Due to an influx of spam, we have had to impose restrictions on new accounts. Please see this wiki page for instructions on how to get full permissions. Sorry for the inconvenience.
Admin message
The migration is almost done, at least the rest should happen in the background. There are still a few technical difference between the old cluster and the new ones, and they are summarized in this issue. Please pay attention to the TL:DR at the end of the comment.
I was reporting this issue on Arch Linux Bug Tracker and with the help of other people using (git bisect and compiling over and over the kernel!) identified the commit, that from then, Rendering has problems. Maybe this helps.
520d05a77b2866eb4cb9e548e1d8c8abcfe60ec5 is the first bad commitcommit 520d05a77b2866eb4cb9e548e1d8c8abcfe60ec5Author: Chris Wilson <chris@chris-wilson.co.uk>Date: Sun Jan 17 09:30:15 2021 +0000 drm/i915/gt: Clear CACHE_MODE prior to clearing residuals commit ef99a60ffd9b918354e038bc5e61f007ff7e901d upstream. Since we do a bare context switch with no restore, the clear residual kernel runs on dirty state, and we must be careful to avoid executing with bad state from context registers inherited from a malicious client. Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2955 Fixes: 09aa9e45863e ("drm/i915/gt: Restore clear-residual mitigations for Ivybridge, Baytrail") Testcase: igt/gem_ctx_isolation # ivb,vlv Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Akeem G Abodunrin <akeem.g.abodunrin@intel.com> Reviewed-by: Akeem G Abodunrin <akeem.g.abodunrin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210117093015.29143-1-chris@chris-wilson.co.uk (cherry picked from commit ace44e13e577c2ae59980e9a6ff5ca253b1cf831) Signed-off-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> drivers/gpu/drm/i915/gt/gen7_renderclear.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
I just commented the changes made in the drm/i915/gt: Clear CACHE_MODE prior to clearing residuals commit, and compiled kernel 5.12.0-rc8 and the problem is gone.
There is definitely a problem here that I'm not understanding because of not being a kernel developer Or a graphics card driver developer.
Check #3081 (closed), I've narrowed it down there some more, but according to Intel, this code shouldn't even be able to cause an issue, so the underlying bug might be somewhere else.
I tried the command, there were some typos so i fixed them echo off | sudo tee /sys/module/i915/parameters/mitigations
The issue still occurs, I don't know how to apply patches but I'll ask a friend to help me.
I'm compiling the 5.10.14 kernel with the patch you provided. I Will tell you if it fixes the issue (Probably tomorrow, 1 AM here and I just started the compilation)
@ickle Did you see my comment about the git bisect result? do you think it's unrelated?
Any updates on this? It's really annoying me that I cannot update from kernel 5.10.12 and the problem is not fixed even in 5.12-rc1
I really am happy to provide any more info or test, but I don't know what more I can do. Maybe coding? Changing the changes of the 520d05a77b2866eb4cb9e548e1d8c8abcfe60ec5 one by one to see which one causes it? I don't have any idea how graphic cards work so I'm afraid a little bit to do that. Any help to fix this is so appreciated :(
Having the same issue on a intel bay-trail system j2900. Linux kernel versions lower than 5.10.12 work just fine on Debian bullseye/testing. This kernel version should not be shipped when this is out of testing phase. Hope this is fixed soon. Been using linux since 2015. I believe this is effecting ivy bridge cpus with the intel hd 2500 and 4000 igpu respectively. Surprisingly my old intel gma 950 hasn’t had issues on my other work computer.