[Impact]
Running stress-ng freezes the screen under wayland on intel iris. Upstream has fixed it in
5aae8a05264c354aa93017d323ce238858f68227 iris: Retry DRM_IOCTL_I915_GEM_EXECBUFFER2 on ENOMEM
646cff13bca8a92b846984d782ef00e57d34d7a1 Revert "iris: Avoid abort() if kernel can't allocate memory"
which need to be backported for 22.2.5
[Test case]
Install the update, then
1. stress-ng --stack 0 --timeout 300
2. check the screen
and note that it shouldn't freeze anymore.
[Where things could go wrong]
This moves checking ENOMEM to the right place, so it's hard to see how it might cause issues.
--
[Steps to reproduce]
(disable systemd-oomd or executing over ssh)
(below command allocates a lot of memory to stress kernel page fault)
1. stress-ng --stack 0 --timeout 300
2. check the screen
[Expected result]
Screen will update slowly when performing the stress-test.
but screen needs back to work after stress.
[Actual result]
Screen freeze after stress test.
[Additional information]
kernel version: vmlinuz-5.17.0-1017-oem
kernel version: vmlinuz-6.0.0-060000rc4drmtip20220910-generic
Mesa version: 22.0.5-0ubuntu0.1
Mutter version: 42.2-0ubuntu1
Gnome-shell version: 42.2-0ubuntu0.2
* Issue happens in Wayland only
* gnome-shell keeps issuing ioctl()
```
(gdb) bt
#0 __GI___ioctl (fd=fd@entry=14, request=request@entry=3223348419) at ../sysdeps/unix/sysv/linux/ioctl.c:36
#1 0x00007fcacab0eb4f in intel_ioctl (arg=0x7fffc5404d80, request=3223348419, fd=14) at ../src/intel/common/intel_gem.h:75
#2 iris_wait_syncobj (timeout_nsec=140736502713808, syncobj=<optimized out>, bufmgr=<optimized out>) at ../src/gallium/drivers/iris/iris_fence.c:229
#3 iris_wait_syncobj (bufmgr=<optimized out>, syncobj=<optimized out>, timeout_nsec=timeout_nsec@entry=9223372036854775807) at ../src/gallium/drivers/iris/iris_fence.c:215
#4 0x00007fcacab8beab in iris_get_query_result (result=0x7fffc5404e50, wait=<optimized out>, query=0x55774159b800, ctx=<optimized out>) at ../src/gallium/drivers/iris/iris_query.c:635
#5 iris_get_query_result (ctx=<optimized out>, query=0x55774159b800, wait=<optimized out>, result=0x7fffc5404e50) at ../src/gallium/drivers/iris/iris_query.c:601
#6 0x00007fcaca405e89 in tc_get_query_result (_pipe=<optimized out>, query=0x55774159b800, wait=<optimized out>, result=0x7fffc5404e50) at ../src/gallium/auxiliary/util/u_threaded_context.c:881
#7 0x00007fcaca0f0e64 in get_query_result (pipe=pipe@entry=0x55773f87bad0, q=q@entry=0x557741a0c790, wait=wait@entry=1 '\001') at ../src/mesa/main/queryobj.c:266
#8 0x00007fcaca0f1b12 in _mesa_wait_query (q=0x557741a0c790, ctx=0x55773f8af980) at ../src/mesa/main/queryobj.c:344
#9 get_query_object (ctx=0x55773f8af980, func=func@entry=0x7fcacaf28283 "glGetQueryObjecti64v", id=<optimized out>, pname=34918, ptype=ptype@entry=5134, buf=0x0, offset=140736502714192)
at ../src/mesa/main/queryobj.c:1174
#10 0x00007fcaca0f2d65 in _mesa_GetQueryObjecti64v (id=<optimized out>, pname=<optimized out>, params=<optimized out>) at ../src/mesa/main/queryobj.c:1257
#11 0x00007fcadfa39b90 in ?? () from /usr/lib/x86_64-linux-gnu/mutter-10/libmutter-cogl-10.so.0
#12 0x00007fcadfa751b9 in cogl_frame_info_get_rendering_duration_ns () from /usr/lib/x86_64-linux-gnu/mutter-10/libmutter-cogl-10.so.0
#13 0x00007fcadf8819c4 in ?? () from /lib/x86_64-linux-gnu/libmutter-10.so.0
#14 0x00007fcadfa6e7a2 in _cogl_onscreen_notify_complete () from /usr/lib/x86_64-linux-gnu/mutter-10/libmutter-cogl-10.so.0
#15 0x00007fcadf969cdd in ?? () from /lib/x86_64-linux-gnu/libmutter-10.so.0
#16 0x00007fcadf96ec7b in ?? () from /lib/x86_64-linux-gnu/libmutter-10.so.0
#17 0x00007fcadf969601 in ?? () from /lib/x86_64-linux-gnu/libmutter-10.so.0
#18 0x00007fcadf98395e in ?? () from /lib/x86_64-linux-gnu/libmutter-10.so.0
#19 0x00007fcadf96962d in ?? () from /lib/x86_64-linux-gnu/libmutter-10.so.0
#20 0x00007fcae0743c24 in g_main_context_dispatch () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
#21 0x00007fcae07986f8 in ?? () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
#22 0x00007fcae0743293 in g_main_loop_run () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
#23 0x00007fcadf8d0849 in meta_context_run_main_loop () from /lib/x86_64-linux-gnu/libmutter-10.so.0
#24 0x000055773f1a4f12 in ?? ()
#25 0x00007fcadf429d90 in __libc_start_call_main (main=main@entry=0x55773f1a4a70, argc=argc@entry=1, argv=argv@entry=0x7fffc54053e8) at ../sysdeps/nptl/libc_start_call_main.h:58
#26 0x00007fcadf429e40 in __libc_start_main_impl (main=0x55773f1a4a70, argc=1, argv=0x7fffc54053e8, init=<optimized out>, fini=<optimized out>, rtld_fini=<optimized out>, stack_end=0x7fffc54053d8)
at ../csu/libc-start.c:392
#27 0x000055773f1a51b5 in ?? ()
```
Other discussion thread:
https://gitlab.freedesktop.org/drm/intel/-/issues/6851
https://gitlab.gnome.org/GNOME/mutter/-/issues/2431
ProblemType: Bug
DistroRelease: Ubuntu 22.04
Package: mutter (not installed)
Uname: Linux 5.18.0-rc2+ x86_64
ApportVersion: 2.20.11-0ubuntu82.1
Architecture: amd64
CasperMD5json:
{
"result": "skip"
}
Date: Mon Sep 19 10:39:59 2022
DistributionChannelDescriptor:
# This is the distribution channel descriptor for the OEM CDs
# For more information see http://wiki.ubuntu.com/DistributionChannelDescriptor
canonical-oem-somerville-jammy-amd64-20220504-33+jellyfish-chansey+X25
InstallationDate: Installed on 2022-09-06 (12 days ago)
InstallationMedia: Ubuntu 22.04 LTS "Jammy Jellyfish" - somerville-jammy-amd64-20220504-33
ProcEnviron:
TERM=rxvt-unicode-256color
PATH=(custom, no user)
XDG_RUNTIME_DIR=<set>
LANG=en_US.UTF-8
SHELL=/bin/bash
SourcePackage: mutter
UpgradeStatus: No upgrade log present (probably fresh install)
This bug is missing log files that will aid in diagnosing the problem. While running an Ubuntu kernel (not a mainline or third-party kernel) please enter the following command in a terminal window:
apport-collect 1990089
and then change the status of the bug to 'Confirmed'.
If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.
This change has been made by an automated script, maintained by the Ubuntu Kernel Team.