forked from mozilla/gecko-dev
-
Notifications
You must be signed in to change notification settings - Fork 1
/
OperatorNewExtensions.h
52 lines (46 loc) · 2.17 KB
/
OperatorNewExtensions.h
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
/* -*- Mode: C++; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* vim: set ts=8 sts=2 et sw=2 tw=80: */
/* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
/* A version of |operator new| that eschews mandatory null-checks. */
#ifndef mozilla_OperatorNewExtensions_h
#define mozilla_OperatorNewExtensions_h
#include "mozilla/Assertions.h"
// Credit goes to WebKit for this implementation, cf.
// https://bugs.webkit.org/show_bug.cgi?id=74676
namespace mozilla {
enum NotNullTag {
KnownNotNull,
};
} // namespace mozilla
/*
* The logic here is a little subtle. [expr.new] states that if the allocation
* function being called returns null, then object initialization must not be
* done, and the entirety of the new expression must return null. Non-throwing
* (noexcept) functions are defined to return null to indicate failure. The
* standard placement operator new is defined in such a way, and so it requires
* a null check, even when that null check would be extraneous. Functions
* declared without such a specification are defined to throw std::bad_alloc if
* they fail, and return a non-null pointer otherwise. We compile without
* exceptions, so any placement new overload we define that doesn't declare
* itself as noexcept must therefore avoid generating a null check. Below is
* just such an overload.
*
* You might think that MOZ_NONNULL might perform the same function, but
* MOZ_NONNULL isn't supported on all of our compilers, and even when it is
* supported, doesn't work on all the versions we support. And even keeping
* those limitations in mind, we can't put MOZ_NONNULL on the global,
* standardized placement new function in any event.
*
* We deliberately don't add MOZ_NONNULL(3) to tag |p| as non-null, to benefit
* hypothetical static analyzers. Doing so makes |MOZ_ASSERT(p)|'s internal
* test vacuous, and some compilers warn about such vacuous tests.
*/
inline void*
operator new(size_t, mozilla::NotNullTag, void* p)
{
MOZ_ASSERT(p);
return p;
}
#endif // mozilla_OperatorNewExtensions_h