This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch committed SH] Fix PR target/41813
Hans-Peter Nilsson <hp@bitrange.com> wrote:
> Just curious, why the naked const_int, why not wrap it in "use"
> or as another operand inside the unspec vector (the latter being
> more canon)? The naked const_int seems like it could confuse
> something, not that there are many passes after dbr.
Maybe. I used this naked const_int because currently that insn
works simply as a unique pattern and there are few passes after
dbr. It should be something you mentioned to avoid the future
problem, though. I'll replace it with more normal one. Thanks
for your suggestion.
Regards,
kaz