[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20091117125735.GA3156@ghostprotocols.net>
Date: Tue, 17 Nov 2009 10:57:35 -0200
From: Arnaldo Carvalho de Melo <acme@...radead.org>
To: Ingo Molnar <mingo@...e.hu>
Cc: linux-kernel@...r.kernel.org,
Frederic Weisbecker <fweisbec@...il.com>,
Mike Galbraith <efault@....de>,
Paul Mackerras <paulus@...ba.org>,
Peter Zijlstra <peterz@...radead.org>
Subject: Re: [PATCH 1/2] perf top: Introduce --hide_{user,kernel}_symbols
Em Tue, Nov 17, 2009 at 07:33:32AM +0100, Ingo Molnar escreveu:
>
> btw., the user symbol printouts of perf top cause some display problems
> - such as below. The DSO printout is way too long to fit into 92 cols,
> wrapping around in my terminal and making much of the output unreadable.
>
> Could we cut it by default please?
Was in my TODO list, just forgot to get back to it, will use that
TIOCWINCH ioctl to get the current width and react to it via the signal
too, etc.
> Also, it would be nice to only print the basename() portion of the DSO.
> Both the /lib64 and the /opt/crossgcc/cross/libexec/ portion is
> repetitive and not really important in understanding perf top output.
> There should perhaps be a 'dso-abs' symbol column that prints out the
> full pathname?
>
> Ingo
>
> ------------->
>
> $ echo $COLUMNS $LINES
> 92 28
>
> $ perf top
>
> ------------------------------------------------------------------------------
> PerfTop: 13676 irqs/sec kernel:14.3% [1000Hz cycles], (all, 16 CPUs)
> ------------------------------------------------------------------------------
>
> samples pcnt function DSO
> _______ _____ ________________________________ ________________
>
> 2315.00 4.6% ht_lookup_with_hash /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 2202.00 4.4% _cpp_lex_direct /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 1490.00 3.0% _int_malloc /lib64/libc-2.10.90.so
> 1172.00 2.3% _cpp_clean_line /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 887.00 1.8% lex_identifier /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 845.00 1.7% clear_page_c [kernel]
> 836.00 1.7% cpp_get_token /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 795.00 1.6% ggc_alloc_stat /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 675.00 1.3% cpp_output_token /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 675.00 1.3% __GI_memset /lib64/libc-2.10.90.so
> 622.00 1.2% _int_free /lib64/libc-2.10.90.so
> 609.00 1.2% page_fault [kernel]
> 568.00 1.1% gimplify_expr /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 567.00 1.1% _cpp_lex_token /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 440.00 0.9% yylex /home/mingo/tip/scripts/genksyms/genksyms
> 423.00 0.8% __GI___libc_malloc /lib64/libc-2.10.90.so
> 417.00 0.8% walk_tree_1 /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
> 416.00 0.8% __GI_memcpy /lib64/libc-2.10.90.so
> 377.00 0.7% enter_macro_context /opt/crossgcc/cross/libexec/gcc/x86_64-linux/4.4.2/cc1
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists