dri-devel Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Michel Dänzer" <michel@daenzer.net>
To: Thomas Schwinge <thomas@codesourcery.com>,
	dri-devel@lists.freedesktop.org
Cc: Thomas Hellstrom <thellstrom@vmware.com>, wolfgang.rupprecht@gmail.com
Subject: Re: Radeon: [TTM] Failed to find memory space for buffer [...] eviction
Date: Mon, 11 Aug 2014 16:40:13 +0900	[thread overview]
Message-ID: <53E8735D.4090402@daenzer.net> (raw)
In-Reply-To: <877g2jm9jn.fsf@schwinge.name>


[-- Attachment #1.1: Type: text/plain, Size: 3574 bytes --]

On 08.08.2014 21:48, Thomas Schwinge wrote:
> 
> I recetly repurposed a BlueMedia OPTIMA II system, Biostar MCP6P M2+
> mainboard, AMD Athlon II X2 215 with 2700 MHz, 8 GiB RAM, Xen setup, for
> use as a desktop machine (the Xen dom0, specifically).  I put in a
> Sapphire Radeon HD 4350 card where I'm connecting the DVI output to a
> portrait-oriented (1200x1920) and the VGA output to a landscape-oriented
> (1440x900) monitor.  I'm running Debian GNU/Linux testing with the
> Cinnamon desktop environment (Gnome Shell derivate of Gnome 2).  This is
> being used as a development and all-purpose system, that is, several
> virtual desktops, with mostly full-screen terminal/Emacs/web browser
> processes running.  This is working nicely.
> 
> However, what happens after (irregularely) several hours of usage is that
> any interactions involving the desktop get executed very sluggishly
> (initially about 1 s of delay, later getting worse), which makes this
> unworkable.  Once this happens, I see the following in dmesg, repeating
> every once in a while:
> 
>     [96450.514927] [TTM] Failed to find memory space for buffer 0xffff88015daef048 eviction
>     [96450.514934] [TTM] No space for ffff88015daef048 (65536 pages, 262144K, 256M)
>     [96450.514936] [TTM]   placement[0]=0x00410002 (1)
>     [96450.514938] [TTM]     has_type: 1
>     [96450.514939] [TTM]     use_type: 1
>     [96450.514940] [TTM]     flags: 0x0000000A
>     [96450.514941] [TTM]     gpu_offset: 0x20000000
>     [96450.514942] [TTM]     size: 262144
>     [96450.514943] [TTM]     available_caching: 0x00070000
>     [96450.514944] [TTM]     default_caching: 0x00010000
>     [96450.514947] [TTM]  0x00000000-0x00000001:        1: used
>     [96450.514949] [TTM]  0x00000001-0x00000011:       16: used
>     [96450.514951] [TTM]  0x00000011-0x00000111:      256: used
>     [96450.514952] [TTM]  0x00000111-0x00000121:       16: used
>     [96450.514953] [TTM]  0x00000121-0x00000122:        1: used
>     [96450.514955] [TTM]  0x00000122-0x00000222:      256: used
>     [96450.514956] [TTM]  0x00000222-0x0000030f:      237: free
>     [96450.514957] [TTM]  0x0000030f-0x00000313:        4: used
>     [96450.514958] [TTM]  0x00000313-0x000061c0:    24237: free
>     [96450.514960] [TTM]  0x000061c0-0x000161c0:    65536: used
>     [96450.514961] [TTM]  0x000161c0-0x000242fe:    57662: free
>     [96450.514962] [TTM]  0x000242fe-0x000342fe:    65536: used
>     [96450.514963] [TTM]  0x000342fe-0x0003d80c:    38158: free
>     [96450.514965] [TTM]  0x0003d80c-0x0003d90c:      256: used
>     [96450.514966] [TTM]  0x0003d90c-0x0003ff94:     9864: free
>     [96450.514967] [TTM]  0x0003ff94-0x0003ffa4:       16: used
>     [96450.514969] [TTM]  0x0003ffa4-0x00040000:       92: free
>     [96450.514970] [TTM]  total: 262144, used 131894 free 130250

Looks like it's failing to evict a buffer object (BO) of size 256MB,
presumably from VRAM.

AFAICT there's only 256MB of GTT available, is this an AGP card? If so,
does radeon.agpmode=-1 on the kernel command line help?

If not, increasing the GTT size further using radeon.gartsize=1024 or
even larger powers of two might help.

Though I wonder if one of the apps you're running really needs a BO of
size 256MB, or maybe just thinks it's a good idea to use a texture of
the maximum size supported by OpenGL...


-- 
Earthling Michel Dänzer            |                  http://www.amd.com
Libre software enthusiast          |                Mesa and X developer


[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 234 bytes --]

[-- Attachment #2: Type: text/plain, Size: 159 bytes --]

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

      reply	other threads:[~2014-08-11  7:40 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-08-08 12:48 Radeon: [TTM] Failed to find memory space for buffer [...] eviction Thomas Schwinge
2014-08-11  7:40 ` Michel Dänzer [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=53E8735D.4090402@daenzer.net \
    --to=michel@daenzer.net \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=thellstrom@vmware.com \
    --cc=thomas@codesourcery.com \
    --cc=wolfgang.rupprecht@gmail.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox