Karl Palsson

I second what Felix says, about rebasing this on current code if at all possible. However, for some of the simpler things like fixing typos, if you can submit them...

libopencm3 should also compile with clang, or anything really, I don't personally want to go very far adding gcc built ins... Further, I'm also of the case that having _code_...

Further, even though I don't really have any motivation to merge this myself, if it _was_ going to be merged, it should be supported at _least_ for all stm32 families,...

I don't see any real reason for this to be a new file with duplicated code. Have a look at how the stm32 family handles usart vs uart. They're both...

I generally like this, but have a couple of comments... a) the "non_cl" part I think is awkward, non_cl are the most common (blanket, sweeping statement I know) so I...

lib/sam/common/pmc_common_3s.c should be in lib/sam/3s/pmc.c as it's specific, not common at all. (and to make that work, the current common/pmc.c I guess should be pmc_common_all.c)

I'm afraid I have no real insight into the usbd implementation itself, but this generally looks good, and I'd be happy to get this in as the first usb we...

I haven't looked at this yet, but as long as there are no _conflicts_ it's ok to have the common files contain a superset. You only need to break them...

The commit " Isochronous IN transfer incomplete callback …" looks good by itself, but this is only for fx07?

probably won't look at this much until I've finished reworking some of the f0/f3 adc basics but valid bug, needs to be addressed. Probably via a bunch of asynch methods