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
next prev parent 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