the `bladerf_channel` enum strings because
`bladerf_channel` is declared as an int, rather than
as an enum.
To resolve this, add a new static method
`bladeRF.str2ch` to handle the conversion
and use it to wrap appropriate `obj.channel`
and `obj.module` arguments in function calls.
Further, replace enum strings
`BLADERF_CORR_LMS_DCOFF...` --> `BLADERF_CORR_DCOFF...`
and
`BLADERF_CORR_FPGA_PHASE...` --> `BLADERF_CORR_PHASE...`
because MATLAB doesn't pick up the #define
which does so for the C code.
v1.0.1 (2018-08-24)
--------------------------------
This version is intended for use with libbladeRF v2.0.1.
Changes:
* host: libbladeRF: Rename clock select option
The bladeRF-micro has an option to use either its onboard VCTCXO or an
external 38.4 MHz clock as the clock source for components on the
board. This change addresses some ambiguity in the naming of the onboard
option by renaming it from 'vctcxo' to 'onboard'.
The valid options are now:
1. onboard
2. external
Massive changes to bladeRF_rx_gui to allow it to work with the new gain methods
and also allow it to properly inform the user of problems setting gains when
the gain mode (AGC) is set to automatic.
Some platforms have power and temperature monitoring, provide a way for them to
be exposed in matlab. To read the temperature of a device run the following:
> b = bladeRF()
> b.misc.temperature
misc.dc_power can be used to read the current power consumption
misc.dc_source be used to determine if the power is coming from USB or AC
bladeRF_XCVR has to catch up to the new way of doing things in libbladeRF. Low
level gain settings like directly RXLNA, RXVGA1, RXVGA2, TXVGA1, and TXVGA2 have
all been removed to make way for the unified gain setting based on
bladerf_set_gain(). Setting the gain value to 60 for RX indicates maximum RX
gain, while a value of 60 for TX indicates a output power of 0dBm.
Adding the new A4 and A9 boards. Currently some decisions are being made by
identifying the generation of the device. In the future this should be replaced
with a feature and capability mask.
Most of the changes that make libbladeRF return -8 still allow scripts to run
without serious issues. This commit should be changed back after a few releases
once developers have had a chance to update their scripts.
* libbladeRF/matlab: change win64 bool type (#543)
This should fix an error in recent Thunk builds. I am not sure what the
side effects of this will be (esp. with older Matlab versions) but this
seems to work now.
* host: libbladeRF_bindings: re-sort libbladeRF_proto.m to match recent MATLAB output
* host: matlab: update for libbladeRF 1.9.0
Add a few missing functions:
bladerf_set_gain_mode
bladerf_get_gain_mode
bladerf_set_smb_mode
bladerf_get_smb_mode
bladerf_trigger_init
bladerf_trigger_arm
bladerf_trigger_fire
bladerf_trigger_state
bladerf_set_vctcxo_tamer_mode
bladerf_get_vctcxo_tamer_mode
bladerf_xb300_set_trx
bladerf_xb300_get_trx
bladerf_xb300_set_amplifier_enable
bladerf_xb300_get_amplifier_enable
bladerf_xb300_get_output_power
bladerf_get_fw_log
bladerf_init_stream
bladerf_stream
bladerf_submit_stream_buffer
bladerf_submit_stream_buffer_nb
bladerf_deinit_stream
bladerf_read_trigger
bladerf_write_trigger
Add a few missing structs and enums, related to above
Also remove the bool_ret/bool_arg vars, as they aren't really needed any
more.
* host: matlab: fixup uint64s, cleanup
Removed the 'pathsep' from the architecture case, since it's seemingly
set by matlab itself
Not pre-allocating ival, to avoid having to count that high
* host: matlab: comment out stream-related protos
Matlab doesn't support FcnPtrs (bladerf_stream_cb) or voidPtrPtrPtr
(void***), so bladerf_init_stream is not usable. If bladerf_init_stream
isn't usable, then anything involving struct bladerf_stream isn't
usable.
Our proto file must provide the absolute path to the thunk library,
whether it is in the current directory or in MATLABPATH (as is the
case with our Windows installer).