Bug 76957

Created attachment 96799 [details]
Xorg log

Created attachment 96800 [details]
glxinfo outputs

Created attachment 96801 [details]
dmesg | grep radeon

Created attachment 96802 [details]
lspci -vvnn

Created attachment 96803 [details]
Syslog excerpt showing a segfault (I think it’s related)

Erratum: my graphic card has 256 MB of RAM.

Thank you,
Benjamin.

Please attach your full dmesg output.  This is not related to bug 66932 as dpm is not enabled on your asic by default.

Created attachment 96809 [details]
full dmesg

The dmesg output reveals these lines:

> [    5.343548] radeon 0000:05:00.0: VRAM: 1024M 0x0000000000000000 - 0x000000003FFFFFFF (1024M used)
> [    5.343550] radeon 0000:05:00.0: GTT: 1024M 0x0000000040000000 - 0x000000007FFFFFFF
> [    5.343552] [drm] Detected VRAM RAM=1024M, BAR=256M

1G VRAM seems to be a wrong value, since my graphic card holds only 256MB, which seems to be seeing as BAR (what’s that stuff?). 

Anyway, could it be the cause of my problem?

Thank you,
Benjamin.

(In reply to comment #9)
> The dmesg output reveals these lines:
> 
> > [    5.343548] radeon 0000:05:00.0: VRAM: 1024M 0x0000000000000000 - 0x000000003FFFFFFF (1024M used)
> > [    5.343550] radeon 0000:05:00.0: GTT: 1024M 0x0000000040000000 - 0x000000007FFFFFFF
> > [    5.343552] [drm] Detected VRAM RAM=1024M, BAR=256M
> 
> 1G VRAM seems to be a wrong value, since my graphic card holds only 256MB,
> which seems to be seeing as BAR (what’s that stuff?). 
> 
> Anyway, could it be the cause of my problem?

Your board has 1 GB of vram.  The max BAR size is 256 MB regardless of how much video ram is actually on the card.  The BAR is the CPU's aperture to vram.  The CPU can only access teh first 256 MB of vram, but the GPU can access the entire amount of vram.

Oups, sorry… Indeed, you are right Alex, this board has 1 GB <http://www.club-3d.com/index.php/products/reader.en/product/radeon-hd-6570-coolstream-edition.html>

Thanks again for the clever explainations.

Thus, it was my last guess… Is there anything else I can do to find the origin of this issue?

Sincerely,
Benjamin.

Does disabling tiling help?  Add:
Option "ColorTiling" "False"
Option "ColorTiling2D" "False"
to the device section of your xorg.conf

Unfortunately, deactivating ColorTiling and/or ColorTiling2D has no effect. I also tried to set "SwapbuffersWait" to "False". 

Actually, as far as I remember, I tried to test most of the options available (http://manpages.debian.org/cgi-bin/man.cgi?sektion=4&query=radeon) two years ago, when I installed the board. Then, I switched to fglrx-driver.

I think it’s not related, but… often, X failed to restart from the virtual console (weird pixelized screen or black screen, or nothing at all).
I had to reboot the computer to double check the result (and then, I did not see any error in the Xorg log). 

Thanks again,
Benjamin.

-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/485.