[llvm] [LV][NFC] precommit test for EVL transform (PR #92203)
Shih-Po Hung via llvm-commits
llvm-commits at lists.llvm.org
Wed May 15 04:27:49 PDT 2024
https://github.com/arcbbb updated https://github.com/llvm/llvm-project/pull/92203
>From 3f3bf370a981e3215269b2b584ba3f75b3221749 Mon Sep 17 00:00:00 2001
From: ShihPo Hung <shihpo.hung at sifive.com>
Date: Tue, 14 May 2024 19:06:11 -0700
Subject: [PATCH 1/2] [LV][NFC] precommit test for EVL transform
A precommit test case to show vector loops generated from EVL transform
- This is a precommit test for https://github.com/llvm/llvm-project/pull/92092
---
.../RISCV/evl-compatible-loops.ll | 262 ++++++++++++++++++
1 file changed, 262 insertions(+)
create mode 100644 llvm/test/Transforms/LoopVectorize/RISCV/evl-compatible-loops.ll
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/evl-compatible-loops.ll b/llvm/test/Transforms/LoopVectorize/RISCV/evl-compatible-loops.ll
new file mode 100644
index 0000000000000..e076ac2ca9f85
--- /dev/null
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/evl-compatible-loops.ll
@@ -0,0 +1,262 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
+; RUN: opt -passes=loop-vectorize -force-tail-folding-style=data-with-evl \
+; RUN: -prefer-predicate-over-epilogue=predicate-dont-vectorize \
+; RUN: -force-target-instruction-cost=1 -mtriple=riscv64 -mattr=+v -S < %s | FileCheck %s
+
+; Check loops having VPWidenIntOrFpInductionRecipe
+define void @foo(ptr noalias %a, i64 %N) {
+; CHECK-LABEL: define void @foo(
+; CHECK-SAME: ptr noalias [[A:%.*]], i64 [[N:%.*]]) #[[ATTR0:[0-9]+]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = sub i64 -1, [[N]]
+; CHECK-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP2:%.*]] = mul i64 [[TMP1]], 2
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i64 [[TMP0]], [[TMP2]]
+; CHECK-NEXT: br i1 [[TMP3]], label [[SCALAR_PH:%.*]], label [[ENTRY:%.*]]
+; CHECK: vector.ph:
+; CHECK-NEXT: [[TMP4:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP5:%.*]] = mul i64 [[TMP4]], 2
+; CHECK-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP7:%.*]] = mul i64 [[TMP6]], 2
+; CHECK-NEXT: [[TMP8:%.*]] = sub i64 [[TMP7]], 1
+; CHECK-NEXT: [[N_RND_UP:%.*]] = add i64 [[N]], [[TMP8]]
+; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP5]]
+; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
+; CHECK-NEXT: [[TMP9:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP10:%.*]] = mul i64 [[TMP9]], 2
+; CHECK-NEXT: [[TMP11:%.*]] = call <vscale x 2 x i64> @llvm.experimental.stepvector.nxv2i64()
+; CHECK-NEXT: [[TMP12:%.*]] = add <vscale x 2 x i64> [[TMP11]], zeroinitializer
+; CHECK-NEXT: [[TMP13:%.*]] = mul <vscale x 2 x i64> [[TMP12]], shufflevector (<vscale x 2 x i64> insertelement (<vscale x 2 x i64> poison, i64 1, i64 0), <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer)
+; CHECK-NEXT: [[INDUCTION:%.*]] = add <vscale x 2 x i64> zeroinitializer, [[TMP13]]
+; CHECK-NEXT: [[TMP14:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP15:%.*]] = mul i64 [[TMP14]], 2
+; CHECK-NEXT: [[TMP16:%.*]] = mul i64 1, [[TMP15]]
+; CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[TMP16]], i64 0
+; CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+; CHECK-NEXT: br label [[FOR_BODY:%.*]]
+; CHECK: vector.body:
+; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[INDEX_EVL_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[VEC_IND:%.*]] = phi <vscale x 2 x i64> [ [[INDUCTION]], [[ENTRY]] ], [ [[VEC_IND_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[TMP17:%.*]] = sub i64 [[N]], [[EVL_BASED_IV]]
+; CHECK-NEXT: [[TMP18:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[TMP17]], i32 2, i1 true)
+; CHECK-NEXT: [[TMP19:%.*]] = add i64 [[EVL_BASED_IV]], 0
+; CHECK-NEXT: [[TMP20:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[TMP19]]
+; CHECK-NEXT: [[TMP21:%.*]] = getelementptr inbounds i64, ptr [[TMP20]], i32 0
+; CHECK-NEXT: call void @llvm.vp.store.nxv2i64.p0(<vscale x 2 x i64> [[VEC_IND]], ptr align 8 [[TMP21]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer), i32 [[TMP18]])
+; CHECK-NEXT: [[TMP22:%.*]] = zext i32 [[TMP18]] to i64
+; CHECK-NEXT: [[INDEX_EVL_NEXT]] = add i64 [[TMP22]], [[EVL_BASED_IV]]
+; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], [[TMP10]]
+; CHECK-NEXT: [[VEC_IND_NEXT]] = add <vscale x 2 x i64> [[VEC_IND]], [[DOTSPLAT]]
+; CHECK-NEXT: [[TMP23:%.*]] = icmp eq i64 [[IV_NEXT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
+; CHECK: middle.block:
+; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]]
+; CHECK: scalar.ph:
+; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY1:%.*]] ]
+; CHECK-NEXT: br label [[FOR_BODY1:%.*]]
+; CHECK: for.body:
+; CHECK-NEXT: [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT1:%.*]], [[FOR_BODY1]] ]
+; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV1]]
+; CHECK-NEXT: store i64 [[IV1]], ptr [[ARRAYIDX]], align 8
+; CHECK-NEXT: [[IV_NEXT1]] = add nuw nsw i64 [[IV1]], 1
+; CHECK-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT1]], [[N]]
+; CHECK-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_COND_CLEANUP]], label [[FOR_BODY1]], !llvm.loop [[LOOP3:![0-9]+]]
+; CHECK: for.cond.cleanup:
+; CHECK-NEXT: ret void
+;
+entry:
+ br label %for.body
+
+for.body:
+ %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds i64, ptr %a, i64 %iv
+ store i64 %iv, ptr %arrayidx, align 8
+ %iv.next = add nuw nsw i64 %iv, 1
+ %exitcond.not = icmp eq i64 %iv.next, %N
+ br i1 %exitcond.not, label %for.cond.cleanup, label %for.body
+
+for.cond.cleanup:
+ ret void
+}
+
+; Check loops having VPWidenPointerInductionRecipe
+define void @foo2(ptr noalias %a, ptr noalias %b, i64 %N) {
+; CHECK-LABEL: define void @foo2(
+; CHECK-SAME: ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], i64 [[N:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = sub i64 -1, [[N]]
+; CHECK-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP2:%.*]] = mul i64 [[TMP1]], 2
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i64 [[TMP0]], [[TMP2]]
+; CHECK-NEXT: br i1 [[TMP3]], label [[SCALAR_PH:%.*]], label [[ENTRY:%.*]]
+; CHECK: vector.ph:
+; CHECK-NEXT: [[TMP4:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP5:%.*]] = mul i64 [[TMP4]], 2
+; CHECK-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP7:%.*]] = mul i64 [[TMP6]], 2
+; CHECK-NEXT: [[TMP8:%.*]] = sub i64 [[TMP7]], 1
+; CHECK-NEXT: [[N_RND_UP:%.*]] = add i64 [[N]], [[TMP8]]
+; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP5]]
+; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
+; CHECK-NEXT: [[TMP9:%.*]] = mul i64 [[N_VEC]], 8
+; CHECK-NEXT: [[IND_END:%.*]] = getelementptr i8, ptr [[B]], i64 [[TMP9]]
+; CHECK-NEXT: [[TMP10:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP11:%.*]] = mul i64 [[TMP10]], 2
+; CHECK-NEXT: br label [[FOR_BODY:%.*]]
+; CHECK: vector.body:
+; CHECK-NEXT: [[POINTER_PHI:%.*]] = phi ptr [ [[B]], [[ENTRY]] ], [ [[PTR_IND:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[INDEX_EVL_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[TMP12:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP13:%.*]] = mul i64 [[TMP12]], 2
+; CHECK-NEXT: [[TMP14:%.*]] = mul i64 [[TMP13]], 1
+; CHECK-NEXT: [[TMP15:%.*]] = mul i64 8, [[TMP14]]
+; CHECK-NEXT: [[TMP16:%.*]] = mul i64 [[TMP13]], 0
+; CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[TMP16]], i64 0
+; CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+; CHECK-NEXT: [[TMP17:%.*]] = call <vscale x 2 x i64> @llvm.experimental.stepvector.nxv2i64()
+; CHECK-NEXT: [[TMP18:%.*]] = add <vscale x 2 x i64> [[DOTSPLAT]], [[TMP17]]
+; CHECK-NEXT: [[VECTOR_GEP:%.*]] = mul <vscale x 2 x i64> [[TMP18]], shufflevector (<vscale x 2 x i64> insertelement (<vscale x 2 x i64> poison, i64 8, i64 0), <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer)
+; CHECK-NEXT: [[TMP19:%.*]] = getelementptr i8, ptr [[POINTER_PHI]], <vscale x 2 x i64> [[VECTOR_GEP]]
+; CHECK-NEXT: [[TMP20:%.*]] = sub i64 [[N]], [[EVL_BASED_IV]]
+; CHECK-NEXT: [[TMP21:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[TMP20]], i32 2, i1 true)
+; CHECK-NEXT: [[TMP22:%.*]] = add i64 [[EVL_BASED_IV]], 0
+; CHECK-NEXT: [[TMP23:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[TMP22]]
+; CHECK-NEXT: [[TMP24:%.*]] = getelementptr inbounds ptr, ptr [[TMP23]], i32 0
+; CHECK-NEXT: call void @llvm.vp.store.nxv2p0.p0(<vscale x 2 x ptr> [[TMP19]], ptr align 8 [[TMP24]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer), i32 [[TMP21]])
+; CHECK-NEXT: [[TMP25:%.*]] = zext i32 [[TMP21]] to i64
+; CHECK-NEXT: [[INDEX_EVL_NEXT]] = add i64 [[TMP25]], [[EVL_BASED_IV]]
+; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], [[TMP11]]
+; CHECK-NEXT: [[PTR_IND]] = getelementptr i8, ptr [[POINTER_PHI]], i64 [[TMP15]]
+; CHECK-NEXT: [[TMP26:%.*]] = icmp eq i64 [[IV_NEXT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[TMP26]], label [[MIDDLE_BLOCK:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
+; CHECK: middle.block:
+; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]]
+; CHECK: scalar.ph:
+; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY1:%.*]] ]
+; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[B]], [[ENTRY1]] ]
+; CHECK-NEXT: br label [[FOR_BODY1:%.*]]
+; CHECK: for.body:
+; CHECK-NEXT: [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT1:%.*]], [[FOR_BODY1]] ]
+; CHECK-NEXT: [[ADDR:%.*]] = phi ptr [ [[INCDEC_PTR:%.*]], [[FOR_BODY1]] ], [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ]
+; CHECK-NEXT: [[INCDEC_PTR]] = getelementptr inbounds i8, ptr [[ADDR]], i64 8
+; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV1]]
+; CHECK-NEXT: store ptr [[ADDR]], ptr [[ARRAYIDX]], align 8
+; CHECK-NEXT: [[IV_NEXT1]] = add nuw nsw i64 [[IV1]], 1
+; CHECK-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT1]], [[N]]
+; CHECK-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_COND_CLEANUP]], label [[FOR_BODY1]], !llvm.loop [[LOOP5:![0-9]+]]
+; CHECK: for.cond.cleanup:
+; CHECK-NEXT: ret void
+;
+entry:
+ br label %for.body
+
+for.body:
+ %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
+ %addr = phi ptr [ %incdec.ptr, %for.body ], [ %b, %entry ]
+ %incdec.ptr = getelementptr inbounds i8, ptr %addr, i64 8
+ %arrayidx = getelementptr inbounds i64, ptr %a, i64 %iv
+ store ptr %addr, ptr %arrayidx, align 8
+ %iv.next = add nuw nsw i64 %iv, 1
+ %exitcond.not = icmp eq i64 %iv.next, %N
+ br i1 %exitcond.not, label %for.cond.cleanup, label %for.body
+
+for.cond.cleanup:
+ ret void
+}
+
+; Check loops having VPVectorPointerRecipe that access in reverse order
+define void @foo3(ptr noalias %a, i64 %N) {
+; CHECK-LABEL: define void @foo3(
+; CHECK-SAME: ptr noalias [[A:%.*]], i64 [[N:%.*]]) #[[ATTR0]] {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: [[TMP0:%.*]] = sub i64 -1, [[N]]
+; CHECK-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP2:%.*]] = mul i64 [[TMP1]], 2
+; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i64 [[TMP0]], [[TMP2]]
+; CHECK-NEXT: br i1 [[TMP3]], label [[SCALAR_PH:%.*]], label [[ENTRY:%.*]]
+; CHECK: vector.ph:
+; CHECK-NEXT: [[TMP4:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP5:%.*]] = mul i64 [[TMP4]], 2
+; CHECK-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP7:%.*]] = mul i64 [[TMP6]], 2
+; CHECK-NEXT: [[TMP8:%.*]] = sub i64 [[TMP7]], 1
+; CHECK-NEXT: [[N_RND_UP:%.*]] = add i64 [[N]], [[TMP8]]
+; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP5]]
+; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
+; CHECK-NEXT: [[TMP9:%.*]] = mul i64 [[N_VEC]], -8
+; CHECK-NEXT: [[IND_END:%.*]] = getelementptr i8, ptr [[A]], i64 [[TMP9]]
+; CHECK-NEXT: [[TRIP_COUNT_MINUS_1:%.*]] = sub i64 [[N]], 1
+; CHECK-NEXT: [[TMP10:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP11:%.*]] = mul i64 [[TMP10]], 2
+; CHECK-NEXT: [[BROADCAST_SPLATINSERT2:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[TRIP_COUNT_MINUS_1]], i64 0
+; CHECK-NEXT: [[BROADCAST_SPLAT3:%.*]] = shufflevector <vscale x 2 x i64> [[BROADCAST_SPLATINSERT2]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+; CHECK-NEXT: [[BROADCAST_SPLATINSERT4:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[N]], i64 0
+; CHECK-NEXT: [[BROADCAST_SPLAT5:%.*]] = shufflevector <vscale x 2 x i64> [[BROADCAST_SPLATINSERT4]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+; CHECK-NEXT: br label [[FOR_BODY:%.*]]
+; CHECK: vector.body:
+; CHECK-NEXT: [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[INDEX_EVL_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[OFFSET_IDX:%.*]] = mul i64 [[EVL_BASED_IV]], -8
+; CHECK-NEXT: [[TMP12:%.*]] = add i64 [[OFFSET_IDX]], 0
+; CHECK-NEXT: [[NEXT_GEP:%.*]] = getelementptr i8, ptr [[A]], i64 [[TMP12]]
+; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[EVL_BASED_IV]], i64 0
+; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+; CHECK-NEXT: [[TMP13:%.*]] = call <vscale x 2 x i64> @llvm.experimental.stepvector.nxv2i64()
+; CHECK-NEXT: [[TMP14:%.*]] = add <vscale x 2 x i64> zeroinitializer, [[TMP13]]
+; CHECK-NEXT: [[VEC_IV:%.*]] = add <vscale x 2 x i64> [[BROADCAST_SPLAT]], [[TMP14]]
+; CHECK-NEXT: [[TMP15:%.*]] = icmp ule <vscale x 2 x i64> [[VEC_IV]], [[BROADCAST_SPLAT3]]
+; CHECK-NEXT: [[TMP16:%.*]] = call i64 @llvm.vscale.i64()
+; CHECK-NEXT: [[TMP17:%.*]] = mul i64 [[TMP16]], 2
+; CHECK-NEXT: [[TMP18:%.*]] = mul i64 0, [[TMP17]]
+; CHECK-NEXT: [[TMP19:%.*]] = sub i64 1, [[TMP17]]
+; CHECK-NEXT: [[TMP20:%.*]] = getelementptr i64, ptr [[NEXT_GEP]], i64 [[TMP18]]
+; CHECK-NEXT: [[TMP21:%.*]] = getelementptr i64, ptr [[TMP20]], i64 [[TMP19]]
+; CHECK-NEXT: [[REVERSE:%.*]] = call <vscale x 2 x i1> @llvm.experimental.vector.reverse.nxv2i1(<vscale x 2 x i1> [[TMP15]])
+; CHECK-NEXT: [[REVERSE6:%.*]] = call <vscale x 2 x i64> @llvm.experimental.vector.reverse.nxv2i64(<vscale x 2 x i64> [[BROADCAST_SPLAT5]])
+; CHECK-NEXT: call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[REVERSE6]], ptr [[TMP21]], i32 8, <vscale x 2 x i1> [[REVERSE]])
+; CHECK-NEXT: [[INDEX_EVL_NEXT]] = add i64 [[EVL_BASED_IV]], [[TMP11]]
+; CHECK-NEXT: [[TMP22:%.*]] = icmp eq i64 [[INDEX_EVL_NEXT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[TMP22]], label [[MIDDLE_BLOCK:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
+; CHECK: middle.block:
+; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]]
+; CHECK: scalar.ph:
+; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY1:%.*]] ]
+; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[A]], [[ENTRY1]] ]
+; CHECK-NEXT: br label [[FOR_BODY1:%.*]]
+; CHECK: for.body:
+; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY1]] ]
+; CHECK-NEXT: [[ADDR:%.*]] = phi ptr [ [[INCDEC_PTR:%.*]], [[FOR_BODY1]] ], [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ]
+; CHECK-NEXT: [[INCDEC_PTR]] = getelementptr inbounds i8, ptr [[ADDR]], i64 -8
+; CHECK-NEXT: store i64 [[N]], ptr [[ADDR]], align 8
+; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
+; CHECK-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], [[N]]
+; CHECK-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_COND_CLEANUP]], label [[FOR_BODY1]], !llvm.loop [[LOOP7:![0-9]+]]
+; CHECK: for.cond.cleanup:
+; CHECK-NEXT: ret void
+;
+entry:
+ br label %for.body
+
+for.body:
+ %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
+ %addr = phi ptr [ %incdec.ptr, %for.body ], [ %a, %entry ]
+ %incdec.ptr = getelementptr inbounds i8, ptr %addr, i64 -8
+ store i64 %N, ptr %addr, align 8
+ %iv.next = add nuw nsw i64 %iv, 1
+ %exitcond.not = icmp eq i64 %iv.next, %N
+ br i1 %exitcond.not, label %for.cond.cleanup, label %for.body
+
+for.cond.cleanup:
+ ret void
+}
+;.
+; CHECK: [[LOOP0]] = distinct !{[[LOOP0]], [[META1:![0-9]+]], [[META2:![0-9]+]]}
+; CHECK: [[META1]] = !{!"llvm.loop.isvectorized", i32 1}
+; CHECK: [[META2]] = !{!"llvm.loop.unroll.runtime.disable"}
+; CHECK: [[LOOP3]] = distinct !{[[LOOP3]], [[META2]], [[META1]]}
+; CHECK: [[LOOP4]] = distinct !{[[LOOP4]], [[META1]], [[META2]]}
+; CHECK: [[LOOP5]] = distinct !{[[LOOP5]], [[META2]], [[META1]]}
+; CHECK: [[LOOP6]] = distinct !{[[LOOP6]], [[META1]], [[META2]]}
+; CHECK: [[LOOP7]] = distinct !{[[LOOP7]], [[META2]], [[META1]]}
+;.
>From feb879030331b7bfc8631aff017db92320215c2b Mon Sep 17 00:00:00 2001
From: ShihPo Hung <shihpo.hung at sifive.com>
Date: Wed, 15 May 2024 04:18:06 -0700
Subject: [PATCH 2/2] Address comments
---
.../RISCV/evl-compatible-loops.ll | 167 +++++-------------
1 file changed, 40 insertions(+), 127 deletions(-)
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/evl-compatible-loops.ll b/llvm/test/Transforms/LoopVectorize/RISCV/evl-compatible-loops.ll
index e076ac2ca9f85..d550856ac4b2c 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/evl-compatible-loops.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/evl-compatible-loops.ll
@@ -1,18 +1,18 @@
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
; RUN: opt -passes=loop-vectorize -force-tail-folding-style=data-with-evl \
; RUN: -prefer-predicate-over-epilogue=predicate-dont-vectorize \
-; RUN: -force-target-instruction-cost=1 -mtriple=riscv64 -mattr=+v -S < %s | FileCheck %s
+; RUN: -mtriple=riscv64 -mattr=+v -S < %s | FileCheck %s
-; Check loops having VPWidenIntOrFpInductionRecipe
-define void @foo(ptr noalias %a, i64 %N) {
-; CHECK-LABEL: define void @foo(
+; Check loops having VPWidenIntOrFpInductionRecipe.
+define void @test_wide_integer_induction(ptr noalias %a, i64 %N) {
+; CHECK-LABEL: define void @test_wide_integer_induction(
; CHECK-SAME: ptr noalias [[A:%.*]], i64 [[N:%.*]]) #[[ATTR0:[0-9]+]] {
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = sub i64 -1, [[N]]
; CHECK-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP2:%.*]] = mul i64 [[TMP1]], 2
; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i64 [[TMP0]], [[TMP2]]
-; CHECK-NEXT: br i1 [[TMP3]], label [[SCALAR_PH:%.*]], label [[ENTRY:%.*]]
+; CHECK-NEXT: br i1 [[TMP3]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[TMP4:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP5:%.*]] = mul i64 [[TMP4]], 2
@@ -33,11 +33,11 @@ define void @foo(ptr noalias %a, i64 %N) {
; CHECK-NEXT: [[TMP16:%.*]] = mul i64 1, [[TMP15]]
; CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[TMP16]], i64 0
; CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
-; CHECK-NEXT: br label [[FOR_BODY:%.*]]
+; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
-; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT: [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[INDEX_EVL_NEXT:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT: [[VEC_IND:%.*]] = phi <vscale x 2 x i64> [ [[INDUCTION]], [[ENTRY]] ], [ [[VEC_IND_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_EVL_NEXT:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_IND:%.*]] = phi <vscale x 2 x i64> [ [[INDUCTION]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP17:%.*]] = sub i64 [[N]], [[EVL_BASED_IV]]
; CHECK-NEXT: [[TMP18:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[TMP17]], i32 2, i1 true)
; CHECK-NEXT: [[TMP19:%.*]] = add i64 [[EVL_BASED_IV]], 0
@@ -46,22 +46,22 @@ define void @foo(ptr noalias %a, i64 %N) {
; CHECK-NEXT: call void @llvm.vp.store.nxv2i64.p0(<vscale x 2 x i64> [[VEC_IND]], ptr align 8 [[TMP21]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer), i32 [[TMP18]])
; CHECK-NEXT: [[TMP22:%.*]] = zext i32 [[TMP18]] to i64
; CHECK-NEXT: [[INDEX_EVL_NEXT]] = add i64 [[TMP22]], [[EVL_BASED_IV]]
-; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], [[TMP10]]
+; CHECK-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP10]]
; CHECK-NEXT: [[VEC_IND_NEXT]] = add <vscale x 2 x i64> [[VEC_IND]], [[DOTSPLAT]]
-; CHECK-NEXT: [[TMP23:%.*]] = icmp eq i64 [[IV_NEXT]], [[N_VEC]]
-; CHECK-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
+; CHECK-NEXT: [[TMP23:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[TMP23]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
-; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY1:%.*]] ]
-; CHECK-NEXT: br label [[FOR_BODY1:%.*]]
+; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
+; CHECK-NEXT: br label [[FOR_BODY:%.*]]
; CHECK: for.body:
-; CHECK-NEXT: [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT1:%.*]], [[FOR_BODY1]] ]
-; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV1]]
-; CHECK-NEXT: store i64 [[IV1]], ptr [[ARRAYIDX]], align 8
-; CHECK-NEXT: [[IV_NEXT1]] = add nuw nsw i64 [[IV1]], 1
-; CHECK-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT1]], [[N]]
-; CHECK-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_COND_CLEANUP]], label [[FOR_BODY1]], !llvm.loop [[LOOP3:![0-9]+]]
+; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]]
+; CHECK-NEXT: store i64 [[IV]], ptr [[ARRAYIDX]], align 8
+; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
+; CHECK-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], [[N]]
+; CHECK-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_COND_CLEANUP]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK: for.cond.cleanup:
; CHECK-NEXT: ret void
;
@@ -80,16 +80,16 @@ for.cond.cleanup:
ret void
}
-; Check loops having VPWidenPointerInductionRecipe
-define void @foo2(ptr noalias %a, ptr noalias %b, i64 %N) {
-; CHECK-LABEL: define void @foo2(
+; Check loops having VPWidenPointerInductionRecipe.
+define void @test_wide_ptr_induction(ptr noalias %a, ptr noalias %b, i64 %N) {
+; CHECK-LABEL: define void @test_wide_ptr_induction(
; CHECK-SAME: ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], i64 [[N:%.*]]) #[[ATTR0]] {
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = sub i64 -1, [[N]]
; CHECK-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP2:%.*]] = mul i64 [[TMP1]], 2
; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i64 [[TMP0]], [[TMP2]]
-; CHECK-NEXT: br i1 [[TMP3]], label [[SCALAR_PH:%.*]], label [[ENTRY:%.*]]
+; CHECK-NEXT: br i1 [[TMP3]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[TMP4:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP5:%.*]] = mul i64 [[TMP4]], 2
@@ -103,11 +103,11 @@ define void @foo2(ptr noalias %a, ptr noalias %b, i64 %N) {
; CHECK-NEXT: [[IND_END:%.*]] = getelementptr i8, ptr [[B]], i64 [[TMP9]]
; CHECK-NEXT: [[TMP10:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP11:%.*]] = mul i64 [[TMP10]], 2
-; CHECK-NEXT: br label [[FOR_BODY:%.*]]
+; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
-; CHECK-NEXT: [[POINTER_PHI:%.*]] = phi ptr [ [[B]], [[ENTRY]] ], [ [[PTR_IND:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT: [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[INDEX_EVL_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[POINTER_PHI:%.*]] = phi ptr [ [[B]], [[VECTOR_PH]] ], [ [[PTR_IND:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
+; CHECK-NEXT: [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_EVL_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP12:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP13:%.*]] = mul i64 [[TMP12]], 2
; CHECK-NEXT: [[TMP14:%.*]] = mul i64 [[TMP13]], 1
@@ -127,25 +127,25 @@ define void @foo2(ptr noalias %a, ptr noalias %b, i64 %N) {
; CHECK-NEXT: call void @llvm.vp.store.nxv2p0.p0(<vscale x 2 x ptr> [[TMP19]], ptr align 8 [[TMP24]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer), i32 [[TMP21]])
; CHECK-NEXT: [[TMP25:%.*]] = zext i32 [[TMP21]] to i64
; CHECK-NEXT: [[INDEX_EVL_NEXT]] = add i64 [[TMP25]], [[EVL_BASED_IV]]
-; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], [[TMP11]]
+; CHECK-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP11]]
; CHECK-NEXT: [[PTR_IND]] = getelementptr i8, ptr [[POINTER_PHI]], i64 [[TMP15]]
-; CHECK-NEXT: [[TMP26:%.*]] = icmp eq i64 [[IV_NEXT]], [[N_VEC]]
-; CHECK-NEXT: br i1 [[TMP26]], label [[MIDDLE_BLOCK:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
+; CHECK-NEXT: [[TMP26:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[TMP26]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
-; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY1:%.*]] ]
-; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[B]], [[ENTRY1]] ]
-; CHECK-NEXT: br label [[FOR_BODY1:%.*]]
+; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
+; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[B]], [[ENTRY]] ]
+; CHECK-NEXT: br label [[FOR_BODY:%.*]]
; CHECK: for.body:
-; CHECK-NEXT: [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT1:%.*]], [[FOR_BODY1]] ]
-; CHECK-NEXT: [[ADDR:%.*]] = phi ptr [ [[INCDEC_PTR:%.*]], [[FOR_BODY1]] ], [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ]
+; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT: [[ADDR:%.*]] = phi ptr [ [[INCDEC_PTR:%.*]], [[FOR_BODY]] ], [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ]
; CHECK-NEXT: [[INCDEC_PTR]] = getelementptr inbounds i8, ptr [[ADDR]], i64 8
-; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV1]]
+; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]]
; CHECK-NEXT: store ptr [[ADDR]], ptr [[ARRAYIDX]], align 8
-; CHECK-NEXT: [[IV_NEXT1]] = add nuw nsw i64 [[IV1]], 1
-; CHECK-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT1]], [[N]]
-; CHECK-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_COND_CLEANUP]], label [[FOR_BODY1]], !llvm.loop [[LOOP5:![0-9]+]]
+; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
+; CHECK-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], [[N]]
+; CHECK-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_COND_CLEANUP]], label [[FOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
; CHECK: for.cond.cleanup:
; CHECK-NEXT: ret void
;
@@ -162,91 +162,6 @@ for.body:
%exitcond.not = icmp eq i64 %iv.next, %N
br i1 %exitcond.not, label %for.cond.cleanup, label %for.body
-for.cond.cleanup:
- ret void
-}
-
-; Check loops having VPVectorPointerRecipe that access in reverse order
-define void @foo3(ptr noalias %a, i64 %N) {
-; CHECK-LABEL: define void @foo3(
-; CHECK-SAME: ptr noalias [[A:%.*]], i64 [[N:%.*]]) #[[ATTR0]] {
-; CHECK-NEXT: entry:
-; CHECK-NEXT: [[TMP0:%.*]] = sub i64 -1, [[N]]
-; CHECK-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
-; CHECK-NEXT: [[TMP2:%.*]] = mul i64 [[TMP1]], 2
-; CHECK-NEXT: [[TMP3:%.*]] = icmp ult i64 [[TMP0]], [[TMP2]]
-; CHECK-NEXT: br i1 [[TMP3]], label [[SCALAR_PH:%.*]], label [[ENTRY:%.*]]
-; CHECK: vector.ph:
-; CHECK-NEXT: [[TMP4:%.*]] = call i64 @llvm.vscale.i64()
-; CHECK-NEXT: [[TMP5:%.*]] = mul i64 [[TMP4]], 2
-; CHECK-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
-; CHECK-NEXT: [[TMP7:%.*]] = mul i64 [[TMP6]], 2
-; CHECK-NEXT: [[TMP8:%.*]] = sub i64 [[TMP7]], 1
-; CHECK-NEXT: [[N_RND_UP:%.*]] = add i64 [[N]], [[TMP8]]
-; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP5]]
-; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
-; CHECK-NEXT: [[TMP9:%.*]] = mul i64 [[N_VEC]], -8
-; CHECK-NEXT: [[IND_END:%.*]] = getelementptr i8, ptr [[A]], i64 [[TMP9]]
-; CHECK-NEXT: [[TRIP_COUNT_MINUS_1:%.*]] = sub i64 [[N]], 1
-; CHECK-NEXT: [[TMP10:%.*]] = call i64 @llvm.vscale.i64()
-; CHECK-NEXT: [[TMP11:%.*]] = mul i64 [[TMP10]], 2
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT2:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[TRIP_COUNT_MINUS_1]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT3:%.*]] = shufflevector <vscale x 2 x i64> [[BROADCAST_SPLATINSERT2]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT4:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[N]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT5:%.*]] = shufflevector <vscale x 2 x i64> [[BROADCAST_SPLATINSERT4]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
-; CHECK-NEXT: br label [[FOR_BODY:%.*]]
-; CHECK: vector.body:
-; CHECK-NEXT: [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[INDEX_EVL_NEXT:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT: [[OFFSET_IDX:%.*]] = mul i64 [[EVL_BASED_IV]], -8
-; CHECK-NEXT: [[TMP12:%.*]] = add i64 [[OFFSET_IDX]], 0
-; CHECK-NEXT: [[NEXT_GEP:%.*]] = getelementptr i8, ptr [[A]], i64 [[TMP12]]
-; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[EVL_BASED_IV]], i64 0
-; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP13:%.*]] = call <vscale x 2 x i64> @llvm.experimental.stepvector.nxv2i64()
-; CHECK-NEXT: [[TMP14:%.*]] = add <vscale x 2 x i64> zeroinitializer, [[TMP13]]
-; CHECK-NEXT: [[VEC_IV:%.*]] = add <vscale x 2 x i64> [[BROADCAST_SPLAT]], [[TMP14]]
-; CHECK-NEXT: [[TMP15:%.*]] = icmp ule <vscale x 2 x i64> [[VEC_IV]], [[BROADCAST_SPLAT3]]
-; CHECK-NEXT: [[TMP16:%.*]] = call i64 @llvm.vscale.i64()
-; CHECK-NEXT: [[TMP17:%.*]] = mul i64 [[TMP16]], 2
-; CHECK-NEXT: [[TMP18:%.*]] = mul i64 0, [[TMP17]]
-; CHECK-NEXT: [[TMP19:%.*]] = sub i64 1, [[TMP17]]
-; CHECK-NEXT: [[TMP20:%.*]] = getelementptr i64, ptr [[NEXT_GEP]], i64 [[TMP18]]
-; CHECK-NEXT: [[TMP21:%.*]] = getelementptr i64, ptr [[TMP20]], i64 [[TMP19]]
-; CHECK-NEXT: [[REVERSE:%.*]] = call <vscale x 2 x i1> @llvm.experimental.vector.reverse.nxv2i1(<vscale x 2 x i1> [[TMP15]])
-; CHECK-NEXT: [[REVERSE6:%.*]] = call <vscale x 2 x i64> @llvm.experimental.vector.reverse.nxv2i64(<vscale x 2 x i64> [[BROADCAST_SPLAT5]])
-; CHECK-NEXT: call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[REVERSE6]], ptr [[TMP21]], i32 8, <vscale x 2 x i1> [[REVERSE]])
-; CHECK-NEXT: [[INDEX_EVL_NEXT]] = add i64 [[EVL_BASED_IV]], [[TMP11]]
-; CHECK-NEXT: [[TMP22:%.*]] = icmp eq i64 [[INDEX_EVL_NEXT]], [[N_VEC]]
-; CHECK-NEXT: br i1 [[TMP22]], label [[MIDDLE_BLOCK:%.*]], label [[FOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
-; CHECK: middle.block:
-; CHECK-NEXT: br i1 true, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]]
-; CHECK: scalar.ph:
-; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY1:%.*]] ]
-; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[A]], [[ENTRY1]] ]
-; CHECK-NEXT: br label [[FOR_BODY1:%.*]]
-; CHECK: for.body:
-; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY1]] ]
-; CHECK-NEXT: [[ADDR:%.*]] = phi ptr [ [[INCDEC_PTR:%.*]], [[FOR_BODY1]] ], [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ]
-; CHECK-NEXT: [[INCDEC_PTR]] = getelementptr inbounds i8, ptr [[ADDR]], i64 -8
-; CHECK-NEXT: store i64 [[N]], ptr [[ADDR]], align 8
-; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
-; CHECK-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], [[N]]
-; CHECK-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_COND_CLEANUP]], label [[FOR_BODY1]], !llvm.loop [[LOOP7:![0-9]+]]
-; CHECK: for.cond.cleanup:
-; CHECK-NEXT: ret void
-;
-entry:
- br label %for.body
-
-for.body:
- %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
- %addr = phi ptr [ %incdec.ptr, %for.body ], [ %a, %entry ]
- %incdec.ptr = getelementptr inbounds i8, ptr %addr, i64 -8
- store i64 %N, ptr %addr, align 8
- %iv.next = add nuw nsw i64 %iv, 1
- %exitcond.not = icmp eq i64 %iv.next, %N
- br i1 %exitcond.not, label %for.cond.cleanup, label %for.body
-
for.cond.cleanup:
ret void
}
@@ -257,6 +172,4 @@ for.cond.cleanup:
; CHECK: [[LOOP3]] = distinct !{[[LOOP3]], [[META2]], [[META1]]}
; CHECK: [[LOOP4]] = distinct !{[[LOOP4]], [[META1]], [[META2]]}
; CHECK: [[LOOP5]] = distinct !{[[LOOP5]], [[META2]], [[META1]]}
-; CHECK: [[LOOP6]] = distinct !{[[LOOP6]], [[META1]], [[META2]]}
-; CHECK: [[LOOP7]] = distinct !{[[LOOP7]], [[META2]], [[META1]]}
;.
More information about the llvm-commits
mailing list