Re: perf report does not get along with perf record -A

[David Ahern]

On 6/25/12 6:10 PM, Martin Rosenberg wrote:
> I'm working under the assumption that if I want to collect statistics from multiple runs of a program, I can do this by running
> perf record prog; perf record -A prog; perf record -A prog
>
> I've found that running perf record twice results in perf report hanging, and running it more than twice results in perf report first spewing a message about invalid formats, then crashing.
> A log of the session, as well as the binary, its input and perf.data can be found at:
>
> http://www.club.cc.cmu.edu/~mjrosenb/ion/perf_crash.tar.gz


Append has been broken for a long time (v2.6.36 last working version).

git bisect points to:

commit 55b44629f599a2305265ae9c77f9d9bcfd6ddc17
Author: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Date:   Tue Nov 30 17:49:46 2010 +0000

    perf session: Use sensible mmap size

    On 64bit we can map the whole file in one go, on 32bit we can at 
least map
    32MB and not map/unmap tiny chunks of the file.

    Base the progress bar on 1/16 of the data size.

    Preparatory patch to get rid of the malloc/memcpy/free of trace data.

    Cc: Ingo Molnar <mingo@xxxxxxx>
    Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
    Cc: Frederic Weisbecker <fweisbec@xxxxxxxxx>
    LKML-Reference: <20101130163820.213687773@xxxxxxxxxxxxx>
    Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
    Signed-off-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>

:040000 040000 a0d7ddfb30b562a548052b12b142b579aa888886 
dd4e368cbdf0ba24126afcc0fe42feea3b68f4a7 M	tools


The commit right before that one (d65132) dumps some errors, but at 
least does not hang -- spinning in the event processing.

David
--
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/