok, will try to send proper patch, once I figure out what to write in
> > > > > the changelog :)
> > > > 
> > > > Thanks in advance!
> > > 
> > > After digging into the "PFN busy" race in CMA (see [1]), I believe we
> > > should just prevent any buddy merging in isolated ranges. This fixes the
> > > race I'm seeing without the need to hold the zone lock for extend
> > > periods of time.
> > 
> > "PFNs busy" can be caused by other type of race, too. I guess that
> > other cases happens more than buddy merging. Do you have any test case for
> > your problem?
> > 
> I don't have any specific test case, but the etnaviv driver manages to
> hit this race quite often. That's because we allocate/free a large
> number of relatively small buffer from CMA, where allocation and free
> regularly happen on different CPUs.
> 
> So while we also have cases where the "PFN busy" is triggered by other
> factors, like pages locked for get_user_pages(), this race is the number
> one source of CMA retries in my workload.
> 
> > If it is indeed a problem, you can avoid it with simple retry
> > MAX_ORDER times on alloc_contig_range(). This is a rather dirty but
> > the reason I suggest it is that there are other type of race in
> > __alloc_contig_range() and retry could help them, too. For example,
> > if some of pages in the requested range isn't attached to the LRU yet
> > or detached from the LRU but not freed to buddy,
> > test_pages_isolated() can be failed.
> 
> While a retry makes sense (if at all just to avoid a CMA allocation
> failure under CMA pressure), I would like to avoid the associated
> overhead for the common path where CMA is just racing with itself. The
> retry should only be needed in situations where we don't have any means
> to control the race, like a concurrent GUP.

Make sense. When I tried to fix merging issue previously, I worried
about side-effect of unmerged buddy so I tried to reduce unmerged
buddy as much as possible. But, your problem can be solved by
implementation that makes unmerged buddy in isolated pageblock and
it is a real problem. I think that now we can bear uncertain side-effect
of unmerged buddy for solving real problem. Please make a patch
and send to the list.

Thanks.

Powered by blists - more mailing lists