Tom Rondeau [Sat, 16 Oct 2010 21:00:11 +0000 (17:00 -0400)]
Work on examples for the synthesize filterbank block. The cleans up the simple example and adds a new example that synthesizes a number of signals and then channelizes them again. It displays the synthesized PSD as well as the PSD and time waveform of one of the channels that's specified in teh code.
Tom Rondeau [Sat, 16 Oct 2010 15:13:53 +0000 (11:13 -0400)]
Adding a FIR filter implemented with its own internal buffer. This one keeps its own delay line and just takes in input samples instead of a pointer to an external buffer.
The synthesis filter is being updated to use the new FIR implementation.
Tom Rondeau [Thu, 7 Oct 2010 00:50:25 +0000 (20:50 -0400)]
Fixed missing set_relative_rate in these two blocks. The others don't actually do it, even though it's counter-intuitive for the pfb_decimate, which is a sync_block (decimation actually care of in the stream_to_streams).
* maint:
Fixed setting of USB_LIBS for *win*
Fix so that non-Darwin OSs use USB_LIBS correctly for linking
New way of checking for various LIBUSB names; for legacy version, make sure the symbol 'usb_debug' exists (so-as to not use the 'compat' version).
Finally, the simple fix for the qtgui issues; also, changed the naming scheme output from _moc to .moc files because I think this looks cleaner.
Removing Waterfall3DPlot. The qwt_plot3d is too much of a hassle to deal with and the plotting is not that stable even when it does work. This does not change the API.
Fixed gitignore to reflect changes in moc/ui file naming.
Fixes a lot of warnings by cleaning up namespace issues.
Finally, the simple fix for the qtgui issues; also, changed the naming scheme output from _moc to .moc files because I think this looks cleaner.
Removing qwtplot3d dependency and fixing a missed include in waterfallGlobalData.h.
Removing Waterfall3DPlot. The qwt_plot3d is too much of a hassle to deal with and the plotting is not that stable even when it does work. This does not change the API.
* tom/qtfix:
Finally, the simple fix for the qtgui issues; also, changed the naming scheme output from _moc to .moc files because I think this looks cleaner.
Removing Waterfall3DPlot. The qwt_plot3d is too much of a hassle to deal with and the plotting is not that stable even when it does work. This does not change the API.
Fixed gitignore to reflect changes in moc/ui file naming.
Fixes a lot of warnings by cleaning up namespace issues.
Finally, the simple fix for the qtgui issues; also, changed the naming scheme output from _moc to .moc files because I think this looks cleaner.
Removing qwtplot3d dependency and fixing a missed include in waterfallGlobalData.h.
Removing Waterfall3DPlot. The qwt_plot3d is too much of a hassle to deal with and the plotting is not that stable even when it does work. This does not change the API.
Merge Michael Dicken's USB fixes after rebasing on maint
Fixed setting of USB_LIBS for *win*
Fix so that non-Darwin OSs use USB_LIBS correctly for linking
New way of checking for various LIBUSB names; for legacy version, make sure the symbol 'usb_debug' exists (so-as to not use the 'compat' version).
Tom Rondeau [Mon, 13 Sep 2010 18:49:00 +0000 (14:49 -0400)]
Removing Waterfall3DPlot. The qwt_plot3d is too much of a hassle to deal with and the plotting is not that stable even when it does work. This does not change the API.
Tom Rondeau [Mon, 13 Sep 2010 18:49:00 +0000 (14:49 -0400)]
Removing Waterfall3DPlot. The qwt_plot3d is too much of a hassle to deal with and the plotting is not that stable even when it does work. This does not change the API.
Tom Rondeau [Sun, 3 Oct 2010 19:00:44 +0000 (15:00 -0400)]
Adding XML output to Python unittests.
gr_unittest.py was updated to handle the path and file creation and run the XML runner
gr_xmlrunner.py is a public domain program written by Sebastian Rittau <srittau@jroger.in-berlin.de> (THANKS!) to extend Python's unittest module to handle XML JUnit format output. This was renamed from xmlrunner.py if any future work is done on it from our project as well as to keep the namespace clean and clear.
qa_skiphead.py is a first test program for the XML output. Creates an outputfile $HOME/.gnuradio/unittest/python/test_skiphead.xml.
Tom Rondeau [Sat, 2 Oct 2010 21:16:04 +0000 (17:16 -0400)]
Moving XML output files from cppunit tests to $HOME/.gnuradio/unittests.
This also adds a new utility gr_unittests.h, which sets up the path for output files.
Tom Rondeau [Thu, 30 Sep 2010 20:08:53 +0000 (16:08 -0400)]
Adding an XML outputter for the CPP Unit tests. This is to a) store the output information but also b) for integration with Hudson for logging and displaying the results during the build stages. This only covers a few cases so far and I need to define a better place to save the output files.
Eric Blossom [Tue, 14 Sep 2010 19:43:45 +0000 (12:43 -0700)]
Merge branch 'maint'
* maint:
Avoid divide by zero in persistence code when using default args
Update config.guess and config.sub in gr-howoto-write-a-block too
Update config.guess and config.sub from canonical home
Update script to pull config.* from git repo
This change introduced a segfault when calling into the gr_prefs module
during the gri_fft helper class constructor. Until this is debugged,
the functionality is being reverted.
Alexandru Csete [Thu, 22 Jul 2010 01:27:46 +0000 (18:27 -0700)]
gnuradio-core: export missing public methods in gr.agc2
Alexandru Csete:
The gr_agc2 block has public methods to set and get the AGC levels and
rates; however, only a few of these interfaces are exported to Python
and none of them are available in GRC.
The attached patch adds the missing public methods to gri_agc2_cc.i
and gri_agc2_ff.i and defines the corresponding callbacks in the
gr_agc2_xx.xml GRC block.
* trondeau/qtclock:
From my reading, we should never need to define the timespec struct (so we will wait until this bites me).
Checking for clock_gettime and timespec defined; use gettimeofday if not, but put it into a timespec structure.
Reformatting highResTimeFunction.h for better readibility.
gr-wxgui: fix non-gl scope sink only updating one channel
This fixes a bug where, when using nongl scopesink, only the real part
of a complex input is shown. After applying this patch, both Ch1 and
Ch2 can be seen on the scope display.
* Adds new scope window trigger, TRIG_MODE_STRIPCHART
* Allows config file selection of whether background window sinks still run
* Allows config file selection of default trigger mode
* Allows config file selection of default waterfall sink color mode
[wxgui]
#Setting this to true prevents the window sinks from stopping when they
#are not displayed, as when they are in a notebook with a hidden tab.
run_always = False
This new method on usrp2.sink_* causes the first TX sample data to be
sent at the FPGA clock time specified, with all further data immediately
following.
u = usrp2.sink_32fc() # or 16sc
...configure sink here...
u.sync_to_pps()
...delay a second for PPS to have happened
u.start_streaming_at(int(100e6)) # start TX stream one second later
...start flowgraph here...
If this function is not called, all transmit data will be sent
immediately (the prior behavior).
* maint:
Fixed WBX RX PLL enable
Refactor WBX and adf4350 to avoid passing usrp pointer
Clean up annoying class structure in wbx
Fixing wbx to use _refclk_freq()
Typo in error message
* fix/wbx_refclock:
Fixed WBX RX PLL enable
Refactor WBX and adf4350 to avoid passing usrp pointer
Clean up annoying class structure in wbx
Fixing wbx to use _refclk_freq()
Typo in error message