1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
|
From cf00b30288c4c81b2c6a5af01c38f236148777a0 Mon Sep 17 00:00:00 2001
From: Mike Hommey <mh@glandium.org>
Date: Tue, 28 Mar 2023 06:13:36 +0900
Subject: [PATCH] Revert "[Passes][VectorCombine] enable early run generally
and try load folds"
This reverts commit 163bb6d64e5f1220777c3ec2a8b58c0666a74d91.
It causes various reftest regressions.
---
llvm/lib/Passes/PassBuilderPipelines.cpp | 7 ++++---
llvm/lib/Transforms/Vectorize/VectorCombine.cpp | 8 ++------
llvm/test/Other/new-pm-defaults.ll | 2 +-
.../Other/new-pm-thinlto-postlink-defaults.ll | 1 -
.../Other/new-pm-thinlto-postlink-pgo-defaults.ll | 1 -
.../new-pm-thinlto-postlink-samplepgo-defaults.ll | 1 -
.../Other/new-pm-thinlto-prelink-pgo-defaults.ll | 1 -
.../new-pm-thinlto-prelink-samplepgo-defaults.ll | 1 -
.../PhaseOrdering/X86/vec-load-combine.ll | 15 +++++++++++----
9 files changed, 18 insertions(+), 19 deletions(-)
diff --git a/llvm/lib/Passes/PassBuilderPipelines.cpp b/llvm/lib/Passes/PassBuilderPipelines.cpp
index eed29c25714b..b925448cd6c0 100644
--- a/llvm/lib/Passes/PassBuilderPipelines.cpp
+++ b/llvm/lib/Passes/PassBuilderPipelines.cpp
@@ -611,9 +611,10 @@ PassBuilder::buildFunctionSimplificationPipeline(OptimizationLevel Level,
// Delete small array after loop unroll.
FPM.addPass(SROAPass(SROAOptions::ModifyCFG));
- // Try vectorization/scalarization transforms that are both improvements
- // themselves and can allow further folds with GVN and InstCombine.
- FPM.addPass(VectorCombinePass(/*TryEarlyFoldsOnly=*/true));
+ // The matrix extension can introduce large vector operations early, which can
+ // benefit from running vector-combine early on.
+ if (EnableMatrix)
+ FPM.addPass(VectorCombinePass(/*TryEarlyFoldsOnly=*/true));
// Eliminate redundancies.
FPM.addPass(MergedLoadStoreMotionPass());
diff --git a/llvm/lib/Transforms/Vectorize/VectorCombine.cpp b/llvm/lib/Transforms/Vectorize/VectorCombine.cpp
index 2e489757ebc1..810a9f92bb7a 100644
--- a/llvm/lib/Transforms/Vectorize/VectorCombine.cpp
+++ b/llvm/lib/Transforms/Vectorize/VectorCombine.cpp
@@ -1715,23 +1715,6 @@ bool VectorCombine::run() {
bool IsFixedVectorType = isa<FixedVectorType>(I.getType());
auto Opcode = I.getOpcode();
- // These folds should be beneficial regardless of when this pass is run
- // in the optimization pipeline.
- // The type checking is for run-time efficiency. We can avoid wasting time
- // dispatching to folding functions if there's no chance of matching.
- if (IsFixedVectorType) {
- switch (Opcode) {
- case Instruction::InsertElement:
- MadeChange |= vectorizeLoadInsert(I);
- break;
- case Instruction::ShuffleVector:
- MadeChange |= widenSubvectorLoad(I);
- break;
- default:
- break;
- }
- }
-
// This transform works with scalable and fixed vectors
// TODO: Identify and allow other scalable transforms
if (isa<VectorType>(I.getType())) {
@@ -1753,9 +1736,11 @@ bool VectorCombine::run() {
if (IsFixedVectorType) {
switch (Opcode) {
case Instruction::InsertElement:
+ MadeChange |= vectorizeLoadInsert(I);
MadeChange |= foldInsExtFNeg(I);
break;
case Instruction::ShuffleVector:
+ MadeChange |= widenSubvectorLoad(I);
MadeChange |= foldShuffleOfBinops(I);
MadeChange |= foldSelectShuffle(I);
break;
diff --git a/llvm/test/Other/new-pm-defaults.ll b/llvm/test/Other/new-pm-defaults.ll
index 13612c3bb459..5f84d28af4a6 100644
--- a/llvm/test/Other/new-pm-defaults.ll
+++ b/llvm/test/Other/new-pm-defaults.ll
@@ -186,7 +186,7 @@
; CHECK-O-NEXT: Running pass: LoopFullUnrollPass
; CHECK-EP-LOOP-END-NEXT: Running pass: NoOpLoopPass
; CHECK-O-NEXT: Running pass: SROAPass on foo
-; CHECK-O23SZ-NEXT: Running pass: VectorCombinePass
+; CHECK-MATRIX: Running pass: VectorCombinePass
; CHECK-O23SZ-NEXT: Running pass: MergedLoadStoreMotionPass
; CHECK-O23SZ-NEXT: Running pass: GVNPass
; CHECK-O23SZ-NEXT: Running analysis: MemoryDependenceAnalysis
diff --git a/llvm/test/Other/new-pm-thinlto-postlink-defaults.ll b/llvm/test/Other/new-pm-thinlto-postlink-defaults.ll
index 3f5d2d5b153d..ea07128c9f6a 100644
--- a/llvm/test/Other/new-pm-thinlto-postlink-defaults.ll
+++ b/llvm/test/Other/new-pm-thinlto-postlink-defaults.ll
@@ -159,7 +159,6 @@
; CHECK-O-NEXT: Running pass: LoopDeletionPass
; CHECK-O-NEXT: Running pass: LoopFullUnrollPass
; CHECK-O-NEXT: Running pass: SROAPass on foo
-; CHECK-O23SZ-NEXT: Running pass: VectorCombinePass
; CHECK-O23SZ-NEXT: Running pass: MergedLoadStoreMotionPass
; CHECK-O23SZ-NEXT: Running pass: GVNPass
; CHECK-O23SZ-NEXT: Running analysis: MemoryDependenceAnalysis
diff --git a/llvm/test/Other/new-pm-thinlto-postlink-pgo-defaults.ll b/llvm/test/Other/new-pm-thinlto-postlink-pgo-defaults.ll
index 29021ceace54..43e943cb6011 100644
--- a/llvm/test/Other/new-pm-thinlto-postlink-pgo-defaults.ll
+++ b/llvm/test/Other/new-pm-thinlto-postlink-pgo-defaults.ll
@@ -121,7 +121,6 @@
; CHECK-O-NEXT: Running pass: LoopDeletionPass
; CHECK-O-NEXT: Running pass: LoopFullUnrollPass
; CHECK-O-NEXT: Running pass: SROAPass on foo
-; CHECK-O23SZ-NEXT: Running pass: VectorCombinePass
; CHECK-O23SZ-NEXT: Running pass: MergedLoadStoreMotionPass
; CHECK-O23SZ-NEXT: Running pass: GVNPass
; CHECK-O23SZ-NEXT: Running analysis: MemoryDependenceAnalysis
diff --git a/llvm/test/Other/new-pm-thinlto-postlink-samplepgo-defaults.ll b/llvm/test/Other/new-pm-thinlto-postlink-samplepgo-defaults.ll
index daf3141a1f2c..78914d1c23b2 100644
--- a/llvm/test/Other/new-pm-thinlto-postlink-samplepgo-defaults.ll
+++ b/llvm/test/Other/new-pm-thinlto-postlink-samplepgo-defaults.ll
@@ -130,7 +130,6 @@
; CHECK-O-NEXT: Running pass: LoopDeletionPass
; CHECK-O-NEXT: Running pass: LoopFullUnrollPass
; CHECK-O-NEXT: Running pass: SROAPass on foo
-; CHECK-O23SZ-NEXT: Running pass: VectorCombinePass
; CHECK-O23SZ-NEXT: Running pass: MergedLoadStoreMotionPass
; CHECK-O23SZ-NEXT: Running pass: GVNPass
; CHECK-O23SZ-NEXT: Running analysis: MemoryDependenceAnalysis
diff --git a/llvm/test/Other/new-pm-thinlto-prelink-pgo-defaults.ll b/llvm/test/Other/new-pm-thinlto-prelink-pgo-defaults.ll
index bfe80902f806..5b62ba39add3 100644
--- a/llvm/test/Other/new-pm-thinlto-prelink-pgo-defaults.ll
+++ b/llvm/test/Other/new-pm-thinlto-prelink-pgo-defaults.ll
@@ -160,7 +160,6 @@
; CHECK-O-NEXT: Running pass: LoopDeletionPass
; CHECK-O-NEXT: Running pass: LoopFullUnrollPass
; CHECK-O-NEXT: Running pass: SROAPass on foo
-; CHECK-O23SZ-NEXT: Running pass: VectorCombinePass
; CHECK-O23SZ-NEXT: Running pass: MergedLoadStoreMotionPass
; CHECK-O23SZ-NEXT: Running pass: GVNPass
; CHECK-O23SZ-NEXT: Running analysis: MemoryDependenceAnalysis
diff --git a/llvm/test/Other/new-pm-thinlto-prelink-samplepgo-defaults.ll b/llvm/test/Other/new-pm-thinlto-prelink-samplepgo-defaults.ll
index c7daf7aa46b1..17475423d696 100644
--- a/llvm/test/Other/new-pm-thinlto-prelink-samplepgo-defaults.ll
+++ b/llvm/test/Other/new-pm-thinlto-prelink-samplepgo-defaults.ll
@@ -124,7 +124,6 @@
; CHECK-O-NEXT: Running pass: IndVarSimplifyPass
; CHECK-O-NEXT: Running pass: LoopDeletionPass
; CHECK-O-NEXT: Running pass: SROAPass on foo
-; CHECK-O23SZ-NEXT: Running pass: VectorCombinePass
; CHECK-O23SZ-NEXT: Running pass: MergedLoadStoreMotionPass
; CHECK-O23SZ-NEXT: Running pass: GVNPass
; CHECK-O23SZ-NEXT: Running analysis: MemoryDependenceAnalysis
diff --git a/llvm/test/Transforms/PhaseOrdering/X86/vec-load-combine.ll b/llvm/test/Transforms/PhaseOrdering/X86/vec-load-combine.ll
index 77cbc70ff369..dd7164febea4 100644
--- a/llvm/test/Transforms/PhaseOrdering/X86/vec-load-combine.ll
+++ b/llvm/test/Transforms/PhaseOrdering/X86/vec-load-combine.ll
@@ -12,13 +12,20 @@ $getAt = comdat any
define dso_local noundef <4 x float> @ConvertVectors_ByRef(ptr noundef nonnull align 16 dereferenceable(16) %0) #0 {
; SSE-LABEL: @ConvertVectors_ByRef(
; SSE-NEXT: [[TMP2:%.*]] = load <4 x float>, ptr [[TMP0:%.*]], align 16
-; SSE-NEXT: [[TMP3:%.*]] = shufflevector <4 x float> [[TMP2]], <4 x float> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 2>
-; SSE-NEXT: ret <4 x float> [[TMP3]]
+; SSE-NEXT: [[TMP3:%.*]] = getelementptr inbounds [4 x float], ptr [[TMP0]], i64 0, i64 1
+; SSE-NEXT: [[TMP4:%.*]] = load <2 x float>, ptr [[TMP3]], align 4
+; SSE-NEXT: [[TMP5:%.*]] = shufflevector <2 x float> [[TMP4]], <2 x float> poison, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+; SSE-NEXT: [[TMP6:%.*]] = shufflevector <4 x float> [[TMP2]], <4 x float> [[TMP5]], <4 x i32> <i32 0, i32 4, i32 5, i32 undef>
+; SSE-NEXT: [[TMP7:%.*]] = shufflevector <4 x float> [[TMP6]], <4 x float> [[TMP5]], <4 x i32> <i32 0, i32 1, i32 2, i32 5>
+; SSE-NEXT: ret <4 x float> [[TMP7]]
;
; AVX-LABEL: @ConvertVectors_ByRef(
; AVX-NEXT: [[TMP2:%.*]] = load <4 x float>, ptr [[TMP0:%.*]], align 16
-; AVX-NEXT: [[TMP3:%.*]] = shufflevector <4 x float> [[TMP2]], <4 x float> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 2>
-; AVX-NEXT: ret <4 x float> [[TMP3]]
+; AVX-NEXT: [[TMP3:%.*]] = getelementptr inbounds [4 x float], ptr [[TMP0]], i64 0, i64 2
+; AVX-NEXT: [[TMP4:%.*]] = load float, ptr [[TMP3]], align 8
+; AVX-NEXT: [[TMP5:%.*]] = insertelement <4 x float> [[TMP2]], float [[TMP4]], i64 2
+; AVX-NEXT: [[TMP6:%.*]] = insertelement <4 x float> [[TMP5]], float [[TMP4]], i64 3
+; AVX-NEXT: ret <4 x float> [[TMP6]]
;
%2 = alloca ptr, align 8
%3 = alloca <4 x float>, align 16
--
2.39.0.1.g6739ec1790
|
