| Summary: | Running apps via Flatpak with DRI device access: [gfxhub] retry page fault (src_id:0 ring:0 vmid:1 pasid:32770, for process X pid 2148 thread X:cs0 pid 2151) in page starting at address 0x0000000107720000 from 27, VM_L2_PROTECTION_FAULT_STATUS:0x00101031 | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | DRI | Reporter: | Dennis Schridde <devurandom> | ||||||||||||||||||
| Component: | DRM/AMDgpu | Assignee: | Default DRI bug account <dri-devel> | ||||||||||||||||||
| Status: | RESOLVED MOVED | QA Contact: | |||||||||||||||||||
| Severity: | not set | ||||||||||||||||||||
| Priority: | not set | CC: | dennis.schridde | ||||||||||||||||||
| Version: | unspecified | ||||||||||||||||||||
| Hardware: | Other | ||||||||||||||||||||
| OS: | All | ||||||||||||||||||||
| Whiteboard: | |||||||||||||||||||||
| i915 platform: | i915 features: | ||||||||||||||||||||
| Attachments: |
|
||||||||||||||||||||
Created attachment 145519 [details]
lspci -vv
Created attachment 145520 [details]
glxinfo
Created attachment 145521 [details]
/proc/cpuinfo
Created attachment 145522 [details]
lshw
Created attachment 145523 [details]
emerge --info mesa libdrm llvm linux-firmware xorg-server
This also affects io.atom.Atom and com.skype.Client. I confirmed again that it does not affect org.zulip.Zulip -- dmesg does not acquire new messages after starting it, browsing in the app (something that is impossible with the other apps, since the whole graphical environment is locked up) and stopping it. Created attachment 145535 [details]
flatpak info (-M) $app
I attached the output of:
```
for app in com.skype.Client com.valvesoftware.Steam io.atom.Atom org.signal.Signal org.zulip.Zulip
flatpak info $app
echo Permissions:
flatpak info -M $app
end
```
What sticks out is that the applications crashing all have `devices=all` or `devices=dri` permissions -- i.e. they can access the GPU directly.
Running /var/lib/flatpak/app/io.atom.Atom/current/active/files/bin/atom-real directly, without Flatpak, locks up the entire system hard -- ctrl+alt+f2, ctrl+alt+del, SysReq+REISUB all do not help and I have to hard reset the machine. No logs are available from this run.
Running /var/lib/flatpak/app/io.atom.Atom/current/active/files/share/atom/atom (the binary started by .../bin/atom-real) will also lock up the system. I installed net-im/signal-desktop-bin-1.27.2::gentoo to my system, natively, using the package manager. Running /usr/sbin/signal-desktop does not lock it up. Running /var/lib/flatpak/app/org.signal.Signal/current/active/files/Signal/signal-desktop also does not lock it up. Ignoring io.atom.Atom and concentrating on org.signal.Signal, I assume that something that Flatpak does to setup its namespaces and cgroups trips OpenGL / Mesa and the Linux kernel's AMDGPU driver... The last kernel where I did not notice such lock ups was 5.1.4 (on the same Gentoo system). I did not use this machine since mid/end May until now, so I did not run any kernels in between 5.1.4 and 5.3.1. This prevents the graphical system lockup: `flatpak run --nodevice=dri org.signal.Signal` For some reason `flatpak run --nodevice=all org.signal.Signal` does not help, it needs to be the more specific `dri`. Created attachment 145562 [details]
dmesg, with backtraces in DCN
`flatpak run --nodevice=all com.valvesoftware.Steam` also makes Steam start.
`flatpak run --nodevice=dri com.valvesoftware.Steam` does not. It appears that `--nodevice` has to match the `device` setting in the app manifest and arithmethic like `all - dri` is not supported.
This time I also found this in dmesg, supposedly from running org.signal.Signal (according to the time: I first tried Signal with --nodevice=all, then with --nodevice=dri, then I tried Steam with --nodevice=dri, and again with --nodevice=all):
[47269.569990] ------------[ cut here ]------------ [1173/5119]
[47269.570063] WARNING: CPU: 7 PID: 1150 at drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_hw_sequencer.c:932 dcn10_verify_allow_pstate_change_high.cold+0xc/0x227 [amdgpu]
[47269.570065] Modules linked in: fuse nf_conntrack_netlink xfrm_user xfrm_algo br_netfilter nf_nat_tftp tun bridge stp llc nft_masq ebtable_nat ip_set nft_chain_nat ebtable_filter ebtables nft_objref nft_reject_inet nf_reject_ipv6 nft_reject nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib xt_MASQUERADE xt_addrtype nf_tables_set iptable_nat ipt_REJECT nf_reject_i
pv4 xt_conntrack iptable_filter xt_CHECKSUM nf_conntrack_tftp nf_conntrack_netbios_ns nf_conntrack_broadcast xt_tcpudp iptable_mangle ip6table_filter ip6table_security ip6table_raw nft_ct ip6table_mangle ip6table_nat nf_nat nf_conntrack bluetooth nf_tables nf_defrag_ipv6 nf_defrag_ipv4 ip6_tables iptable_security iptable_raw ip_tables nfnetlink ecdh_generic ecc
crc16 x_tables bpfilter amdgpu snd_hda_codec_realtek snd_hda_codec_generic kvm_amd ledtrig_audio snd_hda_codec_hdmi gpu_sched ttm ccp snd_hda_intel snd_hda_codec rng_core kvm eeepc_wmi drm_kms_helper snd_hda_core asus_wmi snd_hwdep snd_pcm sparse_keymap
[47269.570085] syscopyarea sysfillrect sysimgblt rfkill sp5100_tco wmi_bmof snd_timer pcspkr irqbypass fb_sys_fops k10temp input_leds drm snd i2c_piix4 led_class soundcore mousedev joydev gpio_amdpt gpio_generic evdev mac_hid acpi_cpufreq nls_iso8859_1 nls_cp437 vfat fat sch_fq_codel sctp lm92 btrfs xor raid6_pq dm_cache_smq dm_cache dm_persistent_data dm_bio_p
rison dm_bufio libcrc32c crc32c_generic sd_mod usbkbd hid_generic usbmouse usbhid hid crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel ahci libahci aesni_intel libata igb aes_x86_64 crypto_simd xhci_pci xhci_hcd cryptd i2c_algo_bit scsi_mod glue_helper dca hwmon wmi pinctrl_amd sunrpc dm_mirror dm_region_hash dm_log dm_mod
[47269.570110] CPU: 7 PID: 1150 Comm: X Tainted: G T 5.3.1-gentoo #7
[47269.570112] Hardware name: System manufacturer System Product Name/ROG STRIX B350-F GAMING, BIOS 5216 09/02/2019
[47269.570154] RIP: 0010:dcn10_verify_allow_pstate_change_high.cold+0xc/0x227 [amdgpu]
[47269.570156] Code: 83 c8 ff e9 09 02 f9 ff 48 c7 c7 98 1e df c0 e8 53 a8 97 de 0f 0b 83 c8 ff e9 f3 01 f9 ff 48 c7 c7 98 1e df c0 e8 3d a8 97 de <0f> 0b 80 bb 9f 01 00 00 00 75 05 e9 fb 24 f9 ff 48 8b 83 f8 02 00
[47269.570157] RSP: 0018:ffffaab7c120f730 EFLAGS: 00010246
[47269.570159] RAX: 0000000000000024 RBX: ffff99fb2cd00000 RCX: 0000000000000006
[47269.570160] RDX: 0000000000000000 RSI: 0000000000000082 RDI: ffff99fb307d64d0
[47269.570160] RBP: ffff99fb2cd00000 R08: 000000007ffffffd R09: 0000000000000545
[47269.570161] R10: 000000000000003b R11: ffffaab7c120f5e0 R12: ffff99fac66601b8
[47269.570162] R13: ffff99fac6661d18 R14: ffff99fac66601b8 R15: ffff99f8ac70c000
[47269.570163] FS: 00007f1c1d65d480(0000) GS:ffff99fb307c0000(0000) knlGS:0000000000000000
[47269.570164] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[47269.570165] CR2: 0000565061480000 CR3: 00000003b6916000 CR4: 00000000003406e0
[47269.570166] Call Trace:
[47269.570212] dcn10_pipe_control_lock.part.0+0x64/0x70 [amdgpu]
[47269.570253] dc_stream_set_cursor_attributes+0x11c/0x170 [amdgpu]
[47269.570297] handle_cursor_update.isra.0+0x1f1/0x350 [amdgpu]
[47269.570341] amdgpu_dm_commit_cursors.isra.0+0x56/0x70 [amdgpu]
[47269.570384] amdgpu_dm_atomic_commit_tail+0x12b9/0x1c50 [amdgpu]
[47269.570388] ? enqueue_task_rt+0x9d/0xc0
[47269.570392] ? ttm_eu_backoff_reservation+0x3f/0x80 [ttm]
[47269.570435] ? dm_plane_helper_prepare_fb+0x19c/0x2d0 [amdgpu]
[47269.570437] ? _cond_resched+0x10/0x20
[47269.570439] ? wait_for_completion_timeout+0x31/0x110
[47269.570440] ? wait_for_completion_interruptible+0x2e/0x140
[47269.570441] ? complete_all+0x2b/0x40
[47269.570447] ? commit_tail+0x37/0x60 [drm_kms_helper]
[47269.570451] commit_tail+0x37/0x60 [drm_kms_helper]
[47269.570456] drm_atomic_helper_commit+0x103/0x110 [drm_kms_helper]
[47269.570461] drm_atomic_helper_update_plane+0xe7/0x100 [drm_kms_helper]
[47269.570470] drm_mode_cursor_universal+0x127/0x240 [drm]
[47269.570478] drm_mode_cursor_common+0xd9/0x230 [drm]
[47269.570486] ? drm_mode_cursor_ioctl+0x70/0x70 [drm]
[47269.570492] drm_ioctl_kernel+0xa2/0xf0 [drm]
[47269.570499] drm_ioctl+0x200/0x380 [drm]
[47269.570507] ? drm_mode_cursor_ioctl+0x70/0x70 [drm]
[47269.570536] amdgpu_drm_ioctl+0x41/0x70 [amdgpu]
[47269.570539] do_vfs_ioctl+0x44a/0x6e0
[47269.570541] ksys_ioctl+0x59/0x90
[47269.570542] __x64_sys_ioctl+0x11/0x20
[47269.570545] do_syscall_64+0x4a/0x110
[47269.570547] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[47269.570548] RIP: 0033:0x7f1c1e2a7597
[47269.570550] Code: 00 00 00 75 0c 48 c7 c0 ff ff ff ff 48 83 c4 18 c3 e8 8d d9 01 00 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d c9 48 0d 00 f7 d8 64 89 01 48
[47269.570551] RSP: 002b:00007fff5dd14a48 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
[47269.570552] RAX: ffffffffffffffda RBX: 00007fff5dd14a80 RCX: 00007f1c1e2a7597
[47269.570553] RDX: 00007fff5dd14a80 RSI: 00000000c02464bb RDI: 000000000000000c
[47269.570554] RBP: 00000000c02464bb R08: 0000565061371710 R09: 0000000000003fff
[47269.570554] R10: 000000000000007f R11: 0000000000000246 R12: 0000565060457200
[47269.570555] R13: 000000000000000c R14: 0000000000000004 R15: 0000565060459fa0
[47269.570557] ---[ end trace 0b5db3fd3a295d4e ]---
[47269.571134] ------------[ cut here ]------------
[47269.571195] WARNING: CPU: 7 PID: 1150 at drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_hw_sequencer.c:932 dcn10_verify_allow_pstate_change_high.cold+0xc/0x227 [amdgpu]
[47269.571196] Modules linked in: fuse nf_conntrack_netlink xfrm_user xfrm_algo br_netfilter nf_nat_tftp tun bridge stp llc nft_masq ebtable_nat ip_set nft_chain_nat ebtable_filter ebtables nft_objref nft_reject_inet nf_reject_ipv6 nft_reject nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib xt_MASQUERADE xt_addrtype nf_tables_set iptable_nat ipt_REJECT nf_reject_i
pv4 xt_conntrack iptable_filter xt_CHECKSUM nf_conntrack_tftp nf_conntrack_netbios_ns nf_conntrack_broadcast xt_tcpudp iptable_mangle ip6table_filter ip6table_security ip6table_raw nft_ct ip6table_mangle ip6table_nat nf_nat nf_conntrack bluetooth nf_tables nf_defrag_ipv6 nf_defrag_ipv4 ip6_tables iptable_security iptable_raw ip_tables nfnetlink ecdh_generic ecc
crc16 x_tables bpfilter amdgpu snd_hda_codec_realtek snd_hda_codec_generic kvm_amd ledtrig_audio snd_hda_codec_hdmi gpu_sched ttm ccp snd_hda_intel snd_hda_codec rng_core kvm eeepc_wmi drm_kms_helper snd_hda_core asus_wmi snd_hwdep snd_pcm sparse_keymap
[47269.571214] syscopyarea sysfillrect sysimgblt rfkill sp5100_tco wmi_bmof snd_timer pcspkr irqbypass fb_sys_fops k10temp input_leds drm snd i2c_piix4 led_class soundcore mousedev joydev gpio_amdpt gpio_generic evdev mac_hid acpi_cpufreq nls_iso8859_1 nls_cp437 vfat fat sch_fq_codel sctp lm92 btrfs xor raid6_pq dm_cache_smq dm_cache dm_persistent_data dm_bio_p
rison dm_bufio libcrc32c crc32c_generic sd_mod usbkbd hid_generic usbmouse usbhid hid crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel ahci libahci aesni_intel libata igb aes_x86_64 crypto_simd xhci_pci xhci_hcd cryptd i2c_algo_bit scsi_mod glue_helper dca hwmon wmi pinctrl_amd sunrpc dm_mirror dm_region_hash dm_log dm_mod
[47269.571236] CPU: 7 PID: 1150 Comm: X Tainted: G W T 5.3.1-gentoo #7
[47269.571237] Hardware name: System manufacturer System Product Name/ROG STRIX B350-F GAMING, BIOS 5216 09/02/2019
[47269.571278] RIP: 0010:dcn10_verify_allow_pstate_change_high.cold+0xc/0x227 [amdgpu]
[47269.571279] Code: 83 c8 ff e9 09 02 f9 ff 48 c7 c7 98 1e df c0 e8 53 a8 97 de 0f 0b 83 c8 ff e9 f3 01 f9 ff 48 c7 c7 98 1e df c0 e8 3d a8 97 de <0f> 0b 80 bb 9f 01 00 00 00 75 05 e9 fb 24 f9 ff 48 8b 83 f8 02 00
[47269.571280] RSP: 0018:ffffaab7c120f750 EFLAGS: 00010246
[47269.571281] RAX: 0000000000000024 RBX: ffff99fb2cd00000 RCX: 0000000000000006
[47269.571282] RDX: 0000000000000000 RSI: 0000000000000082 RDI: ffff99fb307d64d0
[47269.571283] RBP: ffff99fb21f39800 R08: 000000007ffffffd R09: 000000000000057d
[47269.571284] R10: 000000000000003b R11: ffffaab7c120f600 R12: ffff99fb2cd00000
[47269.571284] R13: ffff99fac6661d18 R14: ffff99fac66601b8 R15: ffff99f8ac70c000
[47269.571285] FS: 00007f1c1d65d480(0000) GS:ffff99fb307c0000(0000) knlGS:0000000000000000
[47269.571286] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[47269.571287] CR2: 0000565061480000 CR3: 00000003b6916000 CR4: 00000000003406e0
[47269.571288] Call Trace:
[47269.571331] dc_stream_set_cursor_attributes+0x11c/0x170 [amdgpu]
[47269.571375] handle_cursor_update.isra.0+0x1f1/0x350 [amdgpu]
[47269.571418] amdgpu_dm_commit_cursors.isra.0+0x56/0x70 [amdgpu]
[47269.571461] amdgpu_dm_atomic_commit_tail+0x12b9/0x1c50 [amdgpu]
[47269.571464] ? enqueue_task_rt+0x9d/0xc0
[47269.571468] ? ttm_eu_backoff_reservation+0x3f/0x80 [ttm]
[47269.571510] ? dm_plane_helper_prepare_fb+0x19c/0x2d0 [amdgpu]
[47269.571512] ? _cond_resched+0x10/0x20
[47269.571513] ? wait_for_completion_timeout+0x31/0x110
[47269.571514] ? wait_for_completion_interruptible+0x2e/0x140
[47269.571516] ? complete_all+0x2b/0x40
[47269.571521] ? commit_tail+0x37/0x60 [drm_kms_helper]
[47269.571526] commit_tail+0x37/0x60 [drm_kms_helper]
[47269.571530] drm_atomic_helper_commit+0x103/0x110 [drm_kms_helper]
[47269.571535] drm_atomic_helper_update_plane+0xe7/0x100 [drm_kms_helper]
[47269.571543] drm_mode_cursor_universal+0x127/0x240 [drm]
[47269.571552] drm_mode_cursor_common+0xd9/0x230 [drm]
[47269.571559] ? drm_mode_cursor_ioctl+0x70/0x70 [drm]
[47269.571566] drm_ioctl_kernel+0xa2/0xf0 [drm]
[47269.571572] drm_ioctl+0x200/0x380 [drm]
[47269.571580] ? drm_mode_cursor_ioctl+0x70/0x70 [drm]
[47269.571609] amdgpu_drm_ioctl+0x41/0x70 [amdgpu]
[47269.571612] do_vfs_ioctl+0x44a/0x6e0
[47269.571613] ksys_ioctl+0x59/0x90
[47269.571614] __x64_sys_ioctl+0x11/0x20
[47269.571616] do_syscall_64+0x4a/0x110
[47269.571618] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[47269.571619] RIP: 0033:0x7f1c1e2a7597
[47269.571621] Code: 00 00 00 75 0c 48 c7 c0 ff ff ff ff 48 83 c4 18 c3 e8 8d d9 01 00 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d c9 48 0d 00 f7 d8 64 89 01 48
[47269.571621] RSP: 002b:00007fff5dd14a48 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
[47269.571622] RAX: ffffffffffffffda RBX: 00007fff5dd14a80 RCX: 00007f1c1e2a7597
[47269.571623] RDX: 00007fff5dd14a80 RSI: 00000000c02464bb RDI: 000000000000000c
[47269.571624] RBP: 00000000c02464bb R08: 0000565061371710 R09: 0000000000003fff
[47269.571625] R10: 000000000000007f R11: 0000000000000246 R12: 0000565060457200
[47269.571626] R13: 000000000000000c R14: 0000000000000004 R15: 0000565060459fa0
[47269.571627] ---[ end trace 0b5db3fd3a295d4f ]---
-- 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/drm/amd/issues/920. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
Created attachment 145518 [details] dmesg My system runs fine until I try to start com.valvesoftware.Steam or org.signal.Signal via Flatpak (org.zulip.Zulip works fine). Then the graphical X screen locks up and does not react to input (key presses, mouse clicks) anymore. Moving the mouse cursor is visible. I can move to another VT (ctrl+alt+F2), log in and `flatpak kill ...` the offending application, switch back to the first VT and the system is back to normal without any obvious damage. In dmesg I will find a bunch of messages like these: [ 204.449968] amdgpu 0000:0a:00.0: [gfxhub] retry page fault (src_id:0 ring:0 vmid:1 pasid:32770, for process X pid 2148 thread X:cs0 pid 2151) [ 204.449977] amdgpu 0000:0a:00.0: in page starting at address 0x0000000107720000 from 27 [ 204.449982] amdgpu 0000:0a:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x00101031 Device 0000:0a:00.0 is an AMD Ryzen 5 2400G.