https://bugzilla.novell.com/show_bug.cgi?id=805300 https://bugzilla.novell.com/show_bug.cgi?id=805300#c6 --- Comment #6 from Marek Paśnikowski <novell.com@marekpasnikowski.name> 2013-07-16 21:13:28 UTC --- Recently I installed an AMD RV670 card. I have similar behavior with it. On clean OpenSUSE 13.1 M3 I had weird hangups where I could see the cursor move, but nothing else I did to system worked. Can't really define what my problem was because I did not connect that with this bug and did not observe properly. Now I have a combination of OpenSUSE 12.3 with repos providing current devel versions of kernel and Mesa (3.11 and 9.2, respectively) + radeon.dpm=1 (the new proper power management fresh from AMD). What I experience now is (from dmesg) this: radeon 0000:02:00.0: GPU lockup CP stall for more than 10000msec [15360.084472] radeon 0000:02:00.0: GPU lockup (waiting for 0x0000000000186a34 last fence id 0x0000000000186a31) [15360.084475] [drm:r600_ib_test] *ERROR* radeon: fence wait failed (-35). [15360.084478] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB on GFX ring (-35). [15360.084480] radeon 0000:02:00.0: ib ring test failed (-35). [15360.085498] radeon 0000:02:00.0: GPU softreset: 0x00000019 [15360.085500] radeon 0000:02:00.0: R_008010_GRBM_STATUS = 0xA23334AC [15360.085501] radeon 0000:02:00.0: R_008014_GRBM_STATUS2 = 0x00000003 [15360.085503] radeon 0000:02:00.0: R_000E50_SRBM_STATUS = 0x200000C0 [15360.085505] radeon 0000:02:00.0: R_008674_CP_STALLED_STAT1 = 0x04000000 [15360.085506] radeon 0000:02:00.0: R_008678_CP_STALLED_STAT2 = 0x00000002 [15360.085508] radeon 0000:02:00.0: R_00867C_CP_BUSY_STAT = 0x00008484 [15360.085509] radeon 0000:02:00.0: R_008680_CP_STAT = 0x80818645 [15360.085511] radeon 0000:02:00.0: R_00D034_DMA_STATUS_REG = 0x44C83D57 [15360.135245] radeon 0000:02:00.0: R_008020_GRBM_SOFT_RESET=0x00007FEF [15360.135318] radeon 0000:02:00.0: SRBM_SOFT_RESET=0x00000100 [15360.137408] radeon 0000:02:00.0: R_008010_GRBM_STATUS = 0xA0003030 [15360.137423] radeon 0000:02:00.0: R_008014_GRBM_STATUS2 = 0x00000003 [15360.137427] radeon 0000:02:00.0: R_000E50_SRBM_STATUS = 0x200080C0 [15360.137430] radeon 0000:02:00.0: R_008674_CP_STALLED_STAT1 = 0x00000000 [15360.137433] radeon 0000:02:00.0: R_008678_CP_STALLED_STAT2 = 0x00000000 [15360.137436] radeon 0000:02:00.0: R_00867C_CP_BUSY_STAT = 0x00000000 [15360.137442] radeon 0000:02:00.0: R_008680_CP_STAT = 0x80120000 [15360.137447] radeon 0000:02:00.0: R_00D034_DMA_STATUS_REG = 0x44C83D57 [15360.137453] radeon 0000:02:00.0: GPU reset succeeded, trying to resume [15360.138845] [drm] PCIE gen 2 link speeds already enabled [15360.140518] [drm] PCIE GART of 512M enabled (table at 0x0000000000040000). [15360.140579] radeon 0000:02:00.0: WB enabled [15360.140585] radeon 0000:02:00.0: fence driver on ring 0 use gpu addr 0x0000000020000c00 and cpu addr 0xffff8801101a5c00 [15360.140592] radeon 0000:02:00.0: fence driver on ring 3 use gpu addr 0x0000000020000c0c and cpu addr 0xffff8801101a5c0c [15360.171528] [drm] ring test on 0 succeeded in 1 usecs [15360.171599] [drm] ring test on 3 succeeded in 1 usecs [15360.171617] [drm] ib test on ring 0 succeeded in 0 usecs [15360.171634] [drm] ib test on ring 3 succeeded in 1 usecs So this bug really contains two separate problems: 1. GPU lockup. 2. Failure to reset GPU completely and properly. The HD3870 I currently use recovers properly, without lingering after-effects. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.