[GCov] Emit memset instead of stores in __llvm_gcov_reset

For a very large module, __llvm_gcov_reset can become very large.
__llvm_gcov_reset previously emitted stores to a bunch of globals in one
huge basic block. MemCpyOpt would turn many of these stores into
memsets, and updating MemorySSA would be extremely slow.

Verified that this makes the compile time of certain files go down
drastically (20min -> 5min).

Reviewed By: MaskRay

Differential Revision: https://reviews.llvm.org/D107538

diff --git a/llvm/lib/Transforms/Instrumentation/GCOVProfiling.cpp b/llvm/lib/Transforms/Instrumentation/GCOVProfiling.cpp
index c99f2e6..d029a0c 100644 (file)
--- a/llvm/lib/Transforms/Instrumentation/GCOVProfiling.cpp
+++ b/llvm/lib/Transforms/Instrumentation/GCOVProfiling.cpp
@@ -1373,12 +1373,16 @@ Function *GCOVProfiler::insertReset(
 
   BasicBlock *Entry = BasicBlock::Create(*Ctx, "entry", ResetF);
   IRBuilder<> Builder(Entry);
+  LLVMContext &C = Entry->getContext();
 
   // Zero out the counters.
   for (const auto &I : CountersBySP) {
     GlobalVariable *GV = I.first;
-    Constant *Null = Constant::getNullValue(GV->getValueType());
-    Builder.CreateStore(Null, GV);
+    auto *GVTy = cast<ArrayType>(GV->getValueType());
+    Builder.CreateMemSet(GV, Constant::getNullValue(Type::getInt8Ty(C)),
+                         GVTy->getNumElements() *
+                             GVTy->getElementType()->getScalarSizeInBits() / 8,
+                         GV->getAlign());
   }
 
   Type *RetTy = ResetF->getReturnType();

diff --git a/llvm/test/Transforms/GCOVProfiling/reset.ll b/llvm/test/Transforms/GCOVProfiling/reset.ll
new file mode 100644 (file)
index 0000000..b8869ea
--- /dev/null
+++ b/llvm/test/Transforms/GCOVProfiling/reset.ll
@@ -0,0 +1,37 @@
+; RUN: mkdir -p %t && cd %t
+; RUN: opt < %s -S -passes=insert-gcov-profiling | FileCheck %s
+
+target triple = "x86_64-unknown-linux-gnu"
+
+define void @foo() !dbg !8 {
+entry:
+  ret void, !dbg !12
+}
+
+define void @bar() !dbg !13 {
+entry:
+  ret void, !dbg !14
+}
+
+; CHECK: define internal void @__llvm_gcov_reset()
+; CHECK-NEXT: entry:
+; CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* bitcast ([1 x i64]* @__llvm_gcov_ctr to i8*), i8 0, i64 8, i1 false)
+; CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* bitcast ([1 x i64]* @__llvm_gcov_ctr.1 to i8*), i8 0, i64 8, i1 false)
+
+!llvm.dbg.cu = !{!0}
+!llvm.module.flags = !{!3, !4, !5, !6}
+
+!0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang version 14.0.0", isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug, enums: !2, splitDebugInlining: false, nameTableKind: None)
+!1 = !DIFile(filename: "/tmp/a.c", directory: "/tmp")
+!2 = !{}
+!3 = !{i32 7, !"Dwarf Version", i32 4}
+!4 = !{i32 2, !"Debug Info Version", i32 3}
+!5 = !{i32 1, !"wchar_size", i32 4}
+!6 = !{i32 7, !"uwtable", i32 1}
+!8 = distinct !DISubprogram(name: "foo", scope: !9, file: !9, line: 1, type: !10, scopeLine: 1, flags: DIFlagAllCallsDescribed, spFlags: DISPFlagDefinition | DISPFlagOptimized, unit: !0, retainedNodes: !2)
+!9 = !DIFile(filename: "/tmp/a.c", directory: "")
+!10 = !DISubroutineType(types: !11)
+!11 = !{null}
+!12 = !DILocation(line: 1, column: 13, scope: !8)
+!13 = distinct !DISubprogram(name: "bar", scope: !9, file: !9, line: 2, type: !10, scopeLine: 2, flags: DIFlagAllCallsDescribed, spFlags: DISPFlagDefinition | DISPFlagOptimized, unit: !0, retainedNodes: !2)
+!14 = !DILocation(line: 2, column: 13, scope: !13)