* host: libbladeRF: clang-format fixes
* host: bladeRF-cli: clang-format fixes
* host: libbladeRF: spacing/wording/clang-format
Improve wording of various parts, add missing params, clean up various
wrap anomalies, and make it run clean through clang-format.
* host: libbladeRF: add functions to get/set ADF4002 refclk frequency
* host: bladeRF-cli: add print/set adf_refclk
* host: python: add adf4002_refclk functions
This permits getting and setting the ADF4002 reference clock
frequency.
>>> import bladerf
>>> b = bladerf.BladeRF()
ad9361_init : AD936x Rev 2 successfully initialized
>>> b.adf4002_refclk
10000000
>>> b.adf4002_refclk = 9000000
>>> b.adf4002_refclk
9000000
>>> b.adf4002_refclk = 900000000
[ERROR @
"/home/rtucker/dev/privatebladeRF/host/libraries/libbladeRF/src/board/bladerf2/bladerf2.c":4360]
bladerf_adf4002_set_refclk: bladerf_adf4002_calculate_ratio failed:
Provided parameter was out of the allowable range
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File
"/home/rtucker/dev/privatebladeRF/host/libraries/libbladeRF_bindings/python/bladerf.py",
line 1581, in set_adf4002_refclk
_check_error(ret)
File
"/home/rtucker/dev/privatebladeRF/host/libraries/libbladeRF_bindings/python/bladerf.py",
line 998, in _check_error
raise globals()[errorval.name](err)
bladerf.RangeError: Provided parameter was out of the allowable range
>>> b.adf4002_refclk
9000000
>>> b.adf4002_refclk = 10000000
>>> b.adf4002_refclk
10000000
* host: libbladeRF: make ADF convenience functions static
bladerf_adf4002_set_refclk is much nicer (and safer) than calculating
the R/N ratio and then doing a bladerf_adf4002_configure.
* host: python: update cffi header block
* host: libbladeRF-bindings/python: general updates
This commit adds the following changes:
Overall:
* Adds an _import_header.py tool, which parses libbladeRF.h (or other
headers) into a convenient string ready for ffi.cdef
* Passes pycodestyle and pyflakes checks cleanly
bladerf.py:
* Various new class wrappers for enums, etc
* Subclassed exceptions (e.g. bladerf.UnsupportedError will be raised if
libbladeRF returns BLADERF_ERR_UNSUPPORTED)
* Added functions for bootloader firmware recovery
* More comments all around
* Use properties instead of getters/setters:
>>> b = bladerf.BladeRF()
>>> # OLD:
>>> b.get_devinfo()
>>> b.set_clock_select(ClockSelect.External)
>>> # NEW:
>>> b.devinfo
>>> b.clock_select = ClockSelect.External
* Added more functions, particularly for bladeRF 2
* Added a Channel class, for channel-specific properties:
>>> b = bladerf.BladeRF()
>>> ch = d.Channel(bladerf.CHANNEL_RX(0))
>>> ch.frequency = 2.45e9
bladerf-tool.py:
* General changes to use properties instead of get_/set_ functions
* Also updated to use subclassed exceptions from bladerf.py
* Dump a lot more info with the 'info' command
* Look for bootloader devices with the 'probe' command
* Added 'recover_fw' command, for firmware recovery
* Moved from single-channel paradigm to MIMO paradigm on 'rx', 'tx'
* host: python: implement more methods, various fixes
Implement loopback, rx_mux, ad9361_temperature, and adf4002_* functions
Fix set_clock_select type handling
Use 4-space indent in cffi block, for placating text editors
Add -v option to bladerf-tool.py info, to clean things up a bit
Some are missing, others are too noisy in practice. Uniformly return BLADERF_ERR_UNSUPPORTED.
* host: bladerf2: test for board support in bias_tee
* host: bladeRF2: be quieter on unsupported gain mode query
* host: bladerf2: check for compat on ina219 reads
* host: libbladeRF/bladerf2: fixup remaining board compat checks
Allow user to control which Si53304 clock input to select: onboard
VCTCXO or external UFL clock. The external clock must be present before
changing the active clock, otherwise the host connection to the bladeRF
will drop.
Intended for use by bladeRF-micro, which has a TPS2115A auto power
mux. The function will read the power status bit from the Nios GPIO and
set the human-readable enum corresponding to the active power source.
Because we call bladerf*_initialize from within bladerf*_load_fpga, and
we then do things that do their own locking from within
bladerf*_initialize, problems ensue when loading FPGAs from bladeRF-cli.
This removes the lock from the greater load_fpga call, and moves it to
just around the dev->backend->load_fpga call.
The EVM coming out of the AD9361 is about 1% worse than it should be,
when doing a TX directly after initialization. However, by changing the
frequency and changing it back, the EVM returns to the expected level.
I suspect libad9361 is missing a calibration step somewhere in the
initialization, but we can work around this by doing two frequency
changes during initialization.
In the event a library user needs to test for the existence of a gain
stage, the bladerf_get_gain_stage_range function may now be called with
a null range, and the return value can be checked to determine whether
it exists.
* host: libbladeRF: add bladerf_get_loopback_modes
Returns a pointer to an array of supported loopback modes.
* host: libbladeRF: add bladerf_is_loopback_mode_supported()
Tests if a given loopback mode is supported on a device
generate.bash now allows specification of paths to awk, date, grep, and
so forth; this will permit it to be used in very weird environments.
generate.bash now uses bash built-ins whenever possible.
generate.bash checks the existence of necessary non-built-in commands
before doing anything, and prints a useful error if they're wrong.
Only try doing generate.bash if pandoc and bash are installed.
GCC 7.x adds a warning for implicit fallthroughs in switch cases. We
intentionally do this sometimes, so adding an EXPLICIT_FALLTHROUGH macro
to denote this.
The __attribute__((fallthrough)) on a null statement is only supported
by GCC 7.x as well, hence the macro.
Fixes clang warning:
poke.c:137:24: error: variable 'status' is used uninitialized whenever
'if' condition is false [-Werror ,-Wsometimes-uninitialized]
Realistically, there is no way that 'status' could be used on line 141
without one of f, f2, or f3 being true (thus setting the value), due to
the test on 132. There is no harm in initializing it anyway, though.
Update bladerf_rf_switch_config struct to use "rfic" instead of "ad9361"
for genericness and consistency.
Also clean up missing/incorrect documentation in various places.
Fix bladeRF-cli to use proper struct member names for
bladerf_rf_switch_config.
Fixes missing fields in initializers and misuse of abs() (using abs() on
an unsigned int has no effect, and using it on an int64_t risks
truncation)
Also fix up a missing newline in an error printf