- 21 May, 2014 16 commits
-
-
Namhyung Kim authored
Some fields missed to set default column length so it broke align in --stdio output. Add perf_hpp__reset_width() to set it to a sane default value. Note that this change will ignore -w/--column-widths option for now. Before: $ perf report -F cpu,comm,overhead --stdio ... # CPU Command Overhead # ............... ........ # 0 firefox 2.65% 0 kworker/0:0 1.45% 0 swapper 5.52% 0 synergys 0.92% 1 firefox 4.54% After: # CPU Command Overhead # ... ............... ........ # 0 firefox 2.65% 0 kworker/0:0 1.45% 0 swapper 5.52% 0 synergys 0.92% 1 firefox 4.54% Signed-off-by:Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1400480762-22852-17-git-send-email-namhyung@kernel.orgSigned-off-by:
Jiri Olsa <jolsa@kernel.org>
-
Namhyung Kim authored
When it converted sort entries to hpp formats, it missed se->elide handling, so add it for compatibility. Signed-off-by:
-
Namhyung Kim authored
The --fields option is to allow user setup output field in any order. It can receive any sort keys and following (hpp) fields: overhead, overhead_sys, overhead_us, sample and period If guest profiling is enabled, overhead_guest_{sys,us} will be available too. More more information, please see previous patch "perf report: Add -F option to specify output fields" Signed-off-by: -
Namhyung Kim authored
The hists__filter_entries() function is called when down arrow key is pressed for navigating through the entries in TUI. It has a check for filtering out entries that have very small overhead (under min_pcnt). However it just assumed the entries are sorted by the overhead so when it saw such a small overheaded entry, it just stopped navigating as an optimization. But it's not true anymore due to new --fields and --sort optoin behavior and this case users cannot go down to a next entry if ther's an entry with small overhead in-between. Signed-off-by:
-
Namhyung Kim authored
Currently, what the sort_entry does is just identifying hist entries so that they can be grouped properly. However, with -F option support, it indeed needs to sort entries appropriately to be shown to users. So add ->sort() member to do it. Signed-off-by:
Ingo Molnar <mingo@kernel.org> Link: http://lkml.kernel.org/r/1400480762-22852-13-git-send-email-namhyung@kernel.orgSigned-off-by:
-
Namhyung Kim authored
The -F/--fields option is to allow user setup output field in any order. It can receive any sort keys and following (hpp) fields: overhead, overhead_sys, overhead_us, sample and period If guest profiling is enabled, overhead_guest_{sys,us} will be available too. The output fields also affect sort order unless you give -s/--sort option. And any keys specified on -s option, will also be added to the output field list automatically. $ perf report -F sym,sample,overhead ... # Symbol Samples Overhead # .......................... ............ ........ # [.] __cxa_atexit 2 2.50% [.] __libc_csu_init 4 5.00% [.] __new_exitfn 3 3.75% [.] _dl_check_map_versions 1 1.25% [.] _dl_name_match_p 4 5.00% [.] _dl_setup_hash 1 1.25% [.] _dl_sysdep_start 1 1.25% [.] _init 5 6.25% [.] _setjmp 6 7.50% [.] a 8 10.00% [.] b 8 10.00% [.] brk 1 1.25% [.] c 8 10.00% Note that, the example output above is captured after applying next patch which fixes sort/comparing behavior. Requested-by: -
Namhyung Kim authored
So that it can be set properly prior to set up output fields. That makes easy to handle/warn errors during the setup since it doesn't need to be bothered with the GUI. Signed-off-by:
-
Namhyung Kim authored
The perf uses different default sort orders for different use-cases, and this was scattered throughout the code. Add get_default_sort_ order() function to handle this and change initial value of sort_order to NULL to distinguish it from user-given one. Signed-off-by:
-
Namhyung Kim authored
Add overhead{,_sys,_us,_guest_sys,_guest_us}, sample and period sort keys so that they can be selected with --sort/-s option. $ perf report -s period,comm --stdio ... # Overhead Period Command # ........ ............ ............... # 47.06% 152 swapper 13.93% 45 qemu-system-arm 12.38% 40 synergys 3.72% 12 firefox 2.48% 8 xchat Signed-off-by: -
Namhyung Kim authored
The callback was used by TUI for determining color of folded sign using percent of first field/column. But it cannot be used anymore since it now support dynamic reordering of output field. So move the logic to the hist_browser__show_entry(). Signed-off-by:
-
Namhyung Kim authored
Until now the hpp and sort functions do similar jobs different ways. Since the sort functions converted/wrapped to hpp formats it can do the job in a uniform way. The perf_hpp__sort_list has a list of hpp formats to sort entries and the perf_hpp__list has a list of hpp formats to print output result. To have a backward compatibility, it automatically adds 'overhead' field in front of sort list. And then all of fields in sort list added to the output list (if it's not already there). Signed-off-by:
-
Namhyung Kim authored
Convert output sorting function to use ->sort hpp functions. Signed-off-by:
-
Namhyung Kim authored
Move logic of hist_entry__sort_on_period to __hpp__sort() in order to support event group report. Signed-off-by:
-
Namhyung Kim authored
It wrapped sort entries to hpp functions, so using the hpp sort list to sort entries. Signed-off-by:
-
Namhyung Kim authored
This is a preparation of consolidating management of output field and sort keys. Signed-off-by:
-
Namhyung Kim authored
Those function pointers will be used to sort report output based on the selected fields. This is a preparation of later change. Signed-off-by:
-
- 20 May, 2014 1 commit
-
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf into perf/core Pull perf/core improvements and fixes from Jiri Olsa: * Add libdw DWARF post unwind support for ARM (Jean Pihet) * Consolidate types.h for ARM and ARM64 (Jean Pihet) * Fix possible null pointer dereference in session.c (Masanari Iida) * Cleanup, remove unused variables in map_switch_event() (Dongsheng Yang) * Remove nr_state_machine_bugs in perf latency (Dongsheng Yang) * Remove usage of trace_sched_wakeup(.success) (Peter Zijlstra) Signed-off-by:
-
- 19 May, 2014 3 commits
-
-
Stephane Eranian authored
This patch fixes a bug in precise_store_data_hsw() whereby it would set the data source memory level to the wrong value. As per the the SDM Vol 3b Table 18-41 (Layout of Data Linear Address Information in PEBS Record), when status bit 0 is set this is a L1 hit, otherwise this is a L1 miss. This patch encodes the memory level according to the specification. In V2, we added the filtering on the store events. Only the following events produce L1 information: * MEM_UOPS_RETIRED.STLB_MISS_STORES * MEM_UOPS_RETIRED.LOCK_STORES * MEM_UOPS_RETIRED.SPLIT_STORES * MEM_UOPS_RETIRED.ALL_STORES Cc: mingo@elte.hu Cc: acme@ghostprotocols.net Cc: jolsa@redhat.com Cc: jmario@redhat.com Cc: ak@linux.intel.com Tested-and-Reviewed-by:
Don Zickus <dzickus@redhat.com> Signed-off-by:
Stephane Eranian <eranian@google.com> Signed-off-by:
Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20140515155644.GA3884@quadSigned-off-by:
Thomas Gleixner <tglx@linutronix.de>
-
Peter Zijlstra authored
Vince noticed that we test the (unsigned long) flags field against an (unsigned int) constant. This would allow setting the high bits on 64bit platforms and not get an error. There is nothing that uses the high bits, so it should be entirely harmless, but we don't want userspace to accidentally set them anyway, so fix the constants. Cc: Ingo Molnar <mingo@kernel.org> Cc: Thomas Gleixner <tglx@linutronix.de> Reported-by:
Vince Weaver <vincent.weaver@maine.edu> Tested-by:
-
Borislav Petkov authored
... in 3a497f48 ("perf: Simplify perf_event_exit_task_context()") Signed-off-by:
Borislav Petkov <bp@suse.de> Signed-off-by:
-
- 16 May, 2014 7 commits
-
-
Jean Pihet authored
Adding libdw DWARF post unwind support, which is part of elfutils-devel/libdw-dev package from version 0.158. The new code is contained in unwin-libdw.c object, and implements unwind__get_entries unwind interface function. Signed-off-by:
Jean Pihet <jean.pihet@linaro.org> Reviewed-by:
Will Deacon <will.deacon@arm.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1400229672-16104-4-git-send-email-jean.pihet@linaro.orgSigned-off-by:
-
Jean Pihet authored
Adding dwarf unwind test, that setups live machine data over the perf test thread and does the remote unwind. Need to use -fno-optimize-sibling-calls for test compilation, otherwise 'krava_*' function calls are optimized into jumps and omitted from the stack unwind. So far it was enabled only for x86. Signed-off-by:
-
Jean Pihet authored
Introducing perf_regs_load function, which is going to be used for dwarf unwind test in following patches. It takes single argument as a pointer to the regs dump buffer and populates it with current registers values. Signed-off-by:
-
Jean Pihet authored
Prevents a build breakage since commit d944c4ee 'tools: Consolidate types.h' Signed-off-by:
-
Masanari Iida authored
cppcheck detected following warning: [tools/perf/util/session.c:1628] -> [tools/perf/util/session.c:1632]: (warning) Possible null pointer dereference: session - otherwise it is redundant to check it against null. In order to avoide null pointer, check the pointer before use. Signed-off-by:
Masanari Iida <standby24x7@gmail.com> Link: http://lkml.kernel.org/r/1400087618-13628-1-git-send-email-standby24x7@gmail.comSigned-off-by:
-
Dongsheng Yang authored
In map_switch_event(), we don't care the previous process currently, this patch remove the infomation we get but not used. Signed-off-by:
Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Link: http://lkml.kernel.org/r/1400218625-14613-1-git-send-email-yangds.fnst@cn.fujitsu.comSigned-off-by:
-
Dongsheng Yang authored
As we do not use .success in sched_wakeup event any more, then we can not guarantee that the task when wakeup event happen is out of run queue. So the message of nr_state_machine_bugs is not correct. Signed-off-by:
-
- 12 May, 2014 8 commits
-
-
Peter Zijlstra authored
trace_sched_wakeup(.success) is a dead argument and has been for ages, the only reason its still there is because of brain dead software, which apparently includes perf tools There's a few more instances in pearly snake shit, but that's not supported as far as I care anyhow, so let that bitrot. Signed-off-by:
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf into perf/core Pull perf/core improvements and fixes from Jiri Olsa: * Propagate exit status of a command line workload for record command (Namhyung Kim) * Use tid for finding thread (Namhyung Kim) * Clarify the output of perf sched map plus small sched command fixies (Dongsheng Yang) Signed-off-by:
-
Namhyung Kim authored
I believe that passing pid (instead of tid) as the 3rd arg of the machine__find*_thread() was to find a main thread so that it can search proper map group for symbols. However with the map sharing patch applied, it now can do it in any thread. It fixes a bug when each thread has different name, it only reports a main thread for samples in other threads. Cc: Adrian Hunter <adrian.hunter@intel.com> Acked-by:
David Ahern <dsahern@gmail.com> Acked-by:
-
Namhyung Kim authored
The on_exit() function was only used in perf record but it's gone in previous patch. Signed-off-by:
-
Namhyung Kim authored
Currently perf record doesn't propagate the exit status of a workload given by the command line. But sometimes it'd useful if it's propagated so that a monitoring script can handle errors appropriately. To do that, it moves most of logic out of the exit handlers and run them directly in the __cmd_record(). The only thing needs to be done in the handler is propagating terminating signal so that the shell can terminate its loop properly when Ctrl-C was pressed. Also it cleaned up the resource management code in record__exit(). With this change, perf record returns the child exit status in case of normal termination and send signal to itself when terminated by signal. Example run of Stephane's case: $ perf record true && echo yes || echo no [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.013 MB perf.data (~589 samples) ] yes $ perf record false && echo yes || echo no [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.013 MB perf.data (~589 samples) ] no Jiri's case (error in parent): $ perf record -m 10G true && echo yes || echo no rounding mmap pages size to 17179869184 bytes (4194304 pages) failed to mmap with 12 (Cannot allocate memory) no $ ulimit -n 6 $ perf record sleep 1 && echo yes || echo no failed to create 'go' pipe: Too many open files Couldn't run the workload! no And Peter's case (interrupted by signal): $ while :; do perf record sleep 1; done ^C[ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.014 MB perf.data (~593 samples) ] Reported-by:
Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by:
-
Dongsheng authored
In output of perf sched map, any shortname of thread will be explained at the first time when it appear. Example: *A0 228836.978985 secs A0 => perf:23032 *. A0 228836.979016 secs B0 => swapper:0 . *C0 228836.979099 secs C0 => migration/3:22 *A0 . C0 228836.979115 secs A0 . *. 228836.979115 secs But B0, which is explained as swapper:0 did not appear in the left part of output. Instead, we use '.' as the shortname of swapper:0. So the comment of "B0 => swapper:0" is not easy to understand. This patch clarify the output of perf sched map with not allocating one letter-number shortname for swapper:0 and print ". => swapper:0" as the explanation for swapper:0. Example: *A0 228836.978985 secs A0 => perf:23032 * . A0 228836.979016 secs . => swapper:0 . *B0 228836.979099 secs B0 => migration/3:22 *A0 . B0 228836.979115 secs A0 . * . 228836.979115 secs A0 *C0 . 228836.979225 secs C0 => ksoftirqd/2:18 A0 *D0 . 228836.979236 secs D0 => rcu_sched:7 Signed-off-by: -
Dongsheng authored
Currently, TASK_STATE_TO_CHAR_STR in kernel space is already expanded to RSDTtZXxKWP, but it is still RSDTtZX in perf sched tool. This patch update TASK_STATE_TO_CHAR_STR to the new value in kernel space. Signed-off-by:
-
Dongsheng authored
We should record and process sched:sched_wakeup_new event in perf sched tool, but currently, there is the process function for it, without recording it in record subcommand. This patch add -e sched:sched_wakeup_new to perf sched record. Signed-off-by:
-
- 07 May, 2014 5 commits
-
-
Peter Zijlstra authored
Instead of jumping through hoops to make sure to find (and exit) each event, do it the simple straight fwd way. Signed-off-by:
-
Peter Zijlstra authored
Primarily make perf_event_release_kernel() into put_event(), this will allow kernel space to create per-task inherited events, and is safer in general. Also, document the free_event() assumptions. Signed-off-by:
-
Peter Zijlstra authored
Document and validate the locking assumption of event_sched_in(). Signed-off-by:
-
Peter Zijlstra authored
Commit 38b435b1 ("perf: Fix tear-down of inherited group events") states that we need to destroy groups for inherited events, but it doesn't make any sense to not also destroy groups for normal events. And while it usually makes no difference (the normal events won't leak, and its very likely all the group events will die in quick succession) it does make the code more consistent and closes a potential hole for trouble. Signed-off-by:
-
Peter Zijlstra authored
Make sure all events in a group have the same inherit state. It was possible for group leaders to have inherit set while sibling events would not have inherit set. In this case we'd still inherit the siblings, leading to some non-fatal weirdness. Signed-off-by:
-
