Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

mesa: 18.1.5->18.1.2 [revert] #44383

Closed
wants to merge 2 commits into from
Closed

mesa: 18.1.5->18.1.2 [revert] #44383

wants to merge 2 commits into from

Conversation

mkaito
Copy link
Contributor

@mkaito mkaito commented Aug 2, 2018 •
edited

Motivation for this change

Holy hell, did this ever take debugging. So after nixos-unstable-small was finally unstuck, I eagerly rebased my branch and went onto rebuilding. Just to find that Xorg was segfaulting whenever I got past sddm and into an actual session. See sample log excerpt below.

I took to bisecting the nixpkgs tree, with eternal rebuilds, and eventually isolated the problem: mesa. I know this seems obvious in retrospect, but seeing radeonsi in the trace had me thinking the kernel might be at fault, what with all the new AMDGPU code being merged and me running a fancy Vega card.

Anyway, I tried 18.1.5, found that it did not fix it, and so I reverted that and the bump to 18.1.4 as well. 18.1.2 seems to run fine for me. Let's wait for another release before updating mesa maybe.

Aug 01 23:58:15 cryptbreaker X[1230]: (EE)
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) Backtrace:
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 0: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/bin/X (OsSigHandler+0x29) [0x59e609]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 1: /nix/store/v7hg431d55q30gy7hqlpiji3jnvi8gs3-glibc-2.27/lib/libpthread.so.0 (funlockfile+0x50) [0x7f19a6505f5f]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 2: /run/opengl-driver/lib/dri/radeonsi_dri.so (_mesa_add_parameter+0x223) [0x7f199f104e73]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 3: /run/opengl-driver/lib/dri/radeonsi_dri.so (deserialize_glsl_program+0x5ee) [0x7f199f24e13e]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 4: /run/opengl-driver/lib/dri/radeonsi_dri.so (_Z34shader_cache_read_program_metadataP10gl_contextP17gl_shader_program+0x357) [0x7f199f24f447]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 5: /run/opengl-driver/lib/dri/radeonsi_dri.so (_Z12link_shadersP10gl_contextP17gl_shader_program+0x9d) [0x7f199f1b68ed]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 6: /run/opengl-driver/lib/dri/radeonsi_dri.so (_mesa_glsl_link_shader+0x1ae) [0x7f199f0fd1ce]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 7: /run/opengl-driver/lib/dri/radeonsi_dri.so (link_program_error+0x9d) [0x7f199f01e1fd]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 8: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/lib/xorg/modules/libglamoregl.so (glamor_link_glsl_prog+0xa2) [0x7f19984a8a72]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 9: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/lib/xorg/modules/libglamoregl.so (glamor_build_program+0x4c8) [0x7f19984b6468]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 10: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/lib/xorg/modules/libglamoregl.so (glamor_setup_program_render+0x22b) [0x7f19984b6a0b]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 11: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/lib/xorg/modules/libglamoregl.so (glamor_composite_glyphs+0xb69) [0x7f19984aaf69]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 12: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/bin/X (damageGlyphs+0x3d9) [0x51fae9]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 13: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/bin/X (ProcRenderCompositeGlyphs+0x4fe) [0x51530e]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 14: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/bin/X (Dispatch+0x308) [0x439ae8]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 15: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/bin/X (dix_main+0x3d8) [0x43dab8]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 16: /nix/store/v7hg431d55q30gy7hqlpiji3jnvi8gs3-glibc-2.27/lib/libc.so.6 (__libc_start_main+0xee) [0x7f19a56b7b8e]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) 17: /nix/store/329mvdd6q6xiyq2dp6karcg0yx18n246-xorg-server-1.19.6/bin/X (_start+0x2a) [0x4277da]
Aug 01 23:58:15 cryptbreaker X[1230]: (EE)
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) Segmentation fault at address 0x7f1a0e0e8454
Aug 01 23:58:15 cryptbreaker X[1230]: (EE)
Aug 01 23:58:15 cryptbreaker X[1230]: Fatal server error:
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) Caught signal 11 (Segmentation fault). Server aborting
Aug 01 23:58:15 cryptbreaker X[1230]: (EE)
Aug 01 23:58:15 cryptbreaker X[1230]: Please consult the The X.Org Foundation support
Aug 01 23:58:15 cryptbreaker X[1230]:          at http://wiki.x.org
Aug 01 23:58:15 cryptbreaker X[1230]:  for help.
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) Please also check the log file at "/dev/null" for additional information.
Aug 01 23:58:15 cryptbreaker X[1230]: (EE)
Aug 01 23:58:15 cryptbreaker X[1230]: (II) AIGLX: Suspending AIGLX clients for VT switch
Aug 01 23:58:15 cryptbreaker X[1230]: (EE) Server terminated with error (1). Closing log file.
Things done

I've reverted two commits that bump mesa to 18.1.4 and then to 18.1.5

  • Tested using sandboxing (nix.useSandbox on NixOS, or option sandbox in nix.conf on non-NixOS)
  • Built on platform(s)
    • NixOS
    • macOS
    • other Linux distributions
  • Tested via one or more NixOS test(s) if existing and applicable for the change (look inside nixos/tests)
  • Tested compilation of all pkgs that depend on this change using nix-shell -p nox --run "nox-review wip"
  • Tested execution of all binary files (usually in ./result/bin/)
  • Determined the impact on package closure size (by running nix path-info -S before and after)
  • Fits CONTRIBUTING.md.

Michishige Kaito added 2 commits August 2, 2018 18:52
@lukateras lukateras requested a review from vcunat August 2, 2018 23:26
@matthewbauer
Copy link
Member

Thanks for tracking this down! I think this is okay for now, but eventually we will want to update mesa. We should file an issue in the mesa tracker & maybe someone there will have an idea as to what caused this.

@matthewbauer matthewbauer changed the title mesa: 18.1.5->18.1.2 mesa: 18.1.5->18.1.2 [revert] Aug 3, 2018
@vcunat
Copy link
Member

vcunat commented Aug 3, 2018

Hmm, the AMD situation is getting complex here. amdgpu users actually have problems with 18.1.2 and so far only reverting to 18.0.x helped (and there's how to avoid rebuilds, too): #44183

@mkaito
Copy link
Contributor Author

mkaito commented Aug 3, 2018

I use AMDGPU. In fact, I have little choice. amdgpu-pro hasn't been updated forever and fails to build, so only amdgpu since 4.15 can even run my Vega card.

@mkaito
Copy link
Contributor Author

mkaito commented Aug 3, 2018

I seem to have some issues here and there, such as mpv not being able to use the GPU for decoding, or flickering in Electron apps, but these are not because of 18.1.2. I'd been running 18.1.2 for weeks without an issue, but the recent rebuild of my system off staging seems to have borked something else.

@Chiiruno
Copy link
Contributor

Chiiruno commented Aug 5, 2018 •
edited

It would seem there's no reason to revert now, instead we should look into deleting the mesa cache on upgrade possibly.
See @corngood 's comment near the bottom of #44183 .

@Chiiruno you should be able to use the current mesa just by deleting your cache. To be clear, that's:

sudo rm -rf /root/.cache/mesa_shader_cache
rm -rf ~/.cache/mesa_shader_cache

The problem is, you may have to do it whenever you switch to a new version of mesa, even if you just want to go back from 18.1 to your previously working system config with 18.0. Fixing the cache logic would essentially clear the cache automatically when switching versions.

For anyone hitting this problem, I'd suggest just deleting the cache and moving to 18.1. Hopefully by time there's another breaking update, we'll have a fix in place.

@corngood corngood mentioned this pull request Aug 6, 2018
Merged
9 tasks
@corngood
Copy link
Contributor

corngood commented Aug 6, 2018

@Chiiruno cache fix is in #44575, which should supersede this if there are no problems with it.

@Chiiruno
Copy link
Contributor

Should be fixed by #44575

@Chiiruno
Copy link
Contributor

My bad, whenever it gets merged to master.

@mkaito
Copy link
Contributor Author

mkaito commented Aug 12, 2018 via email

@vcunat
Copy link
Member

vcunat commented Aug 12, 2018

Yeah, and it might explain why different users needed to roll back to different mesa versions.

@mkaito
Copy link
Contributor Author

mkaito commented Aug 13, 2018 •
edited

I've been having some flickering issues in some apps like Slack and Discord, but I'm not sure if that's even related. Everything else works fine, including games.

Nevermind, I'm dumb. Enabling hardware acceleration in those apps fixed it.

@vcunat vcunat closed this Aug 13, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@vcunat vcunat Awaiting requested review from vcunat

Assignees
No one assigned
Labels
10.rebuild-darwin: 1-10 10.rebuild-linux: 501+
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
6 participants
@mkaito @matthewbauer @vcunat @Chiiruno @corngood @GrahamcOfBorg