Created attachment 83189 [details] call trace dmesg System Environment: -------------------------------------------- Kernel: (drm-intel-next-queued)fae5cbff3e3e4097de2e76178d462b90626c2bdb Author: Daniel Vetter <daniel.vetter@ffwll.ch> Date: Sun Jul 21 21:37:09 2013 +0200 drm/i915: clean up crtc timings computation Bug detailed description: -------------------------------------------- Call trace appears after system boots with DP pipe. In boot dmesg, which I will attach in the attachment, you can see two call trace. The first call trace can reproduce every time, but the second call trace appears sporadically. The dmesg I grabbed happened to has the second call trace. Reproduce steps: -------------------------------------------- 1. boot system with DP plugged in 2. grab dmesg
Hm, link train failures, pipe off timeouts ... looks like fun. The second calltrace seems to be an issue in the firewire code, so not our problem.
One for Todd to ponder for sure ...
This should be entertaining. I'll see what I can do with it. Now to find SNB hardware...
(In reply to comment #3) > This should be entertaining. I'll see what I can do with it. Now to find SNB > hardware... Hi, I check this again(using a latest -next-queued kernel) and find this issue happens on SNB, IVB and HSW platforms. Besides, calltrace has nothing to do with DP, it happens even if system booting without a monitor. So I think you can use any of these platforms to reproduce it.. :)
(In reply to comment #4) > (In reply to comment #3) > > This should be entertaining. I'll see what I can do with it. Now to find SNB > > hardware... > > Hi, I check this again(using a latest -next-queued kernel) and find this > issue happens on SNB, IVB and HSW platforms. > Besides, calltrace has nothing to do with DP, it happens even if system > booting without a monitor. So I think you can use any of these platforms to > reproduce it.. :) We have two calltraces in the dmesg here. The first one is [ 1.177979] WARNING: CPU: 1 PID: 1236 at drivers/gpu/drm/i915/intel_display.c:842 ironlake_crtc_disable+0xd6/0x83c [i915]() Thus far that tends to only happen with DP when we try to switch it off. The 2nd calltrace is [ 1.860783] firewire_ohci 0000:04:03.0: failed to read phy reg 2 [ 1.863475] CPU: 2 PID: 1435 Comm: udevd Tainted: G W 3.11.0-rc2_drm-intel-next-queued_fae5cb_20130727_+ #23723 [ 1.866130] Hardware name: /DQ67SW, BIOS SWQ6710H.86A.0060.2011.1220.1805 12/20/2011 [ 1.868786] 0000000000000000 ffff8802302fe000 ffffffff816ed6a8 0000000000000067 [ 1.871438] ffffffffa019929a ffff8802302fe000 ffffffffa0184300 0000000000000016 [ 1.874051] ffffffffa019b4ad ffffffffa019bda0 ffff8802302fe000 ffff8802302fe630 [ 1.876673] Call Trace: [ 1.879231] [<ffffffff816ed6a8>] ? dump_stack+0x41/0x51 [ 1.881843] [<ffffffffa019929a>] ? read_phy_reg+0x71/0x82 [firewire_ohci] [ 1.884455] [<ffffffffa019b4ad>] ? ohci_enable+0x26f/0x535 [firewire_ohci] [ 1.887023] [<ffffffffa017c91d>] ? fw_card_add+0x42/0x8a [firewire_core] [ 1.889594] [<ffffffffa01990bf>] ? pci_probe+0x548/0x6b2 [firewire_ohci] [ 1.892166] [<ffffffff8112cba7>] ? sysfs_do_create_link_sd+0x137/0x188 [ 1.894748] [<ffffffff81363fb9>] ? driver_probe_device+0x19e/0x19e [ 1.897369] [<ffffffff812e1a6d>] ? local_pci_probe+0x16/0x26 [ 1.899939] [<ffffffff812e232b>] ? pci_device_probe+0xbc/0xe0 [ 1.902502] [<ffffffff81363ea5>] ? driver_probe_device+0x8a/0x19e [ 1.905054] [<ffffffff8136400c>] ? __driver_attach+0x53/0x73 [ 1.907646] [<ffffffff813626d0>] ? bus_for_each_dev+0x4e/0x7f [ 1.910242] [<ffffffff813636ca>] ? bus_add_driver+0xe5/0x204 [ 1.912772] [<ffffffff8136452e>] ? driver_register+0x8c/0x104 [ 1.915302] [<ffffffffa019f000>] ? 0xffffffffa019efff [ 1.917855] [<ffffffff8100025a>] ? do_one_initcall+0x78/0xf6 [ 1.920393] [<ffffffff8104ebbf>] ? __blocking_notifier_call_chain+0x4f/0x5d [ 1.922917] [<ffffffff81073a2e>] ? load_module+0x17ec/0x1aba [ 1.925447] [<ffffffff8107151a>] ? free_notes_attrs+0x3f/0x3f [ 1.928027] [<ffffffff81073d8c>] ? SyS_init_module+0x90/0x9b [ 1.930563] [<ffffffff816f7ed2>] ? system_call_fastpath+0x16/0x1b [ 1.933178] firewire_ohci: probe of 0000:04:03.0 failed with error -16 That on is not our bug and you can ignore it (maybe it's fixed in latest upstream already).
I've forgotten to actually add my question ;-) Which of the two types of calltraces do you see on snb-hsw, even without DP?
(In reply to comment #5) > (In reply to comment #4) > > (In reply to comment #3) > > > This should be entertaining. I'll see what I can do with it. Now to find SNB > > > hardware... > > > > Hi, I check this again(using a latest -next-queued kernel) and find this > > issue happens on SNB, IVB and HSW platforms. > > Besides, calltrace has nothing to do with DP, it happens even if system > > booting without a monitor. So I think you can use any of these platforms to > > reproduce it.. :) > > We have two calltraces in the dmesg here. The first one is > > [ 1.177979] WARNING: CPU: 1 PID: 1236 at > drivers/gpu/drm/i915/intel_display.c:842 ironlake_crtc_disable+0xd6/0x83c > [i915]() > > Thus far that tends to only happen with DP when we try to switch it off. > I'm a little confused here..Do you mean this "WARNING" is what we should focus?
(In reply to comment #7) > (In reply to comment #5) > > We have two calltraces in the dmesg here. The first one is > > > > [ 1.177979] WARNING: CPU: 1 PID: 1236 at > > drivers/gpu/drm/i915/intel_display.c:842 ironlake_crtc_disable+0xd6/0x83c > > [i915]() > > > > Thus far that tends to only happen with DP when we try to switch it off. > > > I'm a little confused here..Do you mean this "WARNING" is what we should > focus? Yes, only this WARNING here is the bug we're tracking here, see comment #1
Created attachment 83956 [details] boot without monitor
(In reply to comment #6) > I've forgotten to actually add my question ;-) Which of the two types of > calltraces do you see on snb-hsw, even without DP? I test latest -next-queued kernel on a hsw desktop without any display ports, it seems like we also have two types of calltraces. I grabbed the dmesg. :)
(In reply to comment #6) > I've forgotten to actually add my question ;-) Which of the two types of > calltraces do you see on snb-hsw, even without DP? Oh, I got you. We have the first calltrace even without DP.
(In reply to comment #10) > (In reply to comment #6) > > I've forgotten to actually add my question ;-) Which of the two types of > > calltraces do you see on snb-hsw, even without DP? > > I test latest -next-queued kernel on a hsw desktop without any display > ports, it seems like we also have two types of calltraces. I grabbed the > dmesg. :) Type wrong, only one calltrace , sorry. :)
(In reply to comment #10) > (In reply to comment #6) > > I've forgotten to actually add my question ;-) Which of the two types of > > calltraces do you see on snb-hsw, even without DP? > > I test latest -next-queued kernel on a hsw desktop without any display > ports, it seems like we also have two types of calltraces. I grabbed the > dmesg. :) This is again a different calltrace: [ 1.168965] WARNING: CPU: 1 PID: 1258 at drivers/gpu/drm/i915/i915_gem.c:3440 i915_gem_object_set_cache_level+0x1bd/0x1dc [i915]() Please consult the section about backtraces in the bug filing BKM about how to properly differentiate calltraces in dmesg. We can't just smash all the calltraces into one bug report ;-) Differentiating calltraces properly is imo a very important part of the first-stage bug traiging QA does.
(In reply to comment #13) > [ 1.168965] WARNING: CPU: 1 PID: 1258 at > drivers/gpu/drm/i915/i915_gem.c:3440 > i915_gem_object_set_cache_level+0x1bd/0x1dc [i915]() I've forgotten to add: Please file a new bug report for this issue. On a quick look it looks like a regression, so the bisect result would also be interesting.
It's pretty obvious, is it not? Just an incorrect warn finally barfing.
(Not that I've been seeing it for a couple of years and had been waiting for some one else to finally notice. :)
*** Bug 71905 has been marked as a duplicate of this bug. ***
Is this still happpening? This may have been fixed when Ville did the clean up work on all the various WARNs a while back.
Created attachment 92792 [details] dmesg after boot up with DP The Calltrace still exits when machine boot up with DP pipe. WARNING: CPU: 3 PID: 1222 at drivers/gpu/drm/i915/intel_display.c:857 intel_wait_for_pipe_off+0x166/0x16d [i915]() You can check the details in the attachment.
Can you provide a current dmesg for this, preferably from a -nightly kernel? I think I have an SNB system to test this on and if not, I'll secure one. From what I can see in the original logs, it looks like we're having issues shutting down the pipe, but it's unclear why. Also when we try to train right after the call trace, the pipe is disabled or non-functional (the logs show similar failures as when the main link PHYs are disabled).
(In reply to comment #20) > Can you provide a current dmesg for this, preferably from a -nightly kernel? yes, I check it with drm-intel-nightly_18a766_20140513 kernel, And attach a dmesg that bases on the kernel I said (-nightly_18a766). > I think I have an SNB system to test this on and if not, I'll secure one. > From what I can see in the original logs, it looks like we're having issues > shutting down the pipe, but it's unclear why. Also when we try to train > right after the call trace, the pipe is disabled or non-functional (the logs > show similar failures as when the main link PHYs are disabled). [ 1.374618] ------------[ cut here ]------------ [ 1.374646] WARNING: CPU: 0 PID: 1338 at drivers/gpu/drm/i915/intel_display.c:944 intel_wait_for_pipe_off+0x12e/0x135 [i915]() [ 1.374647] pipe_off wait timed out [ 1.374649] Modules linked in: i915(+) video button drm_kms_helper drm [ 1.374651] CPU: 0 PID: 1338 Comm: udevd Not tainted 3.15.0-rc3_drm-intel-nightly_18a766_20140513+ #2566 [ 1.374652] Hardware name: Dell Inc. OptiPlex 990/0DXWW6, BIOS A02 02/26/2011 [ 1.374655] 00000000 c16a5021 f5bd9a3c c102cf1b f81c8e9c f4f10000 00070008 fffb716b [ 1.374657] f5b695b4 c102cf96 00000009 f5bd9a3c f820bdd6 f5bd9a54 f81c8e9c f820a5c8 [ 1.374660] 000003b0 f820bdd6 f4f10000 f5b695b8 f5af4000 f81ce68e 00000001 f4cae224 [ 1.374660] Call Trace: [ 1.374664] [<c16a5021>] ? dump_stack+0x3e/0x4e [ 1.374667] [<c102cf1b>] ? warn_slowpath_common+0x61/0x74 [ 1.374688] [<f81c8e9c>] ? intel_wait_for_pipe_off+0x12e/0x135 [i915] [ 1.374690] [<c102cf96>] ? warn_slowpath_fmt+0x29/0x2d [ 1.374708] [<f81c8e9c>] ? intel_wait_for_pipe_off+0x12e/0x135 [i915] [ 1.374727] [<f81ce68e>] ? ironlake_crtc_disable+0x85/0x713 [i915] [ 1.374746] [<f81d04ae>] ? __intel_set_mode+0xbe7/0xffd [i915] [ 1.374766] [<f81d2196>] ? intel_set_mode+0x11/0x25 [i915] [ 1.374784] [<f81d2bed>] ? intel_crtc_set_config+0x646/0x8f1 [i915] [ 1.374791] [<f80950e0>] ? drm_mode_set_config_internal+0x39/0x8a [drm] [ 1.374793] [<f80c7322>] ? drm_fb_helper_restore_fbdev_mode+0x7c/0x91 [drm_kms_helper] [ 1.374811] [<f81c5fd1>] ? lpt_disable_clkout_dp+0xa1/0xa1 [i915] [ 1.374813] [<f80c7656>] ? drm_fb_helper_set_par+0x3a/0x59 [drm_kms_helper] [ 1.374816] [<c12bb586>] ? fbcon_init+0x2c3/0x3a3 [ 1.374818] [<c1304e1f>] ? visual_init+0x97/0xe9 [ 1.374820] [<c13063a5>] ? do_bind_con_driver+0x16e/0x275 [ 1.374822] [<c13065b8>] ? do_take_over_console+0x10c/0x134 [ 1.374824] [<c12bad4a>] ? do_fbcon_takeover+0x49/0x88 [ 1.374825] [<c16ac3db>] ? notifier_call_chain+0x20/0x42 [ 1.374828] [<c10429e0>] ? __blocking_notifier_call_chain+0x34/0x4b [ 1.374829] [<c1042a00>] ? blocking_notifier_call_chain+0x9/0xc [ 1.374831] [<c12c3686>] ? register_framebuffer+0x1f6/0x248 [ 1.374834] [<f80c6e22>] ? drm_fb_helper_initial_config+0x355/0x3d4 [drm_kms_helper] [ 1.374858] [<f81f2d90>] ? i915_driver_load+0x959/0xbb7 [i915] [ 1.374860] [<c169ed0d>] ? klist_node_init+0x3a/0x3d [ 1.374862] [<c169ed95>] ? klist_add_tail+0x16/0x30 [ 1.374864] [<c131cd56>] ? device_add+0x45d/0x467 [ 1.374866] [<c100303a>] ? show_stack_log_lvl+0x2e/0x8a [ 1.374871] [<f809128f>] ? drm_minor_register+0xf5/0x149 [drm] [ 1.374875] [<f8091345>] ? drm_dev_register+0x62/0xc3 [drm] [ 1.374880] [<f8092f01>] ? drm_get_pci_dev+0xd7/0x16a [drm] [ 1.374882] [<c12aa8ad>] ? pci_device_probe+0x4e/0x95 [ 1.374884] [<c131e71e>] ? driver_probe_device+0x80/0x180 [ 1.374886] [<c131e862>] ? __driver_attach+0x44/0x5f [ 1.374887] [<c131d52b>] ? bus_for_each_dev+0x37/0x61 [ 1.374889] [<c131e3e2>] ? driver_attach+0x11/0x13 [ 1.374890] [<c131e81e>] ? driver_probe_device+0x180/0x180 [ 1.374892] [<c131e11f>] ? bus_add_driver+0xad/0x167 [ 1.374894] [<c131ecb6>] ? driver_register+0x6b/0x99 [ 1.374895] [<f8224000>] ? 0xf8223fff [ 1.374896] [<c10003de>] ? do_one_initcall+0x6d/0xe7 [ 1.374897] [<c16ac3db>] ? notifier_call_chain+0x20/0x42 [ 1.374899] [<c10429e9>] ? __blocking_notifier_call_chain+0x3d/0x4b [ 1.374901] [<c1071bb7>] ? load_module+0x142a/0x1704 [ 1.374903] [<c16ac3b3>] ? __do_page_fault+0x46c/0x46c [ 1.374905] [<c1071f03>] ? SyS_init_module+0x72/0x8a [ 1.374907] [<c16adeac>] ? sysenter_do_call+0x12/0x22 [ 1.374908] ---[ end trace c9b4fcba76de1ec9 ]---
Created attachment 99003 [details] dmesg of latest -nightly kernel
Please test this patch here, should apply to any recent-ish upstream: http://patchwork.freedesktop.org/patch/24864/
(In reply to comment #23) > Please test this patch here, should apply to any recent-ish upstream: > > http://patchwork.freedesktop.org/patch/24864/ This patch fixes problem "pipe_off wait timed out".But I found this Call Trace. [ 1.423773] usb 2-1: new high-speed USB device number 2 using ehci-pci [ 1.490861] ------------[ cut here ]------------ [ 1.490901] WARNING: CPU: 0 PID: 1262 at drivers/gpu/drm/i915/intel_display.c:860 ironlake_crtc_disable+0x86/0x7f2 [i915]() [ 1.490901] pipe_off wait timed out [ 1.490904] Modules linked in: i915(+) video button drm_kms_helper drm [ 1.490906] CPU: 0 PID: 1262 Comm: udevd Not tainted 3.14.0_kcloud_dde711_20140523+ #51 [ 1.490907] Hardware name: /DQ67SW, BIOS SWQ6710H.86A.0060.2011.1220.1805 12/20/2011 [ 1.490909] 0000000000000000 ffffffff816fd068 ffff88023027b5e0 ffffffff81033e44 [ 1.490910] ffffffffa009b4ed ffff8800b8658000 ffff88023027b630 ffff880234af0000 [ 1.490911] ffff88022f909000 ffffffff81033ea0 ffffffffa00e0245 0000000000000018 [ 1.490912] Call Trace: [ 1.490916] [<ffffffff816fd068>] ? dump_stack+0x41/0x51 [ 1.490918] [<ffffffff81033e44>] ? warn_slowpath_common+0x6f/0x84 [ 1.490939] [<ffffffffa009b4ed>] ? ironlake_crtc_disable+0x86/0x7f2 [i915] [ 1.490940] [<ffffffff81033ea0>] ? warn_slowpath_fmt+0x47/0x49 [ 1.490957] [<ffffffffa009b4ed>] ? ironlake_crtc_disable+0x86/0x7f2 [i915] [ 1.490974] [<ffffffffa00926ee>] ? intel_dump_pipe_config+0x17d/0x311 [i915] [ 1.490991] [<ffffffffa009d13c>] ? __intel_set_mode+0x862/0x118c [i915] [ 1.491008] [<ffffffffa009f682>] ? intel_set_mode+0xb/0x22 [i915] [ 1.491024] [<ffffffffa00a0476>] ? intel_crtc_set_config+0x95c/0x9f6 [i915] [ 1.491030] [<ffffffffa000be6c>] ? drm_mode_set_config_internal+0x48/0xc0 [drm] [ 1.491033] [<ffffffffa00484cf>] ? drm_fb_helper_restore_fbdev_mode+0x90/0xa5 [drm_kms_helper] [ 1.491035] [<ffffffffa0049d7f>] ? drm_fb_helper_set_par+0x42/0x69 [drm_kms_helper] [ 1.491037] [<ffffffff8130717b>] ? fbcon_init+0x2fa/0x400 [ 1.491040] [<ffffffff81352b89>] ? visual_init+0xaf/0x102 [ 1.491042] [<ffffffff8135448a>] ? do_bind_con_driver+0x19e/0x2c2 [ 1.491044] [<ffffffff81354a4b>] ? do_take_over_console+0x125/0x152 [ 1.491046] [<ffffffff81306885>] ? do_fbcon_takeover+0x4e/0x8e [ 1.491047] [<ffffffff817053ec>] ? notifier_call_chain+0x32/0x52 [ 1.491049] [<ffffffff8104cb4e>] ? __blocking_notifier_call_chain+0x43/0x5b [ 1.491051] [<ffffffff812fea77>] ? register_framebuffer+0x265/0x29b [ 1.491053] [<ffffffffa004996f>] ? drm_fb_helper_initial_config+0x3b9/0x45f [drm_kms_helper] [ 1.491070] [<ffffffffa008cc97>] ? gen6_write32+0x22/0x6b [i915] [ 1.491086] [<ffffffffa008cc97>] ? gen6_write32+0x22/0x6b [i915] [ 1.491107] [<ffffffffa00c36cd>] ? i915_driver_load+0xb7b/0xd70 [i915] [ 1.491109] [<ffffffff810428e1>] ? call_usermodehelper_exec+0xcd/0xfb [ 1.491111] [<ffffffff812cb426>] ? kobject_uevent_env+0x3de/0x416 [ 1.491113] [<ffffffff8136cac9>] ? get_device+0xe/0x16 [ 1.491115] [<ffffffff816f5519>] ? klist_add_tail+0x1b/0x3c [ 1.491116] [<ffffffff8136de3b>] ? device_add+0x4ec/0x4fd [ 1.491122] [<ffffffffa000744d>] ? drm_dev_register+0x72/0xe3 [drm] [ 1.491127] [<ffffffffa00093ae>] ? drm_get_pci_dev+0xfc/0x1ba [drm] [ 1.491129] [<ffffffff812eea09>] ? local_pci_probe+0x34/0x77 [ 1.491130] [<ffffffff812eeb0d>] ? pci_device_probe+0xc1/0xe5 [ 1.491132] [<ffffffff8136feaa>] ? driver_probe_device+0x92/0x1b3 [ 1.491134] [<ffffffff8137004f>] ? __driver_attach+0x53/0x73 [ 1.491135] [<ffffffff8136fffc>] ? __device_attach+0x31/0x31 [ 1.491137] [<ffffffff8136e796>] ? bus_for_each_dev+0x6e/0x78 [ 1.491138] [<ffffffff8136f76e>] ? bus_add_driver+0xfb/0x1c4 [ 1.491140] [<ffffffff81370545>] ? driver_register+0x83/0xbb [ 1.491141] [<ffffffffa00fc000>] ? 0xffffffffa00fbfff [ 1.491142] [<ffffffff8100028f>] ? do_one_initcall+0x78/0xf9 [ 1.491144] [<ffffffff817053ec>] ? notifier_call_chain+0x32/0x52 [ 1.491145] [<ffffffff8104cb59>] ? __blocking_notifier_call_chain+0x4e/0x5b [ 1.491148] [<ffffffff8107dacd>] ? load_module+0x1969/0x1aef [ 1.491149] [<ffffffff8107b1dd>] ? store_uevent+0x30/0x30 [ 1.491152] [<ffffffff8107dce1>] ? SyS_init_module+0x8e/0x99 [ 1.491153] [<ffffffff817074e2>] ? system_call_fastpath+0x16/0x1b [ 1.491154] ---[ end trace 313cd642c53a8e83 ]---
Created attachment 99617 [details] dmesg_with_patch
Please test this patch http://patchwork.freedesktop.org/patch/32012/
Ping. Any luck with this patch? It is merged already. So please test latest drm-intel-nightly
(In reply to Rodrigo Vivi from comment #27) > Ping. Any luck with this patch? It is merged already. > So please test latest drm-intel-nightly Test passes on latest testing(e62b9d8268b9f6137681e875dba4d1871120aa8b). Verified.
Closing old verified.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.