a0086add2e
1. Better sorting of scalars to be gathered. Trying to insert constants/arguments/instructions-out-of-loop at first and only then the instructions which are inside the loop. It improves hoisting of invariant insertelements instructions. 2. Better detection of shuffle candidates in gathering function. 3. The cost of insertelement for constants is 0. Part of D57059. Differential Revision: https://reviews.llvm.org/D103458
69 lines
2.9 KiB
LLVM
69 lines
2.9 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
|
|
; RUN: opt < %s -basic-aa -slp-vectorizer -dce -S -mtriple=i386-apple-macosx10.8.0 -mcpu=corei7-avx | FileCheck %s
|
|
|
|
target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128-n8:16:32-S128"
|
|
target triple = "i386-apple-macosx10.9.0"
|
|
|
|
;int foo(int *A, int n, int k) {
|
|
; for (int i=0; i < 10000; i+=4) {
|
|
; A[i] += n;
|
|
; A[i+1] += k;
|
|
; A[i+2] += n;
|
|
; A[i+3] += k;
|
|
; }
|
|
;}
|
|
|
|
define i32 @foo(i32* nocapture %A, i32 %n, i32 %k) {
|
|
; CHECK-LABEL: @foo(
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[TMP0:%.*]] = insertelement <4 x i32> poison, i32 [[N:%.*]], i32 0
|
|
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <4 x i32> [[TMP0]], i32 [[K:%.*]], i32 1
|
|
; CHECK-NEXT: [[SHUFFLE:%.*]] = shufflevector <4 x i32> [[TMP1]], <4 x i32> poison, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
|
|
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
|
|
; CHECK: for.body:
|
|
; CHECK-NEXT: [[I_024:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[ADD10:%.*]], [[FOR_BODY]] ]
|
|
; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[A:%.*]], i32 [[I_024]]
|
|
; CHECK-NEXT: [[TMP2:%.*]] = bitcast i32* [[ARRAYIDX]] to <4 x i32>*
|
|
; CHECK-NEXT: [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* [[TMP2]], align 4
|
|
; CHECK-NEXT: [[TMP4:%.*]] = add nsw <4 x i32> [[TMP3]], [[SHUFFLE]]
|
|
; CHECK-NEXT: [[TMP5:%.*]] = bitcast i32* [[ARRAYIDX]] to <4 x i32>*
|
|
; CHECK-NEXT: store <4 x i32> [[TMP4]], <4 x i32>* [[TMP5]], align 4
|
|
; CHECK-NEXT: [[ADD10]] = add nsw i32 [[I_024]], 4
|
|
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i32 [[ADD10]], 10000
|
|
; CHECK-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END:%.*]]
|
|
; CHECK: for.end:
|
|
; CHECK-NEXT: ret i32 undef
|
|
;
|
|
entry:
|
|
br label %for.body
|
|
|
|
for.body: ; preds = %entry, %for.body
|
|
%i.024 = phi i32 [ 0, %entry ], [ %add10, %for.body ]
|
|
%arrayidx = getelementptr inbounds i32, i32* %A, i32 %i.024
|
|
%0 = load i32, i32* %arrayidx, align 4
|
|
%add = add nsw i32 %0, %n
|
|
store i32 %add, i32* %arrayidx, align 4
|
|
%add121 = or i32 %i.024, 1
|
|
%arrayidx2 = getelementptr inbounds i32, i32* %A, i32 %add121
|
|
%1 = load i32, i32* %arrayidx2, align 4
|
|
%add3 = add nsw i32 %1, %k
|
|
store i32 %add3, i32* %arrayidx2, align 4
|
|
%add422 = or i32 %i.024, 2
|
|
%arrayidx5 = getelementptr inbounds i32, i32* %A, i32 %add422
|
|
%2 = load i32, i32* %arrayidx5, align 4
|
|
%add6 = add nsw i32 %2, %n
|
|
store i32 %add6, i32* %arrayidx5, align 4
|
|
%add723 = or i32 %i.024, 3
|
|
%arrayidx8 = getelementptr inbounds i32, i32* %A, i32 %add723
|
|
%3 = load i32, i32* %arrayidx8, align 4
|
|
%add9 = add nsw i32 %3, %k
|
|
store i32 %add9, i32* %arrayidx8, align 4
|
|
%add10 = add nsw i32 %i.024, 4
|
|
%cmp = icmp slt i32 %add10, 10000
|
|
br i1 %cmp, label %for.body, label %for.end
|
|
|
|
for.end: ; preds = %for.body
|
|
ret i32 undef
|
|
}
|
|
|