https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69943
ktkachov at gcc dot gnu.org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |NEW
Last reconfirmed| |2016-02-24
CC| |ktkachov at gcc dot gnu.org
Component|rtl-optimization |tree-optimization
Ever confirmed|0 |1
--- Comment #1 from ktkachov at gcc dot gnu.org ---
Confirmed the integer addition case on aarch64:
sumi:
add w0, w0, w1
add w2, w0, w2
add w3, w2, w3
add w4, w3, w4
add w5, w4, w5
add w0, w5, w6
add w0, w0, w7
ret
reassoc1 performs the appropriate transformation for the float case.
Perhaps we can teach it to also do it for integer types?
