- 24 Nov, 2014 2 commits
-
-
Namhyung Kim authored
If first level callchain has more than single path like when -g caller option is given, it should show only first one in the path and hide others. But it didn't do it properly and just hindered the output. Before: - 80.33% 11.11% abc2 abc2 [.] main + 86.18% main 13.82% __libc_start_main main After: - 80.33% 11.11% abc2 abc2 [.] main + 86.18% main + 13.82% __libc_start_main Signed-off-by:Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1416816807-6495-2-git-send-email-namhyung@kernel.orgSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Namhyung Kim authored
Currently perf report on TUI doesn't print percent for first-level callchain entry. I guess it (wrongly) assumes that there's only a single callchain in the first level. This patch fixes it by handling the first level callchains same as others - if it's not 100% it should print the percent value. Also it'll affect other callchains in the other way around - if it's 100% (single callchain) it should not print the percentage. Before: - 30.95% 6.84% abc2 abc2 [.] a - a - 70.00% c - 100.00% apic_timer_interrupt smp_apic_timer_interrupt local_apic_timer_interrupt hrtimer_interrupt ... + 30.00% b + __libc_start_main After: - 30.95% 6.84% abc2 abc2 [.] a - 77.90% a - 70.00% c - apic_timer_interrupt smp_apic_timer_interrupt local_apic_timer_interrupt hrtimer_interrupt ... + 30.00% b + 22.10% __libc_start_main Reported-by:
-
- 20 Nov, 2014 1 commit
-
-
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 fixes: - Fallback to kallsyms when using the minimal 'ELF' loader (Arnaldo Carvalho de Melo) - Fix annotation with kcore (Adrian Hunter) - Fix up srcline histogram key formatting (Arnaldo Carvalho de Melo) - Add missing handler for PERF_RECORD_MMAP2 events in 'perf diff' (Kan Liang) User visible changes/new features: - Only print base source file for srcline histogram sort key (Andi Kleen) - Support source line numbers in annotate using a hotkey (Andi Kleen) Infrastructure changes and fixes: - Do not poll events that use the system_wide flag (Adrian Hunter) - Add perf-read-vdso32 and perf-read-vdsox32 to .gitignore (Adrian Hunter) - Only override the default :tid comm entry (Adrian Hunter) - Factor out adding new call chain entries (Andi Kleen) - Use al.addr to set up call chain (Andi Kleen) - Use a common function to resolve symbol or name (Andi Kleen) - Fix ftrace:function event recording (Jiri Olsa) - Move disable_buildid_cache() to util/build-id.c (Namhyung Kim) - Clean up libelf feature support code (Namhyung Kim) - Fix typo in python 'perf test' (WANG Chao) Signed-off-by:
Ingo Molnar <mingo@kernel.org>
-
- 19 Nov, 2014 16 commits
-
-
Adrian Hunter authored
Events may still be ordered even if there are no timestamps e.g. if the data is recorded per-thread. Also synthesized COMM events have a timestamp of zero. Consequently it is better to keep comm entries even if they have a timestamp of zero. However, when a struct thread is created the command string is not known and a comm entry with a string of the form ":<tid>" is used. In that case thread->comm_set is false and the comm entry should be overridden. 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: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1415715423-15563-4-git-send-email-adrian.hunter@intel.comSigned-off-by:
-
Adrian Hunter authored
Recently added executables Add perf-read-vdso32 and perf-read-vdsox32 need to be added to .gitignore. Signed-off-by:
-
Adrian Hunter authored
The system_wide flag causes a selected event to be opened always without a pid. Consequently it will never get a POLLHUP, but it is used for tracking in combination with other events, so it should not need to be polled anyway. Therefore don't add it for polling. Signed-off-by:
-
Jiri Olsa authored
Following patch fails (-EINVAL) ftrace:function with enabled user space callchains: cfa77bc4 perf: Disallow user-space callchains for function trace events We need to follow in perf tool itself and explicitly set the perf_event_attr::exclude_callchain_user flag for ftrace:function event. Reported-by:
Steven Rostedt <rostedt@goodmis.org> Signed-off-by:
Jiri Olsa <jolsa@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> Cc: Steven Rostedt <rostedt@goodmis.org> Link: http://lkml.kernel.org/r/1415899263-24820-1-git-send-email-jolsa@kernel.orgSigned-off-by:
-
Kan Liang authored
Without mmap2, perf diff fails to find the symbol name. The default symbol sort key doesn't work well. Signed-off-by:
Kan Liang <kan.liang@intel.com> Acked-by:
-
Arnaldo Carvalho de Melo authored
Problem introduced in: commit 5b591669 "perf report: Honor column width setting" Where the left justification signal was after the width, which ended up, when the width was, say, 11, always printing: %11.11-s Instead of src:line left justified and limited to 11 chars. Resulting in a like: 70.93% %11.11-s [.] f2 tcall When it should instead be: 70.93% tcall.c:5 [.] f2 tcall Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Borislav Petkov <bp@suse.de> Cc: David Ahern <dsahern@gmail.com> Cc: Don Zickus <dzickus@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-2xnt0vqkoox52etq2qhyetr0@git.kernel.orgSigned-off-by:
-
Andi Kleen authored
With srcline key/sort'ing it's useful to have line numbers in the annotate window. This patch implements this. Use objdump -l to request the line numbers and save them in the line structure. Then the browser displays them for source lines. The line numbers are not displayed by default, but can be toggled on with 'k' There is one unfortunate problem with this setup. For lines not containing source and which are outside functions objdump -l reports line numbers off by a few: it always reports the first line number in the next function even for lines that are outside the function. I haven't found a nice way to detect/correct this. Probably objdump has to be fixed. See https://sourceware.org/bugzilla/show_bug.cgi?id=16433 The line numbers are still useful even with these problems, as most are correct and the ones which are not are nearby. v2: Fix help text. Handle (discriminator...) output in objdump. Left align the line numbers. Signed-off-by:
Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1415844328-4884-9-git-send-email-andi@firstfloor.orgSigned-off-by:
-
Andi Kleen authored
For perf report with --sort srcline only print the base source file name. This makes the results generally fit much better to the screen. The path is usually not that useful anyways because it is often from different systems. Signed-off-by:
-
Andi Kleen authored
Refactor the duplicated code to resolve the symbol name or the address of a symbol into a single function. Used in next patch to add common functionality. Signed-off-by:
-
Andi Kleen authored
Use the relative address, this makes get_srcline work correctly in the end. Signed-off-by:
-
Andi Kleen authored
Move the code to resolve and add a new callchain entry into a new add_callchain_ip function. This will be used in the next patches to add LBRs too. No change in behavior. Signed-off-by:
-
Adrian Hunter authored
Patch "perf tools: Fix build-id matching on vmlinux" breaks annotation with kcore. The problem is that symbol__annotate() first gets the filename based on the build-id which was previously not set. This patch provides a quick fix, however there should probably be only one way to determine the filename. e.g. symbol__annotate() should use the same way as dso__data_fd(). Signed-off-by:
-
WANG Chao authored
Library loading in python syntax should be 'import perf', not 'use perf'. Signed-off-by:
WANG Chao <chaowang@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jean Pihet <jean.pihet@linaro.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.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/1415780826-13250-1-git-send-email-chaowang@redhat.comSigned-off-by:
-
Arnaldo Carvalho de Melo authored
The minimal ELF loader should not return 1 when it manages to read the vmlinux build-id, it should instead return 0, meaning that it hasn't loaded any symbols, since it doesn't parses ELF at all. That way, the main symbol.c routines will understand that it is necessary to continue looking for a file with symbols, and when no libelf is linked, that means it will eventually try kallsyms. Reported-by:
Peter Zijlstra <peterz@infradead.org> Tested-by:
-
Namhyung Kim authored
Current EXTLIBS contains -lelf by default and removes it when libelf is not detected. This is little bit confusing since we can now build perf without libelf so there's no need to handle it differently than other libraries. Signed-off-by:
-
Namhyung Kim authored
Also move static variable no_buildid_cache and check it in the perf_session_cache_build_ids(). Signed-off-by:
-
- 16 Nov, 2014 14 commits
-
-
Peter Zijlstra authored
This patch reorders fields in the perf_sample_data struct in order to minimize the number of cachelines touched in perf_sample_data_init(). It also removes some intializations which are redundant with the code in kernel/events/core.c Signed-off-by:
-
Stephane Eranian authored
Add -I/--intr-regs option to capture machine state registers at interrupt. Add the corresponding man page description Signed-off-by:
Stephane Eranian <eranian@google.com> Signed-off-by:
-
Stephane Eranian authored
This patch updates the sample parsing test with support for the sampling of machine interrupted state. The patch modifies the do_test() code to sahred the sample regts bitmask between user and intr regs. Signed-off-by:
-
Stephane Eranian authored
Add the infrastructure to setup, collect and report the interrupt machine state regs which can be captured by the kernel. Signed-off-by:
-
Stephane Eranian authored
PEBS can capture machine state regs at retiremnt of the sampled instructions. When precise sampling is enabled on an event, PEBS is used, so substitute the interrupted state with the PEBS state. Note that not all registers are captured by PEBS. Those missing are replaced by the interrupt state counter-parts. Signed-off-by:
-
Stephane Eranian authored
Enable capture of interrupted machine state for each sample. Registers to sample are passed per event in the sample_regs_intr bitmask. To sample interrupt machine state, the PERF_SAMPLE_INTR_REGS must be passed in sample_type. The list of available registers is arch dependent and provided by asm/perf_regs.h Registers are laid out as u64 in the order of the bit order of sample_intr_regs. This patch also adds a new ABI version PERF_ATTR_SIZE_VER4 because we extend the perf_event_attr struct with a new u64 field. Reviewed-by:
Jiri Olsa <jolsa@redhat.com> Signed-off-by:
-
Andi Kleen authored
Disallow setting inv/cmask/etc. flags for all PEBS events on these CPUs, except for the UOPS_RETIRED.* events on Nehalem/Westmere, which are needed for cycles:p. This avoids an undefined situation strongly discouraged by the Intle SDM. The PLD_* events were already covered. This follows the earlier changes for Sandy Bridge and alter. Signed-off-by:
-
Andi Kleen authored
My earlier commit: 86a04461 ("perf/x86: Revamp PEBS event selection") made nearly all PEBS on Sandy/IvyBridge/Haswell to reject non zero flags. However this wasn't done for the INST_RETIRED.PREC_DIST event because no suitable macro existed. Now that we have INTEL_FLAGS_UEVENT_CONSTRAINT enforce zero flags for INST_RETIRED.PREC_DIST too. Signed-off-by:
-
Andi Kleen authored
Add a FLAGS_UEVENT_CONSTRAINT macro that allows us to match on event+umask, and in additional all flags. This is needed to ensure the INV and CMASK fields are zero for specific events, as this can cause undefined behavior. Signed-off-by:
-
Andi Kleen authored
Add scaling to MB/s to the memory controller read/write events for Sandy/IvyBridge/Haswell-EP similar to how the client does. This makes the events easier to use from the standard perf tool. Signed-off-by:
-
Ingo Molnar authored
Signed-off-by: -
Andi Kleen authored
There were several reports that on some systems writing the SBOX0 PMU initialization MSR would #GP at boot. This did not happen on all systems -- my two test systems booted fine. Writing the three initialization bits bit-by-bit seems to avoid the problem. So add a special callback to do just that. This replaces an earlier patch that disabled the SBOX. Reported-by:
Alexei Starovoitov <alexei.starovoitov@gmail.com> Reported-and-Tested-by:
Patrick Lu <patrick.lu@intel.com> Signed-off-by:
-
Andi Kleen authored
The counter register offsets for the IRP box PMU for Haswell-EP were incorrect. The offsets actually changed over IvyBridge EP. Fix them to the correct values. For this we need to fork the read function from the IVB and use an own counter array. Tested-by: patrick.lu@intel.com Signed-off-by:
-
Mark Rutland authored
When a CPU hotplugged out, we call perf_remove_from_context() (via perf_event_exit_cpu()) to rip each CPU-bound event out of its PMU's cpu context, but leave siblings grouped together. Freeing of these events is left to the mercy of the usual refcounting. When a CPU-bound event's refcount drops to zero we cross-call to __perf_remove_from_context() to clean it up, detaching grouped siblings. This works when the relevant CPU is online, but will fail if the CPU is currently offline, and we won't detach the event from its siblings before freeing the event, leaving the sibling list corrupt. If the sibling list is later walked (e.g. because the CPU cam online again before a remaining sibling's refcount drops to zero), we will walk the now corrupted siblings list, potentially dereferencing garbage values. Given that the events should never be scheduled again (as we removed them from their context), we can simply detatch siblings when the CPU goes down in the first place. If the CPU comes back online, the redundant call to __perf_remove_from_context() is safe. Reported-by:
Drew Richardson <drew.richardson@arm.com> Signed-off-by:
Mark Rutland <mark.rutland@arm.com> Signed-off-by:
-
- 12 Nov, 2014 1 commit
-
-
Aravind Gopalakrishnan authored
New Fam15h models carry extra feature bits and extend the MSR register space for IBS ops. Adding them here. While at it, add functionality to read IbsBrTarget and OpData4 depending on their availability if user wants a PERF_SAMPLE_RAW. Signed-off-by:
Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com> Acked-by:
Borislav Petkov <bp@suse.de> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Jan Kiszka <jan.kiszka@siemens.com> Cc: Len Brown <len.brown@intel.com> Cc: Fenghua Yu <fenghua.yu@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: <paulus@samba.org> Cc: <acme@kernel.org> Link: http://lkml.kernel.org/r/1415651066-13523-1-git-send-email-Aravind.Gopalakrishnan@amd.comSigned-off-by:
-
- 07 Nov, 2014 1 commit
-
-
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: Infrastructure changes: - Add gzip decompression support for kernel modules (Namhyung Kim) - More prep patches for Intel PT, including a a thread stack and more stuff made available via the database export mechanism (Adrian Hunter) - Optimize checking that tracepoint events are defined in perf script perl/python (Jiri Olsa) - Do not free pevent when deleting tracepoint evsel (Jiri Olsa) - Fix build-id matching for vmlinux (Namhyung Kim) Signed-off-by:
-
- 06 Nov, 2014 4 commits
-
-
Jiri Olsa authored
The libtraceevent library's main handle 'struct pevent' holds pointers of every event that was added to it via functions: pevent_parse_format pevent_parse_event We can't release struct event_format (call pevent_free_format) separately, because that breaks that pointers array mentioned above and another add_event call could end up with segfault. All added events are released within the handle cleanup in pevent_free. Signed-off-by:
-
Jiri Olsa authored
We don't need to maintain cache of 'struct event_format' objects. Currently the 'struct perf_evsel' holds this reference already. Adding events_defined bitmap to keep track of defined events, which is much cheaper than array of pointers. Signed-off-by:
Namhyung Kim <namhyung@gmail.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1414363445-22370-3-git-send-email-jolsa@kernel.orgSigned-off-by:
-
Jiri Olsa authored
We don't need to maintain cache of 'struct event_format' objects. Currently the 'struct perf_evsel' holds this reference already. Adding events_defined bitmap to keep track of defined events, which is much cheaper than array of pointers. Signed-off-by:
-
Jiri Olsa authored
Set a bit and return its old value. Stolen from kernel sources, will be used in next patches. Signed-off-by:
-
- 05 Nov, 2014 1 commit
-
-
Namhyung Kim authored
The previous patch changed kernel dso name from '[kernel.kallsyms]' to vmlinux. However it might add confusion to old users accustomed to the old name. So change the short name to '[kernel.vmlinux]' to reduce such confusion. Before: # Overhead Command Shared Object Symbol # ........ .............. ....................... ............................... # 9.83% swapper vmlinux [k] intel_idle 4.10% awk libc-2.20.so [.] __strcmp_sse2 1.86% sed libc-2.20.so [.] __strcmp_sse2 1.78% netctl-auto libc-2.20.so [.] __strcmp_sse2 1.23% netctl-auto libc-2.20.so [.] __mbrtowc 1.21% firefox libxul.so [.] 0x00000000024b62bd 1.20% swapper vmlinux [k] cpuidle_enter_state 1.03% sleep vmlinux [k] copy_user_generic_unrolled After: # Overhead Command Shared Object Symbol # ........ .............. ....................... ............................... # 9.83% swapper [kernel.vmlinux] [k] intel_idle 4.10% awk libc-2.20.so [.] __strcmp_sse2 1.86% sed libc-2.20.so [.] __strcmp_sse2 1.78% netctl-auto libc-2.20.so [.] __strcmp_sse2 1.23% netctl-auto libc-2.20.so [.] __mbrtowc 1.21% firefox libxul.so [.] 0x00000000024b62bd 1.20% swapper [kernel.vmlinux] [k] cpuidle_enter_state 1.03% sleep [kernel.vmlinux] [k] copy_user_generic_unrolled Signed-off-by:
-
