From: smoch@web.de (Soeren Moch)
Date: Wed, 16 Jan 2013 23:36:35 +0100
Subject: [PATCH v2] mm: dmapool: use provided gfp flags for all
 dma_alloc_coherent() calls
In-Reply-To: <20130116174736.GJ25500@titan.lakedaemon.net>
References: <50F3F289.3090402@web.de>
 <20130115165642.GA25500@titan.lakedaemon.net>
 <20130115175020.GA3764@kroah.com>
 <20130115201617.GC25500@titan.lakedaemon.net>
 <20130115215602.GF25500@titan.lakedaemon.net> <50F5F1B7.3040201@web.de>
 <20130116024014.GH25500@titan.lakedaemon.net> <50F61D86.4020801@web.de>
 <50F66B1B.40301@web.de> <50F6E419.5080007@web.de>
 <20130116174736.GJ25500@titan.lakedaemon.net>
Message-ID: <50F72B73.5070504@web.de>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On 16.01.2013 18:47, Jason Cooper wrote:
> On Wed, Jan 16, 2013 at 06:32:09PM +0100, Soeren Moch wrote:
>> On 16.01.2013 09:55, Soeren Moch wrote:
>>> On 16.01.2013 04:24, Soeren Moch wrote:
>>>> I did not bisect it, but Marek mentioned earlier that commit
>>>> e9da6e9905e639b0f842a244bc770b48ad0523e9 in Linux v3.6-rc1 introduced
>>>> new code for dma allocations. This is probably the root cause for the
>>>> new (mis-)behavior (due to my tests 3.6.0 is not working anymore).
>>>
>>> I don't want to say that Mareks patch is wrong, probably it triggers a
>>> bug somewhere else! (in em28xx?)
>>
>> The em28xx sticks are using isochronous usb transfers. Is there a
>> special handling for that?
>
> I'm looking at that now.  It looks like the em28xx wants (as a maximum)
> 655040 bytes (em28xx-core.c:1088).  There are 5 transfer buffers, with
> 64 max packets and 2047 max packet size (runtime reported max & 0x7ff).
>
> If it actually needs all of that, then the answer may be to just
> increase coherent_pool= when using that driver.  I'll keep digging.

I already tested with 4M coherent pool size and could not see
significant improvement. Would it make sense to further increase the
buffer size?

Regards,
Soeren