remove compare and branch in fast path for c++ operator delete[]

Summary: sdallocx is checking a flag that will never be set (at least in the provided C++ destructor implementation).  This branch will probably only rarely be mispredicted however it removes two instructions in sdallocx and one at the callsite (to zero out flags).

src/jemalloc.c

@@ -2732,7 +2732,7 @@ bool free_fastpath(void *ptr, size_t size, bool size_hint) {
 	tcache_t *tcache = tsd_tcachep_get(tsd);
 
 	alloc_ctx_t alloc_ctx;
-	/* 
+	/*
 	 * If !config_cache_oblivious, we can check PAGE alignment to
 	 * detect sampled objects.  Otherwise addresses are
 	 * randomized, and we have to look it up in the rtree anyway.
@@ -3522,6 +3522,18 @@ je_sdallocx(void *ptr, size_t size, int flags) {
 	LOG("core.sdallocx.exit", "");
 }
 
+void JEMALLOC_NOTHROW
+je_sdallocx_noflags(void *ptr, size_t size) {
+	LOG("core.sdallocx.entry", "ptr: %p, size: %zu, flags: 0", ptr,
+		size);
+
+	if (!free_fastpath(ptr, size, true)) {
+		sdallocx_default(ptr, size, 0);
+	}
+
+	LOG("core.sdallocx.exit", "");
+}
+
 JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW
 JEMALLOC_ATTR(pure)
 je_nallocx(size_t size, int flags) {

src/jemalloc_cpp.cpp

@@ -128,14 +128,14 @@ operator delete(void *ptr, std::size_t size) noexcept {
 	if (unlikely(ptr == nullptr)) {
 		return;
 	}
-	je_sdallocx(ptr, size, /*flags=*/0);
+	je_sdallocx_noflags(ptr, size);
 }
 
 void operator delete[](void *ptr, std::size_t size) noexcept {
 	if (unlikely(ptr == nullptr)) {
 		return;
 	}
-	je_sdallocx(ptr, size, /*flags=*/0);
+	je_sdallocx_noflags(ptr, size);
 }
 
 #endif  // __cpp_sized_deallocation