aboutsummaryrefslogtreecommitdiff
path: root/tools
AgeCommit message (Collapse)Author
2009-09-20perf util: SVG performance improvementsArjan van de Ven
Tweak the output SVG to increase performance in SVG viewers by limiting the different types of font sizes and by smarter transformations on the text. At least with Inkscape this gives a notable performance improvement during zoom and scrolling. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090920181438.3a49cb93@linux.intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-20perf util: Make the timechart SVG width dynamicArjan van de Ven
This patch adds a command line option for timechart that allows the user to specify the width of the SVG file. This patch also makes sure that each second of recording has at least 200 units (pixels at 96 DPI) of width. This impacts recordings longer than 5 seconds; recordings shorter than 5 second will scale up to have a width of 1000 units for the whole recording (as before). Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090920181416.69570c5d@linux.intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-20perf timechart: Show the duration of scheduler delays in the SVGArjan van de Ven
Given that scheduler latencies are the hot thing nowadays, show the duration of said latencies in the SVG in text form. In addition, if the latency is more than 10 msec, pick a brighter yellow color as a way to point these long delays out. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090920181353.796f4509@linux.intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-20perf timechart: Show the name of the waker/wakee in timechartArjan van de Ven
Timechart currently shows thin green lines for sending or receiving wakeups. This patch also prints (in a very small font) the name of the process that is being woken/wakes up this process. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090920181328.68baa978@linux.intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf utils: Use a define for the maximum length of a trace eventArjan van de Ven
As per Ingo's review: use a #define rather than an open coded constant for the maximum length of a trace event for storing in the perf.data file. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Cc: fweisbec@gmail.com Cc: peterz@infradead.org Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <20090919133630.10533d3e@infradead.org> [ add a few comments to nearby functions ] Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf: Add timechart help text and add timechart to "perf help"Arjan van de Ven
As suggested by Ingo, add a timechart man page help text, as well as add it to the "perf help" overview. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Cc: fweisbec@gmail.com Cc: peterz@infradead.org Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <20090919133604.3767fa35@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf utils: Be consistent about minimum text size in the svghelperArjan van de Ven
Be more consistent in the svghelper about the minimum text size by having a global #define for this. There needs to be a minimum text size in order to keep the size of the SVG file within the reach of what current SVG viewers can cope with. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Cc: fweisbec@gmail.com Cc: peterz@infradead.org Cc: Paul Mackerras <paulus@samba.org> Cc: Arjan van de Ven <arjan@infradead.org> LKML-Reference: <20090919133507.7374ef8b@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf timechart: Add "perf timechart record"Arjan van de Ven
Add a command line option to record a trace, similar to "perf sched record". Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Cc: fweisbec@gmail.com Cc: peterz@infradead.org Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <20090919133442.0dc2c7f5@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf: Add the timechart toolArjan van de Ven
timechart is a tool to visualize what is going on in the system. The user makes a trace of what is going on with > perf record --timechart /usr/bin/some_command and then can turn the output of this into an svg file > perf timechart which then can be viewed with any SVG view; inkscape works well enough for me. The idea behind timechart is to create a "infinitely zoomable" picture; something that has high level information on a 1:1 zoom level, but which exposes more details every time you zoom into a specific area. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090912130713.6a77bbc0@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf: Add a SVG helper library fileArjan van de Ven
The timechart tool writes out SVG format output; this patch adds a set of helper functions to abstract dealing with SVG from the core timechart code. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090912130613.677f0516@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf: Add a sample_event type to the event_unionArjan van de Ven
Add a sample_event type to the event_union so that raw samples can be processed easily. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090912130511.411434b5@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf: Allow perf utilities to have "callback" options without argumentsArjan van de Ven
timechart needs to add a "callback" type command line argument that does not take arguments. This patch adds the parse-options.h infrastructure to make this possible. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090912130440.548666c1@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf: Store trace event name/id pairs in perf.dataArjan van de Ven
The trace event name<->id mapping is dynamic for each kernel compile. In order for perf.data to be useable outside the actual system, we thus need to store a table of this mapping for later use. This patch adds this table to perf.data, and provides helper functions for lookup up fields from this table. To avoid mistakes, lookup-from-table is kept completely seprate from lookup-from-local-debugfs. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090912130405.6960d099@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19perf: Add a timestamp to fork eventsArjan van de Ven
perf timechart needs to know when a process forked, in order to be able to visualize properly when tasks start. This patch adds a time field to the event structure, and fills it in appropriately. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20090912130341.51ad2de2@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-19Merge branch 'linus' into perfcounters/coreIngo Molnar
Merge reason: Bring in tracing changes we depend on. Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-18perf sched: Add --input=file option to builtin-sched.cMike Galbraith
perf sched record passes unparsed args on to perf record, so specifying an output file via perf sched record -o FILE (cmd) just works. Ergo, provide an option to specify input file as well. Also add the missing 'map' command to help. Signed-off-by: Mike Galbraith <efault@gmx.de> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1253254944.20589.11.camel@marge.simson.net> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-18perf trace: Sample timestamp and cpu when using record flagLi Zefan
Sample timestamp and cpu just like the -R option. Before: init-0 [-01] 1266874889.17179869184709551615: irq_handler_entry: irq=18 handler=eth0 init-0 [-01] 1266874889.17179869184709551615: irq_handler_entry: irq=18 handler=eth0 init-0 [-01] 1266874889.17179869184709551615: irq_handler_entry: irq=1 handler=i8042 init-0 [-01] 1266874889.17179869184709551615: irq_handler_entry: irq=18 handler=eth0 init-0 [-01] 1266874889.17179869184709551615: irq_handler_entry: irq=1 handler=i8042 After: init-0 [001] 7364.568965353: irq_handler_entry: irq=18 handler=eth0 init-0 [001] 7365.530226877: irq_handler_entry: irq=1 handler=i8042 init-0 [001] 7365.542831563: irq_handler_entry: irq=18 handler=eth0 init-0 [001] 7365.644156299: irq_handler_entry: irq=18 handler=eth0 init-0 [001] 7365.694556201: irq_handler_entry: irq=18 handler=eth0 Signed-off-by: Li Zefan <lizf@cn.fujitsu.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> LKML-Reference: <4AB1F827.8040905@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-18perf tools: Increase MAX_EVENT_LENGTHLi Zefan
The name length of some trace events is longer than 30, like sys_enter_sched_get_priority_max and ext4_mb_discard_preallocations. Passing those events to perf-record will fail, try: # ./perf record -f -e syscalls:sys_enter_sched_get_priority_max -F 1 -a Signed-off-by: Li Zefan <lizf@cn.fujitsu.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> LKML-Reference: <4AB1F4AB.7050205@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-18perf tools: Fix memory leak in read_ftrace_printk()Li Zefan
get_tracing_file() should be paired with put_tracing_file(). Signed-off-by: Li Zefan <lizf@cn.fujitsu.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> LKML-Reference: <4AB1F48F.4070807@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-17perf sched: Determine the number of CPUs automaticallyIngo Molnar
For 'perf sched map' output, determine max_cpu automatically, instead of the static default of 15. [ v2: use sysconf() pointed out by Arjan van de Ven <arjan@infradead.org> ] Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-17perf record: Disable profiling before draining the bufferPeter Zijlstra
I noticed that perf-record continues profiling itself after the child terminated and we're draining the buffer. This can cause a _lot_ of overhead with --all recording - we keep and keep recording, which produces new and new events. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-16perf sched: Add 'perf sched map' scheduling event map printoutIngo Molnar
This prints a textual context-switching outline of workload captured via perf sched record. For example, on a 16 CPU box it outputs: N1 O1 . . . S1 . . . B0 . *I0 C1 . M1 . 23002.773423 secs N1 O1 . *Q0 . S1 . . . B0 . I0 C1 . M1 . 23002.773423 secs N1 O1 . Q0 . S1 . . . B0 . *R1 C1 . M1 . 23002.773485 secs N1 O1 . Q0 . S1 . *S0 . B0 . R1 C1 . M1 . 23002.773478 secs *L0 O1 . Q0 . S1 . S0 . B0 . R1 C1 . M1 . 23002.773523 secs L0 O1 . *. . S1 . S0 . B0 . R1 C1 . M1 . 23002.773531 secs L0 O1 . . . S1 . S0 . B0 . R1 C1 *T1 M1 . 23002.773547 secs T1 => irqbalance:2089 L0 O1 . . . S1 . S0 . *P0 . R1 C1 T1 M1 . 23002.773549 secs *N1 O1 . . . S1 . S0 . P0 . R1 C1 T1 M1 . 23002.773566 secs N1 O1 . . . *J0 . S0 . P0 . R1 C1 T1 M1 . 23002.773571 secs N1 O1 . . . J0 . S0 *B0 P0 . R1 C1 T1 M1 . 23002.773592 secs N1 O1 . . . J0 . *U0 B0 P0 . R1 C1 T1 M1 . 23002.773582 secs N1 O1 . . . *S1 . U0 B0 P0 . R1 C1 T1 M1 . 23002.773604 secs N1 O1 . . . S1 . U0 B0 *. . R1 C1 T1 M1 . 23002.773615 secs N1 O1 . . . S1 . U0 B0 . . *K0 C1 T1 M1 . 23002.773631 secs N1 O1 . *M0 . S1 . U0 B0 . . K0 C1 T1 M1 . 23002.773624 secs N1 O1 . M0 . S1 . U0 *. . . K0 C1 T1 M1 . 23002.773644 secs N1 O1 . M0 . S1 . U0 . . . *R1 C1 T1 M1 . 23002.773662 secs N1 O1 . M0 . S1 . *. . . . R1 C1 T1 M1 . 23002.773648 secs N1 O1 . *. . S1 . . . . . R1 C1 T1 M1 . 23002.773680 secs N1 O1 . . . *L0 . . . . . R1 C1 T1 M1 . 23002.773717 secs *N0 O1 . . . L0 . . . . . R1 C1 T1 M1 . 23002.773709 secs *N1 O1 . . . L0 . . . . . R1 C1 T1 M1 . 23002.773747 secs Columns stand for individual CPUs, from CPU0 to CPU15, and the two-letter shortcuts stand for tasks that are running on a CPU. '*' denotes the CPU that had the event. A dot signals an idle CPU. New tasks are assigned new two-letter shortcuts - when they occur first they are printed. In the above example 'T1' stood for irqbalance: T1 => irqbalance:2089 Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-16perf sched: Make idle thread and comm/pid names more consistentIngo Molnar
Peter noticed that we have 3 ways of referring to the idle thread: [idle]:0 swapper:0 swapper-0 Standardize on 'swapper:0'. Reported-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-16perf sched: Sanity check context switch eventsIngo Molnar
Use 'perf sched latency' to track the current task based on context-switch events, and flag the cases where there's some impossible transition: such as a PID being switched out that was not switched in. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-16perf sched: Account for lost events, increase default bufferingIngo Molnar
Output such lost event and state machine weirdness stats: TOTAL: | 14974.910 ms | 46384 | --------------------------------------------------- INFO: 8.865% lost events (19132 out of 215819, in 8 chunks) INFO: 0.198% state machine bugs (49 out of 24708) (due to lost events?) And increase buffering to -m 1024 (4 MB) by default. Since we use output multiplexing that kind of space is needed. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-14perf sched: Add support for sched:sched_stat_runtime eventsmingo
This allows more precise 'perf sched latency' output: --------------------------------------------------------------------------------------- Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | --------------------------------------------------------------------------------------- ksoftirqd/0-4 | 0.010 ms | 2 | avg: 2.476 ms | max: 2.977 ms | perf-12328 | 15.844 ms | 66 | avg: 1.118 ms | max: 9.979 ms | bdi-default-235 | 0.009 ms | 1 | avg: 0.998 ms | max: 0.998 ms | events/1-8 | 0.020 ms | 2 | avg: 0.998 ms | max: 0.998 ms | events/0-7 | 0.018 ms | 2 | avg: 0.992 ms | max: 0.996 ms | sleep-12329 | 0.742 ms | 3 | avg: 0.906 ms | max: 2.289 ms | sshd-12122 | 0.163 ms | 2 | avg: 0.283 ms | max: 0.562 ms | loop-getpid-lon-12322 | 1023.636 ms | 69 | avg: 0.208 ms | max: 5.996 ms | loop-getpid-lon-12321 | 1038.638 ms | 5 | avg: 0.073 ms | max: 0.171 ms | migration/1-5 | 0.000 ms | 1 | avg: 0.006 ms | max: 0.006 ms | --------------------------------------------------------------------------------------- TOTAL: | 2079.078 ms | 153 | ------------------------------------------------- Also, streamline the code a bit more, add asserts for various state machine failures (they should be debugged if they occur) and fix a few odd ends. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-14perf sched: Print PIDs toomingo
Often it's useful to know the PID of the task as well - print it out too. ( While at it, reformat the output to be a bit more paste-into-commit-logs friendly. ) Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-14perf sched: Fix 'perf sched latency' output on 32-bit systemsIngo Molnar
Before: ----------------------------------------------------------------------------------- Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | ----------------------------------------------------------------------------------- perf |4853313.251 ms | 10 | avg: 0.046 ms | max: 0.337 ms | flush-8:0 |2426659.202 ms | 5 | avg: 0.015 ms | max: 0.016 ms | sleep |485331.966 ms | 1 | avg: 0.012 ms | max: 0.012 ms | ksoftirqd/1 |485331.320 ms | 1 | avg: 0.005 ms | max: 0.005 ms | ----------------------------------------------------------------------------------- TOTAL: |8250635.739 ms | 17 | --------------------------------------------- After: ----------------------------------------------------------------------------------- Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | ----------------------------------------------------------------------------------- perf | 0.206 ms | 10 | avg: 0.046 ms | max: 0.337 ms | flush-8:0 | 2.680 ms | 5 | avg: 0.015 ms | max: 0.016 ms | sleep | 0.662 ms | 1 | avg: 0.012 ms | max: 0.012 ms | ksoftirqd/1 | 0.015 ms | 1 | avg: 0.005 ms | max: 0.005 ms | ----------------------------------------------------------------------------------- TOTAL: | 3.563 ms | 17 | --------------------------------------------- Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-14perf tools: Implement counter output multiplexingIngo Molnar
Finish the -M/--multiplex option implementation: - separate it out from group_fd - correctly set it via the ioctl and dont mmap counters that are multiplexed - modify the perf record event loop to deal with buffer-less counters. - remove the -g option from perf sched record - account for unordered events in perf sched latency - (add -f to perf sched record to ease measurements) - skip idle threads (pid==0) in latency output The result is better latency output by 'perf sched latency': ----------------------------------------------------------------------------------- Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | ----------------------------------------------------------------------------------- ksoftirqd/8 | 0.071 ms | 2 | avg: 0.458 ms | max: 0.913 ms | at-spi-registry | 0.609 ms | 19 | avg: 0.013 ms | max: 0.023 ms | perf | 3.316 ms | 16 | avg: 0.013 ms | max: 0.054 ms | Xorg | 0.392 ms | 19 | avg: 0.011 ms | max: 0.018 ms | sleep | 0.537 ms | 2 | avg: 0.009 ms | max: 0.009 ms | ----------------------------------------------------------------------------------- TOTAL: | 4.925 ms | 58 | --------------------------------------------- Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-14perf tools: Fix processing of randomly serialized sched tracesFrederic Weisbecker
Currently it's possible to meet such too high latency results with 'perf sched latency'. ----------------------------------------------------------------------------------- Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | ----------------------------------------------------------------------------------- xfce4-panel | 0.222 ms | 2 | avg: 4718.345 ms | max: 9436.493 ms | scsi_eh_3 | 3.962 ms | 36 | avg: 55.957 ms | max: 1977.829 ms | The origin is on traces that are sometimes badly serialized across cpus. For example the raw traces that raised such results for xfce4-panel: (1) [init]-0 [000] 1494.663899990: sched_switch: task swapper:0 [140] (R) ==> xfce4-panel:4569 [120] (2) xfce4-panel-4569 [000] 1494.663928373: sched_switch: task xfce4-panel:4569 [120] (S) ==> swapper:0 [140] (3) Xorg-4276 [001] 1494.663860125: sched_wakeup: task xfce4-panel:4569 [120] success=1 [000] (4) Xorg-4276 [001] 1504.098252756: sched_wakeup: task xfce4-panel:4569 [120] success=1 [000] (5) perf-5219 [000] 1504.100353302: sched_switch: task perf:5219 [120] (S) ==> xfce4-panel:4569 [120] The traces are processed in the order they arrive. Then in (2), xfce4-panel sleeps, it is first waken up in (3) and eventually scheduled in (5). The latency reported is then 1504 - 1495 = 9 secs, as reported by perf sched. But this is wrong, we are confident in the fact the traces are nicely serialized while we should actually more trust the timestamps. If we reorder by timestamps we get: (1) Xorg-4276 [001] 1494.663860125: sched_wakeup: task xfce4-panel:4569 [120] success=1 [000] (2) [init]-0 [000] 1494.663899990: sched_switch: task swapper:0 [140] (R) ==> xfce4-panel:4569 [120] (3) xfce4-panel-4569 [000] 1494.663928373: sched_switch: task xfce4-panel:4569 [120] (S) ==> swapper:0 [140] (4) Xorg-4276 [001] 1504.098252756: sched_wakeup: task xfce4-panel:4569 [120] success=1 [000] (5) perf-5219 [000] 1504.100353302: sched_switch: task perf:5219 [120] (S) ==> xfce4-panel:4569 [120] Now the trace make more sense, xfce4-panel is sleeping. Then it is woken up in (1), scheduled in (2) It goes to sleep in (3), woken up in (4) and scheduled in (5). Now, latency captured between (1) and (2) is of 39 us. And between (4) and (5) it is 2.1 ms. Such pattern of bad serializing is the origin of the high latencies reported by perf sched. Basically, we need to check whether wake up time is higher than schedule out time. If it's not the case, we need to tag the current work atom as invalid. Beside that, we may need to work later on a better ordering of the traces given by the kernel. After this patch: xfce4-session | 0.221 ms | 1 | avg: 0.538 ms | max: 0.538 ms | Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-14perf tools: Add an option to multiplex counters in a single channelFrederic Weisbecker
Add an option to multiplex counters output in the channel of the group leader, ie: the first counter opened: -M --multiplex The effect is better serialized samples. This is especially useful for tracepoint samples that need to be well serialized for their post-processing. Also make use of this option in 'perf sched'. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Add 'perf sched trace', improve documentationIngo Molnar
Alias 'perf sched trace' to 'perf trace', for workflow completeness. Add a bit of documentation for perf sched. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Implement the 'perf sched record' subcommandIngo Molnar
Implement the 'perf sched record' subcommand that adds a default list of events, turns on raw sampling and system-wide tracing and passes off the rest of the command to perf record. This is more convenient than having to specify the events all the time. Before: $ perf record -a -R -e sched:sched_switch:r -e sched:sched_stat_wait:r -e sched:sched_stat_sleep:r -e sched:sched_stat_iowait:r -e sched:sched_process_exit:r -e sched:sched_process_fork:r -e sched:sched_wakeup:r -e sched:sched_migrate_task:r -c 1 sleep 1 After: $ perf sched record -f sleep 1 Also fix an assumption in the event string parser that assumed that strings passed in can be modified. (In this case they wont be as they come from a readonly constant section.) Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Clean up PID sorting logicIngo Molnar
Use a sort list for thread atoms insertion as well - instead of hardcoded for PID. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Finish latency => atom rename and misc cleanupsIngo Molnar
- Rename 'latency' field/variable names to the better 'atom' ones - Reduce the number of #include lines and consolidate them - Gather file scope variables at the top of the file - Remove unused bits No change in functionality. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Add 'perf sched latency' and 'perf sched replay'Ingo Molnar
Separate the option parsing cleanly and add two variants: - 'perf sched latency' (can be abbreviated via 'perf sched lat') - 'perf sched replay' (can be abbreviated via 'perf sched rep') Also add a repeat count option to replay and add a separation set of options for replay. Do the sorting setup only in the latency sub-command. Display separate help screens for 'perf sched' and 'perf sched replay -h' - i.e. further separation of the sub-commands. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Implement multidimensional sortingFrederic Weisbecker
Implement multidimensional sorting on perf sched so that you can sort either by number of switches, latency average, latency maximum, runtime. perf sched -l -s avg,max (this is the default) ----------------------------------------------------------------------------------- Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | ----------------------------------------------------------------------------------- gnome-power-man | 0.113 ms | 1 | avg: 4998.531 ms | max: 4998.531 ms | xfdesktop | 1.190 ms | 7 | avg: 136.475 ms | max: 940.933 ms | xfce-mcs-manage | 2.194 ms | 22 | avg: 38.534 ms | max: 735.174 ms | notification-da | 2.749 ms | 31 | avg: 27.436 ms | max: 731.791 ms | xfce4-session | 3.343 ms | 28 | avg: 26.796 ms | max: 734.891 ms | xfwm4 | 3.159 ms | 22 | avg: 12.406 ms | max: 241.333 ms | xchat | 42.789 ms | 214 | avg: 11.886 ms | max: 100.349 ms | xfce4-terminal | 5.386 ms | 22 | avg: 11.414 ms | max: 241.611 ms | firefox | 151.992 ms | 123 | avg: 9.543 ms | max: 153.717 ms | xfce4-panel | 24.324 ms | 47 | avg: 8.189 ms | max: 242.352 ms | :5090 | 6.932 ms | 111 | avg: 8.131 ms | max: 102.665 ms | events/0 | 0.758 ms | 12 | avg: 1.964 ms | max: 21.879 ms | Xorg | 280.558 ms | 340 | avg: 1.864 ms | max: 99.526 ms | geany | 63.391 ms | 295 | avg: 1.099 ms | max: 9.334 ms | reiserfs/0 | 0.039 ms | 2 | avg: 0.854 ms | max: 1.487 ms | kondemand/0 | 8.251 ms | 245 | avg: 0.691 ms | max: 34.372 ms | Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Fix nsec to msec conversionFrederic Weisbecker
We are dividing a time in ns by 1e9. This is a nsec to sec conversion. What we want is msecs. Fix it by dividing by 1e6. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Export the total, max latency and total runtime to thread atoms listFrederic Weisbecker
Add a field in the thread atom list that keeps track of the total and max latencies and also the total runtime. This makes a faster output and also prepares for sorting. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Add involuntarily sleeping task in work atomsFrederic Weisbecker
Currently in perf sched, we are measuring the scheduler wakeup latencies. Now we also want measure the time a task wait to be scheduled after it gets preempted. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Rename struct lat_snapshot to struct work atomsFrederic Weisbecker
To measures the latencies, we capture the sched atoms data into a specific structure named struct lat_snapshot. As this structure can be used for other purposes of scheduler profiling and mirrors what happens in a thread work atom, lets rename it to struct work_atom and propagate this renaming in other functions and structures names to keep it coherent. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Output runtime and context switch totalsIngo Molnar
After: ----------------------------------------------------------------------------------- Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | ----------------------------------------------------------------------------------- make | 0.678 ms | 13 | avg: 0.018 ms | max: 0.050 ms | gcc | 0.014 ms | 2 | avg: 0.320 ms | max: 0.627 ms | gcc | 0.000 ms | 2 | avg: 0.185 ms | max: 0.369 ms | ... ----------------------------------------------------------------------------------- TOTAL: | 21.316 ms | 63 | --------------------------------------------- Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Add runtime statsIngo Molnar
Extend the latency tracking structure with scheduling atom runtime info - and sum it up during per task display. (Also clean up a few details.) Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Display time in milliseconds, reorganize outputIngo Molnar
After: ----------------------------------------------------------------------------------- Task | runtime ms | switches | average delay ms | maximum delay ms | ----------------------------------------------------------------------------------- migration/0 | 0.000 ms | 1 | avg: 0.047 ms | max: 0.047 ms | ksoftirqd/0 | 0.000 ms | 1 | avg: 0.039 ms | max: 0.039 ms | migration/1 | 0.000 ms | 3 | avg: 0.013 ms | max: 0.016 ms | migration/3 | 0.000 ms | 2 | avg: 0.003 ms | max: 0.004 ms | migration/4 | 0.000 ms | 1 | avg: 0.022 ms | max: 0.022 ms | distccd | 0.000 ms | 1 | avg: 0.004 ms | max: 0.004 ms | distccd | 0.000 ms | 1 | avg: 0.014 ms | max: 0.014 ms | distccd | 0.000 ms | 2 | avg: 0.000 ms | max: 0.000 ms | distccd | 0.000 ms | 2 | avg: 0.012 ms | max: 0.019 ms | distccd | 0.000 ms | 1 | avg: 0.002 ms | max: 0.002 ms | as | 0.000 ms | 2 | avg: 0.019 ms | max: 0.019 ms | as | 0.000 ms | 3 | avg: 0.015 ms | max: 0.017 ms | as | 0.000 ms | 1 | avg: 0.009 ms | max: 0.009 ms | perf | 0.000 ms | 1 | avg: 0.001 ms | max: 0.001 ms | gcc | 0.000 ms | 1 | avg: 0.021 ms | max: 0.021 ms | run-mozilla.sh | 0.000 ms | 2 | avg: 0.010 ms | max: 0.017 ms | mozilla-plugin- | 0.000 ms | 1 | avg: 0.006 ms | max: 0.006 ms | gcc | 0.000 ms | 2 | avg: 0.013 ms | max: 0.013 ms | ----------------------------------------------------------------------------------- (The runtime ms column is not filled in yet.) Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Clean up latency and replay sub-commandsIngo Molnar
- Separate the latency and the replay commands more cleanly - Use consistent naming - Display help page on 'perf sched' outlining comments, instead of aborting Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Add sched latency profilingFrederic Weisbecker
Add the -l --latency option that reports statistics about the scheduler latencies. For now, the latencies are measured in the following sequence scope: - task A is sleeping (D or S state) - task B wakes up A ^ | | latency timeframe | | v - task A is scheduled in Start by recording every scheduler events: perf record -e sched:* and then fetch the results: perf sched -l Tasks count total avg max migration/0 2 39849 19924 28826 ksoftirqd/0 7 756383 108054 373014 migration/1 5 45391 9078 10452 ksoftirqd/1 2 399055 199527 359130 events/0 8 4780110 597513 4500250 events/1 9 6353057 705895 2986012 kblockd/0 42 37805097 900121 5077684 The snapshot are in nanoseconds. - Count: number of snapshots taken for the given task - Total: total latencies in nanosec - Avg : average of latency between wake up and sched in - Max : max snapshot latency Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Make it easier to plug in new sub profilersFrederic Weisbecker
Create a sched event structure of handlers in which various sched events reader can plug their own callbacks. This makes easier the addition of new perf sched sub commands. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Fix bad event alignmentFrederic Weisbecker
perf sched raises the following error when it meets a sched switch event: perf: builtin-sched.c:286: register_pid: Assertion `!(pid >= 65536)' failed. Abandon Currently in x86-64, the sched switch events have a hole in the middle of the structure: u16 common_type; u8 common_flags; u8 common_preempt_count; u32 common_pid; u32 common_tgid; char prev_comm[16]; u32 prev_pid; u32 prev_prio; <--- there u64 prev_state; char next_comm[16]; u32 next_pid; u32 next_prio; Gcc inserts a 4 bytes hole there for prev_state to be u64 aligned. And the events are exported to userspace with this hole. But in userspace, from perf sched, we fetch it using a structure that has a new field in the beginning: u32 size. This is because our trace is exported with its size as a field. But now that we have this new field, the hole in the middle disappears because it makes prev_state becoming well aligned. And since we are using a pointer to the raw trace using this struct, instead of reading prev_state, we are reading the hole. We could fix it by keeping the size seperate from the struct but actually there a lot of other potential problems: some fields may be saved as long in a 64 bits system and later read as long in a 32 bits system. Also this direct cast doesn't care about the endianness differences between the host traced machine and the machine in which we do the post processing. So instead of using such dangerous direct casts, fetch the values using the trace parsing API that already takes care of all these problems. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf tools: Allow the specification of all tracepoints at onceFrederic Weisbecker
Currently, when one wants to activate every tracepoint counters of a subsystem from perf record, the current sequence is needed: perf record -e subsys:ev1 -e subsys:ev2 -e subsys:ev3 This may annoy the most patient of us. Now we can just do: perf record -e subsys:* Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13perf sched: Tighten up the codeIngo Molnar
Various small cleanups - removal of debug printks and dead functions, etc. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>