caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[InstCombine] remove multi-use add demanded constant fold

Browse Source 
This was originally part of D133788. There are no visible
regressions. All of the diffs show a large unsigned constant
becoming a small negative constant. This should be better
for analysis (and slightly less compile-time) and codegen.
This commit is contained in:
Sanjay Patel
2022-09-18 14:21:21 -04:00
parent a2842a43a1
commit d6498abc24
12 changed files with 143 additions and 148 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
llvm/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp
View File

@@ -1078,11 +1078,6 @@ Value *InstCombinerImpl::SimplifyMultipleUseDemandedBits(
if (DemandedFromOps.isSubsetOf(LHSKnown.Zero))
return I->getOperand(1);
bool NSW = cast<OverflowingBinaryOperator>(I)->hasNoSignedWrap();
Known = KnownBits::computeForAddSub(true, NSW, LHSKnown, RHSKnown);
if (DemandedMask.isSubsetOf(Known.Zero|Known.One))
return Constant::getIntegerValue(ITy, Known.One);
break;
}
case Instruction::AShr: {

6
llvm/test/Transforms/LoopVectorize/AArch64/intrinsiccost.ll
View File

@@ -24,7 +24,7 @@ define void @saddsat(i16* nocapture readonly %pSrc, i16 signext %offset, i16* no
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 15
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934576
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -16
; CHECK-NEXT: [[CAST_VTC:%.*]] = trunc i64 [[N_VEC]] to i32
; CHECK-NEXT: [[IND_END:%.*]] = sub i32 [[BLOCKSIZE]], [[CAST_VTC]]
; CHECK-NEXT: [[IND_END1:%.*]] = getelementptr i16, i16* [[PSRC:%.*]], i64 [[N_VEC]]
@@ -119,7 +119,7 @@ define void @umin(i8* nocapture readonly %pSrc, i8 signext %offset, i8* nocaptur
; CHECK-NEXT: [[MIN_ITERS_CHECK1:%.*]] = icmp ult i32 [[TMP0]], 31
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK1]], label [[VEC_EPILOG_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934560
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -32
; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <16 x i8> poison, i8 [[OFFSET:%.*]], i64 0
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <16 x i8> [[BROADCAST_SPLATINSERT]], <16 x i8> poison, <16 x i32> zeroinitializer
; CHECK-NEXT: [[BROADCAST_SPLATINSERT6:%.*]] = insertelement <16 x i8> poison, i8 [[OFFSET]], i64 0
@@ -157,7 +157,7 @@ define void @umin(i8* nocapture readonly %pSrc, i8 signext %offset, i8* nocaptur
; CHECK-NEXT: br i1 [[MIN_EPILOG_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]]
; CHECK: vec.epilog.ph:
; CHECK-NEXT: [[VEC_EPILOG_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MAIN_LOOP_ITER_CHECK]] ]
; CHECK-NEXT: [[N_VEC9:%.*]] = and i64 [[TMP2]], 8589934584
; CHECK-NEXT: [[N_VEC9:%.*]] = and i64 [[TMP2]], -8
; CHECK-NEXT: [[CAST_VTC:%.*]] = trunc i64 [[N_VEC9]] to i32
; CHECK-NEXT: [[IND_END:%.*]] = sub i32 [[BLOCKSIZE]], [[CAST_VTC]]
; CHECK-NEXT: [[IND_END12:%.*]] = getelementptr i8, i8* [[PSRC]], i64 [[N_VEC9]]

18
llvm/test/Transforms/LoopVectorize/X86/float-induction-x86.ll
View File

@@ -25,8 +25,8 @@ define void @fp_iv_loop1(float* noalias nocapture %A, i32 %N) #0 {
; AUTO_VEC-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; AUTO_VEC: vector.ph:
; AUTO_VEC-NEXT: [[N_VEC:%.*]] = and i64 [[ZEXT]], 4294967264
; AUTO_VEC-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; AUTO_VEC-NEXT: [[TMP0:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; AUTO_VEC-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; AUTO_VEC-NEXT: [[TMP0:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; AUTO_VEC-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP0]], 1.000000e+00
; AUTO_VEC-NEXT: [[TMP1:%.*]] = add nsw i64 [[ZEXT]], -32
; AUTO_VEC-NEXT: [[TMP2:%.*]] = lshr i64 [[TMP1]], 5
@@ -35,7 +35,7 @@ define void @fp_iv_loop1(float* noalias nocapture %A, i32 %N) #0 {
; AUTO_VEC-NEXT: [[TMP4:%.*]] = icmp ult i64 [[TMP1]], 96
; AUTO_VEC-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK_UNR_LCSSA:%.*]], label [[VECTOR_PH_NEW:%.*]]
; AUTO_VEC: vector.ph.new:
; AUTO_VEC-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP3]], 1152921504606846972
; AUTO_VEC-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP3]], -4
; AUTO_VEC-NEXT: br label [[VECTOR_BODY:%.*]]
; AUTO_VEC: vector.body:
; AUTO_VEC-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH_NEW]] ], [ [[INDEX_NEXT_3:%.*]], [[VECTOR_BODY]] ]
@@ -296,8 +296,8 @@ define double @external_use_with_fast_math(double* %a, i64 %n) {
; AUTO_VEC-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; AUTO_VEC: vector.ph:
; AUTO_VEC-NEXT: [[N_VEC:%.*]] = and i64 [[SMAX]], 9223372036854775792
; AUTO_VEC-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to double
; AUTO_VEC-NEXT: [[TMP0:%.*]] = fmul fast double [[CAST_CRD]], 3.000000e+00
; AUTO_VEC-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to double
; AUTO_VEC-NEXT: [[TMP0:%.*]] = fmul fast double [[CAST_VTC]], 3.000000e+00
; AUTO_VEC-NEXT: [[TMP1:%.*]] = add nsw i64 [[SMAX]], -16
; AUTO_VEC-NEXT: [[TMP2:%.*]] = lshr i64 [[TMP1]], 4
; AUTO_VEC-NEXT: [[TMP3:%.*]] = add nuw nsw i64 [[TMP2]], 1
@@ -305,7 +305,7 @@ define double @external_use_with_fast_math(double* %a, i64 %n) {
; AUTO_VEC-NEXT: [[TMP4:%.*]] = icmp ult i64 [[TMP1]], 48
; AUTO_VEC-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK_UNR_LCSSA:%.*]], label [[VECTOR_PH_NEW:%.*]]
; AUTO_VEC: vector.ph.new:
; AUTO_VEC-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP3]], 2305843009213693948
; AUTO_VEC-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP3]], -4
; AUTO_VEC-NEXT: br label [[VECTOR_BODY:%.*]]
; AUTO_VEC: vector.body:
; AUTO_VEC-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH_NEW]] ], [ [[INDEX_NEXT_3:%.*]], [[VECTOR_BODY]] ]
@@ -556,8 +556,8 @@ define void @fadd_reassoc_FMF(float* nocapture %p, i32 %N) {
; AUTO_VEC-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; AUTO_VEC: vector.ph:
; AUTO_VEC-NEXT: [[N_VEC:%.*]] = and i64 [[TMP0]], 4294967264
; AUTO_VEC-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; AUTO_VEC-NEXT: [[TMP1:%.*]] = fmul reassoc float [[CAST_CRD]], 4.200000e+01
; AUTO_VEC-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; AUTO_VEC-NEXT: [[TMP1:%.*]] = fmul reassoc float [[CAST_VTC]], 4.200000e+01
; AUTO_VEC-NEXT: [[IND_END:%.*]] = fadd reassoc float [[TMP1]], 1.000000e+00
; AUTO_VEC-NEXT: [[TMP2:%.*]] = add nsw i64 [[TMP0]], -32
; AUTO_VEC-NEXT: [[TMP3:%.*]] = lshr i64 [[TMP2]], 5
@@ -566,7 +566,7 @@ define void @fadd_reassoc_FMF(float* nocapture %p, i32 %N) {
; AUTO_VEC-NEXT: [[TMP5:%.*]] = icmp ult i64 [[TMP2]], 32
; AUTO_VEC-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK_UNR_LCSSA:%.*]], label [[VECTOR_PH_NEW:%.*]]
; AUTO_VEC: vector.ph.new:
; AUTO_VEC-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP4]], 1152921504606846974
; AUTO_VEC-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP4]], -2
; AUTO_VEC-NEXT: br label [[VECTOR_BODY:%.*]]
; AUTO_VEC: vector.body:
; AUTO_VEC-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH_NEW]] ], [ [[INDEX_NEXT_1:%.*]], [[VECTOR_BODY]] ]

8
llvm/test/Transforms/LoopVectorize/X86/intrinsiccost.ll
View File

@@ -28,7 +28,7 @@ define void @uaddsat(i16* nocapture readonly %pSrc, i16 signext %offset, i16* no
; CHECK-NEXT: [[MIN_ITERS_CHECK1:%.*]] = icmp ult i32 [[TMP0]], 63
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK1]], label [[VEC_EPILOG_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934528
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -64
; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <16 x i16> poison, i16 [[OFFSET:%.*]], i64 0
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <16 x i16> [[BROADCAST_SPLATINSERT]], <16 x i16> poison, <16 x i32> zeroinitializer
; CHECK-NEXT: [[BROADCAST_SPLATINSERT12:%.*]] = insertelement <16 x i16> poison, i16 [[OFFSET]], i64 0
@@ -84,7 +84,7 @@ define void @uaddsat(i16* nocapture readonly %pSrc, i16 signext %offset, i16* no
; CHECK-NEXT: br i1 [[MIN_EPILOG_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]]
; CHECK: vec.epilog.ph:
; CHECK-NEXT: [[VEC_EPILOG_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MAIN_LOOP_ITER_CHECK]] ]
; CHECK-NEXT: [[N_VEC19:%.*]] = and i64 [[TMP2]], 8589934584
; CHECK-NEXT: [[N_VEC19:%.*]] = and i64 [[TMP2]], -8
; CHECK-NEXT: [[CAST_VTC:%.*]] = trunc i64 [[N_VEC19]] to i32
; CHECK-NEXT: [[IND_END:%.*]] = sub i32 [[BLOCKSIZE]], [[CAST_VTC]]
; CHECK-NEXT: [[IND_END22:%.*]] = getelementptr i16, i16* [[PSRC]], i64 [[N_VEC19]]
@@ -171,7 +171,7 @@ define void @cttz(i8* nocapture readonly %pSrc, i8 signext %offset, i8* nocaptur
; CHECK-NEXT: [[MIN_ITERS_CHECK1:%.*]] = icmp ult i32 [[TMP0]], 127
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK1]], label [[VEC_EPILOG_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934464
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -128
; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <32 x i8> poison, i8 [[OFFSET:%.*]], i64 0
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <32 x i8> [[BROADCAST_SPLATINSERT]], <32 x i8> poison, <32 x i32> zeroinitializer
; CHECK-NEXT: [[BROADCAST_SPLATINSERT12:%.*]] = insertelement <32 x i8> poison, i8 [[OFFSET]], i64 0
@@ -227,7 +227,7 @@ define void @cttz(i8* nocapture readonly %pSrc, i8 signext %offset, i8* nocaptur
; CHECK-NEXT: br i1 [[MIN_EPILOG_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]]
; CHECK: vec.epilog.ph:
; CHECK-NEXT: [[VEC_EPILOG_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MAIN_LOOP_ITER_CHECK]] ]
; CHECK-NEXT: [[N_VEC19:%.*]] = and i64 [[TMP2]], 8589934576
; CHECK-NEXT: [[N_VEC19:%.*]] = and i64 [[TMP2]], -16
; CHECK-NEXT: [[CAST_VTC:%.*]] = trunc i64 [[N_VEC19]] to i32
; CHECK-NEXT: [[IND_END:%.*]] = sub i32 [[BLOCKSIZE]], [[CAST_VTC]]
; CHECK-NEXT: [[IND_END22:%.*]] = getelementptr i8, i8* [[PSRC]], i64 [[N_VEC19]]

208
llvm/test/Transforms/LoopVectorize/float-induction.ll
View File

@@ -30,9 +30,9 @@ define void @fp_iv_loop1_fast_FMF(float %init, float* noalias nocapture %A, i32
; VEC4_INTERL1-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 3
; VEC4_INTERL1-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL1: vector.ph:
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588
; VEC4_INTERL1-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP3:%.*]] = fmul fast float [[FPINC]], [[CAST_CRD]]
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; VEC4_INTERL1-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP3:%.*]] = fmul fast float [[FPINC]], [[CAST_VTC]]
; VEC4_INTERL1-NEXT: [[IND_END:%.*]] = fsub fast float [[INIT:%.*]], [[TMP3]]
; VEC4_INTERL1-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[INIT]], i64 0
; VEC4_INTERL1-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
@@ -88,9 +88,9 @@ define void @fp_iv_loop1_fast_FMF(float %init, float* noalias nocapture %A, i32
; VEC4_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 7
; VEC4_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL2: vector.ph:
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934584
; VEC4_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[FPINC]], [[CAST_CRD]]
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -8
; VEC4_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[FPINC]], [[CAST_VTC]]
; VEC4_INTERL2-NEXT: [[IND_END:%.*]] = fsub fast float [[INIT:%.*]], [[TMP3]]
; VEC4_INTERL2-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[INIT]], i64 0
; VEC4_INTERL2-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
@@ -150,9 +150,9 @@ define void @fp_iv_loop1_fast_FMF(float %init, float* noalias nocapture %A, i32
; VEC1_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; VEC1_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC1_INTERL2: vector.ph:
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; VEC1_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[FPINC]], [[CAST_CRD]]
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; VEC1_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[FPINC]], [[CAST_VTC]]
; VEC1_INTERL2-NEXT: [[IND_END:%.*]] = fsub fast float [[INIT:%.*]], [[TMP3]]
; VEC1_INTERL2-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC1_INTERL2: vector.body:
@@ -203,9 +203,9 @@ define void @fp_iv_loop1_fast_FMF(float %init, float* noalias nocapture %A, i32
; VEC2_INTERL1_PRED_STORE-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; VEC2_INTERL1_PRED_STORE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; VEC2_INTERL1_PRED_STORE: vector.ph:
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP3:%.*]] = fmul fast float [[FPINC]], [[CAST_CRD]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP3:%.*]] = fmul fast float [[FPINC]], [[CAST_VTC]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[IND_END:%.*]] = fsub fast float [[INIT:%.*]], [[TMP3]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <2 x float> poison, float [[INIT]], i64 0
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <2 x float> [[DOTSPLATINSERT]], <2 x float> poison, <2 x i32> zeroinitializer
@@ -289,9 +289,9 @@ define void @fp_iv_loop1_reassoc_FMF(float %init, float* noalias nocapture %A, i
; VEC4_INTERL1-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 3
; VEC4_INTERL1-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL1: vector.ph:
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588
; VEC4_INTERL1-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP3:%.*]] = fmul reassoc float [[FPINC]], [[CAST_CRD]]
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; VEC4_INTERL1-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP3:%.*]] = fmul reassoc float [[FPINC]], [[CAST_VTC]]
; VEC4_INTERL1-NEXT: [[IND_END:%.*]] = fsub reassoc float [[INIT:%.*]], [[TMP3]]
; VEC4_INTERL1-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[INIT]], i64 0
; VEC4_INTERL1-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
@@ -347,9 +347,9 @@ define void @fp_iv_loop1_reassoc_FMF(float %init, float* noalias nocapture %A, i
; VEC4_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 7
; VEC4_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL2: vector.ph:
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934584
; VEC4_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP3:%.*]] = fmul reassoc float [[FPINC]], [[CAST_CRD]]
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -8
; VEC4_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP3:%.*]] = fmul reassoc float [[FPINC]], [[CAST_VTC]]
; VEC4_INTERL2-NEXT: [[IND_END:%.*]] = fsub reassoc float [[INIT:%.*]], [[TMP3]]
; VEC4_INTERL2-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[INIT]], i64 0
; VEC4_INTERL2-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
@@ -409,9 +409,9 @@ define void @fp_iv_loop1_reassoc_FMF(float %init, float* noalias nocapture %A, i
; VEC1_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; VEC1_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC1_INTERL2: vector.ph:
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; VEC1_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP3:%.*]] = fmul reassoc float [[FPINC]], [[CAST_CRD]]
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; VEC1_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP3:%.*]] = fmul reassoc float [[FPINC]], [[CAST_VTC]]
; VEC1_INTERL2-NEXT: [[IND_END:%.*]] = fsub reassoc float [[INIT:%.*]], [[TMP3]]
; VEC1_INTERL2-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC1_INTERL2: vector.body:
@@ -464,9 +464,9 @@ define void @fp_iv_loop1_reassoc_FMF(float %init, float* noalias nocapture %A, i
; VEC2_INTERL1_PRED_STORE-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; VEC2_INTERL1_PRED_STORE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; VEC2_INTERL1_PRED_STORE: vector.ph:
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP3:%.*]] = fmul reassoc float [[FPINC]], [[CAST_CRD]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP3:%.*]] = fmul reassoc float [[FPINC]], [[CAST_VTC]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[IND_END:%.*]] = fsub reassoc float [[INIT:%.*]], [[TMP3]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <2 x float> poison, float [[INIT]], i64 0
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <2 x float> [[DOTSPLATINSERT]], <2 x float> poison, <2 x i32> zeroinitializer
@@ -551,9 +551,9 @@ define void @fp_iv_loop2(float %init, float* noalias nocapture %A, i32 %N) #0 {
; VEC4_INTERL1-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 3
; VEC4_INTERL1-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL1: vector.ph:
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588
; VEC4_INTERL1-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; VEC4_INTERL1-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; VEC4_INTERL1-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP3]], [[INIT:%.*]]
; VEC4_INTERL1-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[INIT]], i64 0
; VEC4_INTERL1-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
@@ -602,9 +602,9 @@ define void @fp_iv_loop2(float %init, float* noalias nocapture %A, i32 %N) #0 {
; VEC4_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 7
; VEC4_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL2: vector.ph:
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934584
; VEC4_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -8
; VEC4_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; VEC4_INTERL2-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP3]], [[INIT:%.*]]
; VEC4_INTERL2-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[INIT]], i64 0
; VEC4_INTERL2-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
@@ -657,9 +657,9 @@ define void @fp_iv_loop2(float %init, float* noalias nocapture %A, i32 %N) #0 {
; VEC1_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; VEC1_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC1_INTERL2: vector.ph:
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; VEC1_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; VEC1_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; VEC1_INTERL2-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP3]], [[INIT:%.*]]
; VEC1_INTERL2-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC1_INTERL2: vector.body:
@@ -709,9 +709,9 @@ define void @fp_iv_loop2(float %init, float* noalias nocapture %A, i32 %N) #0 {
; VEC2_INTERL1_PRED_STORE-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; VEC2_INTERL1_PRED_STORE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; VEC2_INTERL1_PRED_STORE: vector.ph:
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; VEC2_INTERL1_PRED_STORE-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP3]], [[INIT:%.*]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <2 x float> poison, float [[INIT]], i64 0
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <2 x float> [[DOTSPLATINSERT]], <2 x float> poison, <2 x i32> zeroinitializer
@@ -795,13 +795,13 @@ define void @fp_iv_loop3(float %init, float* noalias nocapture %A, float* noalia
; VEC4_INTERL1-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP1]], 3
; VEC4_INTERL1-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL1: vector.ph:
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP3]], 8589934588
; VEC4_INTERL1-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP4:%.*]] = fmul fast float [[CAST_CRD]], -5.000000e-01
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP3]], -4
; VEC4_INTERL1-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP4:%.*]] = fmul fast float [[CAST_VTC]], -5.000000e-01
; VEC4_INTERL1-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP4]], 0x3FB99999A0000000
; VEC4_INTERL1-NEXT: [[CAST_CRD3:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP5:%.*]] = fmul fast float [[TMP0]], [[CAST_CRD3]]
; VEC4_INTERL1-NEXT: [[IND_END4:%.*]] = fadd fast float [[TMP5]], [[INIT:%.*]]
; VEC4_INTERL1-NEXT: [[CAST_VTC2:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP5:%.*]] = fmul fast float [[TMP0]], [[CAST_VTC2]]
; VEC4_INTERL1-NEXT: [[IND_END3:%.*]] = fadd fast float [[TMP5]], [[INIT:%.*]]
; VEC4_INTERL1-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[INIT]], i64 0
; VEC4_INTERL1-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
; VEC4_INTERL1-NEXT: [[DOTSPLATINSERT5:%.*]] = insertelement <4 x float> poison, float [[TMP0]], i64 0
@@ -841,12 +841,12 @@ define void @fp_iv_loop3(float %init, float* noalias nocapture %A, float* noalia
; VEC4_INTERL1: scalar.ph:
; VEC4_INTERL1-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[FOR_BODY_LR_PH]] ]
; VEC4_INTERL1-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 0x3FB99999A0000000, [[FOR_BODY_LR_PH]] ]
; VEC4_INTERL1-NEXT: [[BC_RESUME_VAL2:%.*]] = phi float [ [[IND_END4]], [[MIDDLE_BLOCK]] ], [ [[INIT]], [[FOR_BODY_LR_PH]] ]
; VEC4_INTERL1-NEXT: [[BC_RESUME_VAL4:%.*]] = phi float [ [[IND_END3]], [[MIDDLE_BLOCK]] ], [ [[INIT]], [[FOR_BODY_LR_PH]] ]
; VEC4_INTERL1-NEXT: br label [[FOR_BODY:%.*]]
; VEC4_INTERL1: for.body:
; VEC4_INTERL1-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; VEC4_INTERL1-NEXT: [[Y_012:%.*]] = phi float [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ], [ [[CONV1:%.*]], [[FOR_BODY]] ]
; VEC4_INTERL1-NEXT: [[X_011:%.*]] = phi float [ [[BC_RESUME_VAL2]], [[SCALAR_PH]] ], [ [[ADD:%.*]], [[FOR_BODY]] ]
; VEC4_INTERL1-NEXT: [[X_011:%.*]] = phi float [ [[BC_RESUME_VAL4]], [[SCALAR_PH]] ], [ [[ADD:%.*]], [[FOR_BODY]] ]
; VEC4_INTERL1-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[INDVARS_IV]]
; VEC4_INTERL1-NEXT: store float [[X_011]], float* [[ARRAYIDX]], align 4
; VEC4_INTERL1-NEXT: [[ADD]] = fadd fast float [[X_011]], [[TMP0]]
@@ -877,13 +877,13 @@ define void @fp_iv_loop3(float %init, float* noalias nocapture %A, float* noalia
; VEC4_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP1]], 7
; VEC4_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL2: vector.ph:
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP3]], 8589934584
; VEC4_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP4:%.*]] = fmul fast float [[CAST_CRD]], -5.000000e-01
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP3]], -8
; VEC4_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP4:%.*]] = fmul fast float [[CAST_VTC]], -5.000000e-01
; VEC4_INTERL2-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP4]], 0x3FB99999A0000000
; VEC4_INTERL2-NEXT: [[CAST_CRD3:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP5:%.*]] = fmul fast float [[TMP0]], [[CAST_CRD3]]
; VEC4_INTERL2-NEXT: [[IND_END4:%.*]] = fadd fast float [[TMP5]], [[INIT:%.*]]
; VEC4_INTERL2-NEXT: [[CAST_VTC2:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP5:%.*]] = fmul fast float [[TMP0]], [[CAST_VTC2]]
; VEC4_INTERL2-NEXT: [[IND_END3:%.*]] = fadd fast float [[TMP5]], [[INIT:%.*]]
; VEC4_INTERL2-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[INIT]], i64 0
; VEC4_INTERL2-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
; VEC4_INTERL2-NEXT: [[DOTSPLATINSERT6:%.*]] = insertelement <4 x float> poison, float [[TMP0]], i64 0
@@ -938,12 +938,12 @@ define void @fp_iv_loop3(float %init, float* noalias nocapture %A, float* noalia
; VEC4_INTERL2: scalar.ph:
; VEC4_INTERL2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[FOR_BODY_LR_PH]] ]
; VEC4_INTERL2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 0x3FB99999A0000000, [[FOR_BODY_LR_PH]] ]
; VEC4_INTERL2-NEXT: [[BC_RESUME_VAL2:%.*]] = phi float [ [[IND_END4]], [[MIDDLE_BLOCK]] ], [ [[INIT]], [[FOR_BODY_LR_PH]] ]
; VEC4_INTERL2-NEXT: [[BC_RESUME_VAL4:%.*]] = phi float [ [[IND_END3]], [[MIDDLE_BLOCK]] ], [ [[INIT]], [[FOR_BODY_LR_PH]] ]
; VEC4_INTERL2-NEXT: br label [[FOR_BODY:%.*]]
; VEC4_INTERL2: for.body:
; VEC4_INTERL2-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; VEC4_INTERL2-NEXT: [[Y_012:%.*]] = phi float [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ], [ [[CONV1:%.*]], [[FOR_BODY]] ]
; VEC4_INTERL2-NEXT: [[X_011:%.*]] = phi float [ [[BC_RESUME_VAL2]], [[SCALAR_PH]] ], [ [[ADD:%.*]], [[FOR_BODY]] ]
; VEC4_INTERL2-NEXT: [[X_011:%.*]] = phi float [ [[BC_RESUME_VAL4]], [[SCALAR_PH]] ], [ [[ADD:%.*]], [[FOR_BODY]] ]
; VEC4_INTERL2-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[INDVARS_IV]]
; VEC4_INTERL2-NEXT: store float [[X_011]], float* [[ARRAYIDX]], align 4
; VEC4_INTERL2-NEXT: [[ADD]] = fadd fast float [[X_011]], [[TMP0]]
@@ -974,13 +974,13 @@ define void @fp_iv_loop3(float %init, float* noalias nocapture %A, float* noalia
; VEC1_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP1]], 0
; VEC1_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC1_INTERL2: vector.ph:
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP3]], 8589934590
; VEC1_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP4:%.*]] = fmul fast float [[CAST_CRD]], -5.000000e-01
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP3]], -2
; VEC1_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP4:%.*]] = fmul fast float [[CAST_VTC]], -5.000000e-01
; VEC1_INTERL2-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP4]], 0x3FB99999A0000000
; VEC1_INTERL2-NEXT: [[CAST_CRD3:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP5:%.*]] = fmul fast float [[TMP0]], [[CAST_CRD3]]
; VEC1_INTERL2-NEXT: [[IND_END4:%.*]] = fadd fast float [[TMP5]], [[INIT:%.*]]
; VEC1_INTERL2-NEXT: [[CAST_VTC2:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP5:%.*]] = fmul fast float [[TMP0]], [[CAST_VTC2]]
; VEC1_INTERL2-NEXT: [[IND_END3:%.*]] = fadd fast float [[TMP5]], [[INIT:%.*]]
; VEC1_INTERL2-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC1_INTERL2: vector.body:
; VEC1_INTERL2-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
@@ -1018,12 +1018,12 @@ define void @fp_iv_loop3(float %init, float* noalias nocapture %A, float* noalia
; VEC1_INTERL2: scalar.ph:
; VEC1_INTERL2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[FOR_BODY_LR_PH]] ]
; VEC1_INTERL2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 0x3FB99999A0000000, [[FOR_BODY_LR_PH]] ]
; VEC1_INTERL2-NEXT: [[BC_RESUME_VAL2:%.*]] = phi float [ [[IND_END4]], [[MIDDLE_BLOCK]] ], [ [[INIT]], [[FOR_BODY_LR_PH]] ]
; VEC1_INTERL2-NEXT: [[BC_RESUME_VAL4:%.*]] = phi float [ [[IND_END3]], [[MIDDLE_BLOCK]] ], [ [[INIT]], [[FOR_BODY_LR_PH]] ]
; VEC1_INTERL2-NEXT: br label [[FOR_BODY:%.*]]
; VEC1_INTERL2: for.body:
; VEC1_INTERL2-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; VEC1_INTERL2-NEXT: [[Y_012:%.*]] = phi float [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ], [ [[CONV1:%.*]], [[FOR_BODY]] ]
; VEC1_INTERL2-NEXT: [[X_011:%.*]] = phi float [ [[BC_RESUME_VAL2]], [[SCALAR_PH]] ], [ [[ADD:%.*]], [[FOR_BODY]] ]
; VEC1_INTERL2-NEXT: [[X_011:%.*]] = phi float [ [[BC_RESUME_VAL4]], [[SCALAR_PH]] ], [ [[ADD:%.*]], [[FOR_BODY]] ]
; VEC1_INTERL2-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[INDVARS_IV]]
; VEC1_INTERL2-NEXT: store float [[X_011]], float* [[ARRAYIDX]], align 4
; VEC1_INTERL2-NEXT: [[ADD]] = fadd fast float [[X_011]], [[TMP0]]
@@ -1054,13 +1054,13 @@ define void @fp_iv_loop3(float %init, float* noalias nocapture %A, float* noalia
; VEC2_INTERL1_PRED_STORE-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP1]], 0
; VEC2_INTERL1_PRED_STORE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; VEC2_INTERL1_PRED_STORE: vector.ph:
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP3]], 8589934590
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP4:%.*]] = fmul fast float [[CAST_CRD]], -5.000000e-01
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP3]], -2
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP4:%.*]] = fmul fast float [[CAST_VTC]], -5.000000e-01
; VEC2_INTERL1_PRED_STORE-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP4]], 0x3FB99999A0000000
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_CRD3:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP5:%.*]] = fmul fast float [[TMP0]], [[CAST_CRD3]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[IND_END4:%.*]] = fadd fast float [[TMP5]], [[INIT:%.*]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_VTC2:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP5:%.*]] = fmul fast float [[TMP0]], [[CAST_VTC2]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[IND_END3:%.*]] = fadd fast float [[TMP5]], [[INIT:%.*]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <2 x float> poison, float [[INIT]], i64 0
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <2 x float> [[DOTSPLATINSERT]], <2 x float> poison, <2 x i32> zeroinitializer
; VEC2_INTERL1_PRED_STORE-NEXT: [[DOTSPLATINSERT5:%.*]] = insertelement <2 x float> poison, float [[TMP0]], i64 0
@@ -1100,7 +1100,7 @@ define void @fp_iv_loop3(float %init, float* noalias nocapture %A, float* noalia
; VEC2_INTERL1_PRED_STORE: for.body:
; VEC2_INTERL1_PRED_STORE-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ], [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[FOR_BODY_LR_PH]] ]
; VEC2_INTERL1_PRED_STORE-NEXT: [[Y_012:%.*]] = phi float [ [[CONV1:%.*]], [[FOR_BODY]] ], [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 0x3FB99999A0000000, [[FOR_BODY_LR_PH]] ]
; VEC2_INTERL1_PRED_STORE-NEXT: [[X_011:%.*]] = phi float [ [[ADD:%.*]], [[FOR_BODY]] ], [ [[IND_END4]], [[MIDDLE_BLOCK]] ], [ [[INIT]], [[FOR_BODY_LR_PH]] ]
; VEC2_INTERL1_PRED_STORE-NEXT: [[X_011:%.*]] = phi float [ [[ADD:%.*]], [[FOR_BODY]] ], [ [[IND_END3]], [[MIDDLE_BLOCK]] ], [ [[INIT]], [[FOR_BODY_LR_PH]] ]
; VEC2_INTERL1_PRED_STORE-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[INDVARS_IV]]
; VEC2_INTERL1_PRED_STORE-NEXT: store float [[X_011]], float* [[ARRAYIDX]], align 4
; VEC2_INTERL1_PRED_STORE-NEXT: [[ADD]] = fadd fast float [[X_011]], [[TMP0]]
@@ -1172,9 +1172,9 @@ define void @fp_iv_loop4(float* noalias nocapture %A, i32 %N) {
; VEC4_INTERL1-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 3
; VEC4_INTERL1-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL1: vector.ph:
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588
; VEC4_INTERL1-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; VEC4_INTERL1-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; VEC4_INTERL1-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP3]], 1.000000e+00
; VEC4_INTERL1-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC4_INTERL1: vector.body:
@@ -1220,9 +1220,9 @@ define void @fp_iv_loop4(float* noalias nocapture %A, i32 %N) {
; VEC4_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 7
; VEC4_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL2: vector.ph:
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934584
; VEC4_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -8
; VEC4_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; VEC4_INTERL2-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP3]], 1.000000e+00
; VEC4_INTERL2-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC4_INTERL2: vector.body:
@@ -1272,9 +1272,9 @@ define void @fp_iv_loop4(float* noalias nocapture %A, i32 %N) {
; VEC1_INTERL2-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; VEC1_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC1_INTERL2: vector.ph:
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; VEC1_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; VEC1_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; VEC1_INTERL2-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP3]], 1.000000e+00
; VEC1_INTERL2-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC1_INTERL2: vector.body:
@@ -1324,9 +1324,9 @@ define void @fp_iv_loop4(float* noalias nocapture %A, i32 %N) {
; VEC2_INTERL1_PRED_STORE-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; VEC2_INTERL1_PRED_STORE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; VEC2_INTERL1_PRED_STORE: vector.ph:
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_CRD]], 5.000000e-01
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[TMP3:%.*]] = fmul fast float [[CAST_VTC]], 5.000000e-01
; VEC2_INTERL1_PRED_STORE-NEXT: [[IND_END:%.*]] = fadd fast float [[TMP3]], 1.000000e+00
; VEC2_INTERL1_PRED_STORE-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC2_INTERL1_PRED_STORE: vector.body:
@@ -1389,7 +1389,7 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC4_INTERL1-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL1: vector.ph:
; VEC4_INTERL1-NEXT: [[N_VEC:%.*]] = and i64 [[SMAX]], 9223372036854775804
; VEC4_INTERL1-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL1-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC4_INTERL1: vector.body:
; VEC4_INTERL1-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[PRED_STORE_CONTINUE7:%.*]] ]
@@ -1440,7 +1440,7 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC4_INTERL1-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]]
; VEC4_INTERL1: scalar.ph:
; VEC4_INTERL1-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; VEC4_INTERL1-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[CAST_CRD]], [[MIDDLE_BLOCK]] ], [ 0.000000e+00, [[ENTRY]] ]
; VEC4_INTERL1-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[CAST_VTC]], [[MIDDLE_BLOCK]] ], [ 0.000000e+00, [[ENTRY]] ]
; VEC4_INTERL1-NEXT: br label [[FOR_BODY:%.*]]
; VEC4_INTERL1: for.body:
; VEC4_INTERL1-NEXT: [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_INC:%.*]] ], [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ]
@@ -1467,10 +1467,10 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC4_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC4_INTERL2: vector.ph:
; VEC4_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[SMAX]], 9223372036854775800
; VEC4_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC4_INTERL2-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC4_INTERL2: vector.body:
; VEC4_INTERL2-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[PRED_STORE_CONTINUE17:%.*]] ]
; VEC4_INTERL2-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[PRED_STORE_CONTINUE16:%.*]] ]
; VEC4_INTERL2-NEXT: [[TMP0:%.*]] = sitofp i64 [[INDEX]] to float
; VEC4_INTERL2-NEXT: [[TMP1:%.*]] = or i64 [[INDEX]], 4
; VEC4_INTERL2-NEXT: [[TMP2:%.*]] = getelementptr inbounds float, float* [[A:%.*]], i64 [[INDEX]]
@@ -1478,9 +1478,9 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC4_INTERL2-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, <4 x float>* [[TMP3]], align 4
; VEC4_INTERL2-NEXT: [[TMP4:%.*]] = getelementptr inbounds float, float* [[TMP2]], i64 4
; VEC4_INTERL2-NEXT: [[TMP5:%.*]] = bitcast float* [[TMP4]] to <4 x float>*
; VEC4_INTERL2-NEXT: [[WIDE_LOAD3:%.*]] = load <4 x float>, <4 x float>* [[TMP5]], align 4
; VEC4_INTERL2-NEXT: [[WIDE_LOAD2:%.*]] = load <4 x float>, <4 x float>* [[TMP5]], align 4
; VEC4_INTERL2-NEXT: [[TMP6:%.*]] = fcmp fast oeq <4 x float> [[WIDE_LOAD]], zeroinitializer
; VEC4_INTERL2-NEXT: [[TMP7:%.*]] = fcmp fast oeq <4 x float> [[WIDE_LOAD3]], zeroinitializer
; VEC4_INTERL2-NEXT: [[TMP7:%.*]] = fcmp fast oeq <4 x float> [[WIDE_LOAD2]], zeroinitializer
; VEC4_INTERL2-NEXT: [[TMP8:%.*]] = extractelement <4 x i1> [[TMP6]], i64 0
; VEC4_INTERL2-NEXT: br i1 [[TMP8]], label [[PRED_STORE_IF:%.*]], label [[PRED_STORE_CONTINUE:%.*]]
; VEC4_INTERL2: pred.store.if:
@@ -1489,66 +1489,66 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE]]
; VEC4_INTERL2: pred.store.continue:
; VEC4_INTERL2-NEXT: [[TMP10:%.*]] = extractelement <4 x i1> [[TMP6]], i64 1
; VEC4_INTERL2-NEXT: br i1 [[TMP10]], label [[PRED_STORE_IF4:%.*]], label [[PRED_STORE_CONTINUE5:%.*]]
; VEC4_INTERL2-NEXT: br i1 [[TMP10]], label [[PRED_STORE_IF3:%.*]], label [[PRED_STORE_CONTINUE4:%.*]]
; VEC4_INTERL2: pred.store.if3:
; VEC4_INTERL2-NEXT: [[TMP11:%.*]] = fadd fast float [[TMP0]], 1.000000e+00
; VEC4_INTERL2-NEXT: [[TMP12:%.*]] = or i64 [[INDEX]], 1
; VEC4_INTERL2-NEXT: [[TMP13:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[TMP12]]
; VEC4_INTERL2-NEXT: store float [[TMP11]], float* [[TMP13]], align 4
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE5]]
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE4]]
; VEC4_INTERL2: pred.store.continue4:
; VEC4_INTERL2-NEXT: [[TMP14:%.*]] = extractelement <4 x i1> [[TMP6]], i64 2
; VEC4_INTERL2-NEXT: br i1 [[TMP14]], label [[PRED_STORE_IF6:%.*]], label [[PRED_STORE_CONTINUE7:%.*]]
; VEC4_INTERL2-NEXT: br i1 [[TMP14]], label [[PRED_STORE_IF5:%.*]], label [[PRED_STORE_CONTINUE6:%.*]]
; VEC4_INTERL2: pred.store.if5:
; VEC4_INTERL2-NEXT: [[TMP15:%.*]] = fadd fast float [[TMP0]], 2.000000e+00
; VEC4_INTERL2-NEXT: [[TMP16:%.*]] = or i64 [[INDEX]], 2
; VEC4_INTERL2-NEXT: [[TMP17:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[TMP16]]
; VEC4_INTERL2-NEXT: store float [[TMP15]], float* [[TMP17]], align 4
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE7]]
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE6]]
; VEC4_INTERL2: pred.store.continue6:
; VEC4_INTERL2-NEXT: [[TMP18:%.*]] = extractelement <4 x i1> [[TMP6]], i64 3
; VEC4_INTERL2-NEXT: br i1 [[TMP18]], label [[PRED_STORE_IF8:%.*]], label [[PRED_STORE_CONTINUE9:%.*]]
; VEC4_INTERL2-NEXT: br i1 [[TMP18]], label [[PRED_STORE_IF7:%.*]], label [[PRED_STORE_CONTINUE8:%.*]]
; VEC4_INTERL2: pred.store.if7:
; VEC4_INTERL2-NEXT: [[TMP19:%.*]] = fadd fast float [[TMP0]], 3.000000e+00
; VEC4_INTERL2-NEXT: [[TMP20:%.*]] = or i64 [[INDEX]], 3
; VEC4_INTERL2-NEXT: [[TMP21:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[TMP20]]
; VEC4_INTERL2-NEXT: store float [[TMP19]], float* [[TMP21]], align 4
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE9]]
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE8]]
; VEC4_INTERL2: pred.store.continue8:
; VEC4_INTERL2-NEXT: [[TMP22:%.*]] = extractelement <4 x i1> [[TMP7]], i64 0
; VEC4_INTERL2-NEXT: br i1 [[TMP22]], label [[PRED_STORE_IF10:%.*]], label [[PRED_STORE_CONTINUE11:%.*]]
; VEC4_INTERL2-NEXT: br i1 [[TMP22]], label [[PRED_STORE_IF9:%.*]], label [[PRED_STORE_CONTINUE10:%.*]]
; VEC4_INTERL2: pred.store.if9:
; VEC4_INTERL2-NEXT: [[TMP23:%.*]] = fadd fast float [[TMP0]], 4.000000e+00
; VEC4_INTERL2-NEXT: [[TMP24:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[TMP1]]
; VEC4_INTERL2-NEXT: store float [[TMP23]], float* [[TMP24]], align 4
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE11]]
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE10]]
; VEC4_INTERL2: pred.store.continue10:
; VEC4_INTERL2-NEXT: [[TMP25:%.*]] = extractelement <4 x i1> [[TMP7]], i64 1
; VEC4_INTERL2-NEXT: br i1 [[TMP25]], label [[PRED_STORE_IF12:%.*]], label [[PRED_STORE_CONTINUE13:%.*]]
; VEC4_INTERL2-NEXT: br i1 [[TMP25]], label [[PRED_STORE_IF11:%.*]], label [[PRED_STORE_CONTINUE12:%.*]]
; VEC4_INTERL2: pred.store.if11:
; VEC4_INTERL2-NEXT: [[TMP26:%.*]] = fadd fast float [[TMP0]], 5.000000e+00
; VEC4_INTERL2-NEXT: [[TMP27:%.*]] = or i64 [[INDEX]], 5
; VEC4_INTERL2-NEXT: [[TMP28:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[TMP27]]
; VEC4_INTERL2-NEXT: store float [[TMP26]], float* [[TMP28]], align 4
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE13]]
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE12]]
; VEC4_INTERL2: pred.store.continue12:
; VEC4_INTERL2-NEXT: [[TMP29:%.*]] = extractelement <4 x i1> [[TMP7]], i64 2
; VEC4_INTERL2-NEXT: br i1 [[TMP29]], label [[PRED_STORE_IF14:%.*]], label [[PRED_STORE_CONTINUE15:%.*]]
; VEC4_INTERL2-NEXT: br i1 [[TMP29]], label [[PRED_STORE_IF13:%.*]], label [[PRED_STORE_CONTINUE14:%.*]]
; VEC4_INTERL2: pred.store.if13:
; VEC4_INTERL2-NEXT: [[TMP30:%.*]] = fadd fast float [[TMP0]], 6.000000e+00
; VEC4_INTERL2-NEXT: [[TMP31:%.*]] = or i64 [[INDEX]], 6
; VEC4_INTERL2-NEXT: [[TMP32:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[TMP31]]
; VEC4_INTERL2-NEXT: store float [[TMP30]], float* [[TMP32]], align 4
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE15]]
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE14]]
; VEC4_INTERL2: pred.store.continue14:
; VEC4_INTERL2-NEXT: [[TMP33:%.*]] = extractelement <4 x i1> [[TMP7]], i64 3
; VEC4_INTERL2-NEXT: br i1 [[TMP33]], label [[PRED_STORE_IF16:%.*]], label [[PRED_STORE_CONTINUE17]]
; VEC4_INTERL2-NEXT: br i1 [[TMP33]], label [[PRED_STORE_IF15:%.*]], label [[PRED_STORE_CONTINUE16]]
; VEC4_INTERL2: pred.store.if15:
; VEC4_INTERL2-NEXT: [[TMP34:%.*]] = fadd fast float [[TMP0]], 7.000000e+00
; VEC4_INTERL2-NEXT: [[TMP35:%.*]] = or i64 [[INDEX]], 7
; VEC4_INTERL2-NEXT: [[TMP36:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[TMP35]]
; VEC4_INTERL2-NEXT: store float [[TMP34]], float* [[TMP36]], align 4
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE17]]
; VEC4_INTERL2-NEXT: br label [[PRED_STORE_CONTINUE16]]
; VEC4_INTERL2: pred.store.continue16:
; VEC4_INTERL2-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 8
; VEC4_INTERL2-NEXT: [[TMP37:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
@@ -1558,7 +1558,7 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC4_INTERL2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]]
; VEC4_INTERL2: scalar.ph:
; VEC4_INTERL2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; VEC4_INTERL2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[CAST_CRD]], [[MIDDLE_BLOCK]] ], [ 0.000000e+00, [[ENTRY]] ]
; VEC4_INTERL2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[CAST_VTC]], [[MIDDLE_BLOCK]] ], [ 0.000000e+00, [[ENTRY]] ]
; VEC4_INTERL2-NEXT: br label [[FOR_BODY:%.*]]
; VEC4_INTERL2: for.body:
; VEC4_INTERL2-NEXT: [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_INC:%.*]] ], [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ]
@@ -1585,7 +1585,7 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC1_INTERL2-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; VEC1_INTERL2: vector.ph:
; VEC1_INTERL2-NEXT: [[N_VEC:%.*]] = and i64 [[SMAX]], 9223372036854775806
; VEC1_INTERL2-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC1_INTERL2-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC1_INTERL2: vector.body:
; VEC1_INTERL2-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[PRED_STORE_CONTINUE4:%.*]] ]
@@ -1616,7 +1616,7 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC1_INTERL2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]]
; VEC1_INTERL2: scalar.ph:
; VEC1_INTERL2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; VEC1_INTERL2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[CAST_CRD]], [[MIDDLE_BLOCK]] ], [ 0.000000e+00, [[ENTRY]] ]
; VEC1_INTERL2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi float [ [[CAST_VTC]], [[MIDDLE_BLOCK]] ], [ 0.000000e+00, [[ENTRY]] ]
; VEC1_INTERL2-NEXT: br label [[FOR_BODY:%.*]]
; VEC1_INTERL2: for.body:
; VEC1_INTERL2-NEXT: [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_INC:%.*]] ], [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ]
@@ -1643,7 +1643,7 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC2_INTERL1_PRED_STORE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[FOR_BODY:%.*]], label [[VECTOR_PH:%.*]]
; VEC2_INTERL1_PRED_STORE: vector.ph:
; VEC2_INTERL1_PRED_STORE-NEXT: [[N_VEC:%.*]] = and i64 [[SMAX]], 9223372036854775806
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_CRD:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: [[CAST_VTC:%.*]] = sitofp i64 [[N_VEC]] to float
; VEC2_INTERL1_PRED_STORE-NEXT: br label [[VECTOR_BODY:%.*]]
; VEC2_INTERL1_PRED_STORE: vector.body:
; VEC2_INTERL1_PRED_STORE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[PRED_STORE_CONTINUE3:%.*]] ]
@@ -1676,7 +1676,7 @@ define void @non_primary_iv_float_scalar(float* %A, i64 %N) {
; VEC2_INTERL1_PRED_STORE-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY]]
; VEC2_INTERL1_PRED_STORE: for.body:
; VEC2_INTERL1_PRED_STORE-NEXT: [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_INC:%.*]] ], [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; VEC2_INTERL1_PRED_STORE-NEXT: [[J:%.*]] = phi float [ [[J_NEXT:%.*]], [[FOR_INC]] ], [ [[CAST_CRD]], [[MIDDLE_BLOCK]] ], [ 0.000000e+00, [[ENTRY]] ]
; VEC2_INTERL1_PRED_STORE-NEXT: [[J:%.*]] = phi float [ [[J_NEXT:%.*]], [[FOR_INC]] ], [ [[CAST_VTC]], [[MIDDLE_BLOCK]] ], [ 0.000000e+00, [[ENTRY]] ]
; VEC2_INTERL1_PRED_STORE-NEXT: [[VAR0:%.*]] = getelementptr inbounds float, float* [[A]], i64 [[I]]
; VEC2_INTERL1_PRED_STORE-NEXT: [[VAR1:%.*]] = load float, float* [[VAR0]], align 4
; VEC2_INTERL1_PRED_STORE-NEXT: [[VAR2:%.*]] = fcmp fast oeq float [[VAR1]], 0.000000e+00

6
llvm/test/Transforms/LoopVectorize/if-conversion-nest.ll
View File

@@ -25,7 +25,7 @@ define i32 @foo(i32* nocapture %A, i32* nocapture %B, i32 %n) {
; CHECK-NEXT: [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]]
; CHECK-NEXT: br i1 [[FOUND_CONFLICT]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
@@ -48,7 +48,7 @@ define i32 @foo(i32* nocapture %A, i32* nocapture %B, i32 %n) {
; CHECK-NEXT: store <4 x i32> [[PREDPHI7]], <4 x i32>* [[TMP17]], align 4, !alias.scope !0, !noalias !3
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP5:!llvm.loop !.*]]
; CHECK-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP2]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END_LOOPEXIT:%.*]], label [[SCALAR_PH]]
@@ -76,7 +76,7 @@ define i32 @foo(i32* nocapture %A, i32* nocapture %B, i32 %n) {
; CHECK-NEXT: [[INDVARS_IV_NEXT]] = add i64 [[INDVARS_IV]], 1
; CHECK-NEXT: [[LFTR_WIDEIV:%.*]] = trunc i64 [[INDVARS_IV_NEXT]] to i32
; CHECK-NEXT: [[EXITCOND:%.*]] = icmp eq i32 [[LFTR_WIDEIV]], [[N]]
; CHECK-NEXT: br i1 [[EXITCOND]], label [[FOR_END_LOOPEXIT]], label [[FOR_BODY]], [[LOOP7:!llvm.loop !.*]]
; CHECK-NEXT: br i1 [[EXITCOND]], label [[FOR_END_LOOPEXIT]], label [[FOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]]
; CHECK: for.end.loopexit:
; CHECK-NEXT: br label [[FOR_END]]
; CHECK: for.end:

26
llvm/test/Transforms/LoopVectorize/induction.ll
View File

@@ -63,7 +63,7 @@ define void @multi_int_induction(i32* %A, i32 %N) {
; IND-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; IND-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; IND: vector.ph:
; IND-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; IND-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; IND-NEXT: [[CAST_VTC:%.*]] = trunc i64 [[N_VEC]] to i32
; IND-NEXT: [[IND_END:%.*]] = add i32 [[CAST_VTC]], 190
; IND-NEXT: br label [[VECTOR_BODY:%.*]]
@@ -105,7 +105,7 @@ define void @multi_int_induction(i32* %A, i32 %N) {
; UNROLL-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 3
; UNROLL-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; UNROLL: vector.ph:
; UNROLL-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588
; UNROLL-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; UNROLL-NEXT: [[CAST_VTC:%.*]] = trunc i64 [[N_VEC]] to i32
; UNROLL-NEXT: [[IND_END:%.*]] = add i32 [[CAST_VTC]], 190
; UNROLL-NEXT: br label [[VECTOR_BODY:%.*]]
@@ -202,7 +202,7 @@ define void @multi_int_induction(i32* %A, i32 %N) {
; INTERLEAVE-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 7
; INTERLEAVE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; INTERLEAVE: vector.ph:
; INTERLEAVE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934584
; INTERLEAVE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -8
; INTERLEAVE-NEXT: [[CAST_VTC:%.*]] = trunc i64 [[N_VEC]] to i32
; INTERLEAVE-NEXT: [[IND_END:%.*]] = add i32 [[CAST_VTC]], 190
; INTERLEAVE-NEXT: br label [[VECTOR_BODY:%.*]]
@@ -979,7 +979,7 @@ define float @scalarize_induction_variable_02(float* %a, float* %b, i64 %n) {
; IND-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 8
; IND-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; IND: vector.ph:
; IND-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 4611686018427387902
; IND-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; IND-NEXT: [[IND_END:%.*]] = shl i64 [[N_VEC]], 3
; IND-NEXT: br label [[VECTOR_BODY:%.*]]
; IND: vector.body:
@@ -1039,7 +1039,7 @@ define float @scalarize_induction_variable_02(float* %a, float* %b, i64 %n) {
; UNROLL-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 24
; UNROLL-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; UNROLL: vector.ph:
; UNROLL-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 4611686018427387900
; UNROLL-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; UNROLL-NEXT: [[IND_END:%.*]] = shl i64 [[N_VEC]], 3
; UNROLL-NEXT: br label [[VECTOR_BODY:%.*]]
; UNROLL: vector.body:
@@ -1695,7 +1695,7 @@ define void @scalarize_induction_variable_04(i32* %a, %pair.i32* %p, i32 %n) {
; IND-NEXT: [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]]
; IND-NEXT: br i1 [[FOUND_CONFLICT]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]]
; IND: vector.ph:
; IND-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; IND-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; IND-NEXT: br label [[VECTOR_BODY:%.*]]
; IND: vector.body:
; IND-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
@@ -1757,7 +1757,7 @@ define void @scalarize_induction_variable_04(i32* %a, %pair.i32* %p, i32 %n) {
; UNROLL-NEXT: [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]]
; UNROLL-NEXT: br i1 [[FOUND_CONFLICT]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]]
; UNROLL: vector.ph:
; UNROLL-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588
; UNROLL-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; UNROLL-NEXT: br label [[VECTOR_BODY:%.*]]
; UNROLL: vector.body:
; UNROLL-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
@@ -2582,7 +2582,7 @@ define void @iv_vector_and_scalar_users(%pair.i16* %p, i32 %a, i32 %n) {
; IND-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i32 [[TMP0]], 0
; IND-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; IND: vector.ph:
; IND-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934590
; IND-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; IND-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x i32> poison, i32 [[A:%.*]], i64 0
; IND-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <2 x i32> [[BROADCAST_SPLATINSERT]], <2 x i32> poison, <2 x i32> zeroinitializer
; IND-NEXT: br label [[VECTOR_BODY:%.*]]
@@ -2630,7 +2630,7 @@ define void @iv_vector_and_scalar_users(%pair.i16* %p, i32 %a, i32 %n) {
; UNROLL-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 3
; UNROLL-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; UNROLL: vector.ph:
; UNROLL-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588
; UNROLL-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; UNROLL-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x i32> poison, i32 [[A:%.*]], i64 0
; UNROLL-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <2 x i32> [[BROADCAST_SPLATINSERT]], <2 x i32> poison, <2 x i32> zeroinitializer
; UNROLL-NEXT: [[BROADCAST_SPLATINSERT2:%.*]] = insertelement <2 x i32> poison, i32 [[A]], i64 0
@@ -2754,7 +2754,7 @@ define void @iv_vector_and_scalar_users(%pair.i16* %p, i32 %a, i32 %n) {
; INTERLEAVE-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP0]], 7
; INTERLEAVE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; INTERLEAVE: vector.ph:
; INTERLEAVE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934584
; INTERLEAVE-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -8
; INTERLEAVE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[A:%.*]], i64 0
; INTERLEAVE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
; INTERLEAVE-NEXT: [[BROADCAST_SPLATINSERT2:%.*]] = insertelement <4 x i32> poison, i32 [[A]], i64 0
@@ -3333,7 +3333,7 @@ define i32 @testoverflowcheck() {
; IND-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp eq i8 [[DOTPR_I]], -1
; IND-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; IND: vector.ph:
; IND-NEXT: [[N_VEC:%.*]] = and i32 [[TMP3]], 510
; IND-NEXT: [[N_VEC:%.*]] = and i32 [[TMP3]], -2
; IND-NEXT: [[CAST_VTC:%.*]] = trunc i32 [[N_VEC]] to i8
; IND-NEXT: [[IND_END:%.*]] = add i8 [[DOTPR_I]], [[CAST_VTC]]
; IND-NEXT: [[TMP4:%.*]] = insertelement <2 x i32> <i32 poison, i32 -1>, i32 [[C_PROMOTED_I]], i64 0
@@ -3375,7 +3375,7 @@ define i32 @testoverflowcheck() {
; UNROLL-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ugt i8 [[DOTPR_I]], -4
; UNROLL-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; UNROLL: vector.ph:
; UNROLL-NEXT: [[N_VEC:%.*]] = and i32 [[TMP3]], 508
; UNROLL-NEXT: [[N_VEC:%.*]] = and i32 [[TMP3]], -4
; UNROLL-NEXT: [[CAST_VTC:%.*]] = trunc i32 [[N_VEC]] to i8
; UNROLL-NEXT: [[IND_END:%.*]] = add i8 [[DOTPR_I]], [[CAST_VTC]]
; UNROLL-NEXT: [[TMP4:%.*]] = insertelement <2 x i32> <i32 poison, i32 -1>, i32 [[C_PROMOTED_I]], i64 0
@@ -3469,7 +3469,7 @@ define i32 @testoverflowcheck() {
; INTERLEAVE-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ugt i8 [[DOTPR_I]], -8
; INTERLEAVE-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; INTERLEAVE: vector.ph:
; INTERLEAVE-NEXT: [[N_VEC:%.*]] = and i32 [[TMP3]], 504
; INTERLEAVE-NEXT: [[N_VEC:%.*]] = and i32 [[TMP3]], -8
; INTERLEAVE-NEXT: [[CAST_VTC:%.*]] = trunc i32 [[N_VEC]] to i8
; INTERLEAVE-NEXT: [[IND_END:%.*]] = add i8 [[DOTPR_I]], [[CAST_VTC]]
; INTERLEAVE-NEXT: [[TMP4:%.*]] = insertelement <4 x i32> <i32 poison, i32 -1, i32 -1, i32 -1>, i32 [[C_PROMOTED_I]], i64 0

4
llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll
View File

@@ -1306,7 +1306,7 @@ define void @PR27626_4(i32 *%a, i32 %x, i32 %y, i32 %z, i64 %n) {
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 6
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 9223372036854775804
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; CHECK-NEXT: [[IND_END:%.*]] = shl nuw i64 [[N_VEC]], 1
; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[Y:%.*]], i64 0
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
@@ -1396,7 +1396,7 @@ define void @PR27626_5(i32 *%a, i32 %x, i32 %y, i32 %z, i64 %n) {
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 6
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 9223372036854775804
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4
; CHECK-NEXT: [[TMP3:%.*]] = shl nuw i64 [[N_VEC]], 1
; CHECK-NEXT: [[IND_END:%.*]] = or i64 [[TMP3]], 3
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]

4
llvm/test/Transforms/LoopVectorize/loop-scalars.ll
View File

@@ -70,7 +70,7 @@ define void @scalar_store(i32** %a, i32 *%b, i64 %n) {
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 2
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 9223372036854775806
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; CHECK-NEXT: [[IND_END:%.*]] = shl nuw i64 [[N_VEC]], 1
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
@@ -130,7 +130,7 @@ define void @expansion(i32** %a, i64 *%b, i64 %n) {
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 2
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 9223372036854775806
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -2
; CHECK-NEXT: [[IND_END:%.*]] = shl nuw i64 [[N_VEC]], 1
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:

2
llvm/test/Transforms/LoopVectorize/runtime-check.ll
View File

@@ -28,7 +28,7 @@ define i32 @foo(float* nocapture %a, float* nocapture %b, i32 %n) nounwind uwtab
; CHECK-NEXT: [[DIFF_CHECK:%.*]] = icmp ult i64 [[TMP3]], 16, [[DBG9]]
; CHECK-NEXT: br i1 [[DIFF_CHECK]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]], [[DBG9]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], 8589934588, [[DBG9]]
; CHECK-NEXT: [[N_VEC:%.*]] = and i64 [[TMP2]], -4, [[DBG9]]
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]], [[DBG9]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ], [[DBG9]]

2
llvm/test/Transforms/PhaseOrdering/X86/excessive-unrolling.ll
View File

@@ -183,7 +183,7 @@ define void @test_runtime_trip_count(i32 %N) {
; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i64 [[TMP0]], 28
; CHECK-NEXT: br i1 [[TMP3]], label [[MIDDLE_BLOCK_UNR_LCSSA:%.*]], label [[VECTOR_PH_NEW:%.*]]
; CHECK: vector.ph.new:
; CHECK-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP2]], 9223372036854775800
; CHECK-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP2]], -8
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH_NEW]] ], [ [[INDEX_NEXT_7:%.*]], [[VECTOR_BODY]] ]

2
llvm/test/Transforms/PhaseOrdering/X86/vdiv.ll
View File

@@ -42,7 +42,7 @@ define void @vdiv(double* %x, double* %y, double %a, i32 %N) #0 {
; CHECK-NEXT: [[TMP4:%.*]] = icmp ult i64 [[TMP1]], 16
; CHECK-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK_UNR_LCSSA:%.*]], label [[VECTOR_PH_NEW:%.*]]
; CHECK: vector.ph.new:
; CHECK-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP3]], 2305843009213693950
; CHECK-NEXT: [[UNROLL_ITER:%.*]] = and i64 [[TMP3]], -2
; CHECK-NEXT: [[TMP5:%.*]] = fdiv fast <4 x double> <double 1.000000e+00, double 1.000000e+00, double 1.000000e+00, double 1.000000e+00>, [[BROADCAST_SPLAT]]
; CHECK-NEXT: [[TMP6:%.*]] = fdiv fast <4 x double> <double 1.000000e+00, double 1.000000e+00, double 1.000000e+00, double 1.000000e+00>, [[BROADCAST_SPLAT10]]
; CHECK-NEXT: [[TMP7:%.*]] = fdiv fast <4 x double> <double 1.000000e+00, double 1.000000e+00, double 1.000000e+00, double 1.000000e+00>, [[BROADCAST_SPLAT12]]