Re: rv200 video corruption caused by what? kernel? Xserver? DDX? Mesa?
On Wed, Mar 25, 2020 at 11:36 AM 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. Alex Alex > > # Debian 10 Buster > ii libdrm-amdgpu1:i386 2.4.97-1i386 > ii libdrm-common 2.4.97-1all ii libdrm-radeon1:i386 > 2.4.97-1i386 > ii libdrm2:i386 2.4.97-1i386 > ii libgl1-mesa-dri:i386 18.3.6-2+deb10u1 i386 > ii libva-drm2:i386 2.4.0-1 i386 > ii libdrm-radeon1:i386 2.4.97-1i386 > ii xserver-xorg-core 2:1.20.4-1 i386 > ii xserver-xorg-video-radeon 1:19.0.1-1 i386 > ii linux-image-4.19.0-8-686 4.19.98-1 i386 # OK Xorg video > # Mageia 7 > libdricore1-9.2.4-1.mga4 > libdri-drivers-20.0.1-1.mga7 > libdrm2-2.4.100-1.mga7 > libdrm_amdgpu1-2.4.100-1.mga7 > libdrm_radeon1-2.4.100-1.mga7 > libmesaegl1-20.0.1-1.mga7 > libmesagl1-20.0.1-1.mga7 > libmesaglu1-9.0.0-8.mga7 > mesa-20.0.1-1.mga7 > radeontool-1.6.3-9.mga7 > x11-driver-video-ati-19.1.0-1.mga7 > x11-server-xorg-1.20.7-1.mga7 > kernel-desktop-5.3.13-2.mga7-1-1.mga7 # OK > kernel-desktop-5.4.17-1.mga7-1-1.mga7 # OK Xorg video > # Fedora 30 > libdrm-2.4.100-1.fc30.i686 > mesa-dri-drivers-19.1.8-1.fc30.i686 > mesa-libEGL-19.1.8-1.fc30.i686 > mesa-libgbm-19.1.8-1.fc30.i686 > mesa-libGL-19.1.8-1.fc30.i686 > mesa-libglapi-19.1.8-1.fc30.i686 > mesa-libGLU-9.0.0-17.fc30.i686 > xorg-x11-drv-ati-19.0.1-2.fc30.i686 > xorg-x11-server-Xorg-1.20.6-1.fc30.i686 > kernel-core-5.3.7-200.fc30.i686 # OK > kernel-core-5.4.19-100.fc30.i686# Corrupt Xorg video > # openSUSE Tumbleweed 2020030322 > Mesa-20.0.1-245.1.i586 > Mesa-dri-20.0.1-245.1.i586 > Mesa-gallium-20.0.1-245.1.i586 > Mesa-libEGL1-20.0.1-245.1.i586 > Mesa-libGL1-20.0.1-245.1.i586 > Mesa-libglapi0-20.0.1-245.1.i586 > libdrm2-2.4.100-2.2.i586 > libdrm_amdgpu1-2.4.100-2.2.i586 > libdrm_radeon1-2.4.100-2.2.i586 > xf86-video-ati-19.1.0-2.2.i586 > xorg-x11-server-1.20.7+0-3.1.i586 > kernel-default-5.3.12-2.2.i586 # OK > kernel-default-5.4.14-2.1.i586 # corrupt Xorg video > -- > 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/ > ___ > amd-gfx mailing list > amd-gfx@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/amd-gfx ___ amd-gfx mailing list amd-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Re: rv200 video corruption caused by what? kernel? Xserver? DDX? Mesa?
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/ ___ amd-gfx mailing list amd-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Re: rv200 video corruption caused by what? kernel? Xserver? DDX? Mesa?
Felix Miata composed on 2020-03-25 19:30 (UTC-0400): > 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. >From Mageia kernel-desktop changelogs: update to 5.5.6 20200312: - work around radeon regression introduced in 5.4 series (mga#26237) * Revert "drm/radeon: simplify and cleanup setting the dma mask" * Revert "drm/radeon: handle PCIe root ports with addressing limitations" update to 5.4.6 20200215: - Revert "drm/radeon: simplify and cleanup setting the dma mask" - Revert "drm/radeon: handle PCIe root ports with addressing limitations" -- 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/ ___ amd-gfx mailing list amd-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Re: rv200 video corruption caused by what? kernel? Xserver? DDX? Mesa?
Felix Miata composed on 2020-03-25 20:59 (UTC-0400): > Felix Miata composed on 2020-03-25 19:30 (UTC-0400): >> 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. >>From Mageia kernel-desktop changelogs: > update to 5.5.6 20200312: > - work around radeon regression introduced in 5.4 series (mga#26237) > * Revert "drm/radeon: simplify and cleanup setting the dma mask" > * Revert "drm/radeon: handle PCIe root ports with addressing limitations" > update to 5.4.6 20200215: > - Revert "drm/radeon: simplify and cleanup setting the dma mask" > - Revert "drm/radeon: handle PCIe root ports with addressing limitations" I reported a bug: https://bugzilla.kernel.org/show_bug.cgi?id=206973 [AMD/ATI] RV200 [Radeon 7500/7500 LE] [1002:5157] overwhelming video corruption in Xorg -- 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/ ___ amd-gfx mailing list amd-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx