Indan Zupancic

I can have a look if you can provide a datasheet that describes the timer hardware. It's not an sel4test issue, but an problem with your driver, so it should...

> I've kind of localized the issue already, It looks like the addressing I used (as mentioned in the DTS) not correct for TIMER either or the TIMER is not...

> > Where did you got the register definitions from? > > not sure but I could not find any for now. [..] but structurally I followed the hifive-p550 style...

Me too. I found the documents, but not the right timer info. So at first guess, what's seems missing is configuration of the Timer Clock Control Register?

The timer you need to use is probably: https://developer.spacemit.com/documentation?token=T7TnwVZz1iPBk1kKwAPc6lyKnNb#11.2-timer-%26-watchdog and doesn't match the register layout of the timer in this PR.

Please squash the history into one commit. Ignore the line too long error for the link to the spec, that's okay, but fix the other errors.

You may have tabs instead of spaces and that's what the style checker chokes on.

> I think style fails because of force push as also mentioned in https://github.com/seL4/ci-actions/tree/master/style > > > This breaks for force-push, where the previous head of the branch might not...

> Nice. How hard would it be to generate a test case for that? I guess reliably triggering an IRQ when a thread is in the release queue is not...

> Suggestion: The trap_entry code should check whether the exception was sourced from umode or smode before executing any context saving code. (Even if we don't allow any exceptions to...