This is the sixteenth post in a series tracking the development of Tock, a safe multi-tasking operating system for microcontrollers.
This week has been pretty quiet as we all put our heads down preparing submissions for SenSys and SOSP. Things will start picking back up in a couple weeks, but for now your usual, although abrdiged, installment of activity from the last week.
Some of the Tock developers will be at IPSN in Pittsburgh next week. Say hello if you see them!
More chips and platforms
There have been a fair bit of questions and offers of assistance porting Tock to some readily available hardware platforms, particularly the STM32-based discovery boards. The Tock team is all very supportive of such efforts. On the side, we’ve begun playing around with @japaric’s excellent tutorial (formerly at https://japaric.github.io/discovery) on bring up of the STMF3DISCOVERY.
If you are interested in these efforts, please join the tock-dev mailing list and/or chime in on IRC (#tock on freenode)
@alevy and @ppannuto fixed build failures in Travis CI related to the userspace build system and rustfmt version inconsistencies. The Travis CI problems, in particular, are totally unreplicable locally and seem most consistent with a file system issue of some sort on Travis, but a workaround works for now. The rustfmt version issues will likely resolve themselves in the long run as the rustfmt style guide is stabalized.
@alevy removed the now defunkt Firstorm platform (RIP) from the repository and replaced it with Hail as the default platform. Hail is not only active, but also available for purchase. Make sure to get yours!
@ppannuto fixed an bug in the SAM4L I2C driver that has not yet manifested, but is bad just the same. He found it while trying to debug an issue with multi-master I2C support on the SAM4L that is increasingly seeming like a hardware bug.