From ff7b42c194e0fa23e6a76f0a33a80c0c3af14e7d Mon Sep 17 00:00:00 2001
From: Kazu Hirata <kazu@google.com>
Date: Sun, 24 Nov 2024 21:08:54 -0800
Subject: [memprof] Speed up llvm-profdata (#117446)

CallStackRadixTreeBuilder::build takes the parameter
MemProfFrameIndexes by value, involving copies:

  std::optional<const llvm::DenseMap<FrameIdTy, LinearFrameId>>
    MemProfFrameIndexes

Then "build" makes another copy of MemProfFrameIndexe and passes it to
encodeCallStack for every call stack, which is painfully slow.

This patch changes the type to a pointer so that we don't have to make
a copy every time we pass the argument.

Without this patch, it takes 553 seconds to run "llvm-profdata merge"
on a large MemProf raw profile.  This patch shortenes that down to 67
seconds.
---
 llvm/unittests/ProfileData/MemProfTest.cpp | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

(limited to 'llvm/unittests/ProfileData/MemProfTest.cpp')

diff --git a/llvm/unittests/ProfileData/MemProfTest.cpp b/llvm/unittests/ProfileData/MemProfTest.cpp
index c3f35e4..7b9910e 100644
--- a/llvm/unittests/ProfileData/MemProfTest.cpp
+++ b/llvm/unittests/ProfileData/MemProfTest.cpp
@@ -628,7 +628,7 @@ TEST(MemProf, RadixTreeBuilderEmpty) {
       FrameHistogram =
           llvm::memprof::computeFrameHistogram<FrameId>(MemProfCallStackData);
   llvm::memprof::CallStackRadixTreeBuilder<FrameId> Builder;
-  Builder.build(std::move(MemProfCallStackData), MemProfFrameIndexes,
+  Builder.build(std::move(MemProfCallStackData), &MemProfFrameIndexes,
                 FrameHistogram);
   ASSERT_THAT(Builder.getRadixArray(), testing::IsEmpty());
   const auto Mappings = Builder.takeCallStackPos();
@@ -646,7 +646,7 @@ TEST(MemProf, RadixTreeBuilderOne) {
       FrameHistogram =
           llvm::memprof::computeFrameHistogram<FrameId>(MemProfCallStackData);
   llvm::memprof::CallStackRadixTreeBuilder<FrameId> Builder;
-  Builder.build(std::move(MemProfCallStackData), MemProfFrameIndexes,
+  Builder.build(std::move(MemProfCallStackData), &MemProfFrameIndexes,
                 FrameHistogram);
   EXPECT_THAT(Builder.getRadixArray(), testing::ElementsAreArray({
                                            3U, // Size of CS1,
@@ -673,7 +673,7 @@ TEST(MemProf, RadixTreeBuilderTwo) {
       FrameHistogram =
           llvm::memprof::computeFrameHistogram<FrameId>(MemProfCallStackData);
   llvm::memprof::CallStackRadixTreeBuilder<FrameId> Builder;
-  Builder.build(std::move(MemProfCallStackData), MemProfFrameIndexes,
+  Builder.build(std::move(MemProfCallStackData), &MemProfFrameIndexes,
                 FrameHistogram);
   EXPECT_THAT(Builder.getRadixArray(),
               testing::ElementsAreArray({
@@ -711,7 +711,7 @@ TEST(MemProf, RadixTreeBuilderSuccessiveJumps) {
       FrameHistogram =
           llvm::memprof::computeFrameHistogram<FrameId>(MemProfCallStackData);
   llvm::memprof::CallStackRadixTreeBuilder<FrameId> Builder;
-  Builder.build(std::move(MemProfCallStackData), MemProfFrameIndexes,
+  Builder.build(std::move(MemProfCallStackData), &MemProfFrameIndexes,
                 FrameHistogram);
   EXPECT_THAT(Builder.getRadixArray(),
               testing::ElementsAreArray({
-- 
cgit v1.1