[libcxx-commits] [libcxx] [libc++] Add an ABI flag to optimize mersenne_twister_engine (PR #206423)

[Louis Dionne via libcxx-commits]

================
@@ -34,4 +34,15 @@ static void BM_SeedSeq_Generate(benchmark::State& state) {
 }
 BENCHMARK(BM_SeedSeq_Generate)->Ranges({{1, MAX_SEED_LEN}, {1, MAX_BUFFER_LEN}});
 
+template <class Engine>
+static void BM_engine(benchmark::State& state) {
----------------
ldionne wrote:

Let's also add a benchmark where we call it a hundred times (known statically) to allow the compiler to vectorize further.

https://github.com/llvm/llvm-project/pull/206423