linux/tools/perf
Milian Wolff b38775cf76 perf report: Cache failed lookups of inlined frames
When no inlined frames could be found for a given address, we did not
store this information anywhere. That means we potentially do the costly
inliner lookup repeatedly for cases where we know it can never succeed.

This patch makes dso__parse_addr_inlines always return a valid
inline_node. It will be empty when no inliners are found. This enables
us to cache the empty list in the DSO, thereby improving the performance
when many addresses fail to find the inliners.

For my trivial example, the performance impact is already quite
significant:

Before:

~~~~~
 Performance counter stats for 'perf report --stdio --inline -g srcline -s srcline' (5 runs):

        594.804032      task-clock (msec)         #    0.998 CPUs utilized            ( +-  0.07% )
                53      context-switches          #    0.089 K/sec                    ( +-  4.09% )
                 0      cpu-migrations            #    0.000 K/sec                    ( +-100.00% )
             5,687      page-faults               #    0.010 M/sec                    ( +-  0.02% )
     2,300,918,213      cycles                    #    3.868 GHz                      ( +-  0.09% )
     4,395,839,080      instructions              #    1.91  insn per cycle           ( +-  0.00% )
       939,177,205      branches                  # 1578.969 M/sec                    ( +-  0.00% )
        11,824,633      branch-misses             #    1.26% of all branches          ( +-  0.10% )

       0.596246531 seconds time elapsed                                          ( +-  0.07% )
~~~~~

After:

~~~~~
 Performance counter stats for 'perf report --stdio --inline -g srcline -s srcline' (5 runs):

        113.111405      task-clock (msec)         #    0.990 CPUs utilized            ( +-  0.89% )
                29      context-switches          #    0.255 K/sec                    ( +- 54.25% )
                 0      cpu-migrations            #    0.000 K/sec
             5,380      page-faults               #    0.048 M/sec                    ( +-  0.01% )
       432,378,779      cycles                    #    3.823 GHz                      ( +-  0.75% )
       670,057,633      instructions              #    1.55  insn per cycle           ( +-  0.01% )
       141,001,247      branches                  # 1246.570 M/sec                    ( +-  0.01% )
         2,346,845      branch-misses             #    1.66% of all branches          ( +-  0.19% )

       0.114222393 seconds time elapsed                                          ( +-  1.19% )
~~~~~

Signed-off-by: Milian Wolff <milian.wolff@kdab.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20171019113836.5548-3-milian.wolff@kdab.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-25 10:50:45 -03:00
..
arch perf annotate: Remove arch::cpuid_parse callback 2017-10-23 11:20:54 -03:00
bench perf tools: Use __maybe_unused consistently 2017-06-19 15:27:06 -03:00
Documentation perf list: Fix group description in the man page 2017-10-23 11:20:54 -03:00
jvmti perf jit: fix typo: "incalid" -> "invalid" 2017-06-27 11:55:06 -03:00
pmu-events perf vendor events: Add Goldmont Plus V1 event file 2017-10-23 16:30:54 -03:00
python perf python: Add tracepoint example 2016-07-12 16:23:35 -03:00
scripts perf script python: Add support for sqlite3 to call-graph-from-sql.py 2017-08-15 17:03:38 -03:00
tests perf tests attr: Make hw events optional 2017-10-23 11:20:54 -03:00
trace perf trace beauty madvise: Generate 'behavior' string table from kernel headers 2017-09-21 13:12:59 -03:00
ui perf report: Remove code to handle inline frames from browsers 2017-10-24 09:59:55 -03:00
util perf report: Cache failed lookups of inlined frames 2017-10-25 10:50:45 -03:00
.gitignore perf tools: Ignore generated files pmu-events/{jevents,pmu-events.c} for git 2017-03-13 10:59:36 -03:00
Build perf trace: Only build tools/perf/trace/beauty/ when building 'perf trace' 2017-07-18 23:13:52 -03:00
builtin-annotate.c perf annotate browser: Support --show-nr-samples option 2017-08-18 11:15:09 -03:00
builtin-bench.c perf tools: Remove unused 'prefix' from builtin functions 2017-03-27 11:58:09 -03:00
builtin-buildid-cache.c perf buildid-cache: Cache debuginfo 2017-07-18 23:14:11 -03:00
builtin-buildid-list.c perf tools: Include errno.h where needed 2017-04-19 13:01:51 -03:00
builtin-c2c.c perf tools: Fix leaking rec_argv in error cases 2017-09-18 09:40:21 -03:00
builtin-config.c perf config: Write a config file just once 2017-09-13 09:49:15 -03:00
builtin-data.c perf data: Add doc when no conversion support compiled 2017-07-28 16:30:45 -03:00
builtin-diff.c perf config: Do not die when parsing u64 or int config values 2017-06-27 11:44:58 -03:00
builtin-evlist.c perf tools: Remove unused 'prefix' from builtin functions 2017-03-27 11:58:09 -03:00
builtin-ftrace.c tools include: Adopt strstarts() from the kernel 2017-07-20 15:46:10 -03:00
builtin-help.c tools include: Adopt strstarts() from the kernel 2017-07-20 15:46:10 -03:00
builtin-inject.c perf tools: Add feature header record to pipe-mode 2017-07-18 23:14:36 -03:00
builtin-kallsyms.c perf tools: Including missing inttypes.h header 2017-04-19 13:01:46 -03:00
builtin-kmem.c perf kmem: Perform some cleanup if '--time' is given an invalid value 2017-10-23 16:30:53 -03:00
builtin-kvm.c perf top: Implement multithreading for perf_event__synthesize_threads 2017-10-03 09:27:46 -03:00
builtin-list.c perf list: Add metric groups to perf list 2017-09-13 09:49:13 -03:00
builtin-lock.c perf tools: Include errno.h where needed 2017-04-19 13:01:51 -03:00
builtin-mem.c perf tools: Fix leaking rec_argv in error cases 2017-09-18 09:40:21 -03:00
builtin-probe.c perf buildid-cache: Support binary objects from other namespaces 2017-07-18 23:14:11 -03:00
builtin-record.c perf mmap: Adopt push method from builtin-record.c 2017-10-23 11:20:54 -03:00
builtin-report.c perf report: Group stat values on global event id 2017-08-28 16:44:44 -03:00
builtin-sched.c perf sched timehist: Add pid and tid options 2017-09-13 09:49:12 -03:00
builtin-script.c perf script: Fix error handling path 2017-10-23 16:30:53 -03:00
builtin-stat.c perf stat: Fall weak group back even for EBADF 2017-09-13 09:49:16 -03:00
builtin-timechart.c perf tools: Fix leaking rec_argv in error cases 2017-09-18 09:40:21 -03:00
builtin-top.c perf top: Add option to set the number of thread for event synthesize 2017-10-03 09:27:54 -03:00
builtin-trace.c perf tools: Introduce binary__fprintf() 2017-10-23 16:30:52 -03:00
builtin-version.c perf tools: Remove string.h, unistd.h and sys/stat.h from util.h 2017-04-24 13:43:33 -03:00
builtin.h perf tools: Remove stale prototypes from builtin.h 2017-04-24 13:43:33 -03:00
check-headers.sh perf tools: Do not check ABI headers in a detached tarball build 2017-10-23 16:30:50 -03:00
command-list.txt perf tools: Missing c2c command in command-list 2017-03-13 10:59:31 -03:00
CREDITS
design.txt
Makefile perf build tests: Do parallell builds with 'build-test' 2016-02-04 15:57:00 -03:00
Makefile.config perf tools: Robustify detection of clang binary 2017-08-28 16:44:46 -03:00
Makefile.perf perf trace beauty madvise: Generate 'behavior' string table from kernel headers 2017-09-21 13:12:59 -03:00
MANIFEST perf tools: Get all of tools/{arch,include}/ in the MANIFEST 2017-09-25 10:39:43 -03:00
perf-archive.sh
perf-completion.sh
perf-read-vdso.c
perf-sys.h perf tools: Use default CPUINFO_PROC where it fits 2017-08-17 16:58:21 -03:00
perf-with-kcore.sh
perf.c perf tools: Support running perf binaries with a dash in their name 2017-09-12 12:48:54 -03:00
perf.h perf record: Support direct --user-regs arguments 2017-09-13 09:49:14 -03:00