https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113326
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Keywords| |missed-optimization
Component|tree-optimization |target
Severity|normal |enhancement
Last reconfirmed| |2024-01-11
Ever confirmed|0 |1
Status|UNCONFIRMED |NEW
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Note on aarch64 with SVE, you should be able to generate those constants
without a load, using the index instruction. Though aarch64 does not generate
them currently. So this is definitely target specific and all.
Plus this is simplification happens after the vectorizer which produces:
```
<bb 3> [local count: 252544065]:
# i_13 = PHI <i_10(5), 0(2)>
# ivtmp_3 = PHI <ivtmp_2(5), 16(2)>
# vect_vec_iv_.4_11 = PHI <_16(5), { 0, 1, 2, 3 }(2)>
# vectp_array.6_19 = PHI <vectp_array.6_20(5), &array(2)>
# ivtmp_22 = PHI <ivtmp_23(5), 0(2)>
_16 = vect_vec_iv_.4_11 + { 4, 4, 4, 4 };
vect__1.5_18 = vect_cst__17 >> vect_vec_iv_.4_11;
_1 = value_8(D) >> i_13;
MEM <vector(4) int> [(int *)vectp_array.6_19] = vect__1.5_18;
i_10 = i_13 + 1;
ivtmp_2 = ivtmp_3 - 1;
vectp_array.6_20 = vectp_array.6_19 + 16;
ivtmp_23 = ivtmp_22 + 1;
if (ivtmp_23 < 4)
goto <bb 5>; [75.00%]
else
goto <bb 4>; [25.00%]
```