endless GPU lockups on "ATI Radeon HD 5450" (ChipID = 0x68f9)

Sat Apr 19 10:16:28 UTC 2014

I had about ten minutes of video flickering without restoring image 
before reboot about half an our ago:

Apr 19 13:35:11 zealot kernel: drmn0: error: GPU lockup CP stall for 
more than 10000msec
Apr 19 13:35:11 zealot kernel: drmn0: warning: GPU lockup (waiting for 
0x00000000017dbd48 last fence id 0x00000000017dbd41)
Apr 19 13:35:11 zealot kernel: drmn0: info: Saved 215 dwords of commands 
on ring 0.
Apr 19 13:35:11 zealot kernel: drmn0: info: GPU softreset: 0x00000003
Apr 19 13:35:11 zealot kernel: drmn0: info:   GRBM_STATUS 
= 0xB2703828
Apr 19 13:35:11 zealot kernel: drmn0: info:   GRBM_STATUS_SE0 
= 0x1C000007
Apr 19 13:35:11 zealot kernel: drmn0: info:   GRBM_STATUS_SE1 
= 0x00000007
Apr 19 13:35:11 zealot kernel: drmn0: info:   SRBM_STATUS 
= 0x200000C0
Apr 19 13:35:11 zealot kernel: drmn0: info:   R_008674_CP_STALLED_STAT1 
= 0x00000000
Apr 19 13:35:11 zealot kernel: drmn0: info:   R_008678_CP_STALLED_STAT2 
= 0x40000000
Apr 19 13:35:11 zealot kernel: drmn0: info:   R_00867C_CP_BUSY_STAT 
= 0x00008006
Apr 19 13:35:11 zealot kernel: drmn0: info:   R_008680_CP_STAT 
= 0x80228647
Apr 19 13:35:11 zealot kernel: drmn0: info:   GRBM_SOFT_RESET=0x00007F6B
Apr 19 13:35:11 zealot kernel: drmn0: info:   GRBM_STATUS 
= 0x00003828
Apr 19 13:35:11 zealot kernel: drmn0: info:   GRBM_STATUS_SE0 
= 0x00000007
Apr 19 13:35:11 zealot kernel: drmn0: info:   GRBM_STATUS_SE1 
= 0x00000007
Apr 19 13:35:11 zealot kernel: drmn0: info:   SRBM_STATUS 
= 0x200000C0
Apr 19 13:35:11 zealot kernel: drmn0: info:   R_008674_CP_STALLED_STAT1 
= 0x00000000
Apr 19 13:35:11 zealot kernel: drmn0: info:   R_008678_CP_STALLED_STAT2 
= 0x00000000
Apr 19 13:35:11 zealot kernel: drmn0: info:   R_00867C_CP_BUSY_STAT 
= 0x00000000
Apr 19 13:35:11 zealot kernel: drmn0: info:   R_008680_CP_STAT 
= 0x00000000
Apr 19 13:35:11 zealot kernel: drmn0: info: GPU reset succeeded, trying 
to resume
Apr 19 13:35:11 zealot kernel: info: [drm] probing gen 2 caps for device 
1002:5978 = 2/0
Apr 19 13:35:11 zealot kernel: info: [drm] enabling PCIE gen 2 link 
speeds, disable with radeon.pcie_gen2=0
Apr 19 13:35:11 zealot kernel: info: [drm] PCIE GART of 512M enabled 
(table at 0x0000000000040000).
Apr 19 13:35:11 zealot kernel: drmn0: info: WB enabled
Apr 19 13:35:11 zealot kernel: drmn0: info: fence driver on ring 0 use 
gpu addr 0x0000000020000c00 and cpu addr 0x0xfffff8000bdcbc00
Apr 19 13:35:11 zealot kernel: drmn0: info: fence driver on ring 3 use 
gpu addr 0x0000000020000c0c and cpu addr 0x0xfffff8000bdcbc0c
Apr 19 13:35:11 zealot kernel: info: [drm] ring test on 0 succeeded in 1 
usecs
Apr 19 13:35:11 zealot kernel: info: [drm] ring test on 3 succeeded in 1 
usecs
Apr 19 13:35:21 zealot kernel: drmn0: error: GPU lockup CP stall for 
more than 10000msec
Apr 19 13:35:21 zealot kernel: drmn0: warning: GPU lockup (waiting for 
0x00000000017dbd49 last fence id 0x00000000017dbd42)
Apr 19 13:35:21 zealot kernel: error: [drm:pid2899:r600_ib_test] *ERROR* 
radeon: fence wait failed (-11).
Apr 19 13:35:21 zealot kernel: error: [drm:pid2899:radeon_ib_ring_tests] 
*ERROR* radeon: failed testing IB on GFX ring (-11).
Apr 19 13:35:21 zealot kernel: drmn0: error: ib ring test failed (-11).
Apr 19 13:35:21 zealot kernel: drmn0: info: GPU softreset: 0x00000003
Apr 19 13:35:21 zealot kernel: drmn0: info:   GRBM_STATUS 
= 0xF5703828
Apr 19 13:35:21 zealot kernel: drmn0: info:   GRBM_STATUS_SE0 
= 0xFC000007
Apr 19 13:35:21 zealot kernel: drmn0: info:   GRBM_STATUS_SE1 
= 0x00000007
Apr 19 13:35:21 zealot kernel: drmn0: info:   SRBM_STATUS 
= 0x200000C0
Apr 19 13:35:21 zealot kernel: drmn0: info:   R_008674_CP_STALLED_STAT1 
= 0x00000000
Apr 19 13:35:21 zealot kernel: drmn0: info:   R_008678_CP_STALLED_STAT2 
= 0x40000000
Apr 19 13:35:21 zealot kernel: drmn0: info:   R_00867C_CP_BUSY_STAT 
= 0x00008006
Apr 19 13:35:21 zealot kernel: drmn0: info:   R_008680_CP_STAT 
= 0x80228647
Apr 19 13:35:21 zealot kernel: drmn0: info:   GRBM_SOFT_RESET=0x00007F6B
Apr 19 13:35:21 zealot kernel: drmn0: info:   GRBM_STATUS 
= 0x00003828
Apr 19 13:35:21 zealot kernel: drmn0: info:   GRBM_STATUS_SE0 
= 0x00000007
Apr 19 13:35:21 zealot kernel: drmn0: info:   GRBM_STATUS_SE1 
= 0x00000007
Apr 19 13:35:21 zealot kernel: drmn0: info:   SRBM_STATUS 
= 0x200000C0
Apr 19 13:35:21 zealot kernel: drmn0: info:   R_008674_CP_STALLED_STAT1 
= 0x00000000
Apr 19 13:35:21 zealot kernel: drmn0: info:   R_008678_CP_STALLED_STAT2 
= 0x00000000
Apr 19 13:35:21 zealot kernel: drmn0: info:   R_00867C_CP_BUSY_STAT 
= 0x00000000
Apr 19 13:35:21 zealot kernel: drmn0: info:   R_008680_CP_STAT 
= 0x00000000
Apr 19 13:35:21 zealot kernel: drmn0: info: GPU reset succeeded, trying 
to resume

and such lockups appear somewhat regularly, last time image was restored 
and my desktop worked fine for two or three hours before deadlock 
occured and I had to reset it.

I wonder if 4550 will work more stable? I have one somewhere

-- 
SY, Marat

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4228 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.freebsd.org/pipermail/freebsd-x11/attachments/20140419/e498024c/attachment.bin>