On Wed, Sep 16, 2015 at 8:20 PM, Martin Braun
wrote:
> On 16.09.2015 13:29, West, Nathan wrote:
> > There is a volk_32f_s32f_multiply_32f. It doesn't operate in-place, but
> > almost none of the VOLK kernels do. I think it's safe to give the same
> > output buffer as input buffer. (I've heard tha
Rich,
I started down that path, ran into trouble (VOLK needed Cheetah or
similar, etc.) and got some advice to follow the pybombs route, which
generated its own trouble.
Academically, of course, I think it would be great to iron out the
Windows issues. As I said, if there are suggestions I'd be wi
On Wed, Sep 16, 2015 at 9:04 PM, Robert Durkacz
wrote:
> Marcus Leech wrote "You can put "system dependencies goo" inside every
> Makefile you author..."
>
> I meant to say that I would write makefiles the way gnu make
> recommends that just work on my system.
>
> It would be nice to hear from so
Mike,
Have you stepped through the procedures outlined in the following two links
already:
https://gnuradio.org/redmine/projects/gnuradio/wiki/CygwinInstallMain
https://gnuradio.org/redmine/projects/gnuradio/wiki/CygwinGettingStarted
They are dated, but still helps guide you a fair amount of the
Marcus Leech wrote "You can put "system dependencies goo" inside every
Makefile you author..."
I meant to say that I would write makefiles the way gnu make
recommends that just work on my system.
It would be nice to hear from someone who wants to do the same on a
different system and then we coul
On 16.09.2015 13:29, West, Nathan wrote:
> There is a volk_32f_s32f_multiply_32f. It doesn't operate in-place, but
> almost none of the VOLK kernels do. I think it's safe to give the same
> output buffer as input buffer. (I've heard that doing stuff in-place is
> noticeably better, but I've never t
On 16.09.2015 16:58, Mike Gilmer wrote:
> I wrestled with which way to go originally - figure out how to install
> linux,and in theory have an easier time of gnuradio, or try the
> windows procedure. A co-worker thought the windows path would be
> better but these kinds of problems are exactly why
If I understood the process better I'd be willing to help ferret out
the right procedure for cygwin, but I don't think I'd be very
effective at it. I'm willing to try a few suggestions though... I
mean I'm sure to learn something!
I wrestled with which way to go originally - figure out how to ins
On 16.09.2015 15:38, Chris Kuethe wrote:
> I don't disagree. I'm happy to write the write, I just don't have any
> cygwin machines to test with.
...PyBOMBS2 has hooks for this kind of thing. I'm trying my best to not
make this become the Duke Nukem Forever of GNU Radio :)
M
>
> On Wed, Sep 16,
I don't disagree. I'm happy to write the write, I just don't have any
cygwin machines to test with.
On Wed, Sep 16, 2015 at 3:34 PM, Tom Rondeau wrote:
> On Wed, Sep 16, 2015 at 6:12 PM, Chris Kuethe
> wrote:
>>
>> It's been a decade since I last played with cygwin, but under the hood
>> pybomb
On Wed, Sep 16, 2015 at 6:12 PM, Chris Kuethe
wrote:
> It's been a decade since I last played with cygwin, but under the hood
> pybomb can do things like "apt-get install gcc" or "rpm -i fftw-devel"
> - I'm trying to figure out what the command is to to get cygwin to go
> out and download/install
Do I need to use gnuradio - companion for transmission and reception of the
data packets or the benchmark_tx.py commands from the terminal would be
enough for that to happen. I am unable to sort out. Kindly help.
Regards,
Dave
On Sep 16, 2015 2:59 PM, "Rama V" wrote:
> -- Forwarded messa
It's been a decade since I last played with cygwin, but under the hood
pybomb can do things like "apt-get install gcc" or "rpm -i fftw-devel"
- I'm trying to figure out what the command is to to get cygwin to go
out and download/install some package from the internet, and how you
can query what pac
I'm new to this. ..are you asking me to run that install command. .. or?
Mike
On Sep 16, 2015 5:03 PM, "Chris Kuethe" wrote:
> This is what PyBOMBS does...
>
> https://github.com/gnuradio/pybombs/blob/0382f9253a44135677b656ef08ba438f57f65625/mod_pybombs/sysutils.py#L393
>
> https://github.com/gn
On Wed, 2015-09-16 at 16:29 -0400, West, Nathan wrote:
> On Wed, Sep 16, 2015 at 3:58 PM, Dennis Glatting
> wrote:
> >
> > It would also be nice to have a VOLK kernel that multiples a vector
> > by a constant:
> >
> >void
> >volk_32f_s32f_multiply( float* vecbuffer,
> >
Congratulations, you now know the most of anyone about gnuradio on
windows. Most of us here seem to use some sort of linux host.
PyBOMBS thinks it has to compile gcc, make, binutils, math libraries,
... because it can't find acceptable binaries. Does cygwin have some
sort of command-line package m
This is what PyBOMBS does...
https://github.com/gnuradio/pybombs/blob/0382f9253a44135677b656ef08ba438f57f65625/mod_pybombs/sysutils.py#L393
https://github.com/gnuradio/pybombs/blob/dc593faf9e1557133c5801fe4aa58198e34407db/mod_pybombs/recipe.py
if you could do something like "cygpkg install fftw gc
I'm running a cygwin shell.
Mike
On Wed, Sep 16, 2015 at 4:33 PM, Mike Gilmer wrote:
> Thanks guys.
>
> The output (somewhat abbreviated) is at http://pastebin.com/BqvwB58y
>
> I am running Windows 8.1
>
> Mike
>
> On Wed, Sep 16, 2015 at 3:52 PM, Tom Rondeau wrote:
>> On Wed, Sep 16, 2015 at 3
Thanks guys.
The output (somewhat abbreviated) is at http://pastebin.com/BqvwB58y
I am running Windows 8.1
Mike
On Wed, Sep 16, 2015 at 3:52 PM, Tom Rondeau wrote:
> On Wed, Sep 16, 2015 at 3:48 PM, Chris Kuethe
> wrote:
>>
>> capture the output from "./pybombs install -v -v -v gnuradio", and
On Wed, Sep 16, 2015 at 3:58 PM, Dennis Glatting wrote:
>
>
> It would also be nice to have a VOLK kernel that multiples a vector by a
> constant:
>
>void
>volk_32f_s32f_multiply( float* vecbuffer,
>const float scalar,
>unsigned int n
On Wed, 2015-09-16 at 12:49 -0400, mle...@ripnet.com wrote:
> Sure, but if the flow-graph basically is decimated 10:1 by the time
> it reaches the log10, improvements in log10 are going to have very
> little impact. It is, I would assert, boldly and perhaps brashly,
> that log10 operations almost
On Wed, Sep 16, 2015 at 3:48 PM, Chris Kuethe
wrote:
> capture the output from "./pybombs install -v -v -v gnuradio", and
> stick it on pastebin so we can have a look at it.
>
> It looks like pybombs is trying to recompile make and gcc and goodness
> knows what else... I'm curious about why it de
capture the output from "./pybombs install -v -v -v gnuradio", and
stick it on pastebin so we can have a look at it.
It looks like pybombs is trying to recompile make and gcc and goodness
knows what else... I'm curious about why it decided to do that.
On Wed, Sep 16, 2015 at 9:25 AM, Mike Gilmer
Your problem is "/usr/lib/x86_64-linux-gnu/libgnuradio-osmosdr.so:
undefined reference to `uhd::device::find(uhd::device_addr_t const&)'"
The linker can't find UHD. I see no "-luhd" in your linker flags, so
the error makes sense... I wonder why the configuration process didn't
find UHD, though. Do
(accidentally sent this to the wrong mailing list the firs time,
sorry)
> I admit this seems a bit heath robinson (or rube goldberg for you
> Americans), but thanks for the feedback and the writeup!
For anyone following this, please feel free to interject if there is a
smoother way to go thro
On Wed, Sep 2, 2015 at 8:52 PM, Jeff Long wrote:
> Hi Jeon,
>
> Here's a long-winded answer. You're correct, but I wanted to make sure the
> terminology was clear.
>
> Tag offsets in add_item_tag() are absolute, relative to the beginning of
> the sample stream, not relative to input or output buf
Sure, but if the flow-graph basically is decimated 10:1 by the time it
reaches the log10, improvements in log10 are going to have very little
impact. It is, I would assert, boldly and perhaps brashly,
that log10 operations almost never need to be done at "line rate",
since they are an artifice
The Nyquist theorem says in general your sample rate equals your two sided
bandwidth, i.e. Fs=10 MHz -> BW = 10 MHz. In general, your samples are complex.
As soon as you restrict samples to be real in time, you cut the useable
bandwidth in half. This is because real time samples imply conjugate
On 15.09.2015 20:35, Marcus D. Leech wrote:
> Ordinarily, one does a log10 to convert into engineering units at the
> back of, for example, a power-measurement chain.
>
> There's usually no reason to do that in the middle of a flow-graph,
> where things can stay in linear units.
That's true, but
Gerome,
if you set a sampling rate of 4 MHz, you will have 4 MHz of usable
bandwidth (complex baseband). This is how it's supposed to be.
M
On 16.09.2015 09:27, Gerome Jan L wrote:
> Hi guys,
>
> Actually, that's what I thought too. If I set the UHD: USRP Source
> samp_rate to 400 Msps, I shoul
On 16.09.2015 06:17, Jason Matusiak wrote:
>> PyBOMBS + RFNoC is a bit iffy, you need to manually edit the uhd.lwr
>> file to have it build the rfnoc-devel branch. Then, it *should* work,
>> but I'll admit I haven't tested it myself.
>
> Martin, to wrap a bow on this, here is the steps I've taken
Following the instriuctions on
http://gnuradio.org/redmine/projects/pybombs/wiki/QuickStart
I tried to run the installation: ./pybombs install gnuradio
It appears to start up OK, but after "Loading recipes" it displays
"Installing packages" and gets through several dozen of them until it
display
Note that I kept everything in terms of "real" samples when talking
about sampling rates. Also, I should have made clearer the difference
between rates on receive and transmit. Transmit was not addressed.
For transmit, you know the clock, so you can use a sample rate of >2B
for real or >1B for
Dear all,
I have gnuradio installed on my 14.04 system using PyBOMBS.
Now I would like to try out gqrx but I get the error message pasted below.
I have found many other with this problem on the Internet and they solve it
by removing -mt from the gqrx.lwr file.
However, in my gqrx.lwr file the -mt
You might also want to check out Jeff Long's presentation on sampling rate
at the GRCon15 Intro Day:
http://www.trondeau.com/grcon15-presentations#monday_Long_Sampling
Video: https://www.youtube.com/watch?v=AmY7fCGTKs4
Tom
On Wed, Sep 16, 2015 at 5:57 AM, Marcus Müller
wrote:
> Hi Gerome,
>
> PyBOMBS + RFNoC is a bit iffy, you need to manually edit the uhd.lwr
> file to have it build the rfnoc-devel branch. Then, it *should* work,
> but I'll admit I haven't tested it myself.
Martin, to wrap a bow on this, here is the steps I've taken (and redone
to make sure things seem OK) and it se
This is the use case I had in mind when we were working on the log2 kernel.
Binary logs are the fastest log to compute (for floating point numbers)
because of Marcus' explanation. I even wrote the NEON log2 proto-kernel so
that the computation can be in-lined in other kernels in the future to
avoid
By the way, with a bit of approximation and a little acceptance for
errors in corner cases, log2 of a IEEE-754 float is simply it's exponent
+ log2(mantissa), with the later potentially being implementable using
lookup tables.
Also, VOLK does have a log2, kernel, volk_32f_log2_32f
Best regards,
Ma
Hi Gerome,
I hope the subject I've added suits the topic of your mail well; if not,
feel free to change it when replying)! It's kind of important to have a
proper subject line, as it'll allow people to directly address your mails.
> What does the sampling rate in this usrp block does?
It sets the
39 matches
Mail list logo
