osmocom-sdr

osmocom-sdr@lists.osmocom.org

1 participants
904 discussions

by Harald Welte

[cross-post to many lists, please follow-up-to openbsc(a)lists.osmocom.org] Dear all, time is flying, and I would like to start early with discussions and planning about OsmoCon and OsmoDevCon in 2018. It helps to start early. Side note: We have some pending issues about the events from last year at http://osmocom.org/projects/osmo-dev-con/issues - I've incorporated them in the text below. == OsmoDevCon == For OsmoDevCon, I think it's easy: We keep it as-is. Same procedure as every year, which means: * same venue, same catering options * same concept of 'anyone contributing to Osmocom can apply for registration until all seats are taken' * same idea of inviting some few speaker[s] doing other FOSS mobile communications work to join us The parts that we need to change, IMHO: * don't reduce from 4 to 3 days like last year. Have full 4 days again * sort topics per day / half-day, i.e. have "GSM/UMTS Cellular Infrastructure" days for BTS/BSC/NITB/MSC/HLR/SGSN/GGSN & Co, but then have other days for other projects. This would enable people not interested in the [continued evolvement] of the cellular projects be able to skip those days, or to simply meet in an adjacent room for parallel hacking sessions/discussions * try to be a bit more structured with the schedule in general. The existing approach works for people who attend all the event all day long, but not so well for people with other plans / limited time Any further change requests or topics to discuss? Please note that Pablo Neira has offered to kindly host an OsmoDevCon in Seville (Spain). I've attended a number of netfilter workshops he organized there, and he's doing a great job! However, given the large number of attendees from Berlin (and Germany in general), I think this would make things more complicated, and more expensive for most attendees. If you disagree with that assessment: I'm open for having the discussion, I just thought it's more practical/economic to do it in Berlin. === 10 year Anniversary Party === Given that 2018 marks the 10 year anniversary of Dieter and me hacking with the Siemens BS-11 in 2008, I think the 2018 incarnation deserves some special celebration of some form. I have no concrete idea yet, but for sure we should so something, and it should be at/around OsmoDevCon. And for sure we should have a BS-11 around :) == OsmoCon == The public OsmoCon was welcomed and was a success. However, let's start this discussion with a review of last years event. === Registration === * Registrations came in way too late. Two weeks ahead of the event, we were considering to cancel it. And then within the last few days, we had to turn people down due to limited seating capacity * To make planning more reliable, we see on other option but to significantly raise the registration fee combined with an equally significant discount for early booking === Duration === * Many people requested multiple days rather than just one, in order to make more out of (long distance) travels. This is obvious, but as we had no idea how many people would attend at all (or if we have to cancel due to lack of attendance), planning multiple days in the first incarnation would have been high risk and a multitude of work * I would suggest to expand to two or even three days this week, possibly one days with tutorials and the other day with tech talks * Slightly less crammed schedule due to multiple days === Venue === We recognize this yearso venue was not the best option, due to * Bad ventilation in the basemenet * Difficult to find * No space next to the conference room where people can meet / hang out in parallel to talks (not everyone attends every talk) I still like the "understatement" of the venue. I'd prefer any hostel / non-profit / hackerspace / university over luxurious hotels any time. Going to an expensive venue means more or less automatically more expensive ticket fees, which again is more likely to exclude pure community members without a commercial activity related to Osmocom. So any future venue would ideally: * be able to hold slightly more people than this year * have a second room or large lobby in which people can meet for extended coffee breaks in parallel to some talks, as needed * be slightly easier to find (and we have to put up some signs outside and in the lobby) * have better WiFi and/or wired connectivity === Programme / Format === * less crammed over multiple days * some more "interactive" formats were requested, for users to provide feedback to developers * there was some discussion about topics / speakers in redmine last year, but not too much participation [until it was too late]. * I'd suggest a more formal CfP process with a submission deadline that allows us to publish a preliminary schedule long ahead of the event === Video Recordings === I think they were a big success, and it was a very big surprise that the CCC Video Operations Center was volunteering to help such a small and niche-interest event like OsmoCon. We should make sure that we can repeat this for 2018. == Dates / Frequency == Having OsmoCon and OsmoDevCon back to back becomes somewhat long, if OsmoCon is 2-3 days and OsmoDevCon is 4 days. Basically we're looking at a full week for those of you who would like to attend both events. But then, I think the number of people attending both events is actually not all that big. Without checking the details, I think not more than half of the OsmoDevCon attendees were attending OsmoCon. I would expect that tendency to remain or even increase. I still think it's good to keep them back-to-back. In terms of frequency, I would actually suggest we move to a 6-month cycle rather than a 12-month cycle. There's a lot of development going on at all time. I understand that not everyone is able to attend two events just on Osmocom, especially if it's a spare time / hobby type activity. That's ok, I think there's no problem with attending either of the two only, and catching up by video recordings and/or mail on the other. The qeustion is: Should that second event be developer-oriented or user-oriented? Or again both? Any comments here? Ok, that was a somewhat lengthy e-mail. Please make sure to provide any feedback you may have as early as possible, to increase the chances of your feedback being reflected in the planning. Happy hacking, Harald -- - Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)

7 years, 8 months

[PATCH] Fix a segfault in soapy_source_c

by Vasil Velichkov

When the get_gain_names(chan) returns an empty vector calls set_gain(gain, chan) instead of set_gain(gain, names.front(), chan); (gdb) bt #0 std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string (__str=..., this=0x7fffffffdb10) at /usr/include/c++/7/bits/basic_string.h:424 #1 soapy_source_c::set_if_gain (this=0x555556b85250, gain=32, chan=0) at ./lib/soapy/soapy_source_c.cc:257 #2 0x00007fffdf0b82a4 in source_impl::set_if_gain (this=0x555555bb6f30, gain=32, chan=<optimized out>) at ./lib/source_impl.cc:708 #3 0x00007fffdf3a860d in _wrap_source_sptr_set_if_gain (args=<optimized out>, kwargs=<optimized out>) at ./obj-x86_64-linux-gnu/swig/osmosdr_swigPYTHON_wrap.cxx:23069 #4 0x000055555564f3ca in call_function (oparg=<optimized out>, pp_stack=0x7fffffffdd10) at ../Python/ceval.c:4352 --- lib/soapy/soapy_source_c.cc | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/lib/soapy/soapy_source_c.cc b/lib/soapy/soapy_source_c.cc index a645361..e9e7377 100644 --- a/lib/soapy/soapy_source_c.cc +++ b/lib/soapy/soapy_source_c.cc @@ -253,16 +253,24 @@ double soapy_source_c::get_gain( const std::string & name, size_t chan ) double soapy_source_c::set_if_gain( double gain, size_t chan ) { - //docs specify RF gain is the first element - const std::string name = this->get_gain_names(chan).front(); - return this->set_gain(gain, name, chan); + std::vector<std::string> names = this->get_gain_names(chan); + if (names.length()) { + //docs specify RF gain is the first element + return this->set_gain(gain, names.front(), chan); + } else { + return this->set_gain(gain, chan); + } } double soapy_source_c::set_bb_gain( double gain, size_t chan ) { - //docs specify baseband gain is the last element - const std::string name = this->get_gain_names(chan).back(); - return this->set_gain(gain, name, chan); + std::vector<std::string> names =this->get_gain_names(chan); + if (names.length()) { + //docs specify baseband gain is the last element + return this->set_gain(gain, names.back(), chan); + } else { + return this->set_gain(gain, chan); + } } std::vector< std::string > soapy_source_c::get_antennas( size_t chan ) -- 2.13.6

7 years, 8 months

[PATCH] output calculated ppm to stderr

by Андрей Родионов

Good day, I found an issue, when i call rtl_test from external program with "-p” argument. PPM is written into stdout, which means it will be automatically flushed on interactive shell or when fflush called manually. Neither of this was used. All other output was written to stderr, hence simple patch: diff --git a/src/rtl_test.c b/src/rtl_test.c index 9a6cfda..c7e60c5 100644 --- a/src/rtl_test.c +++ b/src/rtl_test.c @@ -202,7 +202,7 @@ interval += (int64_t)(ppm_now.tv_nsec - ppm_recent.tv_nsec); nsamples_total += nsamples; interval_total += interval; - printf("real sample rate: %i current PPM: %i cumulative PPM: %i\n", + fprintf(stderr, "real sample rate: %i current PPM: %i cumulative PPM: %i\n", (int)((1000000000UL * nsamples) / interval), ppm_report(nsamples, interval), ppm_report(nsamples_total, interval_total)); Best Regards, Andrey

7 years, 9 months

No supported tuner found

by Gabor Mikes

Hi All, I recently purchased a couple of DVB-sticks on Amazon, titled: "RTL-SDR, FM+DAB, DVB-T USB Stick Set with RTL2832U & R820T. Great SDR for SDR#, HDSD" I got "No supported tuner found" when running the rtl_test command. I see others have had the same issue, so when I finally figured it out, I thought some of you might be interested. First of all, this stick has a FC0012 tuner despite the title. Secondly this stick is obliviously wired differently than other sticks with FC0012 because it needs different gpio settings. I added a few lines to the rtlsdr_open() function in librtlsdr.c, just before probing for FCxxx: /* initialise GPIOs */ rtlsdr_set_gpio_output(dev, 0); rtlsdr_set_gpio_output(dev, 3); rtlsdr_set_gpio_output(dev, 4); rtlsdr_set_gpio_output(dev, 5); rtlsdr_set_gpio_output(dev, 6); rtlsdr_set_gpio_bit(dev, 3, 1); rtlsdr_set_gpio_bit(dev, 4, 0); rtlsdr_set_gpio_bit(dev, 6, 0); /* reset tuner before probing */ rtlsdr_set_gpio_bit(dev, 5, 1); rtlsdr_set_gpio_bit(dev, 5, 0); With this modification the tuner comes a live and everything works. Kindest regards, Gabor Mikes

7 years, 9 months

[PATCH] Add ipv6 support to rtl_tcp

by A. Maitland Bottoms

I've prepared this patch in response to Debian bug #870804 https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=870804 It passes the text from the -a and -p options through getaddrinfo() and uses the first result that has a valid socket with a successful bind. While not a complete bind to all possible valid names, it does appear to address the use case of the bug submitter without completely changing the program flow. I submit this for inclusion in the osmocom rtl-sdr repo as it may be of use to others. Thanks for your consideration, -Maitland

7 years, 9 months

redmine e-mails broken from 08/26 through 08/31

by Harald Welte

Dear Osmocom community, from August 26th 06:54 UTC through August 31st 06:22 UTC our Osmocom.org redmine could not send any e-mails. This was due to a configuration file syntax error introduced by me, my apologies. If you rely on redmine e-mail notifications to the issues you have subscribed to, please double-check as related notifications during that interval were unfortunately lost. Best regards, Harald p.s.: The reason for the config change was to enable e-mail notifications from jenkins.osmocom.org (which it now has, if you want to configure e.g. post-build notification actions). -- - Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)

7 years, 9 months

vdl mode2 decoder for RTL SDR

by Thierry Leconte

Hi I just put on github my new project : vdl mode2 decoder for RTL dongle https://github.com/TLeconte/vdlm2dec

7 years, 10 months

SDRSharp/Airspy on Debian

by Oscar Llerena

Hi guys I am attempting to install Airspy (before known as SDRSharp) on my Debian 9.0 and I am following this link: https://rtlsdr.org/softwarelinux When it gets to the line of Airspay Installation it quotes "Don't forget to blacklist the kernel driver as mentioned here <https://github.com/airspy/host/wiki/Troubleshooting%5Dhere%5D.%20%20*%20Dow…> ." which links you to this wiki site https://github.com/airspy/airspyone_host/wiki/Troubleshooting#linux- debianubuntu I was able to install airspy-git, gr-osmosdr, and gqrx but I get lost on the part of blacklist the kernel >= 3.17 What can I do to blacklist the kernell driver? Thanks

7 years, 11 months

GPIO pin integration to osmosdr in gnuradio via argument

by samuel verdon

Hello, I’m trying to switch the state of the pin J2_3 of my xb100 through the argument line of the osmosdr block in gnuradio. To do this, I added this code to the file bladerf_common.cc 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 /* Support of the XB100 to control pin J2_3 through the argument line*/ if ( dict.count("xb100") ) { if (bladerf_expansion_attach(_dev.get(), BLADERF_XB_100)) { std::cout << "Could not attach XB-100" << std::endl; } else { _xb_100_attached = true; std::cout << "Expension XB-100 attached" << std::endl; if ( bladerf_expansion_gpio_dir_masked_write(_dev.get(),BLADERF_XB100_PIN_J2_3, 1)) { std::cout<< "Could not set TRX pin direction" << std::endl; } std::cout << "J2_3 output direction selected" << std::endl; if ( dict["xb100"] == "TX" ) { if(bladerf_expansion_gpio_masked_write (_dev.get(),BLADERF_XB100_PIN_J2_3, BLADERF_XB100_PIN_J2_3)){ std::cout<< "sorry, but I cannot set pin to 1" << std::endl;} } else if(dict["xb100"] == "RX" ) { if(bladerf_expansion_gpio_masked_write (_dev.get(),BLADERF_XB100_PIN_J2_3, 0)){ std::cout<< "sorry, but I cannot set pin to 1" << std::endl;} } else { std::cout<< "Wrong parameter specified, should be TX or RX" << std::endl;} } } In my osmosdr bloc, I added this argument “ xb100=tx” but I don’t have my 5V on my pin :( I don’t understand why the pin is not setted when I start my flowgraph… Could you help me please? Thank you Samuel Verdon

7 years, 11 months

Fwd: Re: Problem with latest version (master) of gr-osmosdr

by John Ackermann N8UR

After a successful build of the Gnuradio plus extras by modifying the build-gnuradio script to point to the osmocom git repository for rtlsdr, I get an error and backtrace/memory map similar to the attached osmosdr-error-output.txt when I exit from, e.g., osmo_fft (the program seems to run fine until exit). Fabian suggested that I post this to the list. I also installed op25 on this system, and it fails at runtime with a bunch of errors that I'm attaching as "op25-error-output.txt"; the end of the sring is: wx._core.PyAssertionError: C++ assertion "!m_frameStatusBar" failed at ../src/common/framecmn.cpp(381) in CreateStatusBar(): recreating status bar in wxFrame I wonder if this error is related, as op25 on a several-month-old gnuradio installation works fine. Thanks for any pointers on how to resolve this issue (or these issues). John -------- Forwarded Message -------- Subject: Re: Problem with latest version (master) of gr-osmosdr Date: Sat, 24 Jun 2017 14:35:45 -0400 From: John Ackermann N8UR <jra(a)febo.com> To: kerel <kerel-fs(a)gmx.de> Here you go, Fabian. Thanks much for looking into this! I have noticed this problem each time I've managed to get a successful build-gnuradio completion on this system (e.g., after I commented out the changes that made the build fail). I have another machine that was updated using build-gnuradio perhaps three months ago, and it does not exhibit this behavior. One difference is that the other machine is running Linux Mint 18.0, while the current one is on 18.1 (both "apt-get dist-upgraded" to current levels at the time of the build). John ---- On 06/24/2017 02:23 PM, kerel wrote: > Hi John, > this seems like an unrelated issue in gr-osmosdr, but I can't tackle it > with the partial? backtrace you provided and can't reproduce it myself. > Could you provide the full backtrace? > > Sincerely, > Fabian

7 years, 12 months

← Newer
1
...
31
32
33
34
35
36
37
...
91
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

osmocom-sdr