OBS Studio began to crash after egl-wayland 1.1.14 was released

[LibertyGM]

If I open the "Windowed projector" in OBS Studio, then OBS crashes. If you roll back to version 1.1.13, then everything works.

For more information, see the video

warning: The Wayland connection experienced a fatal error: Protocol error

20240719_072105.webm

[LibertyGM]

Operating System: Arch Linux KDE Plasma Version: 6.1.3 KDE Frameworks Version: 6.4.0 Qt Version: 6.7.2 Kernel Version: 6.10.0-arch1-1 (64-bit) Graphics Platform: Wayland Processors: 32 × Intel® Core™ i9-14900KF Memory: 62.6 ГиБ of RAM Graphics Processor: NVIDIA GeForce RTX 4080/PCIe/SSE2 Manufacturer: ASUS

[thesword53]

I have the same issue with Firefox

[LibertyGM]

In general, my whole system began to work unstable. If you go back to the old version, then everything is OK.

SDDM sometimes freezes when you turn on your computer. There was no such thing in the old version. In general, many programs began to work unstable after the 1.1.14 update.

[joshrandall8478]

I started having this issue as well

[ptr1337]

I have the same issue with Firefox

Firefox is not related to that. This was a firefox issue, which has been partially solved in the nightly. It seems there is an outstanding issue in mesa too.

[thesword53]

I have the same issue with Firefox

Firefox is not related to that. This was a firefox issue, which has been partially solved in the nightly. It seems there is an outstanding issue in mesa too.

I only have the issue with egl-wayland 1.1.14. With the 1.1.13 version I don't have any Firefox crash.

[MadByteDE]

I have the same issue with Firefox

Firefox is not related to that. This was a firefox issue, which has been partially solved in the nightly. It seems there is an outstanding issue in mesa too.

I only have the issue with egl-wayland 1.1.14. With the 1.1.13 version I don't have any Firefox crash.

Same for me with firefox & KDE on Wayland in general (missing icons in windows, freezes). Also might have something to do with the 555 driver and kwin, didn't read to much into it, but when it happens I get:

Jul 20 10:06:27 user-ms7a34 kwin_wayland_wrapper[1593]: src/gbm_drv_common.c:131: GBM-DRV error (get_bytes_per_component): Unknown or not supported format: 875708754
Jul 20 10:06:27 user-ms7a34 kwin_wayland[1593]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.

So I'm not 100% sure which part is actually causing this.

Linux 6.9.9-zen1-1-zen, KDE Plasma 6.1.3, NVIDIA GeForce RTX 4070, driver 555.58.02

After reverting the update the issue disappeared..

[ptr1337]

@thesword53 @MadByteDE

Yes, we know. There is already an open issue for firefox: https://github.com/NVIDIA/egl-wayland/issues/110#issuecomment-2237853758

Firefox does violate the wayland protocol and due it its crashing. This has been partially fixed in firefox nightly, but still crashes. According the firefox developers this is now a mesa issue, which has been reported. Use firefox for now with MOZ_ENABLE_WAYLAND=0 firefox

[LibertyGM]

@ptr1337 What should I do with OBS? It crashes suddenly during the broadcast. And it also crashes with some actions.

[MadByteDE]

I was able to replicate the issue with OBS on my system. Downgrading the egl-wayland package to the previous version fixes the issue.

@ptr1337 What should I do with OBS? It crashes suddenly during the broadcast. And it also crashes with some actions.

On Arch you could try to downgrade the package if you did not clear your package cache: sudo pacman -U file:///var/cache/pacman/pkg/egl-wayland-2:1.1.13-2-x86_64.pkg.tar.zst

[MadByteDE]

@thesword53 @MadByteDE

Yes, we know. There is already an open issue for firefox: #110 (comment)

Firefox does violate the wayland protocol and due it its crashing. This has been partially fixed in firefox nightly, but still crashes. According the firefox developers this is now a mesa issue, which has been reported. Use firefox for now with MOZ_ENABLE_WAYLAND=0 firefox

That issue does not appear to be related to what I'm experiencing. I'm running KDE6/Wayland/Driver 555/Firefox for some time now & had no issues with it. My post was in response to updating to egl-wayland-1.1.14 and seeing different applications (Firefox, OBS, freetube) to not work correctly and especially the plasmashell or parts of it to freeze and bug out a lot.

As mentioned, downgrading egl-wayland fixes the issues for me.

[LibertyGM]

I was able to replicate the issue with OBS on my system. Downgrading the egl-wayland package to the previous version fixes the issue.

@ptr1337 What should I do with OBS? It crashes suddenly during the broadcast. And it also crashes with some actions.

On Arch you could try to downgrade the package if you did not clear your package cache: sudo pacman -U file:///var/cache/pacman/pkg/egl-wayland-2:1.1.13-2-x86_64.pkg.tar.zst

I downgraded using "downgrade". Tell me, when can I expect a solution (correction) to this problem?

[heftig]

Looks like the egl-wayland code always sets up explicit sync for a window but then does not set sync points if a wlStreamResource was set when creating the surface context. Looks like a bug, but I can barely grasp the code.

Ref. https://bugzilla.mozilla.org/show_bug.cgi?id=1908825#c12

[LibertyGM]

Moonlight-qt also freezes when connecting to sunshine in 90% of cases. There is no such thing on the previous version. Apparently this is due to the innovations regarding the updating of the screen.

[thesword53]

I also noticed SDDM crashes

[testsnake]

That issue does not appear to be related to what I'm experiencing. I'm running KDE6/Wayland/Driver 555/Firefox for some time now & had no issues with it. My post was in response to updating to egl-wayland-1.1.14 and seeing different applications (Firefox, OBS, freetube) to not work correctly and especially the plasmashell or parts of it to freeze and bug out a lot.

I think its the same issue discussed in #117, which should be reopened

[mrdev023]

@ptr1337

Do you have mesa issue link about that ?

If it's a mesa bug, it can be also impact other programs like OBS Studio (it's just a guess) because egl-wayland enable explicit-sync and it's not supported by all applications

[joshrandall8478]

This seems to be fixed by switching to the nvidia-open driver, even with using egl-wayland 1.1.14. im still testing it on kde wayland, and i do still have some visual glitches. it’s still much better than before though, obs doesn’t crash at least.

On Mon, Jul 22, 2024 at 12:26 PM, Emma MB @.***(mailto:On Mon, Jul 22, 2024 at 12:26 PM, Emma MB < wrote:

That issue does not appear to be related to what I'm experiencing. I'm running KDE6/Wayland/Driver 555/Firefox for some time now & had no issues with it. My post was in response to updating to egl-wayland-1.1.14 and seeing different applications (Firefox, OBS, freetube) to not work correctly and especially the plasmashell or parts of it to freeze and bug out a lot.

I think its the same issue discussed in #117, which should be reopened

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>

[mrdev023]

@joshrandall8478 the next version (560) of nvidia will be the open version so it's okay.

Maybe you need to make bug report for the graphical glitches

[joshrandall8478]

Potentially. I haven’t noticed any degradations from switching to the open drivers, and games have performed well so far. I don’t know if the graphical degradations are from this package or from KDE in general. I’m waiting a bit to see what happens in the next update wave.

On Mon, Jul 22, 2024 at 2:17 PM, Florian Richer @.***(mailto:On Mon, Jul 22, 2024 at 2:17 PM, Florian Richer < wrote:

@.***(https://github.com/joshrandall8478) the next version (560) of nvidia will be the open version so it's okay.

Maybe you need to make bug report for the graphical glitches

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

[amshafer]

Please include logs for OBS/SDDM/etc crashes, it's helpful for us to figure out who is affected by what. Please also keep the discussion relevant to the reported explicit sync issues.

There is also a fix for a freeze on KDE with explicit sync that is present in the git release of 1.1.14, but is not present in the latest 555 drivers. I'd suggest using the open egl-wayland from github. We also have a prototype fix for a crash on KDE that will eventually be posted here.

Since there's a lot of confusion going around about the crashes, I'll try to summarize it here. This crash was discussed on the linux-drm-syncobj-v1 protocol MR before it was approved. The problem is that some apps (namely firefox, but also others) are committing a wl_surface from one thread at the same time that the wl_surface is being used by the EGL library. This interrupts Mesa/egl-wayland as they are configuring their surface updates, causing a protocol error because no sync points were set yet. As we were told in the original discussion this is an application bug, it is not specific to Mesa or egl-wayland.

The problem is that applications have never been hit with protocol errors for this behavior before, nor was it widely documented. There are a lot of reports out there coming from apps getting bitten by this, and I'd like to be able to see which issues are caused by the app's wl_surface commit synchronization and which issues are new egl-wayland defects that we need to investigate. So please include any logs and especially any protocol errors mentioned so we can track down which projects need fixes, thanks.

[ToadKing]

There's this log from Firefox Nightly, along with the analysis from a Mozilla dev from here:

newtrimmed.log.gz

There's the related part:

[ 982250.261] {Display Queue} wl_display#1.delete_id(89)
[ 982250.295] {Default Queue} discarded wl_buffer#68.release()
[ 982250.304] {Default Queue} wl_callback#89.done(31369649)
[ 982250.314] {Default Queue}  -> wl_surface#63.frame(new id wl_callback#89)
[ 982250.323] {Default Queue}  -> wl_surface#63.commit()
[ 982251.573] {Default Queue}  -> wl_surface#63.attach(wl_buffer#83, 0, 0)

[ 982251.607]  -> wp_linux_drm_syncobj_surface_v1#70.set_acquire_point(wp_linux_drm_syncobj_timeline_v1#71, 0, 877)
[ 982251.619]  -> wp_linux_drm_syncobj_surface_v1#70.set_release_point(wp_linux_drm_syncobj_timeline_v1#84, 0, 188)

[ 982251.630] {Default Queue}  -> wl_surface#63.damage(0, 0, 2560, 1354)
[ 982251.641] {Default Queue}  -> wl_surface#63.commit()
[ 982251.651]  -> wl_display#1.sync(new id wl_callback#59)
[ 982251.779] {Display Queue} wl_display#1.delete_id(59)
[ 982251.795] wl_callback#59.done(16132)
[ 982256.066] {Display Queue} wl_display#1.delete_id(89)
[ 982256.102] {Default Queue} discarded wl_buffer#97.release()
[ 982256.111] {Default Queue} wl_callback#89.done(31369655)
[ 982256.121] {Default Queue}  -> wl_surface#63.frame(new id wl_callback#89)
[ 982256.129] {Default Queue}  -> wl_surface#63.commit()
[ 982257.479] {Default Queue}  -> wl_surface#63.attach(wl_buffer#91, 0, 0)

[ 982257.524]  -> wp_linux_drm_syncobj_surface_v1#70.set_acquire_point(wp_linux_drm_syncobj_timeline_v1#71, 0, 878)
[ 982257.539]  -> wp_linux_drm_syncobj_surface_v1#70.set_release_point(wp_linux_drm_syncobj_timeline_v1#90, 0, 186)

[ 982257.548] {Default Queue}  -> wl_surface#63.damage(0, 0, 2560, 1354)
[ 982257.582] {Default Queue}  -> wl_surface#63.commit()
[ 982257.593]  -> wl_display#1.sync(new id wl_callback#59)
[ 982257.717] {Display Queue} wl_display#1.delete_id(59)
[ 982257.736] wl_callback#59.done(16132)
[ 982260.449] {Display Queue} wl_display#1.delete_id(89)
[ 982260.484] {Default Queue} discarded wl_buffer#83.release()
[ 982260.493] {Default Queue} wl_callback#89.done(31369655)
[ 982260.503] {Default Queue}  -> wl_surface#63.frame(new id wl_callback#89)
[ 982260.512] {Default Queue}  -> wl_surface#63.commit()
[ 982261.739] {Default Queue}  -> wl_surface#63.attach(wl_buffer#68, 0, 0)

[ 982269.936] {Display Queue} wl_display#1.delete_id(89)
[ 982269.974] {Default Queue} wl_callback#89.done(31369667)

[ 982269.985] {Default Queue}  -> wl_surface#63.frame(new id wl_callback#89)
[ 982269.992] {Default Queue}  -> wl_surface#63.commit()
[ 982270.265] {Display Queue} wl_display#1.error(wp_linux_drm_syncobj_surface_v1#70, 4, "explicit sync is used, but no acquire point is set")

So looks like we're missing the set_acquire_point/set_release_point after wl_surface::attach(). But that code doesn't look like Firefox one but rather Mesa. Firefox doesn't use wp_linux_drm_syncobj_surface_v1 at all.

[LibertyGM]

I'm attaching a log file. Below is the team I used to make it.

I have installed egl-wayland-git.

LC_ALL=C obs &> /disks/D/Liberty/Test/log-egl-wayland-git.txt log-egl-wayland-git.txt

[amshafer]

Oh I forgot to mention, enabling WAYLAND_DEBUG=1 for the logs is also helpful.

Thanks, based on your log OBS has the same protocol error we should file an issue with them to synchronize access to wl_surface.

@ToadKing Your example is what I covered in my last comment: something in firefox is doing a wl_surface.commit() that interrupts what Mesa/egl-wayland are doing and triggering the protocol error. You can see the MR that I linked for the gory details. It seems the firefox developers are already aware of this based on the fix they pushed.

[mahkoh]

OBS issue might be https://github.com/obsproject/obs-studio/issues/11022

[Nodens-]

Firefox keeps crashing momentarily after startup even with the fix patch (it has been backported to Fedora 40). With a different error. This only happens with egl-wayland 1.1.13-4 and 1.1.14-1. With 1.1.13-3 it is fine:

Jul 10 16:01:30 synapse systemd[2772]: Started app-org.mozilla.firefox@413815fe031944ccb1e2cf5f8674b83c.service - Firefox - Web Browser.
Jul 10 16:01:31 synapse audit: BPF prog-id=67 op=LOAD
Jul 10 16:01:31 synapse audit: BPF prog-id=68 op=LOAD
Jul 10 16:01:31 synapse audit: BPF prog-id=69 op=LOAD
Jul 10 16:01:31 synapse systemd[1]: Starting systemd-timedated.service - Time & Date Service...
Jul 10 16:01:31 synapse systemd[1]: Started systemd-timedated.service - Time & Date Service.
Jul 10 16:01:31 synapse audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-timedated comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:01:36 synapse rtkit-daemon[1542]: Successfully made thread 10870 of process 10644 (/usr/lib64/firefox/firefox) owned by '1000' RT at priority 10.
Jul 10 16:01:39 synapse firefox[10644]: [Parent 10644, Main Thread] WARNING: pipe error: Bad file descriptor: file /builddir/build/BUILD/firefox-128.0/ipc/chromium/src/chrome/common/ipc_channel_posix.cc:702
Jul 10 16:01:39 synapse firefox[11265]: Exiting due to channel error.
Jul 10 16:01:44 synapse kernel: [drm:nv_drm_semsurf_fence_wait_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00006500] Attempt to wait on invalid sync FD: 235
Jul 10 16:01:56 synapse kwin_wayland[3079]: This plugin does not support raise()
Jul 10 16:01:56 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:56 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:57 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:58 synapse kwin_wayland[3079]: This plugin does not support raise()
Jul 10 16:01:58 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:58 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:58 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:58 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:58 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:59 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:01:59 synapse kwin_wayland[3079]: kwin_scene_opengl: 0x502: GL_INVALID_OPERATION error generated.
Jul 10 16:02:01 synapse firefox[10644]: [GFX1-]: Detect DeviceReset DeviceResetReason::RESET DeviceResetDetectPlace::WR_POST_UPDATE in Parent process
Jul 10 16:02:01 synapse firefox[10644]: [GFX1-]: Failed to make render context current during destroying.
Jul 10 16:02:05 synapse systemd[1]: systemd-timedated.service: Deactivated successfully.
Jul 10 16:02:05 synapse audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-timedated comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:02:05 synapse audit: BPF prog-id=69 op=UNLOAD
Jul 10 16:02:05 synapse audit: BPF prog-id=68 op=UNLOAD
Jul 10 16:02:05 synapse audit: BPF prog-id=67 op=UNLOAD
Jul 10 16:02:37 synapse wireplumber[3069]: wp-event-dispatcher: <WpAsyncEventHook:0x56383300bbc0> failed: <WpSiStandardLink:0x563833352840> link failed: 1 of 1 PipeWire links failed to activate
Jul 10 16:02:38 synapse firefox[10644]: [Parent 10644, Compositor] WARNING: Call to mmap failed: Bad file descriptor: file /builddir/build/BUILD/firefox-128.0/ipc/chromium/src/base/shared_memory_posix.cc:515
Jul 10 16:02:38 synapse firefox[10644]: ###!!! [Parent][DispatchAsyncMessage] Error: SHMEM_CREATED_MESSAGE Payload error: message could not be deserialized
Jul 10 16:03:06 synapse systemd[2772]: Starting gvfs-metadata.service - Virtual filesystem metadata service...
Jul 10 16:03:06 synapse systemd[2772]: Started gvfs-metadata.service - Virtual filesystem metadata service.
Jul 10 16:03:36 synapse kernel: [drm:nv_drm_semsurf_fence_wait_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00006500] Attempt to wait on invalid sync FD: 226
Jul 10 16:03:53 synapse firefox[10644]: [GFX1-]: Detect DeviceReset DeviceResetReason::RESET DeviceResetDetectPlace::WR_POST_UPDATE in Parent process
Jul 10 16:03:53 synapse firefox[10644]: [GFX1-]: Failed to make render context current during destroying.
Jul 10 16:04:07 synapse firefox[10644]: [Parent 10644, Compositor] WARNING: Call to mmap failed: Bad file descriptor: file /builddir/build/BUILD/firefox-128.0/ipc/chromium/src/base/shared_memory_posix.cc:515
Jul 10 16:04:07 synapse firefox[10644]: ###!!! [Parent][DispatchAsyncMessage] Error: SHMEM_CREATED_MESSAGE Payload error: message could not be deserialized
Jul 10 16:04:50 synapse firefox[10644]: [Parent 10644, Compositor] WARNING: Call to mmap failed: Bad file descriptor: file /builddir/build/BUILD/firefox-128.0/ipc/chromium/src/base/shared_memory_posix.cc:515
Jul 10 16:04:50 synapse firefox[10644]: ###!!! [Parent][DispatchAsyncMessage] Error: SHMEM_CREATED_MESSAGE Payload error: message could not be deserialized
Jul 10 16:04:53 synapse audit[11322]: ANOM_ABEND auid=1000 uid=1000 gid=1000 ses=3 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=11322 comm="WebExtensions" exe="/usr/lib64/firefox/firefox" sig=7 res=1
Jul 10 16:04:53 synapse systemd[1]: Created slice system-drkonqi\x2dcoredump\x2dprocessor.slice - Slice /system/drkonqi-coredump-processor.
Jul 10 16:04:53 synapse systemd[1]: Created slice system-systemd\x2dcoredump.slice - Slice /system/systemd-coredump.
Jul 10 16:04:53 synapse audit: BPF prog-id=70 op=LOAD
Jul 10 16:04:53 synapse audit: BPF prog-id=71 op=LOAD
Jul 10 16:04:53 synapse audit: BPF prog-id=72 op=LOAD
Jul 10 16:04:53 synapse systemd[1]: Started systemd-coredump@0-11966-0.service - Process Core Dump (PID 11966/UID 0).
Jul 10 16:04:53 synapse audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@0-11966-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:04:53 synapse audit: BPF prog-id=73 op=LOAD
Jul 10 16:04:53 synapse audit: BPF prog-id=74 op=LOAD
Jul 10 16:04:53 synapse audit: BPF prog-id=75 op=LOAD
Jul 10 16:04:53 synapse systemd[1]: Started drkonqi-coredump-processor@0-11966-0.service - Pass systemd-coredump journal entries to relevant user for potential DrKonqi handling.
Jul 10 16:04:53 synapse audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=drkonqi-coredump-processor@0-11966-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:04:57 synapse systemd-coredump[11967]: Process 11322 (WebExtensions) of user 1000 dumped core.#012#012Module libpcsclite.so.1 from rpm pcsc-lite-2.0.3-1.fc40.x86_64#012Module legacy.so from rpm openssl-3.2.1-2.fc40.x86_64#012Module libcrypto.so.3 from rpm openssl-3.2.1-2.fc40.x86_64#012Module libopensc.s$
Jul 10 16:04:57 synapse systemd[1]: systemd-coredump@0-11966-0.service: Deactivated successfully.
Jul 10 16:04:57 synapse audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@0-11966-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:04:57 synapse systemd[1]: systemd-coredump@0-11966-0.service: Consumed 4.478s CPU time.
Jul 10 16:04:57 synapse audit: BPF prog-id=72 op=UNLOAD
Jul 10 16:04:57 synapse audit: BPF prog-id=71 op=UNLOAD
Jul 10 16:04:57 synapse audit: BPF prog-id=70 op=UNLOAD
Jul 10 16:04:58 synapse firefox[10644]: [Parent 10644, IPC I/O Parent] WARNING: process 11322 exited on signal 7: file /builddir/build/BUILD/firefox-128.0/ipc/chromium/src/base/process_util_posix.cc:335
Jul 10 16:04:58 synapse systemd[1]: drkonqi-coredump-processor@0-11966-0.service: Deactivated successfully.
Jul 10 16:04:58 synapse audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=drkonqi-coredump-processor@0-11966-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:04:58 synapse systemd[2772]: Started drkonqi-coredump-launcher@69-11968-0.service - Launch DrKonqi for a systemd-coredump crash (PID 11968/UID 0).
Jul 10 16:04:58 synapse drkonqi-coredump-launcher[12055]: Unable to find file for pid 11322 expected at "kcrash-metadata/firefox.dbc88fab335b4d8eadf2c279520fbb33.11322.ini"
Jul 10 16:04:58 synapse drkonqi-coredump-launcher[12055]: Nothing handled the dump :O
Jul 10 16:04:58 synapse audit: BPF prog-id=75 op=UNLOAD
Jul 10 16:04:58 synapse audit: BPF prog-id=74 op=UNLOAD
Jul 10 16:04:58 synapse audit: BPF prog-id=73 op=UNLOAD
Jul 10 16:04:58 synapse abrt-server[12074]: Package 'firefox' isn't signed with proper key
Jul 10 16:04:58 synapse abrt-server[12074]: 'post-create' on '/var/spool/abrt/ccpp-2024-07-10-16:04:58.2594-11322' exited with 1
Jul 10 16:04:58 synapse abrt-server[12074]: Deleting problem directory '/var/spool/abrt/ccpp-2024-07-10-16:04:58.2594-11322'
Jul 10 16:05:07 synapse firefox[11554]: [11554] Sandbox: Unexpected EOF, op 0 flags 00 path /usr/share/fonts/wine-times-new-roman-fonts/times.ttf
Jul 10 16:05:07 synapse firefox[10644]: [10644] Sandbox: failed to send broker response to pid 11554: EBADF
Jul 10 16:05:10 synapse firefox[12032]: [Child 12032, Main Thread] WARNING: Call to mmap failed: Permission denied: file /builddir/build/BUILD/firefox-128.0/ipc/chromium/src/base/shared_memory_posix.cc:515
Jul 10 16:05:10 synapse firefox[12032]: IPDL protocol error: SharedMemory::Map failed
Jul 10 16:05:10 synapse audit[12032]: ANOM_ABEND auid=1000 uid=1000 gid=1000 ses=3 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12032 comm="WebExtensions" exe="/usr/lib64/firefox/firefox" sig=11 res=1
Jul 10 16:05:10 synapse kernel: show_signal_msg: 52 callbacks suppressed
Jul 10 16:05:10 synapse kernel: WebExtensions[12032]: segfault at 0 ip 00007f91c214c5ca sp 00007ffe3fcc1e80 error 6 in libxul.so[7f91c1663000+4dfb000] likely on CPU 15 (core 25, socket 0)
Jul 10 16:05:10 synapse kernel: Code: d4 05 49 89 04 24 31 d2 89 14 25 00 00 00 00 0f 0b 48 8d 35 f6 61 9d 05 48 89 e7 e8 d0 c8 bb 01 48 8b 04 24 49 89 04 24 31 c0 <89> 04 25 00 00 00 00 0f 0b f3 0f 1e fa 48 89 f8 48 85 f6 74 0f 48
Jul 10 16:05:10 synapse audit: BPF prog-id=76 op=LOAD
Jul 10 16:05:10 synapse audit: BPF prog-id=77 op=LOAD
Jul 10 16:05:10 synapse audit: BPF prog-id=78 op=LOAD
Jul 10 16:05:10 synapse systemd[1]: Started systemd-coredump@1-12162-0.service - Process Core Dump (PID 12162/UID 0).
Jul 10 16:05:10 synapse audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@1-12162-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:05:10 synapse audit: BPF prog-id=79 op=LOAD
Jul 10 16:05:10 synapse audit: BPF prog-id=80 op=LOAD
Jul 10 16:05:10 synapse audit: BPF prog-id=81 op=LOAD
Jul 10 16:05:10 synapse systemd[1]: Started drkonqi-coredump-processor@1-12162-0.service - Pass systemd-coredump journal entries to relevant user for potential DrKonqi handling.
Jul 10 16:05:10 synapse audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=drkonqi-coredump-processor@1-12162-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:05:14 synapse systemd-coredump[12163]: Process 12032 (WebExtensions) of user 1000 dumped core.#012#012Module libpcsclite.so.1 from rpm pcsc-lite-2.0.3-1.fc40.x86_64#012Module legacy.so from rpm openssl-3.2.1-2.fc40.x86_64#012Module libcrypto.so.3 from rpm openssl-3.2.1-2.fc40.x86_64#012Module libopensc.s$
Jul 10 16:05:14 synapse systemd[1]: systemd-coredump@1-12162-0.service: Deactivated successfully.
Jul 10 16:05:14 synapse audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@1-12162-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:05:14 synapse systemd[1]: systemd-coredump@1-12162-0.service: Consumed 4.043s CPU time.
Jul 10 16:05:14 synapse audit: BPF prog-id=78 op=UNLOAD
Jul 10 16:05:14 synapse audit: BPF prog-id=77 op=UNLOAD
Jul 10 16:05:14 synapse audit: BPF prog-id=76 op=UNLOAD
Jul 10 16:05:15 synapse firefox[10644]: [Parent 10644, IPC I/O Parent] WARNING: waitid failed pid:12032 errno:10: file /builddir/build/BUILD/firefox-128.0/ipc/chromium/src/base/process_util_posix.cc:314
Jul 10 16:05:15 synapse systemd[1]: drkonqi-coredump-processor@1-12162-0.service: Deactivated successfully.
Jul 10 16:05:15 synapse audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=drkonqi-coredump-processor@1-12162-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 10 16:05:15 synapse systemd[2772]: Started drkonqi-coredump-launcher@70-12164-0.service - Launch DrKonqi for a systemd-coredump crash (PID 12164/UID 0).
Jul 10 16:05:15 synapse drkonqi-coredump-launcher[12219]: Unable to find file for pid 12032 expected at "kcrash-metadata/firefox.dbc88fab335b4d8eadf2c279520fbb33.12032.ini"
Jul 10 16:05:15 synapse drkonqi-coredump-launcher[12219]: Nothing handled the dump :O
Jul 10 16:05:15 synapse audit: BPF prog-id=81 op=UNLOAD
Jul 10 16:05:15 synapse audit: BPF prog-id=80 op=UNLOAD
Jul 10 16:05:15 synapse audit: BPF prog-id=79 op=UNLOAD
Jul 10 16:05:15 synapse abrt-server[12241]: Package 'firefox' isn't signed with proper key
Jul 10 16:05:15 synapse abrt-server[12241]: 'post-create' on '/var/spool/abrt/ccpp-2024-07-10-16:05:15.92121-12032' exited with 1
Jul 10 16:05:15 synapse abrt-server[12241]: Deleting problem directory '/var/spool/abrt/ccpp-2024-07-10-16:05:15.92121-12032'
Jul 10 16:05:49 synapse clamd[2037]: SelfCheck: Database status OK.
Jul 10 16:05:53 synapse firefox[10644]: [Parent 10644, Compositor] WARNING: Call to mmap failed: Bad file descriptor: file /builddir/build/BUILD/firefox-128.0/ipc/chromium/src/base/shared_memory_posix.cc:515
Jul 10 16:05:53 synapse firefox[10644]: ExceptionHandler::GenerateDump attempting to generate:
Jul 10 16:05:53 synapse firefox[10644]: /home/nodens/.mozilla/firefox/7xspfc8h.default-release/minidumps/58773d9f-dc57-b76d-3186-826bed384616.dmp
Jul 10 16:05:53 synapse firefox[10644]: ExceptionHandler::GenerateDump cloned child 12399
Jul 10 16:05:53 synapse firefox[10644]: ExceptionHandler::SendContinueSignalToChild sent continue signal to child
Jul 10 16:05:53 synapse firefox[12399]: ExceptionHandler::WaitForContinueSignal waiting for continue signal...
Jul 10 16:05:53 synapse firefox[10644]: ExceptionHandler::GenerateDump minidump generation
Jul 10 16:05:53 synapse firefox[10644]: succeeded
Jul 10 16:05:53 synapse systemd[2772]: app-org.mozilla.firefox@413815fe031944ccb1e2cf5f8674b83c.service: Main process exited, code=exited, status=11/n/a
Jul 10 16:05:53 synapse firefox[11811]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[11453]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[11984]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[12357]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[11740]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[12202]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[11554]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[10909]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[10925]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[10944]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[11054]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[11078]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[10985]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[10899]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[11106]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[10903]: Exiting due to channel error.
Jul 10 16:05:53 synapse firefox[10836]: Exiting due to channel error.
Jul 10 16:06:05 synapse systemd[2772]: app-org.mozilla.firefox@413815fe031944ccb1e2cf5f8674b83c.service: Failed with result 'exit-code'.
Jul 10 16:06:05 synapse systemd[2772]: app-org.mozilla.firefox@413815fe031944ccb1e2cf5f8674b83c.service: Consumed 5min 24.310s CPU time, 3.5G memory peak, 0B memory swap peak.

[amshafer]

What kernel version is this firefox bad fd crash with? Can you please provide details about that system?

Note that 6.8.x kernels have some DRM syncobj fixes that help explicit sync. Some LTS kernels may have had them backported and work fine but it can be hard to know for sure. The fixes are mentioned in: https://download.nvidia.com/XFree86/Linux-x86_64/560.28.03/README/wayland-issues.html

[Nodens-]

Not running an old kernel, unfortunately.

[$]> inxi -b
System:
  Host: redacted Kernel: 6.9.9-200.fc40.x86_64 arch: x86_64
    bits: 64
  Desktop: KDE Plasma v: 6.1.3 Distro: Fedora Linux 40 (KDE Plasma)
Machine:
  Type: Desktop Mobo: ASUSTeK model: ROG RAMPAGE VI EXTREME ENCORE v: Rev 1.xx
    serial: <superuser required> UEFI: American Megatrends v: 1601
    date: 08/28/2023
CPU:
  Info: 18-core Intel Core i9-10980XE [MT MCP] speed (MHz): avg: 1350
    min/max: 1200/4600:4700:4800
Graphics:
  Device-1: NVIDIA GA102 [GeForce RTX 3080] driver: nvidia v: 555.58.02
  Device-2: Logitech C922 Pro Stream Webcam driver: snd-usb-audio,uvcvideo
    type: USB
  Display: wayland server: X.org v: 1.20.14 with: Xwayland v: 24.1.1
    compositor: kwin_wayland driver: X: loaded: modesetting unloaded: fbdev,vesa
    dri: nouveau gpu: nvidia resolution: 1: 2560x1440 2: 2560x1440
  API: OpenGL v: 4.6.0 compat-v: 4.5 vendor: nvidia mesa v: 555.58.02
    renderer: NVIDIA GeForce RTX 3080/PCIe/SSE2
Network:
  Device-1: Intel Wi-Fi 6 AX200 driver: iwlwifi
  Device-2: Intel Ethernet X710 for 10GbE SFP+ driver: i40e
  Device-3: Intel Ethernet X710 for 10GbE SFP+ driver: i40e
  Device-4: Microsoft Xbox Wireless Adapter for Windows driver: xone-dongle
    type: USB
Drives:
  Local Storage: total: 10.91 TiB used: 10.17 TiB (93.2%)
Info:
  Memory: total: 128 GiB note: est. available: 125.47 GiB
    used: 30.47 GiB (24.3%)
  Processes: 802 Uptime: 5d 41m Shell: Bash inxi: 3.3.34

Let me know if you need any more information.

EDIT: Forgot to mention I am running the open kernel module with GSP Firmware enabled

[Nodens-]

Additional information. The crash happens also on 560.28.3.0 as it is currently packaged in rpmfusion. And what I mean by that is that they have dropped the external egl-wayland and egl-gbm package dependencies and they are packaging the versions included in the official 560.28.3.0 driver release.

[mahkoh]

This is just the same old bug that was not properly fixed in firefox.

[ 543672.346] {mesa vk display queue}  -> wp_linux_drm_syncobj_surface_v1#69.set_acquire_point(wp_linux_drm_syncobj_timeline_v1#95, 0, 60)
<--> missing synchronization with presentation
[ 543672.360] {Display Queue} wl_display#1.delete_id(56)
[ 543672.364] {Default Queue} wl_callback#56.done(328632845)
[ 543672.367] {Default Queue}  -> wl_surface#54.frame(new id wl_callback#56)
[ 543672.370] {Default Queue}  -> wl_surface#54.commit()
<--> missing synchronization with presentation
[ 543672.381] {mesa vk display queue}  -> wp_linux_drm_syncobj_surface_v1#69.set_release_point(wp_linux_drm_syncobj_timeline_v1#96, 0, 60)
[ 543672.387] {mesa vk display queue}  -> wl_surface#54.attach(wl_buffer#94, 0, 0)
[ 543672.399] {mesa vk display queue}  -> wl_surface#54.damage(0, 0, 2147483647, 2147483647)
[ 543672.402] {mesa vk display queue}  -> wl_surface#54.commit()

[joshrandall8478]

So if you install 560, you should remove egl-wayland and egl-gbm?

On Fri, Jul 26, 2024 at 6:33 AM, Nodens- @.***(mailto:On Fri, Jul 26, 2024 at 6:33 AM, Nodens- < wrote:

Additional information. The crash happens also on 560.28.3.0 as it is currently packaged in rpmfusion. And what I mean by that is that they have dropped the external egl-wayland and egl-gbm packages and they are packaging the versions included in the official 560.28.3.0 driver release.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

[ptr1337]

joshrandall8478 commented Jul 26, 2024 via email

No, using egl-wayland 1.14 does work also fine. When using egl-wayland 1.13 OBS for example does not show any input anymore and other issues.

[joshrandall8478]

egl-wayland was rolled back to 1.1.13 on arch due to the issues outlined in this thread. i tried 1.1.14 on the 560 beta (open and proprietary), and OBS was still crashing when i wanted to use window project. Im currently staying on proprietary 555 and 1.1.13 until significant fixes and changes are made.

On Fri, Jul 26, 2024 at 8:13 AM, Peter Jung @.***(mailto:On Fri, Jul 26, 2024 at 8:13 AM, Peter Jung < wrote:

joshrandall8478 commented Jul 26, 2024 via email

No, using egl-wayland 1.14 does work also fine. When using egl-wayland 1.13 OBS for example does not show any input anymore and other issues.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

[hyyoxhk]

I have the same issue with Weston and Qt app

......
[23:53:38.109] Loading module '/usr/local/lib/x86_64-linux-gnu/weston/desktop-shell.so'
[23:53:38.112] launching '/usr/local/libexec/weston-keyboard'
[23:53:38.114] launching '/usr/local/libexec/weston-desktop-shell'
[destroyed object]: error 7: failed to import supplied dmabufs: Arguments are inconsistent (for example, a valid context requires buffers not supplied by a 
[23:53:38.215] panel surface gone
[23:53:38.215] background surface gone
......

Warning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.
QSocketNotifier: Can only be used with threads started with QThread
The Wayland connection experienced a fatal error: Protocol error
Aborted (core dumped)

[hyyoxhk]

I have the same issue with Weston and Qt app

......
[23:53:38.109] Loading module '/usr/local/lib/x86_64-linux-gnu/weston/desktop-shell.so'
[23:53:38.112] launching '/usr/local/libexec/weston-keyboard'
[23:53:38.114] launching '/usr/local/libexec/weston-desktop-shell'
[destroyed object]: error 7: failed to import supplied dmabufs: Arguments are inconsistent (for example, a valid context requires buffers not supplied by a 
[23:53:38.215] panel surface gone
[23:53:38.215] background surface gone
......

Warning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.
QSocketNotifier: Can only be used with threads started with QThread
The Wayland connection experienced a fatal error: Protocol error
Aborted (core dumped)

I solved this issue after updating to version 1.1.12

[Gert-dev]

@amshafer FYI, Martin Stránský posted about his debugging on Firefox today with egl-wayland 1.1.14, still thinks this is related to egl-wayland, and requested someone report his detailed findings back here.

(Hope this helps.)

[amshafer]

Testing of the latest fix would be appreciated as it's been difficult to reproduce every reported crash myself. We've confirmed this fixes some of this issues with firefox but there are still some bad fd crashes the firefox devs are looking into. I'm also curious if this helps anyone else running Qt/OBS/etc. Reproducing crashes has been very intermittent for me so if anyone with more consistent repros can confirm that would be helpful.

[huyz-git]

Testing of the latest fix would be appreciated as it's been difficult to reproduce every reported crash myself. We've confirmed this fixes some of this issues with firefox but there are still some bad fd crashes the firefox devs are looking into. I'm also curious if this helps anyone else running Qt/OBS/etc. Reproducing crashes has been very intermittent for me so if anyone with more consistent repros can confirm that would be helpful.

On both https://github.com/amshafer/egl-wayland/commit/a5182c7390a78ca2f7986cbcd2e1bf38f6be5f47 and 1.1.14, I still get errors on fd related things on moonlight:

Assertion 'close_nointr(fd) != -EBADF' failed at src/basic/fd-util.c:75, function safe_close(). Aborting.
[1]    7856 IOT instruction (core dumped)  moonlight

On 1.1.13.1 no such errors occurred. and this issue can be consistently reproduced in my case. I'm not sure whether this is egl-wayland's issue or moonlight's issue. I'm using 560.28.03 driver.

[amshafer]

Thanks, the badfd errors happening on both firefox and moonlight makes me think this is an egl-wayland issue. I'm guessing it's due to running past the fd limit but am trying to verify the root cause.

[Nodens-]

That would be a bit strange on my system as I'm running with:

[$]> ulimit -a
real-time non-blocking time  (microseconds, -R) unlimited
core file size              (blocks, -c) unlimited
data seg size               (kbytes, -d) unlimited
scheduling priority                 (-e) 0
file size                   (blocks, -f) unlimited
pending signals                     (-i) 513601
max locked memory           (kbytes, -l) 8192
max memory size             (kbytes, -m) unlimited
open files                          (-n) 1048576
pipe size                (512 bytes, -p) 8
POSIX message queues         (bytes, -q) 819200
real-time priority                  (-r) 0
stack size                  (kbytes, -s) 8192
cpu time                   (seconds, -t) unlimited
max user processes                  (-u) 513601
virtual memory              (kbytes, -v) unlimited
file locks                          (-x) unlimited

[$]> cat /proc/sys/fs/file-max
9223372036854775807

[Nodens-]

I also just tested https://github.com/NVIDIA/egl-wayland/pull/123 with 555 driver and Martin's extra debug logging firefox build and it made no difference for the fd crash.

[huyz-git]

Thanks, the badfd errors happening on both firefox and moonlight makes me think this is an egl-wayland issue. I'm guessing it's due to running past the fd limit but am trying to verify the root cause.

After apply https://github.com/NVIDIA/egl-wayland/pull/124 , the fd error on moonlight is gone in my case.

[VictorR2007]

After apply #124 , the fd error on moonlight is gone in my case.

It seems that with the application of patch #124, Firefox Nightly has started to work much better. In half an hour, he never fell down.

[cognitus]

no more crash after last commit

[NoTuxNoBux]

Also confirming the situation has notably improved since 1.1.15. Arch has since shipped the package in the stable repositories.

Am I correct in assuming that for Flatpak applications we now have to wait for a new NVIDIA driver that includes 1.1.15 since org.freedesktop.Platform.GL.nvidia extracts it from there?

[LibertyGM]

1.1.15

The error has been preserved. The problem has remained as it was.

error: v4l2-input: /dev/video2: select timed out
error: v4l2-input: /dev/video2: failed to log status
warning: [obs-browser: 'DonatePay'] Error: Failed to execute 'postMessage' on 'DOMWindow': The target origin provided ('https://www.youtube.com') does not match the recipient window's origin ('https://widget.donatepay.ru'). (https://www.youtube.com/s/player/1c78e434/www-widgetapi.vflset/www-widgetapi.js:211)
warning: [pipewire] Failed to start screencast, denied or cancelled by user
warning: [pipewire] Failed to start screencast, denied or cancelled by user
wp_linux_drm_syncobj_surface_v1#63: error 4: explicit sync is used, but no acquire point is set
warning: The Wayland connection experienced a fatal error: Protocol error

https://github.com/user-attachments/assets/ab94f6d4-1aef-49df-92f5-d8e3e30a7df9

[mrdev023]

1.1.15

The error has been preserved. The problem has remained as it was.

error: v4l2-input: /dev/video2: select timed out
error: v4l2-input: /dev/video2: failed to log status
warning: [obs-browser: 'DonatePay'] Error: Failed to execute 'postMessage' on 'DOMWindow': The target origin provided ('https://www.youtube.com') does not match the recipient window's origin ('https://widget.donatepay.ru'). (https://www.youtube.com/s/player/1c78e434/www-widgetapi.vflset/www-widgetapi.js:211)
warning: [pipewire] Failed to start screencast, denied or cancelled by user
warning: [pipewire] Failed to start screencast, denied or cancelled by user
wp_linux_drm_syncobj_surface_v1#63: error 4: explicit sync is used, but no acquire point is set
warning: The Wayland connection experienced a fatal error: Protocol error

bug.mp4

And with __NV_DISABLE_EXPLICIT_SYNC=1 ? i see the developper add this env var

[LibertyGM]

And with __NV_DISABLE_EXPLICIT_SYNC=1 ? i see the developper add this env var

How exactly do I use this variable? Where should I add it? For egl-wayland or the application I'm running.

[joshrandall8478]

How exactly do I use this variable? Where should I add it? For egl-wayland or the application I'm running.

Regardless it's not a solution. Explicit sync is necessary for xwayland applications to run without stuttering. that being said if applying the environment variable to the application doesn't work, apply it to /etc/environment and see what happens.

[LibertyGM]

How exactly do I use this variable? Where should I add it? For egl-wayland or the application I'm running.

Regardless it's not a solution. Explicit sync is necessary for xwayland applications to run without stuttering. that being said if applying the environment variable to the application doesn't work, apply it to /etc/environment and see what happens.

I am running an application with a variable, for example:

__NV_DISABLE_EXPLICIT_SYNC=1 obs

And everything works fine.