Bug 82886

Summary:	[RadeonSI]GPU Lockup with Linux 3.16 & Mesa Git
Product:	Mesa	Reporter:	mmstickman
Component:	Drivers/Gallium/radeonsi	Assignee:	Default DRI bug account <dri-devel>
Status:	RESOLVED INVALID	QA Contact:
Severity:	normal
Priority:	medium	CC:	albandil
Version:	git
Hardware:	x86-64 (AMD64)
OS:	Linux (All)
Whiteboard:
i915 platform:		i915 features:

Description mmstickman 2014-08-21 03:41:54 UTC

My Radeon HD 7950 GPU locks up and/or Xorg completely crashes whenever the system is graphics card is stressed slightly sometimes (even so much as displaying a web page) -- forcing me to have to reboot the system. I've also been able to trigger it simply by running a game like Tesseract, where it will completely crash Xorg on the main menu.

[104725.380514] radeon 0000:01:00.0: ring 0 stalled for more than 10033msec
[104725.380524] radeon 0000:01:00.0: GPU lockup (waiting for 0x000000000085f665 last fence id 0x000000000085f65a on ring 0)
[104725.921526] radeon 0000:01:00.0: Saved 957 dwords of commands on ring 0.
[104725.921574] radeon 0000:01:00.0: GPU softreset: 0x00000068
[104725.921576] radeon 0000:01:00.0:   GRBM_STATUS               = 0xA0003028
[104725.921578] radeon 0000:01:00.0:   GRBM_STATUS_SE0           = 0x00000006
[104725.921580] radeon 0000:01:00.0:   GRBM_STATUS_SE1           = 0x00000006
[104725.921582] radeon 0000:01:00.0:   SRBM_STATUS               = 0x200000C0
[104725.921617] radeon 0000:01:00.0:   SRBM_STATUS2              = 0x00000000
[104725.921619] radeon 0000:01:00.0:   R_008674_CP_STALLED_STAT1 = 0x00000000
[104725.921621] radeon 0000:01:00.0:   R_008678_CP_STALLED_STAT2 = 0x00010000
[104725.921623] radeon 0000:01:00.0:   R_00867C_CP_BUSY_STAT     = 0x00000002
[104725.921625] radeon 0000:01:00.0:   R_008680_CP_STAT          = 0x80010243
[104725.921626] radeon 0000:01:00.0:   R_00D034_DMA_STATUS_REG   = 0x44C83D57
[104725.921628] radeon 0000:01:00.0:   R_00D834_DMA_STATUS_REG   = 0x44E84246
[104725.921630] radeon 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x00000000
[104725.921632] radeon 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x00000000
[104726.446215] radeon 0000:01:00.0: GRBM_SOFT_RESET=0x0000DDFF
[104726.446269] radeon 0000:01:00.0: SRBM_SOFT_RESET=0x00000140
[104726.447426] radeon 0000:01:00.0:   GRBM_STATUS               = 0x00003028
[104726.447428] radeon 0000:01:00.0:   GRBM_STATUS_SE0           = 0x00000006
[104726.447430] radeon 0000:01:00.0:   GRBM_STATUS_SE1           = 0x00000006
[104726.447432] radeon 0000:01:00.0:   SRBM_STATUS               = 0x20000AC0
[104726.447467] radeon 0000:01:00.0:   SRBM_STATUS2              = 0x00000000
[104726.447469] radeon 0000:01:00.0:   R_008674_CP_STALLED_STAT1 = 0x00000000
[104726.447470] radeon 0000:01:00.0:   R_008678_CP_STALLED_STAT2 = 0x00000000
[104726.447472] radeon 0000:01:00.0:   R_00867C_CP_BUSY_STAT     = 0x00000000
[104726.447474] radeon 0000:01:00.0:   R_008680_CP_STAT          = 0x00000000
[104726.447476] radeon 0000:01:00.0:   R_00D034_DMA_STATUS_REG   = 0x44C83D57
[104726.447478] radeon 0000:01:00.0:   R_00D834_DMA_STATUS_REG   = 0x44C83D57
[104726.447562] radeon 0000:01:00.0: GPU reset succeeded, trying to resume
[104726.496373] [drm] probing gen 2 caps for device 1002:5a16 = 31cd02/0
[104726.496376] [drm] PCIE gen 2 link speeds already enabled
[104726.499844] [drm] PCIE GART of 1024M enabled (table at 0x0000000000276000).
[104726.499980] radeon 0000:01:00.0: WB enabled
[104726.499983] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x00000000c0000c00 and cpu addr 0xffff880420106c00
[104726.499985] radeon 0000:01:00.0: fence driver on ring 1 use gpu addr 0x00000000c0000c04 and cpu addr 0xffff880420106c04
[104726.499987] radeon 0000:01:00.0: fence driver on ring 2 use gpu addr 0x00000000c0000c08 and cpu addr 0xffff880420106c08
[104726.499989] radeon 0000:01:00.0: fence driver on ring 3 use gpu addr 0x00000000c0000c0c and cpu addr 0xffff880420106c0c
[104726.499990] radeon 0000:01:00.0: fence driver on ring 4 use gpu addr 0x00000000c0000c10 and cpu addr 0xffff880420106c10
[104726.500376] radeon 0000:01:00.0: fence driver on ring 5 use gpu addr 0x0000000000075a18 and cpu addr 0xffffc90011cb5a18
[104726.672097] [drm] ring test on 0 succeeded in 1 usecs
[104726.672103] [drm] ring test on 1 succeeded in 1 usecs
[104726.672107] [drm] ring test on 2 succeeded in 1 usecs
[104726.672167] [drm] ring test on 3 succeeded in 2 usecs
[104726.672174] [drm] ring test on 4 succeeded in 1 usecs
[104726.869612] [drm] ring test on 5 succeeded in 2 usecs
[104726.869617] [drm] UVD initialized successfully.
[104736.885653] radeon 0000:01:00.0: ring 0 stalled for more than 10000msec
[104736.885663] radeon 0000:01:00.0: GPU lockup (waiting for 0x000000000085f66b last fence id 0x000000000085f65a on ring 0)
[104736.885689] [drm:r600_ib_test] *ERROR* radeon: fence wait failed (-35).
[104736.885696] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB on GFX ring (-35).
[104736.885700] radeon 0000:01:00.0: ib ring test failed (-35).
[104737.072553] radeon 0000:01:00.0: ring 0 stalled for more than 10186msec
[104737.072563] radeon 0000:01:00.0: GPU lockup (waiting for 0x000000000085f665 last fence id 0x000000000085f65a on ring 0)
[104737.395007] radeon 0000:01:00.0: GPU softreset: 0x00000048
[104737.395010] radeon 0000:01:00.0:   GRBM_STATUS               = 0xA0003028
[104737.395012] radeon 0000:01:00.0:   GRBM_STATUS_SE0           = 0x00000006
[104737.395014] radeon 0000:01:00.0:   GRBM_STATUS_SE1           = 0x00000006
[104737.395016] radeon 0000:01:00.0:   SRBM_STATUS               = 0x200000C0
[104737.395051] radeon 0000:01:00.0:   SRBM_STATUS2              = 0x00000000
[104737.395053] radeon 0000:01:00.0:   R_008674_CP_STALLED_STAT1 = 0x00000000
[104737.395055] radeon 0000:01:00.0:   R_008678_CP_STALLED_STAT2 = 0x00010000
[104737.395057] radeon 0000:01:00.0:   R_00867C_CP_BUSY_STAT     = 0x00400002
[104737.395059] radeon 0000:01:00.0:   R_008680_CP_STAT          = 0x84010243
[104737.395061] radeon 0000:01:00.0:   R_00D034_DMA_STATUS_REG   = 0x44C83D57
[104737.395062] radeon 0000:01:00.0:   R_00D834_DMA_STATUS_REG   = 0x44C83D57
[104737.395065] radeon 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x00000000
[104737.395067] radeon 0000:01:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x00000000
[104737.895759] radeon 0000:01:00.0: GRBM_SOFT_RESET=0x0000DDFF
[104737.895829] radeon 0000:01:00.0: SRBM_SOFT_RESET=0x00000100
[104737.896987] radeon 0000:01:00.0:   GRBM_STATUS               = 0x00003028
[104737.896989] radeon 0000:01:00.0:   GRBM_STATUS_SE0           = 0x00000006
[104737.896991] radeon 0000:01:00.0:   GRBM_STATUS_SE1           = 0x00000006
[104737.896993] radeon 0000:01:00.0:   SRBM_STATUS               = 0x200000C0
[104737.897027] radeon 0000:01:00.0:   SRBM_STATUS2              = 0x00000000
[104737.897067] radeon 0000:01:00.0:   R_008674_CP_STALLED_STAT1 = 0x00000000
[104737.897071] radeon 0000:01:00.0:   R_008678_CP_STALLED_STAT2 = 0x00000000
[104737.897073] radeon 0000:01:00.0:   R_00867C_CP_BUSY_STAT     = 0x00000000
[104737.897075] radeon 0000:01:00.0:   R_008680_CP_STAT          = 0x00000000
[104737.897078] radeon 0000:01:00.0:   R_00D034_DMA_STATUS_REG   = 0x44C83D57
[104737.897080] radeon 0000:01:00.0:   R_00D834_DMA_STATUS_REG   = 0x44C83D57
[104737.897164] radeon 0000:01:00.0: GPU reset succeeded, trying to resume
[104737.914092] [drm] probing gen 2 caps for device 1002:5a16 = 31cd02/0
[104737.914095] [drm] PCIE gen 2 link speeds already enabled
[104737.917563] [drm] PCIE GART of 1024M enabled (table at 0x0000000000276000).
[104737.917737] radeon 0000:01:00.0: WB enabled
[104737.917739] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x00000000c0000c00 and cpu addr 0xffff880420106c00
[104737.917741] radeon 0000:01:00.0: fence driver on ring 1 use gpu addr 0x00000000c0000c04 and cpu addr 0xffff880420106c04
[104737.917743] radeon 0000:01:00.0: fence driver on ring 2 use gpu addr 0x00000000c0000c08 and cpu addr 0xffff880420106c08
[104737.917745] radeon 0000:01:00.0: fence driver on ring 3 use gpu addr 0x00000000c0000c0c and cpu addr 0xffff880420106c0c
[104737.917747] radeon 0000:01:00.0: fence driver on ring 4 use gpu addr 0x00000000c0000c10 and cpu addr 0xffff880420106c10
[104737.918131] radeon 0000:01:00.0: fence driver on ring 5 use gpu addr 0x0000000000075a18 and cpu addr 0xffffc90011cb5a18
[104738.088583] [drm] ring test on 0 succeeded in 1 usecs
[104738.088588] [drm] ring test on 1 succeeded in 1 usecs
[104738.088592] [drm] ring test on 2 succeeded in 1 usecs
[104738.088652] [drm] ring test on 3 succeeded in 2 usecs
[104738.088658] [drm] ring test on 4 succeeded in 1 usecs
[104738.286093] [drm] ring test on 5 succeeded in 2 usecs
[104738.286098] [drm] UVD initialized successfully.
[104738.286119] [drm] ib test on ring 0 succeeded in 0 usecs
[104738.286137] [drm] ib test on ring 1 succeeded in 0 usecs
[104738.286155] [drm] ib test on ring 2 succeeded in 0 usecs
[104738.286173] [drm] ib test on ring 3 succeeded in 0 usecs
[104738.286190] [drm] ib test on ring 4 succeeded in 0 usecs
[104738.457645] [drm:uvd_v1_0_ib_test] *ERROR* radeon: failed to get destroy ib (-22).
[104738.457650] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB on ring 5 (-22).
[104738.457665] [drm:radeon_pm_resume_dpm] *ERROR* radeon: dpm resume failed

Comment 1 Christian König 2014-08-21 09:06:30 UTC


*** This bug has been marked as a duplicate of bug 79980 ***

Comment 2 Michel Dänzer 2014-10-31 03:21:36 UTC

Does this also happen with Mesa 10.2.y?

Comment 3 Jarkko K 2014-12-08 01:49:10 UTC

I think you should try to update kernel and mesa. Both are updated regularly.

Comment 4 Marek Olšák 2015-08-02 11:25:06 UTC

(In reply to Michel Dänzer from comment #2)
> Does this also happen with Mesa 10.2.y?

No feedback after 8 months. Closing.

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.