[Bug 104932] Hang when running X11/Wayland on GFX8/Polaris10/Ellesmere/Rx-480-8GiB (agd5f a5592a6df4f45a018b48f252ad1c498e683e9b9d, hwentland's DC-Patches-Jan-31-2018.mbox applied)
https://bugs.freedesktop.org/show_bug.cgi?id=104932
Bug ID: 104932 Summary: Hang when running X11/Wayland on GFX8/Polaris10/Ellesmere/Rx-480-8GiB (agd5f a5592a6df4f45a018b48f252ad1c498e683e9b9d, hwentland's DC-Patches-Jan-31-2018.mbox applied) Product: DRI Version: XOrg git Hardware: Other OS: All Status: NEW Severity: normal Priority: medium Component: DRM/AMDgpu Assignee: dri-devel@lists.freedesktop.org Reporter: robink@creosotehill.org
Created attachment 137158 --> https://bugs.freedesktop.org/attachment.cgi?id=137158&action=edit Kernel serial log while firing up the Awesome WM
Hello- I've been having (for a while now, and going back a bit in terms of AMDGPU tree commit history) an issue whereby the framebuffer & graphics stack (*usually* not the kernel writ large) will hang upon firing up XOrg's XServer or a Wayland compositor (in my case, GNOME-Shell). What compounded my frustration was that I'd rarely (if ever) see any output from the Xorg server, the Wayland compositor, or the kernel indicating there was any sort of problem, save for a trapped SIGQUIT (trap3) hitting gnome-shell. Finally, in firing up X11 directly w/ Awesome as the WM I obtained some kernel output which will hopefully prove useful in diagnosing this issue. It's below (and attached): amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x00500002 amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0A088002 amdgpu 0000:01:00.0: VM fault (0x02, vmid 5, pasid 32768) at page 5242882, read from 'TC6' (0x54433600) (136) [drm:amdgpu_job_timedout] *ERROR* ring gfx timeout, last signaled seq=45, last emitted seq=48 [drm] IP block:gfx_v8_0 is hung! [drm] GPU recovery disabled.
Doubt it's tremendously useful, but kernel output with gnome-shell (in this case acting as a Wayland compositor) is below and attached as well: traps: gnome-shell[1260] trap int3 ip:7f6e4b956e01 sp:7ffe72a8aa50 error:0 in libglib-2.0.so.0.5200.3[7f6e4b908000+10f000] traps: gnome-shell[1336] trap int3 ip:7f0f89b3ee01 sp:7ffefd8d3df0 error:0 in libglib-2.0.so.0.5200.3[7f0f89af0000+10f000] traps: gnome-shell[1351] trap int3 ip:7f0984cb6e01 sp:7ffd3caa7650 error:0 in libglib-2.0.so.0.5200.3[7f0984c68000+10f000]
Lastly, my Xorg.0.log is attached, but you'll find it's quite boring (no complaints, which is good, but seemingly no useful debugging information, either).
https://bugs.freedesktop.org/show_bug.cgi?id=104932
--- Comment #1 from Robin Kauffman robink@creosotehill.org --- Created attachment 137159 --> https://bugs.freedesktop.org/attachment.cgi?id=137159&action=edit Kernel serial log while firing up gnome-shell as a Wayland compositor
Forgot that Bugzilla nominally allows only one attachment per comment.
https://bugs.freedesktop.org/show_bug.cgi?id=104932
--- Comment #2 from Robin Kauffman robink@creosotehill.org --- Created attachment 137160 --> https://bugs.freedesktop.org/attachment.cgi?id=137160&action=edit Xorg.0.log from firing up X directly with Awesome as a WM (Boring)
Finally, Xorg.0.log
https://bugs.freedesktop.org/show_bug.cgi?id=104932
--- Comment #3 from Robin Kauffman robink@creosotehill.org --- Oops, discovered too late that INT3 is used to set a breakpoint. Regardless, it's still something SIGTRAP-ish (i.e. SIGDUMP-CORE-AND-TERMINATE).
https://bugs.freedesktop.org/show_bug.cgi?id=104932
--- Comment #4 from Michel Dänzer michel@daenzer.net --- Which versions of Mesa & LLVM are you using?
https://bugs.freedesktop.org/show_bug.cgi?id=104932
--- Comment #5 from Robin Kauffman robink@creosotehill.org --- (In reply to Michel Dänzer from comment #4)
Which versions of Mesa & LLVM are you using?
LLVM & Clang 7.0 Git, LLVM commit e0c16f05e9fbc1dcd291814ceab9dbc5, Clang commit e0c16f05e9fbc1dcd291814ceab9dbc5. Both were merged 2018/01/31. Let me know if you need more detail.
https://bugs.freedesktop.org/show_bug.cgi?id=104932
--- Comment #6 from Robin Kauffman robink@creosotehill.org --- (In reply to Michel Dänzer from comment #4)
Which versions of Mesa & LLVM are you using?
My sincere apologies, I neglected to include the version of Mesa et al. Unfortunately, it was reading comprehension that failed me, not attention or due diligence. Here's what's installed:
libdrm Git master, commit 9e34ad590e0e1003a597b8cc790a3f36830ba993, merged 1/31/2018. libclc Git master, commit 13541fc0c779cbadbda793c67a0402558ff92ab6, merged 1/31/2018. Mesa Git master, commit bbef9474fa52d9aba06eeede52558fc5ccb762dd, merged 1/31/2018.
Hopefully this now makes my answer complete enough, and again, apologies for the delay.
https://bugs.freedesktop.org/show_bug.cgi?id=104932
--- Comment #7 from Robin Kauffman robink@creosotehill.org --- (In reply to Michel Dänzer from comment #4)
Which versions of Mesa & LLVM are you using?
I should also add that LLVM & Clang were compiled/installed *prior* to merging libclc & Mesa (probably prior to libdrm as well, but in full honesty I can't remember), and that libdrm, libclc & Mesa were merged in that order.
https://bugs.freedesktop.org/show_bug.cgi?id=104932
Robin Kauffman robink@creosotehill.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |WORKSFORME
--- Comment #8 from Robin Kauffman robink@creosotehill.org --- Hi Again- Well, after (repeatedly) breaking the Law of Sysadminning (changing more than one thing at a time), and upgrading userland (including, but sadly not limited to, the 3D stack) as well as the kernel, I have a working desktop, and what little I can divine from where I ended up is that it may well *not* have been RadeonSI/libdrm/AMDGPU at fault to begin with (almost labeled the resolution NOTOURBUG, but the probability that there might have been some slight issue with things actually pertaining to the RadeonSI/AMDGPU graphics driver is to my mind nonzero, even if scant). I unfortunately neglected to bisect the kernel driver (or parts of userland) to see if I could get things working by going back in time and seeing if there was some commit somewhere that broke things (at least for my already-rickety userland). Such a commit may well exist, but going back to try to find it would be a lengthy endeavor, and given that things more-or-less work for me now, the motivation for doing so has all but dried up. If someone wants to follow-up with a likely cause for the kernel complaint I *did* see earlier, by all means do so, but I have a sneaking suspicion that I likely made things Not Work by having a haphazard and semi-out-of-date userland (in this case, *outside* of libdrm/LLVM/Clang/Mesa/etc). Thanks, and apologies for any time sunk on anyone else's behalf trying to suss this one out.
-Robin K.
dri-devel@lists.freedesktop.org
-
bugzilla-daemon＠freedesktop.org