On Sat, Mar 29, 2025 at 06:51:21PM -0600, Jeff Law wrote:
> > 2025-03-28 Jakub Jelinek <[email protected]>
> >
> > * combine.cc (try_combine): Sets which satisfy set_noop_p can go
> > to i2 unless i3 is a jump and the other set is not.
> Shouldn't this wait for gcc-16? Or can you make a reasonable case that the
> nop moves constitute a code quality regression? OK for gcc-16. OK for
> gcc-15 is those nop moves are a code quality regression.
The intent was not to regress with the other patch posted (which is a P2
wrong-code).
Though on the particular testcase in the end this incremental patch makes no
difference, while with this incremental patch try_combine succeeds on it and
without it fails, seems other attempted combinations or what in the end
result into the same assembly.
Jakub
