From: Sasha Levin <sasha.levin@oracle.com>
To: Joonsoo Kim <iamjoonsoo.kim@lge.com>,
Stefan Strogin <s.strogin@partner.samsung.com>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Andrew Morton <akpm@linux-foundation.org>,
Marek Szyprowski <m.szyprowski@samsung.com>,
Michal Nazarewicz <mina86@mina86.com>,
aneesh.kumar@linux.vnet.ibm.com,
Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
Dmitry Safonov <d.safonov@partner.samsung.com>,
Pintu Kumar <pintu.k@samsung.com>,
Weijie Yang <weijie.yang@samsung.com>,
Laura Abbott <lauraa@codeaurora.org>,
SeongJae Park <sj38.park@gmail.com>, Hui Zhu <zhuhui@xiaomi.com>,
Minchan Kim <minchan@kernel.org>,
Dyasly Sergey <s.dyasly@samsung.com>,
Vyacheslav Tyrtov <v.tyrtov@samsung.com>,
Aleksei Mateosian <a.mateosian@samsung.com>,
gregory.0xf0@gmail.com, gioh.kim@lge.com, pavel@ucw.cz,
stefan.strogin@gmail.com
Subject: Re: [PATCH v4 0/5] mm: cma: add some debug information for CMA
Date: Mon, 16 Mar 2015 21:54:18 -0400 [thread overview]
Message-ID: <5507894A.2080008@oracle.com> (raw)
In-Reply-To: <20150317014306.GB19483@js1304-P5Q-DELUXE>
On 03/16/2015 09:43 PM, Joonsoo Kim wrote:
> On Mon, Mar 16, 2015 at 07:06:55PM +0300, Stefan Strogin wrote:
>> > Hi all.
>> >
>> > Here is the fourth version of a patch set that adds some debugging facility for
>> > CMA.
>> >
>> > This patch set is based on next-20150316.
>> > It is also available on git:
>> > git://github.com/stefanstrogin/linux -b cmainfo-v4
>> >
>> > We want an interface to see a list of currently allocated CMA buffers and some
>> > useful information about them (like /proc/vmallocinfo but for physically
>> > contiguous buffers allocated with CMA).
>> >
>> > For example. We want a big (megabytes) CMA buffer to be allocated in runtime
>> > in default CMA region. If someone already uses CMA then the big allocation
>> > could fail. If it happened then with such an interface we could find who used
>> > CMA at the moment of failure, who caused fragmentation and so on. Ftrace also
>> > would be helpful here, but with ftrace we can see the whole history of
>> > allocations and releases, whereas with this patch set we can see a snapshot of
>> > CMA region with actual information about its allocations.
> Hello,
>
> Hmm... I still don't think that this is really helpful to find root
> cause of fragmentation. Think about following example.
>
> Assume 1024 MB CMA region.
>
> 128 MB allocation * 4
> 1 MB allocation
> 128 MB allocation
> 128 MB release * 4 (first 4)
> try 512 MB allocation
>
> With above sequences, fragmentation happens and 512 MB allocation would
> be failed. We can get information about 1 MB allocation and 128 MB one
> from the buffer list as you suggested, but, fragmentation are related
> to whole sequence of allocation/free history, not snapshot of allocation.
This is solvable by dumping task->comm in the tracepoint patch (1/5), right?
Thanks,
Sasha
--
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>
WARNING: multiple messages have this Message-ID (diff)
From: Sasha Levin <sasha.levin@oracle.com>
To: Joonsoo Kim <iamjoonsoo.kim@lge.com>,
Stefan Strogin <s.strogin@partner.samsung.com>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Andrew Morton <akpm@linux-foundation.org>,
Marek Szyprowski <m.szyprowski@samsung.com>,
Michal Nazarewicz <mina86@mina86.com>,
aneesh.kumar@linux.vnet.ibm.com,
Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
Dmitry Safonov <d.safonov@partner.samsung.com>,
Pintu Kumar <pintu.k@samsung.com>,
Weijie Yang <weijie.yang@samsung.com>,
Laura Abbott <lauraa@codeaurora.org>,
SeongJae Park <sj38.park@gmail.com>, Hui Zhu <zhuhui@xiaomi.com>,
Minchan Kim <minchan@kernel.org>,
Dyasly Sergey <s.dyasly@samsung.com>,
Vyacheslav Tyrtov <v.tyrtov@samsung.com>,
Aleksei Mateosian <a.mateosian@samsung.com>,
gregory.0xf0@gmail.com, gioh.kim@lge.com, pavel@ucw.cz,
stefan.strogin@gmail.com
Subject: Re: [PATCH v4 0/5] mm: cma: add some debug information for CMA
Date: Mon, 16 Mar 2015 21:54:18 -0400 [thread overview]
Message-ID: <5507894A.2080008@oracle.com> (raw)
In-Reply-To: <20150317014306.GB19483@js1304-P5Q-DELUXE>
On 03/16/2015 09:43 PM, Joonsoo Kim wrote:
> On Mon, Mar 16, 2015 at 07:06:55PM +0300, Stefan Strogin wrote:
>> > Hi all.
>> >
>> > Here is the fourth version of a patch set that adds some debugging facility for
>> > CMA.
>> >
>> > This patch set is based on next-20150316.
>> > It is also available on git:
>> > git://github.com/stefanstrogin/linux -b cmainfo-v4
>> >
>> > We want an interface to see a list of currently allocated CMA buffers and some
>> > useful information about them (like /proc/vmallocinfo but for physically
>> > contiguous buffers allocated with CMA).
>> >
>> > For example. We want a big (megabytes) CMA buffer to be allocated in runtime
>> > in default CMA region. If someone already uses CMA then the big allocation
>> > could fail. If it happened then with such an interface we could find who used
>> > CMA at the moment of failure, who caused fragmentation and so on. Ftrace also
>> > would be helpful here, but with ftrace we can see the whole history of
>> > allocations and releases, whereas with this patch set we can see a snapshot of
>> > CMA region with actual information about its allocations.
> Hello,
>
> Hmm... I still don't think that this is really helpful to find root
> cause of fragmentation. Think about following example.
>
> Assume 1024 MB CMA region.
>
> 128 MB allocation * 4
> 1 MB allocation
> 128 MB allocation
> 128 MB release * 4 (first 4)
> try 512 MB allocation
>
> With above sequences, fragmentation happens and 512 MB allocation would
> be failed. We can get information about 1 MB allocation and 128 MB one
> from the buffer list as you suggested, but, fragmentation are related
> to whole sequence of allocation/free history, not snapshot of allocation.
This is solvable by dumping task->comm in the tracepoint patch (1/5), right?
Thanks,
Sasha
next prev parent reply other threads:[~2015-03-17 1:54 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-03-16 16:06 [PATCH v4 0/5] mm: cma: add some debug information for CMA Stefan Strogin
2015-03-16 16:06 ` Stefan Strogin
2015-03-16 16:06 ` [PATCH v4 1/5] mm: cma: add trace events to debug physically-contiguous memory allocations Stefan Strogin
2015-03-16 16:06 ` Stefan Strogin
2015-03-16 20:49 ` Stefan Strogin
2015-03-16 20:49 ` Stefan Strogin
2015-03-16 23:47 ` Steven Rostedt
2015-03-16 23:47 ` Steven Rostedt
2015-03-19 20:18 ` Stefan Strogin
2015-03-19 20:18 ` Stefan Strogin
2015-03-19 20:34 ` Steven Rostedt
2015-03-19 20:34 ` Steven Rostedt
2015-03-20 10:46 ` Stefan Strogin
2015-03-20 10:46 ` Stefan Strogin
2015-03-20 14:31 ` Steven Rostedt
2015-03-20 14:31 ` Steven Rostedt
2015-03-17 7:40 ` Ingo Molnar
2015-03-17 7:40 ` Ingo Molnar
2015-03-19 20:22 ` Stefan Strogin
2015-03-19 20:22 ` Stefan Strogin
2015-03-23 14:04 ` Ingo Molnar
2015-03-23 14:04 ` Ingo Molnar
2015-03-16 16:06 ` [PATCH v4 2/5] mm: cma: add number of pages to debug message in cma_release() Stefan Strogin
2015-03-16 16:06 ` Stefan Strogin
2015-03-16 16:06 ` [PATCH v4 3/5] stacktrace: add seq_print_stack_trace() Stefan Strogin
2015-03-16 16:06 ` Stefan Strogin
2015-03-16 17:33 ` Michal Nazarewicz
2015-03-16 17:33 ` Michal Nazarewicz
2015-03-16 16:06 ` [PATCH v4 4/5] mm: cma: add list of currently allocated CMA buffers to debugfs Stefan Strogin
2015-03-16 16:06 ` Stefan Strogin
2015-03-16 17:51 ` Michal Nazarewicz
2015-03-16 17:51 ` Michal Nazarewicz
2015-03-16 16:09 ` [PATCH v4 5/5] mm: cma: add functions to get region pages counters Stefan Strogin
2015-03-16 16:09 ` Stefan Strogin
2015-03-17 1:43 ` [PATCH v4 0/5] mm: cma: add some debug information for CMA Joonsoo Kim
2015-03-17 1:43 ` Joonsoo Kim
2015-03-17 1:54 ` Sasha Levin [this message]
2015-03-17 1:54 ` Sasha Levin
2015-03-17 2:04 ` Joonsoo Kim
2015-03-17 2:04 ` Joonsoo Kim
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=5507894A.2080008@oracle.com \
--to=sasha.levin@oracle.com \
--cc=a.mateosian@samsung.com \
--cc=akpm@linux-foundation.org \
--cc=aneesh.kumar@linux.vnet.ibm.com \
--cc=d.safonov@partner.samsung.com \
--cc=gioh.kim@lge.com \
--cc=gregory.0xf0@gmail.com \
--cc=iamjoonsoo.kim@lge.com \
--cc=lauraa@codeaurora.org \
--cc=laurent.pinchart@ideasonboard.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=m.szyprowski@samsung.com \
--cc=mina86@mina86.com \
--cc=minchan@kernel.org \
--cc=pavel@ucw.cz \
--cc=pintu.k@samsung.com \
--cc=s.dyasly@samsung.com \
--cc=s.strogin@partner.samsung.com \
--cc=sj38.park@gmail.com \
--cc=stefan.strogin@gmail.com \
--cc=v.tyrtov@samsung.com \
--cc=weijie.yang@samsung.com \
--cc=zhuhui@xiaomi.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.