Oleg Endo <[email protected]> wrote:
> the attached patch addresses PR 50694 and also does some cleanups in
> sh.md.
>
> I haven't run the testsuite with this one since it's basically just text
> replacements, except for the changed cmpgeusi_t insn. However, CSiBE
> didn't show any code size changes with the changed insn. As far as I
> could observe it, the check for the zero special case is already being
> done before insn matching. Or am I missing something there?
>
> If desired I could run the testsuite anyways of course.
The patch makes 4 logically separable changes into one. It'd
be better to split it to 4 independent patches.
The changes except cmpgeusi_t look essentially OK to me. But
please test them as usual for both endian.
The hunk for cmpgeusi_t doesn't look obvious and if that hunk
didn't change anything in CSiBE, what is it for?
I think that your observation is the case, i.e. almost all zero
special cases are already handled. The optimization intended
with the splitter will have a small chance but not zero, though.
Regards,
kaz
