- 27 4月, 2018 12 次提交
-
-
由 Arnaldo Carvalho de Melo 提交于
That returns the a data structure contained the ordered list of kernel modules + the main kernel maps, one more step in removing the MAP__{FUNCTION,VARIABLE} split. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-qsgbxfyaohc80c9ma049dubm@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Takashi Iwai 提交于
The asciidoc package seems behind the recent big wave of python3 conversion, and we were advised to switch to asciidoctor instead. It's almost compatible but some extensions used for perf documentation don't work with it. Here is the patch to cover them, and add the proper support for asciidoctor. Pass USE_ASCIIDOCTOR=yes to make for using asciidoctor instead of asciidoc. The man source and manual attributes are passed via command options. The support for these attributes have been fixed in the latest asciidoctor code. Since asciidoctor can covert to a man page and an HTML directly, we can omit the dependency on xmlto when USE_ASCIIDOCTOR is set. Signed-off-by: NTakashi Iwai <tiwai@suse.de> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180424150456.17353-1-tiwai@suse.deSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
Another step in the road to elliminate the MAP_{FUNCTION,VARIABLE} separation, reducing the exposure to these details in the tools using the symbol APIs. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-8a1hvrqe3r5i0kw865u3uxwt@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
Instead of just returning it in al.sym, allowing for some simplification in its users, and to make it consistent with thread__find_map(). Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-4axi2sigslffdixzxbehvgoj@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
It was returning the searched map just on the addr_location passed, with the function itself returning void. Make it return the map so that we can make the code more compact. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-tzlrrzdeoof4i6ktyqv1t6ks@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
In dc323ce8 ("perf script: Enable printing of branch stack") it first tries to find the map for an address, then the symbol in the DSO backing that map, for that address, well, this is what thread__find_symbol() does, so just use it and make the code shorter. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-03nx3aod955yqnf9l06im28j@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
Out of thread__find_addr_location(..., MAP__FUNCTION, ...), idea here is to continue removing references to MAP__{FUNCTION,VARIABLE} ahead of getting both types of symbols in the same rbtree, as various places do two lookups, looking first at MAP__FUNCTION, then at MAP__VARIABLE. So thread__find_symbol() will eventually do just that, and 'struct symbol' will have the symbol type, for code that cares about that. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-n7528en9e08yd3flzmb26tth@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Hendrik Brueckner 提交于
The output of perf test and perf test list differ because perf test list does not display subtests. Correct this behavior and also let perf test list report subtests. For example: $ ./perf test 2>&1 |wc -l 65 Without this commit: $ ./perf test list 2>&1 |wc -l 57 With this commit: $ ./perf test list 2>&1 |wc -l 65 Signed-off-by: NHendrik Brueckner <brueckner@linux.ibm.com> Reviewed-by: NThomas Richter <tmricht@linux.ibm.com> Acked-by: NJiri Olsa <jolsa@kernel.org> Cc: Michael Petlan <mpetlan@redhat.com> Cc: linux-s390@vger.kernel.org LPU-Reference: 1523605343-11970-1-git-send-email-brueckner@linux.ibm.com Link: https://lkml.kernel.org/n/tip-efb74jw7x2xs2bucp5hf4ilu@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
Out of thread__find_add_map(..., MAP__FUNCTION, ...), idea here is to continue removing references to MAP__{FUNCTION,VARIABLE} ahead of getting both types of symbols in the same rbtree, as various places do two lookups, looking first at MAP__FUNCTION, then at MAP__VARIABLE. So thread__find_map() will eventually do just that, and 'struct symbol' will have the symbol type, for code that cares about that. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-q27xee34l4izpfau49w103s6@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
To further simplify checking if symbols are available for a given map and to reduce the number of users of MAP__{FUNCTION,VARIABLE}. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-iyfoyvbfdti5uehgpjum3qrq@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
To replace longer code sequences in various places. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-tlk3klbkfyjrbfjvryyznfju@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Arnaldo Carvalho de Melo 提交于
Shorter, should be equivalent code, use it. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-q90olng8sfkvrnsrwu7xnul6@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
- 26 4月, 2018 10 次提交
-
-
由 Arnaldo Carvalho de Melo 提交于
Shorter form to figure out if a given map is the kernel one and also reduces the number of code accessing MAP__{FUNCTION,VARIABLE}, that should go away at some point. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-rn8pexelsxpx92ce3elu3wiw@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Jiri Olsa 提交于
Adding support to display visual aid 'length strings' to easily spot the biggest difference in time table. $ perf stat -r 10 --table perf bench sched pipe ... Performance counter stats for './perf bench sched pipe' (5 runs): # Table of individual measurements: 5.189 (-0.293) # 5.189 (-0.294) # 5.186 (-0.296) # 5.663 (+0.181) ## 6.186 (+0.703) #### # Final result: 5.483 +- 0.198 seconds time elapsed ( +- 3.62% ) Suggested-by: NIngo Molnar <mingo@kernel.org> Signed-off-by: NJiri Olsa <jolsa@kernel.org> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180423090823.32309-9-jolsa@kernel.org [ Updated 'perf stat --table' man page entry ] Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Jiri Olsa 提交于
Add --table option to display time for each run (-r option), like: $ perf stat --null -r 5 --table perf bench sched pipe Performance counter stats for './perf bench sched pipe' (5 runs): # Table of individual measurements: 5.379 (-0.176) 5.243 (-0.311) 5.238 (-0.317) 5.536 (-0.019) 6.377 (+0.823) # Final result: 5.555 +- 0.213 seconds time elapsed ( +- 3.83% ) Suggested-by: NIngo Molnar <mingo@kernel.org> Signed-off-by: NJiri Olsa <jolsa@kernel.org> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180423090823.32309-8-jolsa@kernel.org [ Document the new option in 'perf stat's man page ] Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Jiri Olsa 提交于
Ingo suggested to display elapsed time for multirun workload (perf stat -e) with precision based on the precision of the standard deviation. In his own words: > This output is a slightly bit misleading: > Performance counter stats for 'make -j128' (10 runs): > 27.988995256 seconds time elapsed ( +- 0.39% ) > The 9 significant digits in the result, while only 1 is valid, suggests accuracy > where none exists. > It would be better if 'perf stat' would display elapsed time with a precision > adjusted to stddev, it should display at most 2 more significant digits than > the stddev inaccuracy. > I.e. in the above case 0.39% is 0.109, so we only have accuracy for 1 digit, and > so we should only display 3: > 27.988 seconds time elapsed ( +- 0.39% ) Plus a suggestion about the output, which is small enough and connected with the above change that I merged both changes together. > Small output style nit - I think it would be nice if with --repeat the stddev was > also displayed in absolute values, besides percentage: > > 27.988 +- 0.109 seconds time elapsed ( +- 0.39% ) The output is now: Performance counter stats for './perf bench sched pipe' (5 runs): SNIP 13.3667 +- 0.0256 seconds time elapsed ( +- 0.19% ) Suggested-by: NIngo Molnar <mingo@kernel.org> Signed-off-by: NJiri Olsa <jolsa@kernel.org> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180423090823.32309-7-jolsa@kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Jiri Olsa 提交于
Add 'check_2' function to check 2 different files, the 'check' function stays to check files that differs only in the prefix path. In upcoming changes we need to check header files in locations which don't follow the prefix logic. Signed-off-by: NJiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180423090823.32309-3-jolsa@kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Jiri Olsa 提交于
Passing whole string instead of parsing them after. It simplifies things for the next patches, that adds another function call, which makes it hard to pass arguments in the correct shape. Signed-off-by: NJiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180423090823.32309-2-jolsa@kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Ravi Bangoria 提交于
User can remove files from cache using --remove/--purge options but both needs list of files as an argument. It's not convenient when you want to flush out entire cache. Add an option to purge all files from cache. Ex, # perf buildid-cache -l 8a86ef73e44067bca52cc3f6cd3e5446c783391c /tmp/a.out ebe71fdcf4b366518cc154d570a33cd461a51c36 /tmp/a.out.1 # perf buildid-cache -P -v Removing /tmp/a.out (8a86ef73e44067bca52cc3f6cd3e5446c783391c): Ok Removing /tmp/a.out.1 (ebe71fdcf4b366518cc154d570a33cd461a51c36): Ok Purged all: Ok Signed-off-by: NRavi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: NJiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Kate Stewart <kstewart@linuxfoundation.org> Cc: Krister Johansen <kjlx@templeofstupid.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Philippe Ombredanne <pombredanne@nexb.com> Cc: Sihyeon Jang <uneedsihyeon@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/20180417041346.5617-4-ravi.bangoria@linux.vnet.ibm.com [ Initialize 'err' in build_id_cache__purge_all(), to fix build on debian:7, as it can be used uninitialized ] Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Ravi Bangoria 提交于
'perf buildid-cache' allows to add/remove files into cache but there is no option to list all cached files. Add --list option to list all _valid_ cached files. Ex, # perf buildid-cache --add /tmp/a.out # perf buildid-cache -l 8a86ef73e44067bca52cc3f6cd3e5446c783391c /tmp/a.out Signed-off-by: NRavi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: NJiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Kate Stewart <kstewart@linuxfoundation.org> Cc: Krister Johansen <kjlx@templeofstupid.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Philippe Ombredanne <pombredanne@nexb.com> Cc: Sihyeon Jang <uneedsihyeon@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/20180417041346.5617-3-ravi.bangoria@linux.vnet.ibm.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Ingo Molnar 提交于
Merge tag 'perf-urgent-for-mingo-4.17-20180425' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent Pull perf/urgent fixes from Arnaldo Carvalho de Melo: perf stat: - Keep the '/' event modifier separator in fallback, for example when fallbacking from 'cpu/cpu-cycles/' to user level only, where it should become 'cpu/cpu-cycles/u' and not 'cpu/cpu-cycles/:u' (Jiri Olsa) - Fix PMU events parsing rule, improving error reporting for invalid events (Jiri Olsa) - Disable write_backward and other event attributes for !group events in a group, fixing, for instance this group: '{cycles,msr/aperf/}:S' that has leader sampling (:S) and where just the 'cycles', the leader event, should have the write_backward attribute set, in this case it all fails because the PMU where 'msr/aperf/' lives doesn't accepts write_backward style sampling (Jiri Olsa) - Only fall back group read for leader (Kan Liang) - Fix core PMU alias list for x86 platform (Kan Liang) - Print out hint for mixed PMU group error (Kan Liang) - Fix duplicate PMU name for interval print (Kan Liang) Core: - Set main kernel end address properly when reading kernel and module maps (Namhyung Kim) perf mem: - Fix incorrect entries and add missing man options (Sangwon Hong) s/390: - Remove s390 specific strcmp_cpuid_cmp function (Thomas Richter) - Adapt 'perf test' case record+probe_libc_inet_pton.sh for s390 - Fix s390 undefined record__auxtrace_init() return value in 'perf record' (Thomas Richter) Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: NIngo Molnar <mingo@kernel.org>
-
由 Kan Liang 提交于
The SMM freeze feature was introduced since PerfMon V2. But the current code unconditionally enables the feature for all platforms. It can generate #GP exception, if the related FREEZE_WHILE_SMM bit is set for the machine with PerfMon V1. To disable the feature for PerfMon V1, perf needs to - Remove the freeze_on_smi sysfs entry by moving intel_pmu_attrs to intel_pmu, which is only applied to PerfMon V2 and later. - Check the PerfMon version before flipping the SMM bit when starting CPU Fixes: 6089327f ("perf/x86: Add sysfs entry to freeze counters on SMI") Signed-off-by: NKan Liang <kan.liang@linux.intel.com> Signed-off-by: NThomas Gleixner <tglx@linutronix.de> Acked-by: NPeter Zijlstra (Intel) <peterz@infradead.org> Cc: ak@linux.intel.com Cc: eranian@google.com Cc: acme@redhat.com Link: https://lkml.kernel.org/r/1524682637-63219-1-git-send-email-kan.liang@linux.intel.com
-
- 25 4月, 2018 4 次提交
-
-
由 Kan Liang 提交于
PMU name is printed repeatedly for interval print, for example: perf stat --no-merge -e 'unc_m_clockticks' -a -I 1000 # time counts unit events 1.001053069 243,702,144 unc_m_clockticks [uncore_imc_4] 1.001053069 244,268,304 unc_m_clockticks [uncore_imc_2] 1.001053069 244,427,386 unc_m_clockticks [uncore_imc_0] 1.001053069 244,583,760 unc_m_clockticks [uncore_imc_5] 1.001053069 244,738,971 unc_m_clockticks [uncore_imc_3] 1.001053069 244,880,309 unc_m_clockticks [uncore_imc_1] 2.002024821 240,818,200 unc_m_clockticks [uncore_imc_4] [uncore_imc_4] 2.002024821 240,767,812 unc_m_clockticks [uncore_imc_2] [uncore_imc_2] 2.002024821 240,764,215 unc_m_clockticks [uncore_imc_0] [uncore_imc_0] 2.002024821 240,759,504 unc_m_clockticks [uncore_imc_5] [uncore_imc_5] 2.002024821 240,755,992 unc_m_clockticks [uncore_imc_3] [uncore_imc_3] 2.002024821 240,750,403 unc_m_clockticks [uncore_imc_1] [uncore_imc_1] For each print, the PMU name is unconditionally appended to the counter->name. Need to check the counter->name first. If the PMU name is already appended, do nothing. Committer notes: Add and use perf_evsel->uniquified_name bool instead of doing the more expensive strstr(event->name, pmu->name). Signed-off-by: NKan Liang <kan.liang@linux.intel.com> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Agustin Vega-Frias <agustinv@codeaurora.org> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ganapatrao Kulkarni <ganapatrao.kulkarni@cavium.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Shaokun Zhang <zhangshaokun@hisilicon.com> Cc: Will Deacon <will.deacon@arm.com> Fixes: 8c5421c0 ("perf pmu: Display pmu name when printing unmerged events in stat") Link: http://lkml.kernel.org/r/1524594014-79243-5-git-send-email-kan.liang@linux.intel.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Kan Liang 提交于
Perf doesn't support mixed events from different PMUs (except software event) in a group. The perf stat should output <not counted>/<not supported> for all events, but it doesn't. For example, perf stat -e '{cycles,uncore_imc_5/umask=0xF,event=0x4/,instructions}' <not counted> cycles <not supported> uncore_imc_5/umask=0xF,event=0x4/ 1,024,300 instructions If perf fails to open an event, it doesn't error out directly. It will disable some features and retry, until the event is opened or all features are disabled. The disabled features will not be re-enabled. The group read is one of these features. For the example as above, the IMC event and the leader event "cycles" are from different PMUs. Opening the IMC event must fail. The group read feature must be disabled for IMC event and the followed event "instructions". The "instructions" event has the same PMU as the leader "cycles". It can be opened successfully. Since the group read feature has been disabled, the "instructions" event will be read as a single event, which definitely has a value. The group read fallback is still useful for the case which kernel doesn't support group read. It is good enough to be handled only by the leader. For the fallback request from members, it must be caused by an error. The fallback only breaks the semantics of group. Limit the group read fallback only for the leader. Committer testing: On a broadwell t450s notebook: Before: # perf stat -e '{cycles,unc_cbo_cache_lookup.read_i,instructions}' sleep 1 Performance counter stats for 'sleep 1': <not counted> cycles <not supported> unc_cbo_cache_lookup.read_i 818,206 instructions 1.003170887 seconds time elapsed Some events weren't counted. Try disabling the NMI watchdog: echo 0 > /proc/sys/kernel/nmi_watchdog perf stat ... echo 1 > /proc/sys/kernel/nmi_watchdog After: # perf stat -e '{cycles,unc_cbo_cache_lookup.read_i,instructions}' sleep 1 Performance counter stats for 'sleep 1': <not counted> cycles <not supported> unc_cbo_cache_lookup.read_i <not counted> instructions 1.001380511 seconds time elapsed Some events weren't counted. Try disabling the NMI watchdog: echo 0 > /proc/sys/kernel/nmi_watchdog perf stat ... echo 1 > /proc/sys/kernel/nmi_watchdog # Reported-by: NAndi Kleen <ak@linux.intel.com> Signed-off-by: NKan Liang <kan.liang@linux.intel.com> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Agustin Vega-Frias <agustinv@codeaurora.org> Cc: Ganapatrao Kulkarni <ganapatrao.kulkarni@cavium.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Shaokun Zhang <zhangshaokun@hisilicon.com> Cc: Will Deacon <will.deacon@arm.com> Fixes: 82bf311e ("perf stat: Use group read for event groups") Link: http://lkml.kernel.org/r/1524594014-79243-3-git-send-email-kan.liang@linux.intel.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Kan Liang 提交于
Perf doesn't support mixed events from different PMUs (except software event) in a group. For this case, only "<not counted>" or "<not supported>" are printed out. There is no hint which guides users to fix the issue. Checking the PMU type of events to determine if they are from the same PMU. There may be false alarm for the checking. E.g. the core PMU has different PMU type. But it should not happen often. The false alarm can also be tolerated, because: - It only happens on error path. - It just provides a possible solution for the issue. Signed-off-by: NKan Liang <kan.liang@linux.intel.com> Cc: Agustin Vega-Frias <agustinv@codeaurora.org> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ganapatrao Kulkarni <ganapatrao.kulkarni@cavium.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Shaokun Zhang <zhangshaokun@hisilicon.com> Cc: Will Deacon <will.deacon@arm.com> Link: http://lkml.kernel.org/r/1524594014-79243-2-git-send-email-kan.liang@linux.intel.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Kan Liang 提交于
When counting uncore event with alias, core event is mistakenly involved, for example: perf stat --no-merge -e "unc_m_cas_count.all" -C0 sleep 1 Performance counter stats for 'CPU(s) 0': 0 unc_m_cas_count.all [uncore_imc_4] 0 unc_m_cas_count.all [uncore_imc_2] 0 unc_m_cas_count.all [uncore_imc_0] 153,640 unc_m_cas_count.all [cpu] 0 unc_m_cas_count.all [uncore_imc_5] 25,026 unc_m_cas_count.all [uncore_imc_3] 0 unc_m_cas_count.all [uncore_imc_1] 1.001447890 seconds time elapsed The reason is that current implementation doesn't check PMU name of a event when adding its alias into the alias list for core PMU. The uncore event aliases are mistakenly added. This bug was introduced in: commit 14b22ae0 ("perf pmu: Add helper function is_pmu_core to detect PMU CORE devices") Checking the PMU name for all PMUs on X86 and other architectures except ARM. There is no behavior change for ARM. Signed-off-by: NKan Liang <kan.liang@linux.intel.com> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Agustin Vega-Frias <agustinv@codeaurora.org> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ganapatrao Kulkarni <ganapatrao.kulkarni@cavium.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Shaokun Zhang <zhangshaokun@hisilicon.com> Cc: Will Deacon <will.deacon@arm.com> Fixes: 14b22ae0 ("perf pmu: Add helper function is_pmu_core to detect PMU CORE devices") Link: http://lkml.kernel.org/r/1524594014-79243-1-git-send-email-kan.liang@linux.intel.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
- 23 4月, 2018 14 次提交
-
-
由 Thomas Richter 提交于
Command 'perf record' calls: cmd_report() record__auxtrace_init() auxtrace_record__init() On s390 function auxtrace_record__init() returns random return value due to missing initialization. This sometime causes 'perf record' to exit immediately without error message and creating a perf.data file. Fix this by setting error the return code to zero before returning from platform specific functions which may not set the error code in call cases. Signed-off-by: NThomas Richter <tmricht@linux.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Link: http://lkml.kernel.org/r/20180423142940.21143-1-tmricht@linux.ibm.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Sangwon Hong 提交于
Several options were incorrectly described, some lacked describing required arguments while others were simply not documented, fix it. Signed-off-by: NSangwon Hong <qpakzk@gmail.com> Acked-by: NJiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/1524382146-19609-1-git-send-email-qpakzk@gmail.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Jiri Olsa 提交于
.. and other related fields that do not need to be enabled for events that have sampling leader. It fixes the perf top usage Ingo reported broken: # perf top -e '{cycles,msr/aperf/}:S' The 'msr/aperf/' event is configured for write_back sampling, which is not allowed by the MSR PMU, so it fails to create the event. Adjusting related attr test. Reported-by: NIngo Molnar <mingo@kernel.org> Signed-off-by: NJiri Olsa <jolsa@kernel.org> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180423090823.32309-6-jolsa@kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Jiri Olsa 提交于
Currently all the event parsing fails end up in the event_pmu rule, and display misleading help like: $ perf stat -e inst kill event syntax error: 'inst' \___ Cannot find PMU `inst'. Missing kernel support? ... The reason is that the event_pmu is too strong and match also single string. Changing it to force the '/' separators to be part of the rule, and getting the proper error now: $ perf stat -e inst kill event syntax error: 'inst' \___ parser error Run 'perf list' for a list of valid events ... Signed-off-by: NJiri Olsa <jolsa@kernel.org> Reported-by: NIngo Molnar <mingo@kernel.org> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180423090823.32309-5-jolsa@kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Jiri Olsa 提交于
The 'perf stat' fallback for EACCES error sets the exclude_kernel perf_event_attr and tries perf_event_open() again with it. In addition, it also changes the name of the event to reflect that change by adding the 'u' modifier. But it does not take into account the '/' separator, so the event name can end up mangled, like: (note the '/:' characters) $ perf stat -e cpu/cpu-cycles/ kill ... 386,832 cpu/cpu-cycles/:u Adding the code to check on the '/' separator and set the following correct event name: $ perf stat -e cpu/cpu-cycles/ kill ... 388,548 cpu/cpu-cycles/u Signed-off-by: NJiri Olsa <jolsa@kernel.org> Tested-by: NArnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20180423090823.32309-4-jolsa@kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Thomas Richter 提交于
perf test case 58 (record+probe_libc_inet_pton.sh) executed on s390x using kernel 4.16.0rc3 displays this result: # perf trace --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1 probe_libc:inet_pton: (3ffa0240448) __GI___inet_pton (/usr/lib64/libc-2.26.so) gaih_inet (inlined) __GI_getaddrinfo (inlined) main (/usr/bin/ping) __libc_start_main (/usr/lib64/libc-2.26.so) _start (/usr/bin/ping) After I installed kernel 4.16.0 the same tests uses commands: # perf record -e probe_libc:inet_pton/call-graph=dwarf/ -o /tmp/perf.data.abc ping -6 -c 1 ::1 # perf script -i /tmp/perf.data.abc and displays: ping 39048 [006] 84230.381198: probe_libc:inet_pton: (3ffa0240448) 140448 __GI___inet_pton (/usr/lib64/libc-2.26.so) fbde1 gaih_inet (inlined) fe2b9 __GI_getaddrinfo (inlined) 398d main (/usr/bin/ping) Nothing else changed including glibc elfutils and other libraries picked up by the build. The entries for __libc_start_main and _start are missing. I bisected missing __libc_start_main and _start to commit Fixes: 3d20c624 ("perf unwind: Unwind with libdw doesn't take symfs into account") When I undo this commit I get this call stack on s390: [root@s35lp76 perf]# ./perf script -i /tmp/perf.data.abc ping 39048 [006] 84230.381198: probe_libc:inet_pton: (3ffa0240448) 140448 __GI___inet_pton (/usr/lib64/libc-2.26.so) fbde1 gaih_inet (inlined) fe2b9 __GI_getaddrinfo (inlined) 398d main (/usr/bin/ping) 22fbd __libc_start_main (/usr/lib64/libc-2.26.so) 457b _start (/usr/bin/ping) Looks like dwarf functions dwfl_xxx create different call back stack trace when using file /usr/lib/debug/usr/bin/ping-20161105-7.fc27.s390x.debug instead of file /usr/bin/ping. Fix this test case on s390 and do not expect any call back stack entry after the main() function. Also be more robust and accept a leading __GI_ prefix in front of getaddrinfo. On x86 this test case shows the same call stack using both kernel versions 4.16.0rc3 and 4.16.0 and also stops at main: [root@f27 perf]# ./perf script -i /tmp/perf.data.tmr ping 4446 [000] 172.027088: probe_libc:inet_pton: (7fdfa08c93c0) 1393c0 __GI___inet_pton (/usr/lib64/libc-2.26.so) fe60d getaddrinfo (/usr/lib64/libc-2.26.so) 2f40 main (/usr/bin/ping) [root@f27 perf]# Signed-off-by: NThomas Richter <tmricht@linux.ibm.com> Reviewed-by: NHendrik Brueckner <brueckner@linux.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Martin Vuille <jpmv27@aim.com> Link: http://lkml.kernel.org/r/20180423082428.7930-1-tmricht@linux.ibm.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Thomas Richter 提交于
Make the type field in pmu-events/arch/s390/mapfile.cvs more generic to match the created cpuid string for s390. The pattern also checks for the counter first version number and counter second version number ([13]\.[1-5]) and the authorization field which follows. These numbers do not exist in the cpuid identification string when perf commands are executed on a z/VM environment (which does not support CPU counter measurement facility). CPUID string for LPAR: cpuid : IBM,3906,704,M03,3.5,002f CPUID string for z/VM: cpuid : IBM,2964,702,N96 This allows the removal of s390 specific cpuid compare code and uses the common compare function with its regular expression matching algorithm. Signed-off-by: NThomas Richter <tmricht@linux.ibm.com> Reviewed-by: NHendrik Brueckner <brueckner@linux.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Link: http://lkml.kernel.org/r/20180423081745.3672-1-tmricht@linux.ibm.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Namhyung Kim 提交于
map_groups__fixup_end() was called to set the end addresses of kernel and module maps. But now since machine__create_modules() sets the end address of modules properly, the only remaining piece is the kernel map. We can set it with adjacent module's address directly instead of calling map_groups__fixup_end(). If there's no module after the kernel map, the end address will be ~0ULL. Since it also changes the start address of the kernel map, it needs to re-insert the map to the kmaps in order to keep a correct ordering. Kim reported that it caused problems on ARM64. Reported-by: NKim Phillips <kim.phillips@arm.com> Tested-by: NKim Phillips <kim.phillips@arm.com> Signed-off-by: NNamhyung Kim <namhyung@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: kernel-team@lge.com Link: http://lkml.kernel.org/r/20180419235915.GA19067@sejongSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
-
由 Linus Torvalds 提交于
-
git://people.freedesktop.org/~airlied/linux由 Linus Torvalds 提交于
Pull drm fixes from Dave Airlie: "Exynos, i915, vc4, amdgpu fixes. i915: - an oops fix - two race fixes - some gvt fixes amdgpu: - dark screen fix - clk/voltage fix - vega12 smu fix vc4: - memory leak fix exynos just drops some code" * tag 'drm-fixes-for-v4.17-rc2' of git://people.freedesktop.org/~airlied/linux: (23 commits) drm/amd/powerplay: header file interface to SMU update drm/amd/pp: Fix bug voltage can't be OD separately on VI drm/amd/display: Don't program bypass on linear regamma LUT drm/i915: Fix LSPCON TMDS output buffer enabling from low-power state drm/i915/audio: Fix audio detection issue on GLK drm/i915: Call i915_perf_fini() on init_hw error unwind drm/i915/bios: filter out invalid DDC pins from VBT child devices drm/i915/pmu: Inspect runtime PM state more carefully while estimating RC6 drm/i915: Do no use kfree() to free a kmem_cache_alloc() return value drm/exynos: exynos_drm_fb -> drm_framebuffer drm/exynos: Move dma_addr out of exynos_drm_fb drm/exynos: Move GEM BOs to drm_framebuffer drm: Fix HDCP downstream dev count read drm/vc4: Fix memory leak during BO teardown drm/i915/execlists: Clear user-active flag on preemption completion drm/i915/gvt: Add drm_format_mod update drm/i915/gvt: Disable primary/sprite/cursor plane at virtual display initialization drm/i915/gvt: Delete redundant error message in fb_decode.c drm/i915/gvt: Cancel dma map when resetting ggtt entries drm/i915/gvt: Missed to cancel dma map for ggtt entries ...
-
git://people.freedesktop.org/~agd5f/linux由 Dave Airlie 提交于
- Fix a dark screen issue in DC - Fix clk/voltage dependency tracking for wattman - Update SMU interface for vega12 * 'drm-next-4.17' of git://people.freedesktop.org/~agd5f/linux: drm/amd/powerplay: header file interface to SMU update drm/amd/pp: Fix bug voltage can't be OD separately on VI drm/amd/display: Don't program bypass on linear regamma LUT
-
由 Dave Airlie 提交于
Merge tag 'exynos-drm-fixes-for-v4.17-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos into drm-next Remove Exynos specific framebuffer structure and relevant functions. - it removes exynos_drm_fb structure which is a wrapper of drm_framebuffer and unnecessary two exynos specific callback functions, exynos_drm_destory() and exynos_drm_fb_create_handle() because we can reuse existing drm common callback ones instead. * tag 'exynos-drm-fixes-for-v4.17-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos: drm/exynos: exynos_drm_fb -> drm_framebuffer drm/exynos: Move dma_addr out of exynos_drm_fb drm/exynos: Move GEM BOs to drm_framebuffer drm/amdkfd: Deallocate SDMA queues correctly drm/amdkfd: Fix scratch memory with HWS enabled
-
由 Dave Airlie 提交于
Merge tag 'drm-intel-next-fixes-2018-04-19' of git://anongit.freedesktop.org/drm/drm-intel into drm-next - Fix for FDO #105549: Avoid OOPS on bad VBT (Jani) - Fix rare pre-emption race (Chris) - Fix RC6 race against PM transitions (Tvrtko) * tag 'drm-intel-next-fixes-2018-04-19' of git://anongit.freedesktop.org/drm/drm-intel: drm/i915/audio: Fix audio detection issue on GLK drm/i915: Call i915_perf_fini() on init_hw error unwind drm/i915/bios: filter out invalid DDC pins from VBT child devices drm/i915/pmu: Inspect runtime PM state more carefully while estimating RC6 drm/i915: Do no use kfree() to free a kmem_cache_alloc() return value drm/i915/execlists: Clear user-active flag on preemption completion drm/i915/gvt: Add drm_format_mod update drm/i915/gvt: Disable primary/sprite/cursor plane at virtual display initialization drm/i915/gvt: Delete redundant error message in fb_decode.c drm/i915/gvt: Cancel dma map when resetting ggtt entries drm/i915/gvt: Missed to cancel dma map for ggtt entries drm/i915/gvt: Make MI_USER_INTERRUPT nop in cmd parser drm/i915/gvt: Mark expected switch fall-through in handle_g2v_notification drm/i915/gvt: throw error on unhandled vfio ioctls
-
git://anongit.freedesktop.org/drm/drm-misc由 Dave Airlie 提交于
drm-misc-fixes: stable: vc4: Fix memory leak during BO teardown (Daniel) dp: Add i2c retry for LSPCON adapters (Imre) hdcp: Fix device count mask (Ramalingam) Cc: Daniel J Blueman <daniel@quora.org Cc: Imre Deak <imre.deak@intel.com> Cc: Ramalingam C <ramalingam.c@intel.com> * tag 'drm-misc-fixes-2018-04-18-1' of git://anongit.freedesktop.org/drm/drm-misc: drm/i915: Fix LSPCON TMDS output buffer enabling from low-power state drm: Fix HDCP downstream dev count read drm/vc4: Fix memory leak during BO teardown
-