Changelog for hackrf-devel-2017.02.1.git1486841062.0335f1a-2.1.x86_64.rpm :
Mon Feb 13 13:00:00 2017 mardnhAATTgmx.de
- Update to version 2017.02.1.git1486841062.0335f1a:

* Update hackrf_transfer.c

* Update hackrf_transfer.c

* LNA915 updates

* Fixed missing reset of do_exit, which prevented calls to callback when thread is created once again.

* updated toolchain URL

* fixed label in stackup documentation

* Add Travis CI configuration

* Add support for transmitting/receiving from stdin/stdout

* Fix sign comparison warnings

* Update HackRF photo

* Update HackRF One photo size

* Added flags + code to open device by index or serial number

* transferred trademark from Michael Ossmann to Great Scott Gadgets

* Re-generate BOM using mossmann\'s kicad-bom-scripts

* Used delay instead of equivalent code

* Remove unused USB configuration for CPLD update

* added gsg-kicad-lib submodule

* HackRF One: updated for compatibility with KiCad 4.0.2

* Add Windows Compatible ID descriptors

* hackrf_transfer - cast with uint32_t

* Reduce configuration count to 1

* Remove unude set_configuration() code for CPLD updates - now handled under the default configuration

* Use strndup instead of strdup, to handle situations where the reported length and the actual length are different, or not null terminated

* Make the \'do_exit\' flag static

* First receive streaming implementation

* Add OS X to Travis build

* Marzipan: started hardware design by copying HackRF One

* Neapolitan: started hardware design by copying HackRF One

* Neapolitan: fixed file names

* Neapolitan: deleted MCU and other things not needed

* Neapolitan: deleted end plate drawings

* Neapolitan: barrel jack and USB power sources that override USB power on main board

* Neapolitan: header pin assignments

* Neapolitan: added I/O expander

* Neapolitan: barrel jack module

* Neapolitan: experimental 2.5 ppm clock in schematic

* Tidy clang build warnings - Remove transceiver enums from hackrf.h - convert uint64_ts to int64_ts for frequency maths

* Neapolitan: experimental 500 ppb clock in schematic

* Neapolitan: experimental 10 ppb clock in schematic

* Neapolitan: schematic annotation and cleanup

* Neapolitan: module assignment

* Neapolitan: completed initial layout

* Neapolitan: added RF chokes to some supply pins as a noise reduction experiment

* Neapolitan: added decoupling caps for U15 and U18

* Neapolitan: added decoupling caps near pin headers

* Neapolitan: replaced antenna port power MOSFET with current limited power switch

* Build SPI flash and DFU versions of firmware every time

* Add CCCamp 2015 rad1o badge board id to libhackrf

* Update readme to reflect DFU changes

* Start of sync flag

* Added multiple device sync

* Experiments with gpio pins, for hardware sync

* Kind of working hardware sync

* Fix multibyte reads over i2c

* Add initial operacake support

* Fix silly polarity issues

* Kind of working hw sync

* Add Operacake USB API (firmware)

* Add host functions to set operacake ports

* Add support for multiple operacake boards connected to one HackRF

* Tidy unused operacake code and use addresses to refer to boards

* Simplify cmake config for adding tools

* Don\'t bother setting path 1 when tuning RFFC5072

* Speed up RFFC5072 comms by 50%!

* Output RFFC5072 LOCK status

* WIP scanning

* Use vendor request to start scan mode

* Add scan mode function separate to main loop

* Add init_scan call to libhackrf

* Only transfer one block per re-tune

* Change granularity of scan mode to MHz

* Add initial hackrf_scan skeleton

* Fix init_scan request length

* Switch all output from stdout to stderr to allow for piping samples

* Revert \"Only transfer one block per re-tune\"

* Cleanup scan_mode usb transfer logic

* Add magic & scan_freq to the start of each transfer

* Wait for scan_params to actually arrive before starting scan mode

* Rename scan to sweep

* Remove unused helper functions from hackrf_sweep

* Fix missed scan -> sweep changes

* Add some fftw magic to proceedings

* Calculate log power

* Don\'t increment power

* Scale input samples

* Fix input copy loop

* Rearrange fftw output

* Fix up datatypes

* Smaller FFT size

* Write power values to stdout

* Fix messed up buffer skip

* FFT size tweak again

* Skip first block after retune

* DC offset hacks

* Add -f argument for sweep freq range

* Add extra fs/4 hops to overlap band-edges and DC spikes

* Revert \"DC offset hacks\"

* Apply window

* Increase default gains

* Take samples from the end of the buffer

* Reset enter_scan_mode to avoid going straight back into scan_mode

* Frequency list PoC

* Fix up frequency list size

* Extend frequency array to allow 6GHz sweeping

* Add dwell time argument to hackrf sweep - untested because I\'m not able to fire up my HackRF right now

* Change dwell_time to num_samples

* Fix argument parsing for num_samples

* Stop dropping buffers in firmware and allow the host to make the decision.

* Sync ready

* Add vendor request for enabling/disabling gpio-based sync

* Add HackRF tool to change operacake ports

* Remove duplicate hackrf_operacake definition

* removed pre-libhackrf python host code

* Remove duplicate hackrf_sweep definition

* Check for FFTW3 before building as it is required for hackrf_sweep

* Add fftw3 to travis build

* added fftw for osx to travis

* Added fftw dependency to README for host build on Linux. It should be added for Windows too, but I am not qualified to do so properly.

* Remove dummy buffer

* Replace python one liner with script to build dfu files

* updated GCC toolchain URL in firmware README

* Reinstate commands to clean up after DFU process

* Avoided osx build problem by using literal string instead of format string. Fixed minor bug in warning text.

* Build firmware as part of Travis-ci build

* hackrf_transfer: made -n option compatible with -R (repeat TX mode). -n limits the total number of samples transferred via USB. This fixes issue #192.

* Support files larger than 2147483647 bytes on 32-bit platforms. Fixes issue #152.

* fixed operacake indentation warning in firmware

* Support compiling with Visual Studio

* Initial appveyor config

* Fix 7zip command line args

* Use http rather than ftp to download dependencies

* Correct libusb include dir argument to cmake

* Remove unnecessary escaping

* Build Win64 version

* For anti-aliasing, the baseband filter bandwidth is automatically set to the widest available setting that is no more than 75% of the sample rate. This happens every time the sample rate is set. If you want to override the baseband filter selection, you must do so after setting the sample rate.

* use new libhackrf automatic baseband filter selection in hackrf_transfer

* use preferred hackrf_sample_rate_set() instead of older function in hackrf_transfer

* fixed bug in dfu.py that made it only work for hackrf_usb

* have Travis-ci build all firmware, not just hackrf_usb

* fixed unused variable warning in blinky

* fixed indentation warning in sgpio-rx

* Update readme.md to reflect current build instructions
* udev rule
* Visual Studio instructions
* General clean up

* Remove out of date readme files. Up to date information can be found in host/README.md

* Jawbreaker firmware builds

* ifdef out HackRF One rf path pins for jawbreaker build

* Add software controlled reset

* Update submodule to include watchdog reset changes

* Correct argument parsing logic

* Safely call hackrf_close() after resetting the HackRF

* Fixed cmake bug that prevented hackrf_sweep from being built.

* Added -h/-? option for help to all command line tools.

* Fixed hackrf_transfer mode control logic.

* Filled in some missing long options for command line tools that use long options.

* Fixed write bug in hackrf_si5351c.

* Fixed bug that prevented hackrf_max2837 and hackrf_rffc5071 from displaying help when HackRF is not found.

* Clarified hackrf_si5351c control logic

* Removed \"subsequent\" from help for register number options which can now be placed before or after the read/write options. We no longer support chaining operations over multiple registers in a single command.

* Add USB API versioning

* Remove duplicate serial number from hackrf_info

* Turn off VBUS discharge to reduce current draw

* Add issue template to add structure to issue reporting

* Exclude vscode project files

* Replace error messages with more readable strings

* Shhh, silence in the library

* iReplace malloc that we never free with a statically assigned buffer

* Remove experimental firmware code - never used in firmware

* Fixed firmware dwell time bug in sweep mode. Two buffers of sample data are now thrown away at the start of each step in order to allow time for tuning. Fixes #317.

* Changed \"User cancel\" message to \"Exiting\" because it can happen in situations other than a user action. Fixes #319 investigate \"User cancel error\"

* fix(readme): Use correct DFU filename

* Standardize device selection options in hackrf-tools Fixes issue #318

* Add selection by device serial number to rffc5071 and max2837 tools

* removed unused TX options from hackrf_sweep

* Report libusb error messages where possible for USB errors

* Correct usage for -d

* Added csv text output to hackrf_sweep. Added -B option for binary output. Text output is now the default. Removed binary dump of raw samples between FFTs.

* print actual values for all fields in hackrf_sweep csv output

* fixed off-by-one error in hackrf_sweep

* fixed off-by-one error in hackrf_sweep again

* hackrf_transfer initialisation was using unusual syntax

* Remove repeated check for fd==NULL

* Remove unused test for result in opt parsing

* Remove unused error checking code

* Made hackrf_sweep output more like rtl_power\'s.

* consolidated a little code in hackrf_sweep

* too many Hz

* too few Hz

* Select only the best sections of the FFT output. Skip the lowest 1/8 of the band, the middle 1/4 of the band, and the highest 1/8 of the band. This provides full coverage of the selected frequency range because the hopping pattern was designed with this in mind.

* Add printable libhackrf and hackrf tools version strings

* Check function return values when reading registers

* Add dirty flag to host versions

* default frequency range and error checking of frequency range in hackrf_sweep

* Moved sweep mode frequency computation into firmware. Changed from long list of tuning frequencies to short list of ranges.

* support multiple frequency ranges in hackrf_sweep

* fixed firmware bug with sweep ranges that start at high frequencies

* fixed bug in bandwidth filter option error checking in hackrf_transfer

* fixed bug in bandwidth filter option error checking in hackrf_transfer

* Initial commit of combined debug tool

* Support max2837, rffc5071, and si5351

* Remove old separate debug tools

* allow selection of FFT bin width in hackrf_sweep

* fixed FFT bin reordering bug in hackrf_sweep

* don\'t send first buffer to host in sweep mode until it is full

* fft bin selection in hackrf_sweep binary output, similar to text output

* added output description to hackrf_sweep help

* code cleanup

* hackrf_sweep: suppress processing and output until the first expected frequency is seen

* hackrf_sweep: don\'t skip more input than necessary when waiting for valid data

* fixed bug that caused tuning glitches in interleaved sweep mode

* added one shot mode to hackrf_sweep

* made part naming consistent in hackrf_debug

* fixed long option usage examples

* change si5351c usage example to be the most common register that folks are likely to want to read

* added multisynth config to hackrf_debug example usage

* Don\'t build hackrf_operacake utility. Operacake developers can reverse this change locally if needed. The utility will have a new name or will be replaced by functions in other tools before we reinstate it.

* set release string in libhackrf and hackrf-tools even if they are built separately

* hackrf_transfer: switched some output messages from stdout to stderr

* fixed bug in frequency reported to binary output of hackrf_sweep

* Disable EMC clock

* Modify hw sync mode to require an argument - allows it to be ignored for HackRFs with older firmware

* added fields to hackrf_sweep binary output

* Power down or disable all clocks that we aren\'t using

* clarified some things in firmware/README

* hackrf_sweep: fftwf_free

* Reenable some clocks so that HackRF boots

* hackrf_sweep: removed bin width from binary output because a more precise result can be computed from the number of bins in a record

* fixed bug that prevented use of num_samples argument to hackrf_init_sweep()

* updated top level readme

* Reverted 31a37bd841b99154fce7217d2b3d78a24f8aacd9 Output RFFC5072 LOCK status This enabled a debug signal used during sweep mode development on HackRF One, but it broke Jawbreaker.

Sat Feb 20 13:00:00 2016 mardnhAATTgmx.de
- Update to version 2015.07.2.git1453796185.3f4d1a4:
+ Si5351C: Extract low-level driver code.
+ Si5351C: Un-singleton the high- and low-level drivers. Proper.
+ MAX2837: Extract hardware/bus dependent stuff into separate layer.
+ MAX2837: De-singleton the driver.
+ RFFC5071: Separate hardware-specific code.
+ RFFC507x: De-singleton the driver code.
+ W25Q80BV: Extract hardware-specific code into separate layer.
+ W25Q80BV: De-singleton the driver.
+ W25Q80BV: Make flash command details private.
+ W25Q80BV: Separate structures into algorithm and hardware access layers.
+ W25Q80BV: Un-const data argument to program function.
+ W25Q80BV: Move SPI functions to top of file, since they\'re static and used later.
+ W25Q80BV: Consolidate API to perform entire transfers in one call.
+ MAX2837: Remove test code.
+ RFFC5071: Remove debug/test code.
+ RFFC5071: Refactor a bunch.
+ RFFC5071: Delay switching data direction until after target unselected.
+ RFFC5071: Further refactoring to create common SPI interface.
+ RFFC5071: Rename rffc5071_drv->rffc5071_spi.
+ RFFC5071: Further work abstracting SPI details out of driver.
+ MAX2837: Refactoring toward abstracted SPI.
+ MAX2837: Finish SPI abstraction.
+ MAX5864: Abstract SPI, extract target code
+ MAX2837: Header clean-up.
+ W25Q80BV: Finish abstracting SPI code.
+ SSP1: Merge MAX2837 and MAX5864 SPI code to use single SPI interface.
+ SSP1: Fix argument name to match SPI naming.
+ SSP0: Extract SPI interface from W25Q80BV code.
+ SPI: Move SCU/pin config to target layers.
+ SSP: Merge SSP0 and SSP1 code into single unit.
+ MAX2837: Add virtual function for setting device mode.
+ SPI: Remove MAX2837/5864, W25Q80BV driver dependence on target code.
+ SPI: Pull remaining hardware bits out of MAX2837/5864/W25Q80BV device drivers.
+ RFFC5071: Remove extraneous #include.
+ Si5351C: Change write data arg to be const.
+ SI5351C: Rewrite si5351c_write_single() to use si5351c_write().
+ Si5351C: Widen write() data_count argument type to size_t.
+ Si5351C: Extract I2C code into single transfer function.
+ I2C: Finish extracting from Si5351C code.
+ Si5351C: Bring I2C wrapper into main driver.
+ SPI: Interface: Rename init()->start(), add stop().
+ SPI: Rename spi_t to spi_bus_t to match I2C naming.
+ I2C: Interface: Rename init()->start(), add stop().
+ Add \"pin\" interface to abstract GPIO pins.
+ Use new GPIO API to abstract GPIO in various drivers.
+ allow user to specify frequencies in MHz and scientific notation
+ Update build instructions to include pkg-config
+ Update build instructions to include pkg-config
+ Adapt code to signed samples.
+ Change frequency to center of 2.4GHz ISM.
+ Fix calculation of mag^2.
+ Remove extra SGPIO API calls.
+ Configure SGPIO slice mode early.
+ Reduce mag^2 threshold a bit.
+ Push rf_path_set_direction() calls into test functions.
+ rad1o USB idProduct cleanups
+ Add W25Q80BV function to fast-read from device.
+ Use W25Q80BV read function to retrieve SPI flash contents.
+ Only perform W25Q80BV setup once, at startup.
+ Default SPI flash read length to size of SPI flash Fixes #214
+ Comment style fixup to prevent warning in c90 complier mode
+ fix(hackrf-core): Disable unused clock outputs
+ Support both commonly available versions of dfu-util (0.7/0.8)
+ Merge \"abstract buses\" changes with \"fix SPI flash read\"
+ Add error handling to hackrf_trasfer argument parsing
+ Merge \"abstract buses\" and \"fix sgpio example to work with signed samples\"
+ Add dummy byte for w25q80bv fast read command timing
+ Apply C89 patch from guruofquality to support older compilers

Fri Jul 24 14:00:00 2015 mardnhAATTgmx.de
- Update to version 2015.07.2.git1437753263.dd6c8ae:
+ cmake bug fix from roox
+ Move udev rule to libhackrf/

Fri Jul 24 14:00:00 2015 mardnhAATTgmx.de
- Update to version 2015.07.1.git1437710931.c4450ad:
+ Serial number firmware and host-side changes. Very hacky at this point. Among the TODOs:
* Refactor obtaining LPC serial number and chip ID into separate API and header/source files. Remove from main().
* Create a usb_set_serial_number_descriptor() or similar function to be called before USB stack is started.
* Ensure USB serial number descriptor is valid even if code forgets to initialize the serial number before the USB stack is started. May be as simple as providing default initializer for usb_descriptor_string_serial_number[].
* Create a #define/constant for the usb_descriptor_string_serial_number length.
* Identify what\'s causing intermittent crashes in hackrf_transfer when no serial number is specified. I\'m probably misusing getopt.
* Permit serial number without leading zeros so you don\'t have to type as much.
* Add support for serial number argument in other hackrf_
* tools.
* Provide libhackrf support for enumerating multiple HackRFs, so that hackrf_info can list all devices. May require an additional libhackrf function, outside of hackrf_open().
+ Refactor obtaining LPC serial number and chip ID into separate API and header/source files. Remove from main(). Create a usb_set_descriptor_by_serial_number(), called before USB stack is started.
+ Create a #define/constant for the usb_descriptor_string_serial_number length. Clean up a compilation warning, libusb headers do not like the \'const\'.
+ libhackrf: refactor serial number searching to hackrf_open_usb()
+ Permit serial number without leading zeros so you don\'t have to type as much. In fact, permit any suffix of the serial number.
+ Adjust debug printf\'s a bit
+ hackrf_spiflash: added -d for device serial number
+ hackrf_cpldjtag: added -d for device serial number
+ libhackrf: hackrf_device_list() API for device enumeration and opening of any device (with or without serial numbers). hackrf_info uses hackrf_device_list to list devices, with or without serial number API support.
+ hackrf_info: Adjust output to print USB device identification string and device index number, before trying to open.
+ Bug fix, open the correct device in hackrf_device_list_open
+ Do not break hackrf_open() API, provide a hackrf_open_by_serial() instead for the new functionality.
+ libhackrf: The usb firmware reports board type as USB product id. Provide hackrf_usb_board_id_name() to decode that.
+ device enumeration: Return usb_board_ids as an enum
+ hackrf_init: only call libusb_init() once
+ Update README.md
+ Update hackrf.c
+ Wiki images of narrowest baseband filter at 2MHz and 8MHz sampling rate.
+ quick reminder of gain control mappings
+ Fixed pkgdata install path on FreeBSD.
+ Added information on FreeBSD build.
+ Fix discovery of libusb 1.0 for FreeBSD
+ Attempt to detect correct group for udev rule
+ Fix compiling on GCC 4.9 (\'for\' loop initial declarations not allowed with -std=gnu90)
+ explain which gain setting was wrong
+ vga gain must be a multiple of 2, not 8
+ squish invalid gain values to acceptable levels
+ these don\'t need to be errors any more
+ Derp. That\'s what I get for hacking after bedtime...
+ Fix GitHub issue #163 - Detach the kernel driver automatically
+ Remove duplicate call to set configuration + tidy calls to libusb_close()
+ Free config descriptor once we have the number of interfaces
+ Fix issue #113 - CPLD update fails on Windows Using the patch from AATTsupersat
+ Fix inconsistent indentation
+ Add file repeat TX mode [-R].
+ ChunkSize should be size entire file in bytes minus 8 bytes
+ turn off RF amplifier at end of transmission, fixes bug #178
+ Add hackrf_spiflash verbose option, make default not-verbose.
+ Fix KiCad files being detected as Eagle files
+ credit for lines of code belongs in commit log
+ Use rad1o\'s PID
+ Sort readme install instructions
+ bumped .so and .pc version numbers to 0.4
- removed patch hackrf-pr-159-better-udev-handling.diff (included upstream)

Sun Mar 15 13:00:00 2015 mardnhAATTgmx.de
- initial package, version 2014.08.1.git1422378789.50a6622