Sukadev Bhattiprolu <[email protected]> wrote on 05/09/2014
09:46:38 PM:
> From: Sukadev Bhattiprolu <[email protected]>
> To: Arnaldo Carvalho de Melo <[email protected]>,
> Cc: [email protected], Anton Blanchard
> <[email protected]>, [email protected], Michael Ellerman
> <[email protected]>, Maynard Johnson/Rochester/IBM@IBMUS,
> [email protected]
> Date: 05/09/2014 09:46 PM
> Subject: [PATCH 1/1] powerpc/perf: Adjust callchain based on DWARF debug
info
>
> [PATCH 1/1] powerpc/perf: Adjust callchain based on DWARF debug info
Acked-by: Maynard Johnson <[email protected]>
Reviewed and tested. Thanks, Suka.
-Maynard
>
> When saving the callchain on Power, the kernel conservatively saves
excess
> entries in the callchain. A few of these entries are needed in some cases
> but not others.
>
> Eg: the value in the link register (LR) is needed only when it holds the
> return address of a function. At other times it must be ignored.
>
> If the unnecessary entries are not ignored, we end up with duplicate arcs
> in the call-graphs.
>
> Use DWARF debug information to ignore the unnecessary entries.
>
> Callgraph before the patch:
>
> 14.67% 2234 sprintft libc-2.18.so [.] __random
> |
> --- __random
> |
> |--61.12%-- __random
> | |
> | |--97.15%-- rand
> | | do_my_sprintf
> | | main
> | | generic_start_main.isra.0
> | | __libc_start_main
> | | 0x0
> | |
> | --2.85%-- do_my_sprintf
> | main
> | generic_start_main.isra.0
> | __libc_start_main
> | 0x0
> |
> --38.88%-- rand
> |
> |--94.01%-- rand
> | do_my_sprintf
> | main
> | generic_start_main.isra.0
> | __libc_start_main
> | 0x0
> |
> --5.99%-- do_my_sprintf
> main
> generic_start_main.isra.0
> __libc_start_main
> 0x0
>
> Callgraph after the patch:
>
> 14.67% 2234 sprintft libc-2.18.so [.] __random
> |
> --- __random
> |
> |--95.93%-- rand
> | do_my_sprintf
> | main
> | generic_start_main.isra.0
> | __libc_start_main
> | 0x0
> |
> --4.07%-- do_my_sprintf
> main
> generic_start_main.isra.0
> __libc_start_main
> 0x0
>
> TODO: For split-debug info objects like glibc, we can only determine
> the call-frame-address only when both .eh_frame and .debug_info
> sections are available. We should be able to determin the CFA
> even without the .eh_frame section.
>
> Thanks to Ulrich Weigand for help with DWARF debug information.
>
> Fix suggested by Anton Blanchard.
>
> Reported-by: Maynard Johnson <[email protected]>
> Signed-off-by: Sukadev Bhattiprolu <[email protected]>
> ---
> tools/perf/arch/powerpc/Makefile | 1 +
> tools/perf/arch/powerpc/util/adjust-callchain.c | 278 +++++++++++++
> +++++++++++
> tools/perf/config/Makefile | 5 +
> tools/perf/util/callchain.h | 12 +
> tools/perf/util/machine.c | 16 +-
> 5 files changed, 310 insertions(+), 2 deletions(-)
> create mode 100644 tools/perf/arch/powerpc/util/adjust-callchain.c
>
[snip]
_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev
