[Bug 98975] Wasteland 2 Directors Cut: Hangs. GPU fault

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=98975

            Bug ID: 98975
           Summary: Wasteland 2 Directors Cut: Hangs. GPU fault
           Product: Mesa
           Version: git
          Hardware: Other
                OS: All
            Status: NEW
          Severity: normal
          Priority: medium
         Component: Drivers/Gallium/radeonsi
          Assignee: dri-devel at lists.freedesktop.org
          Reporter: sa at whiz.se
        QA Contact: dri-devel at lists.freedesktop.org

I'm having trouble with the game Wasteland 2 Directors Cut.

It sporadically freezes (most often when loading new areas) sometimes right
away after starting a new game sometimes after several hours of gameplay. 

I'm getting a massive log of VM fault, GPU_fault_detected as soon as the game
is launched.

#98183 and #98619 seems similar.

Probably not a regression, it's present with 4.7.0 and Mesa 12.

I also tried recording with apitrace, the trace itself doesn't seem to trigger
the VM fault stuff, but I do get these warnings:


809169: message: major shader compiler issue 29: 0:14(18): warning: `tmpvar_1'
used uninitialized
809169 @1 glCompileShaderARB(shaderObj = 1243)
809169: warning: 0:14(18): warning: `tmpvar_1' used uninitialized

809961: message: shader compiler issue 1: Shader Stats: SGPRS: 80 VGPRS: 16
Code Size: 92 LDS: 0 Scratch: 0 Max Waves: 10 Spilled SGPRs: 0 Spilled VGPRs: 0

System environment:
-- system architecture: 64-bit
-- Linux distribution: Debian unstable
-- GPU: TONGA
-- Model: Asus Strix R9 285 2GB
-- Display connector: DVI
-- xf86-video-amdgpu: 1.2.0
-- xserver: 1.19.0
-- mesa: c45d84ad8349d0c69893458d6c58eb5f6f1609c4
-- drm: 2.4.74
-- kernel: 4.8.0
-- dri3

dmesg:
[  409.259720] VM fault (0x04, vmid 1) at page 1653855, read from 'TC7'
(0x54433700) (68)
[  409.259722] amdgpu 0000:01:00.0: GPU fault detected: 146 0x03188404
[  409.259723] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR  
0x00193C5B
[  409.259724] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS
0x02048004
[  409.259724] VM fault (0x04, vmid 1) at page 1653851, read from 'TC6'
(0x54433600) (72)
[  409.259727] amdgpu 0000:01:00.0: GPU fault detected: 146 0x03188004
[  409.259727] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR  
0x00193C57
[  409.259728] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS
0x02048004
[  409.259729] VM fault (0x04, vmid 1) at page 1653847, read from 'TC6'
(0x54433600) (72)
[  409.259731] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0318c404
[  409.259732] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR  
0x00193C53
[  409.259732] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS
0x02084004
[  409.259733] VM fault (0x04, vmid 1) at page 1653843, read from 'TC10'
(0x54433130) (132)
[  409.259735] amdgpu 0000:01:00.0: GPU fault detected: 146 0x03384004
[  409.259736] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR  
0x00193C53
[  409.259737] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS
0x02044004
[  409.259737] VM fault (0x04, vmid 1) at page 1653843, read from 'TC7'
(0x54433700) (68)
[  409.259739] amdgpu 0000:01:00.0: GPU fault detected: 146 0x03384804
[  409.259740] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR  
0x00193C4B
[  409.259741] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS
0x02004004
[  409.259741] VM fault (0x04, vmid 1) at page 1653835, read from 'TC1'
(0x54433100) (4)
[  409.259743] amdgpu 0000:01:00.0: GPU fault detected: 146 0x03180004
[  409.259744] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR  
0x00193C4B
[  409.259745] amdgpu 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS
0x02084004
[  409.259745] VM fault (0x04, vmid 1) at page 1653835, read from 'TC10'
(0x54433130) (132)

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/dri-devel/attachments/20161202/24332e18/attachment.html>