Re: Implement lbr-as-callgraph v3

From: Jiri Olsa
Date: Wed Jan 15 2014 - 09:00:28 EST

On Mon, Jan 13, 2014 at 05:04:15PM -0800, Andi Kleen wrote:
> This patchkit implements lbr-as-callgraphs in per freport,
> as an alternative way to present LBR information.
>
> Current perf report does a histogram over the branch edges,
> which is useful to look at basic blocks, but doesn't tell
> you anything about the larger control flow behaviour.
>
> This patchkit adds a new option --branch-history that
> adds the branch paths to the callgraph history instead.
>
> This allows to reason about individual branch paths leading
> to specific samples.
>
> Updates to v1:
> - rebased on perf/core
> - fix various issues
> - rename the option to --branch-history
> - various fixes to display the information more concise
>
> Updates to v3:
> - White space changes
> - Consolidate some patches
> - Update some descriptions
>
> Example output:
>
> % perf record -b -g ./tsrc/tcall
> [ perf record: Woken up 1 times to write data ]
> [ perf record: Captured and wrote 0.044 MB perf.data (~1923 samples) ]
> % perf report --branch-history
> ...
> 54.91% tcall.c:6 [.] f2 tcall
> |
> |--65.53%-- f2 tcall.c:5
> | |
> | |--70.83%-- f1 tcall.c:11
> | | f1 tcall.c:10
> | | main tcall.c:18
> | | main tcall.c:18
> | | main tcall.c:17
> | | main tcall.c:17
> | | f1 tcall.c:13
> | | f1 tcall.c:13
> | | f2 tcall.c:7
> | | f2 tcall.c:5
> | | f1 tcall.c:12
> | | f1 tcall.c:12
> | | f2 tcall.c:7
> | | f2 tcall.c:5
> | | f1 tcall.c:11
>
>

the tui output needs some cleanup:

+ 26.38% :0 [.] _IO_file_xsputn@@GLIBC_2.2.5 libc-2.17.so â
+ 20.43% :0 [.] __strlen_sse2 libc-2.17.so â
+ 15.94% yes.c:83 [.] main yes â
+ 12.16% :0 [.] fputs_unlocked libc-2.17.so â
+ 10.27% stdio.h:107 [.] main yes â
+ 10.09% :0 [.] __GI___mempcpy libc-2.17.so â
+ 2.63% yes.c:84 [.] main yes â
+ 1.71% yes[4012f0] [.] fputs_unlocked@plt yes â
+ 0.25% yes.c:85 [.] main yes â
+ 0.03% [kernel.kallsyms][ffffffff81757eb0] [k] system_call [kernel.kallsyms] â
+ 0.02% :0 [.] _IO_file_write@@GLIBC_2.2.5 libc-2.17.so â
+ 0.02% :0 [.] _IO_do_write@@GLIBC_2.2.5 libc-2.17.so â
+ 0.02% [kernel.kallsyms][ffffffff81758b40] [k] apic_timer_interrupt [kernel.kallsyms] â
+ 0.02% :0 [.] _IO_default_xsputn libc-2.17.so â
+ 0.01% :0 [.] _IO_file_overflow@@GLIBC_2.2.5 libc-2.17.so â
+ 0.01% :0 [.] __GI___libc_write libc-2.17.so â
+ 0.00% [kernel.kallsyms][ffffffff8174f020] [k] page_fault [kernel.kallsyms] â
+ 0.00% :0 [.] getenv libc-2.17.so â
+ 0.00% :0 [.] _nl_load_locale_from_archive libc-2.17.so â
+ 0.00% ld-2.17.so[3407c0af52] [.] _dl_setup_hash ld-2.17.so â
+ 0.00% ld-2.17.so[3407c01420] [.] _start ld-2.17.so


when we cannot get source and line number we could output address not ':0':

- 2.63% yes.c:84 [.] main yes â
- main stdio.h:107 â
- main yes.c:84 â
- main yes.c:84 â
- fputs_unlocked :0 â
- fputs_unlocked :0 â
- _IO_file_xsputn@@GLIBC_2.2.5 :0 â
- 99.78% _IO_file_xsputn@@GLIBC_2.2.5 :0 â
__GI___mempcpy :0 â
__GI___mempcpy :0 â
__GI___mempcpy :0 â
__GI___mempcpy :0


at some samples I could not get see source/address info, just:

0.03% [kernel.kallsyms][ffffffff810a7b4f] [k] __srcu_read_lock [kernel.kallsyms]
|
--- __srcu_read_lock
fsnotify
fsnotify
fsnotify
fsnotify
fsnotify
fsnotify
fsnotify
fsnotify

going to check the patches now ;-)

jirka
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/