rv200 video corruption caused by what? kernel? Xserver? DDX? Mesa?

[Felix Miata]

Alex Deucher composed on 2020-03-25 13:22 (UTC-0400):

> Felix Miata wrote:

>> 01:00.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] RV200 [Radeon 7500/7500 LE] [1002:5157]

>> Debian Buster and Mageia 7 are OK, but with Fedora 30 and Tumbleweed with
>> kernels newer than 5.3.x, solid background and mouse pointer are all that
>> paint properly: http://fm.no-ip.com/SS/rv200corrupt.png

>> Is this a known problem? If not, whose bug tracker do I file this on, and
>> which component?

> Can you bisect?  The relevant radeon code in all of these components
> hasn't really been touched in ages.  Maybe related to this:
> https://lkml.org/lkml/2019/12/14/263
> but that was kernel 5.4 not 5.3.
I spent a bunch of time looking for in between kernels to try. In the process I
tried also with RV250 and RV380 without being able to repro. Eventually I stumbled
onto radeon.agpmode=-1, which worked around the corruption well enough on Fedora
and Tumbleweed.

I got luckier and found after reading your lkml URL
https://bugs.mageia.org/show_bug.cgi?id=26237
reported against kernel-desktop-5.5.4-1.mga7-1-1.mga7 to produce corrupt Xorg
video described much like observed here. That bug reports fixed with
kernel-desktop-5.5.6-1.mga7. I found and installed 5.5.4 and confirm it's broken,
and that it's fixed in 5.5.6-2.mga7, which explains why I failed to experience
this in Mageia, where I had jumped from 5.3.13 to 5.4.17.

Apparently whatever Mageia did should be upstreamed if it hasn't been lately.
Fedora's 5.5.10 remains bad.
-- 
Evolution as taught in public schools is religion, not science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/