On Sat, Jan 12, 2019 at 09:50:51PM +0100, Borislav Petkov wrote:
> Hi guys,
>
> my odyssey with the GPU continues. This time it didn't reset itself
> but started spewing a single line about the hardware locking up.
>
> The machine was responsive to sysrq so I was able to write out
> /var/log/messages and reboot.
>
> This is still with 4.20-rc7 but I'm building 5.0-rc1 to see if there's a
> difference.
Well, not really. This time the reset succeeded and the machine is still
alive:
[111333.620619] radeon 0000:1d:00.0: ring 0 stalled for more than 10360msec
[111333.620626] radeon 0000:1d:00.0: GPU lockup (current fence id 0x000000000080f31d last fence id 0x000000000080f416 on ring 0)
[111334.132277] radeon 0000:1d:00.0: ring 0 stalled for more than 10872msec
[111334.132283] radeon 0000:1d:00.0: GPU lockup (current fence id 0x000000000080f31d last fence id 0x000000000080f418 on ring 0)
[111334.199083] radeon 0000:1d:00.0: failed to get a new IB (-35)
[111334.199107] [drm:radeon_cs_ioctl [radeon]] *ERROR* Failed to get ib !
[111334.206116] radeon 0000:1d:00.0: Saved 8121 dwords of commands on ring 0.
[111334.206127] radeon 0000:1d:00.0: GPU softreset: 0x00000008
[111334.206130] radeon 0000:1d:00.0: R_008010_GRBM_STATUS = 0xA0001030
[111334.206132] radeon 0000:1d:00.0: R_008014_GRBM_STATUS2 = 0x00000003
[111334.206135] radeon 0000:1d:00.0: R_000E50_SRBM_STATUS = 0x200000C0
[111334.206137] radeon 0000:1d:00.0: R_008674_CP_STALLED_STAT1 = 0x00000000
[111334.206139] radeon 0000:1d:00.0: R_008678_CP_STALLED_STAT2 = 0x00000000
[111334.206141] radeon 0000:1d:00.0: R_00867C_CP_BUSY_STAT = 0x00020182
[111334.206144] radeon 0000:1d:00.0: R_008680_CP_STAT = 0x80028645
[111334.206146] radeon 0000:1d:00.0: R_00D034_DMA_STATUS_REG = 0x44C83D57
[111334.272194] radeon 0000:1d:00.0: R_008020_GRBM_SOFT_RESET=0x00004001
[111334.272247] radeon 0000:1d:00.0: SRBM_SOFT_RESET=0x00000100
[111334.274336] radeon 0000:1d:00.0: R_008010_GRBM_STATUS = 0xA0003030
[111334.274338] radeon 0000:1d:00.0: R_008014_GRBM_STATUS2 = 0x00000003
[111334.274339] radeon 0000:1d:00.0: R_000E50_SRBM_STATUS = 0x200080C0
[111334.274341] radeon 0000:1d:00.0: R_008674_CP_STALLED_STAT1 = 0x00000000
[111334.274342] radeon 0000:1d:00.0: R_008678_CP_STALLED_STAT2 = 0x00000000
[111334.274344] radeon 0000:1d:00.0: R_00867C_CP_BUSY_STAT = 0x00000000
[111334.274345] radeon 0000:1d:00.0: R_008680_CP_STAT = 0x80100000
[111334.274347] radeon 0000:1d:00.0: R_00D034_DMA_STATUS_REG = 0x44C83D57
[111334.274354] radeon 0000:1d:00.0: GPU reset succeeded, trying to resume
[111334.290030] [drm] PCIE gen 2 link speeds already enabled
[111334.292121] [drm] PCIE GART of 512M enabled (table at 0x0000000000142000).
[111334.292135] radeon 0000:1d:00.0: WB enabled
[111334.292137] radeon 0000:1d:00.0: fence driver on ring 0 use gpu addr 0x0000000020000c00 and cpu addr 0x00000000fb2c042c
[111334.292325] radeon 0000:1d:00.0: fence driver on ring 5 use gpu addr 0x00000000000521d0 and cpu addr 0x0000000014f22c80
[111334.323193] [drm] ring test on 0 succeeded in 0 usecs
[111334.497890] [drm] ring test on 5 succeeded in 1 usecs
[111334.497896] [drm] UVD initialized successfully.
[111334.724316] [drm] ib test on ring 0 succeeded in 0 usecs
[111335.380416] [drm] ib test on ring 5 succeeded
--
Regards/Gruss,
Boris.
Good mailing practices for 400: avoid top-posting and trim the reply.
_______________________________________________
amd-gfx mailing list
amd-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
