| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
Detecting drops by parsing for DD and SS was a flawed method. Tools
should find those programmatically. Plus, the string 'DDC' would
interfere with the regex. Also, we're now using
UHD_LOG_FASTPATH_DISABLE.
|
| |
|
|
|
|
|
| |
- Move filter_* functions out of uhd_test_case
- Reduced some line lengths
|
|
|
|
|
|
|
|
|
|
| |
Fixes #2484
When building uhd host as a static library on Windows, UHD_API is
defined as __declspec(dllimport). This results in MSVC returning error
C2491 during compilation. Added a preprocessor define UHD_STATIC_LIB
in cmake and made the config headers set UHD_API to empty string when
UHD_STATIC_LIB is defined.
|
|
|
|
|
| |
Now matches the FPGA error message (go download, then run
uhd_image_loader).
|
| |
|
|
|
|
|
| |
Change the RB_FIFOSIZE register to store the FIFO size in bytes
instead of log2 of bytes.
|
|
|
|
|
| |
- AMP_LO1_EN_CH1 controls U2, the amp for the external LO1 port, so it must
be set high if channel 2 is using an external LO (external or reimport)
|
|
|
|
|
|
|
|
| |
uhd_images_downloader was mismanaging the image inventory- image
packages entries were not being deleted when the images were
overwritten. Now, each target is saved as a single entry in the
inventory, which is overwritten when a new image package is
downloaded.
|
|
|
|
|
|
| |
This is a common message that will always occur during device init when
there is no RIO device available. Because it looks like an error, it
confuses people and was thus reduced to TRACE.
|
|
|
|
|
|
|
|
|
|
| |
The C/C++ standards don't define what time_t is, only that it is
arithmetic (and real for C11, and integral for C++). It should not be
used in portable software and is only used as the return value for some
libc calls.
A common definition for time_t is int64_t, so we'll switch to that
permanently in our own APIs. System APIs will of course stick with
time_t.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
For USRPs that support user settings (e.g., B2xx, N230), this will
return an object that will allow peeking and poking user-defined
settings registers.
Mock code example:
auto usrp = multi_usrp::make(...);
auto user_settings_iface = usrp->get_user_settings_iface();
user_settings_iface->poke32(0, 23);
|
| |
|
|
|
|
| |
No functional changes. Cleanup only. A little less Boost.
|
| |
|
|
|
|
|
| |
The testsuite now includes more environment variables for source dir,
build dir, and PYTHONPATH.
|
|
|
|
|
|
|
|
|
|
| |
You can now change the time/clock source default through device args:
auto usrp = uhd::usrp::multi_usrp::make(
"type=x300,clock_source=external,time_source=external");
This also enables the use of config files for the clock/time source
implicitly.
|
|
|
|
|
|
|
| |
x300_impl will now use a constrained_device_args_t-derived object to
parse device args.
No API or functional changes.
|
|
|
|
|
| |
- Add default parser helper
- Allow _enforce_discrete() for str_arg
|
|
|
|
|
|
|
| |
Also puts all defaults into the uhd::usrp::x300 namespace.
This commit does some renaming and refactoring, but no functional
changes.
|
|
|
|
| |
- All tests pass except for gpio_test
|
|
|
|
|
|
| |
- Change RX/TX min/max frequency according to AD9361 datasheet
- Fix set_atr_bits to change with rx/tx frequency and antenna independently
- Make AMP switching active high
|
|
|
|
|
|
| |
With Boost 1.64 to 1.65 (which, of course, Ubuntu LTS ships), the
`PyInit_Libraryname` are invisible when one sets the default visibility
to "hidden" (which is reasonable, and which we do).
|
| |
|
|
|
|
|
|
|
|
|
| |
- The flushing mechanism now looks similar to that in noc_shell
- Make use of new flush bit in FIFO control register
- Restrict using the clear bit only after flushing to ensure no
partial packets are introduced in the stream. (clear immediately
empties out FIFOs)
- Changes are backwards compatible with older FPGAs
|
|
|
|
| |
clock_source call
|
|
|
|
|
| |
- Unnecessary use of Boost
- Use of thread_priority.hpp
|
| |
|
| |
|
|
|
|
|
| |
Before, setting these properties in the prop tree would trigger an
exception.
|
|
|
|
| |
To support python api installer
|
|
|
|
|
|
|
|
|
| |
On non-Windows platforms, enable the Python API by default in CMake.
Enabling the Python API in Windows causes build failures- that is, we
currently can't detect if the requirements are met in the Windows
build system. Because of that, we'll leave it disabled by default for
now.
|
|
|
|
|
| |
Adding section in the manual on how to build filesystems and SD card
images.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
This device is the only one using it, and no one will ever use it going
forward.
|
|
|
|
| |
- This bug could cause LO to not lock properly after a set_frequency call
|
| |
|
|
|
|
|
|
| |
This is a fixup of a pretty old commit, but necessary. It adds
a UHD_USRP_MULTI_USRP_TX_LO_CONFIG_API macro for detection of TX LO
controls, which should have been in the original commit.
|
| |
|
| |
|
|
|
|
| |
This will broadcast on all interfaces concurrently, instead of serially.
|
|
|
|
|
|
|
|
|
|
| |
The Python API should now be installed to
`PREFIX/lib/pythonVER/dist-packages/uhd`
where the PREFIX is set by CMake and the Python version is determined
by the Python module `distutils`. This should match user expectations
much more than the previous behavior.
Behavior in Virtualenvs is unchanged.
|
|
|
|
|
|
| |
- Bump the UHD API version and return to development branch status
- Removing artful in list of Ubuntu releases. Support for artful ended
in July 2018.
|
|
|
|
|
|
|
| |
Ethernet buffering is now done so that most of the buffering is done in
the socket buffers and multiple frames are only used to support the
receive side offload of the socket I/O. Eliminates dropped packets at
high full duplex rates.
|
|
|
|
|
| |
send_buff_size is now constrained to input fifo size, and we increase
timeout on getting flow control to reduce CPU usage.
|
|
|
|
|
| |
This is to allow for num_recv_frames=1 and reduce conversions from ticks
to time_spec_t to improve critical path performance.
|