HOST_SYNC is only connected to connector P28, and is therefore not driven (left to float) unless connected to some synchronization signal. Pull it up to keep it steady.
In doing so, I had to switch all unused pins to pull-up, and all input-only and tri-state pins to float. All input/tri-state pins except for HOST_SYNC are tied to the microcontroller and can be pulled up there.
This line is meant to wait for the WEL bit to be set, to signify that
the WRITE_ENABLE command has finished.
The previous code didn't do any masking, so would hang if any other
status bits were set.
Linux issues a `GET_STATUS` request to validate that a USB device has
correctly resumed after an idle suspend; but HackRF has thus far not
implemented `GET_STATUS`. As a result, Linux assumes our USB devices
are failing to resume, and winds up resetting them. Oops. ^-^
Remove "portapack" arg to many static functions, no idea why I thought that was necessary.
Add PortaPack presence detect function using JTAG.
Create a "hackrf_ui_t" type that has function pointers to a specific UI implementation, returned by a UI detection function.
* Try artefacts deployed to gh_pages
* Try to use gh_pages from the travis_artefacts branch
* Try deploying to a different repo
* Try to organise files deployed to github pages
* Test pushing a local dir to master
* Try pushing to original repo
* Be verbose so I can debug it
* Setting env variables
* Oops, environment variables aren't what I thought
* Try to push to nightly repo
* Remove unused cp command
* Copy firmware to archive directory
* Fix pathing to artefacts
* Use TRAVIS_BUID_DIR instead of assuming path
* Use mkdir -p to ensure directories exist
* Put / back in to CPLD path
* Move repo to GSG
* Switch to master branch
* Add nightly deployment
* Fix escaping in sed command
* Allow firmware version styring to be overridden
* Fix some sed commands....
* Switch to master branch for builds
This includes:
* Cmake clean up - thanks @Qyriad
* Windows binaries saved after each appveyor build
* A bump to the Visual Studio version that we use to build it
* An appveyor cygwin script for building firmware, it doesn't work but it seems like someone might pick it up and make it work, or blow it away if we switch to Travis firmware artefacts
