)]}' { "log": [ { "commit": "68d3f1d810500e8b975bdf0b20dd83d060076b4b", "tree": "68d07e9d165aa9fde390dfa5694bb58aa63adb0d", "parents": [ "3aaba20f26f58843e8f20611e5c0b1c06954310f" ], "author": { "name": "Don Zickus", "email": "dzickus@redhat.com", "time": "Tue Aug 31 23:00:07 2010 -0400" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Wed Sep 01 10:02:28 2010 +0200" }, "message": "lockup_detector: Sync touch_*_watchdog back to old semantics\n\nDuring my rewrite, the semantics of touch_nmi_watchdog and\ntouch_softlockup_watchdog changed enough to break some drivers\n(mostly over preemptable regions).\n\nThese are cases where long delays on one CPU (due to\nprint_delay for example) can cause long delays on other\nCPUs - so we must \u0027touch\u0027 the nmi_watchdog flag of those\nother CPUs as well.\n\nThis change brings those touch_*_watchdog() functions back in line\nwith to how they used to work.\n\nSigned-off-by: Don Zickus \u003cdzickus@redhat.com\u003e\nAcked-by: Cyrill Gorcunov \u003cgorcunov@openvz.org\u003e\nCc: peterz@infradead.org\nCc: fweisbec@gmail.com\nLKML-Reference: \u003c1283310009-22168-2-git-send-email-dzickus@redhat.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "c6db67cda735d8ace5f19c3831240e1408679790", "tree": "8512efd772cb3d3f30b49001890c93a22f2db50f", "parents": [ "ede1b4290781ae82ccf0f2ecc6dada8d3dd35779" ], "author": { "name": "Peter Zijlstra", "email": "a.p.zijlstra@chello.nl", "time": "Fri Aug 20 11:49:15 2010 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Mon Aug 23 10:48:05 2010 +0200" }, "message": "watchdog: Don\u0027t throttle the watchdog\n\nStephane reported that when the machine locks up, the regular ticks,\nwhich are responsible to resetting the throttle count, stop too.\n\nHence the NMI watchdog can end up being throttled before it reports on\nthe locked up state, and we end up being sad..\n\nCure this by having the watchdog overflow reset its own throttle count.\n\nReported-by: Stephane Eranian \u003ceranian@google.com\u003e\nTested-by: Stephane Eranian \u003ceranian@google.com\u003e\nCc: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\nSigned-off-by: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nLKML-Reference: \u003c1282215916.1926.4696.camel@laptop\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "eb703f98191a505f78d0066712ad67d5dedc4c90", "tree": "5dc85a0e152c6b4d25816f4ba5f98d3207054c20", "parents": [ "e35e7fb0e9ea557f7504ced6fe4ccf69e44b7f07" ], "author": { "name": "Kulikov Vasiliy", "email": "segooon@gmail.com", "time": "Mon Jul 05 12:00:54 2010 +0400" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Wed Jul 07 08:46:42 2010 +0200" }, "message": "kernel/watchdog: Initialize \u0027result\u0027\n\nVariable on the stack is not initialized to zero, do it\nexplicitly.\n\nThis bug was found by a compiler warning:\n\n kernel/watchdog.c:463: warning: \u0027result\u0027 may be used uninitialized in this function\n\nSigned-off-by: Kulikov Vasiliy \u003csegooon@gmail.com\u003e\nAcked-by: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\nCc: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nCc: Arnaldo Carvalho de Melo \u003cacme@redhat.com\u003e\nCc: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\nCc: Paul Mackerras \u003cpaulus@samba.org\u003e\nCc: Mike Galbraith \u003cefault@gmx.de\u003e\nCc: Steven Rostedt \u003crostedt@goodmis.org\u003e\nLKML-Reference: \u003c1278316854-28442-1-git-send-email-segooon@gmail.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "26e09c6eee14f4827b55137ba0eedc4e77cd50ab", "tree": "9088d86ce87557e4d35a7b5832fdbe1d235516e4", "parents": [ "cafcd80d216bc2136b8edbb794327e495792c666" ], "author": { "name": "Don Zickus", "email": "dzickus@redhat.com", "time": "Mon May 17 18:06:04 2010 -0400" }, "committer": { "name": "Frederic Weisbecker", "email": "fweisbec@gmail.com", "time": "Wed May 19 11:32:14 2010 +0200" }, "message": "lockup_detector: Convert per_cpu to __get_cpu_var for readability\n\nJust a bunch of conversions as suggested by Frederic W.\n__get_cpu_var() provides preemption disabled checks.\n\nPlus it gives more readability as it makes it obvious\nwe are dealing locally now with these vars.\n\nSigned-off-by: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Ingo Molnar \u003cmingo@elte.hu\u003e\nCc: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nCc: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Cyrill Gorcunov \u003cgorcunov@gmail.com\u003e\nLKML-Reference: \u003c1274133966-18415-2-git-send-email-dzickus@redhat.com\u003e\nSigned-off-by: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\n" }, { "commit": "cafcd80d216bc2136b8edbb794327e495792c666", "tree": "4eb819c124684b14d48faf094316e8dd3ba4b73a", "parents": [ "23637d477c1f53acbb176a02c241d60a25888fae" ], "author": { "name": "Don Zickus", "email": "dzickus@redhat.com", "time": "Fri May 14 11:11:21 2010 -0400" }, "committer": { "name": "Frederic Weisbecker", "email": "fweisbec@gmail.com", "time": "Sun May 16 04:25:14 2010 +0200" }, "message": "lockup_detector: Cross arch compile fixes\n\nCombining the softlockup and hardlockup code causes watchdog.c\nto build even without the hardlockup detection support.\n\nSo if an arch, that has the previous and the new nmi watchdog\nimplementations cohabiting, wants to know if the generic one\nis in use, CONFIG_LOCKUP_DETECTOR is not a reliable check.\nWe need to use CONFIG_HARDLOCKUP_DETECTOR instead.\n\nFixes:\n\tkernel/built-in.o: In function `touch_nmi_watchdog\u0027:\n\t(.text+0x449bc): multiple definition of `touch_nmi_watchdog\u0027\n\tarch/sparc/kernel/built-in.o:(.text+0x11b28): first defined here\n\nSigned-off-by: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Ingo Molnar \u003cmingo@elte.hu\u003e\nCc: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nCc: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Cyrill Gorcunov \u003cgorcunov@gmail.com\u003e\nLKML-Reference: \u003c20100514151121.GR15159@redhat.com\u003e\n[ use CONFIG_HARDLOCKUP_DETECTOR instead of CONFIG_PERF_EVENTS_NMI]\nSigned-off-by: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\n" }, { "commit": "23637d477c1f53acbb176a02c241d60a25888fae", "tree": "6d2e9e4a9f6fa2428e0ace29cf764212b823598a", "parents": [ "c01d4323309a90a298fd81cf3a059ee1b12be2e9" ], "author": { "name": "Frederic Weisbecker", "email": "fweisbec@gmail.com", "time": "Sat May 15 23:15:20 2010 +0200" }, "committer": { "name": "Frederic Weisbecker", "email": "fweisbec@gmail.com", "time": "Sun May 16 01:57:42 2010 +0200" }, "message": "lockup_detector: Introduce CONFIG_HARDLOCKUP_DETECTOR\n\nThis new config is deemed to simplify even more the lockup detector\ndependencies and can make it easier to bring a smooth sorting\nbetween archs that support the new generic lockup detector and those\nthat still have their own, especially for those that are in the\nmiddle of this migration.\n\nInstead of checking whether we have CONFIG_LOCKUP_DETECTOR +\nCONFIG_PERF_EVENTS_NMI each time an arch wants to know if it needs\nto build its own lockup detector, take a shortcut with this new\nconfig. It is enabled only if the hardlockup detection part of\nthe whole lockup detector is on.\n\nSigned-off-by: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\nCc: Ingo Molnar \u003cmingo@elte.hu\u003e\nCc: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nCc: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Cyrill Gorcunov \u003cgorcunov@gmail.com\u003e\n" }, { "commit": "0167c781907fcdc3e1f144ef5ce31d402c91eb94", "tree": "522c4acbb5a996ae126f89dce128cccb6cb18ea1", "parents": [ "19cc36c0f0457e5c6629ec24036fbbe8255c88ec" ], "author": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Thu May 13 08:53:33 2010 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Thu May 13 08:53:33 2010 +0200" }, "message": "watchdog: Export touch_softlockup_watchdog\n\nThere are modules that rely on it:\n\n ERROR: \"touch_softlockup_watchdog\" [drivers/video/nvidia/nvidiafb.ko] undefined!\n\nCc: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\nCc: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Cyrill Gorcunov \u003cgorcunov@gmail.com\u003e\nLKML-Reference: \u003c1273713674-8434-1-git-send-regression-fweisbec@gmail.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "d7c547335fa6b0090fa09c46ea0e965ac273a27e", "tree": "66d082cb7acbfc59a2af6bf24130d4832da2919e", "parents": [ "10f9014912a2b1cb59c39cdea777e6d9afa8f17e" ], "author": { "name": "Don Zickus", "email": "dzickus@redhat.com", "time": "Fri May 07 17:11:51 2010 -0400" }, "committer": { "name": "Frederic Weisbecker", "email": "fweisbec@gmail.com", "time": "Wed May 12 23:55:55 2010 +0200" }, "message": "lockup_detector: Separate touch_nmi_watchdog code path from touch_watchdog\n\nWhen I combined the nmi_watchdog (hardlockup) and softlockup code, I\nalso combined the paths the touch_watchdog and touch_nmi_watchdog took.\nThis may not be the best idea as pointed out by Frederic W., that the\ntouch_watchdog case probably should not reset the hardlockup count.\n\nTherefore the patch below falls back to the previous idea of keeping\nthe touch_nmi_watchdog a superset of the touch_watchdog case.\n\nSigned-off-by: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Ingo Molnar \u003cmingo@elte.hu\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Cyrill Gorcunov \u003cgorcunov@gmail.com\u003e\nCc: Eric Paris \u003ceparis@redhat.com\u003e\nCc: Randy Dunlap \u003crandy.dunlap@oracle.com\u003e\nLKML-Reference: \u003c1273266711-18706-9-git-send-email-dzickus@redhat.com\u003e\nSigned-off-by: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\n" }, { "commit": "332fbdbca3f7716c5620970755ae054d213bcc4e", "tree": "ce6fe479c9b362fab1ce9530a0f6545d9f79025d", "parents": [ "58687acba59266735adb8ccd9b5b9aa2c7cd205b" ], "author": { "name": "Don Zickus", "email": "dzickus@redhat.com", "time": "Fri May 07 17:11:45 2010 -0400" }, "committer": { "name": "Frederic Weisbecker", "email": "fweisbec@gmail.com", "time": "Wed May 12 23:55:43 2010 +0200" }, "message": "lockup_detector: Touch_softlockup cleanups and softlockup_tick removal\n\nJust some code cleanup to make touch_softlockup clearer and remove the\nsoftlockup_tick function as it is no longer needed.\n\nAlso remove the /proc softlockup_thres call as it has been changed to\nwatchdog_thres.\n\nSigned-off-by: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Ingo Molnar \u003cmingo@elte.hu\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Cyrill Gorcunov \u003cgorcunov@gmail.com\u003e\nCc: Eric Paris \u003ceparis@redhat.com\u003e\nCc: Randy Dunlap \u003crandy.dunlap@oracle.com\u003e\nLKML-Reference: \u003c1273266711-18706-3-git-send-email-dzickus@redhat.com\u003e\nSigned-off-by: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\n" }, { "commit": "58687acba59266735adb8ccd9b5b9aa2c7cd205b", "tree": "7236582375310b116eedec6facbee87d42e3dd6d", "parents": [ "a9aa1d02de36b450990b0e25a88fc2ff1c3e6b94" ], "author": { "name": "Don Zickus", "email": "dzickus@redhat.com", "time": "Fri May 07 17:11:44 2010 -0400" }, "committer": { "name": "Frederic Weisbecker", "email": "fweisbec@gmail.com", "time": "Wed May 12 23:55:33 2010 +0200" }, "message": "lockup_detector: Combine nmi_watchdog and softlockup detector\n\nThe new nmi_watchdog (which uses the perf event subsystem) is very\nsimilar in structure to the softlockup detector. Using Ingo\u0027s\nsuggestion, I combined the two functionalities into one file:\nkernel/watchdog.c.\n\nNow both the nmi_watchdog (or hardlockup detector) and softlockup\ndetector sit on top of the perf event subsystem, which is run every\n60 seconds or so to see if there are any lockups.\n\nTo detect hardlockups, cpus not responding to interrupts, I\nimplemented an hrtimer that runs 5 times for every perf event\noverflow event. If that stops counting on a cpu, then the cpu is\nmost likely in trouble.\n\nTo detect softlockups, tasks not yielding to the scheduler, I used the\nprevious kthread idea that now gets kicked every time the hrtimer fires.\nIf the kthread isn\u0027t being scheduled neither is anyone else and the\nwarning is printed to the console.\n\nI tested this on x86_64 and both the softlockup and hardlockup paths\nwork.\n\nV2:\n- cleaned up the Kconfig and softlockup combination\n- surrounded hardlockup cases with #ifdef CONFIG_PERF_EVENTS_NMI\n- seperated out the softlockup case from perf event subsystem\n- re-arranged the enabling/disabling nmi watchdog from proc space\n- added cpumasks for hardlockup failure cases\n- removed fallback to soft events if no PMU exists for hard events\n\nV3:\n- comment cleanups\n- drop support for older softlockup code\n- per_cpu cleanups\n- completely remove software clock base hardlockup detector\n- use per_cpu masking on hard/soft lockup detection\n- #ifdef cleanups\n- rename config option NMI_WATCHDOG to LOCKUP_DETECTOR\n- documentation additions\n\nV4:\n- documentation fixes\n- convert per_cpu to __get_cpu_var\n- powerpc compile fixes\n\nV5:\n- split apart warn flags for hard and soft lockups\n\nTODO:\n- figure out how to make an arch-agnostic clock2cycles call\n (if possible) to feed into perf events as a sample period\n\n[fweisbec: merged conflict patch]\n\nSigned-off-by: Don Zickus \u003cdzickus@redhat.com\u003e\nCc: Ingo Molnar \u003cmingo@elte.hu\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Cyrill Gorcunov \u003cgorcunov@gmail.com\u003e\nCc: Eric Paris \u003ceparis@redhat.com\u003e\nCc: Randy Dunlap \u003crandy.dunlap@oracle.com\u003e\nLKML-Reference: \u003c1273266711-18706-2-git-send-email-dzickus@redhat.com\u003e\nSigned-off-by: Frederic Weisbecker \u003cfweisbec@gmail.com\u003e\n" } ] }