--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-15
01:58 ---
Now I see what is the difference between this and PR 18431, = vs .
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16803
--- Additional Comments From nathan at gcc dot gnu dot org 2004-11-12
09:24 ---
We cannot generate better code, without having a different meaning for the
sign_extend action that occurs in the loop.
As zdenek points out, we cannot use dbra, because we cannot tell if the loop
will
--- Additional Comments From nathan at gcc dot gnu dot org 2004-11-11
17:40 ---
FSF HEAD 2004-11-11 gives better code, with the invariants moved out of the
loop,
.main:
ld 9,[EMAIL PROTECTED](2)
lwz 11,0(9)
addi 11,11,20
extsw 11,11
cmpwi 7,11,0
--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-11
18:19 ---
Woops I I filed PR 18431 (which I think is the same problem well the testcases
are the same), I will note
I copied both -m32 and -m64 loops to show where the problem is and with arrays
we get much better
--- Additional Comments From steven at gcc dot gnu dot org 2004-07-30 17:41
---
I doubt LNO would help here, it looks like a backend issue.
But you can give it a try.
--
What|Removed |Added