Here is a dmesg output in a case where the GPU hung twice in a row and the GPU resume failed the second time:
[72317.916119] radeon 0000:01:00.0: GPU lockup CP stall for more than 10000msec [72317.916132] radeon 0000:01:00.0: GPU lockup (waiting for 0x00000000001d62ad last fence id 0x00000000001d62a9) [72317.917267] radeon 0000:01:00.0: GPU softreset [72317.917275] radeon 0000:01:00.0: R_008010_GRBM_STATUS=0xA0003030 [72317.917282] radeon 0000:01:00.0: R_008014_GRBM_STATUS2=0x00000003 [72317.917289] radeon 0000:01:00.0: R_000E50_SRBM_STATUS=0x200000C0 [72317.917303] radeon 0000:01:00.0: R_008020_GRBM_SOFT_RESET=0x00007FEE [72317.932197] radeon 0000:01:00.0: R_008020_GRBM_SOFT_RESET=0x00000001 [72317.948082] radeon 0000:01:00.0: R_008010_GRBM_STATUS=0xA0003030 [72317.948090] radeon 0000:01:00.0: R_008014_GRBM_STATUS2=0x00000003 [72317.948097] radeon 0000:01:00.0: R_000E50_SRBM_STATUS=0x200080C0 [72317.949094] radeon 0000:01:00.0: GPU reset succeed [72317.953205] [drm] PCIE GART of 512M enabled (table at 0x0000000000040000). [72317.953238] radeon 0000:01:00.0: WB enabled [72317.953240] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x0000000010000c00 and cpu addr 0xffc8ec00 [72317.984407] [drm] ring test on 0 succeeded in 1 usecs [72317.984644] [drm] ib test on ring 0 succeeded in 0 usecs [72319.220350] type=1701 audit(1349193168.886:83): auid=4294967295 uid=1000 gid=1000 ses=4294967295 pid=26825 comm="chrome" reason="seccomp" sig=0 syscall=20 compat=0 ip=0xb3510424 code=0x50000 [72329.872338] radeon 0000:01:00.0: GPU lockup CP stall for more than 10000msec [72329.872350] radeon 0000:01:00.0: GPU lockup (waiting for 0x00000000001d63b5 last fence id 0x00000000001d63b3) [72329.873488] radeon 0000:01:00.0: GPU softreset [72329.873496] radeon 0000:01:00.0: R_008010_GRBM_STATUS=0xE57C2030 [72329.873503] radeon 0000:01:00.0: R_008014_GRBM_STATUS2=0x00113303 [72329.873509] radeon 0000:01:00.0: R_000E50_SRBM_STATUS=0x200230C0 [72330.036637] radeon 0000:01:00.0: Wait for MC idle timedout ! [72330.036648] radeon 0000:01:00.0: R_008020_GRBM_SOFT_RESET=0x00007FEE [72330.051531] radeon 0000:01:00.0: R_008020_GRBM_SOFT_RESET=0x00000001 [72330.067415] radeon 0000:01:00.0: R_008010_GRBM_STATUS=0xA0003030 [72330.067421] radeon 0000:01:00.0: R_008014_GRBM_STATUS2=0x00000003 [72330.067428] radeon 0000:01:00.0: R_000E50_SRBM_STATUS=0x2002B0C0 [72330.068435] radeon 0000:01:00.0: GPU reset succeed [72330.232982] radeon 0000:01:00.0: Wait for MC idle timedout ! [72330.396837] radeon 0000:01:00.0: Wait for MC idle timedout ! [72330.399208] [drm] PCIE GART of 512M enabled (table at 0x0000000000040000). [72330.399300] radeon 0000:01:00.0: WB enabled [72330.399304] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x0000000010000c00 and cpu addr 0xffc8ec00 [72330.430520] [drm] ring test on 0 succeeded in 1 usecs [72340.928074] radeon 0000:01:00.0: GPU lockup CP stall for more than 10000msec [72340.928087] radeon 0000:01:00.0: GPU lockup (waiting for 0x00000000001d63b7 last fence id 0x00000000001d63b6) [72340.928095] [drm:r600_ib_test] *ERROR* radeon: fence wait failed (-35). [72340.928101] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB on GFX ring (-35). [72340.928106] [drm:r600_resume] *ERROR* r600 startup failed on resume -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/986524 Title: radeon 0000:02:00.0: GPU lockup CP stall for more than 10000msec To manage notifications about this bug go to: https://bugs.launchpad.net/linux/+bug/986524/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
