- 16 Sep, 2015 3 commits
-
-
Ingo Molnar authored
Conflicts: tools/perf/ui/browsers/hists.c Signed-off-by:Ingo Molnar <mingo@kernel.org>
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo: User visible changes: - Enhance the error reporting of tracepoint event parsing, e.g.: $ oldperf record -e sched:sched_switc usleep 1 event syntax error: 'sched:sched_switc' \___ unknown tracepoint Run 'perf list' for a list of valid events Now we get the much nicer: $ perf record -e sched:sched_switc ls event syntax error: 'sched:sched_switc' \___ can't access trace events Error: No permissions to read /sys/kernel/debug/tracing/events/sched/sched_switc Hint: Try 'sudo mount -o remount,mode=755 /sys/kernel/debug' And after we have those mount point permissions fixed: $ perf record -e sched:sched_switc ls event syntax error: 'sched:sched_switc' \___ unknown tracepoint Error: File /sys/kernel/debug/tracing/events/sched/sched_switc not found. Hint: Perhaps this kernel misses some CONFIG_ setting to enable this feature?. Now its just a matter of using what git uses to suggest alternatives when we make a typo, i.e. that it is just an 'h' missing :-) I.e. basically now the event parsing routing uses the strerror_open() routines introduced by and used in 'perf trace' work. (Jiri Olsa) Infrastructure changes: - Export init/exit_probe_symbol_maps() from 'perf probe' for use in eBPF. (Namhyung Kim) - Free perf_probe_event in cleanup_perf_probe_events(). (Namhyung Kim) - regs_query_register_offset() infrastructure + implementation for x86. First user will be the perf/eBPF code. (Wang Nan) Signed-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by:
-
Ingo Molnar authored
Merge tag 'perf-urgent-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent Pull perf/urgent fixes from Arnaldo Carvalho de Melo: - Fix segfault pressing -> in 'perf top' with no hist entries. (Wang Nan) E.g: perf top -e page-faults --pid 11400 # 11400 generates no page-fault - Fix propagation of thread and cpu maps, that got broken when doing incomplete changes to better support events with a PMU cpu mask, leading to Intel PT to fail with an error like: $ perf record -e intel_pt//u uname Error: The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (sched:sched_switch). Because intel_pt adds that sched:sched_switch evsel to the evlist after the thread/cpu maps were propagated to the evsels, fix it. (Adrian Hunter) Signed-off-by:
-
- 15 Sep, 2015 23 commits
-
-
Adrian Hunter authored
The test titled "Test software clock events have valid period values" was setting cpu/thread maps directly. Make it use the proper function perf_evlist__set_maps() especially now that it also propagates the maps. Signed-off-by:
Adrian Hunter <adrian.hunter@intel.com> Acked-by:
Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Link: http://lkml.kernel.org/r/1441699142-18905-15-git-send-email-adrian.hunter@intel.comSigned-off-by:
-
Adrian Hunter authored
The test titled "Test number of exit event of a simple workload" was setting cpu/thread maps directly. Make it use the proper function perf_evlist__set_maps() especially now that it also propagates the maps. Signed-off-by:
-
Adrian Hunter authored
Fix it by making it call perf_evlist__set_maps() instead of setting the maps itself. Signed-off-by:
-
Adrian Hunter authored
If evsels are added after maps are created, then they won't have any maps propagated to them. Fix that. Signed-off-by:
-
Adrian Hunter authored
Subsequent fixes will need a function that just propagates maps for a single evsel so factor it out. Signed-off-by:
-
Adrian Hunter authored
Since there is a function to set maps, perf_evlist__create_maps() should use it. Signed-off-by:
-
Adrian Hunter authored
Make perf_evlist__set_maps() more resilient by allowing for the possibility that one or another of the maps isn't being changed and therefore should not be "put". Signed-off-by:
-
Adrian Hunter authored
perf_evlist__propagate_maps() cannot easily tell if an evsel has its own cpu map. To make that simpler, keep a copy of the PMU cpu map and adjust the propagation logic accordingly. Signed-off-by:
-
Adrian Hunter authored
perf_evlist__propagate_maps() incorrectly assumes evsel->threads is NULL before reassigning it, but it won't be NULL when perf_evlist__set_maps() is used to set different (or NULL) maps. Thus thread_map__put must be used, which works even if evsel->threads is NULL. Signed-off-by:
-
Adrian Hunter authored
Commit d49e4695 ("perf evsel: Add a backpointer to the evlist a evsel is in") updated perf_evlist__add() but not perf_evlist__splice_list_tail(). This illustrates that it is better if perf_evlist__splice_list_tail() calls perf_evlist__add() instead of duplicating the logic, so do that. This will also simplify a subsequent fix for propagating maps. Signed-off-by:
-
Adrian Hunter authored
Subsequent patches will need to call perf_evlist__propagate_maps without reference to a "target". Add evlist->has_user_cpus to record whether the user has specified which cpus to target (and therefore whether that list of cpus should override the default settings for a selected event i.e. the cpu maps should be propagated) Signed-off-by:
-
Adrian Hunter authored
The validation checks that the values that were just assigned, got assigned i.e. the error can't ever happen. Subsequent patches will call this code in places where errors are not being returned. Changing those code paths to return this non-existent error is counter-productive, so just remove it. That in turn results in perf_evlist__set_maps not needing to return an error, but callers aren't checking it either, so remove that too. Signed-off-by:
-
Adrian Hunter authored
Don't need to check for NULL when "putting" evlist->maps and evlist->threads because the "put" functions already do that. Signed-off-by:
-
Adrian Hunter authored
If evsel->cpus is to be reassigned then the current value must be "put", which works even if it is NULL. Simplify the current logic by moving the "put" next to the assignment. Signed-off-by:
-
Wang Nan authored
regs_query_register_offset() is a helper function which converts register name like "%rax" to offset of a register in 'struct pt_regs', which is required by BPF prologue generator. Since the function is identical, try to reuse the code in arch/x86/kernel/ptrace.c. Comment inside dwarf-regs.c list the differences between this implementation and kernel code. get_arch_regstr() switches to regoffset_table and the old string table is dropped. Signed-off-by:
He Kuang <hekuang@huawei.com> Acked-by:
Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Brendan Gregg <brendan.d.gregg@gmail.com> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: David Ahern <dsahern@gmail.com> Cc: He Kuang <hekuang@huawei.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kaixu Xia <xiakaixu@huawei.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1441523623-152703-20-git-send-email-wangnan0@huawei.comSigned-off-by:
Wang Nan <wangnan0@huawei.com> Signed-off-by:
-
Wang Nan authored
regs_query_register_offset() is a helper function which converts register name like "%rax" to offset of a register in 'struct pt_regs', which is required by BPF prologue generator. PERF_HAVE_ARCH_REGS_QUERY_REGISTER_OFFSET indicates an architecture supports converting name of a register to its offset in 'struct pt_regs'. HAVE_ARCH_REGS_QUERY_REGISTER_OFFSET is introduced as the corresponding CFLAGS of PERF_HAVE_ARCH_REGS_QUERY_REGISTER_OFFSET. Signed-off-by:
-
Jiri Olsa authored
Enhancing parsing events tracepoint error output. Adding more verbose output when the tracepoint is not found or the tracing event path cannot be access. $ sudo perf record -e sched:sched_krava ls event syntax error: 'sched:sched_krava' \___ unknown tracepoint Error: File /sys/kernel/debug/tracing//tracing/events/sched/sched_krava not found. Hint: Perhaps this kernel misses some CONFIG_ setting to enable this feature?. Run 'perf list' for a list of valid events ... $ perf record -e sched:sched_krava ls event syntax error: 'sched:sched_krava' \___ can't access trace events Error: No permissions to read /sys/kernel/debug/tracing//tracing/events/sched/sched_krava Hint: Try 'sudo mount -o remount,mode=755 /sys/kernel/debug' Run 'perf list' for a list of valid events ... Signed-off-by: -
Jiri Olsa authored
Propagate error info from tp_format via ERR_PTR to get it all the way down to the parse-event.c tracepoint adding routines. Following functions now return pointer with encoded error: - tp_format - trace_event__tp_format - perf_evsel__newtp_idx - perf_evsel__newtp This affects several other places in perf, that cannot use pointer check anymore, but must utilize the err.h interface, when getting error information from above functions list. Signed-off-by:
-
Jiri Olsa authored
Pass 'struct parse_events_error *error' to the parse-event.c tracepoint adding path. It will be filled with error data in following patches. Signed-off-by:
Raphael Beamonte <raphael.beamonte@gmail.com> Reviewed-by:
Matt Fleming <matt@codeblueprint.co.uk> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1441615087-13886-4-git-send-email-jolsa@kernel.orgSigned-off-by:
-
Jiri Olsa authored
Adding part of the kernel's <linux/err.h> interface: inline void * __must_check ERR_PTR(long error); inline long __must_check PTR_ERR(__force const void *ptr); inline bool __must_check IS_ERR(__force const void *ptr); It will be used to propagate error through pointers in following patches. Signed-off-by:
-
Namhyung Kim authored
The init/exit_symbols_maps() functions are to setup and cleanup necessary info for probe events. But they need to be called from out of the probe code now, so this patch exports them. However the names are too generic, so change them to have 'probe'. :) Signed-off-by:
Namhyung Kim <namhyung@kernel.org> Acked-by:
-
Namhyung Kim authored
The cleanup_perf_probe_events() frees all resources related to a perf probe event. However it only freed resources in trace probe events, not perf probe events. So call clear_perf_probe_event() too. Reported-by:
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo-2' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo: User visible changes: - Add 'socket' sort entry, to sort by the processor socket in 'perf top' and 'perf report'. (Kan Liang) - Introduce --socket-filter to 'perf report', for filtering by processor socket. (Kan Liang) - Add new "Zoom into Processor Socket" operation in the perf hists browser, used in 'perf top' and 'perf report'. (Kan Liang) - Fix the 'CPU' hist browser column width calculation. (Arnaldo Carvalho de Melo) Infrastructure changes: - 'perf test' fixes for the object code reading entry. (Jan Stancek) - Add processor socket and cpu topology 'perf test' entries. (Kan Liang) - Introduce more sysfs__read_TYPE() helpers. (Arnaldo Carvalho de Melo) - Group cpu information reading functions in tools/lib/api/cpu.[ch], starting with cpu__get_max_freq() from a patchkit by Kan Liang. (Arnaldo Carvalho de Melo) - Retrieve the MSR PMU type from a perf.data file header and store it in struct perf_env. (Kan Liang) - Add tools/include into CTAGS file list. (Jiri Olsa) - Add iterator function for perf tests. (Matt Fleming) - Switch to tracing_patch interface. (Jiri Olsa) Signed-off-by:
-
- 14 Sep, 2015 14 commits
-
-
Wang Nan authored
'perf top' segfaults with following operation: # perf top -e page-faults -p 11400 # 11400 never generates page-fault Then on the resulting empty interface, press right key: # ./perf top -e page-faults -p 11400 perf: Segmentation fault -------- backtrace -------- ./perf[0x535428] /lib64/libc.so.6(+0x3545f)[0x7f0dd360745f] ./perf[0x531d46] ./perf(perf_evlist__tui_browse_hists+0x96)[0x5340d6] ./perf[0x44ba2f] /lib64/libpthread.so.0(+0x81d0)[0x7f0dd49dc1d0] /lib64/libc.so.6(clone+0x6c)[0x7f0dd36b90dc] The bug resides in perf_evsel__hists_browse() that, in the above circumstance browser->selection can be NULL, but code after skip_annotation doesn't consider it. This patch fix it by checking browser->selection before fetching browser->selection->map. Signed-off-by:
-
Kan Liang authored
Add test case for hists socket filter. Signed-off-by:
Kan Liang <kan.liang@intel.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1441377946-44429-5-git-send-email-kan.liang@intel.comSigned-off-by:
-
Kan Liang authored
Currently, users can zoom in/out for threads and dso in 'perf top' and 'perf report'. This patch extends it for the processor sockets. 'S' is the short key to zoom into current Processor Socket. Signed-off-by:
-
Kan Liang authored
Introduce --socket-filter option for 'perf report' to only show entries for a processor socket that match this filter. $ perf report --socket-filter 1 --stdio # To display the perf.data header info, please use --header/--header-only options. # # Total Lost Samples: 0 # # Samples: 752 of event 'cycles' # Event count (approx.): 350995599 # Processor Socket: 1 # # Overhead Command Shared Object Symbol # ........ ......... ................ ................................. # 97.02% test test [.] plusB_c 0.97% test test [.] plusA_c 0.23% swapper [kernel.vmlinux] [k] acpi_idle_do_entry 0.09% rcu_sched [kernel.vmlinux] [k] dyntick_save_progress_counter 0.01% swapper [kernel.vmlinux] [k] task_waking_fair 0.00% swapper [kernel.vmlinux] [k] run_timer_softirq Signed-off-by: -
Kan Liang authored
This patch enable perf report to sort by processor socket: $ perf report --stdio --sort socket,comm,dso,symbol # To display the perf.data header info, please use --header/--header-only options. # # Total Lost Samples: 0 # # Samples: 686 of event 'cycles' # Event count (approx.): 349215462 # # Overhead SOCKET Command Shared Object Symbol # ........ ...... ....... ................ ............................ # 97.05% 000 test test [.] plusB_c 0.98% 000 test test [.] plusA_c 0.93% 001 perf [kernel.vmlinux] [k] smp_call_function_single 0.19% 001 perf [kernel.vmlinux] [k] page_fault 0.19% 001 swapper [kernel.vmlinux] [k] pm_qos_request 0.16% 000 test [kernel.vmlinux] [k] add_mm_counter_fast Signed-off-by: -
Kan Liang authored
This information will come from perf.data files of from the current system, cached when needed, such as when the 'socket' sort order gets introduced. Signed-off-by:
-
Arnaldo Carvalho de Melo authored
The 'struct machine' represents the machine where the samples were/are being collected, and we also have a 'struct perf_env' with extra details about such machine, that we were collecting at 'perf.data' creation time but we also needed when no perf.data file is being used, such as in 'perf top'. So, get those structs closer together, as they provide a bigger picture of the sample's environment. In 'perf session', when the file argument is NULL, we can assume that the tool is sampling the running machine, so point machine->env to the global put in place in previous patches, while set it to the perf_header.env one when reading from a file. This paves the way for machine->env to be used in perf_event__preprocess_sample to populate addr_location.socket. Tested-by:
-
Arnaldo Carvalho de Melo authored
Out of the code to write the cpu topology map in the perf.data file header. Now if one needs the CPU topology map for the running machine, one needs to call perf_env__read_cpu_topology_map(perf_env) and the info will be stored in perf_env.cpu. For now we're using a global perf_env variable, that will have its contents freed after we run a builtin. v2: Check perf_env__read_cpu_topology_map() return in write_cpu_topology() (Kan Liang) Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Borislav Petkov <bp@suse.de> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Kan Liang <kan.liang@intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1441828225-667-5-git-send-email-acme@kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
We have the tools/lib/ sysfs__read_int() for that, avoid code duplication. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Borislav Petkov <bp@suse.de> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-fqg6vt5ku72pbf54ljg6tmoy@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
E.g.: $ ./cpu__get_max_freq 3200000 It does that, as Kan's patch does, by looking at these files: $ cat /sys/devices/system/cpu/online 0-3 $ ./sysfs__read_ull devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq /sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq=3200000 $ I.e. find out the first online CPU, then read its cpufreq info. But do it in tools/lib/api/, so that other tools/ living code can use it, not just perf. Based-on-a-patch-by:
-
Arnaldo Carvalho de Melo authored
To read either an int or an unsigned long long value from the given file. E.g.: $ cat /sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq 3200000 $ ./sysfs__read_ull devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq /sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq=3200000 $ Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Borislav Petkov <bp@suse.de> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Kan Liang <kan.liang@intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-4a12m4d5k8m4qgc1vguocvei@git.kernel.orgSigned-off-by:
-
Kan Liang authored
Get msr pmu type when processing pmu_mappings Signed-off-by:
-
Jiri Olsa authored
Adding tools/include into tags directories, to have include definitions reachable via tags/cscope. Signed-off-by:
-
Arnaldo Carvalho de Melo authored
We have no use for it in evsel.h. Tested-by:
-
