Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: "Agovic, Sanimir" <sanimir.agovic@intel.com>
To: 'Yao Qi' <yao@codesourcery.com>,
	"gdb-patches@sourceware.org"	<gdb-patches@sourceware.org>
Subject: RE: [RFC 3/3] Test on solib load and unload
Date: Wed, 28 Aug 2013 11:31:00 -0000	[thread overview]
Message-ID: <0377C58828D86C4588AEEC42FC3B85A717658713@IRSMSX105.ger.corp.intel.com> (raw)
In-Reply-To: <521D7BCA.10806@codesourcery.com>

lgtm for the python part. I lack some tcl/dejagnu skills to review the other part.

See comments below.

 -Sanimir

> -----Original Message-----
> From: gdb-patches-owner@sourceware.org [mailto:gdb-patches-owner@sourceware.org] On Behalf
> Of Yao Qi
> Sent: Wednesday, August 28, 2013 06:26 AM
> To: gdb-patches@sourceware.org
> Subject: Re: [RFC 3/3] Test on solib load and unload
> 
> diff --git a/gdb/testsuite/gdb.perf/solib.c b/gdb/testsuite/gdb.perf/solib.c
> new file mode 100644
> index 0000000..948b286
> --- /dev/null
> +++ b/gdb/testsuite/gdb.perf/solib.c
> @@ -0,0 +1,79 @@
[...]
> +void
> +do_test (int number)
> +{
> +  void **handles;
> +  char libname[40];
> +  int i;
> +
> +  handles = malloc (sizeof (void *) * number);
> +
> +  for (i = 0; i < number; i++)
> +    {
> +      sprintf (libname, "solib-lib%d", i);
> +      handles[i] = dlopen (libname, RTLD_LAZY);
> +      if (handles[i] == NULL)
> +	{
> +	  printf ("ERROR on dlopen %s\n", libname);
> +	  return;
> +	}
> +    }
> +
> +  for (i = 0; i < number; i++)
> +    {
> +      char funname[20];
> +      void *p;
> +
> +      sprintf (funname, "shr%d", i);
> +      p = dlsym (handles[i], funname);
>
Does dlsym has any perf impact on the debugger?

> diff --git a/gdb/testsuite/gdb.perf/solib.exp b/gdb/testsuite/gdb.perf/solib.exp
> new file mode 100644
> index 0000000..8e7eaf8
> --- /dev/null
> +++ b/gdb/testsuite/gdb.perf/solib.exp
> @@ -0,0 +1,86 @@
[...]
> +
> +set remote_python_file [gdb_remote_download host ${srcdir}/${subdir}/${testfile}.py]
> +
> +# Set sys.path for module perftest.
> +gdb_test_no_output "python import os, sys"
> +gdb_test_no_output "python sys.path.insert\(0,
> os.path.abspath\(\"${srcdir}/${subdir}/lib\"\)\)"
> +
> +gdb_test_no_output "python exec (open ('${remote_python_file}').read ())"
>
The lines above seem pretty generic and could move into an own proc.

> +
> +gdb_test_no_output "python SolibLoadUnload\($SOLIB_NUMBER\)"
> +
> +# Call the convenience function registered by python script.
> +send_gdb "call \$perftest()\n"
>
Not sure if a convenience function is necessary:
python SolibLoadUnload().execute_test()
could do the job as well.

> diff --git a/gdb/testsuite/gdb.perf/solib.py b/gdb/testsuite/gdb.perf/solib.py
> new file mode 100644
> index 0000000..7cc9c4a
> --- /dev/null
> +++ b/gdb/testsuite/gdb.perf/solib.py
> @@ -0,0 +1,48 @@
> +# Copyright (C) 2013 Free Software Foundation, Inc.
> +
> +# This program is free software; 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 3 of the License, or
> +# (at your option) any later version.
> +#
> +# This program 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 program.  If not, see <http://www.gnu.org/licenses/>.
> +
> +# This test case is to test the speed of GDB when it is handling the
> +# shared libraries of inferior are loaded and unloaded.
> +
> +import gdb
> +import time
> +
> +from perftest import perftest
> +
> +class SolibLoadUnload(perftest.SingleVariableTestCase):
> +    def __init__(self, solib_number):
> +        super (SolibLoadUnload, self).__init__ ("solib")
> +        self.solib_number = solib_number
> +
> +    def execute_test(self):
> +        num = self.solib_number
> +        iteration = 5;
> +
> +        # Warm up.
> +        do_test_command = "call do_test (%d)" % num
> +        gdb.execute (do_test_command)
> +        gdb.execute (do_test_command)
> +
> +        while num > 0 and iteration > 0:
> +            do_test_command = "call do_test (%d)" % num
>
This may raise a TypeError if num % 2 != 0

> +
> +            start_time = time.clock()
> +            gdb.execute (do_test_command)
> +            elapsed_time = time.clock() - start_time
> +
> +            self.result.record (num, elapsed_time)
> +
> +            num = num / 2
> +            iteration -= 1
>
You may consider observing solibs loads/unloads to compute the time
between the events.
Can you re-run the sample with turned off garbage collector? It may
cause some jitter if turned on.

 -Sanimir

Intel GmbH
Dornacher Strasse 1
85622 Feldkirchen/Muenchen, Deutschland
Sitz der Gesellschaft: Feldkirchen bei Muenchen
Geschaeftsfuehrer: Christian Lamprechter, Hannes Schwaderer, Douglas Lusk
Registergericht: Muenchen HRB 47456
Ust.-IdNr./VAT Registration No.: DE129385895
Citibank Frankfurt a.M. (BLZ 502 109 00) 600119052

  reply	other threads:[~2013-08-28 11:31 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-08-14 13:01 [RFC] GDB performance testing infrastructure Yao Qi
2013-08-21 20:39 ` Tom Tromey
2013-08-27  6:21   ` Yao Qi
2013-08-27 13:49 ` Agovic, Sanimir
2013-08-28  3:04   ` Yao Qi
2013-09-19  0:36     ` Doug Evans
2013-08-28  4:17 ` [RFC 0/3] GDB Performance testing Yao Qi
2013-08-28  4:17   ` [RFC 3/3] Test on solib load and unload Yao Qi
2013-08-28  4:27     ` Yao Qi
2013-08-28 11:31       ` Agovic, Sanimir [this message]
2013-09-03  1:59         ` Yao Qi
2013-09-03  6:33           ` Agovic, Sanimir
2013-09-02 15:24       ` Blanc, Nicolas
2013-09-03  2:04         ` Yao Qi
2013-09-03  7:50           ` Blanc, Nicolas
2013-09-19 22:45       ` Doug Evans
2013-09-20 19:19         ` Tom Tromey
2013-10-05  0:34           ` Doug Evans
2013-10-07 16:31             ` Tom Tromey
2013-09-22  6:25         ` Yao Qi
2013-09-23  0:14           ` Doug Evans
2013-09-24  2:31             ` Yao Qi
2013-10-05  0:37               ` Doug Evans
2013-09-20 19:14       ` Tom Tromey
2013-08-28  4:17   ` [RFC 2/3] Perf test framework Yao Qi
2013-08-28  9:57     ` Agovic, Sanimir
2013-09-03  1:45       ` Yao Qi
2013-09-03  6:38         ` Agovic, Sanimir
2013-09-19 19:09     ` Doug Evans
2013-09-20  8:04       ` Yao Qi
2013-09-20 16:51         ` Doug Evans
2013-09-22  2:54           ` Yao Qi
2013-09-22 23:14             ` Doug Evans
2013-09-20 17:12         ` Doug Evans
2013-08-28  4:17   ` [RFC 1/3] New make target 'check-perf' and new dir gdb.perf Yao Qi
2013-08-28  9:40     ` Agovic, Sanimir
2013-09-19 17:47     ` Doug Evans
2013-09-20 19:00       ` Tom Tromey
2013-09-20 18:59     ` Tom Tromey
2013-09-19 17:25   ` [RFC 0/3] GDB Performance testing Doug Evans

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=0377C58828D86C4588AEEC42FC3B85A717658713@IRSMSX105.ger.corp.intel.com \
    --to=sanimir.agovic@intel.com \
    --cc=gdb-patches@sourceware.org \
    --cc=yao@codesourcery.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox