Hi Peter, > Hmmm, should not the ndelay before the loop also be split up in two like > this, with one ndelay(... / 2) on either side of the (possible) set_sda. > Not that it should matter, since SDA is presumably stuck low. But what if > it isn't? I agree it would be better. > I would also change the while (...) to > > for (i = 0; i < RECOVERY_CLK_CNT * 2; i++) Yeah, could be done as well. > but both these "issues" are perhaps not related to this patch... I also think these should be handled incrementally. > Reviewed-by: Peter Rosin <peda@xxxxxxxxxx> Thanks for the review! Regards, Wolfram
Attachment:
signature.asc
Description: PGP signature
