diff options
| author | Florian Hahn <flo@fhahn.com> | 2023-08-29 11:42:17 +0100 |
|---|---|---|
| committer | Florian Hahn <flo@fhahn.com> | 2023-08-29 11:42:18 +0100 |
| commit | 96e83d3705c76e7da28551a698cb651d352a9b2b (patch) | |
| tree | e3fcb4e83e8028817844ce8baecd21c162da622d | |
| parent | 3a1409f93da32bf626f76257e0aac71716f2f67e (diff) | |
| download | llvm-96e83d3705c76e7da28551a698cb651d352a9b2b.zip llvm-96e83d3705c76e7da28551a698cb651d352a9b2b.tar.gz llvm-96e83d3705c76e7da28551a698cb651d352a9b2b.tar.bz2 | |
[LV] Use IRBuilder to create and optimize middle-block compare.
Split off from D150398 to avoid builder-related diff changes there.
Using IRBuilder to create ICmps simplifies the result if both operands
are constants.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D158332
68 files changed, 372 insertions, 1047 deletions
diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp index 9184f97..c11ac8e 100644 --- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp +++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp @@ -3069,15 +3069,16 @@ BasicBlock *InnerLoopVectorizer::completeLoopSkeleton() { // 3) Otherwise, construct a runtime check. if (!Cost->requiresScalarEpilogue(VF.isVector()) && !Cost->foldTailByMasking()) { - Instruction *CmpN = CmpInst::Create(Instruction::ICmp, CmpInst::ICMP_EQ, - Count, VectorTripCount, "cmp.n", - LoopMiddleBlock->getTerminator()); - // Here we use the same DebugLoc as the scalar loop latch terminator instead // of the corresponding compare because they may have ended up with // different line numbers and we want to avoid awkward line stepping while // debugging. Eg. if the compare has got a line number inside the loop. - CmpN->setDebugLoc(ScalarLatchTerm->getDebugLoc()); + // TODO: At the moment, CreateICmpEQ will simplify conditions with constant + // operands. Perform simplification directly on VPlan once the branch is + // modeled there. + IRBuilder<> B(LoopMiddleBlock->getTerminator()); + B.SetCurrentDebugLocation(ScalarLatchTerm->getDebugLoc()); + Value *CmpN = B.CreateICmpEQ(Count, VectorTripCount, "cmp.n"); cast<BranchInst>(LoopMiddleBlock->getTerminator())->setCondition(CmpN); } diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/epilog-vectorization-widen-inductions.ll b/llvm/test/Transforms/LoopVectorize/AArch64/epilog-vectorization-widen-inductions.ll index 1215962..24d6d2d 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/epilog-vectorization-widen-inductions.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/epilog-vectorization-widen-inductions.ll @@ -45,8 +45,7 @@ define void @test_widen_ptr_induction(ptr %ptr.start.1) { ; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 10000 ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], {{!llvm.loop ![0-9]+}} ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 10001, 10000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK: vec.epilog.iter.check: ; CHECK-NEXT: [[IND_END6:%.*]] = getelementptr i8, ptr [[PTR_START_1]], i64 10000 ; CHECK-NEXT: br i1 true, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] @@ -74,8 +73,7 @@ define void @test_widen_ptr_induction(ptr %ptr.start.1) { ; CHECK-NEXT: [[TMP25:%.*]] = icmp eq i64 [[INDEX_NEXT12]], 10000 ; CHECK-NEXT: br i1 [[TMP25]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], {{!llvm.loop ![0-9]+}} ; CHECK: vec.epilog.middle.block: -; CHECK-NEXT: [[CMP_N8:%.*]] = icmp eq i64 10001, 10000 -; CHECK-NEXT: br i1 [[CMP_N8]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL4:%.*]] = phi i64 [ 10000, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 10000, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-NEXT: [[BC_RESUME_VAL7:%.*]] = phi ptr [ [[IND_END5]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[IND_END6]], [[VEC_EPILOG_ITER_CHECK]] ], [ [[PTR_START_1]], [[ITER_CHECK]] ] @@ -420,8 +418,7 @@ define void @test_widen_extended_induction(ptr %dst) { ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i32 [[INDEX_NEXT]], 10000 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], {{!llvm.loop ![0-9]+}} ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 10000, 10000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK: vec.epilog.iter.check: ; CHECK-NEXT: br i1 true, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; CHECK: vec.epilog.ph: @@ -445,8 +442,7 @@ define void @test_widen_extended_induction(ptr %dst) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i32 [[INDEX_NEXT9]], 10000 ; CHECK-NEXT: br i1 [[TMP13]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], {{!llvm.loop ![0-9]+}} ; CHECK: vec.epilog.middle.block: -; CHECK-NEXT: [[CMP_N3:%.*]] = icmp eq i32 10000, 10000 -; CHECK-NEXT: br i1 [[CMP_N3]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL2:%.*]] = phi i8 [ 16, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 16, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_SCEVCHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -504,8 +500,7 @@ define void @test_widen_truncated_induction(ptr %A) { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 10000 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], {{!llvm.loop ![0-9]+}} ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 10000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK: vec.epilog.iter.check: ; CHECK-NEXT: br i1 true, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; CHECK: vec.epilog.ph: @@ -528,8 +523,7 @@ define void @test_widen_truncated_induction(ptr %A) { ; CHECK-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT8]], 10000 ; CHECK-NEXT: br i1 [[TMP11]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], {{!llvm.loop ![0-9]+}} ; CHECK: vec.epilog.middle.block: -; CHECK-NEXT: [[CMP_N3:%.*]] = icmp eq i64 10000, 10000 -; CHECK-NEXT: br i1 [[CMP_N3]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL2:%.*]] = phi i64 [ 10000, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 10000, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/pr60831-sve-inv-store-crash.ll b/llvm/test/Transforms/LoopVectorize/AArch64/pr60831-sve-inv-store-crash.ll index 9526dd1..3f6d1b7 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/pr60831-sve-inv-store-crash.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/pr60831-sve-inv-store-crash.ll @@ -139,8 +139,7 @@ define void @test_loop2(i64 %n, ptr %dst) { ; CHECK-NEXT: [[TMP52:%.*]] = icmp eq i64 [[INDEX_NEXT]], 992 ; CHECK-NEXT: br i1 [[TMP52]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1001, 992 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK: vec.epilog.iter.check: ; CHECK-NEXT: br i1 false, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; CHECK: vec.epilog.ph: @@ -181,8 +180,7 @@ define void @test_loop2(i64 %n, ptr %dst) { ; CHECK-NEXT: [[TMP81:%.*]] = icmp eq i64 [[INDEX_NEXT3]], 1000 ; CHECK-NEXT: br i1 [[TMP81]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: vec.epilog.middle.block: -; CHECK-NEXT: [[CMP_N1:%.*]] = icmp eq i64 1001, 1000 -; CHECK-NEXT: br i1 [[CMP_N1]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 992, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/sve-epilog-vect.ll b/llvm/test/Transforms/LoopVectorize/AArch64/sve-epilog-vect.ll index 3df9f31..6982b3f 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/sve-epilog-vect.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/sve-epilog-vect.ll @@ -155,8 +155,7 @@ define void @main_vf_vscale_x_16(ptr %A) #0 { ; CHECK-VF8-NEXT: [[TMP22:%.*]] = icmp eq i64 [[INDEX_NEXT3]], 1024 ; CHECK-VF8-NEXT: br i1 [[TMP22]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] ; CHECK-VF8: vec.epilog.middle.block: -; CHECK-VF8-NEXT: [[CMP_N1:%.*]] = icmp eq i64 1024, 1024 -; CHECK-VF8-NEXT: br i1 [[CMP_N1]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-VF8-NEXT: br i1 true, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK-VF8: vec.epilog.scalar.ph: ; CHECK-VF8-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-VF8-NEXT: br label [[FOR_BODY:%.*]] @@ -256,8 +255,7 @@ define void @main_vf_vscale_x_2(ptr %A) #0 vscale_range(8, 8) { ; CHECK-NEXT: [[TMP22:%.*]] = icmp eq i64 [[INDEX_NEXT3]], 1024 ; CHECK-NEXT: br i1 [[TMP22]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: vec.epilog.middle.block: -; CHECK-NEXT: [[CMP_N1:%.*]] = icmp eq i64 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N1]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -326,8 +324,7 @@ define void @main_vf_vscale_x_2(ptr %A) #0 vscale_range(8, 8) { ; CHECK-VF8-NEXT: [[TMP22:%.*]] = icmp eq i64 [[INDEX_NEXT3]], 1024 ; CHECK-VF8-NEXT: br i1 [[TMP22]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK-VF8: vec.epilog.middle.block: -; CHECK-VF8-NEXT: [[CMP_N1:%.*]] = icmp eq i64 1024, 1024 -; CHECK-VF8-NEXT: br i1 [[CMP_N1]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-VF8-NEXT: br i1 true, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK-VF8: vec.epilog.scalar.ph: ; CHECK-VF8-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-VF8-NEXT: br label [[FOR_BODY:%.*]] @@ -503,8 +500,7 @@ define void @test_pr57912_pointer_induction(ptr %start) #0 { ; CHECK-VF8-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT9]], 10000 ; CHECK-VF8-NEXT: br i1 [[TMP18]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP9:![0-9]+]] ; CHECK-VF8: vec.epilog.middle.block: -; CHECK-VF8-NEXT: [[CMP_N6:%.*]] = icmp eq i64 10000, 10000 -; CHECK-VF8-NEXT: br i1 [[CMP_N6]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-VF8-NEXT: br i1 true, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK-VF8: vec.epilog.scalar.ph: ; CHECK-VF8-NEXT: [[BC_RESUME_VAL2:%.*]] = phi i64 [ 10000, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-VF8-NEXT: [[BC_RESUME_VAL5:%.*]] = phi ptr [ [[IND_END3]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[IND_END4]], [[VEC_EPILOG_ITER_CHECK]] ], [ [[START]], [[ITER_CHECK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/synthesize-mask-for-call.ll b/llvm/test/Transforms/LoopVectorize/AArch64/synthesize-mask-for-call.ll index 849b934..cbded74 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/synthesize-mask-for-call.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/synthesize-mask-for-call.ll @@ -194,8 +194,7 @@ define void @test_v4_v4m(ptr noalias %a, ptr readonly %b) #3 { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -251,8 +250,7 @@ define void @test_v2_v4m(ptr noalias %a, ptr readonly %b) #3 { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -308,8 +306,7 @@ define void @test_v2_v4(ptr noalias %a, ptr readonly %b) #3 { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/type-shrinkage-insertelt.ll b/llvm/test/Transforms/LoopVectorize/AArch64/type-shrinkage-insertelt.ll index 7d84889..56a39d3 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/type-shrinkage-insertelt.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/type-shrinkage-insertelt.ll @@ -51,8 +51,7 @@ define void @test0(ptr noalias %M3, ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP29:%.*]] = icmp eq i64 [[INDEX_NEXT]], 16 ; CHECK-NEXT: br i1 [[TMP29]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 16, 16 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_INC1286_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_INC1286_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[IF_THEN1165_US:%.*]] @@ -148,8 +147,7 @@ define void @test1(ptr noalias %M3, ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP31:%.*]] = icmp eq i64 [[INDEX_NEXT]], 16 ; CHECK-NEXT: br i1 [[TMP31]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 16, 16 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_INC1286_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_INC1286_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[IF_THEN1165_US:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/ARM/tail-folding-not-allowed.ll b/llvm/test/Transforms/LoopVectorize/ARM/tail-folding-not-allowed.ll index 7b9820d..f7b656d2 100644 --- a/llvm/test/Transforms/LoopVectorize/ARM/tail-folding-not-allowed.ll +++ b/llvm/test/Transforms/LoopVectorize/ARM/tail-folding-not-allowed.ll @@ -33,8 +33,7 @@ define void @trunc_not_allowed_different_vec_elemns(ptr noalias nocapture %A, pt ; CHECK-NEXT: [[TMP12:%.*]] = icmp eq i32 [[INDEX_NEXT]], 428 ; CHECK-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 431, 428 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 428, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -144,8 +143,7 @@ define void @narrowing_load_not_allowed(ptr noalias nocapture %A, ptr noalias no ; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i32 [[INDEX_NEXT]], 424 ; CHECK-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 431, 424 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 424, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -214,8 +212,7 @@ define void @trunc_not_allowed(ptr noalias nocapture %A, ptr noalias nocapture r ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i32 [[INDEX_NEXT]], 428 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 431, 428 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 428, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -287,8 +284,7 @@ define void @strides_different_direction(ptr noalias nocapture %A, ptr noalias n ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i32 [[INDEX_NEXT]], 428 ; CHECK-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 431, 428 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 428, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -352,8 +348,7 @@ define void @too_many_loop_blocks(ptr noalias nocapture %A, ptr noalias nocaptur ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i32 [[INDEX_NEXT]], 428 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 431, 428 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 428, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -464,8 +459,7 @@ define void @fptrunc_not_allowed(ptr noalias nocapture %A, ptr noalias nocapture ; CHECK-NEXT: [[TMP12:%.*]] = icmp eq i32 [[INDEX_NEXT]], 428 ; CHECK-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 431, 428 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 428, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/divrem.ll b/llvm/test/Transforms/LoopVectorize/RISCV/divrem.ll index 8bff72e..a6ab8d1 100644 --- a/llvm/test/Transforms/LoopVectorize/RISCV/divrem.ll +++ b/llvm/test/Transforms/LoopVectorize/RISCV/divrem.ll @@ -78,8 +78,7 @@ define void @vector_udiv(ptr noalias nocapture %a, i64 %v, i64 %n) { ; FIXED-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] @@ -183,8 +182,7 @@ define void @vector_sdiv(ptr noalias nocapture %a, i64 %v, i64 %n) { ; FIXED-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] @@ -288,8 +286,7 @@ define void @vector_urem(ptr noalias nocapture %a, i64 %v, i64 %n) { ; FIXED-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] @@ -393,8 +390,7 @@ define void @vector_srem(ptr noalias nocapture %a, i64 %v, i64 %n) { ; FIXED-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] @@ -516,8 +512,7 @@ define void @predicated_udiv(ptr noalias nocapture %a, i64 %v, i64 %n) { ; FIXED-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] @@ -651,8 +646,7 @@ define void @predicated_sdiv(ptr noalias nocapture %a, i64 %v, i64 %n) { ; FIXED-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] @@ -779,8 +773,7 @@ define void @predicated_udiv_by_constant(ptr noalias nocapture %a, i64 %n) { ; FIXED-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] @@ -907,8 +900,7 @@ define void @predicated_sdiv_by_constant(ptr noalias nocapture %a, i64 %n) { ; FIXED-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] @@ -1038,8 +1030,7 @@ define void @predicated_sdiv_by_minus_one(ptr noalias nocapture %a, i64 %n) { ; FIXED-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXED-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP18:![0-9]+]] ; FIXED: middle.block: -; FIXED-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; FIXED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXED-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXED: scalar.ph: ; FIXED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXED-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-accesses.ll b/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-accesses.ll index 8f6e8d0..d6196c8 100644 --- a/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-accesses.ll +++ b/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-accesses.ll @@ -28,8 +28,7 @@ define void @load_store_factor2_i32(ptr %p) { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -153,8 +152,7 @@ define void @load_store_factor3_i32(ptr %p) { ; CHECK-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -414,8 +412,7 @@ define void @combine_load_factor2_i32(ptr noalias %p, ptr noalias %q) { ; CHECK-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -484,8 +481,7 @@ define void @combine_load_factor2_i64(ptr noalias %p, ptr noalias %q) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/safe-dep-distance.ll b/llvm/test/Transforms/LoopVectorize/RISCV/safe-dep-distance.ll index a266ae6..90c64e3 100644 --- a/llvm/test/Transforms/LoopVectorize/RISCV/safe-dep-distance.ll +++ b/llvm/test/Transforms/LoopVectorize/RISCV/safe-dep-distance.ll @@ -94,8 +94,7 @@ define void @test_may_clobber(ptr %p) { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 200 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 200, 200 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 200, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/short-trip-count.ll b/llvm/test/Transforms/LoopVectorize/RISCV/short-trip-count.ll index 4c99477..31d0b2a 100644 --- a/llvm/test/Transforms/LoopVectorize/RISCV/short-trip-count.ll +++ b/llvm/test/Transforms/LoopVectorize/RISCV/short-trip-count.ll @@ -18,8 +18,7 @@ define void @small_trip_count_min_vlen_128(ptr nocapture %a) nounwind vscale_ran ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4 ; CHECK-NEXT: br i1 true, label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 4, 4 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 4, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -69,8 +68,7 @@ define void @small_trip_count_min_vlen_32(ptr nocapture %a) nounwind vscale_rang ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4 ; CHECK-NEXT: br i1 true, label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 4, 4 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 4, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/uniform-load-store.ll b/llvm/test/Transforms/LoopVectorize/RISCV/uniform-load-store.ll index 5c31dd3..89c3219 100644 --- a/llvm/test/Transforms/LoopVectorize/RISCV/uniform-load-store.ll +++ b/llvm/test/Transforms/LoopVectorize/RISCV/uniform-load-store.ll @@ -75,8 +75,7 @@ define void @uniform_load(ptr noalias nocapture %a, ptr noalias nocapture %b, i6 ; FIXEDLEN-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXEDLEN-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; FIXEDLEN: middle.block: -; FIXEDLEN-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, 1024 -; FIXEDLEN-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXEDLEN-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXEDLEN: scalar.ph: ; FIXEDLEN-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] @@ -252,8 +251,7 @@ define i64 @uniform_load_outside_use(ptr noalias nocapture %a, ptr noalias nocap ; FIXEDLEN-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXEDLEN-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; FIXEDLEN: middle.block: -; FIXEDLEN-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, 1024 -; FIXEDLEN-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXEDLEN-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXEDLEN: scalar.ph: ; FIXEDLEN-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] @@ -412,8 +410,7 @@ define void @conditional_uniform_load(ptr noalias nocapture %a, ptr noalias noca ; FIXEDLEN-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXEDLEN-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; FIXEDLEN: middle.block: -; FIXEDLEN-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, 1024 -; FIXEDLEN-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXEDLEN-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXEDLEN: scalar.ph: ; FIXEDLEN-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] @@ -638,8 +635,7 @@ define void @uniform_load_unaligned(ptr noalias nocapture %a, ptr noalias nocapt ; FIXEDLEN-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXEDLEN-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; FIXEDLEN: middle.block: -; FIXEDLEN-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, 1024 -; FIXEDLEN-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXEDLEN-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXEDLEN: scalar.ph: ; FIXEDLEN-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] @@ -814,8 +810,7 @@ define void @uniform_store(ptr noalias nocapture %a, ptr noalias nocapture %b, i ; FIXEDLEN-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXEDLEN-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; FIXEDLEN: middle.block: -; FIXEDLEN-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, 1024 -; FIXEDLEN-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXEDLEN-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXEDLEN: scalar.ph: ; FIXEDLEN-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] @@ -1007,8 +1002,7 @@ define void @uniform_store_of_loop_varying(ptr noalias nocapture %a, ptr noalias ; FIXEDLEN-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXEDLEN-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; FIXEDLEN: middle.block: -; FIXEDLEN-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, 1024 -; FIXEDLEN-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXEDLEN-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXEDLEN: scalar.ph: ; FIXEDLEN-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] @@ -1249,8 +1243,7 @@ define void @conditional_uniform_store(ptr noalias nocapture %a, ptr noalias noc ; FIXEDLEN-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXEDLEN-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; FIXEDLEN: middle.block: -; FIXEDLEN-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, 1024 -; FIXEDLEN-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXEDLEN-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXEDLEN: scalar.ph: ; FIXEDLEN-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] @@ -1473,8 +1466,7 @@ define void @uniform_store_unaligned(ptr noalias nocapture %a, ptr noalias nocap ; FIXEDLEN-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; FIXEDLEN-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; FIXEDLEN: middle.block: -; FIXEDLEN-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, 1024 -; FIXEDLEN-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; FIXEDLEN-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; FIXEDLEN: scalar.ph: ; FIXEDLEN-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1024, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/constant-fold.ll b/llvm/test/Transforms/LoopVectorize/X86/constant-fold.ll index 648ba0e..61cae9c 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/constant-fold.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/constant-fold.ll @@ -27,8 +27,7 @@ define void @f1() { ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 2 ; CHECK-NEXT: br i1 true, label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 2, 2 -; CHECK-NEXT: br i1 [[CMP_N]], label [[BB3:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[BB3:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 2, [[MIDDLE_BLOCK]] ], [ 0, [[BB1:%.*]] ] ; CHECK-NEXT: br label [[BB2:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/gather_scatter.ll b/llvm/test/Transforms/LoopVectorize/X86/gather_scatter.ll index 43634bb..e9541c1 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/gather_scatter.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/gather_scatter.ll @@ -44,30 +44,7 @@ define void @foo1(ptr noalias %in, ptr noalias %out, ptr noalias %trigger, ptr n ; AVX512-NEXT: call void @llvm.masked.store.v16f32.p0(<16 x float> [[TMP8]], ptr [[TMP10]], i32 4, <16 x i1> [[TMP3]]) ; AVX512-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 16 ; AVX512-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 -; AVX512-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] -; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; AVX512: for.body: -; AVX512-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; AVX512-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; AVX512-NEXT: [[TMP12:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; AVX512-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP12]], 0 -; AVX512-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; AVX512: if.then: -; AVX512-NEXT: [[ARRAYIDX3:%.*]] = getelementptr inbounds i32, ptr [[INDEX]], i64 [[INDVARS_IV]] -; AVX512-NEXT: [[TMP13:%.*]] = load i32, ptr [[ARRAYIDX3]], align 4 -; AVX512-NEXT: [[IDXPROM4:%.*]] = sext i32 [[TMP13]] to i64 -; AVX512-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[IN]], i64 [[IDXPROM4]] -; AVX512-NEXT: [[TMP14:%.*]] = load float, ptr [[ARRAYIDX5]], align 4 -; AVX512-NEXT: [[ADD:%.*]] = fadd float [[TMP14]], 5.000000e-01 -; AVX512-NEXT: [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[OUT]], i64 [[INDVARS_IV]] -; AVX512-NEXT: store float [[ADD]], ptr [[ARRAYIDX7]], align 4 -; AVX512-NEXT: br label [[FOR_INC]] -; AVX512: for.inc: -; AVX512-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1 -; AVX512-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT]], 4096 -; AVX512-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] +; AVX512-NEXT: br i1 [[TMP11]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; AVX512: for.end: ; AVX512-NEXT: ret void ; @@ -93,30 +70,7 @@ define void @foo1(ptr noalias %in, ptr noalias %out, ptr noalias %trigger, ptr n ; FVW2-NEXT: call void @llvm.masked.store.v2f32.p0(<2 x float> [[TMP8]], ptr [[TMP10]], i32 4, <2 x i1> [[TMP3]]) ; FVW2-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 2 ; FVW2-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 -; FVW2-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] -; FVW2: middle.block: -; FVW2-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; FVW2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; FVW2: for.body: -; FVW2-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; FVW2-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; FVW2-NEXT: [[TMP12:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; FVW2-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP12]], 0 -; FVW2-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; FVW2: if.then: -; FVW2-NEXT: [[ARRAYIDX3:%.*]] = getelementptr inbounds i32, ptr [[INDEX]], i64 [[INDVARS_IV]] -; FVW2-NEXT: [[TMP13:%.*]] = load i32, ptr [[ARRAYIDX3]], align 4 -; FVW2-NEXT: [[IDXPROM4:%.*]] = sext i32 [[TMP13]] to i64 -; FVW2-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[IN]], i64 [[IDXPROM4]] -; FVW2-NEXT: [[TMP14:%.*]] = load float, ptr [[ARRAYIDX5]], align 4 -; FVW2-NEXT: [[ADD:%.*]] = fadd float [[TMP14]], 5.000000e-01 -; FVW2-NEXT: [[ARRAYIDX7:%.*]] = getelementptr inbounds float, ptr [[OUT]], i64 [[INDVARS_IV]] -; FVW2-NEXT: store float [[ADD]], ptr [[ARRAYIDX7]], align 4 -; FVW2-NEXT: br label [[FOR_INC]] -; FVW2: for.inc: -; FVW2-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1 -; FVW2-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT]], 4096 -; FVW2-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] +; FVW2-NEXT: br i1 [[TMP11]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; FVW2: for.end: ; FVW2-NEXT: ret void ; @@ -180,27 +134,7 @@ define void @foo2(ptr noalias %in, ptr noalias %out, ptr noalias %trigger, ptr n ; AVX512-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 16 ; AVX512-NEXT: [[VEC_IND_NEXT]] = add <16 x i64> [[VEC_IND]], <i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256> ; AVX512-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; AVX512-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] -; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; AVX512: for.body: -; AVX512-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; AVX512-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; AVX512-NEXT: [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; AVX512-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP6]], 0 -; AVX512-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; AVX512: if.then: -; AVX512-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr [[IN]], i64 [[INDVARS_IV]], i32 1 -; AVX512-NEXT: [[TMP7:%.*]] = load float, ptr [[B]], align 4 -; AVX512-NEXT: [[ADD:%.*]] = fadd float [[TMP7]], 5.000000e-01 -; AVX512-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[OUT]], i64 [[INDVARS_IV]] -; AVX512-NEXT: store float [[ADD]], ptr [[ARRAYIDX5]], align 4 -; AVX512-NEXT: br label [[FOR_INC]] -; AVX512: for.inc: -; AVX512-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; AVX512-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; AVX512-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP5:![0-9]+]] +; AVX512-NEXT: br i1 [[TMP5]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] ; AVX512: for.end: ; AVX512-NEXT: ret void ; @@ -242,27 +176,7 @@ define void @foo2(ptr noalias %in, ptr noalias %out, ptr noalias %trigger, ptr n ; FVW2-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 2 ; FVW2-NEXT: [[VEC_IND_NEXT]] = add <2 x i64> [[VEC_IND]], <i64 32, i64 32> ; FVW2-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; FVW2-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] -; FVW2: middle.block: -; FVW2-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; FVW2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; FVW2: for.body: -; FVW2-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; FVW2-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; FVW2-NEXT: [[TMP18:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; FVW2-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP18]], 0 -; FVW2-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; FVW2: if.then: -; FVW2-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr [[IN]], i64 [[INDVARS_IV]], i32 1 -; FVW2-NEXT: [[TMP19:%.*]] = load float, ptr [[B]], align 4 -; FVW2-NEXT: [[ADD:%.*]] = fadd float [[TMP19]], 5.000000e-01 -; FVW2-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[OUT]], i64 [[INDVARS_IV]] -; FVW2-NEXT: store float [[ADD]], ptr [[ARRAYIDX5]], align 4 -; FVW2-NEXT: br label [[FOR_INC]] -; FVW2: for.inc: -; FVW2-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; FVW2-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; FVW2-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP5:![0-9]+]] +; FVW2-NEXT: br i1 [[TMP17]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] ; FVW2: for.end: ; FVW2-NEXT: ret void ; @@ -327,27 +241,7 @@ define void @foo3(ptr noalias %in, ptr noalias %out, ptr noalias %trigger) { ; AVX512-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 16 ; AVX512-NEXT: [[VEC_IND_NEXT]] = add <16 x i64> [[VEC_IND]], <i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256> ; AVX512-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; AVX512-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] -; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; AVX512: for.body: -; AVX512-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; AVX512-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; AVX512-NEXT: [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; AVX512-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP6]], 0 -; AVX512-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; AVX512: if.then: -; AVX512-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr [[IN]], i64 [[INDVARS_IV]], i32 1 -; AVX512-NEXT: [[TMP7:%.*]] = load float, ptr [[B]], align 4 -; AVX512-NEXT: [[ADD:%.*]] = fadd float [[TMP7]], 5.000000e-01 -; AVX512-NEXT: [[B6:%.*]] = getelementptr inbounds [[STRUCT_OUT]], ptr [[OUT]], i64 [[INDVARS_IV]], i32 1 -; AVX512-NEXT: store float [[ADD]], ptr [[B6]], align 4 -; AVX512-NEXT: br label [[FOR_INC]] -; AVX512: for.inc: -; AVX512-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; AVX512-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; AVX512-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP7:![0-9]+]] +; AVX512-NEXT: br i1 [[TMP5]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; AVX512: for.end: ; AVX512-NEXT: ret void ; @@ -389,27 +283,7 @@ define void @foo3(ptr noalias %in, ptr noalias %out, ptr noalias %trigger) { ; FVW2-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; FVW2-NEXT: [[VEC_IND_NEXT]] = add <2 x i64> [[VEC_IND]], <i64 32, i64 32> ; FVW2-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; FVW2-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] -; FVW2: middle.block: -; FVW2-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; FVW2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; FVW2: for.body: -; FVW2-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; FVW2-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; FVW2-NEXT: [[TMP18:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; FVW2-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP18]], 0 -; FVW2-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; FVW2: if.then: -; FVW2-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr [[IN]], i64 [[INDVARS_IV]], i32 1 -; FVW2-NEXT: [[TMP19:%.*]] = load float, ptr [[B]], align 4 -; FVW2-NEXT: [[ADD:%.*]] = fadd float [[TMP19]], 5.000000e-01 -; FVW2-NEXT: [[B6:%.*]] = getelementptr inbounds [[STRUCT_OUT]], ptr [[OUT]], i64 [[INDVARS_IV]], i32 1 -; FVW2-NEXT: store float [[ADD]], ptr [[B6]], align 4 -; FVW2-NEXT: br label [[FOR_INC]] -; FVW2: for.inc: -; FVW2-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; FVW2-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; FVW2-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP7:![0-9]+]] +; FVW2-NEXT: br i1 [[TMP17]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; FVW2: for.end: ; FVW2-NEXT: ret void ; @@ -461,27 +335,7 @@ define void @foo2_addrspace(ptr addrspace(1) noalias %in, ptr addrspace(1) noali ; AVX512-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 16 ; AVX512-NEXT: [[VEC_IND_NEXT]] = add <16 x i64> [[VEC_IND]], <i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256> ; AVX512-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; AVX512-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] -; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; AVX512: for.body: -; AVX512-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; AVX512-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; AVX512-NEXT: [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; AVX512-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP6]], 0 -; AVX512-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; AVX512: if.then: -; AVX512-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr addrspace(1) [[IN]], i64 [[INDVARS_IV]], i32 1 -; AVX512-NEXT: [[TMP7:%.*]] = load float, ptr addrspace(1) [[B]], align 4 -; AVX512-NEXT: [[ADD:%.*]] = fadd float [[TMP7]], 5.000000e-01 -; AVX512-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr addrspace(1) [[OUT]], i64 [[INDVARS_IV]] -; AVX512-NEXT: store float [[ADD]], ptr addrspace(1) [[ARRAYIDX5]], align 4 -; AVX512-NEXT: br label [[FOR_INC]] -; AVX512: for.inc: -; AVX512-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; AVX512-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; AVX512-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP9:![0-9]+]] +; AVX512-NEXT: br i1 [[TMP5]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; AVX512: for.end: ; AVX512-NEXT: ret void ; @@ -523,27 +377,7 @@ define void @foo2_addrspace(ptr addrspace(1) noalias %in, ptr addrspace(1) noali ; FVW2-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 2 ; FVW2-NEXT: [[VEC_IND_NEXT]] = add <2 x i64> [[VEC_IND]], <i64 32, i64 32> ; FVW2-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; FVW2-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] -; FVW2: middle.block: -; FVW2-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; FVW2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; FVW2: for.body: -; FVW2-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; FVW2-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; FVW2-NEXT: [[TMP18:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; FVW2-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP18]], 0 -; FVW2-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; FVW2: if.then: -; FVW2-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr addrspace(1) [[IN]], i64 [[INDVARS_IV]], i32 1 -; FVW2-NEXT: [[TMP19:%.*]] = load float, ptr addrspace(1) [[B]], align 4 -; FVW2-NEXT: [[ADD:%.*]] = fadd float [[TMP19]], 5.000000e-01 -; FVW2-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr addrspace(1) [[OUT]], i64 [[INDVARS_IV]] -; FVW2-NEXT: store float [[ADD]], ptr addrspace(1) [[ARRAYIDX5]], align 4 -; FVW2-NEXT: br label [[FOR_INC]] -; FVW2: for.inc: -; FVW2-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; FVW2-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; FVW2-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP9:![0-9]+]] +; FVW2-NEXT: br i1 [[TMP17]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; FVW2: for.end: ; FVW2-NEXT: ret void ; @@ -594,27 +428,7 @@ define void @foo2_addrspace2(ptr addrspace(1) noalias %in, ptr addrspace(0) noal ; AVX512-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 16 ; AVX512-NEXT: [[VEC_IND_NEXT]] = add <16 x i64> [[VEC_IND]], <i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256> ; AVX512-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; AVX512-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] -; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; AVX512: for.body: -; AVX512-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; AVX512-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; AVX512-NEXT: [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; AVX512-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP6]], 0 -; AVX512-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; AVX512: if.then: -; AVX512-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr addrspace(1) [[IN]], i64 [[INDVARS_IV]], i32 1 -; AVX512-NEXT: [[TMP7:%.*]] = load float, ptr addrspace(1) [[B]], align 4 -; AVX512-NEXT: [[ADD:%.*]] = fadd float [[TMP7]], 5.000000e-01 -; AVX512-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[OUT]], i64 [[INDVARS_IV]] -; AVX512-NEXT: store float [[ADD]], ptr [[ARRAYIDX5]], align 4 -; AVX512-NEXT: br label [[FOR_INC]] -; AVX512: for.inc: -; AVX512-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; AVX512-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; AVX512-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP11:![0-9]+]] +; AVX512-NEXT: br i1 [[TMP5]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; AVX512: for.end: ; AVX512-NEXT: ret void ; @@ -656,27 +470,7 @@ define void @foo2_addrspace2(ptr addrspace(1) noalias %in, ptr addrspace(0) noal ; FVW2-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 2 ; FVW2-NEXT: [[VEC_IND_NEXT]] = add <2 x i64> [[VEC_IND]], <i64 32, i64 32> ; FVW2-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; FVW2-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] -; FVW2: middle.block: -; FVW2-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; FVW2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; FVW2: for.body: -; FVW2-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; FVW2-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; FVW2-NEXT: [[TMP18:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; FVW2-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP18]], 0 -; FVW2-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; FVW2: if.then: -; FVW2-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr addrspace(1) [[IN]], i64 [[INDVARS_IV]], i32 1 -; FVW2-NEXT: [[TMP19:%.*]] = load float, ptr addrspace(1) [[B]], align 4 -; FVW2-NEXT: [[ADD:%.*]] = fadd float [[TMP19]], 5.000000e-01 -; FVW2-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[OUT]], i64 [[INDVARS_IV]] -; FVW2-NEXT: store float [[ADD]], ptr [[ARRAYIDX5]], align 4 -; FVW2-NEXT: br label [[FOR_INC]] -; FVW2: for.inc: -; FVW2-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; FVW2-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; FVW2-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP11:![0-9]+]] +; FVW2-NEXT: br i1 [[TMP17]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; FVW2: for.end: ; FVW2-NEXT: ret void ; @@ -727,27 +521,7 @@ define void @foo2_addrspace3(ptr addrspace(0) noalias %in, ptr addrspace(1) noal ; AVX512-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 16 ; AVX512-NEXT: [[VEC_IND_NEXT]] = add <16 x i64> [[VEC_IND]], <i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256, i64 256> ; AVX512-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; AVX512-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] -; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; AVX512: for.body: -; AVX512-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; AVX512-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; AVX512-NEXT: [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; AVX512-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP6]], 0 -; AVX512-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; AVX512: if.then: -; AVX512-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr [[IN]], i64 [[INDVARS_IV]], i32 1 -; AVX512-NEXT: [[TMP7:%.*]] = load float, ptr [[B]], align 4 -; AVX512-NEXT: [[ADD:%.*]] = fadd float [[TMP7]], 5.000000e-01 -; AVX512-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr addrspace(1) [[OUT]], i64 [[INDVARS_IV]] -; AVX512-NEXT: store float [[ADD]], ptr addrspace(1) [[ARRAYIDX5]], align 4 -; AVX512-NEXT: br label [[FOR_INC]] -; AVX512: for.inc: -; AVX512-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; AVX512-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; AVX512-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP13:![0-9]+]] +; AVX512-NEXT: br i1 [[TMP5]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]] ; AVX512: for.end: ; AVX512-NEXT: ret void ; @@ -789,27 +563,7 @@ define void @foo2_addrspace3(ptr addrspace(0) noalias %in, ptr addrspace(1) noal ; FVW2-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX1]], 2 ; FVW2-NEXT: [[VEC_IND_NEXT]] = add <2 x i64> [[VEC_IND]], <i64 32, i64 32> ; FVW2-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 256 -; FVW2-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] -; FVW2: middle.block: -; FVW2-NEXT: [[CMP_N:%.*]] = icmp eq i64 256, 256 -; FVW2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; FVW2: for.body: -; FVW2-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 4096, [[MIDDLE_BLOCK]] ] -; FVW2-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[TRIGGER]], i64 [[INDVARS_IV]] -; FVW2-NEXT: [[TMP18:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; FVW2-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP18]], 0 -; FVW2-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; FVW2: if.then: -; FVW2-NEXT: [[B:%.*]] = getelementptr inbounds [[STRUCT_IN]], ptr [[IN]], i64 [[INDVARS_IV]], i32 1 -; FVW2-NEXT: [[TMP19:%.*]] = load float, ptr [[B]], align 4 -; FVW2-NEXT: [[ADD:%.*]] = fadd float [[TMP19]], 5.000000e-01 -; FVW2-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr addrspace(1) [[OUT]], i64 [[INDVARS_IV]] -; FVW2-NEXT: store float [[ADD]], ptr addrspace(1) [[ARRAYIDX5]], align 4 -; FVW2-NEXT: br label [[FOR_INC]] -; FVW2: for.inc: -; FVW2-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 16 -; FVW2-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 4080 -; FVW2-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP13:![0-9]+]] +; FVW2-NEXT: br i1 [[TMP17]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]] ; FVW2: for.end: ; FVW2-NEXT: ret void ; @@ -899,16 +653,16 @@ define void @test_gather_not_profitable_pr48429(i32 %d, ptr readonly noalias %pt ; AVX512-NEXT: [[TMP17:%.*]] = getelementptr i8, ptr [[POINTER_PHI]], <16 x i64> <i64 0, i64 64, i64 128, i64 192, i64 256, i64 320, i64 384, i64 448, i64 512, i64 576, i64 640, i64 704, i64 768, i64 832, i64 896, i64 960> ; AVX512-NEXT: [[TMP18:%.*]] = getelementptr inbounds float, ptr [[NEXT_GEP]], i64 [[IDXPROM]] ; AVX512-NEXT: [[TMP19:%.*]] = getelementptr inbounds float, ptr [[TMP18]], i32 0 -; AVX512-NEXT: [[WIDE_LOAD:%.*]] = load <16 x float>, ptr [[TMP19]], align 4, !alias.scope !14 -; AVX512-NEXT: call void @llvm.masked.scatter.v16f32.v16p0(<16 x float> [[WIDE_LOAD]], <16 x ptr> [[TMP17]], i32 4, <16 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>), !alias.scope !17, !noalias !19 +; AVX512-NEXT: [[WIDE_LOAD:%.*]] = load <16 x float>, ptr [[TMP19]], align 4, !alias.scope !8 +; AVX512-NEXT: call void @llvm.masked.scatter.v16f32.v16p0(<16 x float> [[WIDE_LOAD]], <16 x ptr> [[TMP17]], i32 4, <16 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>), !alias.scope !11, !noalias !13 ; AVX512-NEXT: [[TMP20:%.*]] = getelementptr float, ptr [[NEXT_GEP]], i32 0 -; AVX512-NEXT: [[WIDE_LOAD8:%.*]] = load <16 x float>, ptr [[TMP20]], align 4, !alias.scope !21 +; AVX512-NEXT: [[WIDE_LOAD8:%.*]] = load <16 x float>, ptr [[TMP20]], align 4, !alias.scope !15 ; AVX512-NEXT: [[TMP21:%.*]] = getelementptr inbounds float, <16 x ptr> [[TMP17]], i64 1 -; AVX512-NEXT: call void @llvm.masked.scatter.v16f32.v16p0(<16 x float> [[WIDE_LOAD8]], <16 x ptr> [[TMP21]], i32 4, <16 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>), !alias.scope !17, !noalias !19 +; AVX512-NEXT: call void @llvm.masked.scatter.v16f32.v16p0(<16 x float> [[WIDE_LOAD8]], <16 x ptr> [[TMP21]], i32 4, <16 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>), !alias.scope !11, !noalias !13 ; AVX512-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 16 ; AVX512-NEXT: [[PTR_IND]] = getelementptr i8, ptr [[POINTER_PHI]], i64 1024 ; AVX512-NEXT: [[TMP22:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] -; AVX512-NEXT: br i1 [[TMP22]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] +; AVX512-NEXT: br i1 [[TMP22]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; AVX512: middle.block: ; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP3]], [[N_VEC]] ; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END]], label [[VEC_EPILOG_ITER_CHECK:%.*]] @@ -940,16 +694,16 @@ define void @test_gather_not_profitable_pr48429(i32 %d, ptr readonly noalias %pt ; AVX512-NEXT: [[TMP29:%.*]] = getelementptr i8, ptr [[POINTER_PHI22]], <8 x i64> <i64 0, i64 64, i64 128, i64 192, i64 256, i64 320, i64 384, i64 448> ; AVX512-NEXT: [[TMP30:%.*]] = getelementptr inbounds float, ptr [[NEXT_GEP21]], i64 [[IDXPROM]] ; AVX512-NEXT: [[TMP31:%.*]] = getelementptr inbounds float, ptr [[TMP30]], i32 0 -; AVX512-NEXT: [[WIDE_LOAD24:%.*]] = load <8 x float>, ptr [[TMP31]], align 4, !alias.scope !23 -; AVX512-NEXT: call void @llvm.masked.scatter.v8f32.v8p0(<8 x float> [[WIDE_LOAD24]], <8 x ptr> [[TMP29]], i32 4, <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>), !alias.scope !26, !noalias !28 +; AVX512-NEXT: [[WIDE_LOAD24:%.*]] = load <8 x float>, ptr [[TMP31]], align 4, !alias.scope !17 +; AVX512-NEXT: call void @llvm.masked.scatter.v8f32.v8p0(<8 x float> [[WIDE_LOAD24]], <8 x ptr> [[TMP29]], i32 4, <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>), !alias.scope !20, !noalias !22 ; AVX512-NEXT: [[TMP32:%.*]] = getelementptr float, ptr [[NEXT_GEP21]], i32 0 -; AVX512-NEXT: [[WIDE_LOAD25:%.*]] = load <8 x float>, ptr [[TMP32]], align 4, !alias.scope !30 +; AVX512-NEXT: [[WIDE_LOAD25:%.*]] = load <8 x float>, ptr [[TMP32]], align 4, !alias.scope !24 ; AVX512-NEXT: [[TMP33:%.*]] = getelementptr inbounds float, <8 x ptr> [[TMP29]], i64 1 -; AVX512-NEXT: call void @llvm.masked.scatter.v8f32.v8p0(<8 x float> [[WIDE_LOAD25]], <8 x ptr> [[TMP33]], i32 4, <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>), !alias.scope !26, !noalias !28 +; AVX512-NEXT: call void @llvm.masked.scatter.v8f32.v8p0(<8 x float> [[WIDE_LOAD25]], <8 x ptr> [[TMP33]], i32 4, <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>), !alias.scope !20, !noalias !22 ; AVX512-NEXT: [[INDEX_NEXT26]] = add nuw i64 [[INDEX20]], 8 ; AVX512-NEXT: [[PTR_IND23]] = getelementptr i8, ptr [[POINTER_PHI22]], i64 512 ; AVX512-NEXT: [[TMP34:%.*]] = icmp eq i64 [[INDEX_NEXT26]], [[N_VEC12]] -; AVX512-NEXT: br i1 [[TMP34]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP31:![0-9]+]] +; AVX512-NEXT: br i1 [[TMP34]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP25:![0-9]+]] ; AVX512: vec.epilog.middle.block: ; AVX512-NEXT: [[CMP_N19:%.*]] = icmp eq i64 [[TMP3]], [[N_VEC12]] ; AVX512-NEXT: br i1 [[CMP_N19]], label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] @@ -969,7 +723,7 @@ define void @test_gather_not_profitable_pr48429(i32 %d, ptr readonly noalias %pt ; AVX512-NEXT: [[INCDEC_PTR]] = getelementptr inbounds float, ptr [[PTR_ADDR_012]], i64 1 ; AVX512-NEXT: [[ADD_PTR6]] = getelementptr inbounds float, ptr [[DEST_ADDR_011]], i64 16 ; AVX512-NEXT: [[CMP_NOT:%.*]] = icmp eq ptr [[INCDEC_PTR]], [[ADD_PTR]] -; AVX512-NEXT: br i1 [[CMP_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] +; AVX512-NEXT: br i1 [[CMP_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP26:![0-9]+]] ; AVX512: for.end: ; AVX512-NEXT: ret void ; @@ -1031,22 +785,22 @@ define void @test_gather_not_profitable_pr48429(i32 %d, ptr readonly noalias %pt ; FVW2-NEXT: [[NEXT_GEP10:%.*]] = getelementptr i8, ptr [[DEST]], i64 [[TMP20]] ; FVW2-NEXT: [[TMP21:%.*]] = getelementptr inbounds float, ptr [[NEXT_GEP]], i64 [[IDXPROM]] ; FVW2-NEXT: [[TMP22:%.*]] = getelementptr inbounds float, ptr [[TMP21]], i32 0 -; FVW2-NEXT: [[WIDE_LOAD:%.*]] = load <2 x float>, ptr [[TMP22]], align 4, !alias.scope !14 +; FVW2-NEXT: [[WIDE_LOAD:%.*]] = load <2 x float>, ptr [[TMP22]], align 4, !alias.scope !8 ; FVW2-NEXT: [[TMP23:%.*]] = extractelement <2 x float> [[WIDE_LOAD]], i32 0 -; FVW2-NEXT: store float [[TMP23]], ptr [[NEXT_GEP9]], align 4, !alias.scope !17, !noalias !19 +; FVW2-NEXT: store float [[TMP23]], ptr [[NEXT_GEP9]], align 4, !alias.scope !11, !noalias !13 ; FVW2-NEXT: [[TMP24:%.*]] = extractelement <2 x float> [[WIDE_LOAD]], i32 1 -; FVW2-NEXT: store float [[TMP24]], ptr [[NEXT_GEP10]], align 4, !alias.scope !17, !noalias !19 +; FVW2-NEXT: store float [[TMP24]], ptr [[NEXT_GEP10]], align 4, !alias.scope !11, !noalias !13 ; FVW2-NEXT: [[TMP25:%.*]] = getelementptr float, ptr [[NEXT_GEP]], i32 0 -; FVW2-NEXT: [[WIDE_LOAD11:%.*]] = load <2 x float>, ptr [[TMP25]], align 4, !alias.scope !21 +; FVW2-NEXT: [[WIDE_LOAD11:%.*]] = load <2 x float>, ptr [[TMP25]], align 4, !alias.scope !15 ; FVW2-NEXT: [[TMP26:%.*]] = getelementptr inbounds float, ptr [[NEXT_GEP9]], i64 1 ; FVW2-NEXT: [[TMP27:%.*]] = getelementptr inbounds float, ptr [[NEXT_GEP10]], i64 1 ; FVW2-NEXT: [[TMP28:%.*]] = extractelement <2 x float> [[WIDE_LOAD11]], i32 0 -; FVW2-NEXT: store float [[TMP28]], ptr [[TMP26]], align 4, !alias.scope !17, !noalias !19 +; FVW2-NEXT: store float [[TMP28]], ptr [[TMP26]], align 4, !alias.scope !11, !noalias !13 ; FVW2-NEXT: [[TMP29:%.*]] = extractelement <2 x float> [[WIDE_LOAD11]], i32 1 -; FVW2-NEXT: store float [[TMP29]], ptr [[TMP27]], align 4, !alias.scope !17, !noalias !19 +; FVW2-NEXT: store float [[TMP29]], ptr [[TMP27]], align 4, !alias.scope !11, !noalias !13 ; FVW2-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; FVW2-NEXT: [[TMP30:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] -; FVW2-NEXT: br i1 [[TMP30]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] +; FVW2-NEXT: br i1 [[TMP30]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; FVW2: middle.block: ; FVW2-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP3]], [[N_VEC]] ; FVW2-NEXT: br i1 [[CMP_N]], label [[FOR_END]], label [[SCALAR_PH]] @@ -1066,7 +820,7 @@ define void @test_gather_not_profitable_pr48429(i32 %d, ptr readonly noalias %pt ; FVW2-NEXT: [[INCDEC_PTR]] = getelementptr inbounds float, ptr [[PTR_ADDR_012]], i64 1 ; FVW2-NEXT: [[ADD_PTR6]] = getelementptr inbounds float, ptr [[DEST_ADDR_011]], i64 16 ; FVW2-NEXT: [[CMP_NOT:%.*]] = icmp eq ptr [[INCDEC_PTR]], [[ADD_PTR]] -; FVW2-NEXT: br i1 [[CMP_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP23:![0-9]+]] +; FVW2-NEXT: br i1 [[CMP_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP17:![0-9]+]] ; FVW2: for.end: ; FVW2-NEXT: ret void ; diff --git a/llvm/test/Transforms/LoopVectorize/X86/imprecise-through-phis.ll b/llvm/test/Transforms/LoopVectorize/X86/imprecise-through-phis.ll index 1c7f3a7..8c33e7b 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/imprecise-through-phis.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/imprecise-through-phis.ll @@ -86,8 +86,7 @@ define double @sumIfVector(ptr nocapture readonly %arr) { ; SSE-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; SSE: middle.block: ; SSE-NEXT: [[TMP7:%.*]] = call fast double @llvm.vector.reduce.fadd.v2f64(double -0.000000e+00, <2 x double> [[PREDPHI]]) -; SSE-NEXT: [[CMP_N:%.*]] = icmp eq i32 32, 32 -; SSE-NEXT: br i1 [[CMP_N]], label [[DONE:%.*]], label [[SCALAR_PH]] +; SSE-NEXT: br i1 true, label [[DONE:%.*]], label [[SCALAR_PH]] ; SSE: scalar.ph: ; SSE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 32, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; SSE-NEXT: [[BC_MERGE_RDX:%.*]] = phi double [ 0.000000e+00, [[ENTRY]] ], [ [[TMP7]], [[MIDDLE_BLOCK]] ] @@ -134,8 +133,7 @@ define double @sumIfVector(ptr nocapture readonly %arr) { ; AVX-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; AVX: middle.block: ; AVX-NEXT: [[TMP7:%.*]] = call fast double @llvm.vector.reduce.fadd.v4f64(double -0.000000e+00, <4 x double> [[PREDPHI]]) -; AVX-NEXT: [[CMP_N:%.*]] = icmp eq i32 32, 32 -; AVX-NEXT: br i1 [[CMP_N]], label [[DONE:%.*]], label [[SCALAR_PH]] +; AVX-NEXT: br i1 true, label [[DONE:%.*]], label [[SCALAR_PH]] ; AVX: scalar.ph: ; AVX-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 32, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; AVX-NEXT: [[BC_MERGE_RDX:%.*]] = phi double [ 0.000000e+00, [[ENTRY]] ], [ [[TMP7]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/X86/limit-vf-by-tripcount.ll b/llvm/test/Transforms/LoopVectorize/X86/limit-vf-by-tripcount.ll index 93c78f4..e302301 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/limit-vf-by-tripcount.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/limit-vf-by-tripcount.ll @@ -24,8 +24,7 @@ define void @test_tc_17_no_epilogue_vectorization(ptr noalias %src, ptr noalias ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 16 ; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 17, 16 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -79,8 +78,7 @@ define void @test_tc_18(ptr noalias %src, ptr noalias %dst) { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 16 ; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 18, 16 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK: vec.epilog.iter.check: ; CHECK-NEXT: br i1 false, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; CHECK: vec.epilog.ph: @@ -99,8 +97,7 @@ define void @test_tc_18(ptr noalias %src, ptr noalias %dst) { ; CHECK-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT4]], 18 ; CHECK-NEXT: br i1 [[TMP11]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; CHECK: vec.epilog.middle.block: -; CHECK-NEXT: [[CMP_N1:%.*]] = icmp eq i64 18, 18 -; CHECK-NEXT: br i1 [[CMP_N1]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 18, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 16, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -154,8 +151,7 @@ define void @test_tc_19(ptr noalias %src, ptr noalias %dst) { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 16 ; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 19, 16 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK: vec.epilog.iter.check: ; CHECK-NEXT: br i1 false, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; CHECK: vec.epilog.ph: @@ -174,8 +170,7 @@ define void @test_tc_19(ptr noalias %src, ptr noalias %dst) { ; CHECK-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT4]], 18 ; CHECK-NEXT: br i1 [[TMP11]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: vec.epilog.middle.block: -; CHECK-NEXT: [[CMP_N1:%.*]] = icmp eq i64 19, 18 -; CHECK-NEXT: br i1 [[CMP_N1]], label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 18, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 16, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -227,8 +222,7 @@ define void @test_tc_20(ptr noalias %src, ptr noalias %dst) { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 20 ; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 20, 20 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 20, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/load-deref-pred.ll b/llvm/test/Transforms/LoopVectorize/X86/load-deref-pred.ll index 2011d9b..09be042 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/load-deref-pred.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/load-deref-pred.ll @@ -75,8 +75,7 @@ define i32 @test_explicit_pred(i64 %len) { ; CHECK-NEXT: [[BIN_RDX13:%.*]] = add <4 x i32> [[TMP22]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX14:%.*]] = add <4 x i32> [[TMP23]], [[BIN_RDX13]] ; CHECK-NEXT: [[TMP25:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX14]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP25]], [[MIDDLE_BLOCK]] ] @@ -235,8 +234,7 @@ define i32 @test_explicit_pred_generic(i64 %len, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP78]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP79]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP81:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP81]], [[MIDDLE_BLOCK]] ] @@ -420,8 +418,7 @@ define i32 @test_invariant_address(i64 %len, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX7:%.*]] = add <4 x i32> [[TMP102]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX8:%.*]] = add <4 x i32> [[TMP103]], [[BIN_RDX7]] ; CHECK-NEXT: [[TMP105:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX8]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP105]], [[MIDDLE_BLOCK]] ] @@ -712,8 +709,7 @@ define i32 @test_step_narrower_than_access(i64 %len, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX37:%.*]] = add <4 x i32> [[TMP150]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX38:%.*]] = add <4 x i32> [[TMP151]], [[BIN_RDX37]] ; CHECK-NEXT: [[TMP153:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX38]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP153]], [[MIDDLE_BLOCK]] ] @@ -1046,8 +1042,7 @@ define i32 @test_non_zero_start(i64 %len, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP78]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP79]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP81:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 3072, 3072 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 1024, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP81]], [[MIDDLE_BLOCK]] ] @@ -1295,8 +1290,7 @@ define i32 @test_non_unit_stride(i64 %len, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX7:%.*]] = add <4 x i32> [[TMP118]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX8:%.*]] = add <4 x i32> [[TMP119]], [[BIN_RDX7]] ; CHECK-NEXT: [[TMP121:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX8]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 2048, 2048 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP121]], [[MIDDLE_BLOCK]] ] @@ -1456,8 +1450,7 @@ define i32 @neg_off_by_many(i64 %len, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP78]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP79]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP81:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP81]], [[MIDDLE_BLOCK]] ] @@ -1617,8 +1610,7 @@ define i32 @neg_off_by_one_iteration(i64 %len, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP78]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP79]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP81:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP81]], [[MIDDLE_BLOCK]] ] @@ -1778,8 +1770,7 @@ define i32 @neg_off_by_one_byte(i64 %len, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP78]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP79]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP81:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP81]], [[MIDDLE_BLOCK]] ] @@ -2116,8 +2107,7 @@ define i32 @test_allocsize(i64 %len, ptr %test_base) nofree nosync { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP78]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP79]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP81:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP81]], [[MIDDLE_BLOCK]] ] @@ -2278,8 +2268,7 @@ define i32 @test_allocsize_array(i64 %len, ptr %test_base) nofree nosync { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP78]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP79]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP81:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP81]], [[MIDDLE_BLOCK]] ] @@ -2450,8 +2439,7 @@ define i32 @test_allocsize_cond_deref(i1 %allzero, ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP78]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP79]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP81:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[PREHEADER]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[PREHEADER]] ], [ [[TMP81]], [[MIDDLE_BLOCK]] ] @@ -2557,8 +2545,7 @@ define i32 @test_stride_three(i64 %len, ptr %test_base) { ; CHECK-NEXT: br i1 [[TMP30]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP30:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP31:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP29]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 35, 32 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 96, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP31]], [[MIDDLE_BLOCK]] ] @@ -2656,8 +2643,7 @@ define i32 @test_non_unit_stride_four(i64 %len, ptr %test_base) { ; CHECK-NEXT: br i1 [[TMP30]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP31:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP29]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 27, 24 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 96, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP31]], [[MIDDLE_BLOCK]] ] @@ -2755,8 +2741,7 @@ define i32 @test_non_unit_stride_five(i64 %len, ptr %test_base) { ; CHECK-NEXT: br i1 [[TMP30]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP34:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP31:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP29]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 22, 20 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP31]], [[MIDDLE_BLOCK]] ] @@ -2878,8 +2863,7 @@ define i32 @neg_test_non_unit_stride_off_by_four_bytes(i64 %len, ptr %test_base) ; CHECK-NEXT: br i1 [[TMP38]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP36:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP39:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP37]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 52, 52 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 104, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP39]], [[MIDDLE_BLOCK]] ] @@ -3093,8 +3077,7 @@ define i32 @test_non_unit_stride_with_first_iteration_step_access(i64 %len, ptr ; CHECK-NEXT: [[BIN_RDX7:%.*]] = add <4 x i32> [[TMP134]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX8:%.*]] = add <4 x i32> [[TMP135]], [[BIN_RDX7]] ; CHECK-NEXT: [[TMP137:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX8]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 152, 144 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 288, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP137]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/X86/masked_load_store.ll b/llvm/test/Transforms/LoopVectorize/X86/masked_load_store.ll index a4b962f..40ac668 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/masked_load_store.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/masked_load_store.ll @@ -51,8 +51,7 @@ define void @foo1(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX1-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT]], 10000 ; AVX1-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; AVX1: middle.block: -; AVX1-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 10000 -; AVX1-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX1-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX1: scalar.ph: ; AVX1-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 10000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX1-NEXT: br label [[FOR_BODY:%.*]] @@ -145,8 +144,7 @@ define void @foo1(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX2-NEXT: [[TMP38:%.*]] = icmp eq i64 [[INDEX_NEXT]], 9984 ; AVX2-NEXT: br i1 [[TMP38]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; AVX2: middle.block: -; AVX2-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 9984 -; AVX2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX2-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX2: scalar.ph: ; AVX2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 9984, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX2-NEXT: br label [[FOR_BODY:%.*]] @@ -241,8 +239,7 @@ define void @foo1(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX512-NEXT: [[TMP38:%.*]] = icmp eq i64 [[INDEX_NEXT]], 9984 ; AVX512-NEXT: br i1 [[TMP38]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 9984 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; AVX512-NEXT: br i1 false, label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; AVX512: vec.epilog.iter.check: ; AVX512-NEXT: br i1 false, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; AVX512: vec.epilog.ph: @@ -266,8 +263,7 @@ define void @foo1(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX512-NEXT: [[TMP48:%.*]] = icmp eq i64 [[INDEX_NEXT15]], 10000 ; AVX512-NEXT: br i1 [[TMP48]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] ; AVX512: vec.epilog.middle.block: -; AVX512-NEXT: [[CMP_N11:%.*]] = icmp eq i64 10000, 10000 -; AVX512-NEXT: br i1 [[CMP_N11]], label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] +; AVX512-NEXT: br i1 true, label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] ; AVX512: vec.epilog.scalar.ph: ; AVX512-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 10000, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 9984, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MEMCHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; AVX512-NEXT: br label [[FOR_BODY:%.*]] @@ -354,8 +350,7 @@ define void @foo1_addrspace1(ptr addrspace(1) nocapture %A, ptr addrspace(1) noc ; AVX1-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT]], 10000 ; AVX1-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; AVX1: middle.block: -; AVX1-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 10000 -; AVX1-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX1-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX1: scalar.ph: ; AVX1-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 10000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX1-NEXT: br label [[FOR_BODY:%.*]] @@ -448,8 +443,7 @@ define void @foo1_addrspace1(ptr addrspace(1) nocapture %A, ptr addrspace(1) noc ; AVX2-NEXT: [[TMP38:%.*]] = icmp eq i64 [[INDEX_NEXT]], 9984 ; AVX2-NEXT: br i1 [[TMP38]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; AVX2: middle.block: -; AVX2-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 9984 -; AVX2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX2-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX2: scalar.ph: ; AVX2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 9984, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX2-NEXT: br label [[FOR_BODY:%.*]] @@ -544,8 +538,7 @@ define void @foo1_addrspace1(ptr addrspace(1) nocapture %A, ptr addrspace(1) noc ; AVX512-NEXT: [[TMP38:%.*]] = icmp eq i64 [[INDEX_NEXT]], 9984 ; AVX512-NEXT: br i1 [[TMP38]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 9984 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; AVX512-NEXT: br i1 false, label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; AVX512: vec.epilog.iter.check: ; AVX512-NEXT: br i1 false, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; AVX512: vec.epilog.ph: @@ -569,8 +562,7 @@ define void @foo1_addrspace1(ptr addrspace(1) nocapture %A, ptr addrspace(1) noc ; AVX512-NEXT: [[TMP48:%.*]] = icmp eq i64 [[INDEX_NEXT15]], 10000 ; AVX512-NEXT: br i1 [[TMP48]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; AVX512: vec.epilog.middle.block: -; AVX512-NEXT: [[CMP_N11:%.*]] = icmp eq i64 10000, 10000 -; AVX512-NEXT: br i1 [[CMP_N11]], label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] +; AVX512-NEXT: br i1 true, label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] ; AVX512: vec.epilog.scalar.ph: ; AVX512-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 10000, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 9984, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MEMCHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; AVX512-NEXT: br label [[FOR_BODY:%.*]] @@ -667,8 +659,7 @@ define void @foo2(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX1-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], 10000 ; AVX1-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; AVX1: middle.block: -; AVX1-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 10000 -; AVX1-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX1-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX1: scalar.ph: ; AVX1-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 10000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX1-NEXT: br label [[FOR_BODY:%.*]] @@ -766,8 +757,7 @@ define void @foo2(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX2-NEXT: [[TMP42:%.*]] = icmp eq i64 [[INDEX_NEXT]], 9984 ; AVX2-NEXT: br i1 [[TMP42]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; AVX2: middle.block: -; AVX2-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 9984 -; AVX2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX2-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX2: scalar.ph: ; AVX2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 9984, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX2-NEXT: br label [[FOR_BODY:%.*]] @@ -867,8 +857,7 @@ define void @foo2(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX512-NEXT: [[TMP42:%.*]] = icmp eq i64 [[INDEX_NEXT]], 9984 ; AVX512-NEXT: br i1 [[TMP42]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 9984 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; AVX512-NEXT: br i1 false, label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; AVX512: vec.epilog.iter.check: ; AVX512-NEXT: br i1 false, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; AVX512: vec.epilog.ph: @@ -893,8 +882,7 @@ define void @foo2(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX512-NEXT: [[TMP53:%.*]] = icmp eq i64 [[INDEX_NEXT15]], 10000 ; AVX512-NEXT: br i1 [[TMP53]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP9:![0-9]+]] ; AVX512: vec.epilog.middle.block: -; AVX512-NEXT: [[CMP_N11:%.*]] = icmp eq i64 10000, 10000 -; AVX512-NEXT: br i1 [[CMP_N11]], label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] +; AVX512-NEXT: br i1 true, label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] ; AVX512: vec.epilog.scalar.ph: ; AVX512-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 10000, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 9984, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MEMCHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; AVX512-NEXT: br label [[FOR_BODY:%.*]] @@ -1034,8 +1022,7 @@ define void @foo3(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX-NEXT: [[TMP40:%.*]] = icmp eq i64 [[INDEX_NEXT]], 10000 ; AVX-NEXT: br i1 [[TMP40]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; AVX: middle.block: -; AVX-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 10000 -; AVX-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX: scalar.ph: ; AVX-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 10000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX-NEXT: br label [[FOR_BODY:%.*]] @@ -1135,8 +1122,7 @@ define void @foo3(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX512-NEXT: [[TMP40:%.*]] = icmp eq i64 [[INDEX_NEXT]], 9984 ; AVX512-NEXT: br i1 [[TMP40]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP19:![0-9]+]] ; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 9984 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX512-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX512: scalar.ph: ; AVX512-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 9984, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX512-NEXT: br label [[FOR_BODY:%.*]] @@ -1261,8 +1247,7 @@ define void @foo4(ptr nocapture %A, ptr nocapture readonly %B, ptr nocapture rea ; AVX512-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 624 ; AVX512-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP29:![0-9]+]] ; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 625, 624 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX512-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX512: scalar.ph: ; AVX512-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 9984, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; AVX512-NEXT: br label [[FOR_BODY:%.*]] @@ -1454,8 +1439,7 @@ define void @foo6(ptr nocapture readonly %in, ptr nocapture %out, i32 %size, ptr ; AVX2-NEXT: [[TMP48:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 ; AVX2-NEXT: br i1 [[TMP48]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP26:![0-9]+]] ; AVX2: middle.block: -; AVX2-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; AVX2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX2-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX2: scalar.ph: ; AVX2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ -1, [[MIDDLE_BLOCK]] ], [ 4095, [[ENTRY:%.*]] ], [ 4095, [[VECTOR_MEMCHECK]] ] ; AVX2-NEXT: br label [[FOR_BODY:%.*]] @@ -1579,8 +1563,7 @@ define void @foo6(ptr nocapture readonly %in, ptr nocapture %out, i32 %size, ptr ; AVX512-NEXT: [[TMP48:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 ; AVX512-NEXT: br i1 [[TMP48]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP39:![0-9]+]] ; AVX512: middle.block: -; AVX512-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; AVX512-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; AVX512-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; AVX512: scalar.ph: ; AVX512-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ -1, [[MIDDLE_BLOCK]] ], [ 4095, [[ENTRY:%.*]] ], [ 4095, [[VECTOR_MEMCHECK]] ] ; AVX512-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/metadata-enable.ll b/llvm/test/Transforms/LoopVectorize/X86/metadata-enable.ll index 60b7398..a1dd890 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/metadata-enable.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/metadata-enable.ll @@ -1197,8 +1197,7 @@ define i32 @nopragma(ptr noalias nocapture %a, ptr noalias nocapture readonly %b ; O1VEC2-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 64 ; O1VEC2-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; O1VEC2: middle.block: -; O1VEC2-NEXT: [[CMP_N:%.*]] = icmp eq i64 64, 64 -; O1VEC2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; O1VEC2-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; O1VEC2: scalar.ph: ; O1VEC2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 64, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; O1VEC2-NEXT: br label [[FOR_BODY:%.*]] @@ -1237,8 +1236,7 @@ define i32 @nopragma(ptr noalias nocapture %a, ptr noalias nocapture readonly %b ; OzVEC2-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 64 ; OzVEC2-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; OzVEC2: middle.block: -; OzVEC2-NEXT: [[CMP_N:%.*]] = icmp eq i64 64, 64 -; OzVEC2-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; OzVEC2-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; OzVEC2: scalar.ph: ; OzVEC2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 64, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; OzVEC2-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/optsize.ll b/llvm/test/Transforms/LoopVectorize/X86/optsize.ll index 6a514c4..ce6dd52 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/optsize.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/optsize.ll @@ -234,8 +234,7 @@ define void @scev4stride1(ptr noalias nocapture %a, ptr noalias nocapture readon ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i32 [[INDEX_NEXT]], 256 ; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 256, 256 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_END_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 256, [[MIDDLE_BLOCK]] ], [ 0, [[FOR_BODY_PREHEADER:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -274,8 +273,7 @@ define void @scev4stride1(ptr noalias nocapture %a, ptr noalias nocapture readon ; AUTOVF-NEXT: [[TMP5:%.*]] = icmp eq i32 [[INDEX_NEXT]], 256 ; AUTOVF-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; AUTOVF: middle.block: -; AUTOVF-NEXT: [[CMP_N:%.*]] = icmp eq i32 256, 256 -; AUTOVF-NEXT: br i1 [[CMP_N]], label [[FOR_END_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; AUTOVF-NEXT: br i1 true, label [[FOR_END_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; AUTOVF: scalar.ph: ; AUTOVF-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 256, [[MIDDLE_BLOCK]] ], [ 0, [[FOR_BODY_PREHEADER:%.*]] ] ; AUTOVF-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/pr34438.ll b/llvm/test/Transforms/LoopVectorize/X86/pr34438.ll index 96046dd..cc60359 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/pr34438.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/pr34438.ll @@ -29,8 +29,7 @@ define void @small_tc(ptr noalias nocapture %A, ptr noalias nocapture readonly % ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 8 ; CHECK-NEXT: br i1 true, label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP1:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 8, 8 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 8, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/pr36524.ll b/llvm/test/Transforms/LoopVectorize/X86/pr36524.ll index 18ac690..337759e 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/pr36524.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/pr36524.ll @@ -35,8 +35,7 @@ define void @foo(ptr %ptr, ptr %ptr.2) { ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 80 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 80, 80 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 80, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; CHECK-NEXT: [[BC_RESUME_VAL2:%.*]] = phi i64 [ 82, [[MIDDLE_BLOCK]] ], [ 2, [[ENTRY]] ], [ 2, [[VECTOR_MEMCHECK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/X86/pr51366-sunk-instruction-used-outside-of-loop.ll b/llvm/test/Transforms/LoopVectorize/X86/pr51366-sunk-instruction-used-outside-of-loop.ll index 52f5934..c139afa 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/pr51366-sunk-instruction-used-outside-of-loop.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/pr51366-sunk-instruction-used-outside-of-loop.ll @@ -37,8 +37,7 @@ define ptr @test(ptr noalias %src, ptr noalias %dst) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label %vector.body ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label %exit, label %scalar.ph +; CHECK-NEXT: br i1 true, label %exit, label %scalar.ph ; CHECK: exit: ; CHECK-NEXT: [[GEP_LCSSA:%.*]] = phi ptr [ %gep.src, %loop.latch ], [ [[TMP2]], %middle.block ] ; CHECK-NEXT: ret ptr [[GEP_LCSSA]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/reduction-fastmath.ll b/llvm/test/Transforms/LoopVectorize/X86/reduction-fastmath.ll index f9116c0..6c2cb10 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/reduction-fastmath.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/reduction-fastmath.ll @@ -75,8 +75,7 @@ define float @reduction_sum_float_fastmath(i32 %n, ptr %array) { ; CHECK: middle.block: ; CHECK-NEXT: [[BIN_RDX:%.*]] = fadd fast <4 x float> [[TMP7]], [[TMP6]] ; CHECK-NEXT: [[TMP9:%.*]] = call fast float @llvm.vector.reduce.fadd.v4f32(float -0.000000e+00, <4 x float> [[BIN_RDX]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_PREHEADER]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi float [ 0.000000e+00, [[LOOP_PREHEADER]] ], [ [[TMP9]], [[MIDDLE_BLOCK]] ] @@ -145,8 +144,7 @@ define float @reduction_sum_float_only_reassoc(i32 %n, ptr %array) { ; CHECK: middle.block: ; CHECK-NEXT: [[BIN_RDX:%.*]] = fadd reassoc <4 x float> [[TMP7]], [[TMP6]] ; CHECK-NEXT: [[TMP9:%.*]] = call reassoc float @llvm.vector.reduce.fadd.v4f32(float -0.000000e+00, <4 x float> [[BIN_RDX]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_PREHEADER]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi float [ -0.000000e+00, [[LOOP_PREHEADER]] ], [ [[TMP9]], [[MIDDLE_BLOCK]] ] @@ -215,8 +213,7 @@ define float @reduction_sum_float_only_reassoc_and_contract(i32 %n, ptr %array) ; CHECK: middle.block: ; CHECK-NEXT: [[BIN_RDX:%.*]] = fadd reassoc contract <4 x float> [[TMP7]], [[TMP6]] ; CHECK-NEXT: [[TMP9:%.*]] = call reassoc contract float @llvm.vector.reduce.fadd.v4f32(float -0.000000e+00, <4 x float> [[BIN_RDX]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_PREHEADER]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi float [ -0.000000e+00, [[LOOP_PREHEADER]] ], [ [[TMP9]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/X86/strided_load_cost.ll b/llvm/test/Transforms/LoopVectorize/X86/strided_load_cost.ll index e27e44b..1917023 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/strided_load_cost.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/strided_load_cost.ll @@ -62,8 +62,7 @@ define i32 @matrix_row_col(ptr nocapture readonly %data, i32 %i, i32 %j) local_u ; CHECK-NEXT: br i1 [[TMP37]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP38:%.*]] = call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> [[TMP36]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 100, 96 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 96, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP38]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/X86/uniform_mem_op.ll b/llvm/test/Transforms/LoopVectorize/X86/uniform_mem_op.ll index 5505b17..c5c80c3 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/uniform_mem_op.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/uniform_mem_op.ll @@ -21,8 +21,7 @@ define i32 @uniform_load(ptr align(4) %addr) { ; CHECK-NEXT: [[TMP1:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 ; CHECK-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4097, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -77,8 +76,7 @@ define i32 @uniform_load2(ptr align(4) %addr) { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP3]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP4]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP6:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4097, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP6]], [[MIDDLE_BLOCK]] ] @@ -135,8 +133,7 @@ define i32 @uniform_address(ptr align(4) %addr, i32 %byte_offset) { ; CHECK-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 ; CHECK-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4097, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -183,8 +180,7 @@ define void @uniform_store_uniform_value(ptr align(4) %addr) { ; CHECK-NEXT: [[TMP0:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 ; CHECK-NEXT: br i1 [[TMP0]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4097, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -241,8 +237,7 @@ define void @uniform_store_varying_value(ptr align(4) %addr) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4097, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -323,8 +318,7 @@ define void @uniform_copy(ptr %A, ptr %B) { ; CHECK-NEXT: [[TMP1:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4096 ; CHECK-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP17:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4097, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -436,8 +430,7 @@ define i32 @test_count_bits(ptr %test_base) { ; CHECK-NEXT: [[BIN_RDX13:%.*]] = add <4 x i32> [[TMP38]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX14:%.*]] = add <4 x i32> [[TMP39]], [[BIN_RDX13]] ; CHECK-NEXT: [[TMP41:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX14]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP41]], [[MIDDLE_BLOCK]] ] @@ -515,8 +508,7 @@ define i32 @uniform_load_global() { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP3]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP4]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP6:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4097, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP6]], [[MIDDLE_BLOCK]] ] @@ -577,8 +569,7 @@ define i32 @uniform_load_constexpr() { ; CHECK-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[TMP3]], [[BIN_RDX]] ; CHECK-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[TMP4]], [[BIN_RDX10]] ; CHECK-NEXT: [[TMP6:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4097, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP6]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/X86/vect.omp.force.small-tc.ll b/llvm/test/Transforms/LoopVectorize/X86/vect.omp.force.small-tc.ll index c95870b..3c2ea22 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/vect.omp.force.small-tc.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/vect.omp.force.small-tc.ll @@ -39,8 +39,7 @@ define void @vectorized(ptr noalias nocapture %A, ptr noalias nocapture readonly ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 20 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP1:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 20, 20 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 20, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] @@ -175,8 +174,7 @@ define void @vectorized2(ptr noalias nocapture %A, ptr noalias nocapture readonl ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 16 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 16, 16 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/X86/x86-predication.ll b/llvm/test/Transforms/LoopVectorize/X86/x86-predication.ll index 0c2cbc8..9135572 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/x86-predication.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/x86-predication.ll @@ -52,29 +52,7 @@ define i32 @predicated_sdiv_masked_load(ptr %a, ptr %b, i32 %x, i1 %c) { ; CHECK-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP19:%.*]] = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> [[TMP17]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 10000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; CHECK: for.body: -; CHECK-NEXT: [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 10000, [[MIDDLE_BLOCK]] ] -; CHECK-NEXT: [[R:%.*]] = phi i32 [ [[T7:%.*]], [[FOR_INC]] ], [ [[TMP19]], [[MIDDLE_BLOCK]] ] -; CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds i32, ptr [[A]], i64 [[I]] -; CHECK-NEXT: [[T1:%.*]] = load i32, ptr [[T0]], align 4 -; CHECK-NEXT: br i1 [[C]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; CHECK: if.then: -; CHECK-NEXT: [[T2:%.*]] = getelementptr inbounds i32, ptr [[B]], i64 [[I]] -; CHECK-NEXT: [[T3:%.*]] = load i32, ptr [[T2]], align 4 -; CHECK-NEXT: [[T4:%.*]] = sdiv i32 [[T3]], [[X]] -; CHECK-NEXT: [[T5:%.*]] = add nsw i32 [[T4]], [[T1]] -; CHECK-NEXT: br label [[FOR_INC]] -; CHECK: for.inc: -; CHECK-NEXT: [[T6:%.*]] = phi i32 [ [[T1]], [[FOR_BODY]] ], [ [[T5]], [[IF_THEN]] ] -; CHECK-NEXT: [[T7]] = add i32 [[R]], [[T6]] -; CHECK-NEXT: [[I_NEXT]] = add nuw nsw i64 [[I]], 1 -; CHECK-NEXT: [[COND:%.*]] = icmp eq i64 [[I_NEXT]], 10000 -; CHECK-NEXT: br i1 [[COND]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP2:![0-9]+]] -; CHECK: for.end: -; CHECK-NEXT: [[T8:%.*]] = phi i32 [ [[T7]], [[FOR_INC]] ], [ [[TMP19]], [[MIDDLE_BLOCK]] ] -; CHECK-NEXT: ret i32 [[T8]] +; CHECK-NEXT: ret i32 [[TMP19]] ; ; SINK-GATHER-LABEL: @predicated_sdiv_masked_load( ; SINK-GATHER-NEXT: entry: @@ -174,8 +152,7 @@ define i32 @predicated_sdiv_masked_load(ptr %a, ptr %b, i32 %x, i1 %c) { ; SINK-GATHER-NEXT: br i1 [[TMP48]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; SINK-GATHER: middle.block: ; SINK-GATHER-NEXT: [[TMP49:%.*]] = call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> [[TMP47]]) -; SINK-GATHER-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 10000 -; SINK-GATHER-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; SINK-GATHER-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; SINK-GATHER: scalar.ph: ; SINK-GATHER-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 10000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; SINK-GATHER-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP49]], [[MIDDLE_BLOCK]] ] @@ -197,7 +174,7 @@ define i32 @predicated_sdiv_masked_load(ptr %a, ptr %b, i32 %x, i1 %c) { ; SINK-GATHER-NEXT: [[T7]] = add i32 [[R]], [[T6]] ; SINK-GATHER-NEXT: [[I_NEXT]] = add nuw nsw i64 [[I]], 1 ; SINK-GATHER-NEXT: [[COND:%.*]] = icmp eq i64 [[I_NEXT]], 10000 -; SINK-GATHER-NEXT: br i1 [[COND]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP2:![0-9]+]] +; SINK-GATHER-NEXT: br i1 [[COND]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] ; SINK-GATHER: for.end: ; SINK-GATHER-NEXT: [[T8:%.*]] = phi i32 [ [[T7]], [[FOR_INC]] ], [ [[TMP49]], [[MIDDLE_BLOCK]] ] ; SINK-GATHER-NEXT: ret i32 [[T8]] @@ -283,7 +260,7 @@ define i32 @scalarize_and_sink_gather(ptr %a, i1 %c, i32 %x, i64 %n) { ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; CHECK-NEXT: [[VEC_IND_NEXT]] = add <2 x i64> [[VEC_IND]], <i64 2, i64 2> ; CHECK-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] -; CHECK-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] +; CHECK-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP20:%.*]] = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> [[TMP18]]) ; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[SMAX]], [[N_VEC]] @@ -307,7 +284,7 @@ define i32 @scalarize_and_sink_gather(ptr %a, i1 %c, i32 %x, i64 %n) { ; CHECK-NEXT: [[T6]] = add i32 [[R]], [[T5]] ; CHECK-NEXT: [[I_NEXT]] = add nuw nsw i64 [[I]], 1 ; CHECK-NEXT: [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]] -; CHECK-NEXT: br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP5:![0-9]+]] +; CHECK-NEXT: br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: for.end: ; CHECK-NEXT: [[T7:%.*]] = phi i32 [ [[T6]], [[FOR_INC]] ], [ [[TMP20]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: ret i32 [[T7]] diff --git a/llvm/test/Transforms/LoopVectorize/create-induction-resume.ll b/llvm/test/Transforms/LoopVectorize/create-induction-resume.ll index 656a870..d92fbe3 100644 --- a/llvm/test/Transforms/LoopVectorize/create-induction-resume.ll +++ b/llvm/test/Transforms/LoopVectorize/create-induction-resume.ll @@ -53,8 +53,7 @@ define void @test(i32 %arg, i32 %L1.limit, i32 %L2.switch, i1 %c) { ; CHECK-NEXT: [[TMP4:%.*]] = icmp eq i64 [[INDEX_NEXT]], 12 ; CHECK-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 12, 12 -; CHECK-NEXT: br i1 [[CMP_N]], label [[L2_HEADER_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[L2_HEADER_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 1, [[L2_INNER_HEADER_PREHEADER]] ] ; CHECK-NEXT: [[BC_RESUME_VAL2:%.*]] = phi i64 [ 13, [[MIDDLE_BLOCK]] ], [ 1, [[L2_INNER_HEADER_PREHEADER]] ] diff --git a/llvm/test/Transforms/LoopVectorize/dont-fold-tail-for-const-TC.ll b/llvm/test/Transforms/LoopVectorize/dont-fold-tail-for-const-TC.ll index cfdb2c8..23945f9 100644 --- a/llvm/test/Transforms/LoopVectorize/dont-fold-tail-for-const-TC.ll +++ b/llvm/test/Transforms/LoopVectorize/dont-fold-tail-for-const-TC.ll @@ -30,8 +30,7 @@ define dso_local void @constTC(ptr noalias nocapture %A) optsize { ; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1800 ; CHECK-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 1800, 1800 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 1800, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/epilog-vectorization-reductions.ll b/llvm/test/Transforms/LoopVectorize/epilog-vectorization-reductions.ll index 02c6475..7a3c7d6 100644 --- a/llvm/test/Transforms/LoopVectorize/epilog-vectorization-reductions.ll +++ b/llvm/test/Transforms/LoopVectorize/epilog-vectorization-reductions.ll @@ -216,8 +216,7 @@ define i16 @reduction_or_trunc(ptr noalias nocapture %ptr) { ; CHECK-NEXT: [[TMP9:%.*]] = trunc <4 x i32> [[TMP8]] to <4 x i16> ; CHECK-NEXT: [[TMP10:%.*]] = call i16 @llvm.vector.reduce.or.v4i16(<4 x i16> [[TMP9]]) ; CHECK-NEXT: [[TMP11:%.*]] = zext i16 [[TMP10]] to i32 -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 256, 256 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-NEXT: br i1 true, label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK: vec.epilog.iter.check: ; CHECK-NEXT: br i1 true, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] ; CHECK: vec.epilog.ph: @@ -244,8 +243,7 @@ define i16 @reduction_or_trunc(ptr noalias nocapture %ptr) { ; CHECK-NEXT: [[TMP22:%.*]] = trunc <4 x i32> [[TMP21]] to <4 x i16> ; CHECK-NEXT: [[TMP23:%.*]] = call i16 @llvm.vector.reduce.or.v4i16(<4 x i16> [[TMP22]]) ; CHECK-NEXT: [[TMP24:%.*]] = zext i16 [[TMP23]] to i32 -; CHECK-NEXT: [[CMP_N1:%.*]] = icmp eq i32 256, 256 -; CHECK-NEXT: br i1 [[CMP_N1]], label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 256, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 256, [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX6:%.*]] = phi i32 [ 0, [[ITER_CHECK]] ], [ [[TMP11]], [[VEC_EPILOG_ITER_CHECK]] ], [ [[TMP24]], [[VEC_EPILOG_MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/first-order-recurrence-chains.ll b/llvm/test/Transforms/LoopVectorize/first-order-recurrence-chains.ll index 371dc7f..c663d2b 100644 --- a/llvm/test/Transforms/LoopVectorize/first-order-recurrence-chains.ll +++ b/llvm/test/Transforms/LoopVectorize/first-order-recurrence-chains.ll @@ -18,7 +18,6 @@ define i16 @test_chained_first_order_recurrences_1(ptr %ptr) { ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP8]], label %middle.block, label %vector.body ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT2:%.*]] = extractelement <4 x i16> [[TMP4]], i32 3 @@ -62,7 +61,6 @@ define i16 @test_chained_first_order_recurrences_2(ptr %ptr) { ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP8]], label %middle.block, label %vector.body, !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[TMP4]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i16> [[TMP4]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT2:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 3 @@ -109,7 +107,6 @@ define i16 @test_chained_first_order_recurrences_3(ptr %ptr) { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP10]], label %middle.block, label %vector.body, !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT3:%.*]] = extractelement <4 x i16> [[TMP4]], i32 3 @@ -222,7 +219,6 @@ define i16 @test_chained_first_order_recurrences_3_reordered_1(ptr %ptr) { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP10]], label %middle.block, label %vector.body, !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT7:%.*]] = extractelement <4 x i16> [[TMP5]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI8:%.*]] = extractelement <4 x i16> [[TMP5]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT3:%.*]] = extractelement <4 x i16> [[TMP4]], i32 3 @@ -274,7 +270,6 @@ define i16 @test_chained_first_order_recurrences_3_reordered_2(ptr %ptr) { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP10]], label %middle.block, label %vector.body, !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT3:%.*]] = extractelement <4 x i16> [[TMP4]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI4:%.*]] = extractelement <4 x i16> [[TMP4]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT7:%.*]] = extractelement <4 x i16> [[TMP5]], i32 3 @@ -326,7 +321,6 @@ define i16 @test_chained_first_order_recurrences_3_for2_no_other_uses(ptr %ptr) ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP10]], label %middle.block, label %vector.body, !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT3:%.*]] = extractelement <4 x i16> [[TMP4]], i32 3 @@ -377,7 +371,6 @@ define i16 @test_chained_first_order_recurrences_3_for1_for2_no_other_uses(ptr % ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP10]], label %middle.block, label %vector.body, !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT3:%.*]] = extractelement <4 x i16> [[TMP4]], i32 3 @@ -427,7 +420,6 @@ define double @test_chained_first_order_recurrence_sink_users_1(ptr %ptr) { ; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i64 [[INDEX_NEXT]], 996 ; CHECK-NEXT: br i1 [[TMP9]], label %middle.block, label %vector.body, !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 996 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x double> [[WIDE_LOAD]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x double> [[WIDE_LOAD]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT2:%.*]] = extractelement <4 x double> [[TMP4]], i32 3 @@ -496,10 +488,9 @@ define i64 @test_first_order_recurrences_and_induction(ptr %ptr) { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP5]], label %middle.block, label %vector.body ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i64> [[VEC_IND]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i64> [[VEC_IND]], i32 2 -; CHECK-NEXT: br i1 [[CMP_N]] +; CHECK-NEXT: br i1 true entry: br label %loop @@ -537,10 +528,9 @@ define i64 @test_first_order_recurrences_and_induction2(ptr %ptr) { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP5]], label %middle.block, label %vector.body ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i64> [[VEC_IND]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i64> [[VEC_IND]], i32 2 -; CHECK-NEXT: br i1 [[CMP_N]] +; CHECK-NEXT: br i1 true ; entry: br label %loop @@ -579,10 +569,9 @@ define ptr @test_first_order_recurrences_and_pointer_induction1(ptr %ptr) { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP5]], label %middle.block, label %vector.body ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x ptr> [[TMP0]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x ptr> [[TMP0]], i32 2 -; CHECK-NEXT: br i1 [[CMP_N]], +; CHECK-NEXT: br i1 true ; entry: br label %loop @@ -624,10 +613,9 @@ define ptr @test_first_order_recurrences_and_pointer_induction2(ptr %ptr) { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP5]], label %middle.block, label %vector.body ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x ptr> [[TMP0]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x ptr> [[TMP0]], i32 2 -; CHECK-NEXT: br i1 [[CMP_N]], +; CHECK-NEXT: br i1 true ; entry: br label %loop @@ -672,14 +660,13 @@ define double @test_resinking_required(ptr %p, ptr noalias %a, ptr noalias %b) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 0 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP28:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 0, 0 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x double> [[BROADCAST_SPLAT]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x double> [[BROADCAST_SPLAT]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT5:%.*]] = extractelement <4 x double> [[BROADCAST_SPLAT4]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI6:%.*]] = extractelement <4 x double> [[BROADCAST_SPLAT4]], i32 2 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT9:%.*]] = extractelement <4 x double> [[TMP4]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI10:%.*]] = extractelement <4 x double> [[TMP4]], i32 2 -; CHECK-NEXT: br i1 [[CMP_N]], label %End, label %scalar.ph +; CHECK-NEXT: br i1 true, label %End, label %scalar.ph ; Entry: br label %Loop diff --git a/llvm/test/Transforms/LoopVectorize/first-order-recurrence-complex.ll b/llvm/test/Transforms/LoopVectorize/first-order-recurrence-complex.ll index daf37a3..9567107 100644 --- a/llvm/test/Transforms/LoopVectorize/first-order-recurrence-complex.ll +++ b/llvm/test/Transforms/LoopVectorize/first-order-recurrence-complex.ll @@ -38,9 +38,8 @@ define void @can_sink_after_store(i32 %x, ptr %ptr, i64 %tc) local_unnamed_addr ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1996 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1999, 1996 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i32> [[WIDE_LOAD]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ [[DOTPRE]], [[PREHEADER]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1997, [[MIDDLE_BLOCK]] ], [ 1, [[PREHEADER]] ] @@ -119,9 +118,8 @@ define void @sink_sdiv(i32 %x, ptr %ptr, i64 %tc) local_unnamed_addr #0 { ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1996 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1999, 1996 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i32> [[WIDE_LOAD]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ [[DOTPRE]], [[PREHEADER]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1997, [[MIDDLE_BLOCK]] ], [ 1, [[PREHEADER]] ] @@ -201,9 +199,8 @@ define void @can_sink_with_additional_user(i32 %x, ptr %ptr, i64 %tc) { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1996 ; CHECK-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1999, 1996 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i32> [[WIDE_LOAD]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ [[DOTPRE]], [[PREHEADER]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1997, [[MIDDLE_BLOCK]] ], [ 1, [[PREHEADER]] ] @@ -392,10 +389,9 @@ define void @instruction_with_2_FOR_operands(ptr noalias %A, ptr noalias %B, ptr ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1001, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x float> [[BROADCAST_SPLAT3]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT4:%.*]] = extractelement <4 x float> [[BROADCAST_SPLAT]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[BB74:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[BB74:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT5:%.*]] = phi float [ 1.000000e+00, [[BB:%.*]] ], [ [[VECTOR_RECUR_EXTRACT4]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi float [ 0.000000e+00, [[BB]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] @@ -471,10 +467,9 @@ define void @instruction_with_2_FOR_operands_and_multiple_other_uses(ptr noalias ; CHECK-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1001, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x float> [[BROADCAST_SPLAT3]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT4:%.*]] = extractelement <4 x float> [[BROADCAST_SPLAT]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT5:%.*]] = phi float [ 0.000000e+00, [[BB:%.*]] ], [ [[VECTOR_RECUR_EXTRACT4]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi float [ 0.000000e+00, [[BB]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] @@ -565,10 +560,9 @@ define void @instruction_with_2_FOR_operands_and_multiple_other_uses_chain(ptr n ; CHECK-NEXT: [[TMP15:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1001, 1000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x float> [[BROADCAST_SPLAT3]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT4:%.*]] = extractelement <4 x float> [[BROADCAST_SPLAT]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT5:%.*]] = phi float [ 0.000000e+00, [[BB:%.*]] ], [ [[VECTOR_RECUR_EXTRACT4]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi float [ 0.000000e+00, [[BB]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] @@ -1069,10 +1063,9 @@ define void @test_for_sink_instruction_after_same_incoming_1(ptr %ptr) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 996 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 996 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x double> [[WIDE_LOAD]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT2:%.*]] = extractelement <4 x double> [[WIDE_LOAD]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT3:%.*]] = phi double [ 2.000000e+01, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT2]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi double [ 1.000000e+01, [[ENTRY]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] @@ -1138,10 +1131,9 @@ define void @test_for_sink_instruction_after_same_incoming_2(ptr %ptr) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 996 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 996 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x double> [[WIDE_LOAD]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT2:%.*]] = extractelement <4 x double> [[WIDE_LOAD]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT3:%.*]] = phi double [ 1.000000e+01, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT2]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi double [ 2.000000e+01, [[ENTRY]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/first-order-recurrence-multiply-recurrences.ll b/llvm/test/Transforms/LoopVectorize/first-order-recurrence-multiply-recurrences.ll index 3d07e2c..937e01e 100644 --- a/llvm/test/Transforms/LoopVectorize/first-order-recurrence-multiply-recurrences.ll +++ b/llvm/test/Transforms/LoopVectorize/first-order-recurrence-multiply-recurrences.ll @@ -121,10 +121,9 @@ define void @test_pr54223_sink_after_insertion_order(ptr noalias %a, ptr noalias ; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i64 [[INDEX_NEXT]], 10000 ; CHECK-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 10000, 10000 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x float> [[BROADCAST_SPLAT]], i32 3 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT4:%.*]] = extractelement <4 x float> [[BROADCAST_SPLAT3]], i32 3 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT5:%.*]] = phi float [ 0.000000e+00, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT4]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi float [ 0.000000e+00, [[ENTRY]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/first-order-recurrence.ll b/llvm/test/Transforms/LoopVectorize/first-order-recurrence.ll index 7dd8fc7..7b242f5 100644 --- a/llvm/test/Transforms/LoopVectorize/first-order-recurrence.ll +++ b/llvm/test/Transforms/LoopVectorize/first-order-recurrence.ll @@ -1285,8 +1285,7 @@ define i64 @constant_folded_previous_value() { ; UNROLL-NO-IC-NEXT: [[TMP1:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; UNROLL-NO-IC-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; UNROLL-NO-IC: middle.block: -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ 1, [[MIDDLE_BLOCK]] ] ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1316,8 +1315,7 @@ define i64 @constant_folded_previous_value() { ; UNROLL-NO-VF-NEXT: [[TMP2:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; UNROLL-NO-VF-NEXT: br i1 [[TMP2]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; UNROLL-NO-VF: middle.block: -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[TMP1]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-VF-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1346,8 +1344,7 @@ define i64 @constant_folded_previous_value() { ; SINK-AFTER-NEXT: [[TMP1:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; SINK-AFTER-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; SINK-AFTER: middle.block: -; SINK-AFTER-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; SINK-AFTER-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; SINK-AFTER-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; SINK-AFTER: scalar.ph: ; SINK-AFTER-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ 1, [[MIDDLE_BLOCK]] ] ; SINK-AFTER-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1406,10 +1403,9 @@ define i32 @extract_second_last_iteration(ptr %cval, i32 %x) { ; UNROLL-NO-IC-NEXT: [[TMP4:%.*]] = icmp eq i32 [[INDEX_NEXT]], 96 ; UNROLL-NO-IC-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; UNROLL-NO-IC: middle.block: -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i32 96, 96 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i32> [[TMP1]], i32 3 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i32> [[TMP1]], i32 2 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 96, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1442,8 +1438,7 @@ define i32 @extract_second_last_iteration(ptr %cval, i32 %x) { ; UNROLL-NO-VF-NEXT: [[TMP4:%.*]] = icmp eq i32 [[INDEX_NEXT]], 96 ; UNROLL-NO-VF-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; UNROLL-NO-VF: middle.block: -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i32 96, 96 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[TMP3]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-VF-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 96, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1478,10 +1473,9 @@ define i32 @extract_second_last_iteration(ptr %cval, i32 %x) { ; SINK-AFTER-NEXT: [[TMP2:%.*]] = icmp eq i32 [[INDEX_NEXT]], 96 ; SINK-AFTER-NEXT: br i1 [[TMP2]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; SINK-AFTER: middle.block: -; SINK-AFTER-NEXT: [[CMP_N:%.*]] = icmp eq i32 96, 96 ; SINK-AFTER-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i32> [[TMP0]], i32 3 ; SINK-AFTER-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i32> [[TMP0]], i32 2 -; SINK-AFTER-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; SINK-AFTER-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; SINK-AFTER: scalar.ph: ; SINK-AFTER-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; SINK-AFTER-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 96, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1610,9 +1604,8 @@ define i32 @PR33613(ptr %b, double %j, i32 %d) { ; UNROLL-NO-IC: middle.block: ; UNROLL-NO-IC-NEXT: [[BIN_RDX:%.*]] = add <4 x i32> [[TMP49]], [[TMP48]] ; UNROLL-NO-IC-NEXT: [[TMP51:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX]]) -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i64 10240, 10240 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x double> [[TMP39]], i32 3 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi double [ [[J]], [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[B]], [[ENTRY]] ] @@ -1673,8 +1666,7 @@ define i32 @PR33613(ptr %b, double %j, i32 %d) { ; UNROLL-NO-VF-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; UNROLL-NO-VF: middle.block: ; UNROLL-NO-VF-NEXT: [[BIN_RDX:%.*]] = add i32 [[TMP15]], [[TMP14]] -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i64 10240, 10240 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi double [ [[J]], [[ENTRY:%.*]] ], [ [[TMP7]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-VF-NEXT: [[BC_RESUME_VAL:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[B]], [[ENTRY]] ] @@ -1746,9 +1738,8 @@ define i32 @PR33613(ptr %b, double %j, i32 %d) { ; SINK-AFTER-NEXT: br i1 [[TMP25]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; SINK-AFTER: middle.block: ; SINK-AFTER-NEXT: [[TMP26:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP24]]) -; SINK-AFTER-NEXT: [[CMP_N:%.*]] = icmp eq i64 10240, 10240 ; SINK-AFTER-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x double> [[TMP19]], i32 3 -; SINK-AFTER-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; SINK-AFTER-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; SINK-AFTER: scalar.ph: ; SINK-AFTER-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi double [ [[J]], [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; SINK-AFTER-NEXT: [[BC_RESUME_VAL:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[B]], [[ENTRY]] ] @@ -2618,10 +2609,9 @@ define void @sink_dead_inst(ptr %a) { ; UNROLL-NO-IC-NEXT: [[TMP18:%.*]] = icmp eq i32 [[INDEX_NEXT]], 40 ; UNROLL-NO-IC-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]] ; UNROLL-NO-IC: middle.block: -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i32 43, 40 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[TMP9]], i32 3 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT3:%.*]] = extractelement <4 x i32> [[TMP5]], i32 3 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT4:%.*]] = phi i32 [ -27, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT3]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i16 [ 0, [[ENTRY]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] @@ -2671,8 +2661,7 @@ define void @sink_dead_inst(ptr %a) { ; UNROLL-NO-VF-NEXT: [[TMP12:%.*]] = icmp eq i32 [[INDEX_NEXT]], 42 ; UNROLL-NO-VF-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]] ; UNROLL-NO-VF: middle.block: -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i32 43, 42 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[SCALAR_RECUR_INIT2:%.*]] = phi i32 [ -27, [[ENTRY:%.*]] ], [ [[TMP5]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-VF-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i16 [ 0, [[ENTRY]] ], [ [[TMP7]], [[MIDDLE_BLOCK]] ] @@ -2720,10 +2709,9 @@ define void @sink_dead_inst(ptr %a) { ; SINK-AFTER-NEXT: [[TMP9:%.*]] = icmp eq i32 [[INDEX_NEXT]], 40 ; SINK-AFTER-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]] ; SINK-AFTER: middle.block: -; SINK-AFTER-NEXT: [[CMP_N:%.*]] = icmp eq i32 43, 40 ; SINK-AFTER-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[TMP4]], i32 3 ; SINK-AFTER-NEXT: [[VECTOR_RECUR_EXTRACT2:%.*]] = extractelement <4 x i32> [[TMP2]], i32 3 -; SINK-AFTER-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; SINK-AFTER-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; SINK-AFTER: scalar.ph: ; SINK-AFTER-NEXT: [[SCALAR_RECUR_INIT3:%.*]] = phi i32 [ -27, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT2]], [[MIDDLE_BLOCK]] ] ; SINK-AFTER-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i16 [ 0, [[ENTRY]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] @@ -3534,10 +3522,9 @@ define i32 @sink_after_dead_inst(ptr %A.ptr) { ; UNROLL-NO-IC-NEXT: [[TMP14:%.*]] = icmp eq i32 [[INDEX_NEXT]], 16 ; UNROLL-NO-IC-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; UNROLL-NO-IC: middle.block: -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i32 16, 16 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i32> [[TMP7]], i32 3 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i32> [[TMP7]], i32 2 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 16, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -3585,8 +3572,7 @@ define i32 @sink_after_dead_inst(ptr %A.ptr) { ; UNROLL-NO-VF-NEXT: [[TMP10:%.*]] = icmp eq i32 [[INDEX_NEXT]], 16 ; UNROLL-NO-VF-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; UNROLL-NO-VF: middle.block: -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i32 16, 16 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[TMP7]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-VF-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 16, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -3632,10 +3618,9 @@ define i32 @sink_after_dead_inst(ptr %A.ptr) { ; SINK-AFTER-NEXT: [[TMP7:%.*]] = icmp eq i32 [[INDEX_NEXT]], 16 ; SINK-AFTER-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; SINK-AFTER: middle.block: -; SINK-AFTER-NEXT: [[CMP_N:%.*]] = icmp eq i32 16, 16 ; SINK-AFTER-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i32> [[TMP3]], i32 3 ; SINK-AFTER-NEXT: [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i32> [[TMP3]], i32 2 -; SINK-AFTER-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; SINK-AFTER-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; SINK-AFTER: scalar.ph: ; SINK-AFTER-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; SINK-AFTER-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 16, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -3705,9 +3690,8 @@ define void @unused_recurrence(ptr %a) { ; UNROLL-NO-IC-NEXT: [[TMP6:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1024 ; UNROLL-NO-IC-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP34:![0-9]+]] ; UNROLL-NO-IC: middle.block: -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i32 1028, 1024 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[TMP3]], i32 3 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i16 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 997, [[MIDDLE_BLOCK]] ], [ -27, [[ENTRY]] ] @@ -3743,8 +3727,7 @@ define void @unused_recurrence(ptr %a) { ; UNROLL-NO-VF-NEXT: [[TMP6:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1028 ; UNROLL-NO-VF-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP34:![0-9]+]] ; UNROLL-NO-VF: middle.block: -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i32 1028, 1028 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i16 [ 0, [[ENTRY:%.*]] ], [ [[TMP5]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-VF-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 1001, [[MIDDLE_BLOCK]] ], [ -27, [[ENTRY]] ] @@ -3777,9 +3760,8 @@ define void @unused_recurrence(ptr %a) { ; SINK-AFTER-NEXT: [[TMP3:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1028 ; SINK-AFTER-NEXT: br i1 [[TMP3]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP34:![0-9]+]] ; SINK-AFTER: middle.block: -; SINK-AFTER-NEXT: [[CMP_N:%.*]] = icmp eq i32 1028, 1028 ; SINK-AFTER-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[TMP1]], i32 3 -; SINK-AFTER-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; SINK-AFTER-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; SINK-AFTER: scalar.ph: ; SINK-AFTER-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i16 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; SINK-AFTER-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 1001, [[MIDDLE_BLOCK]] ], [ -27, [[ENTRY]] ] diff --git a/llvm/test/Transforms/LoopVectorize/float-minmax-instruction-flag.ll b/llvm/test/Transforms/LoopVectorize/float-minmax-instruction-flag.ll index ac60eb5..01ff00d 100644 --- a/llvm/test/Transforms/LoopVectorize/float-minmax-instruction-flag.ll +++ b/llvm/test/Transforms/LoopVectorize/float-minmax-instruction-flag.ll @@ -68,8 +68,7 @@ define float @minloopattr(ptr nocapture readonly %arg) #0 { ; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP6:%.*]] = call float @llvm.vector.reduce.fmin.v4f32(<4 x float> [[TMP4]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 65536, 65536 -; CHECK-NEXT: br i1 [[CMP_N]], label [[OUT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[OUT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 65537, [[MIDDLE_BLOCK]] ], [ 1, [[TOP:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi float [ [[T]], [[TOP]] ], [ [[TMP6]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/if-pred-non-void.ll b/llvm/test/Transforms/LoopVectorize/if-pred-non-void.ll index 51c5885..e7e7c58 100644 --- a/llvm/test/Transforms/LoopVectorize/if-pred-non-void.ll +++ b/llvm/test/Transforms/LoopVectorize/if-pred-non-void.ll @@ -37,7 +37,7 @@ define void @test(ptr nocapture %asd, ptr nocapture %aud, ; CHECK-NEXT: [[BOUND120:%.*]] = icmp ult ptr [[AUR]], [[SCEVGEP2]] ; CHECK-NEXT: [[FOUND_CONFLICT21:%.*]] = and i1 [[BOUND019]], [[BOUND120]] ; CHECK-NEXT: [[CONFLICT_RDX22:%.*]] = or i1 [[CONFLICT_RDX18]], [[FOUND_CONFLICT21]] -; CHECK-NEXT: br i1 [[CONFLICT_RDX22]], label [[SCALAR_PH:%.*]], label [[VECTOR_BODY:%.*]] +; CHECK-NEXT: br i1 [[CONFLICT_RDX22]], label [[FOR_BODY:%.*]], label [[VECTOR_BODY:%.*]] ; CHECK: vector.body: ; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ [[INDEX_NEXT:%.*]], [[PRED_UREM_CONTINUE27:%.*]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0 @@ -123,17 +123,11 @@ define void @test(ptr nocapture %asd, ptr nocapture %aud, ; CHECK-NEXT: store <2 x i32> [[PREDPHI30]], ptr [[TMP60]], align 4, !alias.scope !15 ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; CHECK-NEXT: [[TMP61:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 -; CHECK-NEXT: br i1 [[TMP61]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] -; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] -; CHECK: scalar.ph: -; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 128, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] -; CHECK-NEXT: br label [[FOR_BODY:%.*]] +; CHECK-NEXT: br i1 [[TMP61]], label [[FOR_COND_CLEANUP:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; CHECK: for.cond.cleanup: ; CHECK-NEXT: ret void ; CHECK: for.body: -; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[IF_END:%.*]] ] +; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[IF_END:%.*]] ], [ 0, [[ENTRY]] ] ; CHECK-NEXT: [[ISD:%.*]] = getelementptr inbounds i32, ptr [[ASD]], i64 [[INDVARS_IV]] ; CHECK-NEXT: [[IUD:%.*]] = getelementptr inbounds i32, ptr [[AUD]], i64 [[INDVARS_IV]] ; CHECK-NEXT: [[ISR:%.*]] = getelementptr inbounds i32, ptr [[ASR]], i64 [[INDVARS_IV]] @@ -277,8 +271,7 @@ define void @test(ptr nocapture %asd, ptr nocapture %aud, ; UNROLL-NO-VF-NEXT: [[TMP46:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 ; UNROLL-NO-VF-NEXT: br i1 [[TMP46]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; UNROLL-NO-VF: middle.block: -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 128, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; UNROLL-NO-VF-NEXT: br label [[FOR_BODY:%.*]] @@ -372,7 +365,7 @@ define void @test_scalar2scalar(ptr nocapture %asd, ptr nocapture %bsd) { ; CHECK-NEXT: [[BOUND0:%.*]] = icmp ult ptr [[ASD]], [[SCEVGEP1]] ; CHECK-NEXT: [[BOUND1:%.*]] = icmp ult ptr [[BSD]], [[SCEVGEP]] ; CHECK-NEXT: [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]] -; CHECK-NEXT: br i1 [[FOUND_CONFLICT]], label [[SCALAR_PH:%.*]], label [[VECTOR_BODY:%.*]] +; CHECK-NEXT: br i1 [[FOUND_CONFLICT]], label [[FOR_BODY:%.*]], label [[VECTOR_BODY:%.*]] ; CHECK: vector.body: ; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ [[INDEX_NEXT:%.*]], [[PRED_SDIV_CONTINUE4:%.*]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0 @@ -416,17 +409,11 @@ define void @test_scalar2scalar(ptr nocapture %asd, ptr nocapture %bsd) { ; CHECK-NEXT: store <2 x i32> [[PREDPHI]], ptr [[TMP26]], align 4, !alias.scope !20, !noalias !23 ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; CHECK-NEXT: [[TMP27:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 -; CHECK-NEXT: br i1 [[TMP27]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP25:![0-9]+]] -; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] -; CHECK: scalar.ph: -; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 128, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] -; CHECK-NEXT: br label [[FOR_BODY:%.*]] +; CHECK-NEXT: br i1 [[TMP27]], label [[FOR_COND_CLEANUP:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP25:![0-9]+]] ; CHECK: for.cond.cleanup: ; CHECK-NEXT: ret void ; CHECK: for.body: -; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[IF_END:%.*]] ] +; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[IF_END:%.*]] ], [ 0, [[ENTRY]] ] ; CHECK-NEXT: [[ISD:%.*]] = getelementptr inbounds i32, ptr [[ASD]], i64 [[INDVARS_IV]] ; CHECK-NEXT: [[LSD:%.*]] = load i32, ptr [[ISD]], align 4 ; CHECK-NEXT: [[ISD_B:%.*]] = getelementptr inbounds i32, ptr [[BSD]], i64 [[INDVARS_IV]] @@ -499,8 +486,7 @@ define void @test_scalar2scalar(ptr nocapture %asd, ptr nocapture %bsd) { ; UNROLL-NO-VF-NEXT: [[TMP24:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 ; UNROLL-NO-VF-NEXT: br i1 [[TMP24]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP25:![0-9]+]] ; UNROLL-NO-VF: middle.block: -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 128, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; UNROLL-NO-VF-NEXT: br label [[FOR_BODY:%.*]] @@ -564,7 +550,7 @@ define void @pr30172(ptr nocapture %asd, ptr nocapture %bsd) !dbg !5 {; ; CHECK-NEXT: [[BOUND0:%.*]] = icmp ult ptr [[ASD]], [[SCEVGEP1]] ; CHECK-NEXT: [[BOUND1:%.*]] = icmp ult ptr [[BSD]], [[SCEVGEP]] ; CHECK-NEXT: [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]] -; CHECK-NEXT: br i1 [[FOUND_CONFLICT]], label [[SCALAR_PH:%.*]], label [[VECTOR_BODY:%.*]] +; CHECK-NEXT: br i1 [[FOUND_CONFLICT]], label [[FOR_BODY:%.*]], label [[VECTOR_BODY:%.*]] ; CHECK: vector.body: ; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ [[INDEX_NEXT:%.*]], [[PRED_SDIV_CONTINUE4:%.*]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0 @@ -613,17 +599,11 @@ define void @pr30172(ptr nocapture %asd, ptr nocapture %bsd) !dbg !5 {; ; CHECK-NEXT: store <2 x i32> [[PREDPHI]], ptr [[TMP31]], align 4, !alias.scope !29, !noalias !32 ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; CHECK-NEXT: [[TMP32:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 -; CHECK-NEXT: br i1 [[TMP32]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP36:![0-9]+]] -; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] -; CHECK: scalar.ph: -; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 128, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] -; CHECK-NEXT: br label [[FOR_BODY:%.*]] +; CHECK-NEXT: br i1 [[TMP32]], label [[FOR_COND_CLEANUP:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP36:![0-9]+]] ; CHECK: for.cond.cleanup: ; CHECK-NEXT: ret void ; CHECK: for.body: -; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[IF_END:%.*]] ] +; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[IF_END:%.*]] ], [ 0, [[ENTRY]] ] ; CHECK-NEXT: [[ISD:%.*]] = getelementptr inbounds i32, ptr [[ASD]], i64 [[INDVARS_IV]] ; CHECK-NEXT: [[LSD:%.*]] = load i32, ptr [[ISD]], align 4 ; CHECK-NEXT: [[ISD_B:%.*]] = getelementptr inbounds i32, ptr [[BSD]], i64 [[INDVARS_IV]] @@ -708,8 +688,7 @@ define void @pr30172(ptr nocapture %asd, ptr nocapture %bsd) !dbg !5 {; ; UNROLL-NO-VF-NEXT: [[TMP34:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 ; UNROLL-NO-VF-NEXT: br i1 [[TMP34]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP36:![0-9]+]] ; UNROLL-NO-VF: middle.block: -; UNROLL-NO-VF-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; UNROLL-NO-VF-NEXT: br i1 [[CMP_N]], label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-VF-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-VF: scalar.ph: ; UNROLL-NO-VF-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 128, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; UNROLL-NO-VF-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/if-pred-stores.ll b/llvm/test/Transforms/LoopVectorize/if-pred-stores.ll index d967b7f..926168c 100644 --- a/llvm/test/Transforms/LoopVectorize/if-pred-stores.ll +++ b/llvm/test/Transforms/LoopVectorize/if-pred-stores.ll @@ -34,24 +34,7 @@ define i32 @test(ptr nocapture %f) #0 { ; UNROLL: pred.store.continue2: ; UNROLL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; UNROLL-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 -; UNROLL-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] -; UNROLL: middle.block: -; UNROLL-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; UNROLL-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; UNROLL: for.body: -; UNROLL-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 128, [[MIDDLE_BLOCK]] ] -; UNROLL-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[F]], i64 [[INDVARS_IV]] -; UNROLL-NEXT: [[TMP11:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; UNROLL-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP11]], 100 -; UNROLL-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; UNROLL: if.then: -; UNROLL-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], 20 -; UNROLL-NEXT: store i32 [[ADD]], ptr [[ARRAYIDX]], align 4 -; UNROLL-NEXT: br label [[FOR_INC]] -; UNROLL: for.inc: -; UNROLL-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1 -; UNROLL-NEXT: [[EXITCOND:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT]], 128 -; UNROLL-NEXT: br i1 [[EXITCOND]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] +; UNROLL-NEXT: br i1 [[TMP10]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; UNROLL: for.end: ; UNROLL-NEXT: ret i32 0 ; @@ -86,8 +69,7 @@ define i32 @test(ptr nocapture %f) #0 { ; UNROLL-NOSIMPLIFY-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 ; UNROLL-NOSIMPLIFY-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; UNROLL-NOSIMPLIFY: middle.block: -; UNROLL-NOSIMPLIFY-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; UNROLL-NOSIMPLIFY-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NOSIMPLIFY-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NOSIMPLIFY: scalar.ph: ; UNROLL-NOSIMPLIFY-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 128, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; UNROLL-NOSIMPLIFY-NEXT: br label [[FOR_BODY:%.*]] @@ -139,24 +121,7 @@ define i32 @test(ptr nocapture %f) #0 { ; VEC: pred.store.continue2: ; VEC-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; VEC-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 128 -; VEC-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] -; VEC: middle.block: -; VEC-NEXT: [[CMP_N:%.*]] = icmp eq i64 128, 128 -; VEC-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; VEC: for.body: -; VEC-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC:%.*]] ], [ 128, [[MIDDLE_BLOCK]] ] -; VEC-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[F]], i64 [[INDVARS_IV]] -; VEC-NEXT: [[TMP14:%.*]] = load i32, ptr [[ARRAYIDX]], align 4 -; VEC-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[TMP14]], 100 -; VEC-NEXT: br i1 [[CMP1]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; VEC: if.then: -; VEC-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP14]], 20 -; VEC-NEXT: store i32 [[ADD]], ptr [[ARRAYIDX]], align 4 -; VEC-NEXT: br label [[FOR_INC]] -; VEC: for.inc: -; VEC-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1 -; VEC-NEXT: [[EXITCOND:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT]], 128 -; VEC-NEXT: br i1 [[EXITCOND]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] +; VEC-NEXT: br i1 [[TMP13]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; VEC: for.end: ; VEC-NEXT: ret i32 0 ; @@ -233,7 +198,7 @@ define void @bug18724(i1 %cond, ptr %ptr, i1 %cond.2, i64 %v.1, i32 %v.2) { ; UNROLL-NEXT: [[PREDPHI4]] = select i1 [[TMP14]], i32 [[VEC_PHI1]], i32 [[TMP12]] ; UNROLL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; UNROLL-NEXT: [[TMP15:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] -; UNROLL-NEXT: br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] +; UNROLL-NEXT: br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] ; UNROLL: middle.block: ; UNROLL-NEXT: [[BIN_RDX:%.*]] = add i32 [[PREDPHI4]], [[PREDPHI]] ; UNROLL-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP4]], [[N_VEC]] @@ -390,7 +355,7 @@ define void @bug18724(i1 %cond, ptr %ptr, i1 %cond.2, i64 %v.1, i32 %v.2) { ; VEC-NEXT: [[PREDPHI]] = select <2 x i1> [[TMP17]], <2 x i32> [[VEC_PHI]], <2 x i32> [[TMP16]] ; VEC-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; VEC-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] -; VEC-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] +; VEC-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] ; VEC: middle.block: ; VEC-NEXT: [[TMP19:%.*]] = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> [[PREDPHI]]) ; VEC-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP4]], [[N_VEC]] @@ -477,26 +442,7 @@ define void @minimal_bit_widths(i1 %c) { ; UNROLL: pred.store.continue3: ; UNROLL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; UNROLL-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 -; UNROLL-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] -; UNROLL: middle.block: -; UNROLL-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; UNROLL-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; UNROLL: for.body: -; UNROLL-NEXT: [[TMP0:%.*]] = phi i64 [ [[TMP6:%.*]], [[FOR_INC:%.*]] ], [ 1000, [[MIDDLE_BLOCK]] ] -; UNROLL-NEXT: [[TMP1:%.*]] = phi i64 [ [[TMP7:%.*]], [[FOR_INC]] ], [ 0, [[MIDDLE_BLOCK]] ] -; UNROLL-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr undef, i64 [[TMP0]] -; UNROLL-NEXT: [[TMP3:%.*]] = load i8, ptr [[TMP2]], align 1 -; UNROLL-NEXT: br i1 [[C]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; UNROLL: if.then: -; UNROLL-NEXT: [[TMP4:%.*]] = zext i8 [[TMP3]] to i32 -; UNROLL-NEXT: [[TMP5:%.*]] = trunc i32 [[TMP4]] to i8 -; UNROLL-NEXT: store i8 [[TMP5]], ptr [[TMP2]], align 1 -; UNROLL-NEXT: br label [[FOR_INC]] -; UNROLL: for.inc: -; UNROLL-NEXT: [[TMP6]] = add nuw nsw i64 [[TMP0]], 1 -; UNROLL-NEXT: [[TMP7]] = add i64 [[TMP1]], -1 -; UNROLL-NEXT: [[TMP8:%.*]] = icmp eq i64 [[TMP7]], 0 -; UNROLL-NEXT: br i1 [[TMP8]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] +; UNROLL-NEXT: br i1 [[TMP10]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; UNROLL: for.end: ; UNROLL-NEXT: ret void ; @@ -531,8 +477,7 @@ define void @minimal_bit_widths(i1 %c) { ; UNROLL-NOSIMPLIFY-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; UNROLL-NOSIMPLIFY-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; UNROLL-NOSIMPLIFY: middle.block: -; UNROLL-NOSIMPLIFY-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; UNROLL-NOSIMPLIFY-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NOSIMPLIFY-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NOSIMPLIFY: scalar.ph: ; UNROLL-NOSIMPLIFY-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; UNROLL-NOSIMPLIFY-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 0, [[MIDDLE_BLOCK]] ], [ 1000, [[ENTRY]] ] @@ -590,26 +535,7 @@ define void @minimal_bit_widths(i1 %c) { ; VEC: pred.store.continue3: ; VEC-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 ; VEC-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 -; VEC-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] -; VEC: middle.block: -; VEC-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VEC-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[FOR_BODY:%.*]] -; VEC: for.body: -; VEC-NEXT: [[TMP0:%.*]] = phi i64 [ [[TMP6:%.*]], [[FOR_INC:%.*]] ], [ 1000, [[MIDDLE_BLOCK]] ] -; VEC-NEXT: [[TMP1:%.*]] = phi i64 [ [[TMP7:%.*]], [[FOR_INC]] ], [ 0, [[MIDDLE_BLOCK]] ] -; VEC-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr undef, i64 [[TMP0]] -; VEC-NEXT: [[TMP3:%.*]] = load i8, ptr [[TMP2]], align 1 -; VEC-NEXT: br i1 [[C]], label [[IF_THEN:%.*]], label [[FOR_INC]] -; VEC: if.then: -; VEC-NEXT: [[TMP4:%.*]] = zext i8 [[TMP3]] to i32 -; VEC-NEXT: [[TMP5:%.*]] = trunc i32 [[TMP4]] to i8 -; VEC-NEXT: store i8 [[TMP5]], ptr [[TMP2]], align 1 -; VEC-NEXT: br label [[FOR_INC]] -; VEC: for.inc: -; VEC-NEXT: [[TMP6]] = add nuw nsw i64 [[TMP0]], 1 -; VEC-NEXT: [[TMP7]] = add i64 [[TMP1]], -1 -; VEC-NEXT: [[TMP8:%.*]] = icmp eq i64 [[TMP7]], 0 -; VEC-NEXT: br i1 [[TMP8]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] +; VEC-NEXT: br i1 [[TMP14]], label [[FOR_END:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; VEC: for.end: ; VEC-NEXT: ret void ; @@ -659,7 +585,7 @@ define void @minimal_bit_widths_with_aliasing_store(i1 %c, ptr %ptr) { ; UNROLL-NEXT: [[TMP6]] = add nuw nsw i64 [[TMP0]], 1 ; UNROLL-NEXT: [[TMP7]] = add i64 [[TMP1]], -1 ; UNROLL-NEXT: [[TMP8:%.*]] = icmp eq i64 [[TMP7]], 0 -; UNROLL-NEXT: br i1 [[TMP8]], label [[FOR_END:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]] +; UNROLL-NEXT: br i1 [[TMP8]], label [[FOR_END:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; UNROLL: for.end: ; UNROLL-NEXT: ret void ; @@ -696,8 +622,7 @@ define void @minimal_bit_widths_with_aliasing_store(i1 %c, ptr %ptr) { ; UNROLL-NOSIMPLIFY-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 0 ; UNROLL-NOSIMPLIFY-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; UNROLL-NOSIMPLIFY: middle.block: -; UNROLL-NOSIMPLIFY-NEXT: [[CMP_N:%.*]] = icmp eq i64 0, 0 -; UNROLL-NOSIMPLIFY-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NOSIMPLIFY-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NOSIMPLIFY: scalar.ph: ; UNROLL-NOSIMPLIFY-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; UNROLL-NOSIMPLIFY-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -741,7 +666,7 @@ define void @minimal_bit_widths_with_aliasing_store(i1 %c, ptr %ptr) { ; VEC-NEXT: [[TMP6]] = add nuw nsw i64 [[TMP0]], 1 ; VEC-NEXT: [[TMP7]] = add i64 [[TMP1]], -1 ; VEC-NEXT: [[TMP8:%.*]] = icmp eq i64 [[TMP7]], 0 -; VEC-NEXT: br i1 [[TMP8]], label [[FOR_END:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]] +; VEC-NEXT: br i1 [[TMP8]], label [[FOR_END:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; VEC: for.end: ; VEC-NEXT: ret void ; diff --git a/llvm/test/Transforms/LoopVectorize/induction-multiple-uses-in-same-instruction.ll b/llvm/test/Transforms/LoopVectorize/induction-multiple-uses-in-same-instruction.ll index 692bedc..6e390d2 100644 --- a/llvm/test/Transforms/LoopVectorize/induction-multiple-uses-in-same-instruction.ll +++ b/llvm/test/Transforms/LoopVectorize/induction-multiple-uses-in-same-instruction.ll @@ -24,8 +24,7 @@ define void @multiple_iv_uses_in_same_instruction(ptr %ptr) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 100, 100 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/induction.ll b/llvm/test/Transforms/LoopVectorize/induction.ll index 9048d74..fb9edf4 100644 --- a/llvm/test/Transforms/LoopVectorize/induction.ll +++ b/llvm/test/Transforms/LoopVectorize/induction.ll @@ -2783,8 +2783,7 @@ define i32 @i8_loop() nounwind readnone ssp uwtable { ; CHECK-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP28:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> [[TMP0]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 256, 256 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 1, [[ENTRY]] ], [ [[TMP2]], [[MIDDLE_BLOCK]] ] @@ -2855,8 +2854,7 @@ define i32 @i8_loop() nounwind readnone ssp uwtable { ; UNROLL-NO-IC: middle.block: ; UNROLL-NO-IC-NEXT: [[BIN_RDX:%.*]] = and <2 x i32> [[TMP1]], [[TMP0]] ; UNROLL-NO-IC-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> [[BIN_RDX]]) -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i32 256, 256 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; UNROLL-NO-IC-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 1, [[ENTRY]] ], [ [[TMP3]], [[MIDDLE_BLOCK]] ] @@ -2922,8 +2920,7 @@ define i32 @i16_loop() nounwind readnone ssp uwtable { ; CHECK-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP30:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> [[TMP0]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 65536, 65536 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 1, [[ENTRY]] ], [ [[TMP2]], [[MIDDLE_BLOCK]] ] @@ -2994,8 +2991,7 @@ define i32 @i16_loop() nounwind readnone ssp uwtable { ; UNROLL-NO-IC: middle.block: ; UNROLL-NO-IC-NEXT: [[BIN_RDX:%.*]] = and <2 x i32> [[TMP1]], [[TMP0]] ; UNROLL-NO-IC-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> [[BIN_RDX]]) -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i32 65536, 65536 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; UNROLL-NO-IC-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 1, [[ENTRY]] ], [ [[TMP3]], [[MIDDLE_BLOCK]] ] @@ -3065,8 +3061,7 @@ define i32 @max_i32_backedgetaken() nounwind readnone ssp uwtable { ; CHECK-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> [[TMP0]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 0, 0 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 1, [[ENTRY]] ], [ [[TMP2]], [[MIDDLE_BLOCK]] ] @@ -3090,7 +3085,7 @@ define i32 @max_i32_backedgetaken() nounwind readnone ssp uwtable { ; IND: vector.body: ; IND-NEXT: br i1 poison, label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; IND: middle.block: -; IND-NEXT: br i1 poison, label [[EXIT:%.*]], label [[SCALAR_PH]] +; IND-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; IND: scalar.ph: ; IND-NEXT: br label [[LOOP:%.*]] ; IND: loop: @@ -3109,7 +3104,7 @@ define i32 @max_i32_backedgetaken() nounwind readnone ssp uwtable { ; UNROLL: vector.body: ; UNROLL-NEXT: br i1 poison, label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; UNROLL: middle.block: -; UNROLL-NEXT: br i1 poison, label [[EXIT:%.*]], label [[SCALAR_PH]] +; UNROLL-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; UNROLL: scalar.ph: ; UNROLL-NEXT: br label [[LOOP:%.*]] ; UNROLL: loop: @@ -3137,8 +3132,7 @@ define i32 @max_i32_backedgetaken() nounwind readnone ssp uwtable { ; UNROLL-NO-IC: middle.block: ; UNROLL-NO-IC-NEXT: [[BIN_RDX:%.*]] = and <2 x i32> [[TMP1]], [[TMP0]] ; UNROLL-NO-IC-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> [[BIN_RDX]]) -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i32 0, 0 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; UNROLL-NO-IC-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 1, [[ENTRY]] ], [ [[TMP3]], [[MIDDLE_BLOCK]] ] @@ -3162,7 +3156,7 @@ define i32 @max_i32_backedgetaken() nounwind readnone ssp uwtable { ; INTERLEAVE: vector.body: ; INTERLEAVE-NEXT: br i1 poison, label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; INTERLEAVE: middle.block: -; INTERLEAVE-NEXT: br i1 poison, label [[EXIT:%.*]], label [[SCALAR_PH]] +; INTERLEAVE-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; INTERLEAVE: scalar.ph: ; INTERLEAVE-NEXT: br label [[LOOP:%.*]] ; INTERLEAVE: loop: @@ -5156,8 +5150,7 @@ define i32 @PR32419(i32 %a, i16 %b) { ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP48:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP17:%.*]] = call i32 @llvm.vector.reduce.or.v2i32(<2 x i32> [[TMP15]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 20, 20 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 0, [[MIDDLE_BLOCK]] ], [ -20, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ [[A]], [[ENTRY]] ], [ [[TMP17]], [[MIDDLE_BLOCK]] ] @@ -5377,8 +5370,7 @@ define i32 @PR32419(i32 %a, i16 %b) { ; UNROLL-NO-IC: middle.block: ; UNROLL-NO-IC-NEXT: [[BIN_RDX:%.*]] = or <2 x i32> [[TMP29]], [[TMP28]] ; UNROLL-NO-IC-NEXT: [[TMP31:%.*]] = call i32 @llvm.vector.reduce.or.v2i32(<2 x i32> [[BIN_RDX]]) -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i32 20, 20 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 0, [[MIDDLE_BLOCK]] ], [ -20, [[ENTRY:%.*]] ] ; UNROLL-NO-IC-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ [[A]], [[ENTRY]] ], [ [[TMP31]], [[MIDDLE_BLOCK]] ] @@ -5590,9 +5582,8 @@ define i64 @trunc_with_first_order_recurrence() { ; CHECK-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP50:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP12:%.*]] = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> [[TMP10]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 113, 112 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <2 x i32> [[VEC_IND2]], i32 1 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 42, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 113, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY]] ] @@ -5811,9 +5802,8 @@ define i64 @trunc_with_first_order_recurrence() { ; UNROLL-NO-IC: middle.block: ; UNROLL-NO-IC-NEXT: [[BIN_RDX:%.*]] = add <2 x i64> [[TMP21]], [[TMP20]] ; UNROLL-NO-IC-NEXT: [[TMP23:%.*]] = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> [[BIN_RDX]]) -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i64 113, 112 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <2 x i32> [[STEP_ADD5]], i32 1 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 42, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 113, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY]] ] @@ -5982,9 +5972,8 @@ define void @pr52460_first_order_recurrence_truncated_iv(ptr noalias %src, ptr % ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP52:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 100, 100 ; CHECK-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <2 x i32> [[VEC_IND]], i32 1 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -6110,9 +6099,8 @@ define void @pr52460_first_order_recurrence_truncated_iv(ptr noalias %src, ptr % ; UNROLL-NO-IC-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], 100 ; UNROLL-NO-IC-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP52:![0-9]+]] ; UNROLL-NO-IC: middle.block: -; UNROLL-NO-IC-NEXT: [[CMP_N:%.*]] = icmp eq i64 100, 100 ; UNROLL-NO-IC-NEXT: [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <2 x i32> [[STEP_ADD]], i32 1 -; UNROLL-NO-IC-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; UNROLL-NO-IC-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; UNROLL-NO-IC: scalar.ph: ; UNROLL-NO-IC-NEXT: [[SCALAR_RECUR_INIT:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ] ; UNROLL-NO-IC-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] diff --git a/llvm/test/Transforms/LoopVectorize/instruction-only-used-outside-of-loop.ll b/llvm/test/Transforms/LoopVectorize/instruction-only-used-outside-of-loop.ll index 8f0dc81..5f5cd78 100644 --- a/llvm/test/Transforms/LoopVectorize/instruction-only-used-outside-of-loop.ll +++ b/llvm/test/Transforms/LoopVectorize/instruction-only-used-outside-of-loop.ll @@ -20,8 +20,7 @@ define i32 @one_direct_branch(ptr %src) { ; CHECK-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP5:%.*]] = extractelement <4 x i32> [[TMP3]], i32 3 -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -79,8 +78,7 @@ define i32 @two_direct_branch(ptr %src) { ; CHECK-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP5:%.*]] = extractelement <4 x i32> [[TMP3]], i32 3 -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -152,8 +150,7 @@ define i32 @cond_branch(i32 %a, ptr %src) { ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP7:%.*]] = extractelement <4 x i32> [[PREDPHI]], i32 3 -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -218,8 +215,7 @@ define i32 @optimizable_trunc_used_outside() { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/is_fpclass.ll b/llvm/test/Transforms/LoopVectorize/is_fpclass.ll index c711c73..fa28831 100644 --- a/llvm/test/Transforms/LoopVectorize/is_fpclass.ll +++ b/llvm/test/Transforms/LoopVectorize/is_fpclass.ll @@ -19,8 +19,7 @@ define void @d() { ; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 0 ; CHECK-NEXT: br i1 [[TMP5]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 0, 0 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/lcssa-crashes.ll b/llvm/test/Transforms/LoopVectorize/lcssa-crashes.ll index 827e4f3..b193482 100644 --- a/llvm/test/Transforms/LoopVectorize/lcssa-crashes.ll +++ b/llvm/test/Transforms/LoopVectorize/lcssa-crashes.ll @@ -16,8 +16,7 @@ define void @test() { ; CHECK-NEXT: [[TMP1:%.*]] = icmp eq i64 [[INDEX_NEXT]], 0 ; CHECK-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1, 0 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END_I_I_I:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_END_I_I_I:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[FOR_BODY_LR_PH_I_I_I]] ] ; CHECK-NEXT: br label [[FOR_BODY_I_I_I:%.*]] @@ -129,8 +128,7 @@ define i32 @pr57508(ptr %src) { ; CHECK-NEXT: [[TMP0:%.*]] = icmp eq i64 [[INDEX_NEXT]], 2000 ; CHECK-NEXT: br i1 [[TMP0]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 2001, 2000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 2000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i32 [ 2000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] diff --git a/llvm/test/Transforms/LoopVectorize/load-deref-pred-align.ll b/llvm/test/Transforms/LoopVectorize/load-deref-pred-align.ll index f988874..ba98391 100644 --- a/llvm/test/Transforms/LoopVectorize/load-deref-pred-align.ll +++ b/llvm/test/Transforms/LoopVectorize/load-deref-pred-align.ll @@ -50,8 +50,7 @@ define i16 @test_access_size_not_multiple_of_align(i64 %len, ptr %test_base) { ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP18:%.*]] = call i16 @llvm.vector.reduce.add.v2i16(<2 x i16> [[TMP16]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i16 [ 0, [[ENTRY]] ], [ [[TMP18]], [[MIDDLE_BLOCK]] ] @@ -149,8 +148,7 @@ define i32 @test_access_size_multiple_of_align_but_offset_by_1(i64 %len, ptr %te ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[TMP18:%.*]] = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> [[TMP16]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4096, 4096 -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[LOOP_EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 4096, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP18]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/load-of-struct-deref-pred.ll b/llvm/test/Transforms/LoopVectorize/load-of-struct-deref-pred.ll index 2568057..7ddb864 100644 --- a/llvm/test/Transforms/LoopVectorize/load-of-struct-deref-pred.ll +++ b/llvm/test/Transforms/LoopVectorize/load-of-struct-deref-pred.ll @@ -34,8 +34,7 @@ define void @accesses_to_struct_dereferenceable(ptr noalias %dst) { ; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i64 [[INDEX_NEXT]], 32000 ; CHECK-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 32000, 32000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 32000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] @@ -157,8 +156,7 @@ define void @accesses_to_struct_may_not_be_dereferenceable_due_to_loop_bound(ptr ; CHECK-NEXT: [[TMP31:%.*]] = icmp eq i64 [[INDEX_NEXT]], 32000 ; CHECK-NEXT: br i1 [[TMP31]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 32001, 32000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 32000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] @@ -281,8 +279,7 @@ define void @accesses_to_struct_may_not_be_dereferenceable_access_size(ptr noali ; CHECK-NEXT: [[TMP32:%.*]] = icmp eq i64 [[INDEX_NEXT]], 32000 ; CHECK-NEXT: br i1 [[TMP32]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 32000, 32000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 32000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/optimal-epilog-vectorization.ll b/llvm/test/Transforms/LoopVectorize/optimal-epilog-vectorization.ll index 098b5a6..393490a 100644 --- a/llvm/test/Transforms/LoopVectorize/optimal-epilog-vectorization.ll +++ b/llvm/test/Transforms/LoopVectorize/optimal-epilog-vectorization.ll @@ -495,8 +495,7 @@ define void @induction_resume_value_requires_non_trivial_scev_expansion(ptr %dst ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 84 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP11:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 84, 84 -; CHECK-NEXT: br i1 [[CMP_N]], label [[OUTER_LATCH]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-NEXT: br i1 true, label [[OUTER_LATCH]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK: vec.epilog.iter.check: ; CHECK-NEXT: [[IND_END5:%.*]] = mul i8 84, [[INDUCTION_IV]] ; CHECK-NEXT: br i1 true, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] @@ -527,8 +526,7 @@ define void @induction_resume_value_requires_non_trivial_scev_expansion(ptr %dst ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT19]], 84 ; CHECK-NEXT: br i1 [[TMP13]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; CHECK: vec.epilog.middle.block: -; CHECK-NEXT: [[CMP_N7:%.*]] = icmp eq i64 84, 84 -; CHECK-NEXT: br i1 [[CMP_N7]], label [[OUTER_LATCH]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[OUTER_LATCH]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK: vec.epilog.scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL3:%.*]] = phi i64 [ 85, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 85, [[VEC_EPILOG_ITER_CHECK]] ], [ 1, [[ITER_CHECK]] ] ; CHECK-NEXT: [[BC_RESUME_VAL6:%.*]] = phi i8 [ [[IND_END4]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[IND_END5]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK]] ] @@ -587,8 +585,7 @@ define void @induction_resume_value_requires_non_trivial_scev_expansion(ptr %dst ; CHECK-PROFITABLE-BY-DEFAULT-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 84 ; CHECK-PROFITABLE-BY-DEFAULT-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; CHECK-PROFITABLE-BY-DEFAULT: middle.block: -; CHECK-PROFITABLE-BY-DEFAULT-NEXT: [[CMP_N:%.*]] = icmp eq i64 84, 84 -; CHECK-PROFITABLE-BY-DEFAULT-NEXT: br i1 [[CMP_N]], label [[OUTER_LATCH]], label [[VEC_EPILOG_ITER_CHECK:%.*]] +; CHECK-PROFITABLE-BY-DEFAULT-NEXT: br i1 true, label [[OUTER_LATCH]], label [[VEC_EPILOG_ITER_CHECK:%.*]] ; CHECK-PROFITABLE-BY-DEFAULT: vec.epilog.iter.check: ; CHECK-PROFITABLE-BY-DEFAULT-NEXT: [[IND_END5:%.*]] = mul i8 84, [[INDUCTION_IV]] ; CHECK-PROFITABLE-BY-DEFAULT-NEXT: br i1 true, label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]] @@ -619,8 +616,7 @@ define void @induction_resume_value_requires_non_trivial_scev_expansion(ptr %dst ; CHECK-PROFITABLE-BY-DEFAULT-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT19]], 84 ; CHECK-PROFITABLE-BY-DEFAULT-NEXT: br i1 [[TMP13]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK-PROFITABLE-BY-DEFAULT: vec.epilog.middle.block: -; CHECK-PROFITABLE-BY-DEFAULT-NEXT: [[CMP_N7:%.*]] = icmp eq i64 84, 84 -; CHECK-PROFITABLE-BY-DEFAULT-NEXT: br i1 [[CMP_N7]], label [[OUTER_LATCH]], label [[VEC_EPILOG_SCALAR_PH]] +; CHECK-PROFITABLE-BY-DEFAULT-NEXT: br i1 true, label [[OUTER_LATCH]], label [[VEC_EPILOG_SCALAR_PH]] ; CHECK-PROFITABLE-BY-DEFAULT: vec.epilog.scalar.ph: ; CHECK-PROFITABLE-BY-DEFAULT-NEXT: [[BC_RESUME_VAL3:%.*]] = phi i64 [ 85, [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ 85, [[VEC_EPILOG_ITER_CHECK]] ], [ 1, [[ITER_CHECK]] ] ; CHECK-PROFITABLE-BY-DEFAULT-NEXT: [[BC_RESUME_VAL6:%.*]] = phi i8 [ [[IND_END4]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[IND_END5]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/pointer-induction-unroll.ll b/llvm/test/Transforms/LoopVectorize/pointer-induction-unroll.ll index 36728c2..4109477 100644 --- a/llvm/test/Transforms/LoopVectorize/pointer-induction-unroll.ll +++ b/llvm/test/Transforms/LoopVectorize/pointer-induction-unroll.ll @@ -62,8 +62,7 @@ define void @non_constant_scalar_expansion(i32 %0, ptr %call) { ; STRIDED-NEXT: [[TMP20:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4294967264 ; STRIDED-NEXT: br i1 [[TMP20]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; STRIDED: middle.block: -; STRIDED-NEXT: [[CMP_N:%.*]] = icmp eq i64 4294967267, 4294967264 -; STRIDED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; STRIDED-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; STRIDED: scalar.ph: ; STRIDED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ -2, [[MIDDLE_BLOCK]] ], [ 30, [[ENTRY:%.*]] ] ; STRIDED-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ null, [[ENTRY]] ] diff --git a/llvm/test/Transforms/LoopVectorize/pointer-induction.ll b/llvm/test/Transforms/LoopVectorize/pointer-induction.ll index 885c7d7..27f6f7b 100644 --- a/llvm/test/Transforms/LoopVectorize/pointer-induction.ll +++ b/llvm/test/Transforms/LoopVectorize/pointer-induction.ll @@ -252,8 +252,7 @@ define void @non_constant_vector_expansion(i32 %0, ptr %call) { ; STRIDED-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 100 ; STRIDED-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; STRIDED: middle.block: -; STRIDED-NEXT: [[CMP_N:%.*]] = icmp eq i64 101, 100 -; STRIDED-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; STRIDED-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; STRIDED: scalar.ph: ; STRIDED-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; STRIDED-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ null, [[ENTRY]] ] diff --git a/llvm/test/Transforms/LoopVectorize/pr39417-optsize-scevchecks.ll b/llvm/test/Transforms/LoopVectorize/pr39417-optsize-scevchecks.ll index 31ced78..d04cca0 100644 --- a/llvm/test/Transforms/LoopVectorize/pr39417-optsize-scevchecks.ll +++ b/llvm/test/Transforms/LoopVectorize/pr39417-optsize-scevchecks.ll @@ -66,8 +66,7 @@ define void @scev4stride1(ptr noalias nocapture %a, ptr noalias nocapture readon ; CHECK-NEXT: [[TMP24:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1024 ; CHECK-NEXT: br i1 [[TMP24]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop !0 ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 1024, 1024 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END_LOOPEXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[FOR_END_LOOPEXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK: for.body: ; CHECK: for.end.loopexit: diff --git a/llvm/test/Transforms/LoopVectorize/pr44488-predication.ll b/llvm/test/Transforms/LoopVectorize/pr44488-predication.ll index b91a96c..f1075e3 100644 --- a/llvm/test/Transforms/LoopVectorize/pr44488-predication.ll +++ b/llvm/test/Transforms/LoopVectorize/pr44488-predication.ll @@ -44,8 +44,7 @@ define i16 @test_true_and_false_branch_equal() { ; CHECK-NEXT: [[TMP12:%.*]] = icmp eq i32 [[INDEX_NEXT]], 12 ; CHECK-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 12, 12 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 111, [[MIDDLE_BLOCK]] ], [ 99, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[FOR_BODY:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/pr47343-expander-lcssa-after-cfg-update.ll b/llvm/test/Transforms/LoopVectorize/pr47343-expander-lcssa-after-cfg-update.ll index 9831135..b3b6d3e 100644 --- a/llvm/test/Transforms/LoopVectorize/pr47343-expander-lcssa-after-cfg-update.ll +++ b/llvm/test/Transforms/LoopVectorize/pr47343-expander-lcssa-after-cfg-update.ll @@ -52,8 +52,7 @@ define void @f() { ; CHECK-NEXT: [[TMP2:%.*]] = icmp eq i32 [[INDEX_NEXT]], 500 ; CHECK-NEXT: br i1 [[TMP2]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 500, 500 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 500, [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_PREHEADER]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/pr50686.ll b/llvm/test/Transforms/LoopVectorize/pr50686.ll index 9cf4f9c..6ce491e 100644 --- a/llvm/test/Transforms/LoopVectorize/pr50686.ll +++ b/llvm/test/Transforms/LoopVectorize/pr50686.ll @@ -38,8 +38,7 @@ define void @m(ptr nocapture %p, ptr nocapture %p2, i32 %q) { ; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i64 [[INDEX_NEXT]], 60 ; CHECK-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 63, 60 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END17:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_END17:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 60, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_MEMCHECK]] ] ; CHECK-NEXT: br label [[FOR_COND5:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/pr55167-fold-tail-live-out.ll b/llvm/test/Transforms/LoopVectorize/pr55167-fold-tail-live-out.ll index b9e0eb3..e968d45 100644 --- a/llvm/test/Transforms/LoopVectorize/pr55167-fold-tail-live-out.ll +++ b/llvm/test/Transforms/LoopVectorize/pr55167-fold-tail-live-out.ll @@ -36,8 +36,7 @@ define i32 @test(i32 %a, i1 %c.1, i1 %c.2 ) #0 { ; CHECK: middle.block: ; CHECK-NEXT: [[TMP9:%.*]] = extractelement <2 x i32> [[PREDPHI5]], i32 1 ; CHECK-NEXT: [[TMP10:%.*]] = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> [[PREDPHI7]]) -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 176, 176 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 182, [[MIDDLE_BLOCK]] ], [ 6, [[BB:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ 35902, [[BB]] ], [ [[TMP10]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/pr58811-scev-expansion.ll b/llvm/test/Transforms/LoopVectorize/pr58811-scev-expansion.ll index 3f97c0c..64fdefb 100644 --- a/llvm/test/Transforms/LoopVectorize/pr58811-scev-expansion.ll +++ b/llvm/test/Transforms/LoopVectorize/pr58811-scev-expansion.ll @@ -40,9 +40,8 @@ define void @test1_pr58811() { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i32 [[INDEX_NEXT]], 196 ; CHECK-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 199, 196 ; CHECK-NEXT: [[IND_ESCAPE:%.*]] = mul i32 195, [[INDUCTION_IV_LCSSA]] -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_3_PREHEADER:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOP_3_PREHEADER:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 196, [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_2_PREHEADER]] ] ; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i32 [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_2_PREHEADER]] ] @@ -145,9 +144,8 @@ define void @test2_pr58811() { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i32 [[INDEX_NEXT]], 196 ; CHECK-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 199, 196 ; CHECK-NEXT: [[IND_ESCAPE:%.*]] = mul i32 195, [[INDUCTION_IV_LCSSA]] -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_4_PREHEADER:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOP_4_PREHEADER:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 196, [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_3_PREHEADER]] ] ; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i32 [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_3_PREHEADER]] ] @@ -233,9 +231,8 @@ define void @test3_pr58811() { ; CHECK-NEXT: [[TMP12:%.*]] = icmp eq i32 [[INDEX_NEXT]], 196 ; CHECK-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 199, 196 ; CHECK-NEXT: [[IND_ESCAPE:%.*]] = mul i32 195, [[TMP3]] -; CHECK-NEXT: br i1 [[CMP_N]], label [[LOOP_4_PREHEADER:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[LOOP_4_PREHEADER:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i16 [ 196, [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_3_PREHEADER]] ] ; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i32 [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 0, [[LOOP_3_PREHEADER]] ] diff --git a/llvm/test/Transforms/LoopVectorize/reduction-with-invariant-store.ll b/llvm/test/Transforms/LoopVectorize/reduction-with-invariant-store.ll index b953e21..022830d 100644 --- a/llvm/test/Transforms/LoopVectorize/reduction-with-invariant-store.ll +++ b/llvm/test/Transforms/LoopVectorize/reduction-with-invariant-store.ll @@ -25,8 +25,7 @@ target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f3 ; CHECK: middle.block: ; CHECK-NEXT: [[TMP6:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP4]]) ; CHECK-NEXT: store i32 [[TMP6]], ptr [[GEP_DST:%.*]], align 4 -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH:%.*]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH:%.*]] define void @reduc_store(ptr %dst, ptr readonly %src) { entry: %gep.dst = getelementptr inbounds i32, ptr %dst, i64 42 @@ -221,8 +220,7 @@ for.end: ; CHECK: middle.block: ; CHECK-NEXT: [[TMP36:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP34]]) ; CHECK-NEXT: store i32 [[TMP36]], ptr [[GEP_DST:%.*]], align 4 -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH:%.*]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH:%.*]] define void @reduc_store_inside_unrolled(ptr %dst, ptr readonly %src) { entry: %gep.dst = getelementptr inbounds i32, ptr %dst, i64 42 @@ -491,8 +489,7 @@ define void @test_drop_poison_generating_dead_recipe(ptr %dst) { ; CHECK: middle.block: ; CHECK-NEXT: [[TMP2:%.*]] = call i64 @llvm.vector.reduce.add.v4i64(<4 x i64> [[TMP0]]) ; CHECK-NEXT: store i64 [[TMP2]], ptr [[DST:%.*]], align 8 -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 363, 360 -; CHECK-NEXT: br i1 [[CMP_N]], label %exit, label %scalar.ph +; CHECK-NEXT: br i1 false, label %exit, label %scalar.ph ; CHECK: scalar.ph: ; entry: diff --git a/llvm/test/Transforms/LoopVectorize/runtime-check-needed-but-empty.ll b/llvm/test/Transforms/LoopVectorize/runtime-check-needed-but-empty.ll index b5f4755..b6c6eff 100644 --- a/llvm/test/Transforms/LoopVectorize/runtime-check-needed-but-empty.ll +++ b/llvm/test/Transforms/LoopVectorize/runtime-check-needed-but-empty.ll @@ -31,8 +31,7 @@ define void @test(ptr %A, i32 %x) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], undef ; CHECK-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 undef, undef -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 undef, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ undef, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_SCEVCHECK]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/scev-exit-phi-invalidation.ll b/llvm/test/Transforms/LoopVectorize/scev-exit-phi-invalidation.ll index ae3d536..521af74 100644 --- a/llvm/test/Transforms/LoopVectorize/scev-exit-phi-invalidation.ll +++ b/llvm/test/Transforms/LoopVectorize/scev-exit-phi-invalidation.ll @@ -19,8 +19,7 @@ define void @test_pr63368(i1 %c, ptr %A) { ; CHECK-NEXT: [[TMP1:%.*]] = icmp eq i32 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP1]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 100, 100 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT_1:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT_1:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_1_HEADER:%.*]] @@ -58,7 +57,7 @@ define void @test_pr63368(i1 %c, ptr %A) { ; CHECK-NEXT: [[N_VEC:%.*]] = sub i32 [[TMP2]], [[N_MOD_VF]] ; CHECK-NEXT: [[IND_END:%.*]] = trunc i32 [[N_VEC]] to i8 ; CHECK-NEXT: br label [[VECTOR_BODY7:%.*]] -; CHECK: vector.body7: +; CHECK: vector.body6: ; CHECK-NEXT: [[INDEX8:%.*]] = phi i32 [ 0, [[VECTOR_PH4]] ], [ [[INDEX_NEXT9:%.*]], [[VECTOR_BODY7]] ] ; CHECK-NEXT: [[OFFSET_IDX:%.*]] = trunc i32 [[INDEX8]] to i8 ; CHECK-NEXT: [[TMP14:%.*]] = add i8 [[OFFSET_IDX]], 0 diff --git a/llvm/test/Transforms/LoopVectorize/scev-predicate-reasoning.ll b/llvm/test/Transforms/LoopVectorize/scev-predicate-reasoning.ll index ab080c7..6ae6645 100644 --- a/llvm/test/Transforms/LoopVectorize/scev-predicate-reasoning.ll +++ b/llvm/test/Transforms/LoopVectorize/scev-predicate-reasoning.ll @@ -47,8 +47,7 @@ define void @step_direction_unknown(i32 %arg, ptr %dst) { ; CHECK-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], 0 ; CHECK-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 0, 0 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 0, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_SCEVCHECK]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -116,8 +115,7 @@ define void @integer_induction_wraps_scev_predicate_known(i32 %x, ptr %call, ptr ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 4294967264 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 4294967267, 4294967264 -; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ -2, [[MIDDLE_BLOCK]] ], [ 30, [[ENTRY:%.*]] ], [ 30, [[VECTOR_SCEVCHECK]] ] ; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[START]], [[ENTRY]] ], [ [[START]], [[VECTOR_SCEVCHECK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/select-reduction-start-value-may-be-undef-or-poison.ll b/llvm/test/Transforms/LoopVectorize/select-reduction-start-value-may-be-undef-or-poison.ll index f691a2a..16ab454 100644 --- a/llvm/test/Transforms/LoopVectorize/select-reduction-start-value-may-be-undef-or-poison.ll +++ b/llvm/test/Transforms/LoopVectorize/select-reduction-start-value-may-be-undef-or-poison.ll @@ -28,8 +28,7 @@ define i64 @pr62565_incoming_value_known_undef(i64 %a, ptr %src) { ; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <2 x i64> [[TMP4]], undef ; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[RDX_SELECT_CMP]]) ; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP6]], i64 [[A]], i64 undef -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 32, 32 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 33, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i64 [ undef, [[ENTRY]] ], [ [[RDX_SELECT]], [[MIDDLE_BLOCK]] ] @@ -93,8 +92,7 @@ define i64 @pr62565_incoming_value_known_poison(i64 %a, ptr %src) { ; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <2 x i64> [[TMP4]], poison ; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[RDX_SELECT_CMP]]) ; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP6]], i64 [[A]], i64 poison -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 32, 32 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 33, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i64 [ poison, [[ENTRY]] ], [ [[RDX_SELECT]], [[MIDDLE_BLOCK]] ] @@ -162,8 +160,7 @@ define i64 @pr62565_incoming_value_may_be_poison(i64 %a, ptr %src, i64 %start) { ; CHECK-NEXT: [[RDX_SELECT_CMP:%.*]] = icmp ne <2 x i64> [[TMP4]], [[DOTSPLAT]] ; CHECK-NEXT: [[TMP6:%.*]] = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> [[RDX_SELECT_CMP]]) ; CHECK-NEXT: [[RDX_SELECT:%.*]] = select i1 [[TMP6]], i64 [[A]], i64 [[START]] -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 32, 32 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 33, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i64 [ [[START]], [[ENTRY]] ], [ [[RDX_SELECT]], [[MIDDLE_BLOCK]] ] diff --git a/llvm/test/Transforms/LoopVectorize/single-value-blend-phis.ll b/llvm/test/Transforms/LoopVectorize/single-value-blend-phis.ll index 250d0b7..47271ba 100644 --- a/llvm/test/Transforms/LoopVectorize/single-value-blend-phis.ll +++ b/llvm/test/Transforms/LoopVectorize/single-value-blend-phis.ll @@ -34,8 +34,7 @@ define void @single_incoming_phi_no_blend_mask(i64 %a, i64 %b) { ; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i64 [[INDEX_NEXT]], 32 ; CHECK-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 32, 32 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 32, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] @@ -126,8 +125,7 @@ define void @single_incoming_phi_with_blend_mask(i64 %a, i64 %b) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 32 ; CHECK-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 32, 32 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 32, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] @@ -220,8 +218,7 @@ define void @multiple_incoming_phi_with_blend_mask(i64 %a, ptr noalias %dst) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 32 ; CHECK-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 32, 32 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 32, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] @@ -323,8 +320,7 @@ define void @single_incoming_needs_predication(i64 %a, i64 %b) { ; CHECK-NEXT: [[TMP22:%.*]] = icmp eq i64 [[INDEX_NEXT]], 64 ; CHECK-NEXT: br i1 [[TMP22]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 64, 64 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 64, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] @@ -401,8 +397,7 @@ define void @duplicated_incoming_blocks_blend(i32 %x, ptr %ptr) { ; CHECK-NEXT: [[TMP3:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP3]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/strided-accesses-interleave-only.ll b/llvm/test/Transforms/LoopVectorize/strided-accesses-interleave-only.ll index 3782ed5..bc6fb48 100644 --- a/llvm/test/Transforms/LoopVectorize/strided-accesses-interleave-only.ll +++ b/llvm/test/Transforms/LoopVectorize/strided-accesses-interleave-only.ll @@ -22,8 +22,7 @@ define void @test_variable_stride(ptr %dst, i32 %scale) { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/trunc-shifts.ll b/llvm/test/Transforms/LoopVectorize/trunc-shifts.ll index b6e8722..d7fd074 100644 --- a/llvm/test/Transforms/LoopVectorize/trunc-shifts.ll +++ b/llvm/test/Transforms/LoopVectorize/trunc-shifts.ll @@ -26,8 +26,7 @@ define void @test_pr47927_lshr_const_shift_ops(ptr %dst, i32 %f) { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i32 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 100, 100 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -87,8 +86,7 @@ define void @test_shl_const_shift_ops(ptr %dst, i32 %f) { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i32 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 100, 100 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -148,8 +146,7 @@ define void @test_ashr_const_shift_ops(ptr %dst, i32 %f) { ; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i32 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 100, 100 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -209,8 +206,7 @@ define void @test_shl_const_shifted_op(ptr %dst, i32 %f) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i32 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 100, 100 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -275,8 +271,7 @@ define void @test_lshr_by_18(ptr %A) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i32 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 100, 100 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -344,8 +339,7 @@ define void @test_lshr_by_4(ptr %A) { ; CHECK-NEXT: [[TMP11:%.*]] = icmp eq i32 [[INDEX_NEXT]], 100 ; CHECK-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 100, 100 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ 100, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1.ll b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1.ll index c47c3fc..4cee3e3 100644 --- a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1.ll +++ b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1.ll @@ -26,8 +26,7 @@ define void @ld_div1_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -87,8 +86,7 @@ define void @ld_div2_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -153,8 +151,7 @@ define void @ld_div3_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -224,8 +221,7 @@ define void @ld_div1_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -288,8 +284,7 @@ define void @ld_div2_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; CHECK-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -359,8 +354,7 @@ define void @ld_div3_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -430,8 +424,7 @@ define void @ld_div1_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -501,8 +494,7 @@ define void @ld_div2_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -565,8 +557,7 @@ define void @ld_div3_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP18:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -625,8 +616,7 @@ define void @ld_div1_step1_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 998 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP20:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 998 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 999, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -692,8 +682,7 @@ define void @ld_div2_step1_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 998 ; CHECK-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 998 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 999, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -759,8 +748,7 @@ define void @ld_div3_step1_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 998 ; CHECK-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 998 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 999, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -831,8 +819,7 @@ define void @ld_div1_step2_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 498 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP26:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 498 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -896,8 +883,7 @@ define void @ld_div2_step2_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT]], 498 ; CHECK-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP28:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 498 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -968,8 +954,7 @@ define void @ld_div3_step2_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 498 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP30:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 498 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -1040,8 +1025,7 @@ define void @ld_div1_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -1112,8 +1096,7 @@ define void @ld_div2_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP34:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -1177,8 +1160,7 @@ define void @ld_div3_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP36:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -1241,8 +1223,7 @@ define void @test_step_is_not_invariant(ptr %A) { ; CHECK-NEXT: [[TMP11:%.*]] = icmp eq i32 [[INDEX_NEXT]], 56 ; CHECK-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP38:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 56, 56 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 56, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_and.ll b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_and.ll index 3fb1bc7..7f8b33e 100644 --- a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_and.ll +++ b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_and.ll @@ -26,8 +26,7 @@ define void @ld_and_neg1_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -87,8 +86,7 @@ define void @ld_and_neg2_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -153,8 +151,7 @@ define void @ld_and_neg3_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -224,8 +221,7 @@ define void @ld_and_neg1_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -289,8 +285,7 @@ define void @ld_and_neg2_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; CHECK-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -360,8 +355,7 @@ define void @ld_and_neg1_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -431,8 +425,7 @@ define void @ld_and_neg2_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -498,8 +491,7 @@ define void @ld_and_neg2_step1_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 998 ; CHECK-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 998 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 999, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -570,8 +562,7 @@ define void @ld_and_neg2_step2_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 498 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP18:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 498 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -642,8 +633,7 @@ define void @ld_and_neg2_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP20:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -714,8 +704,7 @@ define void @ld_and_neg3_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_div_urem.ll b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_div_urem.ll index ff1a647..098e29e 100644 --- a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_div_urem.ll +++ b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_div_urem.ll @@ -59,8 +59,7 @@ define void @ld_div2_urem3_1(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP38:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP38]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -153,8 +152,7 @@ define void @ld_div2_urem3_2(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP39:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP39]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -245,8 +243,7 @@ define void @ld_div4(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP37:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP37]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] @@ -307,8 +304,7 @@ define void @ld_div8_urem3(ptr noalias %A, ptr noalias %B) { ; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_lshr.ll b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_lshr.ll index 72d41ff..f797729 100644 --- a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_lshr.ll +++ b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction1_lshr.ll @@ -27,8 +27,7 @@ define void @ld_lshr0_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF2-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -67,8 +66,7 @@ define void @ld_lshr0_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF4-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -128,8 +126,7 @@ define void @ld_lshr1_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF2-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -183,8 +180,7 @@ define void @ld_lshr1_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP21:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF4-NEXT: br i1 [[TMP21]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -244,8 +240,7 @@ define void @ld_lshr2_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF2-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -285,8 +280,7 @@ define void @ld_lshr2_step1_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF4-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -356,8 +350,7 @@ define void @ld_lshr0_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF2-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -424,8 +417,7 @@ define void @ld_lshr0_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP30:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF4-NEXT: br i1 [[TMP30]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -488,8 +480,7 @@ define void @ld_lshr1_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF2-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -541,8 +532,7 @@ define void @ld_lshr1_step2_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF4-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -612,8 +602,7 @@ define void @ld_lshr0_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -680,8 +669,7 @@ define void @ld_lshr0_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP30:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP30]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -751,8 +739,7 @@ define void @ld_lshr1_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -819,8 +806,7 @@ define void @ld_lshr1_step3_start0_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP30:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP30]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -887,8 +873,7 @@ define void @ld_lshr1_step1_start1_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT]], 998 ; VF2-NEXT: br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 998 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 999, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -943,8 +928,7 @@ define void @ld_lshr1_step1_start1_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP21:%.*]] = icmp eq i64 [[INDEX_NEXT]], 996 ; VF4-NEXT: br i1 [[TMP21]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 996 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -1008,8 +992,7 @@ define void @ld_lshr1_step2_start1_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP11:%.*]] = icmp eq i64 [[INDEX_NEXT]], 498 ; VF2-NEXT: br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP18:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 498 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -1062,8 +1045,7 @@ define void @ld_lshr1_step2_start1_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 496 ; VF4-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP18:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 496 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 993, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -1134,8 +1116,7 @@ define void @ld_lshr1_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP20:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -1203,8 +1184,7 @@ define void @ld_lshr1_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP31:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP31]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP20:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] @@ -1275,8 +1255,7 @@ define void @ld_lshr2_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: br label [[LOOP:%.*]] @@ -1344,8 +1323,7 @@ define void @ld_lshr2_step3_start1_ind1(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP31:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP31]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: br label [[LOOP:%.*]] diff --git a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction2.ll b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction2.ll index 0461718..b8401ca 100644 --- a/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction2.ll +++ b/llvm/test/Transforms/LoopVectorize/uniform_across_vf_induction2.ll @@ -36,8 +36,7 @@ define void @ld_div1_step1_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP15:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF2-NEXT: br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -100,8 +99,7 @@ define void @ld_div1_step1_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP23:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF4-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -173,8 +171,7 @@ define void @ld_div2_step1_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF2-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -237,8 +234,7 @@ define void @ld_div2_step1_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP23:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF4-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -316,8 +312,7 @@ define void @ld_div3_step1_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP15:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF2-NEXT: br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -380,8 +375,7 @@ define void @ld_div3_step1_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP23:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000 ; VF4-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 1000, 1000 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -464,8 +458,7 @@ define void @ld_div1_step2_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF2-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 500, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -541,8 +534,7 @@ define void @ld_div1_step2_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP32:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF4-NEXT: br i1 [[TMP32]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 500, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -625,8 +617,7 @@ define void @ld_div2_step2_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF2-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 500, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -702,8 +693,7 @@ define void @ld_div2_step2_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP32:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF4-NEXT: br i1 [[TMP32]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 500, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -786,8 +776,7 @@ define void @ld_div3_step2_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF2-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 500, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -863,8 +852,7 @@ define void @ld_div3_step2_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP32:%.*]] = icmp eq i64 [[INDEX_NEXT]], 500 ; VF4-NEXT: br i1 [[TMP32]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 500, 500 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 1000, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 500, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -947,8 +935,7 @@ define void @ld_div1_step3_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1024,8 +1011,7 @@ define void @ld_div1_step3_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP32:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP32]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1108,8 +1094,7 @@ define void @ld_div2_step3_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1185,8 +1170,7 @@ define void @ld_div2_step3_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP32:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP32]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1269,8 +1253,7 @@ define void @ld_div3_step3_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP18:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1346,8 +1329,7 @@ define void @ld_div3_step3_start0_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP32:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP32]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP18:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1426,8 +1408,7 @@ define void @ld_div1_step1_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP15:%.*]] = icmp eq i64 [[INDEX_NEXT]], 998 ; VF2-NEXT: br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP20:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 998 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 999, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 998, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1491,8 +1472,7 @@ define void @ld_div1_step1_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP23:%.*]] = icmp eq i64 [[INDEX_NEXT]], 996 ; VF4-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP20:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 996 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1571,8 +1551,7 @@ define void @ld_div2_step1_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP15:%.*]] = icmp eq i64 [[INDEX_NEXT]], 998 ; VF2-NEXT: br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 998 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 999, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 998, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1636,8 +1615,7 @@ define void @ld_div2_step1_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP23:%.*]] = icmp eq i64 [[INDEX_NEXT]], 996 ; VF4-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 996 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1716,8 +1694,7 @@ define void @ld_div3_step1_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP15:%.*]] = icmp eq i64 [[INDEX_NEXT]], 998 ; VF2-NEXT: br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 998 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 999, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 998, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1781,8 +1758,7 @@ define void @ld_div3_step1_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP23:%.*]] = icmp eq i64 [[INDEX_NEXT]], 996 ; VF4-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 999, 996 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 996, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1866,8 +1842,7 @@ define void @ld_div1_step2_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], 498 ; VF2-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP26:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 498 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 498, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -1944,8 +1919,7 @@ define void @ld_div1_step2_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP33:%.*]] = icmp eq i64 [[INDEX_NEXT]], 496 ; VF4-NEXT: br i1 [[TMP33]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP26:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 496 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 993, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 496, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2029,8 +2003,7 @@ define void @ld_div2_step2_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], 498 ; VF2-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP28:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 498 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 498, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2107,8 +2080,7 @@ define void @ld_div2_step2_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP33:%.*]] = icmp eq i64 [[INDEX_NEXT]], 496 ; VF4-NEXT: br i1 [[TMP33]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP28:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 496 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 993, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 496, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2192,8 +2164,7 @@ define void @ld_div3_step2_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], 498 ; VF2-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP30:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 498 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 498, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2270,8 +2241,7 @@ define void @ld_div3_step2_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP33:%.*]] = icmp eq i64 [[INDEX_NEXT]], 496 ; VF4-NEXT: br i1 [[TMP33]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP30:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 499, 496 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 993, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 496, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2355,8 +2325,7 @@ define void @ld_div1_step3_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2433,8 +2402,7 @@ define void @ld_div1_step3_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP33:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP33]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP32:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2518,8 +2486,7 @@ define void @ld_div2_step3_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP34:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2596,8 +2563,7 @@ define void @ld_div2_step3_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP33:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP33]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP34:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2681,8 +2647,7 @@ define void @ld_div3_step3_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF2-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF2-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP36:![0-9]+]] ; VF2: middle.block: -; VF2-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF2-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF2-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF2: scalar.ph: ; VF2-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF2-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] @@ -2759,8 +2724,7 @@ define void @ld_div3_step3_start1_ind2(ptr noalias %A, ptr noalias %B) { ; VF4-NEXT: [[TMP33:%.*]] = icmp eq i64 [[INDEX_NEXT]], 332 ; VF4-NEXT: br i1 [[TMP33]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP36:![0-9]+]] ; VF4: middle.block: -; VF4-NEXT: [[CMP_N:%.*]] = icmp eq i64 333, 332 -; VF4-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; VF4-NEXT: br i1 false, label [[EXIT:%.*]], label [[SCALAR_PH]] ; VF4: scalar.ph: ; VF4-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ 997, [[MIDDLE_BLOCK]] ], [ 1, [[ENTRY:%.*]] ] ; VF4-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ 332, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ] diff --git a/llvm/test/Transforms/LoopVectorize/widen-gep-all-indices-invariant.ll b/llvm/test/Transforms/LoopVectorize/widen-gep-all-indices-invariant.ll index 54e274c..31021d5 100644 --- a/llvm/test/Transforms/LoopVectorize/widen-gep-all-indices-invariant.ll +++ b/llvm/test/Transforms/LoopVectorize/widen-gep-all-indices-invariant.ll @@ -22,8 +22,7 @@ define void @pr63340(ptr %A, ptr %B) { ; CHECK-NEXT: [[TMP4:%.*]] = icmp eq i32 [[INDEX_NEXT]], 128 ; CHECK-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: -; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 128, 128 -; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-NEXT: br i1 true, label [[EXIT:%.*]], label [[SCALAR_PH]] ; CHECK: scalar.ph: ; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i8 [ -128, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; CHECK-NEXT: br label [[LOOP_HEADER:%.*]] |