Re: [Discuss-gnuradio] Profiling GNU Radio flow graphs

address@hidden:~/Public$ valgrind --tool=callgrind ./dial_tone.py

==30385== Callgrind, a call-graph generating cache profiler

==30385== Using Valgrind-3.6.0.SVN-Debian and LibVEX; rerun with -h for copyright info

==30385== Command: ./dial_tone.py

==30385==

==30385== For interactive control, run 'callgrind_control -h'.

address@hidden:~/Public$ ls -l

total 12

-rw------- 1 yangqing yangqing 0 2012-08-27 16:29 callgrind.out.30385 ** the size of profile data is 0? **

-rwxr-xr-x 1 yangqing yangqing 2006 2012-07-05 16:37 dial_tone.py

-rwxr-xr-x 1 yangqing yangqing 249 2012-08-27 16:22 mainloop.py

drwxr-xr-x 25 yangqing yangqing 4096 2012-08-27 00:23 oprofile

Then I try Oprofile, but also failed :(

address@hidden:~/Public$ sudo opcontrol --init

address@hidden:~/Public$ sudo opcontrol --setup --no-vmlinux

address@hidden:~/Public$ sudo opcontrol --start

ATTENTION: Use of opcontrol is discouraged. Please see the man page for operf.

Using default event: CPU_CLK_UNHALTED:100000:0:1:1

Using 2.6+ OProfile kernel interface.

Using log file /var/lib/oprofile/samples/oprofiled.log

Daemon started.

Profiler running.

address@hidden:~/Public$ sudo opcontrol --reset

Signalling daemon... done

address@hidden:~/Public$ ./dial_tone.py

address@hidden:~/Public$ sudo opcontrol --dump

address@hidden:~/Public$ sudo opcontrol --shutdown

Stopping profiling.

Killing daemon.

address@hidden:~/Public$ opreport -l dial_tone.py

Using /var/lib/oprofile/samples/ for samples directory.

error: no sample files found: profile specification too strict ? ** can't find the profile data? **

and I try

address@hidden:~/Public$ opreport -l|less

CPU: Intel Core/i7, speed 2.794e+06 MHz (estimated)

Counted CPU_CLK_UNHALTED events (Clock cycles when not halted) with a unit mask of 0x00 (No unit mask) count 100000

samples % image name app name symbol name

97293 44.4076 no-vmlinux no-vmlinux /no-vmlinux

43324 19.7744 nvidia_drv.so nvidia_drv.so /usr/lib/nvidia-173/xorg/nvidia_drv.so

23189 10.5842 chromium-browser chromium-browser /usr/lib/chromium-browser/chromium-browser

8264 3.7719 libpixman-1.so.0.18.4 libpixman-1.so.0.18.4 /usr/lib/libpixman-1.so.0.18.4

5975 2.7272 libglib-2.0.so.0.2600.1 libglib-2.0.so.0.2600.1 /lib/libglib-2.0.so.0.2600.1

4409 2.0124 libgobject-2.0.so.0.2600.1 libgobject-2.0.so.0.2600.1 /usr/lib/libgobject-2.0.so.0.2600.1

3811 1.7395 libcairo.so.2.11000.0 libcairo.so.2.11000.0 /usr/lib/libcairo.so.2.11000.0

3001 1.3698 libpangoft2-1.0.so.0.2800.2 libpangoft2-1.0.so.0.2800.2 /usr/lib/libpangoft2-1.0.so.0.2800.2

2714 1.2388 python2.6 python2.6 /usr/bin/python2.6

2083 0.9507 libdbus-1.so.3.5.2 libdbus-1.so.3.5.2 /lib/libdbus-1.so.3.5.2

2003 0.9142 libwfb.so libwfb.so /usr/lib/xorg/modules/libwfb.so

1969 0.8987 Xorg Xorg /usr/bin/Xorg

1894 0.8645 libgtk-x11-2.0.so.0.2200.0 libgtk-x11-2.0.so.0.2200.0 /usr/lib/libgtk-x11-2.0.so.0.2200.0

1732 0.7905 libgnuradio-core-3.5.2git.so.0.0.0 libgnuradio-core-3.5.2git.so.0.0.0 gr_sig_source_f::work(int, std::vector<void const*, std::allocat

or<void const*> >&, std::vector<void*, std::allocator<void*> >&)

1578 0.7202 libgdk-x11-2.0.so.0.2200.0 libgdk-x11-2.0.so.0.2200.0 /usr/lib/libgdk-x11-2.0.so.0.2200.0

1378 0.6290 libpango-1.0.so.0.2800.2 libpango-1.0.so.0.2800.2 /usr/lib/libpango-1.0.so.0.2800.2

962 0.4391 [vdso] (tgid:1237 range:0xb77ec000-0xb77ed000) Xorg [vdso] (tgid:1237 range:0xb77ec000-0xb77ed000)

802 0.3661 libpthread-2.12.1.so libpthread-2.12.1.so pthread_mutex_lock

690 0.3149 libc-2.12.1.so libc-2.12.1.so __memcpy_ssse3_rep

640 0.2921 libpthread-2.12.1.so libpthread-2.12.1.so __pthread_mutex_unlock_usercnt

459 0.2095 libX11.so.6.3.0 libX11.so.6.3.0 /usr/lib/libX11.so.6.3.0

444 0.2027 libQtGui.so.4.7.0 libQtGui.so.4.7.0 /usr/lib/libQtGui.so.4.7.0

409 0.1867 libc-2.12.1.so libc-2.12.1.so _int_malloc

354 0.1616 librt-2.12.1.so librt-2.12.1.so clock_gettime

341 0.1556 libc-2.12.1.so libc-2.12.1.so __memset_sse2_rep

337 0.1538 libgnuradio-audio-3.5.2git.so.0.0.0 libgnuradio-audio-3.5.2git.so.0.0.0 audio_alsa_sink::work_s32(int, std::vector<void const*, std::allocator<void const*> >&, std::vector<void*, std::allocator<void*> >&)

332 0.1515 anon (tgid:25090 range:0x4fc0a000-0x4fcff000) chromium-browser anon (tgid:25090 range:0x4fc0a000-0x4fcff000)

322 0.1470 libc-2.12.1.so libc-2.12.1.so __strcmp_sse4_2

237 0.1082 oprofiled oprofiled odb_update_node_with_offset

230 0.1050 libpangocairo-1.0.so.0.2800.2 libpangocairo-1.0.so.0.2800.2 /usr/lib/libpangocairo-1.0.so.0.2800.2

224 0.1022 libpthread-2.12.1.so libpthread-2.12.1.so pthread_getspecific

223 0.1018 libstdc++.so.6.0.14 libstdc++.so.6.0.14 /usr/lib/libstdc++.so.6.0.14

214 0.0977 libxcb.so.1.1.0 libxcb.so.1.1.0 /usr/lib/libxcb.so.1.1.0

210 0.0959 metacity metacity /usr/bin/metacity

204 0.0931 libc-2.12.1.so libc-2.12.1.so fgetc

180 0.0822 ld-2.12.1.so ld-2.12.1.so do_lookup_x

168 0.0767 libc-2.12.1.so libc-2.12.1.so __i686.get_pc_thunk.bx

165 0.0753 libc-2.12.1.so libc-2.12.1.so _IO_vfscanf

163 0.0744 ibus-daemon ibus-daemon /usr/bin/ibus-daemon

... ....

I can't find my thread of dial_tone.py. I guess I use Oprofile in the wrong way, could you give me some tips?

Sincerely,

Yang, Qing

Information Engineering, CUHK

2012/8/24 Felix Wunsch <address@hidden>

Hi,

using oprofile is quite easy. Basically you configure your profiler, start it, start your application, kill it after some time, kill the profiler and look at the results. You don't have to set any special compiler flags. However, if you want to get annotated source, you need to compile with debug symbols. All this can be done in a simple shell script what makes it very convenient.

For more details I'd like to refer you to the oprofile manual [1] as it's good to read and not too extensive.

In my case, most CPU time is used by memmove (about 40%!). Unfortunately, I wasn't able to figure out where it gets called from.

Best regards,
Felix

[1] http://oprofile.sourceforge.net/doc/index.html

Am 24.08.2012 08:28, schrieb Qing Yang:

Hi Felix,

Currently I also need to profile and optimize my system. Now I just add some some sentences to print the processing time of each block, but this is definitely not a good method. Could you describe your profiling method in more details, perhaps your results can be a reference for me.

Hi Tim,

Have you tried this Kcachegrind on GNU Radio flow graphs? Does it works well? I want to see the profile of each gr module (in python).

--

Yang, Qing
Information Engineering, CUHK

2012/8/24 <address@hidden>

Hi all,

I am currently trying to optimize the performance of my DRM transmitter
and for this purpose I want to profile my flow graphs.

After some googling and searching the archives I stumbled upon oprofile
which looks quite nice to me. However, a first try did not really provide
very significant results. But that could also be due to misconfiguration,
I did not read the manual very carefully...

Just wanted to know if there are other/better solutions for profiling you
would recommend. Any comments are appreciated!

Best regards,
Felix

_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio

From:	Qing Yang
Subject:	Re: [Discuss-gnuradio] Profiling GNU Radio flow graphs
Date:	Mon, 27 Aug 2012 16:46:08 +0800