[LV] Improve a test, regen with UTC #130092
Merged
Conversation
This file contains 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-llvm-transforms Author: Ramkumar Ramachandra (artagnon) ChangesFull diff: https://github.com/llvm/llvm-project/pull/130092.diff 1 Files Affected:
diff --git a/llvm/test/Transforms/LoopVectorize/X86/reduction-crash.ll b/llvm/test/Transforms/LoopVectorize/X86/reduction-crash.ll
index bd11562c07ff3..5f11f1ed7089f 100644
--- a/llvm/test/Transforms/LoopVectorize/X86/reduction-crash.ll
+++ b/llvm/test/Transforms/LoopVectorize/X86/reduction-crash.ll
@@ -1,35 +1,107 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
; RUN: opt -S -aa-pipeline= -passes=loop-vectorize -mcpu=prescott < %s | FileCheck %s
target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32-n8:16:32-S128"
target triple = "i386-apple-darwin"
-; PR15344
-define void @test1(ptr nocapture %arg, i32 %arg1, i1 %arg2) nounwind {
-; CHECK-LABEL: @test1(
-; CHECK: preheader
-; CHECK: insertelement <2 x double> zeroinitializer, double %tmp, i32 0
-; CHECK: vector.memcheck
+define void @pr15344(ptr noalias %ar, ptr noalias %ar2, i32 %exit.limit, i1 %cond) {
+; CHECK-LABEL: define void @pr15344(
+; CHECK-SAME: ptr noalias [[AR:%.*]], ptr noalias [[AR2:%.*]], i32 [[EXIT_LIMIT:%.*]], i1 [[COND:%.*]]) #[[ATTR0:[0-9]+]] {
+; CHECK-NEXT: [[ENTRY:.*:]]
+; CHECK-NEXT: br label %[[PH:.*]]
+; CHECK: [[PH]]:
+; CHECK-NEXT: [[LD:%.*]] = load double, ptr null, align 8
+; CHECK-NEXT: br i1 [[COND]], label %[[LOOP_PREHEADER:.*]], label %[[EXIT:.*]]
+; CHECK: [[LOOP_PREHEADER]]:
+; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[EXIT_LIMIT]], 10
+; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_MEMCHECK:.*]]
+; CHECK: [[VECTOR_MEMCHECK]]:
+; CHECK-NEXT: [[TMP0:%.*]] = shl i32 [[EXIT_LIMIT]], 2
+; CHECK-NEXT: [[SCEVGEP:%.*]] = getelementptr i8, ptr [[AR2]], i32 [[TMP0]]
+; CHECK-NEXT: [[TMP1:%.*]] = shl i32 [[EXIT_LIMIT]], 3
+; CHECK-NEXT: [[SCEVGEP1:%.*]] = getelementptr i8, ptr [[AR]], i32 [[TMP1]]
+; CHECK-NEXT: [[BOUND0:%.*]] = icmp ult ptr [[AR2]], [[SCEVGEP1]]
+; CHECK-NEXT: [[BOUND1:%.*]] = icmp ult ptr [[AR]], [[SCEVGEP]]
+; CHECK-NEXT: [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]]
+; CHECK-NEXT: br i1 [[FOUND_CONFLICT]], label %[[SCALAR_PH]], label %[[VECTOR_PH:.*]]
+; CHECK: [[VECTOR_PH]]:
+; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i32 [[EXIT_LIMIT]], 4
+; CHECK-NEXT: [[N_VEC:%.*]] = sub i32 [[EXIT_LIMIT]], [[N_MOD_VF]]
+; CHECK-NEXT: [[TMP2:%.*]] = insertelement <2 x double> zeroinitializer, double [[LD]], i32 0
+; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
+; CHECK: [[VECTOR_BODY]]:
+; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <2 x double> [ [[TMP2]], %[[VECTOR_PH]] ], [ [[TMP4:%.*]], %[[VECTOR_BODY]] ]
+; CHECK-NEXT: [[VEC_PHI2:%.*]] = phi <2 x double> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP5:%.*]], %[[VECTOR_BODY]] ]
+; CHECK-NEXT: [[TMP3:%.*]] = add i32 [[INDEX]], 0
+; CHECK-NEXT: [[TMP4]] = fadd fast <2 x double> [[VEC_PHI]], splat (double 1.000000e+00)
+; CHECK-NEXT: [[TMP5]] = fadd fast <2 x double> [[VEC_PHI2]], splat (double 1.000000e+00)
+; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds float, ptr [[AR2]], i32 [[TMP3]]
+; CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i32 0
+; CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds float, ptr [[TMP6]], i32 2
+; CHECK-NEXT: store <2 x float> splat (float 2.000000e+00), ptr [[TMP7]], align 4, !alias.scope [[META0:![0-9]+]], !noalias [[META3:![0-9]+]]
+; CHECK-NEXT: store <2 x float> splat (float 2.000000e+00), ptr [[TMP8]], align 4, !alias.scope [[META0]], !noalias [[META3]]
+; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
+; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[TMP9]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
+; CHECK: [[MIDDLE_BLOCK]]:
+; CHECK-NEXT: [[BIN_RDX:%.*]] = fadd fast <2 x double> [[TMP5]], [[TMP4]]
+; CHECK-NEXT: [[TMP10:%.*]] = call fast double @llvm.vector.reduce.fadd.v2f64(double 0.000000e+00, <2 x double> [[BIN_RDX]])
+; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 [[EXIT_LIMIT]], [[N_VEC]]
+; CHECK-NEXT: br i1 [[CMP_N]], label %[[EXIT_LOOPEXIT:.*]], label %[[SCALAR_PH]]
+; CHECK: [[SCALAR_PH]]:
+; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi double [ [[TMP10]], %[[MIDDLE_BLOCK]] ], [ [[LD]], %[[LOOP_PREHEADER]] ], [ [[LD]], %[[VECTOR_MEMCHECK]] ]
+; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], %[[MIDDLE_BLOCK]] ], [ 0, %[[LOOP_PREHEADER]] ], [ 0, %[[VECTOR_MEMCHECK]] ]
+; CHECK-NEXT: br label %[[LOOP:.*]]
+; CHECK: [[LOOP]]:
+; CHECK-NEXT: [[RDX:%.*]] = phi double [ [[FADD:%.*]], %[[LOOP]] ], [ [[BC_MERGE_RDX]], %[[SCALAR_PH]] ]
+; CHECK-NEXT: [[IV:%.*]] = phi i32 [ [[IV_NEXT:%.*]], %[[LOOP]] ], [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ]
+; CHECK-NEXT: [[GEP_AR:%.*]] = getelementptr inbounds [16 x double], ptr [[AR]], i32 0, i32 [[IV]]
+; CHECK-NEXT: [[LD2:%.*]] = load double, ptr [[GEP_AR]], align 4
+; CHECK-NEXT: [[IV_NEXT]] = add nsw i32 [[IV]], 1
+; CHECK-NEXT: [[FADD]] = fadd fast double [[RDX]], 1.000000e+00
+; CHECK-NEXT: [[GEP_AR2:%.*]] = getelementptr inbounds float, ptr [[AR2]], i32 [[IV]]
+; CHECK-NEXT: store float 2.000000e+00, ptr [[GEP_AR2]], align 4
+; CHECK-NEXT: [[EXIT_COND:%.*]] = icmp eq i32 [[IV_NEXT]], [[EXIT_LIMIT]]
+; CHECK-NEXT: br i1 [[EXIT_COND]], label %[[EXIT_LOOPEXIT]], label %[[LOOP]], !llvm.loop [[LOOP8:![0-9]+]]
+; CHECK: [[EXIT_LOOPEXIT]]:
+; CHECK-NEXT: [[FADD_LCSSA:%.*]] = phi double [ [[FADD]], %[[LOOP]] ], [ [[TMP10]], %[[MIDDLE_BLOCK]] ]
+; CHECK-NEXT: br label %[[EXIT]]
+; CHECK: [[EXIT]]:
+; CHECK-NEXT: [[RET:%.*]] = phi double [ [[LD]], %[[PH]] ], [ [[FADD_LCSSA]], %[[EXIT_LOOPEXIT]] ]
+; CHECK-NEXT: ret void
+;
+entry:
+ br label %ph
-bb:
- br label %bb2
+ph:
+ %ld = load double, ptr null, align 8
+ br i1 %cond, label %loop, label %exit
-bb2: ; preds = %bb
- %tmp = load double, ptr null, align 8
- br i1 %arg2, label %bb3, label %bb12
+loop:
+ %rdx = phi double [ %fadd, %loop ], [ %ld, %ph ]
+ %iv = phi i32 [ %iv.next, %loop ], [ 0, %ph ]
+ %gep.ar = getelementptr inbounds [16 x double], ptr %ar, i32 0, i32 %iv
+ %ld2 = load double, ptr %gep.ar, align 4
+ %iv.next = add nsw i32 %iv, 1
+ %fadd = fadd fast double %rdx, 1.0
+ %gep.ar2 = getelementptr inbounds float, ptr %ar2, i32 %iv
+ store float 2.0, ptr %gep.ar2, align 4
+ %exit.cond = icmp eq i32 %iv.next, %exit.limit
+ br i1 %exit.cond, label %exit, label %loop
-bb3: ; preds = %bb3, %bb2
- %tmp4 = phi double [ %tmp9, %bb3 ], [ %tmp, %bb2 ]
- %tmp5 = phi i32 [ %tmp8, %bb3 ], [ 0, %bb2 ]
- %tmp6 = getelementptr inbounds [16 x double], ptr undef, i32 0, i32 %tmp5
- %tmp7 = load double, ptr %tmp6, align 4
- %tmp8 = add nsw i32 %tmp5, 1
- %tmp9 = fadd fast double %tmp4, undef
- %tmp10 = getelementptr inbounds float, ptr %arg, i32 %tmp5
- store float undef, ptr %tmp10, align 4
- %tmp11 = icmp eq i32 %tmp8, %arg1
- br i1 %tmp11, label %bb12, label %bb3
-
-bb12: ; preds = %bb3, %bb2
- %tmp13 = phi double [ %tmp, %bb2 ], [ %tmp9, %bb3 ]
+exit:
+ %ret = phi double [ %ld, %ph ], [ %fadd, %loop ]
ret void
}
+;.
+; CHECK: [[META0]] = !{[[META1:![0-9]+]]}
+; CHECK: [[META1]] = distinct !{[[META1]], [[META2:![0-9]+]]}
+; 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]]}
+;.
|
|
Gentle ping. |
fhahn
reviewed
Mar 25, 2025
artagnon
force-pushed
the
lv-reduction-crash-utc
branch
from
6cee18b to
e57f686
Compare
fhahn
approved these changes
Mar 31, 2025
artagnon
force-pushed
the
lv-reduction-crash-utc
branch
from
e57f686 to
225f150
Compare
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/10/builds/2603 Here is the relevant piece of the build log for the reference |
Please ignore this fail. Not sure what happened here. Looking into it. |
Ankur-0429
pushed a commit
to Ankur-0429/llvm-project
that referenced
this pull request
Apr 2, 2025
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.
No description provided.