/*
- * Copyright 2015 Facebook, Inc.
+ * Copyright 2016 Facebook, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
// Functions to provide smarter use of jemalloc, if jemalloc is being used.
// http://www.canonware.com/download/jemalloc/jemalloc-latest/doc/jemalloc.html
-#ifndef FOLLY_MALLOC_H_
-#define FOLLY_MALLOC_H_
+#pragma once
/**
* Define various MALLOCX_* macros normally provided by jemalloc. We define
#define MALLOCX_ZERO (static_cast<int>(0x40))
#endif
-// If using fbstring from libstdc++, then just define stub code
-// here to typedef the fbstring type into the folly namespace.
+// If using fbstring from libstdc++ (see comment in FBString.h), then
+// just define stub code here to typedef the fbstring type into the
+// folly namespace.
// This provides backwards compatibility for code that explicitly
// includes and uses fbstring.
#if defined(_GLIBCXX_USE_FB) && !defined(_LIBSTDCXX_FBSTRING)
#include <folly/detail/Malloc.h>
+#include <folly/portability/BitsFunctexcept.h>
#include <string>
__attribute__((__weak__));
extern "C" void dallocx(void*, int)
__attribute__((__weak__));
+extern "C" void sdallocx(void*, size_t, int)
+__attribute__((__weak__));
extern "C" size_t nallocx(size_t, int)
__attribute__((__weak__));
extern "C" int mallctl(const char*, void*, size_t*, void*, size_t)
__attribute__((__weak__));
#include <bits/functexcept.h>
+
#define FOLLY_HAVE_MALLOC_H 1
-#else
+
+#else // !defined(_LIBSTDCXX_FBSTRING)
+
#include <folly/detail/Malloc.h> /* nolint */
-#include <folly/Portability.h>
+#include <folly/portability/BitsFunctexcept.h> /* nolint */
+
#endif
// for malloc_usable_size
-// NOTE: FreeBSD 9 doesn't have malloc.h. It's defitions
+// NOTE: FreeBSD 9 doesn't have malloc.h. Its definitions
// are found in stdlib.h.
#if FOLLY_HAVE_MALLOC_H
#include <malloc.h>
#include <cassert>
#include <cstddef>
+#include <cstdint>
#include <cstdlib>
#include <cstring>
namespace folly {
#endif
-bool usingJEMallocSlow();
+// Cannot depend on Portability.h when _LIBSTDCXX_FBSTRING.
+// Disabled for nvcc because it fails on attributes on lambdas.
+#if defined(__GNUC__) && !defined(__NVCC__)
+#define FOLLY_MALLOC_NOINLINE __attribute__((__noinline__))
+#else
+#define FOLLY_MALLOC_NOINLINE
+#endif
/**
* Determine if we are using jemalloc or not.
*/
-inline bool usingJEMalloc() {
+inline bool usingJEMalloc() noexcept {
// Checking for rallocx != NULL is not sufficient; we may be in a dlopen()ed
// module that depends on libjemalloc, so rallocx is resolved, but the main
- // program might be using a different memory allocator. Look at the
- // implementation of usingJEMallocSlow() for the (hacky) details.
- static const bool result = usingJEMallocSlow();
+ // program might be using a different memory allocator.
+ // How do we determine that we're using jemalloc? In the hackiest
+ // way possible. We allocate memory using malloc() and see if the
+ // per-thread counter of allocated memory increases. This makes me
+ // feel dirty inside. Also note that this requires jemalloc to have
+ // been compiled with --enable-stats.
+ static const bool result = [] () FOLLY_MALLOC_NOINLINE noexcept {
+ // Some platforms (*cough* OSX *cough*) require weak symbol checks to be
+ // in the form if (mallctl != nullptr). Not if (mallctl) or if (!mallctl)
+ // (!!). http://goo.gl/xpmctm
+ if (mallocx == nullptr || rallocx == nullptr || xallocx == nullptr
+ || sallocx == nullptr || dallocx == nullptr || sdallocx == nullptr
+ || nallocx == nullptr || mallctl == nullptr
+ || mallctlnametomib == nullptr || mallctlbymib == nullptr) {
+ return false;
+ }
+
+ // "volatile" because gcc optimizes out the reads from *counter, because
+ // it "knows" malloc doesn't modify global state...
+ /* nolint */ volatile uint64_t* counter;
+ size_t counterLen = sizeof(uint64_t*);
+
+ if (mallctl("thread.allocatedp", static_cast<void*>(&counter), &counterLen,
+ nullptr, 0) != 0) {
+ return false;
+ }
+
+ if (counterLen != sizeof(uint64_t*)) {
+ return false;
+ }
+
+ uint64_t origAllocated = *counter;
+
+ // Static because otherwise clever compilers will find out that
+ // the ptr is not used and does not escape the scope, so they will
+ // just optimize away the malloc.
+ static void* ptr = malloc(1);
+ if (!ptr) {
+ // wtf, failing to allocate 1 byte
+ return false;
+ }
+
+ return (origAllocated != *counter);
+ }();
+
return result;
}
-/**
- * For jemalloc's size classes, see
- * http://www.canonware.com/download/jemalloc/jemalloc-latest/doc/jemalloc.html
- */
inline size_t goodMallocSize(size_t minSize) noexcept {
+ if (minSize == 0) {
+ return 0;
+ }
+
if (!usingJEMalloc()) {
// Not using jemalloc - no smarts
return minSize;
}
- size_t goodSize;
- if (minSize <= 64) {
- // Choose smallest allocation to be 64 bytes - no tripping over
- // cache line boundaries, and small string optimization takes care
- // of short strings anyway.
- goodSize = 64;
- } else if (minSize <= 512) {
- // Round up to the next multiple of 64; we don't want to trip over
- // cache line boundaries.
- goodSize = (minSize + 63) & ~size_t(63);
- } else {
- // Boundaries between size classes depend on numerious factors, some of
- // which can even be modified at run-time. Determine the good allocation
- // size by calling nallocx() directly.
- goodSize = nallocx(minSize, 0);
- }
- assert(nallocx(goodSize, 0) == goodSize);
- return goodSize;
+
+ return nallocx(minSize, 0);
}
// We always request "good" sizes for allocation, so jemalloc can
} // folly
#endif // !defined(_GLIBCXX_USE_FB) || defined(_LIBSTDCXX_FBSTRING)
-
-#endif // FOLLY_MALLOC_H_