[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.

1 files changed, 1 insertions, 1 deletions

diff --git a/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp b/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp
index 8f22a50..63f4e34 100644
--- a/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp
+++ b/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp
@@ -4235,7 +4235,7 @@ static DenseMap<CallStackId, LinearCallStackId> writeMemoryProfileRadixTree(
   CallStackRadixTreeBuilder<LinearFrameId> Builder;
   // We don't need a MemProfFrameIndexes map as we have already converted the
   // full stack id hash to a linear offset into the StackIds array.
-  Builder.build(std::move(CallStacks), /*MemProfFrameIndexes=*/std::nullopt,
+  Builder.build(std::move(CallStacks), /*MemProfFrameIndexes=*/nullptr,
                 FrameHistogram);
   Stream.EmitRecord(bitc::FS_CONTEXT_RADIX_TREE_ARRAY, Builder.getRadixArray(),
                     RadixAbbrev);