------- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-01 02:36
-------
The corresponding asm for the loop:
with -fno-ivopts:
L11:
ld r11,8(r10)
srd r2,r4,r7
addi r0,r7,-6
rldicl r2,r2,0,58
rldicl r7,r0,0,32
ld r9,0(r11)
sldi r6,r2,3
addi r11,r10,24
stw r2,16(r11)
addi r8,r9,8
std r9,24(r10)
mr r10,r11
add r0,r8,r6
std r0,8(r11)
bdnz L8
without:
L11:
add r11,r8,r30
mulli r0,r3,-6 <--- this is really not a good idea inside the loop
add r10,r5,r8
addi r3,r3,1
ld r2,8(r11)
addi r7,r11,24
add r0,r0,r12
ld r9,0(r2)
srd r0,r4,r0
rldicl r0,r0,0,58
addi r6,r9,8
sldi r11,r0,3
stdx r9,r8,r5
stw r0,16(r10)
addi r8,r8,24
add r2,r6,r11
std r2,8(r10
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18241
