[LV][EVL] Attach a new metadata on EVL vectorized loops #131000
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@llvm/pr-subscribers-vectorizers @llvm/pr-subscribers-llvm-transforms Author: Min-Yih Hsu (mshockwave) ChangesThis patch attaches a new metadata, This approach is much safer than, said IR pattern matching to figure out if a loop is EVL-vectorized. Patch is 81.67 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/131000.diff 11 Files Affected:
diff --git a/llvm/lib/Transforms/Vectorize/VPlan.cpp b/llvm/lib/Transforms/Vectorize/VPlan.cpp
index e595347d62bf5..7ef06957d5322 100644
--- a/llvm/lib/Transforms/Vectorize/VPlan.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlan.cpp
@@ -1013,6 +1013,32 @@ void VPlan::execute(VPTransformState *State) {
Value *Val = State->get(PhiR->getBackedgeValue(), NeedsScalar);
cast<PHINode>(Phi)->addIncoming(Val, VectorLatchBB);
}
+
+ // Check if it's EVL-vectorized and mark the corresponding metadata.
+ // Note that we could have done this during the codegen of
+ // ExplictVectorLength, but the enclosing vector loop was not in a good shape
+ // for us to attach the metadata.
+ bool IsEVLVectorized = llvm::any_of(*Header, [](const VPRecipeBase &Recipe) {
+ // Looking for the ExplictVectorLength VPInstruction.
+ if (const auto *VI = dyn_cast<VPInstruction>(&Recipe))
+ return VI->getOpcode() == VPInstruction::ExplicitVectorLength;
+ return false;
+ });
+ if (IsEVLVectorized) {
+ // VPTransformState::CurrentParentLoop has already been reset
+ // at this moment.
+ Loop *L = State->LI->getLoopFor(VectorLatchBB);
+ assert(L);
+ LLVMContext &Context = State->Builder.getContext();
+ MDNode *LoopID = L->getLoopID();
+ auto *IsEVLVectorizedMD = MDNode::get(
+ Context,
+ {MDString::get(Context, "llvm.loop.isvectorized.withevl"),
+ ConstantAsMetadata::get(ConstantInt::get(Context, APInt(32, 1)))});
+ MDNode *NewLoopID = makePostTransformationMetadata(Context, LoopID, {},
+ {IsEVLVectorizedMD});
+ L->setLoopID(NewLoopID);
+ }
}
InstructionCost VPlan::cost(ElementCount VF, VPCostContext &Ctx) {
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/truncate-to-minimal-bitwidth-evl-crash.ll b/llvm/test/Transforms/LoopVectorize/RISCV/truncate-to-minimal-bitwidth-evl-crash.ll
index ba7158eb02d90..7a28574740348 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/truncate-to-minimal-bitwidth-evl-crash.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/truncate-to-minimal-bitwidth-evl-crash.ll
@@ -53,7 +53,7 @@ define void @truncate_to_minimal_bitwidths_widen_cast_recipe(ptr %src) {
; CHECK-NEXT: store i8 [[CONV36]], ptr null, align 1
; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1
; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], 1
-; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP3:![0-9]+]]
+; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: [[EXIT]]:
; CHECK-NEXT: ret void
;
@@ -77,8 +77,9 @@ exit: ; preds = %loop
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: [[LOOP0]] = distinct !{[[LOOP0]], [[META1:![0-9]+]], [[META2:![0-9]+]], [[META3:![0-9]+]]}
+; CHECK: [[META1]] = !{!"llvm.loop.isvectorized.withevl", i32 1}
+; CHECK: [[META2]] = !{!"llvm.loop.isvectorized", i32 1}
+; CHECK: [[META3]] = !{!"llvm.loop.unroll.runtime.disable"}
+; CHECK: [[LOOP4]] = distinct !{[[LOOP4]], [[META3]], [[META2]]}
;.
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/type-info-cache-evl-crash.ll b/llvm/test/Transforms/LoopVectorize/RISCV/type-info-cache-evl-crash.ll
index c95414db18bef..8543964968c5a 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/type-info-cache-evl-crash.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/type-info-cache-evl-crash.ll
@@ -80,7 +80,7 @@ define void @type_info_cache_clobber(ptr %dstv, ptr %src, i64 %wide.trip.count)
; CHECK-NEXT: store i16 [[CONV36]], ptr null, align 2
; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1
; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], [[WIDE_TRIP_COUNT]]
-; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP8:![0-9]+]]
+; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP9:![0-9]+]]
; CHECK: [[EXIT]]:
; CHECK-NEXT: ret void
;
@@ -114,8 +114,9 @@ exit:
; CHECK: [[META2]] = distinct !{[[META2]], !"LVerDomain"}
; CHECK: [[META3]] = !{[[META4:![0-9]+]]}
; CHECK: [[META4]] = distinct !{[[META4]], [[META2]]}
-; CHECK: [[LOOP5]] = distinct !{[[LOOP5]], [[META6:![0-9]+]], [[META7:![0-9]+]]}
-; CHECK: [[META6]] = !{!"llvm.loop.isvectorized", i32 1}
-; CHECK: [[META7]] = !{!"llvm.loop.unroll.runtime.disable"}
-; CHECK: [[LOOP8]] = distinct !{[[LOOP8]], [[META6]]}
+; CHECK: [[LOOP5]] = distinct !{[[LOOP5]], [[META6:![0-9]+]], [[META7:![0-9]+]], [[META8:![0-9]+]]}
+; CHECK: [[META6]] = !{!"llvm.loop.isvectorized.withevl", i32 1}
+; CHECK: [[META7]] = !{!"llvm.loop.isvectorized", i32 1}
+; CHECK: [[META8]] = !{!"llvm.loop.unroll.runtime.disable"}
+; CHECK: [[LOOP9]] = distinct !{[[LOOP9]], [[META7]]}
;.
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/vectorize-force-tail-with-evl-bin-unary-ops-args.ll b/llvm/test/Transforms/LoopVectorize/RISCV/vectorize-force-tail-with-evl-bin-unary-ops-args.ll
index e7181f7f30c77..d962258662728 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/vectorize-force-tail-with-evl-bin-unary-ops-args.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/vectorize-force-tail-with-evl-bin-unary-ops-args.ll
@@ -65,7 +65,7 @@ define void @test_and(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP3:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP4:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -144,7 +144,7 @@ define void @test_or(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -159,7 +159,7 @@ define void @test_or(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP5:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP6:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -238,7 +238,7 @@ define void @test_xor(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -253,7 +253,7 @@ define void @test_xor(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP7:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP8:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -332,7 +332,7 @@ define void @test_shl(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP9:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -347,7 +347,7 @@ define void @test_shl(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP9:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP10:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -426,7 +426,7 @@ define void @test_lshr(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP11:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -441,7 +441,7 @@ define void @test_lshr(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP11:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP12:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -520,7 +520,7 @@ define void @test_ashr(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP13:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -535,7 +535,7 @@ define void @test_ashr(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP13:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP14:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -614,7 +614,7 @@ define void @test_add(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP15:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -629,7 +629,7 @@ define void @test_add(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP15:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP16:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -708,7 +708,7 @@ define void @test_sub(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP17:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -723,7 +723,7 @@ define void @test_sub(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP17:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP18:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -802,7 +802,7 @@ define void @test_mul(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP18:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP19:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -817,7 +817,7 @@ define void @test_mul(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP19:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP20:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -896,7 +896,7 @@ define void @test_sdiv(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP20:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP21:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -911,7 +911,7 @@ define void @test_sdiv(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP21:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP22:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -990,7 +990,7 @@ define void @test_udiv(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP22:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP23:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -1005,7 +1005,7 @@ define void @test_udiv(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP23:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP24:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -1084,7 +1084,7 @@ define void @test_srem(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP9]]
; IF-EVL-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
-; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP24:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP25:![0-9]+]]
; IF-EVL: [[MIDDLE_BLOCK]]:
; IF-EVL-NEXT: br i1 true, label %[[FINISH_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; IF-EVL: [[SCALAR_PH]]:
@@ -1099,7 +1099,7 @@ define void @test_srem(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds i8, ptr [[B]], i64 [[LEN]]
; IF-EVL-NEXT: store i8 [[TMP]], ptr [[ARRAYIDX1]], align 1
; IF-EVL-NEXT: [[DOTNOT:%.*]] = icmp eq i64 [[DEC]], 100
-; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP25:![0-9]+]]
+; IF-EVL-NEXT: br i1 [[DOTNOT]], label %[[FINISH_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP26:![0-9]+]]
; IF-EVL: [[FINISH_LOOPEXIT]]:
; IF-EVL-NEXT: ret void
;
@@ -1178,7 +1178,7 @@ define void @test_urem(ptr nocapture %a, ptr nocapture readonly %b) {
; IF-EVL-NEXT: [[INDEX_EVL_NEXT]] = add nuw i64 [[TMP18]], [[EVL_BASED_IV]]
; IF-EVL-NEXT: [[INDEX_NEXT]] = add nuw i64...
[truncated]
|
wangpc-pp
reviewed
Mar 13, 2025
fhahn
reviewed
Mar 13, 2025
| // Note that we could have done this during the codegen of | ||
| // ExplictVectorLength, but the enclosing vector loop was not in a good shape | ||
| // for us to attach the metadata. | ||
| bool IsEVLVectorized = llvm::any_of(*Header, [](const VPRecipeBase &Recipe) { |
There was a problem hiding this comment.
Suggested change
| bool IsEVLVectorized = llvm::any_of(*Header, [](const VPRecipeBase &Recipe) { | |
| bool IsEVLVectorized = any_of(*Header, [](const VPRecipeBase &Recipe) { |
Comment on lines
1027
to
1040
| if (IsEVLVectorized) { | ||
| // VPTransformState::CurrentParentLoop has already been reset | ||
| // at this moment. | ||
| Loop *L = State->LI->getLoopFor(VectorLatchBB); | ||
| assert(L); | ||
| LLVMContext &Context = State->Builder.getContext(); | ||
| MDNode *LoopID = L->getLoopID(); | ||
| auto *IsEVLVectorizedMD = MDNode::get( | ||
| Context, | ||
| {MDString::get(Context, "llvm.loop.isvectorized.withevl"), | ||
| ConstantAsMetadata::get(ConstantInt::get(Context, APInt(32, 1)))}); | ||
| MDNode *NewLoopID = makePostTransformationMetadata(Context, LoopID, {}, | ||
| {IsEVLVectorizedMD}); | ||
| L->setLoopID(NewLoopID); |
There was a problem hiding this comment.
Can we do this at the same place we set the other loop metadata?
There was a problem hiding this comment.
Ofc, I moved it to LoopVectorizeHints, same place where we set the isvectorized metadata. Because of that we now check if a VPlan is EVL-vectorized in LoopVectorizationPlanner::executePlan (instead of VPlan::execute)
lukel97
reviewed
Mar 13, 2025
fhahn
reviewed
Mar 20, 2025
Comment on lines
128
to
132
| enum TailFoldingKind { | ||
| TFK_Unspecified = -1, | ||
| /// Tail folding with explicit vector length intrinsics. | ||
| TFK_EVL = 0 | ||
| }; |
There was a problem hiding this comment.
Does that mean it would also be possible to set the tail-folding style with the hints?
Might not be a good idea to open a two-way street.
My original suggestion was mostly meant to move it to
// 2.6. Maintain Loop Hints
// Keep all loop hints from the original loop on the vector loop (we'll
In LoopVectorize.cpp
There was a problem hiding this comment.
Does that mean it would also be possible to set the tail-folding style with the hints?
Might not be a good idea to open a two-way street.
Currently LV doesn't use this metadata to make any decision w.r.t. tail-folding style, but I agree it does make such impression.
While I can move the logics to LoopVectorize.cpp as you pointed out, I'm afraid the raw metadata form -- llvm.loop.isvectorized.tailfoldingstyle = 0 -- is not really self-described, especially when accessing it (and that's one of the reasons why I wrapped it with LoopVectorizeHints::isEVLVectorized). IMHO, there are two potential ways to move forward:
- Use a string value to indicate tail-folding style in the metadata:
llvm.loop.isvectorized.tailfoldingstyle = "evl". Which was what Luke originally suggested - Use boolean-value metadata so that it's more intuitive when accessing it. Which was my original approach.
Of course, I'm open for other suggestions. What do you think @lukel97 @fhahn ?
There was a problem hiding this comment.
Does that mean it would also be possible to set the tail-folding style with the hints?
Might not be a good idea to open a two-way street.Currently LV doesn't use this metadata to make any decision w.r.t. tail-folding style, but I agree it does make such impression. While I can move the logics to LoopVectorize.cpp as you pointed out, I'm afraid the raw metadata form --
llvm.loop.isvectorized.tailfoldingstyle = 0-- is not really self-described, especially when accessing it (and that's one of the reasons why I wrapped it withLoopVectorizeHints::isEVLVectorized). IMHO, there are two potential ways to move forward:
- Use a string value to indicate tail-folding style in the metadata:
llvm.loop.isvectorized.tailfoldingstyle = "evl". Which was what Luke originally suggested- Use boolean-value metadata so that it's more intuitive when accessing it. Which was my original approach.
Of course, I'm open for other suggestions. What do you think @lukel97 @fhahn ?
I'm settling down with option (1): Use a string value to indicate tail-folding style in the metadata, llvm.loop.isvectorized.tailfoldingstyle = "evl" as Luke originally suggested.
Also, I'm now setting the metadata in LoopVectorize.cpp (instead of LoopVectorizeHints) as @fhahn suggested.
mshockwave
force-pushed
the
patch/evl-vectorized-metadata
branch
from
0ffcd14 to
727c02a
Compare
wangpc-pp
reviewed
Apr 9, 2025
And set the metadata in LoopVectorizationHints.
And use the string value to represent tail-folding style (again).
mshockwave
force-pushed
the
patch/evl-vectorized-metadata
branch
from
727c02a to
54d01d8
Compare
|
The merge conflicts are resolved now. |
alexey-bataev
approved these changes
May 6, 2025
fhahn
reviewed
May 6, 2025
|
|
||
| // Check if it's EVL-vectorized and mark the corresponding metadata. | ||
| bool IsEVLVectorized = | ||
| llvm::any_of(*HeaderVPBB, [](const VPRecipeBase &Recipe) { |
There was a problem hiding this comment.
just for future reference, llvm:: should not be needed here
GeorgeARM
pushed a commit
to GeorgeARM/llvm-project
that referenced
this pull request
May 7, 2025
This patch attaches a new metadata, `llvm.loop.isvectorized.withevl`, on loops vectorized with explicit vector length. This will help other optimizations down in the pipeline that focus on EVL-vectorized loop This approach is much safer than, said IR pattern matching to figure out if a loop is EVL-vectorized or not.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This patch attaches a new metadata,
llvm.loop.isvectorized.withevl, on loops vectorized with explicit vector length. This will help other optimizations down in the pipeline that focus on EVL-vectorized loop, like the (renovated) EVLIndVarSimplifyPass, of which patch I will post shortly.This approach is much safer than, said IR pattern matching to figure out if a loop is EVL-vectorized.