https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82636
--- Comment #5 from Michael Meissner ---
Created attachment 42425
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=42425&action=edit
Post reload dump file
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82636
--- Comment #4 from Michael Meissner ---
Created attachment 42424
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=42424&action=edit
Reload dump file
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82636
--- Comment #3 from Michael Meissner ---
Created attachment 42423
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=42423&action=edit
LRA dump file
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82636
--- Comment #2 from Michael Meissner ---
Created attachment 42422
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=42422&action=edit
Simpler test case that does not use asm
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82636
--- Comment #1 from Michael Meissner ---
This looks like a reload bug. I see the same thing with automatically
generated fmas:
--> cat foo06c.c
__ieee128
__fmaf128_power9 (__ieee128 x, __ieee128 y, __ieee128 z)
{
return (x * y) + z;
}
After
