On Mon, Feb 14, 2022 at 8:50 AM Christoph Hellwig <hch@xxxxxxxxxxxxx> wrote:
>
> I like the series a lot.
>
> Superficial comments:
>
> for nds32 is there any good reason why __get_user / __set_user check
> the address limit directly? Maybe we should unify this and make it work
> like the other architectures.
I've done that now, and am glad I did, because I uncovered that
put_user() was actually missing the check that got added to __get_user()
by accident.
> With "uaccess: add generic __{get,put}_kernel_nofault" we should be able
> to remove HAVE_GET_KERNEL_NOFAULT entirely and just check if the helpers
> are already defined in linux/uaccess.h.
Good idea, changed now as well.
> The new generic __access_ok, and the 3 fixed up version early on
> have a whole lot of superflous braces.
I'd prefer to leave those in, the logic is complex enough that I'd
rather make sure this is completely obvious to readers.
I got a few built bot reports about missing __user annotations that were
uncovered by the added type checking in access_ok, fixed those
as well.
I'm doing another test build after the last changes, will send it out in a bit
if there are no further regressions.
Arnd
