Created attachment 119351 [details] complete dmesg Here is an extract of the dmeg log. We are several to be bitten by the same bug as described here: http://forum.xbmc.org/showthread.php?tid=174854&pid=1577781#pid1577781 [drm] Found smc ucode version: 0x00010200 [ 40.117359] BUG: unable to handle kernel paging request at ffffec2000000900 [ 40.117377] IP: [<ffffffff810a796f>] kfree+0x41/0xe2 [ 40.117396] PGD 0 [ 40.117404] Oops: 0000 [#1] PREEMPT SMP [ 40.117413] Modules linked in: radeon cfbfillrect cfbimgblt cfbcopyarea i2c_algo_bit backlight drm_kms_helper ttm drm fb fbdev [ 40.117445] CPU: 1 PID: 3029 Comm: Xorg Not tainted 3.13.0-rc5 #6 [ 40.117454] Hardware name: To be filled by O.E.M. To be filled by O.E.M./Inagua CRB, BIOS 4.6.4 06/15/2011 [ 40.117463] task: ffff88013ba5f2c0 ti: ffff88009d1ce000 task.ti: ffff88009d1ce000 [ 40.117470] RIP: 0010:[<ffffffff810a796f>] [<ffffffff810a796f>] kfree+0x41/0xe2 [ 40.117484] RSP: 0018:ffff88009d1cfa40 EFLAGS: 00010286 [ 40.117492] RAX: ffffea0000000000 RBX: 0000100000024414 RCX: 0000000000000000 [ 40.117499] RDX: 0000000000010005 RSI: 0000000000005f78 RDI: 0000100000024414 [ 40.117505] RBP: 0000000000000c00 R08: ffffec2000000900 R09: 0000000000000c00 [ 40.117512] R10: ffffffffa00a7da5 R11: ffffffffa00a7f08 R12: ffff8800a2c01200 [ 40.117518] R13: ffff88013b298000 R14: 0000000000000000 R15: ffff88013aaa6b00 [ 40.117526] FS: 00007f94df8fe980(0000) GS:ffff88013fd00000(0000) knlGS:0000000000000000 [ 40.117533] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 40.117539] CR2: ffffec2000000900 CR3: 00000000a36e3000 CR4: 00000000000007e0 [ 40.117544] Stack: [ 40.117549] ffff88013b298000 ffffffffa00f11e2 00007c44a00f62be 00007d2c00007c48 [ 40.117559] ffff8800a2c01200 0000000000000003 0000000000000000 ffffffffa01750b1 [ 40.117569] 0000100000024414 0000188000024414 0000180000028488 ffffffff00024414 [ 40.117579] Call Trace: [ 40.117686] [<ffffffffa00f11e2>] ? evergreen_hdmi_setmode+0x3cc/0x7e1 [radeon] [ 40.117705] [<ffffffffa0071732>] ? drm_crtc_helper_set_mode+0x30a/0x40a [drm_kms_helper] [ 40.117726] [<ffffffffa007248c>] ? drm_crtc_helper_set_config+0x5ff/0x842 [drm_kms_helper] [ 40.117798] [<ffffffffa00b4392>] ? radeon_crtc_set_config+0x41/0xe8 [radeon] [ 40.117843] [<ffffffffa001d3e7>] ? drm_mode_set_config_internal+0x44/0xac [drm] [ 40.117885] [<ffffffffa001f8cb>] ? drm_mode_setcrtc+0x3e8/0x499 [drm] [ 40.117917] [<ffffffffa00148b3>] ? drm_ioctl+0x26f/0x394 [drm] [ 40.117957] [<ffffffffa001f4e3>] ? drm_mode_setplane+0x335/0x335 [drm] [ 40.118015] [<ffffffffa0096045>] ? radeon_drm_ioctl+0x45/0x77 [radeon] [ 40.118028] [<ffffffff810b962a>] ? do_vfs_ioctl+0x356/0x421 [ 40.118039] [<ffffffff810b9743>] ? SyS_ioctl+0x4e/0x7b [ 40.118054] [<ffffffff8147a322>] ? system_call_fastpath+0x16/0x1b [ 40.118059] Code: 00 00 00 80 ff 77 00 00 49 01 f8 48 0f 42 05 b9 86 96 00 49 01 c0 48 b8 00 00 00 00 00 ea ff ff 49 c1 e8 0c 49 c1 e0 06 49 01 c0 <49> 8b 00 f6 c4 80 74 04 4d 8b 40 30 49 8b 00 a8 80 75 22 49 f7 [ 40.118149] RIP [<ffffffff810a796f>] kfree+0x41/0xe2 [ 40.118160] RSP <ffff88009d1cfa40> [ 40.118165] CR2: ffffec2000000900 [ 40.118173] ---[ end trace e7d41f7cbae99794 ]---
Created attachment 119361 [details] dmidecode output
The board was given to me by AMD when trying to promote their APU some time ago. I used it using proprietary fglrx driver since, but as I have a zotac AD02 board with same APU that works well, I wanted to try it on this board also. So same software setup works with a similar processor but on a different hardware. I use Linux 3.13-rc5 debian 64 bits (unstable + experimental). Setup is described here <http://forum.xbmc.org/showthread.php?tid=174854&pid=1575163#pid1575163>. I upgraded to mesa 10.0.1.
Duplicate of bug 72283. This should be fixed in this commit: http://cgit.freedesktop.org/~airlied/linux/commit/?h=drm-fixes&id=b67ce39a30976171e7b96b30a94a0216ab89df97 which should show up in Linus tree soon.
Thanks. Indeed it fixes the bug that carsh the kernel buy I still have this: [drm] ib test on ring 0 succeeded in 0 usecs [ 27.400271] [drm] ib test on ring 3 succeeded in 1 usecs [ 37.421526] radeon 0000:00:01.0: GPU lockup CP stall for more than 10001msec [ 37.421559] radeon 0000:00:01.0: GPU lockup (waiting for 0x0000000000000004 last fence id 0x0000000000000002 on ring 5) [ 37.421570] [drm:uvd_v1_0_ib_test] *ERROR* radeon: fence wait failed (-35). [ 37.441641] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB on ring 5 (-35). [ 37.441664] [drm] Found smc ucode version: 0x00010200 [ 38.411143] [drm:dce4_afmt_write_speaker_allocation] *ERROR* Couldn't read Speaker Allocation Data Block: 0 [ 42.249879] [drm] Disabling audio 0 support Does this mean I will get no audio via HDMI?
(In reply to Eric Valette from comment #4) > Thanks. Indeed it fixes the bug that carsh the kernel buy I still have this: > > [drm] ib test on ring 0 succeeded in 0 usecs > [ 27.400271] [drm] ib test on ring 3 succeeded in 1 usecs > [ 37.421526] radeon 0000:00:01.0: GPU lockup CP stall for more than > 10001msec > [ 37.421559] radeon 0000:00:01.0: GPU lockup (waiting for > 0x0000000000000004 last fence id 0x0000000000000002 on ring 5) > [ 37.421570] [drm:uvd_v1_0_ib_test] *ERROR* radeon: fence wait failed > (-35). > [ 37.441641] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB > on ring 5 (-35). Make sure you have the uvd and latest rlc ucode installed in order to use UVD. > [ 37.441664] [drm] Found smc ucode version: 0x00010200 > [ 38.411143] [drm:dce4_afmt_write_speaker_allocation] *ERROR* Couldn't > read Speaker Allocation Data Block: 0 This error is harmless, it's just telling you that the monitor provides invalid speak allocation information. Audio support should still work. > [ 42.249879] [drm] Disabling audio 0 support Can you attach your full dmesg log? You should only see this message during suspend, shutdown, or gpu reset. This is unrelated to patch. > > Does this mean I will get no audio via HDMI? You should still get audio unless the GPU is hung and the driver is not able to reset it.
(In reply to Alex Deucher from comment #5) > Make sure you have the uvd and latest rlc ucode installed in order to use > UVD. I should. > > [ 42.249879] [drm] Disabling audio 0 support > Can you attach your full dmesg log? You should only see this message during > suspend, shutdown, or gpu reset. This is unrelated to patch. > > > > > Does this mean I will get no audio via HDMI? > > You should still get audio unless the GPU is hung and the driver is not able > to reset it. I will reattach with patch applied.
Created attachment 119371 [details] dmesg with suggested patch aplied
root@stb-fusion-3:/lib/firmware/radeon# md5sum * 59375dccb37f974c045575cd9428009a ARUBA_me.bin b3072fac01a6eab4711c18148c8bc305 ARUBA_pfp.bin 246d1c75a5946829f6864dbd5f71d850 ARUBA_rlc.bin 158f8e21ccf228ef063888c4f637fbf0 BARTS_mc.bin 8012e24b187c6b1ba17fa48691c3b048 BARTS_me.bin b08d560e8f57d700fd67957584e0567c BARTS_pfp.bin 24a4c72d0bc120ffd2283e428faf432b BARTS_smc.bin 44ec9d529b6fb44d4dd0a219e3218a1e BONAIRE_ce.bin ef4e1c28226020f29718c1b4a71e4936 BONAIRE_mc.bin 16a295b3cfe280ea070727713049a2d9 BONAIRE_me.bin e2a1fb791002c7ce24f770d234700104 BONAIRE_mec.bin 48db59feaf30154dc5183301781ee7c5 BONAIRE_pfp.bin 85eabd2f0f48679eeade573c471814ad BONAIRE_rlc.bin 0f6501d69df393af36f8f3bcb59d3835 BONAIRE_sdma.bin 303438f5daec8e0661a3d1272606c558 BONAIRE_uvd.bin 25d61fad839b30b263f52328c1f678fb BTC_rlc.bin 158f8e21ccf228ef063888c4f637fbf0 CAICOS_mc.bin 8012e24b187c6b1ba17fa48691c3b048 CAICOS_me.bin 87b95689bb03323faf917bda6aa1cd11 CAICOS_pfp.bin 03d4c15eeda157c96819088253acb46a CAICOS_smc.bin b8f97a70b25104e3ca24b8b8ade19997 CAYMAN_mc.bin 5b4feb3f418fa1725ae7ea2633071118 CAYMAN_me.bin 53671bbdd823e4b14dbaab63bd5f248f CAYMAN_pfp.bin 0c8ca68a18efff6e890cd5ea176c052a CAYMAN_rlc.bin 1884c8c5e6e6af4f088c38ae25721f42 CAYMAN_smc.bin 2b244d41832f46382bfbb8994522dcdd CEDAR_me.bin 23915e382ea0d2f2491a19146ca3001c CEDAR_pfp.bin e8770d3d588f24dc6f1a8609c9db3467 CEDAR_rlc.bin e8618d8a65add54200e73f5580fc48d0 CEDAR_smc.bin fa937b6596298b4bbc9edb6df4adca2a CYPRESS_me.bin 2dca2882a14e1d6a43792f786471ec51 CYPRESS_pfp.bin e8770d3d588f24dc6f1a8609c9db3467 CYPRESS_rlc.bin aeb83918c9fb268b0a4cbb03f2dfab3f CYPRESS_smc.bin fb23b281dcc94a035d374e709c9842bd CYPRESS_uvd.bin a5f07f65a9ef260c0077021ecae43dc7 HAINAN_ce.bin 3bbdb66a8d049cf2b7f85ebfe4d8df94 HAINAN_mc.bin 9545cef078ac83b037e1727c06ee6af2 HAINAN_me.bin ba3d0e27b8cbcdb24181040595255d3e HAINAN_pfp.bin 3519612cd874d840a510d575559d6b9b HAINAN_rlc.bin 9a39456f0001671d1d6d9dc30a581fe0 HAINAN_smc.bin fa937b6596298b4bbc9edb6df4adca2a JUNIPER_me.bin 2dca2882a14e1d6a43792f786471ec51 JUNIPER_pfp.bin e8770d3d588f24dc6f1a8609c9db3467 JUNIPER_rlc.bin 2dbce2e58ef5b9c79a1fd2e671d78f35 JUNIPER_smc.bin 44ec9d529b6fb44d4dd0a219e3218a1e KABINI_ce.bin b1469ac001eaf8d5a04d91395c5257f8 KABINI_me.bin c6f8cda051fea873ce8e306afb9f20c5 KABINI_mec.bin 92bbe966f67d6998cc96f150e3db2df5 KABINI_pfp.bin 24c0f737db80a07d784a226036aac9da KABINI_rlc.bin 0f6501d69df393af36f8f3bcb59d3835 KABINI_sdma.bin a5f07f65a9ef260c0077021ecae43dc7 OLAND_ce.bin 3bbdb66a8d049cf2b7f85ebfe4d8df94 OLAND_mc.bin 9545cef078ac83b037e1727c06ee6af2 OLAND_me.bin 417f193fd055a6842d5a4cad2ef624e1 OLAND_pfp.bin 466d29f573fefcb60bae26b8c867d6e5 OLAND_rlc.bin 42069d2e8978b87a0b9319a2caa32d41 OLAND_smc.bin 7d9ff6962e7bcc10b6eecd811d029dc8 PALM_me.bin 3f9d2af72e73d44aec16a496e7fc7fef PALM_pfp.bin a5f07f65a9ef260c0077021ecae43dc7 PITCAIRN_ce.bin 96b18c6f7c74ad4cecb04fca967ca433 PITCAIRN_mc.bin 5e899b3ff3e128453784b8fdacb947bb PITCAIRN_me.bin 6a1f860df54aa4d462339322ba363092 PITCAIRN_pfp.bin 3d2c150b3626419131bbc9a5864c7f1d PITCAIRN_rlc.bin b4b17dd30f14ceab88446c20796767d5 PITCAIRN_smc.bin f4f27d17dc204e11632cf98a8294650d R100_cp.bin 52a30faef239f286f497d95be7d2194c R200_cp.bin a05f4d9e10f0cdbfa3f96300048d300f R300_cp.bin c33ab57e1cc74b4c63bb09bc4d7bc1a2 R420_cp.bin 5a097d5e86c991f54806e88ad6882585 R520_cp.bin f2432caf487c4b586a2c391435f3749c R600_me.bin 448dbf1df580c31a0e55de22bb076be3 R600_pfp.bin f74a5163948bde215be6b689ca24afde R600_rlc.bin 5d186be14cc2cc328d02698ae4317a1b R700_rlc.bin 9334c37ae709f8faa6120c3ad7a5adb7 REDWOOD_me.bin 23915e382ea0d2f2491a19146ca3001c REDWOOD_pfp.bin e8770d3d588f24dc6f1a8609c9db3467 REDWOOD_rlc.bin 33480e5daef82d4039cabcc111917478 REDWOOD_smc.bin 801f81f19823e42e83f932d7ab73ab25 RS600_cp.bin 16ed338f3be50b60fd163ce91cd36915 RS690_cp.bin cafce27d4a77066d3980b1751cd7fd24 RS780_me.bin 0f7efc627708e22928dc1ef8da1646d8 RS780_pfp.bin 05ac3fb7f6dd64f3d99f59d2c6ab5d66 RV610_me.bin d9b7b8c30048a060b6d27fbf566108bd RV610_pfp.bin 05ac3fb7f6dd64f3d99f59d2c6ab5d66 RV620_me.bin d9b7b8c30048a060b6d27fbf566108bd RV620_pfp.bin 3f2a89200db525a69d79c84458111a7d RV630_me.bin d9b7b8c30048a060b6d27fbf566108bd RV630_pfp.bin 3f2a89200db525a69d79c84458111a7d RV635_me.bin d9b7b8c30048a060b6d27fbf566108bd RV635_pfp.bin 6491f874139f311383e7d2e9ac0411f2 RV670_me.bin d9b7b8c30048a060b6d27fbf566108bd RV670_pfp.bin a3e162705012e104727b2487bd690837 RV710_me.bin 89d2d6c1d169d0c2019f2388def7df56 RV710_pfp.bin 3e08d61531b186e66abbe8ca4b7aac90 RV710_smc.bin 7aa399a248c0d42fba9439ae0fbc5d90 RV710_uvd.bin 9fa1130a453e2a95a0a2de836cd96260 RV730_me.bin 89d2d6c1d169d0c2019f2388def7df56 RV730_pfp.bin 9fb755c1d51474635887122169ce77cc RV730_smc.bin 855a930fa529f9b945413165b825c617 RV740_smc.bin eaf386f2ae6d70779e9cb44da7bcad3f RV770_me.bin 0d6cf0e479dcaf69f48322a74ddf90ea RV770_pfp.bin 5e6e079252159d1960080e170eb96e4c RV770_smc.bin 5844be40ff36dcc30d161765e1a46e31 SUMO2_me.bin 3804aabfa24cc8a45b2a579b3398b96b SUMO2_pfp.bin 5844be40ff36dcc30d161765e1a46e31 SUMO_me.bin 1d569f6fe2e5bd262739789ebe089996 SUMO_pfp.bin 687e72d53413710b0a3e9330333b2dbe SUMO_rlc.bin 51d9e0e2247c313c5bfc8fa7bb5b213d SUMO_uvd.bin a5f07f65a9ef260c0077021ecae43dc7 TAHITI_ce.bin 96b18c6f7c74ad4cecb04fca967ca433 TAHITI_mc.bin 5e899b3ff3e128453784b8fdacb947bb TAHITI_me.bin 6a1f860df54aa4d462339322ba363092 TAHITI_pfp.bin 8e3f8b42b798737b6888e89050e37c0e TAHITI_rlc.bin 69d0115a4a07ba98b5ee56e41aac1c8f TAHITI_smc.bin 201877fa59f2fe4d896d5e6b6c1d2e1c TAHITI_uvd.bin 158f8e21ccf228ef063888c4f637fbf0 TURKS_mc.bin 8012e24b187c6b1ba17fa48691c3b048 TURKS_me.bin 25f26ba407a9bb13528b903c617209c8 TURKS_pfp.bin 4fe0f4dafe21f0efa6301a888eed4470 TURKS_smc.bin a5f07f65a9ef260c0077021ecae43dc7 VERDE_ce.bin 96b18c6f7c74ad4cecb04fca967ca433 VERDE_mc.bin a291d177203e882872ba809f82010077 VERDE_me.bin 8929a87c20f87426578518e3fafa12f2 VERDE_pfp.bin f8ee65f13adc45fe229a48128b7cd8f2 VERDE_rlc.bin 2443ed77790c7ba390db43903b8eebd5 VERDE_smc.bin root@stb-fusion-3:/lib/firmware/radeon#
I had a chance to look at the display on the tv set after my kids went to bed, here is whar happens: 1) I get on long black immediately after boot as if normal vga display does not work, 2) Once init is completed, I see the kdm loggin, 3) I see the begining of kde initialization and at one point the screen goes black, 4) After a GPU rest I see for one second the normal desktop, 5) screen goes black forever Looking at the dmesg, I have a pile of until it fails to reset I guess [ 52.398524] radeon 0000:00:01.0: GPU lockup CP stall for more than 10000msec [ 52.398545] radeon 0000:00:01.0: GPU lockup (waiting for 0x0000000000000006 last fence id 0x0000000000000002 on ring 5) [ 52.398556] [drm:uvd_v1_0_ib_test] *ERROR* radeon: fence wait failed (-35). [ 52.418619] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB on ring 5 (-35). [ 52.418632] [drm] Found smc ucode version: 0x00010200 [ 52.419416] [drm:dce4_afmt_write_speaker_allocation] *ERROR* Couldn't read Speaker Allocation Data Block: 0
Created attachment 119511 [details] dmesg when trying to log via kdm
Something is causing the GPU to hang repeatedly and the driver resets it repeatedly. I'd suggest updating your userspace acceleration drivers (xf86-video-ati, mesa).
I have mesa 10.0.0.1 already and x11 7.2.0. same user space works well on similar hardware. I can wait...
The first hang is caused when enabling radeon framebufer that does not work
Try removing the uvd firmware. That will cause the driver to skip UVD init. Maybe UVD is causing the problem.
Will do. Kids are watching TV... Sound like an initial is at I on problem because I saw a lot of people having the same behavior after resume. BTW Merry Christmas!
initialization problem (tablet!)
So It did a simple 'mv SUMO_uvd.bin.mov SUMO_uvd.bin.mov' in the /lib/firmware/radeon directory, and now X11 works correctly. Problem is that without uvd, I can't use a E350 as an HTPC as it will never decode blueray quality in software. Attached the new dmseg.
Created attachment 119571 [details] Dmesg without SUMO uvd firmware loading
Any thing I can try?