Hi, A weird bug affects my system since 3.15 pre-rc: sometimes the system becomes unresponsive, for example when I try to open the logout/shutdown/reboot KDE menu from the "K" button. It takes ~30 seconds to open the menu and in the meanwhile the system starts to behave strangely: for example I can't load any web page in Chrome until the menu finally appears. Here is full dmesg: http://bpaste.net/show/262088/ [ 360.726656] INFO: task pulseaudio:3190 blocked for more than 120 seconds. [ 360.726659] Not tainted 3.15.0-rc4-core-avx-i #1 [ 360.726660] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 360.726661] pulseaudio D ffff88042f354500 0 3190 3189 0x00000000 [ 360.726663] ffff880410913610 0000000000000002 ffff8804128e4b60 ffff880410913fd8 [ 360.726665] 0000000000014500 0000000000014500 ffff880418e98000 ffff8804173ac890 [ 360.726666] ffff8804173ac938 ffff8804173ac9f8 0000000000000000 0000000000000000 [ 360.726668] Call Trace: [...] I tought the problem had something to do with pulseaudio, so I removed it but nothing really changed: Full dmesg: http://bpaste.net/show/262089/ [ 240.533969] INFO: task kded4:2839 blocked for more than 120 seconds. [ 240.534029] INFO: task kmixctrl:3181 blocked for more than 120 seconds. [ 240.534116] INFO: task kmix:3201 blocked for more than 120 seconds. [ 360.710646] INFO: task kded4:2839 blocked for more than 120 seconds. [ 360.710724] INFO: task kmixctrl:3181 blocked for more than 120 seconds. [ 360.710803] INFO: task kmix:3201 blocked for more than 120 seconds. [ 480.887194] INFO: task kded4:2839 blocked for more than 120 seconds. [ 480.887280] INFO: task kmixctrl:3181 blocked for more than 120 seconds. [ 480.887360] INFO: task kmix:3201 blocked for more than 120 seconds. [ 601.063803] INFO: task kded4:2839 blocked for more than 120 seconds. I believe it's a drm related bug because I remember having the very same problem when I applied my own drm backports (http://www.linuxsystems.it/linux-drm-graphic-stack-backports/) on top of 3.14. But I have the same problem with both my AMD Radeon HD7950 and my Intel HD 4000 which is quite strange for a drm bug... 3.14.0 works fine.
I just noticed I have the very same problem with nomodeset+vesa so it has nothing to do with drm, the problem is elsewhere in the kernel.
Best bet is to bisect.
# git bisect start # git bisect good 455c6fdbd219161bd09b1165f11699d6d73de11c # git bisect bad c9eaa447e77efe77b7fa4c953bd62de8297fd6c5 # git bisect bad c9eaa447e77efe77b7fa4c953bd62de8297fd6c5 Bisecting: 5578 revisions left to test after this (roughly 13 steps) [cd6362befe4cc7bf589a5236d2a780af2d47bcc9] Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next # git bisect bad Bisecting: 3353 revisions left to test after this (roughly 12 steps) [c12e69c6aaf785fd307d05cb6f36ca0e7577ead7] Merge tag 'staging-3.15-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging # git bisect bad Bisecting: 2255 revisions left to test after this (roughly 11 steps) [4b1779c2cf030c68aefe939d946475e4136c1895] Merge tag 'pci-v3.15-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci # git bisect good Bisecting: 1127 revisions left to test after this (roughly 10 steps) [bea8052074691c70526ae6a555d25f111c0aad0a] staging: comedi: c6xdigio: introduce c6xdigio_get_encoder_bits() # git bisect skip Bisecting: 1127 revisions left to test after this (roughly 10 steps) [04cc79a048ee215ec39af05d61f1fc8a4ab3d8c1] ALSA: seq: Use standard printk helpers # git bisect bad Bisecting: 30 revisions left to test after this (roughly 5 steps) [60c5772b50fe8ce947685cfe3a80b9bf6ea92831] ALSA: pci: Convert to snd_card_new() with a device pointer # git bisect bad Bisecting: 14 revisions left to test after this (roughly 4 steps) [8e54b4accdf0cc30a0246fdc4cced0c2f463254a] ALSA: hda - Rename ASUS subwoofer quirks # git bisect bad Bisecting: 6 revisions left to test after this (roughly 3 steps) [70540e245c727114d0477f00ec7d83a2d2fe666d] ALSA: hda/conexant - Add analog loopback mixing to CX20549 # git bisect good Bisecting: 3 revisions left to test after this (roughly 2 steps) [3d692451eaf213e44f343aa2be06501b56ab2077] ALSA: hda - Do not accept responses from non-existing codecs # git bisect good Bisecting: 1 revision left to test after this (roughly 1 step) [400362f1d8dcfda3562e80e88cfc2a92cffaf9bf] ALSA: usb-audio: Resume mixer values properly # git bisect bad Bisecting: 0 revisions left to test after this (roughly 0 steps) [63f4b3a475e84c5aa77f720f0f66989c686a14fb] Merge branch 'topic/hda' into for-next # git bisect good 400362f1d8dcfda3562e80e88cfc2a92cffaf9bf is the first bad commit commit 400362f1d8dcfda3562e80e88cfc2a92cffaf9bf Author: Takashi Iwai <tiwai@suse.de> Date: Mon Jan 20 16:51:16 2014 +0100 ALSA: usb-audio: Resume mixer values properly Implement reset_resume callback so that the mixer values are properly restored. Still no boot quirks are called, so it might not work well on some devices. Signed-off-by: Takashi Iwai <tiwai@suse.de> :040000 040000 b11fe2b1239594a6528dd33b5fa0f5c13d45278e ffec5159801c47f8a289e587bc39e217fea51937 M sound Please revert that little bastard: 400362f1d8dcfda3562e80e88cfc2a92cffaf9bf :-)
The bug was already fixed in the recent commits 1ee23fe07ee83a38ecee927e701f762888ada942 and 1c53e7253ed8769a00afa0f06777d731dbe1ba6f.
I confirm the fix works, thanks.