search

OpenELEC / OpenELEC.tv

OpenELEC - The living room PC for everyone
http://openelec.tv
1.6k stars 881 forks source link

ASIC hang back again #2296

Closed WebSpider closed 11 years ago

WebSpider commented 11 years ago

In OE 3.0.2 I've just experienced the ASIC hang again (as referenced in #315 for the first time) on a Zotac ZBOX AD02 (AMD E350)

May 15 20:51:41 openelec user.warn kernel: [ 5312.144619] <6>[fglrx] ASIC hang happened May 15 20:51:41 openelec user.warn kernel: [ 5312.144642] Pid: 944, comm: xbmc.bin Tainted: P O 3.7.10 #1 May 15 20:51:41 openelec user.warn kernel: [ 5312.144645] Call Trace: May 15 20:51:41 openelec user.warn kernel: [ 5312.144817] [] ? firegl_hardwareHangRecovery+0x1c/0x60 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.144932] [] ? _ZN18mmEnginesContainer9timestampEP26_QS_MM_TIMESTAMP_PACKET_INP27_QS_MM_TIMESTAMP_PACKET_OUT+0x184/0x1d0 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.144944] [] ? __cache_alloc+0x3c/0x543 May 15 20:51:41 openelec user.warn kernel: [ 5312.145077] [] ? firegl_trace+0x72/0x1e0 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145183] [] ? _ZN15QS_PRIVATE_CORE25escapeMultiMediaInterfaceEP21_QS_QUERY_API_CALL_INPvjS2_j+0xd3/0xe0 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145324] [] ? Z8uCWDDEQCmjjPvjS+0xffc/0x1200 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145342] [] ? should_resched+0x5/0x23 May 15 20:51:41 openelec user.warn kernel: [ 5312.145350] [] ? _cond_resched+0x6/0x1b May 15 20:51:41 openelec user.warn kernel: [ 5312.145424] [] ? firegl_cmmqs_CWDDE_32+0x36f/0x480 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145497] [] ? firegl_cmmqs_CWDDE32+0x8e/0x140 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145506] [] ? ns_capable+0x35/0x48 May 15 20:51:41 openelec user.warn kernel: [ 5312.145577] [] ? firegl_cmmqs_disabledriver+0x120/0x120 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145641] [] ? firegl_ioctl+0x1ed/0x250 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145727] [] ? ip_firegl_unlocked_ioctl+0x6/0xa [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145785] [] ? ip_firegl_mmap+0x62/0x62 [fglrx] May 15 20:51:41 openelec user.warn kernel: [ 5312.145793] [] ? vfs_ioctl+0x1e/0x31 May 15 20:51:41 openelec user.warn kernel: [ 5312.145798] [] ? do_vfs_ioctl+0x3ed/0x42f May 15 20:51:41 openelec user.warn kernel: [ 5312.145804] [] ? mmdrop+0xd/0x1c May 15 20:51:41 openelec user.warn kernel: [ 5312.145810] [] ? finish_task_switch+0x4e/0x75 May 15 20:51:41 openelec user.warn kernel: [ 5312.145815] [] ? schedule+0x360/0x377 May 15 20:51:41 openelec user.warn kernel: [ 5312.145820] [] ? sys_ioctl+0x3b/0x65 May 15 20:51:41 openelec user.warn kernel: [ 5312.145827] [] ? system_call_fastpath+0x16/0x1b May 15 20:51:41 openelec user.warn kernel: [ 5312.145837] pubdev:0xffffffffa04cf780, num of device:1 , name:fglrx, major 9, minor 0. May 15 20:51:41 openelec user.warn kernel: [ 5312.145842] device 0 : 0xffff8800548b4000 . May 15 20:51:41 openelec user.warn kernel: [ 5312.145847] Asic ID:0x9802, revision:0x23, MMIOReg:0xffffc90011100000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145852] FB phys addr: 0xb0000000, MC :0xf00000000, Total FB size :0x18000000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145858] gart table MC:0xf0fe6a000, Physical:0xbfe6a000, size:0x195000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145864] mc_node :FB, total 1 zones May 15 20:51:41 openelec user.warn kernel: [ 5312.145869] MC start:0xf00000000, Physical:0xb0000000, size:0x10000000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145875] Mapped heap -- Offset:0x0, size:0xfe6a000, reference count:11, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145881] Mapped heap -- Offset:0x0, size:0x1000000, reference count:1, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145886] Mapped heap -- Offset:0xfe6a000, size:0x196000, reference count:1, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145890] mc_node :INV_FB, total 1 zones May 15 20:51:41 openelec user.warn kernel: [ 5312.145895] MC start:0xf10000000, Physical:0xc0000000, size:0x8000000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145900] Mapped heap -- Offset:0x7ff4000, size:0xc000, reference count:1, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145905] mc_node :GART_USWC, total 3 zones May 15 20:51:41 openelec user.warn kernel: [ 5312.145909] MC start:0x22eb0000, Physical:0x0, size:0x1f800000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145915] Mapped heap -- Offset:0x2810000, size:0x800000, reference count:2, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145921] Mapped heap -- Offset:0x2010000, size:0x800000, reference count:3, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145927] Mapped heap -- Offset:0x10000, size:0x2000000, reference count:14, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145931] mc_node :GART_CACHEABLE, total 3 zones May 15 20:51:41 openelec user.warn kernel: [ 5312.145935] MC start:0x10400000, Physical:0x0, size:0x12ab0000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145941] Mapped heap -- Offset:0x200000, size:0x1000000, reference count:2, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145947] Mapped heap -- Offset:0x0, size:0x200000, reference count:3, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145952] Mapped heap -- Offset:0xef000, size:0x11000, reference count:1, mapping count:0, May 15 20:51:41 openelec user.warn kernel: [ 5312.145956] mc_node :PEER_FB_GART, total 1 zones May 15 20:51:41 openelec user.warn kernel: [ 5312.145961] MC start:0x426b0000, Physical:0x0, size:0x1000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145967] GRBM : 0xe7734828, SRBM : 0x20004840 . May 15 20:51:41 openelec user.warn kernel: [ 5312.145973] CP_RB_BASE : 0x22ec00, CP_RB_RPTR : 0x28d70 , CP_RB_WPTR :0x28d70. May 15 20:51:41 openelec user.warn kernel: [ 5312.145979] CP_IB1_BUFSZ:0x1938, CP_IB1_BASE_HI:0x0, CP_IB1_BASE_LO:0x2339b000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145985] last submit IB buffer -- MC :0x2339b000,phys:0x4d829000. May 15 20:51:41 openelec user.warn kernel: [ 5312.145991] Dump the trace queue. May 15 20:51:41 openelec user.warn kernel: [ 5312.145993] End of dump

fritsch commented 11 years ago

Just to make sure, this is a standard OE 3.0.2 with no additional selfbuild patches, right? Can you somehow reproduce (should be rather difficult to do it deterministic) - or can you also supply xbmc.log? Not nice at all.

@FernetMenta: Asic hang seems back.

FernetMenta commented 11 years ago

That was my fear. See my comments here: https://github.com/xbmc/xbmc/pull/2540 There is some work in progress that addresses this: https://github.com/xbmc/xbmc/pull/2681 https://github.com/xbmc/xbmc/pull/2309

Does this occur during video playback or while in the gui?

EDIT: that does NOT mean that 2540 introduced the problem again. Seems like fences are the only reliable way to get this under control.

WebSpider commented 11 years ago

This is a completely standard OE302. No patches, from autoupdater, Fusion x64.

This hang occured when back in the GUI, right after video playback of a 720p mkv stream from CIFS. I will play the stream again to see if it is related to that particular file, will let you know.

WebSpider commented 11 years ago

XBMC logs can be found here:

http://pastebin.com/3jqNv8Mj

WebSpider commented 11 years ago

Unfortunately, just playing the same file, ending up in the same menu doesn't reproduce this. Does anyone have better ideas on how to reproduce this issue?

fritsch commented 11 years ago

@WebSpider: Work is done to get that working. But it is a really long standing issue and solving it to 100% needs a whole lot of code within xbmc.

stefansaraev commented 11 years ago

no fglrx bugs will be fixed. fglrx is a history now.