Semi-frequent fglrx ASIC hangs in 11.11 & 11.12

[chort]

I've only been getting these since updating to Cat 11.11 on Ubuntu 11.04/64. I upgraded to 11.12, but it's still happening

It seems to be triggered when ocl-hashcat puts a very high load on the GPU(s) (i.e. multiple ocl-hashcat* processes with -n 160 or higher).


Dec 17 21:22:41 hydra kernel: [ 2725.374197] [fglrx] ASIC hang happened
Dec 17 21:22:41 hydra kernel: [ 2725.374203] Pid: 3746, comm: oclhl64 Tainted: P 2.6.38-13-generic #52-Ubuntu
Dec 17 21:22:41 hydra kernel: [ 2725.374205] Call Trace:
Dec 17 21:22:41 hydra kernel: [ 2725.374263] [<ffffffffa00b5e7e>] ? KCL_DEBUG_OsDump+0xe/0x10 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374289] [<ffffffffa00c340c>] ? firegl_hardwareHangRecovery+0x1c/0x50 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374332] [<ffffffffa0158c99>] ? _ZN4Asic9WaitUntil15ResetASICIfHungEv+0x9/0x10 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374372] [<ffffffffa0158c3c>] ? _ZN4Asic9WaitUntil15WaitForCompleteEv+0x9c/0xf0 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374412] [<ffffffffa0153618>] ? _ZN4Asic19PM4ElapsedTimeStampEj14_LARGE_INTEGER12_QS_CP_RING_+0x148/0x170 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374441] [<ffffffffa00e0022>] ? firegl_trace+0x72/0x1e0 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374468] [<ffffffffa00e0022>] ? firegl_trace+0x72/0x1e0 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374496] [<ffffffffa00e0022>] ? firegl_trace+0x72/0x1e0 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374535] [<ffffffffa014c963>] ? _ZN15QS_PRIVATE_CORE27multiVpuPM4ElapsedTimeStampEj14_LARGE_INTEGER12_QS_CP_RING_+0x33/0x50 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374575] [<ffffffffa0145204>] ? _Z19uQSTimeStampRetiredmjj14_LARGE_INTEGER+0x74/0x80 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374613] [<ffffffffa014108d>] ? _Z8uCWDDEQCmjjPvjS_+0x54d/0x10c0 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374619] [<ffffffff8108d81e>] ? down+0x2e/0x50
Dec 17 21:22:41 hydra kernel: [ 2725.374646] [<ffffffffa00e25f2>] ? firegl_cmmqs_CWDDE_32+0x332/0x440 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374674] [<ffffffffa00e0f20>] ? firegl_cmmqs_CWDDE32+0x70/0x100 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374702] [<ffffffffa00e0eb0>] ? firegl_cmmqs_CWDDE32+0x0/0x100 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374727] [<ffffffffa00befad>] ? firegl_ioctl+0x1ed/0x250 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374750] [<ffffffffa00afd6e>] ? ip_firegl_unlocked_ioctl+0xe/0x20 [fglrx]
Dec 17 21:22:41 hydra kernel: [ 2725.374755] [<ffffffff811764ef>] ? do_vfs_ioctl+0x8f/0x360
Dec 17 21:22:41 hydra kernel: [ 2725.374759] [<ffffffff8109d08b>] ? sys_futex+0x7b/0x180
Dec 17 21:22:41 hydra kernel: [ 2725.374762] [<ffffffff81176851>] ? sys_ioctl+0x91/0xa0
Dec 17 21:22:41 hydra kernel: [ 2725.374766] [<ffffffff8100c002>] ? system_call_fastpath+0x16/0x1b
Dec 17 21:22:41 hydra kernel: [ 2725.374770] pubdev:0xffffffffa033e740, num of device:2 , name:fglrx, major 8, minor 92.
Dec 17 21:22:41 hydra kernel: [ 2725.374772] device 0 : 0xffff8803fdddc000 .
Dec 17 21:22:41 hydra kernel: [ 2725.374774] Asic ID:0x9640, revision:0x2, MMIOReg:0xffffc90014580000.
Dec 17 21:22:41 hydra kernel: [ 2725.374776] FB phys addr: 0xc0000000, MC :0xf00000000, Total FB size :0x20000000.
Dec 17 21:22:41 hydra kernel: [ 2725.374778] gart table MC:0xf0fbfd000, Physical:0xcfbfd000, size:0x402000.
Dec 17 21:22:41 hydra kernel: [ 2725.374780] mc_node :FB, total 1 zones
Dec 17 21:22:41 hydra kernel: [ 2725.374781] MC start:0xf00000000, Physical:0xc0000000, size:0x10000000.
Dec 17 21:22:41 hydra kernel: [ 2725.374784] Mapped heap -- Offset:0x0, size:0xfbfd000, reference count:44, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374786] Mapped heap -- Offset:0x0, size:0x1000000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374788] Mapped heap -- Offset:0xfbfd000, size:0x403000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374790] mc_node :INV_FB, total 1 zones
Dec 17 21:22:41 hydra kernel: [ 2725.374791] MC start:0xf10000000, Physical:0xd0000000, size:0x10000000.
Dec 17 21:22:41 hydra kernel: [ 2725.374793] Mapped heap -- Offset:0xfff4000, size:0xc000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374795] mc_node :GART_USWC, total 2 zones
Dec 17 21:22:41 hydra kernel: [ 2725.374796] MC start:0x40100000, Physical:0x0, size:0x50000000.
Dec 17 21:22:41 hydra kernel: [ 2725.374798] Mapped heap -- Offset:0x0, size:0x2000000, reference count:17, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374800] mc_node :GART_CACHEABLE, total 3 zones
Dec 17 21:22:41 hydra kernel: [ 2725.374801] MC start:0x10400000, Physical:0x0, size:0x2fd00000.
Dec 17 21:22:41 hydra kernel: [ 2725.374803] Mapped heap -- Offset:0x2e00000, size:0x200000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374805] Mapped heap -- Offset:0x3d00000, size:0x500000, reference count:2, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374807] Mapped heap -- Offset:0x3a00000, size:0x300000, reference count:2, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374809] Mapped heap -- Offset:0x3300000, size:0x300000, reference count:2, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374812] Mapped heap -- Offset:0x3000000, size:0x300000, reference count:2, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374814] Mapped heap -- Offset:0x2c00000, size:0x200000, reference count:2, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374816] Mapped heap -- Offset:0x2900000, size:0x300000, reference count:4, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374818] Mapped heap -- Offset:0x2600000, size:0x300000, reference count:3, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374820] Mapped heap -- Offset:0x2300000, size:0x300000, reference count:4, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374822] Mapped heap -- Offset:0x2000000, size:0x300000, reference count:4, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374824] Mapped heap -- Offset:0x1800000, size:0x200000, reference count:2, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374826] Mapped heap -- Offset:0x1600000, size:0x200000, reference count:4, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374828] Mapped heap -- Offset:0x700000, size:0x500000, reference count:4, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374830] Mapped heap -- Offset:0x1b00000, size:0x500000, reference count:3, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374832] Mapped heap -- Offset:0x1100000, size:0x500000, reference count:11, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374834] Mapped heap -- Offset:0xc00000, size:0x500000, reference count:5, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374837] Mapped heap -- Offset:0x200000, size:0x500000, reference count:4, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374839] Mapped heap -- Offset:0x0, size:0x200000, reference count:10, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374841] Mapped heap -- Offset:0xef000, size:0x11000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374846] GRBM : 0x3828, SRBM : 0x20000040 .
Dec 17 21:22:41 hydra kernel: [ 2725.374850] CP_RB_BASE : 0x401000, CP_RB_RPTR : 0x1da90 , CP_RB_WPTR :0x1da90.
Dec 17 21:22:41 hydra kernel: [ 2725.374855] CP_IB1_BUFSZ:0x0, CP_IB1_BASE_HI:0x0, CP_IB1_BASE_LO:0x4051d000.
Dec 17 21:22:41 hydra kernel: [ 2725.374856] last submit IB buffer -- MC :0x4051d000,phys:0x3f4d46000.
Dec 17 21:22:41 hydra kernel: [ 2725.374859] device 1 : 0xffff8803fb050000 .
Dec 17 21:22:41 hydra kernel: [ 2725.374861] Asic ID:0x6898, revision:0x2, MMIOReg:0xffffc900144c0000.
Dec 17 21:22:41 hydra kernel: [ 2725.374863] FB phys addr: 0xd0000000, MC :0xf00000000, Total FB size :0x40000000.
Dec 17 21:22:41 hydra kernel: [ 2725.374865] gart table MC:0xf0f8fd000, Physical:0xdf8fd000, size:0x402000.
Dec 17 21:22:41 hydra kernel: [ 2725.374866] mc_node :FB, total 1 zones
Dec 17 21:22:41 hydra kernel: [ 2725.374868] MC start:0xf00000000, Physical:0xd0000000, size:0xfd00000.
Dec 17 21:22:41 hydra kernel: [ 2725.374870] Mapped heap -- Offset:0x0, size:0xf8fd000, reference count:24, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374872] Mapped heap -- Offset:0x0, size:0x1000000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374874] Mapped heap -- Offset:0xf8fd000, size:0x403000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374876] mc_node :INV_FB, total 1 zones
Dec 17 21:22:41 hydra kernel: [ 2725.374877] MC start:0xf0fd00000, Physical:0xdfd00000, size:0x30300000.
Dec 17 21:22:41 hydra kernel: [ 2725.374879] Mapped heap -- Offset:0x302f4000, size:0xc000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374881] mc_node :GART_USWC, total 2 zones
Dec 17 21:22:41 hydra kernel: [ 2725.374882] MC start:0x40100000, Physical:0x0, size:0x50000000.
Dec 17 21:22:41 hydra kernel: [ 2725.374884] Mapped heap -- Offset:0x0, size:0x2000000, reference count:13, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374886] mc_node :GART_CACHEABLE, total 3 zones
Dec 17 21:22:41 hydra kernel: [ 2725.374887] MC start:0x10400000, Physical:0x0, size:0x2fd00000.
Dec 17 21:22:41 hydra kernel: [ 2725.374889] Mapped heap -- Offset:0x1500000, size:0x500000, reference count:2, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374891] Mapped heap -- Offset:0x700000, size:0x500000, reference count:3, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374893] Mapped heap -- Offset:0xc00000, size:0x500000, reference count:2, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374895] Mapped heap -- Offset:0x200000, size:0x500000, reference count:3, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374897] Mapped heap -- Offset:0x0, size:0x200000, reference count:9, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374899] Mapped heap -- Offset:0xef000, size:0x11000, reference count:1, mapping count:0,
Dec 17 21:22:41 hydra kernel: [ 2725.374904] GRBM : 0x3828, SRBM : 0x200000c0 .
Dec 17 21:22:41 hydra kernel: [ 2725.374907] CP_RB_BASE : 0x401000, CP_RB_RPTR : 0x1ec70 , CP_RB_WPTR :0x1ec70.
Dec 17 21:22:41 hydra kernel: [ 2725.374911] CP_IB1_BUFSZ:0x0, CP_IB1_BASE_HI:0x0, CP_IB1_BASE_LO:0x40229000.
Dec 17 21:22:41 hydra kernel: [ 2725.374913] last submit IB buffer -- MC :0x40229000,phys:0x3f1ed1000.
Dec 17 21:22:41 hydra kernel: [ 2725.374915] Dump the trace queue.
Dec 17 21:22:41 hydra kernel: [ 2725.374916] End of dump

[atom]

never saw that error before. someone else have this error?

[chort]

"Asic ID:0x9640" seems to be device codename "SUMO," which would be the HD 6550D that's built-in to my A8-3850 "APU" (CPU + GPU). I can't tell from the trace whether Asic ID:0x9640 was the one that froze, or if it was Asic ID:0x6898 (HD 5870).

I found a bug on the unofficial ATI site about it, same symptoms:
http://ati.cchtml.com/show_bug.cgi?id=230

Interestingly the first report there is for Cat 11.4, but I never had this problem with 11.4. It only started when I updated to 11.11, still happens with 11.12.