)]}' { "log": [ { "commit": "46ac22bab42cc868b9c1d0e915ddbc8e8065a44d", "tree": "407d6a38b698b0e071f6d32c0244d8af2c618fa1", "parents": [ "2087a1ad822cd3a68b73338457047fcc54da726b" ], "author": { "name": "Ankita Garg", "email": "ankita@in.ibm.com", "time": "Tue Jul 01 14:30:06 2008 +0530" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Fri Jul 04 12:50:23 2008 +0200" }, "message": "sched: fix accounting in task delay accounting \u0026 migration\n\nOn Thu, Jun 19, 2008 at 12:27:14PM +0200, Peter Zijlstra wrote:\n\u003e On Thu, 2008-06-05 at 10:50 +0530, Ankita Garg wrote:\n\u003e\n\u003e \u003e Thanks Peter for the explanation...\n\u003e \u003e\n\u003e \u003e I agree with the above and that is the reason why I did not see weird\n\u003e \u003e values with cpu_time. But, run_delay still would suffer skews as the end\n\u003e \u003e points for delta could be taken on different cpus due to migration (more\n\u003e \u003e so on RT kernel due to the push-pull operations). With the below patch,\n\u003e \u003e I could not reproduce the issue I had seen earlier. After every dequeue,\n\u003e \u003e we take the delta and start wait measurements from zero when moved to a\n\u003e \u003e different rq.\n\u003e\n\u003e OK, so task delay delay accounting is broken because it doesn\u0027t take\n\u003e migration into account.\n\u003e\n\u003e What you\u0027ve done is make it symmetric wrt enqueue, and account it like\n\u003e\n\u003e cpu0 cpu1\n\u003e\n\u003e enqueue\n\u003e \u003cwait-d1\u003e\n\u003e dequeue\n\u003e enqueue\n\u003e \u003cwait-d2\u003e\n\u003e run\n\u003e\n\u003e Where you add both d1 and d2 to the run_delay,.. right?\n\u003e\n\nThanks for reviewing the patch. The above is exactly what I have done.\n\n\u003e This seems like a good fix, however it looks like the patch will break\n\u003e compilation in !CONFIG_SCHEDSTATS \u0026\u0026 !CONFIG_TASK_DELAY_ACCT, of it\n\u003e failing to provide a stub for sched_info_dequeue() in that case.\n\nFixed. Pl. find the new patch below.\n\nSigned-off-by: Ankita Garg \u003cankita@in.ibm.com\u003e\nAcked-by: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Gregory Haskins \u003cghaskins@novell.com\u003e\nCc: rostedt@goodmis.org\nCc: suresh.b.siddha@intel.com\nCc: aneesh.kumar@linux.vnet.ibm.com\nCc: dhaval@linux.vnet.ibm.com\nCc: vatsa@linux.vnet.ibm.com\nCc: David Bahi \u003cDBahi@novell.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "d4abc238c9f4df8b3216f3e883f5d0a07b7ac75a", "tree": "38b571d9a787a62ee88de8f1a2d1c39b3a9d83b8", "parents": [ "d819c49da624e3ee09b2844603d58265039eecdd" ], "author": { "name": "Bharath Ravi", "email": "bharathravi1@gmail.com", "time": "Mon Jun 16 15:11:01 2008 +0530" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Thu Jun 19 14:15:28 2008 +0200" }, "message": "sched, delay accounting: fix incorrect delay time when constantly waiting on runqueue\n\nThis patch corrects the incorrect value of per process run-queue wait\ntime reported by delay statistics. The anomaly was due to the following\nreason. When a process leaves the CPU and immediately starts waiting for\nCPU on the runqueue (which means it remains in the TASK_RUNNABLE state),\nthe time of re-entry into the run-queue is never recorded. Due to this,\nthe waiting time on the runqueue from this point of re-entry upto the\nnext time it hits the CPU is not accounted for. This is solved by\nrecording the time of re-entry of a process leaving the CPU in the\nsched_info_depart() function IF the process will go back to waiting on\nthe run-queue. This IF condition is verified by checking whether the\nprocess is still in the TASK_RUNNABLE state.\n\nThe patch was tested on 2.6.26-rc6 using two simple CPU hog programs.\nThe values noted prior to the fix did not account for the time spent on\nthe runqueue waiting. After the fix, the correct values were reported\nback to user space.\n\nSigned-off-by: Bharath Ravi \u003cbharathravi1@gmail.com\u003e\nSigned-off-by: Madhava K R \u003cmadhavakr@gmail.com\u003e\nCc: dhaval@linux.vnet.ibm.com\nCc: vatsa@in.ibm.com\nCc: balbir@in.ibm.com\nAcked-by: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "c6fba5451a84143f34056a465e72ba187fcc651c", "tree": "b552b99533703a0e63b809f6fa77f81255c384bc", "parents": [ "3f33a7ce9567ded582af1ab71f9802165fe12f09" ], "author": { "name": "Adrian Bunk", "email": "bunk@kernel.org", "time": "Wed May 14 16:22:59 2008 -0700" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Thu May 29 11:25:15 2008 +0200" }, "message": "show_schedstat(): fix memleak\n\nThe Coverity checker spotted a memleak introduced by commit\n39106dcf85285e78f3b290022122c76f851379b8 (cpumask: use new cpus_scnprintf\nfunction).\n\nIt seems the kfree() got lost between v2 and v3 of this patch...\n\nSigned-off-by: Adrian Bunk \u003cbunk@kernel.org\u003e\nCc: Mike Travis \u003ctravis@sgi.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "39106dcf85285e78f3b290022122c76f851379b8", "tree": "7fe93aaf6a433920b1c31725f42db30799deaa55", "parents": [ "fb0f330e62d71f7c535251438068199af320cf73" ], "author": { "name": "Mike Travis", "email": "travis@sgi.com", "time": "Tue Apr 08 11:43:03 2008 -0700" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Sat Apr 19 19:44:59 2008 +0200" }, "message": "cpumask: use new cpus_scnprintf function\n\n * Cleaned up references to cpumask_scnprintf() and added new\n cpulist_scnprintf() interfaces where appropriate.\n\n * Fix some small bugs (or code efficiency improvments) for various uses\n of cpumask_scnprintf.\n\n * Clean up some checkpatch errors.\n\nSigned-off-by: Mike Travis \u003ctravis@sgi.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "f95e0d1c2ad668c77aa4b272c076faf3aa0d631c", "tree": "83751c7a0eb007ad6fb992f3046932ac57039e0b", "parents": [ "c1a89740da168d3431f2f4e7c3b03daacbb55be1" ], "author": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Wed Nov 28 15:52:56 2007 +0100" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Wed Nov 28 15:52:56 2007 +0100" }, "message": "sched: clean up kernel/sched_stat.h\n\nclean up kernel/sched_stat.h.\n\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "9a41785cc43d88397f787a651ed7286a33f8462f", "tree": "ab3efa693c573f2b40dbf4d0e3f219f2dc20da22", "parents": [ "b2be5e96dc0b5a179cf4cb98e65cfb605752ca26" ], "author": { "name": "Balbir Singh", "email": "balbir@linux.vnet.ibm.com", "time": "Fri Nov 09 22:39:37 2007 +0100" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Fri Nov 09 22:39:37 2007 +0100" }, "message": "sched: fix delay accounting regression\n\nFix the delay accounting regression introduced by commit\n75d4ef16a6aa84f708188bada182315f80aab6fa. rq no longer has sched_info\ndata associated with it. task_struct sched_info structure is used by delay\naccounting to provide back statistics to user space.\n\nalso remove direct use of sched_clock() (which is not a valid thing to\ndo anymore) and use rq-\u003eclock instead.\n\nSigned-off-by: Balbir Singh \u003cbalbir@linux.vnet.ibm.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "480b9434c542ddf2833aaed3dabba71bc0b787b5", "tree": "78c2638ac583cc57165ee1393ebbbbbe367f46fb", "parents": [ "cc4ea79588e688ea9b1161650979a194dd709169" ], "author": { "name": "Ken Chen", "email": "kenchen@google.com", "time": "Thu Oct 18 21:32:56 2007 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Thu Oct 18 21:32:56 2007 +0200" }, "message": "sched: reduce schedstat variable overhead a bit\n\nschedstat is useful in investigating CPU scheduler behavior. Ideally,\nI think it is beneficial to have it on all the time. However, the\ncost of turning it on in production system is quite high, largely due\nto number of events it collects and also due to its large memory\nfootprint.\n\nMost of the fields probably don\u0027t need to be full 64-bit on 64-bit\narch. Rolling over 4 billion events will most like take a long time\nand user space tool can be made to accommodate that. I\u0027m proposing\nkernel to cut back most of variable width on 64-bit system. (note,\nthe following patch doesn\u0027t affect 32-bit system).\n\nSigned-off-by: Ken Chen \u003ckenchen@google.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "2d72376b3af1e7d4d4515ebfd0f4383f2e92c343", "tree": "a9f36173883f1309640f3a1e58b03a53422262d8", "parents": [ "2b1e315dd2822c99793485f9e53a73459fb399c1" ], "author": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Mon Oct 15 17:00:12 2007 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Mon Oct 15 17:00:12 2007 +0200" }, "message": "sched: clean up schedstats, cnt -\u003e count\n\nrename all \u0027cnt\u0027 fields and variables to the less yucky \u0027count\u0027 name.\n\nyuckage noticed by Andrew Morton.\n\nno change in code, other than the /proc/sched_debug bkl_count string got\na bit larger:\n\n text data bss dec hex filename\n 38236 3506 24 41766 a326 sched.o.before\n 38240 3506 24 41770 a32a sched.o.after\n\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\nReviewed-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n" }, { "commit": "75d4ef16a6aa84f708188bada182315f80aab6fa", "tree": "580c044338e3525f39038b1fe703c2a6138f9357", "parents": [ "30cfdcfc5f180fc21a3dad6ae3b7b2a9ee112186" ], "author": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Mon Oct 15 17:00:08 2007 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Mon Oct 15 17:00:08 2007 +0200" }, "message": "sched: fix delay accounting performance regression\n\nfix delay accounting performance regression - those sched_clock()\ncalls are not needed.\n\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\nSigned-off-by: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nReviewed-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n" }, { "commit": "c3c7011969274768818842b0a08ec45d88f45b4f", "tree": "9e37522521d347a5c4cc1412007116e0fd138554", "parents": [ "9c2172459a47c99adf9c968180a8a57d9ff84efa" ], "author": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Thu Aug 02 17:41:40 2007 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Thu Aug 02 17:41:40 2007 +0200" }, "message": "[PATCH] sched: add schedstat_set() API\n\nadd the schedstat_set() API, to allow the reduction of\nCONFIG_SCHEDSTAT related #ifdefs. No code changed.\n\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "172ba844a8851c3edd13c0a979cdf46bd5e3cc1a", "tree": "5e1bfd820c8e68fc28450688f166f4136351e1e1", "parents": [ "b27f03d4bdc145a09fb7b0c0e004b29f1ee555fa" ], "author": { "name": "Balbir Singh", "email": "balbir@linux.vnet.ibm.com", "time": "Mon Jul 09 18:52:00 2007 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Mon Jul 09 18:52:00 2007 +0200" }, "message": "sched: update delay-accounting to use CFS\u0027s precise stats\n\nupdate delay-accounting to use CFS\u0027s precise stats.\n\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" }, { "commit": "425e0968a25fa3f111f9919964cac079738140b5", "tree": "0b5165947477446d1a3680bf970689ded93c2931", "parents": [ "1df21055e34b6a68d62cf0c524b9e52deebd7ead" ], "author": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Mon Jul 09 18:51:58 2007 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@elte.hu", "time": "Mon Jul 09 18:51:58 2007 +0200" }, "message": "sched: move code into kernel/sched_stats.h\n\ncreate sched_stats.h and move sched.c schedstats code into it.\nThis cleans up sched.c a bit.\n\nno code changes are caused by this patch.\n\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n" } ] }