Amdgpu crashes running Steam snap on RX 9070 XT with Ubuntu 25.04

Ubuntu Version: 25.04

Desktop Environment (if applicable): GNOME

Problem Description:

I just purchased a RX 9070 XT as an upgrade from my RX 5700 XT, and I’m trying it out on my existing Ubuntu 25.04 installation. When I try to launch Steam, GNOME shell seems to freeze for a few seconds, and Steam never successfully launches.

Relevant System Information:

ASRock Steel Legend AMD RX 9070 XT GPU.

Kernel: 6.14.0-15-generic #15-Ubuntu

I’m running Mesa 25:

[I] oac@station ~> apt list --installed | grep mesa
libegl-mesa0/plucky,now 25.0.3-1ubuntu2 amd64 [installed,automatic]
libegl1-mesa-dev/plucky,now 25.0.3-1ubuntu2 amd64 [installed,automatic]
libgl1-mesa-dri/plucky,now 25.0.3-1ubuntu2 amd64 [installed,automatic]
libglu1-mesa/plucky,now 9.0.2-1.1build1 amd64 [installed,automatic]
libglx-mesa0/plucky,now 25.0.3-1ubuntu2 amd64 [installed,automatic]
mesa-libgallium/plucky,now 25.0.3-1ubuntu2 amd64 [installed,automatic]
mesa-vdpau-drivers/plucky,now 25.0.3-1ubuntu2 amd64 [installed,automatic]
mesa-vulkan-drivers/plucky,now 25.0.3-1ubuntu2 amd64 [installed,automatic]

And I have libdrm-radeon1/plucky,now 2.4.124-2 amd64 [installed,automatic]

Additionally, here is my system info from GNOME Settings:

Hardware Information:

  • Hardware Model: Gigabyte Technology Co., Ltd. B650 GAMING X AX V2
  • Memory: 32.0 GiB
  • Processor: AMD Ryzen™ 7 9700X × 16
  • Graphics: AMD Radeon™ Graphics
  • Graphics 1: AMD Radeon™ Graphics
  • Disk Capacity: 4.0 TB

Software Information:

  • Firmware Version: F30
  • OS Name: Ubuntu 25.04
  • OS Build: (null)
  • OS Type: 64-bit
  • GNOME Version: 48
  • Windowing System: Wayland
  • Kernel Version: Linux 6.14.0-15-generic

Screenshots or Error Messages:

Steam errors from snap run steam:

'gfx1201' is not a recognized processor for this target (ignoring processor)
'gfx1201' is not a recognized processor for this target (ignoring processor)
amd: LLVM doesn't support gfx1201, bailing out...
glx: failed to create dri3 screen
failed to load driver: radeonsi
05/19 14:28:26 minidumps folder is set to /tmp/dumps
05/19 14:28:26 Init: Installing breakpad exception handler for appid(steamsysinfo)/version(1745876290)/tid(20045)
Running query: 1 - GpuTopology
Response: gpu_topology {
  gpus {
    id: 1
    name: "AMD Radeon Graphics (RADV RAPHAEL_MENDOCINO)"
    vram_size_bytes: 11028205568
    driver_id: k_EGpuDriverId_MesaRadv
    driver_version_major: 24
    driver_version_minor: 2
    driver_version_patch: 3
  }
  gpus {
    id: 2
    name: "AMD Radeon Graphics (RADV GFX1201)"
    vram_size_bytes: 17095983104
    driver_id: k_EGpuDriverId_MesaRadv
    driver_version_major: 24
    driver_version_minor: 2
    driver_version_patch: 3
  }
  default_gpu_id: 1
}

Exit code: 0

Importantly, note that the mesa version detected is 24.2.3, not the 25.0.3 I have installed on my system.

Errors from dmesg:

[  225.612907] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:7 pasid:32787)
[  225.612915] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20313 thread steamwebh:zfq0 pid 20340)
[  225.612917] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x000042000006c000 from client 10
[  225.612918] amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x007012B0
[  225.612920] amdgpu 0000:03:00.0: amdgpu: 	 Faulty UTCL2 client ID: SQC (inst) (0x9)
[  225.612921] amdgpu 0000:03:00.0: amdgpu: 	 MORE_FAULTS: 0x0
[  225.612922] amdgpu 0000:03:00.0: amdgpu: 	 WALKER_ERROR: 0x0
[  225.612922] amdgpu 0000:03:00.0: amdgpu: 	 PERMISSION_FAULTS: 0xb
[  225.612923] amdgpu 0000:03:00.0: amdgpu: 	 MAPPING_ERROR: 0x0
[  225.612924] amdgpu 0000:03:00.0: amdgpu: 	 RW: 0x0
[  236.091135] amdgpu 0000:03:00.0: amdgpu: Dumping IP State
[  236.092318] amdgpu 0000:03:00.0: amdgpu: Dumping IP State Completed
[  236.092434] amdgpu 0000:03:00.0: amdgpu: ring gfx_0.0.0 timeout, but soft recovered
[  236.092753] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:7 pasid:32787)
[  236.092757] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20313 thread steamwebh:zfq0 pid 20340)
[  236.092758] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000c1000006c000 from client 10
[  236.092761] amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x007012B1
[  236.092762] amdgpu 0000:03:00.0: amdgpu: 	 Faulty UTCL2 client ID: SQC (inst) (0x9)
[  236.092763] amdgpu 0000:03:00.0: amdgpu: 	 MORE_FAULTS: 0x1
[  236.092765] amdgpu 0000:03:00.0: amdgpu: 	 WALKER_ERROR: 0x0
[  236.092766] amdgpu 0000:03:00.0: amdgpu: 	 PERMISSION_FAULTS: 0xb
[  236.092767] amdgpu 0000:03:00.0: amdgpu: 	 MAPPING_ERROR: 0x0
[  236.092768] amdgpu 0000:03:00.0: amdgpu: 	 RW: 0x0
[  236.092776] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:7 pasid:32787)
[  236.092778] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20313 thread steamwebh:zfq0 pid 20340)
[  236.092779] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000e4000006c000 from client 10
[  236.092787] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:7 pasid:32787)
[  236.092788] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20313 thread steamwebh:zfq0 pid 20340)
[  236.092789] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000b70000000000 from client 10
[  236.092797] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:7 pasid:32787)
[  236.092799] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20313 thread steamwebh:zfq0 pid 20340)
[  236.092800] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x000042000006c000 from client 10
[  236.092807] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:7 pasid:32787)
[  236.092809] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20313 thread steamwebh:zfq0 pid 20340)
[  236.092810] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000c10000000000 from client 10
[  236.606712] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  236.606719] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  236.606720] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000c2000006b000 from client 10
[  236.606721] amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x001012B0
[  236.606722] amdgpu 0000:03:00.0: amdgpu: 	 Faulty UTCL2 client ID: SQC (inst) (0x9)
[  236.606723] amdgpu 0000:03:00.0: amdgpu: 	 MORE_FAULTS: 0x0
[  236.606723] amdgpu 0000:03:00.0: amdgpu: 	 WALKER_ERROR: 0x0
[  236.606724] amdgpu 0000:03:00.0: amdgpu: 	 PERMISSION_FAULTS: 0xb
[  236.606724] amdgpu 0000:03:00.0: amdgpu: 	 MAPPING_ERROR: 0x0
[  236.606725] amdgpu 0000:03:00.0: amdgpu: 	 RW: 0x0
[  246.844005] amdgpu 0000:03:00.0: amdgpu: ring gfx_0.0.0 timeout, but soft recovered
[  246.844103] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844106] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844108] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000c1000006d000 from client 10
[  246.844109] amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x001012B1
[  246.844111] amdgpu 0000:03:00.0: amdgpu: 	 Faulty UTCL2 client ID: SQC (inst) (0x9)
[  246.844114] amdgpu 0000:03:00.0: amdgpu: 	 MORE_FAULTS: 0x1
[  246.844115] amdgpu 0000:03:00.0: amdgpu: 	 WALKER_ERROR: 0x0
[  246.844116] amdgpu 0000:03:00.0: amdgpu: 	 PERMISSION_FAULTS: 0xb
[  246.844116] amdgpu 0000:03:00.0: amdgpu: 	 MAPPING_ERROR: 0x0
[  246.844119] amdgpu 0000:03:00.0: amdgpu: 	 RW: 0x0
[  246.844126] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844130] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844131] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000a4000006b000 from client 10
[  246.844138] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844139] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844140] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000b7000006d000 from client 10
[  246.844146] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844147] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844147] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000c2000006d000 from client 10
[  246.844154] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844155] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844156] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000c1000006d000 from client 10
[  246.844168] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844168] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844169] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000a4000006d000 from client 10
[  246.844176] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844177] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844178] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000b7000006d000 from client 10
[  246.844182] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844183] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844183] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000c2000006d000 from client 10
[  246.844188] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844189] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844189] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000c1000006d000 from client 10
[  246.844204] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:1 pasid:32789)
[  246.844206] amdgpu 0000:03:00.0: amdgpu:  in process steamwebhelper pid 20511 thread steamwebh:zfq0 pid 20526)
[  246.844206] amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000a40000000000 from client 10

(there are lots more of those in the dmesg logs)

What I’ve Tried:

I see this previous post about 24.10: 9070 XT on Ubuntu 24.10 - #2 by YamiYukiSenpai

Steam snap doesn’t use the system’s Mesa libraries.

It’s that they’re using Mesa from a Snap. The issue, however, is that the gaming-graphics-core22 hasn’t been updated to Mesa 25

I haven’t tried using the steam .deb yet. I’m about to reboot into Bazzite (based on Fedora 42) and see if everything is alright there.

I also saw this: https://forums.linuxmint.com/viewtopic.php?t=441828#firmware but I assume the firmware is already present in the 6.14 kernel which ships with Ubuntu 25.04.

If Bazzite works, I’ll try installing the Steam .deb package instead, and see if the situation improves on Ubuntu 25.04.

It does seem like the Steam snap mesa 24 is the problem, however, and it should be updated to ship mesa 25 as soon as possible.

Hmm, the Tuba snap also fails to run:

[I] oac@station ~> snap run tuba
/usr/share/libdrm/amdgpu.ids: No such file or directory
/usr/share/libdrm/amdgpu.ids: No such file or directory
Gdk-Message: 15:07:18.594: Error 71 (Protocol error) dispatching to Wayland display.

So this is a problem beyond just the Steam snap.

Looking around, I see a similar issue from 2020 around a missing amdgpu.ics file: AMD OSS graphics drivers missing access to amdgpu.ids file - #14 by James-Carroll - snap - snapcraft.io

The solution was to manually add the following to the snap.yaml:

layout:
  /usr/share/libdrm:
    bind: $SNAP/usr/share/libdrm

However, the tuba snap already hasa layout defined for /usr/share/libdrm:

layout:
  /usr/lib/x86_64-linux-gnu/webkit2gtk-4.0:
    bind: $SNAP/gnome-platform/usr/lib/x86_64-linux-gnu/webkit2gtk-4.0
  /usr/lib/x86_64-linux-gnu/webkit2gtk-4.1:
    bind: $SNAP/gnome-platform/usr/lib/x86_64-linux-gnu/webkit2gtk-4.1
  /usr/share/xml/iso-codes:
    bind: $SNAP/gnome-platform/usr/share/xml/iso-codes
  /usr/share/libdrm:
    bind: $SNAP/gpu-2404/libdrm
  /usr/share/drirc.d:
    symlink: $SNAP/gpu-2404/drirc.d
  /usr/share/X11/XErrorDB:
    symlink: $SNAP/gpu-2404/X11/XErrorDB

Looking some more at snap.yaml, I see that gpu-2404 is defined as follows:

  gpu-2404:
    interface: content
    target: $SNAP/gpu-2404
    default-provider: mesa-2404

So it’s provided by mesa-2404, which has snap info mesa-2404:

tracking:     latest/stable
refresh-date: 2025-02-19
channels:
  latest/stable:    24.2.8         2025-02-19 (495) 217MB -
  latest/candidate: ↑                                     
  latest/beta:      24.2.8-snap172 2025-04-02 (699) 304MB -
  latest/edge:      ↑                                     
installed:          24.2.8                    (495) 217MB -

Again, mesa 24, which doesn’t support AMD 9000 series cards.

Would it be possible to upgrade the mesa-2404 snap to mesa 25, or would that potentially introduce incompatibilities with Ubuntu 24.04? There isn’t any newer mesa snap available:

[I] oac@station ~> snap search mesa
Name           Version                Publisher   Notes  Summary
mesa-core22    23.2.1                 canonical✓  -      mesa libraries for core22 snaps
mesa-2404      24.2.8                 canonical✓  -      Mesa libraries for core24 snaps
nvidia-core22  535.230.02+mesa23.2.1  canonical✓  -      NVIDIA and Mesa libraries for core22 snaps

Alternatively, would it be possible to release a separate track of the mesa-2404 snap, or some other snap which provides gpu-2404, which would use mesa 25 instead? Otherwise I’m not sure how lots of graphical snaps will be able to run on machines with AMD RX 9000 series graphics cards until Ubuntu 26.04 is released.

You might need to switch to one of the kisak or obiaf channels for mesa-2404 …

1 Like

Did you read this? There seems to be some work behind the scenes. I’m not on Ubuntu right now and don’t have Snap installed, but there seems to be a hidden channel with Mesa 25.

2 Likes

I’ll give that a try, thanks! What’s the difference between kisak and obiaf? Are those channels documented somewhere?

Thanks for the heads up! It seems like the beta/kisak channel has expired, as now refreshing to it installs the same version as latest/beta, which is 24.2.8-snap172:

tracking:     latest/beta/kisak
refresh-date: today at 16:58 CDT
channels:
  latest/stable:    24.2.8         2025-02-19 (495) 217MB -
  latest/candidate: ↑                                     
  latest/beta:      24.2.8-snap172 2025-04-02 (699) 304MB -
  latest/edge:      ↑                                     
installed:          24.2.8-snap172            (699) 304MB -

Also @ogra, I poked around but couldn’t construct a channel name with oibaf that was recognized. Do you know the channel name to use?

I’m assuming the mesa-2404 latest/beta/kisak channel expired, but even if it hadn’t, I don’t think the steam snap would be improved, since it uses gaming-graphics-core22, which is also still on mesa 24.

Do you know if there are plans to make similar channels for gaming-graphics-core22, or to migrate steam or other snaps to mesa-2404? @YamiYukiSenpai has opened issues for these suggestions:

CC @saviq, could you perhaps check if the beta/kisak channel of mesa-2404 closed or was otherwise pushed to incorrectly? It seems it gives the same revision as latest/beta, which has 24.2.8-snap172:

tracking:     latest/beta/kisak
refresh-date: today at 16:58 CDT
channels:
  latest/stable:    24.2.8         2025-02-19 (495) 217MB -
  latest/candidate: ↑                                     
  latest/beta:      24.2.8-snap172 2025-04-02 (699) 304MB -
  latest/edge:      ↑                                     
installed:          24.2.8-snap172            (699) 304MB 

Checking the latest build logs from mesa-2404-kisak : Snap packages : “Mir development team” team (which was from 2025-05-06), they succeeded and built version 25.0.5:

Packed mesa-2404_25.0.5-snap173_amd64.snap

So I’m confused why that newer version isn’t in the beta/kisak channel.

1 Like

I assume @jibel might be able to point us towards someone who’d be working on this…
(Not sure what the current plans are for maintenance of the snap)

2 Likes