cli: rationalize return codes for running applets. * Allow applets and tools to return non-zero codes using `return` statement in the runner function. * Return 1 on `GlasgowAppletError` being raised. * Return 130 (SIGINT return code) on applet interruption.
by whitequark 1 year 5 months
90999d3f
cli: fix typo.
by whitequark 1 year 5 months
1c055a48
meetings: add minutes from 2023-07-29
by attie 1 year 5 months
6b4dbfdf
firmware,software: set high byte of bcdDevice to API level. Before this commit, bcdDevice can have two possible values: * 0x00RR for devices without firmware * 0x01RR for devices with firmware (where RR is the encoded revision, e.g. 0x32 for revC2) Finding out the API level requires opening a device and sending the REQ_USB_API_LEVEL control request. The API level is also 1 at the moment. After this commit, the high byte of bcdDevice is either zero (if the firmware is not loaded), or equal to the current API level (if the firmware is loaded). Conveniently, with the API level before this commit the rules are already satisfied. REQ_USB_API_LEVEL is eliminated and finding out the API level can now be done without opening the device. The API level is increased to 2. Compatibility considerations: * The encoded hardware revision can be 0x00 in the FX2 API block in case of really old devices. The firmware will fix it up on load. * There is a range of firmware between the commit that set high byte of bcdDevice to 0x01 (3b9b39bc29) and the commit that introduced API levels (30dc11b2fd) where the firmware may not actually comply with API level 1. Because of this, this commit bumps the API level. * A device with no FX2 configuraiton block will have bcdDevice set to 0xA000. (Likely, FX2 die revision A.) However, this is not relevant since the VID/PID are different. * In the future, REQ_USB_API_LEVEL can be removed. However, it is left in for now, for compatibility with old software checkouts.
by whitequark 1 year 5 months
4afe459f
firmware: fix git merge resolution error with 3a783fad and 5783d828. This has caused the FX2 LED to be stuck on even in LED test mode.
by whitequark 1 year 5 months
6f2a8b42
README: Update the doco to match PR#353 ("native"->"system")
by Rod Whitby 1 year 5 months
e1da219b
firmware: breathe FX2 LED when USB address is not assigned. Fixes #257.
by whitequark 1 year 5 months
3a783fad
firmware,software: make iManufacturer configurable in factory flow. Fixes #347.
by whitequark 1 year 5 months
5fcedc48
firmware: modify iProduct to say "Glasgow Interface Explorer". Previously it was "Glasgow Debug Tool".
by whitequark 1 year 5 months
610d8c97
applet.selftest: add `leds` mode. Fixes #235.
by whitequark 1 year 5 months
55bb30f5
Report a bug