Attention is currently required from: Thomas Heijligen, David Hendricks, Edward O'Callaghan.
Patch set 19:
We should also align this with the mmio version of this functions.
MMIO access is usually done with some offset too, but when it's with
specific endianness than we usually use pci_mmio_* functions. Maybe
those are worth to adapt? If we start to make changes that need an
update for a lot of calls, we probably should discuss that a little
more. One thing that comes to mind is the backwards parameters of
Is this more like what you think?
Yes that is what I had in mind :)
Patch Set #19, Line 111:
const base, const size_t offset
The latter 2 `const` are superfluous in a forward declaration.
It's not meaningful to the callers, they shouldn't care if the
callee can change its variables. Please drop them here (same
applies to the commit message, I guess).
Patch Set #19, Line 26:
base + offset
Technically this is wrong because arithmetic is undefined for
`void *`. So we should add another cast (to a pointer of some-
thing with size 1), e.g.
return le_to_cpu8 (*(uint8_t *)((char *)base + offset));
Casting to `uintptr_t` should also work.
To view, visit change 31016. To unsubscribe, or for help writing mail filters, visit settings.