On Wed, 5 Apr 2023, at 14:26, Joel Stanley wrote: > On Wed, 5 Apr 2023 at 04:34, Jeremy Kerr <jk@xxxxxxxxxxxxxxxxxxxx> wrote: >> >> Hi Joel, >> >> > Given we have acks on the bindings, I think it's safe to send the >> > device tree changes now so we can merge what you have in the upcoming >> > merge window. If there's changes we can modify or revert. >> >> OK, I'll get those into shape. >> >> There is one dependency on this though, and unfortunately it requires >> solving *two* of the known-hard problems in computer science: >> >> Do we start at i3c0 or i3c1? > > I'd vote for i3c0, like we do for ethernet, i2c, spi, etc. > > It is annoying that aspeed chooses to start counting at 1, but at > least we would be consistent with other parts of the kernel. > > Starting at 1 reminds me of doing matlab assignments at uni, so that's > another reason to avoid it. I've heard rumors that Aspeed will number their devices from zero in the future :) Andrew
