[scudo] [GWP-ASan] Add GWP-ASan variant of scudo benchmarks.

GWP-ASan is the "production" variant as compiled by compiler-rt, and it's useful to be able to benchmark changes in GWP-ASan or Scudo's GWP-ASan hooks across versions. GWP-ASan is sampled, and sampled allocations are much slower, but given the amount of allocations that happen under test here - we actually get a reasonable representation of GWP-ASan's negligent performance impact between runs.

Reviewed By: cryptoad

Differential Revision: https://reviews.llvm.org/D101865

GitOrigin-RevId: 8936608e6f4dbd2a80acde660849cd87ef5c9d26
diff --git a/benchmarks/malloc_benchmark.cpp b/benchmarks/malloc_benchmark.cpp
index ce48dc0..661fff4 100644
--- a/benchmarks/malloc_benchmark.cpp
+++ b/benchmarks/malloc_benchmark.cpp
@@ -13,15 +13,22 @@
 #include "benchmark/benchmark.h"
 
 #include <memory>
+#include <vector>
+
+void *CurrentAllocator;
+template <typename Config> void PostInitCallback() {
+  reinterpret_cast<scudo::Allocator<Config> *>(CurrentAllocator)->initGwpAsan();
+}
 
 template <typename Config> static void BM_malloc_free(benchmark::State &State) {
-  using AllocatorT = scudo::Allocator<Config>;
+  using AllocatorT = scudo::Allocator<Config, PostInitCallback<Config>>;
   auto Deleter = [](AllocatorT *A) {
     A->unmapTestOnly();
     delete A;
   };
   std::unique_ptr<AllocatorT, decltype(Deleter)> Allocator(new AllocatorT,
                                                            Deleter);
+  CurrentAllocator = Allocator.get();
   Allocator->reset();
 
   const size_t NBytes = State.range(0);
@@ -55,18 +62,19 @@
 
 template <typename Config>
 static void BM_malloc_free_loop(benchmark::State &State) {
-  using AllocatorT = scudo::Allocator<Config>;
+  using AllocatorT = scudo::Allocator<Config, PostInitCallback<Config>>;
   auto Deleter = [](AllocatorT *A) {
     A->unmapTestOnly();
     delete A;
   };
   std::unique_ptr<AllocatorT, decltype(Deleter)> Allocator(new AllocatorT,
                                                            Deleter);
+  CurrentAllocator = Allocator.get();
   Allocator->reset();
 
   const size_t NumIters = State.range(0);
   size_t PageSize = scudo::getPageSizeCached();
-  void *Ptrs[NumIters];
+  std::vector<void *> Ptrs(NumIters);
 
   for (auto _ : State) {
     size_t SizeLog2 = 0;
diff --git a/combined.h b/combined.h
index 146408a..03a85ec 100644
--- a/combined.h
+++ b/combined.h
@@ -51,8 +51,7 @@
   typedef typename Params::template TSDRegistryT<ThisT> TSDRegistryT;
 
   void callPostInitCallback() {
-    static pthread_once_t OnceControl = PTHREAD_ONCE_INIT;
-    pthread_once(&OnceControl, PostInitCallback);
+    pthread_once(&PostInitNonce, PostInitCallback);
   }
 
   struct QuarantineCallback {
@@ -952,6 +951,7 @@
   SecondaryT Secondary;
   QuarantineT Quarantine;
   TSDRegistryT TSDRegistry;
+  pthread_once_t PostInitNonce = PTHREAD_ONCE_INIT;
 
 #ifdef GWP_ASAN_HOOKS
   gwp_asan::GuardedPoolAllocator GuardedAlloc;