- 29 Apr, 2015 35 commits
-
-
Jiri Olsa authored
Allowing flex parser to report back event parsing error, like: $ perf record -e cycles,cache-mises ls event syntax error: '..es,cache-mises' \___ parser error ... Signed-off-by:Jiri Olsa <jolsa@kernel.org> Cc: David Ahern <dsahern@gmail.com> 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/1429729824-13932-3-git-send-email-jolsa@kernel.orgSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Jiri Olsa authored
Adding support to return error information from parse_events function. Following struct will be populated by parse_events function on return: struct parse_events_error { int idx; char *str; char *help; }; where 'idx' is the position in the string where the parsing failed, 'str' contains dynamically allocated error string describing the error and 'help' is optional help string. The change contains reporting function, which currently does not display anything. The code changes to supply error data for specific event types are coming in next patches. However this is what the expected output is: $ sudo perf record -e 'sched:krava' ls event syntax error: 'sched:krava' \___ unknown tracepoint ... $ perf record -e 'cpu/even=0x1/' ls event syntax error: 'cpu/even=0x1/' \___ unknown term valid terms: pc,any,inv,edge,cmask,event,in_tx,ldlat,umask,in_tx_cp,offcore_rsp,config,config1,config2,name,period,branch_type ... $ perf record -e cycles,cache-mises ls event syntax error: '..es,cache-mises' \___ parser error ... The output functions cut the beginning of the event string so the error starts up to 10th character and cut the end of the string of it crosses the terminal width. Signed-off-by: -
Adrian Hunter authored
Add new AUX area member (aux_watermark) of struct perf_event_attr to debug prints and byte swapping. Signed-off-by:
Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1428594864-29309-27-git-send-email-adrian.hunter@intel.comSigned-off-by:
-
Adrian Hunter authored
Instruction tracing will typically have access to information about the instruction being executed for a particular ip sample. Some of that information will be available in the 'flags' member of struct perf_sample. With the addition of transactions events synthesis to Instruction Tracing options, there is a need to be able easily to see the flags because they show whether the ip is at the start, commit or abort of a tranasaction. Consequently add an option to display the flags. The flags are "bcrosyiABEx" which stand for branch, call, return, conditional, system, asynchronous, interrupt, transaction abort, trace begin, trace end, and in transaction, respectively. Example using Intel PT: perf script -fip,time,event,sym,addr,flags ... 1288.721584105: branches:u: bo 401146 main => 401152 main 1288.721584105: transactions: x 0 401164 main 1288.721584105: branches:u: bx 40117c main => 40119b main 1288.721584105: branches:u: box 4011a4 main => 40117e main 1288.721584105: branches:u: bcx 401187 main => 401094 g ... 1288.721591645: branches:u: bx 4010c4 g => 4010cb g 1288.721591645: branches:u: brx 4010cc g => 401189 main 1288.721591645: transactions: 0 4011a6 main 1288.721593199: branches:u: b 4011a9 main => 4011af main 1288.721593199: branches:u: bo 4011bc main => 40113e main 1288.721593199: branches:u: b 401150 main => 40115a main 1288.721593199: transactions: x 0 401164 main 1288.721593199: branches:u: bx 40117c main => 40119b main 1288.721593199: branches:u: box 4011a4 main => 40117e main 1288.721593199: branches:u: bcx 401187 main => 40105e f ... 1288.722284747: branches:u: brx 401093 f => 401189 main 1288.722284747: branches:u: box 4011a4 main => 40117e main 1288.722284747: branches:u: bcx 401187 main => 40105e f 1288.722285883: transactions: bA 0 401071 f 1288.722285883: branches:u: bA 401071 f => 40116a main 1288.722285883: branches:u: bE 40116a main => 0 [unknown] 1288.722297174: branches:u: bB 0 [unknown] => 40116a main ... Signed-off-by:
-
Adrian Hunter authored
Add support for decoding an AUX area assuming it contains instruction tracing data. The AUX area tracing events are stripped and replaced by synthesized events. Signed-off-by:
-
Adrian Hunter authored
New AUX area tracing events must be re-piped by default. Signed-off-by:
-
Adrian Hunter authored
Add support for decoding an AUX area assuming it contains instruction tracing data. Signed-off-by:
-
Adrian Hunter authored
Add a member to struct dso that can be used by Instruction Trace implementations to hold a cache for decoded instructions. Signed-off-by:
-
Adrian Hunter authored
Decoding AUX area data may involve walking object code. Rather than repetitively decoding the same instructions, a cache can be used to cache the results. This patch implements a fairly generic hashtable with a 32-bit key that could be used for other purposes as well. Signed-off-by:
-
Adrian Hunter authored
Provide hooks so that an AUX area decoder can process AUX area tracing events. Signed-off-by:
-
Adrian Hunter authored
In order to process AUX area tracing data in time order, the queue with data with the lowest timestamp must be processed first. Provide a heap to keep track of which queue that is. As with the queues, a decoder does not have to use the heap, but Intel BTS and Intel PT will use it. Signed-off-by:
-
Adrian Hunter authored
Provide functions to queue AUX area tracing data buffers for processing. A AUX area decoder need not use the queues, however Intel BTS and Intel PT will use them. There is one queue for each of the mmap buffers that were used for recording. Because those mmaps were associated with per-cpu or per-thread contexts, the data is time-ordered with respect to those contexts. Signed-off-by:
-
Adrian Hunter authored
Add functions to synthesize, count and print AUX area tracing error events. Signed-off-by:
-
Adrian Hunter authored
It is assumed that AUX area decoding will synthesize events for consumption by other tools. At this time, the main use of AUX area tracing will be to capture instruction trace (aka processor trace) data. The nature of instruction tracing suggests the initial inclusion of options for "instructions" and "branches" events, but more could be added as needed. Signed-off-by:
-
Adrian Hunter authored
Hook into session processing so that AUX area decoding can synthesize events transparently to the tools. The advantages of transparent decoding are that tools can be used directly with perf.data files containing AUX area tracing data, which is easier for the user and more efficient than having a separate decoding tool. This will work as follows: 1. Tools will feed auxtrace events to the decoder using perf_tool->auxtrace() (support for that still to come). 2. The decoder can process side-band events as needed due to the auxtrace->process_event() hook. 3. The decoder can deliver synthesized events into the event stream using perf_session__deliver_synth_event(). Note the expectation is that decoding will work on data that is time-ordered with respect to the per-cpu or per-thread contexts that were recorded. Signed-off-by:
-
Adrian Hunter authored
Errors encountered when decoding an AUX area trace need to be reported to the user. However the "user" might be a script or another tool, so provide a new user event to capture those errors. Signed-off-by:
-
Adrian Hunter authored
Extend the -m option so that the number of mmap pages for AUX area tracing can be specified by adding a comma followed by the number of pages. Signed-off-by:
-
Adrian Hunter authored
Amend the perf record tool to read the AUX area tracing mmap and synthesize AUX area tracing events. Signed-off-by:
-
Adrian Hunter authored
Add support for reading from the AUX area tracing mmap and synthesizing AUX area tracing events. This patch introduces an abstraction for recording AUX area data. Recording is initialized by auxtrace_record__init() which is a weak function to be implemented by the architecture to provide recording callbacks. Recording is mainly handled by auxtrace_mmap__read() and perf_event__synthesize_auxtrace() but there are callbacks for miscellaneous needs including validating and processing user options, populating private data in auxtrace_info_event, and freeing the structure when finished. Signed-off-by:
-
Adrian Hunter authored
Add two user events for AUX area tracing. PERF_RECORD_AUXTRACE_INFO contains metadata, consisting primarily the type of the AUX area tracing data plus some amount of architecture-specific information. There should be only one PERF_RECORD_AUXTRACE_INFO event. PERF_RECORD_AUXTRACE identifies AUX area tracing data copied from the mmapped AUX area tracing region. The actual data is not part of the event but immediately follows it. Signed-off-by:
-
Adrian Hunter authored
This patch supports the addition to the kernel of AUX area buffers that can be mmapped separately from the perf-events buffer. The AUX buffer can be configured to contain hardware-produced trace information. The first implementation will support Intel BTS and Intel PT. One auxtrace buffer is mmapped per perf-events buffer. If the requested auxtrace buffer size is zero, which it will be until further support is added, then no auxtrace mmapping is attempted. Signed-off-by:
-
Adrian Hunter authored
Add a feature to indicate that a perf.data file contains AUX area data. Signed-off-by:
-
Wang Nan authored
When converting int values, perf first extractes it to a ulonglong, then feeds it to babeltrace as a signed value. For negative 32 bit values (for example, return values of failed syscalls), the extracted data should be something like 0xfffffffe (-2). It becomes a large int64 value. Babeltrace denies to insert it with bt_ctf_field_signed_integer_set_value() because it is larger than 0x7fffffff, the largest positive value a 32 bit int can be. This patch introduces adjust_signedness(), which fills high bits of ulonglong with 1 if the value is negative. Signed-off-by:
Wang Nan <wangnan0@huawei.com> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: He Kuang <hekuang@huawei.com> Cc: Jeremie Galarneau <jgalar@efficios.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Tom Zanussi <tzanussi@gmail.com> Link: http://lkml.kernel.org/r/1429372220-6406-8-git-send-email-jolsa@kernel.orgSigned-off-by:
-
Wang Nan authored
Some parameters of syscall tracepoints named as 'nr', 'event', etc. When dealing with them, perf convert to ctf meets some problem: 1. If a parameter with name 'nr', it will duplicate syscall's common field 'nr'. One such syscall is io_submit(). 2. If a parameter with name 'event', it is denied to be inserted because 'event' is a CTF spec keyword[1]. One such syscall is epoll_ctl. This patch appends '_dupl_X' suffix to avoid problem 1, prepend a '_' prefix to avoid problem 2. [1] http://diamon.org/docs/ctf/v1.8.2/#specC.1.2Signed-off-by: -
Jiri Olsa authored
Adding support to limit the size of ordered_events queue, so we could control allocation size of perf data files without proper finished round events. Reviewed-by:
David Ahern <dsahern@gmail.com> Signed-off-by:
Namhyung Kim <namhyung@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: He Kuang <hekuang@huawei.com> Cc: Jeremie Galarneau <jgalar@efficios.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1429372220-6406-5-git-send-email-jolsa@kernel.orgSigned-off-by:
-
Jiri Olsa authored
For big data files the size of data allocated for stream instance could get really high. It's needed to flush the data out of the stream once in a while. Unfortunately there's no size indication in the stream object, so we govern the flush based on the number of stored events. Current flush limit is set ot 100000 events. Reviewed-by:
-
Sebastian Andrzej Siewior authored
Currently we store the data into single data strea/file. The cpu if data is stored within the event sample. The lttng puts the CPU number that belongs to the event into the packet context instead into the event. This patch makes sure that the trace produce by perf does look the same way. We now use one stream per-CPU. Having it all in one stream increased the total size of the resulting file. The test went from 416KiB (with perf_cpu event member) to 24MiB due to the required (and pointless) flush. With the per-cpu streams the total size went up to 588KiB. Signed-off-by:
Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: He Kuang <hekuang@huawei.com> Cc: Jeremie Galarneau <jgalar@efficios.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1429372220-6406-3-git-send-email-jolsa@kernel.orgSigned-off-by:
-
Jiri Olsa authored
Adding build tests for following make commands: $ make -C <kernelsrc> tools/perf $ make -C <kernelsrc>/tools perf Signed-off-by:
-
Jiri Olsa authored
Several fixes were needed to allow following builds: $ make tools/tmon $ make -C <kernelsrc> tools/perf $ make -C <kernelsrc>/tools perf - some of the tools (perf) use same make variables as in kernel build, unsetting srctree and objtree - using original $(O) for O variable - perf build does not follow the descend function setup invoking it via it's own make rule I tried the rest of the tools/Makefile targets and they seem to work now. Reported-by:
Brian Norris <computersforpeace@gmail.com> Signed-off-by:
-
Jiri Olsa authored
The perf build handles its dependencies by itself. Also renaming libapi libapikfs to libapi as it got changed just recently. Signed-off-by:
-
Jiri Olsa authored
Introduce an 'alias' field to 'struct format_field' to be able to use alternative name for the field. It is initialized with same string pointer as 'name' field. The free logic checks the 'alias' pointer being reset by user and frees it. This will be handy when converting data into CTF, where each field within event needs to have a unique name (while this is not required for tracepoint). Converter can easily assign unique name into the format_field struct. Suggested-by:
Steven Rostedt <rostedt@goodmis.org> Signed-off-by:
-
He Kuang authored
Show message when errors occurred during conversion setup and conversion process. Before this patch: $ ./perf data convert --to-ctf=ctf $ echo $? 255 After this patch: $ ./perf data convert --to-ctf=ctf Error during conversion setup. Signed-off-by:
He Kuang <hekuang@huawei.com> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jeremie Galarneau <jgalar@efficios.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-xvhr1vf7zav9kkeo9w1hv4uk@git.kernel.org Link: http://lkml.kernel.org/r/1429372220-6406-2-git-send-email-jolsa@kernel.orgSigned-off-by:
-
Namhyung Kim authored
The period_ratio_delta, period_ratio and wdiff are never by used at the same time. Instead, Just one of them is accessed according to a comparison method. So make it union to reduce memory footprint. Signed-off-by:
-
Namhyung Kim authored
It's not used anywhere, let's get rid of it. Signed-off-by:
-
David Ahern authored
0d68bc92 breaks compiles on RHEL6/OL6: cc1: warnings being treated as errors builtin-kmem.c: In function ‘search_page_alloc_stat’: builtin-kmem.c:322: error: declaration of ‘stat’ shadows a global declaration node = &parent->rb_left; /usr/include/sys/stat.h:455: error: shadowed declaration is here builtin-kmem.c: In function ‘perf_evsel__process_page_alloc_event’: builtin-kmem.c:378: error: declaration of ‘stat’ shadows a global declaration /usr/include/sys/stat.h:455: error: shadowed declaration is here builtin-kmem.c: In function ‘perf_evsel__process_page_free_event’: builtin-kmem.c:431: error: declaration of ‘stat’ shadows a global declaration /usr/include/sys/stat.h:455: error: shadowed declaration is here Rename local variable to pstat to avoid the name conflict. Signed-off-by:
David Ahern <david.ahern@oracle.com> Link: http://lkml.kernel.org/r/1429033773-31383-1-git-send-email-david.ahern@oracle.comSigned-off-by:
-
- 14 Apr, 2015 1 commit
-
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo-2' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo: New features: - Analyze page allocator events in 'perf kmem' (Namhyung Kim) User visible changes: - Fix retprobe 'perf probe' handling when failing to find needed debuginfo (He Kuang) - lazy_line probe fixes in 'perf probe' (Naohiro Aota, He Kuang) Infrastructure changes: - Record pfn instead of pointer to struct page in tracepoints (Namhyung Kim) Signed-off-by:
Ingo Molnar <mingo@kernel.org>
-
- 13 Apr, 2015 4 commits
-
-
He Kuang authored
The first argument passed to find_probe_point_lazy() should be CU die, which will be passed to die_walk_lines() when lazy_line matches. Currently, when we probe with lazy_line pattern to file without function name, NULL pointer is passed and causes a segment fault. Can be reproduced as following: $ perf probe -k vmlinux --add='fs/super.c;s->s_count=1;' [ 1958.984658] perf[1020]: segfault at 10 ip 00007fc6e10d8c71 sp 00007ffcbfaaf900 error 4 in libdw-0.161.so[7fc6e10ce000+34000] Segmentation fault After this patch: $ perf probe -k vmlinux --add='fs/super.c;s->s_count=1;' Added new event: probe:_stext (on @fs/super.c) You can now use it in all perf tools, such as: perf record -e probe:_stext -aR sleep 1 Signed-off-by:Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1428925290-5623-3-git-send-email-hekuang@huawei.comSigned-off-by:
-
Naohiro Aota authored
If we use lazy matching, it failed to open a souce file if perf command is invoked outside of compilation directory: $ perf probe -a '__schedule;clear_*' Failed to open kernel/sched/core.c: No such file or directory Error: Failed to add events. (-2) OTOH, other commands like "probe -L" can solve the souce directory by themselves. Let's make it possible for lazy matching too! Signed-off-by:
Naohiro Aota <naota@elisp.net> Acked-by:
-
He Kuang authored
When perf probe searched in a debuginfo file and failed, it tried with an alternative, in function get_alternative_probe_event(): memcpy(tmp, &pev->point, sizeof(*tmp)); memset(&pev->point, 0, sizeof(pev->point)); In this case, it drops the retprobe flag and forgets to set it back in find_alternative_probe_point(), so the problem occurs. Can be reproduced as following: $ perf probe -v -k vmlinux --add='sys_write%return' ... Added new event: Writing event: p:probe/sys_write _stext+1584952 probe:sys_write (on sys_write%return) $ cat /sys/kernel/debug/tracing/kprobe_events p:probe/sys_write _stext+1584952 After this patch: $ perf probe -v -k vmlinux --add='sys_write%return' Added new event: Writing event: r:probe/sys_write SyS_write+0 probe:sys_write (on sys_write%return) $ cat /sys/kernel/debug/tracing/kprobe_events r:probe/sys_write SyS_write Signed-off-by: -
Namhyung Kim authored
The perf kmem command records and analyze kernel memory allocation only for SLAB objects. This patch implement a simple page allocator analyzer using kmem:mm_page_alloc and kmem:mm_page_free events. It adds two new options of --slab and --page. The --slab option is for analyzing SLAB allocator and that's what perf kmem currently does. The new --page option enables page allocator events and analyze kernel memory usage in page unit. Currently, 'stat --alloc' subcommand is implemented only. If none of these --slab nor --page is specified, --slab is implied. First run 'perf kmem record' to generate a suitable perf.data file: # perf kmem record --page sleep 5 Then run 'perf kmem stat' to postprocess the perf.data file: # perf kmem stat --page --alloc --line 10 ------------------------------------------------------------------------------- PFN | Total alloc (KB) | Hits | Order | Mig.type | GFP flags ------------------------------------------------------------------------------- 4045014 | 16 | 1 | 2 | RECLAIM | 00285250 4143980 | 16 | 1 | 2 | RECLAIM | 00285250 3938658 | 16 | 1 | 2 | RECLAIM | 00285250 4045400 | 16 | 1 | 2 | RECLAIM | 00285250 3568708 | 16 | 1 | 2 | RECLAIM | 00285250 3729824 | 16 | 1 | 2 | RECLAIM | 00285250 3657210 | 16 | 1 | 2 | RECLAIM | 00285250 4120750 | 16 | 1 | 2 | RECLAIM | 00285250 3678850 | 16 | 1 | 2 | RECLAIM | 00285250 3693874 | 16 | 1 | 2 | RECLAIM | 00285250 ... | ... | ... | ... | ... | ... ------------------------------------------------------------------------------- SUMMARY (page allocator) ======================== Total allocation requests : 44,260 [ 177,256 KB ] Total free requests : 117 [ 468 KB ] Total alloc+freed requests : 49 [ 196 KB ] Total alloc-only requests : 44,211 [ 177,060 KB ] Total free-only requests : 68 [ 272 KB ] Total allocation failures : 0 [ 0 KB ] Order Unmovable Reclaimable Movable Reserved CMA/Isolated ----- ------------ ------------ ------------ ------------ ------------ 0 32 . 44,210 . . 1 . . . . . 2 . 18 . . . 3 . . . . . 4 . . . . . 5 . . . . . 6 . . . . . 7 . . . . . 8 . . . . . 9 . . . . . 10 . . . . . Signed-off-by:
-
