From: Jan Chaloupka <jchaloup-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
To: "Michael Kerrisk (man-pages)"
<mtk.manpages-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Cc: linux-man-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
libc-alpha-9JcytcrH/bA+uJoB2kUjGw@public.gmane.org,
myllynen-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org,
drepper-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org,
Peter Schiffer <pschiffe-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Subject: Re: memusage (1) man page
Date: Wed, 03 Sep 2014 08:49:57 +0200 [thread overview]
Message-ID: <5406BA15.5030702@redhat.com> (raw)
In-Reply-To: <54049318.4070404-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
On 09/01/2014 05:39 PM, Michael Kerrisk (man-pages) wrote:
> Hello Jan,
>
> On 09/01/2014 09:18 AM, Jan Chaloupka wrote:
>> Hi,
>>
>> memusage man page.
> Thank you for submitting this page on Peter's behalf. I've merged it.
>
> I edited both pages quite substantially, doing some clean up
> for man-pages style, and also adding many details to the memusage.1
> page. And--because I added various extra explanatory text about the
> output of memusage(1) and showed some of the sample output that
> results from memusage(1)--it seemed more sensible to have the EXAMPLE
> in the memusage.1 page, so I moved it from memusagestat.1 to here.
>
> You can find both updated pages in Git. I've also appended the
> revised memusage.1 page here, in case you or Peter would like to
> check it over.
Thank you very much, Michael.
> Cheers,
>
> Michael
>
> .\" Copyright (c) 2013, Peter Schiffer <pschiffe-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> .\" and Copyright (C) 2014, Michael Kerrisk <mtk.manpages-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> .\"
> .\" %%%LICENSE_START(GPLv2+_DOC_FULL)
> .\" This is free documentation; you can redistribute it and/or
> .\" modify it under the terms of the GNU General Public License as
> .\" published by the Free Software Foundation; either version 2 of
> .\" the License, or (at your option) any later version.
> .\"
> .\" The GNU General Public License's references to "object code"
> .\" and "executables" are to be interpreted as the output of any
> .\" document formatting or typesetting system, including
> .\" intermediate and printed output.
> .\"
> .\" This manual is distributed in the hope that it will be useful,
> .\" but WITHOUT ANY WARRANTY; without even the implied warranty of
> .\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> .\" GNU General Public License for more details.
> .\"
> .\" You should have received a copy of the GNU General Public
> .\" License along with this manual; if not, see
> .\" <http://www.gnu.org/licenses/>.
> .\" %%%LICENSE_END
> .TH MEMUSAGE 1 2013-04-10 "GNU" "Linux user manual"
> .SH NAME
> memusage \- profile memory usage of a program
> .SH SYNOPSIS
> .BR memusage " [\fIoption\fR]... \fIprogram\fR [\fIprogramoption\fR]..."
> .SH DESCRIPTION
> .B memusage
> is a bash script which profiles memory usage of the program,
> .IR program .
> It preloads the
> .B libmemusage.so
> library into the caller's environment (via the
> .B LD_PRELOAD
> environment variable; see
> .BR ld.so (8)).
> The
> .B libmemusage.so
> library traces memory allocation by intercepting calls to
> .BR malloc (3),
> .BR calloc (3),
> .BR free (3),
> and
> .BR realloc (3);
> optionally, calls to
> .BR mmap (2),
> .BR mremap (2),
> and
> .BR munmap (2)
> can also be intercepted.
> .PP
> .B memusage
> can output the collected data in textual form, or it can use
> .BR memusagestat (1)
> (see the
> .B -p
> option, below)
> to create a PNG file containing graphical representation
> of the collected data.
> .SS Memory usage summary
> The "Memory usage summary" line output by
> .BR memusage
> contains three fields:
> .RS 4
> .TP
> \fBheap total\fR
> Sum of \fIsize\fR arguments of all
> .BR malloc (3)
> calls,
> products of arguments (\fInmemb\fR*\fIsize\fR) of all
> .BR calloc (3)
> calls,
> and sum of \fIlength\fR arguments of all
> .BR mmap (2)
> calls.
> In the case of
> .BR realloc (3)
> and
> .BR mremap (2),
> if the new size of an allocation is larger than the previous size,
> the sum of all such differences (new size minus old size) is added.
> .TP
> .B "heap peak"
> Maximum of all \fIsize\fR arguments of
> .BR malloc (3),
> all products of \fInmemb\fR*\fIsize\fR of
> .BR calloc (3),
> all \fIsize\fR arguments of
> .BR realloc (3),
> .I length
> arguments of
> .BR mmap (2),
> and
> \fInew_size\fR arguments of
> .BR mremap (2).
> .TP
> .B "stack peak"
> Before the first call to any monitored function,
> the stack pointer address (base stack pointer) is saved.
> After each function call, the actual stack pointer address is read and
> the difference from the base stack pointer computed.
> The maximum of these differences is then the stack peak.
> .RE
> .PP
> Immediately following this summary line, a table shows the number calls,
> total memory allocated or deallocated,
> and number of failed calls for each intercepted function.
> For
> .BR realloc (3)
> and
> .BR mremap (2),
> the additional field "nomove" shows reallocations that
> changed the address of a block,
> and the additional "dec" field shows reallocations that
> decreased the size of the block.
> For
> .BR realloc (3),
> the additional field "free" shows reallocations that
> caused a block to be freed (i.e., the reallocated size was 0).
>
>
> The "realloc/total memory" of the table output by
> .B memusage
> does not reflect cases where
> .BR realloc (3)
> is used to reallocate a block of memory
> to have a smaller size than previously.
> This can cause sum of all "total memory" cells (excluding "free")
> to be larger than the "free/total memory" cell.
> .SS Histogram for block sizes
> The "Histogram for block sizes" provides a breakdown of memory
> allocations into various bucket sizes.
> .SH OPTIONS
> .TP
> .BI \-n\ name \fR,\ \fB\-\-progname= name
> Name of the program file to profile.
> .TP
> .BI \-p\ file \fR,\ \fB\-\-png= file
> Generate PNG graphic and store it in
> .IR file .
> .TP
> .BI \-d\ file \fR,\ \fB\-\-data= file
> Generate binary data file and store it in
> .IR file .
> .TP
> .BI \-u\fR,\ \fB\-\-unbuffered
> Do not buffer output.
> .TP
> .BI \-b\ size \fR,\ \fB\-\-buffer= size
> Collect
> .I size
> entries before writing them out.
> .TP
> .BI \fB\-\-no-timer
> Disable timer-based
> .RB ( SIGPROF )
> sampling of stack pointer value.
> .TP
> .BI \-m\fR,\ \fB\-\-mmap
> Also trace
> .BR mmap (2),
> .BR mremap (2),
> and
> .BR mmap (2).
> .TP
> .BI \-?\fR,\ \fB\-\-help
> Print help and exit.
> .TP
> .BI \fB\-\-usage
> Print a short usage message and exit.
> .TP
> .BI \-V\fR,\ \fB\-\-version
> Print version information and exit.
> .TP
> The following options only apply when generating graphical output:
> .TP
> .BI \-t\fR,\ \fB\-\-time\-based
> Use time (rather than number of function calls) as the scale for the X axis.
> .TP
> .BI \-T\fR,\ \fB\-\-total
> Also draw a graph of total memory use.
> .TP
> .BI \fB\-\-title= name
> Use
> .I name
> as the title of the graph.
> .TP
> .BI \-x\ size \fR,\ \fB\-\-x\-size= size
> Make the graph
> .I size
> pixels wide.
> .TP
> .BI \-y\ size \fR,\ \fB\-\-y\-size= size
> Make the graph
> .I size
> pixels high.
> .SH EXIT STATUS
> Exit status is equal to the exit status of profiled program.
> .SH EXAMPLE
> Below is a simple program that reallocates a block of
> memory in cycles that rise to a peak before then cyclically
> reallocating the memory in smaller blocks that return to zero.
> After compiling the program and running the following commands,
> a graph of the memory usage of the program can be found in the file
> .IR memusage.png :
>
> .in +4n
> .nf
> $ \fBmemusage --data=memusage.dat ./a.out\fP
> \&...
> Memory usage summary: heap total: 45200, heap peak: 6440, stack peak: 224
> total calls total memory failed calls
> malloc| 1 400 0
> realloc| 40 44800 0 (nomove:40, dec:19, free:0)
> calloc| 0 0 0
> free| 1 440
> Histogram for block sizes:
> 192-207 1 2% ================
> \&...
> 2192-2207 1 2% ================
> 2240-2255 2 4% =================================
> 2832-2847 2 4% =================================
> 3440-3455 2 4% =================================
> 4032-4047 2 4% =================================
> 4640-4655 2 4% =================================
> 5232-5247 2 4% =================================
> 5840-5855 2 4% =================================
> 6432-6447 1 2% ================
> $ \fBmemusagestat memusage.dat memusage.png\fP
> .fi
> .in
> .SS Program source
> .nf
> #include <stdio.h>
> #include <stdlib.h>
>
> #define CYCLES 20
>
> int
> main(int argc, char *argv[])
> {
> int i, j;
> int *p;
>
> printf("malloc: %zd\\n", sizeof(int) * 100);
> p = malloc(sizeof(int) * 100);
>
> for (i = 0; i < CYCLES; i++) {
> if (i < CYCLES / 2)
> j = i;
> else
> j--;
>
> printf("realloc: %zd\\n", sizeof(int) * (j * 50 + 110));
> p = realloc(p, sizeof(int) * (j * 50 + 100));
>
> printf("realloc: %zd\\n", sizeof(int) * ((j+1) * 150 + 110));
> p = realloc(p, sizeof(int) * ((j + 1) * 150 + 110));
> }
>
> free(p);
> exit(EXIT_SUCCESS);
> }
> .fi
> .SH BUGS
> To report bugs, see
> .UR http://www.gnu.org/software/libc/bugs.html
> .UE
> .SH SEE ALSO
> .BR memusagestat (1),
> .BR mtrace (1)
> .BR ld.so (8)
>
>
--
Jan Chaloupka
------------------------------
* Software Engineer *
* ENG Base Operating Systems *
* Red Hat Czech, s. r. o. *
* UTC+1 (CET), jchaloup *
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
prev parent reply other threads:[~2014-09-03 6:49 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-09-01 7:18 memusage (1) man page Jan Chaloupka
[not found] ` <54041DD1.3010009-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2014-09-01 15:39 ` Michael Kerrisk (man-pages)
[not found] ` <54049318.4070404-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2014-09-03 6:49 ` Jan Chaloupka [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=5406BA15.5030702@redhat.com \
--to=jchaloup-h+wxahxf7alqt0dzr+alfa@public.gmane.org \
--cc=drepper-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
--cc=libc-alpha-9JcytcrH/bA+uJoB2kUjGw@public.gmane.org \
--cc=linux-man-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=mtk.manpages-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
--cc=myllynen-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=pschiffe-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.