Would you need to see the assembly code for that? Would you mind tell me why?
As a starting point it was good for me to read out the GNU Radio documentation of Dawei Shen: http://www.nd.edu/~jnl/sdr/docs/ Hoo Chang. ----- Original Message ---- From: Mikyung Han <[EMAIL PROTECTED]> To: Eric Blossom <[EMAIL PROTECTED]> Cc: discuss-gnuradio@gnu.org Sent: Tuesday, June 24, 2008 12:53:07 AM Subject: Re: [Discuss-gnuradio] oprofile callgraph question On Mon, Jun 23, 2008 at 6:37 PM, Eric Blossom <[EMAIL PROTECTED]> wrote: On Mon, Jun 23, 2008 at 05:42:16PM -0400, Mikyung Han wrote: > BTW, any suggestion for finding out who caused calling the symbols such as > __ieee754_atan2 > and __kernel_cosf, __kernel_sinf, __ieee754_rem_pio2f from libm-2.5.so? What kind of a machine are you running on? uname -a gives 2.6.18-53.1.19.el5 #1 SMP Tue Apr 22 03:01:10 EDT 2008 x86_64 x86_64 x86_64 GNU/Linux The first is of course the guts of atan2. I'm surprised to see it. I think the only place it could be getting called is by way of the complex arg(z) function. We're replace all the others with calls to gr_fast_atan2 Yeah, I think I used 3.1.1 for atsc example. [EMAIL PROTECTED] trunk]$ dgs . '\barg[ \t]*(' ../gnuradio-core/src/lib/general/gr_remez.cc: * Change: ColumnVector x=arg(i).vector_value(); ../gnuradio-core/src/lib/general/gr_remez.cc: * to: ColumnVector x(arg(i).vector_value()); ../gnuradio-core/src/lib/general/gr_mpsk_receiver_cc.cc: return -arg(sample*conj(d_constellation[d_current_const_point])); ../gnuradio-core/src/lib/general/gr_ofdm_frame_sink.cc: float angle = arg(accum_error); ../gnuradio-core/src/lib/general/gr_complex_to_xxx.cc: // out[i] = std::arg (in[i]); ../gnuradio-core/src/lib/general/gr_quadrature_demod_cf.cc: // out[i] = d_gain * arg (product); The last 3 are from the guts of cosf and sinf. What's your test case? Besides, atsc example, I tried profiling benchmark_ofdm_rx/tx.py - profiling benchmark_ofdm_rx.py now gives samples % app name symbol name 226875 22.8492 libgnuradio-core.so.0.0.0 .fcomplex_dotprod_sse64_loop2 79310 7.9875 libgnuradio-core.so.0.0.0 .float_dotprod_sse64_loop2 61142 6.1578 libm-2.5.so __ieee754_rem_pio2f 57980 5.8393 libm-2.5.so __kernel_rem_pio2f 31631 3.1856 libgnuradio-core.so.0.0.0 gr_fir_ccf_simd::filter(std::complex<float> const*) 30052 3.0266 libm-2.5.so __kernel_sinf 26588 2.6778 libm-2.5.so sincosf 25214 2.5394 libm-2.5.so __kernel_cosf 23153 2.3318 libgnuradio-core.so.0.0.0 gr_fast_atan2f(float, float) 22000 2.2157 libgnuradio-core.so.0.0.0 gr_single_threaded_scheduler::main_loop() 20853 2.1002 libgnuradio-core.so.0.0.0 .cleanup 13089 1.3182 libmozjs.so (no symbols) 12536 1.2625 libgnuradio-core.so.0.0.0 gr_ofdm_frame_acquisition::coarse_freq_comp(int, int) 12384 1.2472 libgnuradio-core.so.0.0.0 gr_fft_filter_ccc::work(int, std::vector<void const*, std::allocator< void const*> >&, std::vector<void*, std::allocator<void*> >&) 12080 1.2166 libm-2.5.so __ieee754_hypotf 11982 1.2067 libgnuradio-core.so.0.0.0 gr_ofdm_frame_acquisition::general_work(int, std::vector<int, std::al locator<int> >&, std::vector<void const*, std::allocator<void const*> >&, std::vector<void*, std::allocator<void* > >&) 11453 1.1535 libgnuradio-core.so.0.0.0 gr_fir_ccf_generic::filterN(std::complex<float>*, std::complex<float> const*, unsigned long) 10726 1.0802 libgnuradio-core.so.0.0.0 gr_multiply_cc::work(int, std::vector<void const*, std::allocator<voi d const*> >&, std::vector<void*, std::allocator<void*> >&) 10250 1.0323 libgnuradio-core.so.0.0.0 gr_peak_detector_fb::work(int, std::vector<void const*, std::allocato r<void const*> >&, std::vector<void*, std::allocator<void*> >&) 10135 1.0207 libfftw3f.so.3.1.2 t2fv_32 - for profiling benchmark_ofdm_tx.py gives: 21303 10.8694 libpython2.4.so.1.0 (no symbols) 18889 9.6377 vmlinux-2.6.18-53.1.19.el5 mwait_idle 8283 4.2262 libgnuradio-core.so.0.0.0 gr_single_threaded_scheduler::main_loop() 6579 3.3568 libgnuradio-core.so.0.0.0 gr_multiply_const_cc::work(int, std::vector<void const*, std::allocator<void const*> > &, std::vector<void*, std::allocator<void*> >&) 6266 3.1971 vmlinux-2.6.18-53.1.19.el5 apic_timer_interrupt 5521 2.8170 ehci_hcd (no symbols) 5106 2.6052 libc-2.5.so memcpy 5011 2.5568 libc-2.5.so strcmp 4356 2.2226 libfftw3f.so.3.1.2 t2bv_64 3691 1.8833 libcrypto.so.0.9.8b (no symbols) 3471 1.7710 libgnuradio-core.so.0.0.0 gr_ofdm_mapper_bcv::work(int, std::vector<void const*, std::allocator<void const*> >&, std::vector<void*, std::allocator<void*> >&) 2769 1.4128 vmlinux-2.6.18-53.1.19.el5 schedule 2675 1.3649 oprofiled (no symbols) 2534 1.2929 _usrp1.so usrp1_sink_c::copy_to_usrp_buffer(std::vector<void const*, std::allocator<void const*> >&, int, int, int&, void*, int, int&) 2403 1.2261 libX11.so.6.2.0 (no symbols) 2212 1.1286 libgnuradio-core.so.0.0.0 gr_ofdm_cyclic_prefixer::work(int, std::vector<void const*, std::allocator<void const* > >&, std::vector<void*, std::allocator<void*> >&) 1978 1.0092 libgnuradio-core.so.0.0.0 gr_block_detail::input(unsigned int) > Also .cleanup? ..cleanup is a just a label toward the bottom of the SIMD code. It's unlikely you're getting any serious amount of hits there, it's not in a loop. Eric I guess I need to study the mapping from python code to c files and assembly codes. What would you suggest doing, reading, etc in order to figure out this mapping and skeleton of ofdm example? My eventual goal is to firgure out which piece of code is taking most of the resources for ofdm. Thanks! Mikie
_______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
