Commit Graph

372 Commits

Author SHA1 Message Date
Steve Markgraf f427883320 rtl_tcp: Initialize listensocket
Older versions of GCC will complain that it can be used
uninitialized - which is not the case, but it breaks our Jenkins
build as we build with -Werror.
2020-02-26 23:52:02 +01:00
A. Maitland Bottoms 8985b45e1f rtl_tcp: Add IPv6 support
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.
2020-02-26 22:55:23 +01:00
A. Maitland Bottoms b281473156 Use udev uaccess rules 2020-02-26 22:39:45 +01:00
A. Maitland Bottoms f2a9a81c4b rtl_fm/rtl_power: Improve scanning range parsing 2020-02-26 22:16:29 +01:00
A. Maitland Bottoms b2ee24eff5 CMake: support for libusb on GNU/Hurd
Debian builds on hurd-i386 with a variant of libusb.
2020-02-26 22:16:29 +01:00
A. Maitland Bottoms 222517b506 Improve librtlsdr.pc file
librtlsdr.pc should declare -lusb-1.0 in Libs.private section
to exclude usb library from dynamic linking.
References to libusb headers are not needed in Cflags, since these
headers are not used by external rtlsdr API, but this is optional.
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=784912
2020-02-26 22:16:29 +01:00
Benjamin Larsson 1dff8e0b4c lib: enable better UHF reception (>862MHz) for FC0013 2020-02-18 20:52:20 +01:00
Steve Markgraf be2e4f899a set CMake policy CMP0075 if it exists
Otherwise newer versions of CMake are throwing a warning.
2020-01-19 16:56:51 +01:00
leonsal b5af355b1d Fixed issues compiling on Windows with MSVC, CMake and NMake (#61)
When trying to build a simple program which uses librtlsdr
as a subproject on Windows, CMake reported several problems
which were solved by:
- Added complete name of libusb in FindLibUSB module.
- Replaced CMAKE_SOURCE_DIR to PROJECT_SOURCE_DIR in src/CMakeLists.txt.
- Replaced header file <afxres.h> in src/rtlsdr.rc.in (only present when windows MFC is
  installed) by <windows.h> which defines the same constants.
2019-12-03 20:47:17 +01:00
leonsal bd73737c49 allow building librtlsdr as CMake subproject
Replace CMAKE_SOURCE_DIR by PROJECT_SOURCE_DIR in main CMakeLists.txt
to fix CMake errors when building librtlsdr as a subproject.
2019-11-12 21:09:05 +01:00
Adrian Chadd 43b257854c rtl_biast: Add rtl_biast
This is an import of the rtl_biast command line tool from the
rtlsdrblog github repository.  It's easier to include it here than
try to package up the separate application because they both
wish to install dynamic libraries.
2019-11-12 20:46:02 +01:00
Adrian Chadd 5d0735f5df lib: Add GPIO version of the bias tee configuration API
rtl_biast allows for non-default GPIO pins to be used.
Add an API call which allows for that.
2019-11-12 20:32:35 +01:00
Brian Waters 9d05150a64 Fix building librtlsdr on OpenBSD
Gets rid of librt, which doesn't exist on OpenBSD. The version of
librtlsdr in the OpenBSD ports tree is extremely old (~2013), so this
should help some users.

Tested against tag 0.6.0, but it should apply just fine to HEAD.
2019-11-11 14:28:46 +01:00
Steve Markgraf 3c263b7451 rtl_eeprom: fix warnings
Account for \0 string terminator when calling strncpy().

Fixes the following GCC 9 warning:
warning: ‘__builtin_strncpy’ specified
bound 256 equals destination size
2019-11-01 02:18:54 +01:00
Steve Markgraf 81833a1cf6 lib: disable usbfs zero-copy support by default
Although we added a detection mechanism for the presence of the Kernel
bug earlier, reading from the incorrectly mapped memory might cause a
bus error on some ARM systems.

With the overall performance benefit being rather minimal for the
data rates of rtl-sdr, disable zero-copy by default.
2019-07-16 23:49:48 +02:00
Steve Markgraf be1d1206bf lib: fix memory leak in rtlsdr_open()
Thanks to Vincent Perrier for reporting the bug.
2019-07-16 23:31:31 +02:00
Oliver Smith 7f2ae12548 contrib/jenkins.sh: run "make maintainer-clean"
Related: OS#3047
Signed-off-by: Oliver Smith <osmith@sysmocom.de>
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2019-07-16 23:19:46 +02:00
Steve Markgraf f68bb2fa77 lib: Add workaround for Linux usbfs mmap() bug
The Linux Kernel has a bug on ARM/ARM64 systems where the USB CMA
memory is incorrectly mapped to userspace, breaking zerocopy.

When the Kernel allocates the memory, it clears it with memset().
If the mapping worked correctly, we should have zeroed out buffers,
if it doesn't, we get random Kernel memory. We now check for this,
and fall back to buffers in userspace if that's the case.

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-10-07 01:25:07 +02:00
Steve Markgraf c4452b6d5a change version to 0.6git
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-10-07 01:00:49 +02:00
Steve Markgraf 1f0eafe604 change version to 0.6.0
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-08-26 14:54:51 +02:00
David Woodhouse 9047eef701 Fix inline functions to use 'static inline'
With just 'inline', if the compiler decides not to inline them, it isn't
required to emit them at all. For some targets with -Os that is causing
build failures, see https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86360.

Perhaps we might consider using '__attribute__((always_inline))' for
GCC builds, but 'static inline' is a good start.

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-07-03 23:59:01 +02:00
Harald Welte 8a26abaf79 Add Debian packaging information
This allows (among others) to build rtl-sdr nightly packages
for a variety of distributions as part of the network:osmocom:nightly
builds
2018-06-03 15:19:53 +02:00
Harald Welte 9573e68d10 README: Update homepage/wiki URL from old trac to redmine
Signed-off-by: Harald Welte <laforge@gnumonks.org>
2018-06-03 14:31:39 +02:00
Harald Welte 8e9d4e6f2a Fix "make uninstall" for documentation
During "make uninstall", the documentatin was not deleted properly,
if DESTDIR was used.  The existing commands expanded to:

make[3]: Entering directory '/home/osmocom-build/jenkins/workspace/master-rtl-sdr/a1/default/a2/default/a3/default/label/osmocom-master-debian9/librtlsdr-0.5.4.5-5c30/_build/sub'
cd /tmp/am-dc-24621/ && rm -rf /home/osmocom-build/jenkins/workspace/master-rtl-sdr/a1/default/a2/default/a3/default/label/osmocom-master-debian9/librtlsdr-0.5.4.5-5c30/_inst/share/doc/librtlsdr/librtlsdr-0.5.4.5-5c30/html

This fixes the "make distcheck" step of rtl-sdr, and hence should also
fix the long-standing jenkins build verification for rtl-sdr
at https://jenkins.osmocom.org/jenkins/job/master-rtl-sdr/

Signed-off-by: Harald Welte <laforge@gnumonks.org>
2018-06-03 13:24:42 +02:00
Steve Markgraf 5c3073f417 tuner_r82xx: turn off loop-through, remove dead code
As pointed out by Carl Laufer on the mailing list,
turning the loop-through output off reduces the
current consumption by 10-20mA which in turn reduces
the heat a bit. So far there seem to be no devices
that have anything connected to the loop-through output.

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-05-25 00:16:38 +02:00
Pete Zaitcev ce9595ef53 rtl_adsb: fix hanging upon a signal in Fedora 27
This code stayed unchanged for many years, but for some reason
rtl_adsb started hanging upon exit:

*b66116a5164b69281eacc42ae950;
^CSignal caught, exiting!
 <------ hangs here forever

Examining it with gdb reveals that the demod thread waits
peacefully on the condition variable, which we're trying to
destroy. Either the signals killed all threads before, or
condition variables were possible to destroy while other
threads still waited on them.

The easiest fix appears to be just cancel the demod thread
and wait for it to exit before proceeding for the door.

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-05-23 23:20:26 +02:00
Steve Markgraf a854ae8b48 lib: use USB zero-copy transfers if possible
Decreases CPU load especially for embedded machines.
Requires Linux >= 4.6 and libusb >= 1.0.21. If this
is not the case or the allocation fails, we will
fall back to buffers allocated in userspace.

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-05-01 14:23:37 +02:00
Martin Hauke 326b0e3acc Fix build with MinGW
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-05-01 14:13:00 +02:00
Steve Markgraf d93819d29d change version to 0.6git
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-05-01 12:46:31 +02:00
Steve Markgraf 9ca8780df7 change version to 0.5.4
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-04-29 00:30:45 +02:00
Steve Markgraf c403ef0bc1 rtl_test: fix build on Mac OS
Thanks to mot for reporting.
closes #3130

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-04-26 23:51:15 +02:00
Harald Welte 566a1b942f Makefile.am: Fix "permission denied" during "make distcheck"
This fixes the following error during "make distcheck":
	echo 0.5.3.24-2a36 > ../../.version-t && mv ../../.version-t ../../.version
	/bin/sh: ../../.version-t: Permission denied
2018-04-19 00:09:15 +02:00
Harald Welte 2a36e5dd65 add contrib/jenkins.sh for build testing 2018-04-18 23:59:17 +02:00
Harald Welte f7f3e3f6d7 configure.ac: Add --enable-sanitize and --enable-werror
this is "osmocom standard" these days, and is used by our CI scripts
2018-04-18 23:59:17 +02:00
Harald Welte 5a9864c56d fix "-Wshift-negative-value" compiler warning on ggc-7.3.0 2018-04-18 23:59:17 +02:00
Harald Welte e92b11bfb7 cosmetic: Fix "implicit fall-through" gcc 7.3.0 warnings 2018-04-18 23:59:17 +02:00
Steve Markgraf 4520f001d8 lib: fix FC0012 reset GPIO
Since a typo in rtlsdr_set_gpio_output() was fixed,
FC0012 tuners were not detected anymore, as the reset pin
is actually 4, not 5.

Thanks to David Basden et al for reporting the bug.

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-02-20 22:24:46 +01:00
Kacper Michajłow a03198c5b7 rtl_test: Support PPM error measurement on Windows.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2018-02-20 22:24:40 +01:00
Hoernchen b04c2f9f03 fix for msvc14 2017-09-19 19:26:49 +02:00
hayati ayguen 18bf26989c Add option '-T' for activating bias-T in rtl_(adsb|fm|power|tcp)
* added debug output when activating
Signed-off-by: hayati ayguen <h_ayguen@web.de>
* fixed options argument
Signed-off-by: Fabian P. Schmidt <kerel-fs@gmx.de>
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2017-06-13 01:20:32 +02:00
Dimitri Stolnikov fa3a113b77 rtl_tcp: add command for bias tee control 2017-06-12 00:04:05 +02:00
Lucas Teske 2be1612e60 lib: Use GPIO P0 to toggle an (optional) bias-t
At least one distributor of rtl-sdr dongles (rtl-sdr.com) added
a bias-t to their dongles which could be toggled via GPIO P0 of the
RTL2832U chip.

source: http://www.rtl-sdr.com/rtl-sdr-blog-v-3-dongles-user-guide/

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2017-06-09 15:06:23 +02:00
Lucas Teske ba64a7459a lib: fix direction bit in GPIO code
source: http://lea.hamradio.si/~s57uuu/mischam/rtlsdr/ports.html

* Removed unnecessary comment of old code.
Signed-off-by: Fabian P. Schmidt <kerel-fs@gmx.de>
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2017-06-09 15:06:09 +02:00
Steve Markgraf e3e6ee23b7 lib: add new HanfTek dongle
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2016-11-27 12:19:44 +01:00
Hoernchen e3c03f738f lib: check for libusb init failure 2015-06-09 01:53:23 +02:00
Hoernchen 5c376fc79c tools: allow built-in functions for newer versions of MSVC 2015-06-09 01:37:46 +02:00
Steve Markgraf 8b4d755ba1 Merge pull request #20 from lcl7/correct-e4k_reg_write
Correct return code of e4k_reg_write().
2015-05-16 00:48:37 +02:00
Jiří Pinkava 92df068dc5 New functions rtlsdr_set_tuner_bandwidth()
Signed-off-by: Steve Markgraf <steve@steve-m.de>
2015-05-16 00:38:01 +02:00
Jiří Pinkava d892279085 Allow setting bandwidth for R820T
This improves SDR performence for nearby channel interference.
As a sideeffect also improves dynamic range becase ADC is not overloaded
by onwanted singlas.

Signed-off-by: Steve Markgraf <steve@steve-m.de>
2015-05-16 00:37:17 +02:00
lcl c5dc459fc5 Correct return code of e4k_reg_write().
In contradiction with the documentation, e4k_reg_write() was
returning the value 2 upon success (the number of bytes written),
not zero as documented.
Patched to return zero upon success.

As a consequence, this corrects also the return value of
e4k_reg_set_mask() and e4k_dc_offset_calibrate().
2015-05-08 11:10:30 +02:00