| Bug ID | 104181 |
|---|---|
| Summary | [NV50/G94] system freeze on youtube with epiphany 3.26 |
| Product | xorg |
| Version | unspecified |
| Hardware | x86-64 (AMD64) |
| OS | Linux (All) |
| Status | NEW |
| Severity | normal |
| Priority | medium |
| Component | Driver/nouveau |
| Assignee | nouveau@lists.freedesktop.org |
| Reporter | cyril.picard@gmail.com |
| QA Contact | xorg-team@lists.x.org |
Created attachment 136057 [det=
ails]
kernel log
Nouveau crashes each time I try to watch a youtube video with epiphany 3.26.
Kernel log attached.
archlinux 64bit
linux 4.14.4-1
libdrm 2.4.88-1
mesa 17.2.6-1
xf86-video-nouveau 1.0.15-2
| What | Removed | Added |
|---|---|---|
| Priority | medium | high |
| CC | guido@trentalancia.com | |
| Severity | normal | major |
It happens also with more recent versions, such as any kernel = ranging from 4.10.5 to very recent 4.17.10, latest mesa (18.1.5) and libdrm (2.4.93). Aug 13 21:42:50 vortex kernel: [ 216.549941] nouveau 0000:01:00.0: fifo: DMA_PUSHER - ch 8 [ositorWorkQueue[5050]] get 002015efe4 put 0020160044 ib_= get 000000f7 ib_put 000000f8 state 80007698 (err: INVALID_CMD) push 00400040 Aug 13 21:44:02 vortex kernel: [ 287.658593] nouveau 0000:01:00.0: ositorWorkQueue[5150]: multiple instances of buffer 3 on validation list Aug 13 21:44:02 vortex kernel: [ 287.658600] nouveau 0000:01:00.0: ositorWorkQueue[5150]: validate_init Aug 13 21:44:02 vortex kernel: [ 287.658602] nouveau 0000:01:00.0: ositorWorkQueue[5150]: validate: -22 Aug 13 21:44:02 vortex kernel: [ 287.661076] nouveau 0000:01:00.0: ositorWorkQueue[5150]: multiple instances of buffer 3 on validation list Aug 13 21:44:02 vortex kernel: [ 287.661085] nouveau 0000:01:00.0: ositorWorkQueue[5150]: validate_init Aug 13 21:44:02 vortex kernel: [ 287.661089] nouveau 0000:01:00.0: ositorWorkQueue[5150]: validate: -22 Aug 13 21:44:02 vortex kernel: [ 287.698020] show_signal_msg: 441 callbacks suppressed Aug 13 21:44:02 vortex kernel: [ 287.698028] gstglcontext[5213]: segfault = at 7f64f24b1000 ip 00007f644f46d137 sp 00007f63e0f11408 error 4 in nouveau_dri.so[7f644ebae000+c01000] Aug 13 21:44:10 vortex kernel: [ 296.368787] ositorWorkQueue[5234]: segfau= lt at 8 ip 00007f9a41da9e3c sp 00007f9a7e5d6330 error 4 in libdrm_nouveau.so.2.0.0[7f9a41da5000+8000]
However, I think I have a different chipset, it is reported in= Xorg logfile as "NVIDIA NVA5", while in the kernel dmesg as "NVIDIA GT216&qu= ot;...
It might be related to a similar crash reported in similar bug: https://bugs.freedesktop.org/show_b= ug.cgi?id=3D70390#c24
The crash does not happen when the video is played directly wi= th gstreamer (e.g. gst-play-1.0 from a terminal).
Maybe the following output is more meaningful (kernel 4.18): [ 860.276334] WARNING: CPU: 1 PID: 3296 at drivers/gpu/drm/nouveau/nvif/vmm.c:71 nvif_vmm_put+0x65/0x70 [nouveau] [ 860.276338] Modules linked in: xt_iprange ipv6 crc_ccitt arc4 brcmsmac cordic brcmutil mac80211 sha256_ssse3 sha256_generic cfg80211 acer_wmi sparse_keymap rfkill iTCO_wdt iTCO_vendor_support joydev pcc_cpufreq acpi_cpufreq snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel snd_hda_codec bcma snd_hda_core snd_hwdep snd_pcm snd_timer snd soundcore mei_me i2c_i801 pcspkr mei lpc_ich mfd_core dm_crypt serio_raw atl1c nouveau mxm_wmi i2c_algo_bit wmi drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops ttm video drm i2c_core [ 860.276369] CPU: 1 PID: 3296 Comm: kworker/1:2 Not tainted 4.18.0 #3 [ 860.276371] Hardware name: Acer Aspire 5745G/JV51_CP, BIOS V1.19 22/03/2= 011 [ 860.276389] Workqueue: events nouveau_cli_work [nouveau] [ 860.276402] RIP: 0010:nvif_vmm_put+0x65/0x70 [nouveau] [ 860.276404] Code: 00 00 48 89 e2 be 02 00 00 00 48 c7 04 24 00 00 00 00 = 48 89 44 24 08 e8 b9 e6 ff ff 85 c0 75 0a 48 c7 43 08 00 00 00 00 eb b7 <0f= > 0b eb f2 e8 22 64 e5 e9 66 90 53 48 83 ec 20 65 48 8b 04 25 28 [ 860.276428] RSP: 0018:ffffa1344253fdf0 EFLAGS: 00010282 [ 860.276430] RAX: 00000000fffffffe RBX: ffffa1344253fe18 RCX: 0000000000000000 [ 860.276431] RDX: 0000000000000000 RSI: ffffa1344253fd60 RDI: ffffa1344253fe00 [ 860.276433] RBP: ffffa1344253fe48 R08: 0000000000af4000 R09: 0000000000000000 [ 860.276434] R10: fffffc22023aa7c0 R11: ffff933153403a00 R12: ffff93314f563260 [ 860.276436] R13: dead000000000200 R14: dead000000000100 R15: ffff93314f563270 [ 860.276438] FS: 0000000000000000(0000) GS:ffff933157c40000(0000) knlGS:0000000000000000 [ 860.276439] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 860.276441] CR2: 00007f8462b05ff0 CR3: 000000012d00a003 CR4: 00000000000206e0 [ 860.276442] Call Trace: [ 860.276462] nouveau_vma_del+0x6b/0xb0 [nouveau] [ 860.276479] nouveau_gem_object_delete_work+0x31/0x60 [nouveau] [ 860.276497] nouveau_cli_work+0x71/0x100 [nouveau] [ 860.276502] process_one_work+0x1bb/0x330 [ 860.276505] worker_thread+0x28/0x3c0 [ 860.276507] ? set_worker_desc+0xb0/0xb0 [ 860.276509] kthread+0x109/0x120 [ 860.276511] ? kthread_create_worker_on_cpu+0x70/0x70 [ 860.276515] ret_from_fork+0x35/0x40 [ 860.276517] ---[ end trace 4129b6c001bb82f1 ]--- [ 860.342881] nouveau 0000:01:00.0: fifo: DMA_PUSHER - ch 8 [ositorWorkQueue[5997]] get 0000090664 put 0000091400 ib_get 000002b2 ib_put 000002b9 state 80006efc (err: INVALID_CMD) push 00400040 [ 860.342911] nouveau 0000:01:00.0: fifo: DMA_PUSHER - ch 8 [ositorWorkQueue[5997]] get 0000602004 put 000060200c ib_get 000002b3 ib_put 000002b9 state 80000000 (err: INVALID_CMD) push 00400040 [ 860.343162] nouveau 0000:01:00.0: gr: DATA_ERROR 00000004 [INVALID_VALUE] [ 860.343168] nouveau 0000:01:00.0: gr: 00100000 [] ch 8 [003f55d000 ositorWorkQueue[5997]] subc 3 class 8597 mthd 0ef4 data 0fac6881
At the moment, the only solution is to add the following line = to /etc/X11/xorg.conf (under Section "Device"): Option "NoAccel" "true" because setting the kernel module option "noaccel=3D1" also leads= to a crash due to some other silly bug !
There are no reports earlier than year 2013, so this bug has p= robably been introduced that year (and never fixed since then)...
I was wrong in my previous comment (Comment 7)... I have now verified more carefully and realized that this bug has been arou= nd at least since kernel version 3.0 (August 2011) ! At this point, it might even be that this bug has existed since the initial release of the driver...
| What | Removed | Added |
|---|---|---|
| Resolution | --- | MOVED |
| Status | NEW | RESOLVED |
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this = link to our GitLab instance: https://gitlab.freedesktop.org/xorg/driver/xf86-video-nouveau/iss= ues/389.