1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
|
; RUN: opt -passes='default<Oz>' -S < %s | FileCheck %s
; Forcing vectorization should allow for more aggressive loop-rotation with
; -Oz, because LV requires rotated loops. Make sure the loop in @foo is
; vectorized with -Oz.
target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128"
target triple = "arm64-apple-ios5.0.0"
define void @foo(ptr noalias nocapture %ptrA, ptr noalias nocapture readonly %ptrB, i64 %size) {
; CHECK-LABEL: @foo(
; CHECK: fmul <4 x float>
;
entry:
br label %for.cond
for.cond: ; preds = %for.body, %entry
%indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ]
%exitcond = icmp eq i64 %indvars.iv, %size
br i1 %exitcond, label %for.cond.cleanup, label %for.body
for.body: ; preds = %for.cond
%arrayidx = getelementptr inbounds float, ptr %ptrB, i64 %indvars.iv
%0 = load float, ptr %arrayidx, align 4
%arrayidx2 = getelementptr inbounds float, ptr %ptrA, i64 %indvars.iv
%1 = load float, ptr %arrayidx2, align 4
%mul3 = fmul float %0, %1
store float %mul3, ptr %arrayidx2, align 4
%indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
br label %for.cond, !llvm.loop !0
for.cond.cleanup: ; preds = %for.cond
ret void
}
!0 = distinct !{!0, !1}
!1 = !{!"llvm.loop.vectorize.enable", i1 true}
|
