blob: a0a13b2b5aac0373f6fcdcc6e02f3e235ed608d8 [file] [log] [blame]
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -basicaa -slp-vectorizer -slp-threshold=-100 -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(double *A, int k) {
; double A0;
; double A1;
; if (k) {
; A0 = 3;
; A1 = 5;
; } else {
; A0 = A[10];
; A1 = A[11];
; }
; A[0] = A0;
; A[1] = A1;
;}
define i32 @foo(double* nocapture %A, i32 %k) {
; CHECK-LABEL: @foo(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TOBOOL:%.*]] = icmp eq i32 [[K:%.*]], 0
; CHECK-NEXT: br i1 [[TOBOOL]], label [[IF_ELSE:%.*]], label [[IF_END:%.*]]
; CHECK: if.else:
; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds double, double* [[A:%.*]], i64 10
; CHECK-NEXT: [[TMP0:%.*]] = bitcast double* [[ARRAYIDX]] to <2 x double>*
; CHECK-NEXT: [[TMP1:%.*]] = load <2 x double>, <2 x double>* [[TMP0]], align 8
; CHECK-NEXT: br label [[IF_END]]
; CHECK: if.end:
; CHECK-NEXT: [[TMP2:%.*]] = phi <2 x double> [ [[TMP1]], [[IF_ELSE]] ], [ <double 3.000000e+00, double 5.000000e+00>, [[ENTRY:%.*]] ]
; CHECK-NEXT: [[TMP3:%.*]] = bitcast double* [[A]] to <2 x double>*
; CHECK-NEXT: store <2 x double> [[TMP2]], <2 x double>* [[TMP3]], align 8
; CHECK-NEXT: ret i32 undef
;
entry:
%tobool = icmp eq i32 %k, 0
br i1 %tobool, label %if.else, label %if.end
if.else: ; preds = %entry
%arrayidx = getelementptr inbounds double, double* %A, i64 10
%0 = load double, double* %arrayidx, align 8
%arrayidx1 = getelementptr inbounds double, double* %A, i64 11
%1 = load double, double* %arrayidx1, align 8
br label %if.end
if.end: ; preds = %entry, %if.else
%A0.0 = phi double [ %0, %if.else ], [ 3.000000e+00, %entry ]
%A1.0 = phi double [ %1, %if.else ], [ 5.000000e+00, %entry ]
store double %A0.0, double* %A, align 8
%arrayidx3 = getelementptr inbounds double, double* %A, i64 1
store double %A1.0, double* %arrayidx3, align 8
ret i32 undef
}
;int foo(double * restrict B, double * restrict A, int n, int m) {
; double R=A[1];
; double G=A[0];
; for (int i=0; i < 100; i++) {
; R += 10;
; G += 10;
; R *= 4;
; G *= 4;
; R += 4;
; G += 4;
; }
; B[0] = G;
; B[1] = R;
; return 0;
;}
define i32 @foo2(double* noalias nocapture %B, double* noalias nocapture %A, i32 %n, i32 %m) #0 {
; CHECK-LABEL: @foo2(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = bitcast double* [[A:%.*]] to <2 x double>*
; CHECK-NEXT: [[TMP1:%.*]] = load <2 x double>, <2 x double>* [[TMP0]], align 8
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
; CHECK: for.body:
; CHECK-NEXT: [[I_019:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[TMP2:%.*]] = phi <2 x double> [ [[TMP1]], [[ENTRY]] ], [ [[TMP5:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[TMP3:%.*]] = fadd <2 x double> <double 1.000000e+01, double 1.000000e+01>, [[TMP2]]
; CHECK-NEXT: [[TMP4:%.*]] = fmul <2 x double> <double 4.000000e+00, double 4.000000e+00>, [[TMP3]]
; CHECK-NEXT: [[TMP5]] = fadd <2 x double> <double 4.000000e+00, double 4.000000e+00>, [[TMP4]]
; CHECK-NEXT: [[INC]] = add nsw i32 [[I_019]], 1
; CHECK-NEXT: [[EXITCOND:%.*]] = icmp eq i32 [[INC]], 100
; CHECK-NEXT: br i1 [[EXITCOND]], label [[FOR_END:%.*]], label [[FOR_BODY]]
; CHECK: for.end:
; CHECK-NEXT: [[TMP6:%.*]] = bitcast double* [[B:%.*]] to <2 x double>*
; CHECK-NEXT: store <2 x double> [[TMP5]], <2 x double>* [[TMP6]], align 8
; CHECK-NEXT: ret i32 0
;
entry:
%arrayidx = getelementptr inbounds double, double* %A, i64 1
%0 = load double, double* %arrayidx, align 8
%1 = load double, double* %A, align 8
br label %for.body
for.body: ; preds = %for.body, %entry
%i.019 = phi i32 [ 0, %entry ], [ %inc, %for.body ]
%G.018 = phi double [ %1, %entry ], [ %add5, %for.body ]
%R.017 = phi double [ %0, %entry ], [ %add4, %for.body ]
%add = fadd double %R.017, 1.000000e+01
%add2 = fadd double %G.018, 1.000000e+01
%mul = fmul double %add, 4.000000e+00
%mul3 = fmul double %add2, 4.000000e+00
%add4 = fadd double %mul, 4.000000e+00
%add5 = fadd double %mul3, 4.000000e+00
%inc = add nsw i32 %i.019, 1
%exitcond = icmp eq i32 %inc, 100
br i1 %exitcond, label %for.end, label %for.body
for.end: ; preds = %for.body
store double %add5, double* %B, align 8
%arrayidx7 = getelementptr inbounds double, double* %B, i64 1
store double %add4, double* %arrayidx7, align 8
ret i32 0
}
; float foo3(float *A) {
;
; float R = A[0];
; float G = A[1];
; float B = A[2];
; float Y = A[3];
; float P = A[4];
; for (int i=0; i < 121; i+=3) {
; R+=A[i+0]*7;
; G+=A[i+1]*8;
; B+=A[i+2]*9;
; Y+=A[i+3]*10;
; P+=A[i+4]*11;
; }
;
; return R+G+B+Y+P;
; }
define float @foo3(float* nocapture readonly %A) #0 {
; CHECK-LABEL: @foo3(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = load float, float* [[A:%.*]], align 4
; CHECK-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds float, float* [[A]], i64 1
; CHECK-NEXT: [[TMP1:%.*]] = bitcast float* [[ARRAYIDX1]] to <4 x float>*
; CHECK-NEXT: [[TMP2:%.*]] = load <4 x float>, <4 x float>* [[TMP1]], align 4
; CHECK-NEXT: [[REORDER_SHUFFLE:%.*]] = shufflevector <4 x float> [[TMP2]], <4 x float> undef, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; CHECK-NEXT: [[TMP3:%.*]] = extractelement <4 x float> [[REORDER_SHUFFLE]], i32 3
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
; CHECK: for.body:
; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[R_052:%.*]] = phi float [ [[TMP0]], [[ENTRY]] ], [ [[ADD6:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[TMP4:%.*]] = phi float [ [[TMP3]], [[ENTRY]] ], [ [[TMP12:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[TMP5:%.*]] = phi float [ [[TMP0]], [[ENTRY]] ], [ [[TMP14:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[TMP6:%.*]] = phi <4 x float> [ [[REORDER_SHUFFLE]], [[ENTRY]] ], [ [[TMP19:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[MUL:%.*]] = fmul float [[TMP5]], 7.000000e+00
; CHECK-NEXT: [[ADD6]] = fadd float [[R_052]], [[MUL]]
; CHECK-NEXT: [[TMP7:%.*]] = add nsw i64 [[INDVARS_IV]], 2
; CHECK-NEXT: [[ARRAYIDX14:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[TMP7]]
; CHECK-NEXT: [[TMP8:%.*]] = load float, float* [[ARRAYIDX14]], align 4
; CHECK-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 3
; CHECK-NEXT: [[ARRAYIDX19:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[INDVARS_IV_NEXT]]
; CHECK-NEXT: [[TMP9:%.*]] = bitcast float* [[ARRAYIDX19]] to <2 x float>*
; CHECK-NEXT: [[TMP10:%.*]] = load <2 x float>, <2 x float>* [[TMP9]], align 4
; CHECK-NEXT: [[REORDER_SHUFFLE1:%.*]] = shufflevector <2 x float> [[TMP10]], <2 x float> undef, <2 x i32> <i32 1, i32 0>
; CHECK-NEXT: [[TMP11:%.*]] = insertelement <4 x float> <float 1.100000e+01, float 1.000000e+01, float 9.000000e+00, float undef>, float [[TMP4]], i32 3
; CHECK-NEXT: [[TMP12]] = extractelement <2 x float> [[REORDER_SHUFFLE1]], i32 0
; CHECK-NEXT: [[TMP13:%.*]] = insertelement <4 x float> undef, float [[TMP12]], i32 0
; CHECK-NEXT: [[TMP14]] = extractelement <2 x float> [[REORDER_SHUFFLE1]], i32 1
; CHECK-NEXT: [[TMP15:%.*]] = insertelement <4 x float> [[TMP13]], float [[TMP14]], i32 1
; CHECK-NEXT: [[TMP16:%.*]] = insertelement <4 x float> [[TMP15]], float [[TMP8]], i32 2
; CHECK-NEXT: [[TMP17:%.*]] = insertelement <4 x float> [[TMP16]], float 8.000000e+00, i32 3
; CHECK-NEXT: [[TMP18:%.*]] = fmul <4 x float> [[TMP11]], [[TMP17]]
; CHECK-NEXT: [[TMP19]] = fadd <4 x float> [[TMP6]], [[TMP18]]
; CHECK-NEXT: [[TMP20:%.*]] = trunc i64 [[INDVARS_IV_NEXT]] to i32
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i32 [[TMP20]], 121
; CHECK-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END:%.*]]
; CHECK: for.end:
; CHECK-NEXT: [[TMP21:%.*]] = extractelement <4 x float> [[TMP19]], i32 3
; CHECK-NEXT: [[ADD28:%.*]] = fadd float [[ADD6]], [[TMP21]]
; CHECK-NEXT: [[TMP22:%.*]] = extractelement <4 x float> [[TMP19]], i32 2
; CHECK-NEXT: [[ADD29:%.*]] = fadd float [[ADD28]], [[TMP22]]
; CHECK-NEXT: [[TMP23:%.*]] = extractelement <4 x float> [[TMP19]], i32 1
; CHECK-NEXT: [[ADD30:%.*]] = fadd float [[ADD29]], [[TMP23]]
; CHECK-NEXT: [[TMP24:%.*]] = extractelement <4 x float> [[TMP19]], i32 0
; CHECK-NEXT: [[ADD31:%.*]] = fadd float [[ADD30]], [[TMP24]]
; CHECK-NEXT: ret float [[ADD31]]
;
entry:
%0 = load float, float* %A, align 4
%arrayidx1 = getelementptr inbounds float, float* %A, i64 1
%1 = load float, float* %arrayidx1, align 4
%arrayidx2 = getelementptr inbounds float, float* %A, i64 2
%2 = load float, float* %arrayidx2, align 4
%arrayidx3 = getelementptr inbounds float, float* %A, i64 3
%3 = load float, float* %arrayidx3, align 4
%arrayidx4 = getelementptr inbounds float, float* %A, i64 4
%4 = load float, float* %arrayidx4, align 4
br label %for.body
for.body: ; preds = %for.body, %entry
%indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
%P.056 = phi float [ %4, %entry ], [ %add26, %for.body ]
%Y.055 = phi float [ %3, %entry ], [ %add21, %for.body ]
%B.054 = phi float [ %2, %entry ], [ %add16, %for.body ]
%G.053 = phi float [ %1, %entry ], [ %add11, %for.body ]
%R.052 = phi float [ %0, %entry ], [ %add6, %for.body ]
%5 = phi float [ %1, %entry ], [ %11, %for.body ]
%6 = phi float [ %0, %entry ], [ %9, %for.body ]
%mul = fmul float %6, 7.000000e+00
%add6 = fadd float %R.052, %mul
%mul10 = fmul float %5, 8.000000e+00
%add11 = fadd float %G.053, %mul10
%7 = add nsw i64 %indvars.iv, 2
%arrayidx14 = getelementptr inbounds float, float* %A, i64 %7
%8 = load float, float* %arrayidx14, align 4
%mul15 = fmul float %8, 9.000000e+00
%add16 = fadd float %B.054, %mul15
%indvars.iv.next = add nuw nsw i64 %indvars.iv, 3
%arrayidx19 = getelementptr inbounds float, float* %A, i64 %indvars.iv.next
%9 = load float, float* %arrayidx19, align 4
%mul20 = fmul float %9, 1.000000e+01
%add21 = fadd float %Y.055, %mul20
%10 = add nsw i64 %indvars.iv, 4
%arrayidx24 = getelementptr inbounds float, float* %A, i64 %10
%11 = load float, float* %arrayidx24, align 4
%mul25 = fmul float %11, 1.100000e+01
%add26 = fadd float %P.056, %mul25
%12 = trunc i64 %indvars.iv.next to i32
%cmp = icmp slt i32 %12, 121
br i1 %cmp, label %for.body, label %for.end
for.end: ; preds = %for.body
%add28 = fadd float %add6, %add11
%add29 = fadd float %add28, %add16
%add30 = fadd float %add29, %add21
%add31 = fadd float %add30, %add26
ret float %add31
}
; Make sure the order of phi nodes of different types does not prevent
; vectorization of same typed phi nodes.
define float @sort_phi_type(float* nocapture readonly %A) {
; CHECK-LABEL: @sort_phi_type(
; CHECK-NEXT: entry:
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
; CHECK: for.body:
; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = phi <4 x float> [ <float 1.000000e+01, float 1.000000e+01, float 1.000000e+01, float 1.000000e+01>, [[ENTRY]] ], [ [[TMP9:%.*]], [[FOR_BODY]] ]
; CHECK-NEXT: [[TMP1:%.*]] = extractelement <4 x float> [[TMP0]], i32 0
; CHECK-NEXT: [[TMP2:%.*]] = insertelement <4 x float> undef, float [[TMP1]], i32 0
; CHECK-NEXT: [[TMP3:%.*]] = extractelement <4 x float> [[TMP0]], i32 1
; CHECK-NEXT: [[TMP4:%.*]] = insertelement <4 x float> [[TMP2]], float [[TMP3]], i32 1
; CHECK-NEXT: [[TMP5:%.*]] = extractelement <4 x float> [[TMP0]], i32 3
; CHECK-NEXT: [[TMP6:%.*]] = insertelement <4 x float> [[TMP4]], float [[TMP5]], i32 2
; CHECK-NEXT: [[TMP7:%.*]] = extractelement <4 x float> [[TMP0]], i32 2
; CHECK-NEXT: [[TMP8:%.*]] = insertelement <4 x float> [[TMP6]], float [[TMP7]], i32 3
; CHECK-NEXT: [[TMP9]] = fmul <4 x float> <float 8.000000e+00, float 9.000000e+00, float 1.000000e+02, float 1.110000e+02>, [[TMP8]]
; CHECK-NEXT: [[INDVARS_IV_NEXT]] = add nsw i64 [[INDVARS_IV]], 4
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i64 [[INDVARS_IV_NEXT]], 128
; CHECK-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END:%.*]]
; CHECK: for.end:
; CHECK-NEXT: [[TMP10:%.*]] = extractelement <4 x float> [[TMP9]], i32 0
; CHECK-NEXT: [[TMP11:%.*]] = extractelement <4 x float> [[TMP9]], i32 1
; CHECK-NEXT: [[ADD29:%.*]] = fadd float [[TMP10]], [[TMP11]]
; CHECK-NEXT: [[TMP12:%.*]] = extractelement <4 x float> [[TMP9]], i32 2
; CHECK-NEXT: [[ADD30:%.*]] = fadd float [[ADD29]], [[TMP12]]
; CHECK-NEXT: [[TMP13:%.*]] = extractelement <4 x float> [[TMP9]], i32 3
; CHECK-NEXT: [[ADD31:%.*]] = fadd float [[ADD30]], [[TMP13]]
; CHECK-NEXT: ret float [[ADD31]]
;
entry:
br label %for.body
for.body: ; preds = %for.body, %entry
%Y = phi float [ 1.000000e+01, %entry ], [ %mul10, %for.body ]
%indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
%B = phi float [ 1.000000e+01, %entry ], [ %mul15, %for.body ]
%G = phi float [ 1.000000e+01, %entry ], [ %mul20, %for.body ]
%R = phi float [ 1.000000e+01, %entry ], [ %mul25, %for.body ]
%mul10 = fmul float %Y, 8.000000e+00
%mul15 = fmul float %B, 9.000000e+00
%mul20 = fmul float %R, 10.000000e+01
%mul25 = fmul float %G, 11.100000e+01
%indvars.iv.next = add nsw i64 %indvars.iv, 4
%cmp = icmp slt i64 %indvars.iv.next, 128
br i1 %cmp, label %for.body, label %for.end
for.end: ; preds = %for.body
%add28 = fadd float 1.000000e+01, %mul10
%add29 = fadd float %mul10, %mul15
%add30 = fadd float %add29, %mul20
%add31 = fadd float %add30, %mul25
ret float %add31
}
define void @test(x86_fp80* %i1, x86_fp80* %i2, x86_fp80* %o) {
; CHECK-LABEL: @test(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[I1_0:%.*]] = load x86_fp80, x86_fp80* [[I1:%.*]], align 16
; CHECK-NEXT: [[I1_GEP1:%.*]] = getelementptr x86_fp80, x86_fp80* [[I1]], i64 1
; CHECK-NEXT: [[I1_1:%.*]] = load x86_fp80, x86_fp80* [[I1_GEP1]], align 16
; CHECK-NEXT: br i1 undef, label [[THEN:%.*]], label [[END:%.*]]
; CHECK: then:
; CHECK-NEXT: [[I2_GEP0:%.*]] = getelementptr inbounds x86_fp80, x86_fp80* [[I2:%.*]], i64 0
; CHECK-NEXT: [[I2_0:%.*]] = load x86_fp80, x86_fp80* [[I2_GEP0]], align 16
; CHECK-NEXT: [[I2_GEP1:%.*]] = getelementptr inbounds x86_fp80, x86_fp80* [[I2]], i64 1
; CHECK-NEXT: [[I2_1:%.*]] = load x86_fp80, x86_fp80* [[I2_GEP1]], align 16
; CHECK-NEXT: br label [[END]]
; CHECK: end:
; CHECK-NEXT: [[PHI0:%.*]] = phi x86_fp80 [ [[I1_0]], [[ENTRY:%.*]] ], [ [[I2_0]], [[THEN]] ]
; CHECK-NEXT: [[PHI1:%.*]] = phi x86_fp80 [ [[I1_1]], [[ENTRY]] ], [ [[I2_1]], [[THEN]] ]
; CHECK-NEXT: store x86_fp80 [[PHI0]], x86_fp80* [[O:%.*]], align 16
; CHECK-NEXT: [[O_GEP1:%.*]] = getelementptr inbounds x86_fp80, x86_fp80* [[O]], i64 1
; CHECK-NEXT: store x86_fp80 [[PHI1]], x86_fp80* [[O_GEP1]], align 16
; CHECK-NEXT: ret void
;
; Test that we correctly recognize the discontiguous memory in arrays where the
; size is less than the alignment, and through various different GEP formations.
; We disable the vectorization of x86_fp80 for now.
entry:
%i1.0 = load x86_fp80, x86_fp80* %i1, align 16
%i1.gep1 = getelementptr x86_fp80, x86_fp80* %i1, i64 1
%i1.1 = load x86_fp80, x86_fp80* %i1.gep1, align 16
br i1 undef, label %then, label %end
then:
%i2.gep0 = getelementptr inbounds x86_fp80, x86_fp80* %i2, i64 0
%i2.0 = load x86_fp80, x86_fp80* %i2.gep0, align 16
%i2.gep1 = getelementptr inbounds x86_fp80, x86_fp80* %i2, i64 1
%i2.1 = load x86_fp80, x86_fp80* %i2.gep1, align 16
br label %end
end:
%phi0 = phi x86_fp80 [ %i1.0, %entry ], [ %i2.0, %then ]
%phi1 = phi x86_fp80 [ %i1.1, %entry ], [ %i2.1, %then ]
store x86_fp80 %phi0, x86_fp80* %o, align 16
%o.gep1 = getelementptr inbounds x86_fp80, x86_fp80* %o, i64 1
store x86_fp80 %phi1, x86_fp80* %o.gep1, align 16
ret void
}