[MergeFuncs] Use sizeWithoutDebug to decide if we create a thunk #68627
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 ChangesI noticed that when we determine the size of the function to figure out if its profitable, we include debug instructions which can end up making larger functions than necessary. Full diff: https://github.com/llvm/llvm-project/pull/68627.diff 3 Files Affected:
diff --git a/llvm/lib/Transforms/IPO/MergeFunctions.cpp b/llvm/lib/Transforms/IPO/MergeFunctions.cpp
index 89ddd7b6adebbec..146228db4644045 100644
--- a/llvm/lib/Transforms/IPO/MergeFunctions.cpp
+++ b/llvm/lib/Transforms/IPO/MergeFunctions.cpp
@@ -653,7 +653,7 @@ static bool canCreateThunkFor(Function *F) {
// Don't merge tiny functions using a thunk, since it can just end up
// making the function larger.
if (F->size() == 1) {
- if (F->front().size() <= 2) {
+ if (F->front().sizeWithoutDebug() <= 2) {
LLVM_DEBUG(dbgs() << "canCreateThunkFor: " << F->getName()
<< " is too small to bother creating a thunk for\n");
return false;
diff --git a/llvm/test/Transforms/MergeFunc/mergefunc-preserve-debug-info.ll b/llvm/test/Transforms/MergeFunc/mergefunc-preserve-debug-info.ll
index 13d94f8cf6a257c..20763dc1374591a 100644
--- a/llvm/test/Transforms/MergeFunc/mergefunc-preserve-debug-info.ll
+++ b/llvm/test/Transforms/MergeFunc/mergefunc-preserve-debug-info.ll
@@ -1,3 +1,4 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 2
; RUN: opt -passes='default<O0>,mergefunc' -S -mergefunc-preserve-debug-info < %s | FileCheck %s --check-prefix=OPTIMIZATION_LEVEL_0
; RUN: opt -passes='default<O2>,mergefunc' -S -mergefunc-preserve-debug-info < %s | FileCheck %s --check-prefix=OPTIMIZATION_LEVEL_2
@@ -43,6 +44,48 @@
; Function Attrs: nounwind uwtable
define i32 @maxA(i32 %x, i32 %y) !dbg !6 {
+; OPTIMIZATION_LEVEL_0-LABEL: define i32 @maxA
+; OPTIMIZATION_LEVEL_0-SAME: (i32 [[X:%.*]], i32 [[Y:%.*]]) !dbg [[DBG6:![0-9]+]] {
+; OPTIMIZATION_LEVEL_0-NEXT: entry:
+; OPTIMIZATION_LEVEL_0-NEXT: [[X_ADDR:%.*]] = alloca i32, align 4
+; OPTIMIZATION_LEVEL_0-NEXT: [[Y_ADDR:%.*]] = alloca i32, align 4
+; OPTIMIZATION_LEVEL_0-NEXT: [[I:%.*]] = alloca i32, align 4
+; OPTIMIZATION_LEVEL_0-NEXT: [[M:%.*]] = alloca i32, align 4
+; OPTIMIZATION_LEVEL_0-NEXT: [[J:%.*]] = alloca i32, align 4
+; OPTIMIZATION_LEVEL_0-NEXT: store i32 [[X]], ptr [[X_ADDR]], align 4
+; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr [[X_ADDR]], metadata [[META11:![0-9]+]], metadata !DIExpression()), !dbg [[DBG12:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: store i32 [[Y]], ptr [[Y_ADDR]], align 4
+; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr [[Y_ADDR]], metadata [[META13:![0-9]+]], metadata !DIExpression()), !dbg [[DBG14:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr [[I]], metadata [[META15:![0-9]+]], metadata !DIExpression()), !dbg [[DBG16:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr [[M]], metadata [[META17:![0-9]+]], metadata !DIExpression()), !dbg [[DBG18:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr [[J]], metadata [[META19:![0-9]+]], metadata !DIExpression()), !dbg [[DBG20:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: [[TMP0:%.*]] = load i32, ptr [[X_ADDR]], align 4, !dbg [[DBG21:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: [[TMP1:%.*]] = load i32, ptr [[Y_ADDR]], align 4, !dbg [[DBG23:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP0]], [[TMP1]], !dbg [[DBG24:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: br i1 [[CMP]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]], !dbg [[DBG25:![0-9]+]]
+; OPTIMIZATION_LEVEL_0: if.then:
+; OPTIMIZATION_LEVEL_0-NEXT: [[TMP2:%.*]] = load i32, ptr [[X_ADDR]], align 4, !dbg [[DBG26:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: store i32 [[TMP2]], ptr [[M]], align 4, !dbg [[DBG27:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: br label [[IF_END:%.*]], !dbg [[DBG28:![0-9]+]]
+; OPTIMIZATION_LEVEL_0: if.else:
+; OPTIMIZATION_LEVEL_0-NEXT: [[TMP3:%.*]] = load i32, ptr [[Y_ADDR]], align 4, !dbg [[DBG29:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: store i32 [[TMP3]], ptr [[M]], align 4, !dbg [[DBG30:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: br label [[IF_END]]
+; OPTIMIZATION_LEVEL_0: if.end:
+; OPTIMIZATION_LEVEL_0-NEXT: [[TMP4:%.*]] = load i32, ptr [[M]], align 4, !dbg [[DBG31:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: ret i32 [[TMP4]], !dbg [[DBG32:![0-9]+]]
+;
+; OPTIMIZATION_LEVEL_2-LABEL: define i32 @maxA
+; OPTIMIZATION_LEVEL_2-SAME: (i32 [[X:%.*]], i32 [[Y:%.*]]) local_unnamed_addr #[[ATTR0:[0-9]+]] !dbg [[DBG6:![0-9]+]] {
+; OPTIMIZATION_LEVEL_2-NEXT: entry:
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.value(metadata i32 [[X]], metadata [[META11:![0-9]+]], metadata !DIExpression()), !dbg [[DBG12:![0-9]+]]
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.value(metadata i32 [[Y]], metadata [[META13:![0-9]+]], metadata !DIExpression()), !dbg [[DBG12]]
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.declare(metadata ptr undef, metadata [[META14:![0-9]+]], metadata !DIExpression()), !dbg [[DBG15:![0-9]+]]
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.declare(metadata ptr undef, metadata [[META16:![0-9]+]], metadata !DIExpression()), !dbg [[DBG17:![0-9]+]]
+; OPTIMIZATION_LEVEL_2-NEXT: [[X_Y:%.*]] = tail call i32 @llvm.smax.i32(i32 [[X]], i32 [[Y]])
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.value(metadata i32 [[X_Y]], metadata [[META18:![0-9]+]], metadata !DIExpression()), !dbg [[DBG12]]
+; OPTIMIZATION_LEVEL_2-NEXT: ret i32 [[X_Y]], !dbg [[DBG19:![0-9]+]]
+;
entry:
%x.addr = alloca i32, align 4
%y.addr = alloca i32, align 4
@@ -81,26 +124,30 @@ declare void @llvm.dbg.declare(metadata, metadata, metadata)
; Function Attrs: nounwind uwtable
define i32 @maxB(i32 %x, i32 %y) !dbg !34 {
+; OPTIMIZATION_LEVEL_0-LABEL: define i32 @maxB
+; OPTIMIZATION_LEVEL_0-SAME: (i32 [[X:%.*]], i32 [[Y:%.*]]) !dbg [[DBG33:![0-9]+]] {
+; OPTIMIZATION_LEVEL_0-NEXT: entry:
+; OPTIMIZATION_LEVEL_0-NEXT: [[X_ADDR:%.*]] = alloca i32, align 4
+; OPTIMIZATION_LEVEL_0-NEXT: [[Y_ADDR:%.*]] = alloca i32, align 4
+; OPTIMIZATION_LEVEL_0-NEXT: store i32 [[X]], ptr [[X_ADDR]], align 4
+; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr [[X_ADDR]], metadata [[META34:![0-9]+]], metadata !DIExpression()), !dbg [[DBG35:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: store i32 [[Y]], ptr [[Y_ADDR]], align 4
+; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr [[Y_ADDR]], metadata [[META36:![0-9]+]], metadata !DIExpression()), !dbg [[DBG37:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: [[TMP0:%.*]] = tail call i32 @maxA(i32 [[X]], i32 [[Y]]), !dbg [[DBG38:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: ret i32 [[TMP0]], !dbg [[DBG38]]
+;
+; OPTIMIZATION_LEVEL_2-LABEL: define i32 @maxB
+; OPTIMIZATION_LEVEL_2-SAME: (i32 [[X:%.*]], i32 [[Y:%.*]]) local_unnamed_addr #[[ATTR0]] !dbg [[DBG20:![0-9]+]] {
+; OPTIMIZATION_LEVEL_2-NEXT: entry:
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.value(metadata i32 [[X]], metadata [[META21:![0-9]+]], metadata !DIExpression()), !dbg [[DBG22:![0-9]+]]
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.value(metadata i32 [[Y]], metadata [[META23:![0-9]+]], metadata !DIExpression()), !dbg [[DBG22]]
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.declare(metadata ptr undef, metadata [[META24:![0-9]+]], metadata !DIExpression()), !dbg [[DBG25:![0-9]+]]
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.declare(metadata ptr undef, metadata [[META26:![0-9]+]], metadata !DIExpression()), !dbg [[DBG27:![0-9]+]]
+; OPTIMIZATION_LEVEL_2-NEXT: [[X_Y:%.*]] = tail call i32 @llvm.smax.i32(i32 [[X]], i32 [[Y]])
+; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.value(metadata i32 [[X_Y]], metadata [[META28:![0-9]+]], metadata !DIExpression()), !dbg [[DBG22]]
+; OPTIMIZATION_LEVEL_2-NEXT: ret i32 [[X_Y]], !dbg [[DBG29:![0-9]+]]
+;
-; OPTIMIZATION_LEVEL_0: define i32 @maxB(i32 %x, i32 %y)
-; OPTIMIZATION_LEVEL_0-NEXT: entry:
-; OPTIMIZATION_LEVEL_0-NEXT: %x.addr = alloca i32, align 4
-; OPTIMIZATION_LEVEL_0-NEXT: %y.addr = alloca i32, align 4
-; OPTIMIZATION_LEVEL_0-NEXT: store i32 %x, ptr %x.addr, align 4
-; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr %x.addr, metadata !{{[0-9]+}}, metadata !DIExpression()), !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_0-NEXT: store i32 %y, ptr %y.addr, align 4
-; OPTIMIZATION_LEVEL_0-NEXT: call void @llvm.dbg.declare(metadata ptr %y.addr, metadata !{{[0-9]+}}, metadata !DIExpression()), !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_0-NEXT: %0 = tail call i32 @maxA(i32 %x, i32 %y), !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_0-NEXT: ret i32 %0, !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_0-NEXT: }
-
-; OPTIMIZATION_LEVEL_2: define i32 @maxB(i32 %x, i32 %y)
-; OPTIMIZATION_LEVEL_2-NEXT: entry:
-; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.value(metadata i32 %x, metadata !{{[0-9]+}}, metadata !DIExpression()), !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_2-NEXT: call void @llvm.dbg.value(metadata i32 %y, metadata !{{[0-9]+}}, metadata !DIExpression()), !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_2-NEXT: %0 = tail call i32 @maxA(i32 %x, i32 %y) #{{[0-9]+}}, !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_2-NEXT: ret i32 %0, !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_2-NEXT: }
entry:
%x.addr = alloca i32, align 4
@@ -137,18 +184,19 @@ if.end: ; preds = %if.else, %if.then
; Function Attrs: nounwind uwtable
define void @f() !dbg !57 {
+; OPTIMIZATION_LEVEL_0-LABEL: define void @f
+; OPTIMIZATION_LEVEL_0-SAME: () !dbg [[DBG39:![0-9]+]] {
+; OPTIMIZATION_LEVEL_0-NEXT: entry:
+; OPTIMIZATION_LEVEL_0-NEXT: [[CALL:%.*]] = call i32 @maxA(i32 3, i32 4), !dbg [[DBG42:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: [[CALL1:%.*]] = call i32 @maxB(i32 1, i32 9), !dbg [[DBG43:![0-9]+]]
+; OPTIMIZATION_LEVEL_0-NEXT: ret void, !dbg [[DBG44:![0-9]+]]
+;
+; OPTIMIZATION_LEVEL_2-LABEL: define void @f
+; OPTIMIZATION_LEVEL_2-SAME: () local_unnamed_addr #[[ATTR2:[0-9]+]] !dbg [[DBG30:![0-9]+]] {
+; OPTIMIZATION_LEVEL_2-NEXT: entry:
+; OPTIMIZATION_LEVEL_2-NEXT: ret void, !dbg [[DBG33:![0-9]+]]
+;
entry:
-
-; OPTIMIZATION_LEVEL_0: define void @f()
-; OPTIMIZATION_LEVEL_0-NEXT: entry:
-; OPTIMIZATION_LEVEL_0-NEXT: %call = call i32 @maxA(i32 3, i32 4), !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_0-NEXT: %call1 = call i32 @maxB(i32 1, i32 9), !dbg !{{[0-9]+}}
-; OPTIMIZATION_LEVEL_0-NEXT: ret void, !dbg !{{[0-9]+}}
-
-; OPTIMIZATION_LEVEL_2: define void @f()
-; OPTIMIZATION_LEVEL_2-NEXT: entry:
-; OPTIMIZATION_LEVEL_2-NEXT: ret void, !dbg !{{[0-9]+}}
-
%call = call i32 @maxA(i32 3, i32 4), !dbg !60
%call1 = call i32 @maxB(i32 1, i32 9), !dbg !61
ret void, !dbg !62
diff --git a/llvm/test/Transforms/MergeFunc/no-merge-debug-thunks.ll b/llvm/test/Transforms/MergeFunc/no-merge-debug-thunks.ll
new file mode 100644
index 000000000000000..80467c7aee407f8
--- /dev/null
+++ b/llvm/test/Transforms/MergeFunc/no-merge-debug-thunks.ll
@@ -0,0 +1,61 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 2
+;; Make sure debug instructions are not counted when deciding to merge functions
+; RUN: opt -S -passes=mergefunc < %s | FileCheck %s
+
+target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+
+; Function Attrs: nounwind readnone
+define hidden i32 @f(i32 %t) {
+; CHECK-LABEL: define hidden i32 @f
+; CHECK-SAME: (i32 [[T:%.*]]) {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: call void @llvm.dbg.value(metadata i32 [[T]], metadata [[META6:![0-9]+]], metadata !DIExpression()), !dbg [[DBG12:![0-9]+]]
+; CHECK-NEXT: call void @llvm.dbg.value(metadata i32 [[T]], metadata [[META6]], metadata !DIExpression()), !dbg [[DBG12]]
+; CHECK-NEXT: ret i32 0
+;
+entry:
+ call void @llvm.dbg.value(metadata i32 %t, metadata !12, metadata !DIExpression()), !dbg !13
+ call void @llvm.dbg.value(metadata i32 %t, metadata !12, metadata !DIExpression()), !dbg !13
+ ret i32 0
+}
+
+; Function Attrs: nounwind readnone
+define hidden i32 @f_thunk(i32 %t) {
+; CHECK-LABEL: define hidden i32 @f_thunk
+; CHECK-SAME: (i32 [[T:%.*]]) {
+; CHECK-NEXT: entry:
+; CHECK-NEXT: call void @llvm.dbg.value(metadata i32 [[T]], metadata [[META6]], metadata !DIExpression()), !dbg [[DBG12]]
+; CHECK-NEXT: call void @llvm.dbg.value(metadata i32 [[T]], metadata [[META6]], metadata !DIExpression()), !dbg [[DBG12]]
+; CHECK-NEXT: ret i32 0
+;
+entry:
+ call void @llvm.dbg.value(metadata i32 %t, metadata !12, metadata !DIExpression()), !dbg !13
+ call void @llvm.dbg.value(metadata i32 %t, metadata !12, metadata !DIExpression()), !dbg !13
+ ret i32 0
+}
+
+; Function Attrs: nounwind readnone speculatable
+declare void @llvm.dbg.value(metadata, metadata, metadata)
+
+!llvm.dbg.cu = !{!0}
+!llvm.module.flags = !{!3, !4, !5}
+
+!0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug, enums: !2)
+!1 = !DIFile(filename: "no-merge-debug-thunks.c", directory: "/tmp")
+!2 = !{}
+!3 = !{i32 2, !"Dwarf Version", i32 4}
+!4 = !{i32 2, !"Debug Info Version", i32 3}
+!5 = !{i32 1, !"wchar_size", i32 4}
+!7 = distinct !DISubprogram(name: "test", scope: !1, file: !1, line: 3, type: !8, isLocal: false, isDefinition: true, scopeLine: 3, flags: DIFlagPrototyped, isOptimized: true, unit: !0, retainedNodes: !11)
+!8 = !DISubroutineType(types: !9)
+!9 = !{!10, !10}
+!10 = !DIBasicType(name: "int", size: 32, encoding: DW_ATE_signed)
+!11 = !{!12}
+!12 = !DILocalVariable(name: "t", arg: 1, scope: !7, file: !1, line: 3, type: !10)
+!13 = !DILocation(line: 3, column: 14, scope: !7)
+!14 = !DILocation(line: 4, column: 12, scope: !7)
+!16 = distinct !DISubprogram(name: "_start", scope: !1, file: !1, line: 7, type: !17, isLocal: false, isDefinition: true, scopeLine: 7, isOptimized: true, unit: !0, retainedNodes: !2)
+!17 = !DISubroutineType(types: !18)
+!18 = !{!10}
+!19 = !DILocation(line: 8, column: 3, scope: !16)
+!20 = !DILocation(line: 9, column: 3, scope: !16)
|
oskarwirga
force-pushed
the
merge-thunk-debug
branch
from
d4e901d
to
ac3268d
Compare
I noticed that when we determine the size of the function to figure out if its profitable, we include debug instructions which can end up making larger functions than necessary.
oskarwirga
force-pushed
the
merge-thunk-debug
branch
from
ac3268d
to
93841e1
Compare
|
I modified the check from: to: as internal testing has shown that it reduces binary size. |
I'm not convinced this is a good idea. This means you allow merging via thunk for functions that only have a single real instruction (apart from the return). This might minimize binary size, but introducing an extra call seems like a bad size/performance tradeoff for that case. |
Sorry for overzealously landing before you had a chance to re-review. Without this change, I actually saw a binary size regression and I figured people opting in to function merging were seeking out minimal binary size. Before this change, we were inconsistent on how we merged functions. The test case I added was from a direct observation where I noticed we were merging functions like |
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.
I noticed that when we determine the size of the function to figure out if its profitable, we include debug instructions which can end up making larger functions than necessary.