2
0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-12-17 17:53:56 +08:00
linux-next/tools/perf/Documentation
Kan Liang 94ba462d69 perf diff: Support for different binaries
Currently, the perf diff only works with same binaries. That's because
it compares the symbol start address. It doesn't work if the perf.data
comes from different binaries. This patch matches the symbol names.

Actually, perf diff once intended to compare the symbol names.  The
commit as below can look for a pair by name.

604c5c9297 (perf diff: Change the default sort order to "dso,symbol")
However, at that time, perf diff used a global list of dsos. That means
the binaries which has same name can only be loaded once. That's a
problem for comparing different binaries.

For example, we have an old binary and an updated binary. They very
likely have same name and most of the functions, so only dsos from old
binary will be loaded. When processing the data from updated binary,
perf still use the symbol information from old binary. That's wrong.

Then the commit as below used IP to replace symbol name.
9c443dfdd3 ("perf diff: Fix support for all --sort combinations")
>From that time, perf diff starts to compare the symbol address.

The global dsos is discarded from a patch in 2010.
a1645ce12a ("perf: 'perf kvm' tool for monitoring guest performance
from host")
However, at that time, perf diff already compared by address. So perf
diff cannot work for different binaries as well.

This patch actually rolls back the perf diff to original design. The
document is also changed, so everybody knows the original design is to
compare the symbol names.

Here are some examples:

The only difference between example_v1.c and example_v2.c is the
location of f2 and f3. There is no change in behavior, but the previous
perf diff display the wrong differential profile.

example_v1.c
noinline void f3(void)
{
        volatile int i;
        for (i = 0; i < 10000;) {

                if(i%2)
                        i++;
                else
                        i++;
        }
}

noinline void f2(void)
{
        volatile int a = 100, b, c;
        for (b = 0; b < 10000; b++)
                c = a * b;

}

noinline void f1(void)
{
                f2();
                f3();
}

int main()
{
        int i;
        for (i = 0; i < 100000; i++)
                f1();
}

example_v2.c
noinline void f2(void)
{
        volatile int a = 100, b, c;
        for (b = 0; b < 10000; b++)
                c = a * b;
}

noinline void f3(void)
{
        volatile int i;
        for (i = 0; i < 10000;) {
                if(i%2)
                        i++;
                else
                        i++;
        }
}

noinline void f1(void)
{
                f2();
                f3();
}

int main()
{
        int i;
        for (i = 0; i < 100000; i++)
                f1();
}

[lk@localhost perf_diff]$ gcc example_v1.c -o example
[lk@localhost perf_diff]$ perf record -o example_v1.data ./example
[ perf record: Woken up 4 times to write data ]
[ perf record: Captured and wrote 0.813 MB example_v1.data (~35522 samples) ]

[lk@localhost perf_diff]$ gcc example_v2.c -o example
[lk@localhost perf_diff]$ perf record -o example_v2.data ./example
[ perf record: Woken up 4 times to write data ]
[ perf record: Captured and wrote 0.824 MB example_v2.data (~36015 samples) ]

Old perf diff result:

[lk@localhost perf_diff]$ perf diff example_v1.data example_v2.data
 Event 'cycles'
 Baseline    Delta  Shared Object     Symbol
 ........  .......  ................  ...............................

                     [kernel.vmlinux]  [k] __perf_event_task_sched_out
     0.00%           [kernel.vmlinux]  [k] apic_timer_interrupt
                     [kernel.vmlinux]  [k] idle_cpu
                     [kernel.vmlinux]  [k] intel_pstate_timer_func
                     [kernel.vmlinux]  [k] native_read_msr_safe
     0.00%           [kernel.vmlinux]  [k] native_read_tsc
     0.00%           [kernel.vmlinux]  [k] native_write_msr_safe
                     [kernel.vmlinux]  [k] ntp_tick_length
     0.00%           [kernel.vmlinux]  [k] rb_erase
     0.00%           [kernel.vmlinux]  [k] tick_sched_timer
     0.00%           [kernel.vmlinux]  [k] unmap_single_vma
     0.00%           [kernel.vmlinux]  [k] update_wall_time
     0.00%           example           [.] f1
    46.24%           example           [.] f2
    53.71%   -7.55%  example           [.] f3
            +53.81%  example           [.] f3
     0.02%           example           [.] main

New perf diff result:

[lk@localhost perf_diff]$ perf diff example_v1.data example_v2.data
                     [kernel.vmlinux]  [k] __perf_event_task_sched_out
     0.00%           [kernel.vmlinux]  [k] apic_timer_interrupt
                     [kernel.vmlinux]  [k] idle_cpu
                     [kernel.vmlinux]  [k] intel_pstate_timer_func
                     [kernel.vmlinux]  [k] native_read_msr_safe
     0.00%           [kernel.vmlinux]  [k] native_read_tsc
     0.00%           [kernel.vmlinux]  [k] native_write_msr_safe
                     [kernel.vmlinux]  [k] ntp_tick_length
     0.00%           [kernel.vmlinux]  [k] rb_erase
     0.00%           [kernel.vmlinux]  [k] tick_sched_timer
     0.00%           [kernel.vmlinux]  [k] unmap_single_vma
     0.00%           [kernel.vmlinux]  [k] update_wall_time
     0.00%           example           [.] f1
    46.24%   -0.08%  example           [.] f2
    53.71%   +0.11%  example           [.] f3
     0.02%           example           [.] main

Signed-off-by: Kan Liang <kan.liang@intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Link: http://lkml.kernel.org/r/1423460384-11645-1-git-send-email-kan.liang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-02-27 10:08:38 -03:00
..
android.txt perf tools: Add info about cross compiling for Android ARM 2012-10-31 12:17:49 -02:00
asciidoc.conf perf_counter tools: Move from Documentation/perf_counter/ to tools/perf/ 2009-06-06 20:33:43 +02:00
Build.txt perf build: Add build documentation 2015-02-12 17:53:00 -03:00
examples.txt perf record: Remove -f/--force option 2013-07-08 17:37:25 -03:00
jit-interface.txt perf symbols: Add description of JIT interface 2012-08-13 14:55:02 -03:00
Makefile perf tools: Implement summary output for 'make install' 2013-10-11 12:18:11 -03:00
manpage-1.72.xsl perf_counter tools: Move from Documentation/perf_counter/ to tools/perf/ 2009-06-06 20:33:43 +02:00
manpage-base.xsl perf_counter tools: Move from Documentation/perf_counter/ to tools/perf/ 2009-06-06 20:33:43 +02:00
manpage-bold-literal.xsl perf_counter tools: Move from Documentation/perf_counter/ to tools/perf/ 2009-06-06 20:33:43 +02:00
manpage-normal.xsl perf_counter tools: Move from Documentation/perf_counter/ to tools/perf/ 2009-06-06 20:33:43 +02:00
manpage-suppress-sp.xsl perf_counter tools: Move from Documentation/perf_counter/ to tools/perf/ 2009-06-06 20:33:43 +02:00
perf-annotate.txt perf annotate: Add basic support to event group view 2013-03-15 13:06:05 -03:00
perf-archive.txt perf archive: Remove duplicated 'runs' in man page 2013-12-09 15:21:45 -03:00
perf-bench.txt perf bench: Add --repeat option 2014-06-19 16:13:15 -03:00
perf-buildid-cache.txt perf buildid-cache: Add new buildid cache if update target is not cached 2015-02-27 10:08:37 -03:00
perf-buildid-list.txt perf report: Accept fifos as input file 2011-12-23 17:01:03 -02:00
perf-data.txt perf data: Add perf data to CTF conversion support 2015-02-25 16:13:12 -03:00
perf-diff.txt perf diff: Support for different binaries 2015-02-27 10:08:38 -03:00
perf-evlist.txt perf evlist: Add --group option 2013-01-31 13:07:49 -03:00
perf-help.txt perf_counter tools: Move from Documentation/perf_counter/ to tools/perf/ 2009-06-06 20:33:43 +02:00
perf-inject.txt perf inject: Add --kallsyms parameter 2014-07-25 12:08:34 -03:00
perf-kmem.txt perf report: Accept fifos as input file 2011-12-23 17:01:03 -02:00
perf-kvm.txt perf Documentation: Fix typos in perf/Documentation 2014-10-15 17:39:02 -03:00
perf-list.txt perf tools: Document parameterized and symbolic events 2015-01-21 13:24:33 -03:00
perf-lock.txt perf lock: Account for lock average wait time 2013-10-09 11:24:01 -03:00
perf-mem.txt perf mem: Enable sampling loads and stores simultaneously 2015-01-21 13:24:31 -03:00
perf-probe.txt perf probe: Update man page 2015-02-06 11:46:36 +01:00
perf-record.txt perf record: Support recording running/enabled time 2015-02-25 12:42:23 -03:00
perf-report.txt perf report: Add --branch-history option 2014-12-01 20:00:31 -03:00
perf-sched.txt perf report: Accept fifos as input file 2011-12-23 17:01:03 -02:00
perf-script-perl.txt perf Documentation: Fix typos in perf/Documentation 2014-10-15 17:39:02 -03:00
perf-script-python.txt perf Documentation: Fix typos in perf/Documentation 2014-10-15 17:39:02 -03:00
perf-script.txt perf tools: Remove EOL whitespaces 2015-01-21 13:24:31 -03:00
perf-stat.txt perf tools: Document parameterized and symbolic events 2015-01-21 13:24:33 -03:00
perf-test.txt perf Documentation: Fix typos in perf/Documentation 2014-10-15 17:39:02 -03:00
perf-timechart.txt perf timechart: Add more options to IO mode 2014-07-10 00:22:54 +02:00
perf-top.txt perf tools: Disable kernel symbol demangling by default 2014-09-17 17:08:09 -03:00
perf-trace.txt perf trace: Add man page entry for --event 2015-02-22 22:22:07 -03:00
perf.txt perf data: Add perf data to CTF conversion support 2015-02-25 16:13:12 -03:00
perfconfig.example perf annotate browser: Read perf config file for settings 2012-05-29 22:06:30 -03:00