Closed Bug 1784813 Opened 2 years ago Closed 1 year ago

Gnome Xwayland 2:1.20.11-1+rpt1+deb11u3,Raspberry Pi 4 OS (bullseye): Intermittent startup crash in __GI_getenv ("GLVND_DISALLOW_PATCHING") at getenv.c:84 with address 0xe5e5e5e5e5e5e5e5. Seems fixed with Debian testing/bookworm (xwayland 22.1.5).

Categories

(Core :: Graphics, defect)

ARM64
Linux
defect

Tracking

()

RESOLVED WORKSFORME
Tracking Status
firefox105 --- wontfix
firefox109 --- affected

People

(Reporter: jan, Unassigned)

References

(Blocks 1 open bug)

Details

(Keywords: crash, nightly-community)

Crash Data

I'm filing this bug because this crash address looks bad:

Crash Address 0xe5e5e5e5e5e5e5e5

(Martin Stránský [:stransky] (ni? me) from bug 1656415 comment #7)

0xe5e5e5e5e5e5e5e5 means already released/freed memory.

This crash occured with try build from bug 1738814 comment 35
on Gnome Wayland on 64 bit Raspberry Pi OS (Debian bullseye) + apt install -t testing libegl-mesa0 --no-install-recommends (Mesa 22.2.0~rc1) from bookworm
after closing Firefox and starting it again via ./firefox -P 2022-08-14-1738814c35.

This Nightly uses Xwayland because libgtk-3-common 3.24.24-4+rpt4+deb11u2 is lower than Nightly's Wayland min version limit of 3.24.30.
I don't have set the MOZ_ENABLE_WAYLAND=1 env var.

This is the second time this crash occurs.
The first time was 3 days ago with a different build: bp-e0118d4f-03a1-4b4f-be2d-943660220812


Crash report: https://crash-stats.mozilla.org/report/index/09f08039-c9d7-4704-a29c-834a00220814

Reason: SIGSEGV / SEGV_MAPERR

Top 10 frames of crashing thread:

0 libc.so.6 libc.so.6@0x000000000003653c 
1 libc.so.6 libc.so.6@0x0000000000036518 
2 libGLdispatch.so.0 libGLdispatch.so.0@0x0000000000040cd8 
3 libGLdispatch.so.0 libGLdispatch.so.0@0x0000000000040ec4 
4 libEGL.so.1 libEGL.so.1@0x0000000000003274 
5 libxul.so libxul.so@0x00000000038e9bb8 
6 libxul.so libxul.so@0x00000000038c9f14 
7 libxul.so libxul.so@0x00000000038b89a4 
8 libxul.so libxul.so@0x00000000038bdf14 
9 libxul.so libxul.so@0x00000000038c9cc8 

There's a high chance that this is another Mesa bug, but having symbols would help a lot. Can you install debug symbols for the Mesa packages?

As for FF, here's a try build with "upload symbols", hope it works: https://treeherder.mozilla.org/jobs?repo=try&revision=3bd1b1d443d00db4769e185f46e4cf1eb11388d7

Same build as in comment 0:

pi@raspberrypi:~/Downloads/2022-08-14-1738814c35/target/firefox $ gdb --args ./firefox -P 2022-08-14-1738814c35
GNU gdb (Debian 10.1-1.7) 10.1.90.20210103-git
Copyright (C) 2021 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Type "show copying" and "show warranty" for details.
This GDB was configured as "aarch64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<https://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
    <http://www.gnu.org/software/gdb/documentation/>.

For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from ./firefox...
(No debugging symbols found in ./firefox)
(gdb) r
Starting program: /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/firefox -P 2022-08-14-1738814c35
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/aarch64-linux-gnu/libthread_db.so.1".
process 15525 is executing new program: /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/firefox-bin
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/aarch64-linux-gnu/libthread_db.so.1".
[New Thread 0x7ff77ff100 (LWP 15528)]
[Thread 0x7ff77ff100 (LWP 15528) exited]
[Detaching after fork from child process 15529]
[Detaching after fork from child process 15530]
[Detaching after vfork from child process 15531]
[Detaching after vfork from child process 15532]
[Detaching after vfork from child process 15533]
[New Thread 0x7ff77ff100 (LWP 15534)]
[New Thread 0x7febdff100 (LWP 15535)]
[Detaching after vfork from child process 15536]
[New Thread 0x7feb59d100 (LWP 15537)]
[New Thread 0x7fe6db9100 (LWP 15538)]
[New Thread 0x7fe6d69100 (LWP 15539)]
[New Thread 0x7f66fff100 (LWP 15540)]
[New Thread 0x7f66faf100 (LWP 15541)]
[New Thread 0x7f66f5f100 (LWP 15542)]
[New Thread 0x7f66dfd100 (LWP 15543)]
[New Thread 0x7f66bff100 (LWP 15544)]
[Thread 0x7f66dfd100 (LWP 15543) exited]
get_gles_status useGles: 1
[New Thread 0x7f66dad100 (LWP 15546)]
[New Thread 0x7f66d5d100 (LWP 15547)]
get_gles_status versionString: OpenGL ES 3.1 Mesa 22.2.0-rc1
[New Thread 0x7f669ef100 (LWP 15548)]
[Detaching after fork from child process 15549]
[New Thread 0x7feb54d100 (LWP 15550)]
[New Thread 0x7f661a1100 (LWP 15551)]
[New Thread 0x7f65721100 (LWP 15552)]
[New Thread 0x7f65513100 (LWP 15553)]
[New Thread 0x7f65305100 (LWP 15554)]
[New Thread 0x7f650f7100 (LWP 15555)]
[GFX1-]: No GPUs detected via PCI
[Thread 0x7f669ef100 (LWP 15548) exited]
[New Thread 0x7f645ff100 (LWP 15556)]
[New Thread 0x7f645af100 (LWP 15557)]
[New Thread 0x7f6451f100 (LWP 15558)]
[New Thread 0x7f669ef100 (LWP 15559)]
[New Thread 0x7f644cf100 (LWP 15560)]
[New Thread 0x7f640ff100 (LWP 15561)]
[New Thread 0x7f63cff100 (LWP 15562)]
[New Thread 0x7f638ff100 (LWP 15563)]
[New Thread 0x7f636ef100 (LWP 15564)]
[New Thread 0x7f632ff100 (LWP 15565)]
[New Thread 0x7f630ef100 (LWP 15566)]
[New Thread 0x7f642bf100 (LWP 15567)]
[New Thread 0x7f63ea1100 (LWP 15568)]
ATTENTION: default value of option mesa_glthread overridden by environment.
[New Thread 0x7f576a6100 (LWP 15570)]
Thread 24 "Renderer" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7f6451f100 (LWP 15558)]
__GI_getenv (name=0x7f62e9509a "GLVND_DISALLOW_PATCHING") at getenv.c:84
84	getenv.c: No such file or directory.
(gdb) bt full
#0  __GI_getenv (name=0x7f62e9509a "GLVND_DISALLOW_PATCHING") at getenv.c:84
Python Exception <class 'gdb.MemoryError'> Cannot access memory at address 0xe5e5e5e5e5e5e5e5: 
        ep_start = #1  0x0000007f62d95cdc in  () at /usr/lib/aarch64-linux-gnu/libGLdispatch.so.0
#2  0x0000007f62d95ec8 in __glDispatchMakeCurrent () at /usr/lib/aarch64-linux-gnu/libGLdispatch.so.0
#3  0x0000007f64705278 in  () at /usr/lib/aarch64-linux-gnu/libEGL.so
#4  0x0000007fefd19bbc in mozilla::gl::GLLibraryEGL::fMakeCurrent(void*, void*, void*, void*) const ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#5  0x0000007fefcf9f18 in mozilla::gl::GLContextEGL::MakeCurrentImpl() const ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#6  0x0000007fefce89a8 in mozilla::gl::GLContext::MakeCurrent(bool) const ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#7  0x0000007fefcedf18 in mozilla::gl::GLContext::InitImpl() () at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#8  0x0000007fefcf9ccc in mozilla::gl::GLContextEGL::Init() () at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#9  0x0000007fefcf9194 in mozilla::gl::GLContextEGL::CreateGLContext(std::shared_ptr<mozilla::gl::EglDisplay>, mozilla::gl::GLContextDesc const&, void*, void*, bool, nsTSubstring<char>*) () at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#10 0x0000007fefcf70f4 in mozilla::gl::GLContextEGLFactory::CreateImpl(void*, bool, bool) ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#11 0x0000007fefcfb164 in mozilla::gl::GLContextProviderEGL::CreateForCompositorWidget(mozilla::widget::CompositorWidget*, bool, bool) ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#12 0x0000007feff7e2ac in mozilla::wr::RenderThread::CreateSingletonGL(nsTSubstring<char>&) ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#13 0x0000007feff7a844 in mozilla::wr::RenderThread::InitDeviceTask() ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#14 0x0000007fef9a2d84 in mozilla::detail::runnable_args_base<(mozilla::detail::RunnableResult)0>::Run() ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#15 0x0000007fef3a13c4 in nsThread::ProcessNextEvent(bool, bool*) () at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#16 0x0000007fef3a5150 in NS_ProcessNextEvent(nsIThread*, bool) () at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#17 0x0000007fef9fc624 in mozilla::ipc::MessagePumpForNonMainThreads::Run(base::MessagePump::Delegate*) ()
    at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#18 0x0000007fef9b1808 in MessageLoop::Run() () at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#19 0x0000007fef39eae4 in nsThread::ThreadFunc(void*) () at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libxul.so
#20 0x0000007ff7a54094 in _pt_root () at /home/pi/Downloads/2022-08-14-1738814c35/target/firefox/libnspr4.so
#21 0x00000055556186ec in set_alt_signal_stack_and_start(PthreadCreateParams*) ()
#22 0x0000007ff7f8df3c in start_thread (arg=0x0) at pthread_create.c:481
        ret = <optimized out>
        pd = 0x0
        unwind_buf = 
              {cancel_jmp_buf = {{jmp_buf = {547143938304, 549755796888, 549755796614, 0, 549755796615, 0, 549755796680, 547143938304, 547143940240, 549616342944, 547143936080, 17247675391755635550, 0, 17247675389317119454, 1065353216, 0, 0, 0, 0, 0, 0, 0}, mask_was_saved = 0}}, priv = {pad = {0x0, 0x0, 0x0, 0x0}, data = {prev = 0x0, cleanup = 0x0, canceltype = 0}}}
        not_first_call = <optimized out>
#23 0x0000007ff7bd3cdc in thread_start () at ../sysdeps/unix/sysv/linux/aarch64/clone.S:79
(gdb) 

Is this a problem caused by trying to read and set env vars almost at the same time? I have seen this a few days ago: https://github.com/rust-lang/rust/issues/90308

See Also: → 1696370
Crash Signature: [@ libc.so.6@0x3653c | libc.so.6@0x36518 | libGLdispatch.so.0@0x40cd8] → [@ libc.so.6@0x3653c | libc.so.6@0x36518 | libGLdispatch.so.0@0x40cd8] [@ __GI_getenv ]
Summary: Gnome Wayland, Raspberry Pi 4: Crash in [@ libc.so.6@0x3653c | libc.so.6@0x36518 | libGLdispatch.so.0@0x40cd8] → Gnome Wayland, Raspberry Pi 4: Intermittent startup crash in __GI_getenv ("GLVND_DISALLOW_PATCHING") at getenv.c:84 with address 0xe5e5e5e5e5e5e5e5

Before trying to reproduce again (comment 2),
I had added deb http://debug.mirrors.debian.org/debian-debug testing-debug main to /etc/apt/sources.list
and installed sudo apt install -t testing libegl-mesa0-dbgsym=22.2.0~rc1-1 libgbm-dbgsym=22.2.0~rc1-1 libgbm1-dbgsym=22.2.0~rc1-1 libgl1-mesa-dri-dbgsym=22.2.0~rc1-1 libglapi-mesa-dbgsym=22.2.0~rc1-1 libglx-mesa0-dbgsym=22.2.0~rc1-1 (http://debug.mirrors.debian.org/debian-debug/pool/main/m/mesa/).

(In reply to Robert Mader [:rmader] from comment #1)

There's a high chance that this is another Mesa bug, but having symbols would help a lot. Can you install debug symbols for the Mesa packages?

As for FF, here's a try build with "upload symbols", hope it works: https://treeherder.mozilla.org/jobs?repo=try&revision=3bd1b1d443d00db4769e185f46e4cf1eb11388d7

pi@raspberrypi:~/Downloads/2022-08-15-1784813/target/firefox $ ./firefox -P 2022-08-15-1784813
[GFX1-]: No GPUs detected via PCI
ATTENTION: default value of option mesa_glthread overridden by environment.
ExceptionHandler::GenerateDump cloned child 17336
ExceptionHandler::SendContinueSignalToChild sent continue signal to child
ExceptionHandler::WaitForContinueSignal waiting for continue signal...

bp-bc5f4885-99ef-49f1-8c4f-930e00220815

#12 0x0000007feff7e2ac in mozilla::wr::RenderThread::CreateSingletonGL(nsTSubstring<char>&) ()

Do I read this correctly that this happens randomly on startup? As IIRC this call should only happen once.

Edit: ah, only now read the updated title which says exactly that :)

Sometimes I get a main process crash on Firefox startup, most often not.

Is glvnd actually installed and can you deinstall it?

Summary: Gnome Wayland, Raspberry Pi 4: Intermittent startup crash in __GI_getenv ("GLVND_DISALLOW_PATCHING") at getenv.c:84 with address 0xe5e5e5e5e5e5e5e5 → Gnome Xwayland, Raspberry Pi 4: Intermittent startup crash in __GI_getenv ("GLVND_DISALLOW_PATCHING") at getenv.c:84 with address 0xe5e5e5e5e5e5e5e5

Err, does this only happen on Xwayland? In that case it could be a Xwayland bug due to the old Gnome and Xwayland version. We really need bookworm :)

Xwayland:
ATTENTION: default value of option mesa_glthread overridden by environment. is usually printed twice on startup.
If the crash occurs, then it occurs after the first one, the second one is not printed.

(Darkspirit from comment #3)

Is this a problem caused by trying to read and set env vars almost at the same time? I have seen this a few days ago: https://github.com/rust-lang/rust/issues/90308

Could bug 1741400 and bug 1748363 be related?

See Also: → 1741400, 1748363

(In reply to Robert Mader [:rmader] from comment #9)

Err, does this only happen on Xwayland?

Yes, it seems so.

  • X11: 3x ATTENTION: default value of option mesa_glthread overridden by environment. on startup, no crash.
  • Xwayland: 2x ATTENTION: default value of option mesa_glthread overridden by environment. on startup, sometimes a crash after the first one.
  • Wayland: No ATTENTION, no crash.

(In reply to Robert Mader [:rmader] from comment #8)

Is glvnd actually installed

libGLdispatch.so.0 is libglvnd0.

and can you deinstall it?

I don't know how. Many packages including xserver depend on https://packages.debian.org/stable/libegl1 which depends on libglvnd0 and libegl-mesa0.

See Also: → 1794309

Since the crash volume is low (less than 5 per week), the severity is downgraded to S3. Feel free to change it back if you think the bug is still critical.

For more information, please visit auto_nag documentation.

Severity: S2 → S3

This appears to have disappeared

Flags: needinfo?(robert.mader)

I will re-test.

Flags: needinfo?(robert.mader) → needinfo?(jan)

I have upgraded 64 bit Raspberry Pi OS (Debian bullseye) to Debian testing (bookworm).
I start Nightly with MOZ_ENABLE_WAYLAND=0 environment variable to force Xwayland.

I am able to reproduce the startup crash with:

With both builds, I get ATTENTION: default value of option mesa_glthread overridden by environment. three times now.
If the startup crash occurs, it occurs after the second one. The third is not printed then.

Crash Signature: [@ libc.so.6@0x3653c | libc.so.6@0x36518 | libGLdispatch.so.0@0x40cd8] [@ __GI_getenv ] → [@ libc.so.6@0x3653c | libc.so.6@0x36518 | libGLdispatch.so.0@0x40cd8] [@ __GI_getenv ] [@ libc.so.6@0x3d9e4 | libGLdispatch.so.0@0x40c98 ]
Flags: needinfo?(jan)

gdb looks almost the same as in comment 2:

pi@raspberrypi:~/Downloads $ MOZ_ENABLE_WAYLAND=0 gdb --args 2022-12-01/target/firefox/firefox -P 2022-12-01
GNU gdb (Debian 12.1-4) 12.1
Copyright (C) 2022 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Type "show copying" and "show warranty" for details.
This GDB was configured as "aarch64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<https://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
    <http://www.gnu.org/software/gdb/documentation/>.

For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from 2022-12-01/target/firefox/firefox...
(No debugging symbols found in 2022-12-01/target/firefox/firefox)
(gdb) r
Starting program: /home/pi/Downloads/2022-12-01/target/firefox/firefox -P 2022-12-01
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/aarch64-linux-gnu/libthread_db.so.1".
process 88262 is executing new program: /home/pi/Downloads/2022-12-01/target/firefox/firefox-bin
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/aarch64-linux-gnu/libthread_db.so.1".
[New Thread 0x7ff77ff140 (LWP 88267)]
[Thread 0x7ff77ff140 (LWP 88267) exited]
[Detaching after fork from child process 88268]
[Detaching after fork from child process 88278]
[New Thread 0x7ff77ff140 (LWP 88279)]
[New Thread 0x7fea49f140 (LWP 88280)]
[New Thread 0x7fe9c6f140 (LWP 88282)]
[New Thread 0x7fe536b140 (LWP 88283)]
[New Thread 0x7f655ff140 (LWP 88284)]
[New Thread 0x7f655af140 (LWP 88285)]
[New Thread 0x7f6555f140 (LWP 88286)]
[New Thread 0x7f6550f140 (LWP 88287)]
[New Thread 0x7f64e2e140 (LWP 88288)]
[New Thread 0x7f64dde140 (LWP 88289)]
[Thread 0x7f64e2e140 (LWP 88288) exited]
[New Thread 0x7f64bce140 (LWP 88290)]
[New Thread 0x7f64b7e140 (LWP 88291)]
[New Thread 0x7f649ff140 (LWP 88292)]
[Detaching after fork from child process 88293]
[New Thread 0x7f641ef140 (LWP 88295)]
[Thread 0x7f649ff140 (LWP 88292) exited]
[New Thread 0x7f64180140 (LWP 88309)]
[New Thread 0x7f649ff140 (LWP 88310)]
[New Thread 0x7f6371d140 (LWP 88311)]
[New Thread 0x7f6350f140 (LWP 88312)]
[New Thread 0x7f63301140 (LWP 88313)]
[GFX1-]: glxtest: DRM render node not clearly detectable. Falling back to using the only one that was found.
[GFX1-]: glxtest: VA-API test failed: failed to initialise VAAPI connection.
ATTENTION: default value of option mesa_glthread overridden by environment.
[New Thread 0x7f568be140 (LWP 88317)]
[New Thread 0x7f56b9f140 (LWP 88318)]
[New Thread 0x7f560ae140 (LWP 88319)]
[New Thread 0x7f56b4f140 (LWP 88320)]
[New Thread 0x7f5601e140 (LWP 88321)]
[New Thread 0x7f55e0e140 (LWP 88322)]
[New Thread 0x7f559ff140 (LWP 88323)]
[New Thread 0x7f555ff140 (LWP 88324)]
[New Thread 0x7f551ff140 (LWP 88325)]
[New Thread 0x7f54dff140 (LWP 88326)]
[New Thread 0x7f549ff140 (LWP 88327)]
[New Thread 0x7f545ff140 (LWP 88328)]
[New Thread 0x7f55bfe140 (LWP 88329)]
[New Thread 0x7f557ef140 (LWP 88331)]
ATTENTION: default value of option mesa_glthread overridden by environment.
[New Thread 0x7f53abe140 (LWP 88332)]

Thread 25 "Renderer" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7f56b4f140 (LWP 88320)]
__GI_getenv (name=0x7f540b00a2 "GLVND_DISALLOW_PATCHING") at ./stdlib/getenv.c:84
84	./stdlib/getenv.c: No such file or directory.
(gdb) bt full
#0  __GI_getenv (name=0x7f540b00a2 "GLVND_DISALLOW_PATCHING") at ./stdlib/getenv.c:84
Python Exception <class 'gdb.MemoryError'>: Cannot access memory at address 0xe5e5e5e5e5e5e5e5
        ep_start = #1  0x0000007f53fb0c9c in  () at /lib/aarch64-linux-gnu/libGLdispatch.so.0
#2  0x0000007f53fb0e84 in __glDispatchMakeCurrent () at /lib/aarch64-linux-gnu/libGLdispatch.so.0
#3  0x0000007f55b33338 in  () at /lib/aarch64-linux-gnu/libEGL.so
#4  0x0000007fec28c270 in mozilla::gl::GLLibraryEGL::fMakeCurrent(void*, void*, void*, void*) const ()
    at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#5  0x0000007fec282844 in mozilla::gl::GLContextEGL::MakeCurrentImpl() const () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#6  0x0000007fef1a1e48 in mozilla::gl::GLContext::InitImpl() () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#7  0x0000007fec27cdac in mozilla::gl::GLContext::Init() () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#8  0x0000007fef1a6df8 in mozilla::gl::GLContextEGL::Init() () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#9  0x0000007fec281900 in mozilla::gl::GLContextEGL::CreateGLContext(std::shared_ptr<mozilla::gl::EglDisplay>, mozilla::gl::GLContextDesc const&, void*, void*, bool, nsTSubstring<char>*) () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#10 0x0000007fec280594 in mozilla::gl::GLContextEGLFactory::CreateImpl(void*, bool, bool) ()
    at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#11 0x0000007fec282008 in mozilla::gl::GLContextEGLFactory::Create(void*, bool) ()
    at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#12 0x0000007fef29cff8 in mozilla::wr::RenderThread::CreateSingletonGL(nsTSubstring<char>&) ()
    at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#13 0x0000007fec3ca6b8 in mozilla::wr::RenderThread::InitDeviceTask() () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#14 0x0000007feefcf7b0 in mozilla::detail::runnable_args_base<(mozilla::detail::RunnableResult)0>::Run() ()
    at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#15 0x0000007fee747620 in NS_ProcessNextEvent(nsIThread*, bool) () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#16 0x0000007fee7ab8d4 in mozilla::ipc::MessagePumpForNonMainThreads::Run(base::MessagePump::Delegate*) ()
    at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#17 0x0000007feefd4e88 in MessageLoop::Run() () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#18 0x0000007feeda6b40 in nsThread::ThreadFunc(void*) () at /home/pi/Downloads/2022-12-01/target/firefox/libxul.so
#19 0x0000007ff6fbe680 in _pt_root () at /home/pi/Downloads/2022-12-01/target/firefox/libnspr4.so
#20 0x00000055555f9cb8 in set_alt_signal_stack_and_start(PthreadCreateParams*) ()
#21 0x0000007ff7b1edd8 in start_thread (arg=0x7fffffbb07) at ./nptl/pthread_create.c:442
        ret = <optimized out>
        pd = 0x7fffffbb07

                      unwind_buf = {cancel_jmp_buf = {{jmp_buf = {546915545408, 549755796488, 549755796230, 0, 322048, 549755796231, 546915221504, 546915545408, 546915221504, 546915547264, 546915543264, 12088575589521454630, 0, 12088575592189358722, 2147483648000032767, 0, 0, 0, 0, 0, 0, 0}, mask_was_saved = 0}}, priv = {pad = {0x0, 0x0, 0x0, 0x0}, data = {prev = 0x0, cleanup = 0x0, canceltype = 0}}}
        not_first_call = 0
#22 0x0000007ff7b87e9c in thread_start () at ../sysdeps/unix/sysv/linux/aarch64/clone.S:79
(gdb) exit
A debugging session is active.

	Inferior 1 [process 88262] will be killed.

Quit anyway? (y or n) y

Correction:
In comment 15 + 16 I still had old xwayland and mutter-common from archive.raspberrypi.org (modified Debian stable/bullseye).
Crash doesn't seem to be reproducible with xwayland (2:22.1.5-1), libglvnd0 (1.5.0-1), mutter (43.0-2) from Debian testing/bookworm.
Debian bookworm will be released in summer 2023.

xwayland changelogs:
21.0.99.901 https://lists.freedesktop.org/archives/xorg/2021-February/060615.html
21.1.0 https://lists.x.org/archives/xorg-announce/2021-March/003076.html
21.1.1 https://lists.freedesktop.org/archives/xorg/2021-April/060679.html
21.1.2 https://lists.freedesktop.org/archives/xorg/2021-July/060731.html
21.1.3 https://lists.freedesktop.org/archives/xorg/2021-November/060808.html
21.1.4: https://lists.freedesktop.org/archives/xorg/2021-December/060841.html
22.0.99.901 (aka 22.1.0 rc1) https://lists.freedesktop.org/archives/xorg/2022-January/060884.html
22.0.99.902 (aka 22.1.0 rc2) https://lists.freedesktop.org/archives/xorg/2022-February/060901.html
22.1.0 https://lists.freedesktop.org/archives/xorg/2022-February/060924.html
22.1.1 https://lists.freedesktop.org/archives/xorg/2022-March/060943.html
22.1.2 https://lists.freedesktop.org/archives/xorg/2022-May/060993.html
22.1.3 https://lists.freedesktop.org/archives/xorg/2022-July/061037.html
22.1.4 https://lists.freedesktop.org/archives/xorg/2022-October/061146.html
22.1.5 https://lists.freedesktop.org/archives/xorg/2022-November/061165.html

libglvnd changelog: https://github.com/NVIDIA/libglvnd/compare/v1.3.2...v1.5.0

Status: NEW → RESOLVED
Closed: 1 year ago
Resolution: --- → WORKSFORME
Summary: Gnome Xwayland, Raspberry Pi 4: Intermittent startup crash in __GI_getenv ("GLVND_DISALLOW_PATCHING") at getenv.c:84 with address 0xe5e5e5e5e5e5e5e5 → Gnome Xwayland 2:1.20.11-1+rpt1+deb11u3,Raspberry Pi 4 OS (bullseye): Intermittent startup crash in __GI_getenv ("GLVND_DISALLOW_PATCHING") at getenv.c:84 with address 0xe5e5e5e5e5e5e5e5. Seems fixed with Debian testing/bookworm (xwayland 22.1.5).
You need to log in before you can comment on or make changes to this bug.