public inbox for linux-media@vger.kernel.org
 help / color / mirror / Atom feed
From: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
To: "Gabor Z. Papp" <gzpapp.lists@gmail.com>
Cc: linux-media@vger.kernel.org
Subject: Re: em28xx problem with 3.10-4.0
Date: Wed, 17 Jun 2015 08:02:49 -0300	[thread overview]
Message-ID: <20150617080249.379eeb10@recife.lan> (raw)
In-Reply-To: <x6oakedev9@gzp>

Em Wed, 17 Jun 2015 08:32:26 +0200
"Gabor Z. Papp" <gzpapp.lists@gmail.com> escreveu:

> * Mauro Carvalho Chehab <mchehab@osg.samsung.com>:
> 
> | Nothing. You just ran out of continuous memory. This driver
> | requires long chunks of continuous memory for USB data transfer.
> 
> And there is no way to preset some mem?
> Or do something to get the driver work again?
> I don't think I'm using too much memory.
> 
> $ free
>              total       used       free     shared    buffers     cached
> Mem:       2073656     625696    1447960          0      21072     231096
> -/+ buffers/cache:     373528    1700128
> Swap:      1004056          0    1004056

>From your error logs, it failed to allocate the 3rd buffer (of a total of 5
buffers) with a continuous block of 165.120 bytes on the DMA range.

In order words, your system needs to have at least 5 non-fragmented buffers
with 256KB each, on a memory region where the CPU can do DMA (e. g. 
outside the high memory area).

I'm not a memory management specialist, but I guess you could try to change
some sysctl parameters or use a different memory allocator in order to avoid
memory fragmentation.

If you're a C programmer, an option would be to change the driver's code
to optimize it for low memory usage, for example, to reduce the buffer size
and increasing the number of buffers (at the cost of requiring more CPU
and/or reducing the maximum size of the image). Another alternative would be
to reserve the memory at the time the driver gets loaded.

Regards,
Mauro

      reply	other threads:[~2015-06-17 11:02 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-06-11  8:11 em28xx problem with 3.10-4.0 Gabor Z. Papp
2015-06-16  6:54 ` Gabor Z. Papp
2015-06-16  9:20   ` Mauro Carvalho Chehab
2015-06-16 11:09     ` Andy Furniss
2015-06-17  6:32     ` Gabor Z. Papp
2015-06-17 11:02       ` Mauro Carvalho Chehab [this message]

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=20150617080249.379eeb10@recife.lan \
    --to=mchehab@osg.samsung.com \
    --cc=gzpapp.lists@gmail.com \
    --cc=linux-media@vger.kernel.org \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox