[Bug 1185595] New: Kernel 5.12.0-1 Xorg on QEMU/KVM and dirver video QXL crash
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 Bug ID: 1185595 Summary: Kernel 5.12.0-1 Xorg on QEMU/KVM and dirver video QXL crash Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: x86-64 OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Kernel Assignee: kernel-bugs@opensuse.org Reporter: gstoduto@gmail.com QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- Host: openSUSE thumbleweed last upgrade 20210430 video driver official Nvidia 460.73.01 cpu AMD vm GDM/Gnome Guest: openSUSE thumbleweed last upgrade 20210430 video driver QXL screen Spice vm GDM/Gnome After installing the new kernel i have the following error when I enter an Xorg section and the crash server with a Segmentation fault: 159.934] (II) event5 - spice vdagent tablet: device is a pointer [ 159.934] (II) event5 - spice vdagent tablet: device removed [ 159.934] (**) Option "config_info" "udev:/sys/devices/virtual/input/input13/event5" [ 159.934] (II) XINPUT: Adding extended input device "spice vdagent tablet" (type: MOUSE, id 10) [ 159.935] (**) Option "AccelerationScheme" "none" [ 159.935] (**) spice vdagent tablet: (accel) selected scheme none/0 [ 159.935] (**) spice vdagent tablet: (accel) acceleration factor: 2.000 [ 159.935] (**) spice vdagent tablet: (accel) acceleration threshold: 4 [ 159.935] (II) event5 - spice vdagent tablet: is tagged by udev as: Mouse [ 159.935] (II) event5 - spice vdagent tablet: device is a pointer [ 220.389] (EE) qxl(0): error doing QXL_ALLOC [ 220.389] (EE) [ 220.389] (EE) Backtrace: [ 220.389] (EE) 0: /usr/bin/Xorg.bin (xorg_backtrace+0x7d) [0x55fb822483cd] [ 220.389] (EE) 1: /usr/bin/Xorg.bin (0x55fb82085000+0x1cbe55) [0x55fb82250e55] [ 220.390] (EE) 2: /lib64/libpthread.so.0 (0x7f1e35ba6000+0x13a30) [0x7f1e35bb9a30] [ 220.390] (EE) 3: /usr/lib64/xorg/modules/drivers/qxl_drv.so (0x7f1e351f3000+0x8d35) [0x7f1e351fbd35] [ 220.390] (EE) 4: /usr/lib64/xorg/modules/drivers/qxl_drv.so (0x7f1e351f3000+0x91cb) [0x7f1e351fc1cb] [ 220.390] (EE) 5: /usr/lib64/xorg/modules/drivers/qxl_drv.so (0x7f1e351f3000+0x156a1) [0x7f1e352086a1] [ 220.390] (EE) 6: /usr/bin/Xorg.bin (miCopyRegion+0x97) [0x55fb82221897] [ 220.390] (EE) 7: /usr/bin/Xorg.bin (miDoCopy+0x466) [0x55fb8222b326] [ 220.390] (EE) 8: /usr/lib64/xorg/modules/drivers/qxl_drv.so (0x7f1e351f3000+0x158da) [0x7f1e352088da] [ 220.390] (EE) 9: /usr/bin/Xorg.bin (0x55fb82085000+0x149586) [0x55fb821ce586] [ 220.390] (EE) 10: /usr/bin/Xorg.bin (0x55fb82085000+0xffc02) [0x55fb82184c02] [ 220.390] (EE) 11: /usr/bin/Xorg.bin (0x55fb82085000+0x4af0d) [0x55fb820cff0d] [ 220.390] (EE) 12: /lib64/libc.so.6 (__libc_start_main+0xd5) [0x7f1e359feb25] [ 220.390] (EE) 13: /usr/bin/Xorg.bin (_start+0x2e) [0x55fb820d02ce] [ 220.390] (EE) [ 220.390] (EE) Segmentation fault at address 0x0 [ 220.390] (EE) Fatal server error: [ 220.390] (EE) Caught signal 11 (Segmentation fault). Server aborting [ 220.390] (EE) After Segmentation fault GDM restarts and the Xorg section closes. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 http://bugzilla.opensuse.org/show_bug.cgi?id=1185595#c1 Takashi Iwai <tiwai@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |gstoduto@gmail.com, | |tiwai@suse.com Flags| |needinfo?(gstoduto@gmail.co | |m) --- Comment #1 from Takashi Iwai <tiwai@suse.com> --- Do you see any relevant kernel crash / oops or such? Otherwise, the segfault is rather a user-space problem. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 Takashi Iwai <tiwai@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |sndirsch@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 http://bugzilla.opensuse.org/show_bug.cgi?id=1185595#c2 --- Comment #2 from Giuseppe Stoduto <gstoduto@gmail.com> --- I don't. I did tests and changing QXL with Vertio the problem does not occur. Even changing the kernel with the old kernel 5.11 the problem does not occur -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 http://bugzilla.opensuse.org/show_bug.cgi?id=1185595#c3 --- Comment #3 from Takashi Iwai <tiwai@suse.com> --- OK, thanks. FWIW, my test VM with XFCE works fine on 5.12 guest kernel with both bochs and QXL. So it doesn't seem hitting always with those drivers. The fact that 5.11 kernel worked means some changes in DRM side causes the incompatibility that blows up the X driver. Although X driver itself shouldn't cause a segfault, the kernel regression has to be fixed in anyway. So we need fixes in both side. BTW, does it happen also with Wayland session? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 http://bugzilla.opensuse.org/show_bug.cgi?id=1185595#c4 --- Comment #4 from Stefan Dirsch <sndirsch@suse.com> --- Hmm. Isn't there are a KMS driver, so you no longer need qxl X driver, can just uninstall xf86-video-qxl package and use modesetting X driver? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 http://bugzilla.opensuse.org/show_bug.cgi?id=1185595#c5 --- Comment #5 from Stefan Dirsch <sndirsch@suse.com> --- Just verified. There is definitely a "qxl" kernel module. So please uninstall xf86-video-qxl package and try again. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 http://bugzilla.opensuse.org/show_bug.cgi?id=1185595#c6 --- Comment #6 from Giuseppe Stoduto <gstoduto@gmail.com> --- I uninstalled the driver xf86-video-qxl package now does not give me more segment fault but the system remains locked for a few minutes. Looking at the root logs there is this error: kernel: f 4026531847#1389: failed to wait on release 7 after spincount 301 kernel: f 4026531847#1389: failed to wait on release 7 after spincount 301 kernel: f 4026531847#1389: failed to wait on release 7 after spincount 301 kernel: f 4026531847#1389: failed to wait on release 7 after spincount 301 kernel: [TTM] Buffer eviction failed kernel: qxl 0000:00:01.0: object_init failed for (3149824, 0x00000001) kernel: [drm:qxl_alloc_bo_reserved [qxl]] *ERROR* failed to allocate VRAM BO this on the guest machine I also created a file in xorg.conf with this content to be sure to start the system with modesetting enabled Section "Device" Identifier "modesetting" Driver "modesetting" Option "PreferColorMode" "true" Option "AccelMethod" "none" EndSection Section "Screen" Identifier "modesetting" Device "modesetting" EndSection Same problem and error on Wayland. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 http://bugzilla.opensuse.org/show_bug.cgi?id=1185595#c7 Stefan Dirsch <sndirsch@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Summary|Kernel 5.12.0-1 Xorg on |Kernel 5.12.0-1 Xorg on |QEMU/KVM and dirver video |QEMU/KVM and driver video |QXL crash |QXL crash --- Comment #7 from Stefan Dirsch <sndirsch@suse.com> --- Thanks. This looks definitely like a kernel problem. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1185595 David Nessl <drn@nessl.name> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |drn@nessl.name -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply＠suse.com