beenull/ladybird
1
0
Fork 0
mirror of https://github.com/LadybirdBrowser/ladybird.git synced 2024-11-26 09:30:24 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 7
ladybird/AK/Platform.h

206 lines
4.7 KiB
C
Raw Normal View History

Meta: Add license header to source files As suggested by Joshua, this commit adds the 2-clause BSD license as a comment block to the top of every source file. For the first pass, I've just added myself for simplicity. I encourage everyone to add themselves as copyright holders of any file they've added or modified in some significant way. If I've added myself in error somewhere, feel free to replace it with the appropriate copyright holder instead. Going forward, all new source files should include a license header.
2020-01-18 08:38:21 +00:00
/*
AK: Make it possible to not `using` AK classes into the global namespace This patch adds the `USING_AK_GLOBALLY` macro which is enabled by default, but can be overridden by build flags. This is a step towards integrating Jakt and AK types.
2022-11-26 11:18:30 +00:00
* Copyright (c) 2018-2022, Andreas Kling <kling@serenityos.org>
AK+Everywhere: Add AK_COMPILER_{GCC,CLANG} and use them most places Doesn't use them in libc headers so that those don't have to pull in AK/Platform.h. AK_COMPILER_GCC is set _only_ for gcc, not for clang too. (__GNUC__ is defined in clang builds as well.) Using AK_COMPILER_GCC simplifies things some. AK_COMPILER_CLANG isn't as much of a win, other than that it's consistent with AK_COMPILER_GCC.
2022-10-04 19:04:13 +00:00
* Copyright (c) 2022, Nico Weber <thakis@chromium.org>
Meta: Add license header to source files As suggested by Joshua, this commit adds the 2-clause BSD license as a comment block to the top of every source file. For the first pass, I've just added myself for simplicity. I encourage everyone to add themselves as copyright holders of any file they've added or modified in some significant way. If I've added myself in error somewhere, feel free to replace it with the appropriate copyright holder instead. Going forward, all new source files should include a license header.
2020-01-18 08:38:21 +00:00
*
Everything: Move to SPDX license identifiers in all files. SPDX License Identifiers are a more compact / standardized way of representing file license information. See: https://spdx.dev/resources/use/#identifiers This was done with the `ambr` search and replace tool. ambr --no-parent-ignore --key-from-file --rep-from-file key.txt rep.txt *
2021-04-22 08:24:48 +00:00
* SPDX-License-Identifier: BSD-2-Clause
Meta: Add license header to source files As suggested by Joshua, this commit adds the 2-clause BSD license as a comment block to the top of every source file. For the first pass, I've just added myself for simplicity. I encourage everyone to add themselves as copyright holders of any file they've added or modified in some significant way. If I've added myself in error somewhere, feel free to replace it with the appropriate copyright holder instead. Going forward, all new source files should include a license header.
2020-01-18 08:38:21 +00:00
*/
AK: Add Platform.h with an ARCH() macro. You can currently use this to detect the CPU architecture like so: #if ARCH(I386) ... #elif ARCH(X86_64) ... #else ... #endif This will be helpful for separating out architecture-specific code blocks.
2019-07-09 13:46:22 +00:00
#pragma once
AK: Make it possible to not `using` AK classes into the global namespace This patch adds the `USING_AK_GLOBALLY` macro which is enabled by default, but can be overridden by build flags. This is a step towards integrating Jakt and AK types.
2022-11-26 11:18:30 +00:00
#ifndef USING_AK_GLOBALLY
# define USING_AK_GLOBALLY 1
#endif
AK: Add Platform.h with an ARCH() macro. You can currently use this to detect the CPU architecture like so: #if ARCH(I386) ... #elif ARCH(X86_64) ... #else ... #endif This will be helpful for separating out architecture-specific code blocks.
2019-07-09 13:46:22 +00:00
#ifdef __x86_64__
AK: Use endianness flags to determine if conversion is necessary
2020-02-19 14:58:54 +00:00
# define AK_ARCH_X86_64 1
AK: Add Platform.h with an ARCH() macro. You can currently use this to detect the CPU architecture like so: #if ARCH(I386) ... #elif ARCH(X86_64) ... #else ... #endif This will be helpful for separating out architecture-specific code blocks.
2019-07-09 13:46:22 +00:00
#endif
AK: Add support for ARCH(AARCH64) to Platform.h
2021-10-11 22:38:16 +00:00
#ifdef __aarch64__
# define AK_ARCH_AARCH64 1
#endif
Everywhere: Add support for compilation under emscripten Co-Authored-By: Andrew Kaster <akaster@serenityos.org>
2022-11-20 03:23:14 +00:00
#ifdef __wasm32__
# define AK_ARCH_WASM32 1
#endif
AK+LibJS: Handle NaN-boxing pointers on AArch64 JS::Value stores 48 bit pointers to separately allocated objects in its payload. On x86-64, canonical addresses have their top 16 bits set to the same value as bit 47, effectively meaning that the value has to be sign-extended to get the pointer. AArch64, however, expects the topmost bits to be all zeros. This commit gates sign extension behind `#if ARCH(X86_64)`, and adds an `#error` for unsupported architectures, so that we do not forget to think about pointer handling when porting to a new architecture. Fixes #15290 Fixes SerenityOS/ladybird#56
2022-09-20 16:09:33 +00:00
#if (defined(__SIZEOF_POINTER__) && __SIZEOF_POINTER__ == 8) || defined(_WIN64)
# define AK_ARCH_64_BIT
#else
# define AK_ARCH_32_BIT
#endif
AK+Everywhere: Add AK_COMPILER_{GCC,CLANG} and use them most places Doesn't use them in libc headers so that those don't have to pull in AK/Platform.h. AK_COMPILER_GCC is set _only_ for gcc, not for clang too. (__GNUC__ is defined in clang builds as well.) Using AK_COMPILER_GCC simplifies things some. AK_COMPILER_CLANG isn't as much of a win, other than that it's consistent with AK_COMPILER_GCC.
2022-10-04 19:04:13 +00:00
#if defined(__clang__)
# define AK_COMPILER_CLANG
#elif defined(__GNUC__)
# define AK_COMPILER_GCC
#endif
AK: Add more AK_OS_FOO macros, including AK_OS_SERENITY
2022-10-01 12:23:49 +00:00
#if defined(__serenity__)
# define AK_OS_SERENITY
#endif
#if defined(__linux__)
# define AK_OS_LINUX
#endif
AK: Add CLOCK_*_COARSE aliases for darwin and BSD variants The coarse clocks in time.h are a linux extension that we've adopted. MacOS and the BSDs don't have it, so we need an alias in a platform header for Lagom builds.
2020-12-27 22:38:32 +00:00
#if defined(__APPLE__) && defined(__MACH__)
# define AK_OS_MACOS
# define AK_OS_BSD_GENERIC
#endif
AK: Add more AK_OS_FOO macros, including AK_OS_SERENITY
2022-10-01 12:23:49 +00:00
#if defined(__FreeBSD__)
# define AK_OS_BSD_GENERIC
# define AK_OS_FREEBSD
#endif
#if defined(__NetBSD__)
# define AK_OS_BSD_GENERIC
# define AK_OS_NETBSD
#endif
#if defined(__OpenBSD__)
# define AK_OS_BSD_GENERIC
# define AK_OS_OPENBSD
#endif
#if defined(__DragonFly__)
AK: Add CLOCK_*_COARSE aliases for darwin and BSD variants The coarse clocks in time.h are a linux extension that we've adopted. MacOS and the BSDs don't have it, so we need an alias in a platform header for Lagom builds.
2020-12-27 22:38:32 +00:00
# define AK_OS_BSD_GENERIC
AK: Add more AK_OS_FOO macros, including AK_OS_SERENITY
2022-10-01 12:23:49 +00:00
# define AK_OS_DRAGONFLY
AK: Add CLOCK_*_COARSE aliases for darwin and BSD variants The coarse clocks in time.h are a linux extension that we've adopted. MacOS and the BSDs don't have it, so we need an alias in a platform header for Lagom builds.
2020-12-27 22:38:32 +00:00
#endif
Lagom: Win32 support baby steps This is the initial port of Lagom to win32. This will enable developers to use Lagom as an alternative to vanilla STL/StandardC++Library - which gives a much richer environment (think QtCore - but modern). My main incentive - is to have a native Windows Ladybird working. I am starting with AK, which does not yet fully compile (on mingw). When AK is compiling (currently fails building StringBuffer.cpp) - I will continue to LibCore and then the rest of the user space libraries (excluding the GUI, which will be another different effort). Most of the code is happily stollen from Andrew Kaster's fork - he deserves the credit. Co-authored-by: Andrew Kaster <akaster@serenityos.org>
2022-09-20 07:15:12 +00:00
#if defined(_WIN32) || defined(_WIN64)
# define AK_OS_WINDOWS
#endif
AK: Add support for building on Android with API version >= 30
2022-07-11 07:04:39 +00:00
#if defined(__ANDROID__)
# define STR(x) __STR(x)
# define __STR(x) #x
# if __ANDROID_API__ < 30
# pragma message "Invalid android API " STR(__ANDROID_API__)
# error "Build configuration not tested on configured Android API version"
# endif
# undef STR
# undef __STR
# define AK_OS_ANDROID
#endif
Everywhere: Add support for compilation under emscripten Co-Authored-By: Andrew Kaster <akaster@serenityos.org>
2022-11-20 03:23:14 +00:00
#if defined(__EMSCRIPTEN__)
# define AK_OS_EMSCRIPTEN
#endif
AK: Add Platform.h with an ARCH() macro. You can currently use this to detect the CPU architecture like so: #if ARCH(I386) ... #elif ARCH(X86_64) ... #else ... #endif This will be helpful for separating out architecture-specific code blocks.
2019-07-09 13:46:22 +00:00
#define ARCH(arch) (defined(AK_ARCH_##arch) && AK_ARCH_##arch)
AK: Remove i686 support
2022-10-04 00:32:20 +00:00
#if ARCH(X86_64)
Kernel: Add VALIDATE_IS_X86 macro
2021-12-01 17:10:36 +00:00
# define VALIDATE_IS_X86()
#else
# define VALIDATE_IS_X86() static_assert(false, "Trying to include x86 only header on non x86 platform");
#endif
Kernel: Add VALIDATE_IS_AARCH64 guard macro
2022-10-16 14:55:42 +00:00
#if ARCH(AARCH64)
# define VALIDATE_IS_AARCH64()
#else
# define VALIDATE_IS_AARCH64() static_assert(false, "Trying to include aarch64 only header on non aarch64 platform");
#endif
AK+Everywhere: Add AK_COMPILER_{GCC,CLANG} and use them most places Doesn't use them in libc headers so that those don't have to pull in AK/Platform.h. AK_COMPILER_GCC is set _only_ for gcc, not for clang too. (__GNUC__ is defined in clang builds as well.) Using AK_COMPILER_GCC simplifies things some. AK_COMPILER_CLANG isn't as much of a win, other than that it's consistent with AK_COMPILER_GCC.
2022-10-04 19:04:13 +00:00
#if !defined(AK_COMPILER_CLANG) && !defined(__CLION_IDE_) && !defined(__CLION_IDE__)
AK: Use conditionally trivial special member functions This commit makes use of the conditionally trivial special member functions introduced in C++20. Basically, `Optional` and `Variant` inherits whether its wrapped type is trivially copy constructible, trivially copy assignable or trivially destructible. This lets the compiler optimize optimize a large number of their use cases. The constraints have been applied to `Optional`'s converting constructors too in order to make the API more explicit. This feature is not supported by Clang yet, so we use conditional compilation so that Lagom can be built on macOS. Once Clang has P0848R3 support, these can be removed.
2021-07-02 15:42:50 +00:00
# define AK_HAS_CONDITIONALLY_TRIVIAL
#endif
AK: Add ALWAYS_INLINE, NEVER_INLINE and FLATTEN macros It's tedious to write (and look at) [[gnu::always_inline]] etc. :^)
2020-04-30 09:43:25 +00:00
#ifdef ALWAYS_INLINE
AK: Add Endian.h header to replace NetworkOrdered.h.
2020-08-25 13:11:15 +00:00
# undef ALWAYS_INLINE
AK: Add ALWAYS_INLINE, NEVER_INLINE and FLATTEN macros It's tedious to write (and look at) [[gnu::always_inline]] etc. :^)
2020-04-30 09:43:25 +00:00
#endif
AK: Use `__attribute__((name))` for functions everywhere Clang enforces the ordering that attributes specified with the `[[attr_name]]` syntax must comes before those defines as `__attribute__((attr_name))`. We don't want to deal with that, so we should stick to a single syntax (for functions, at least). This commit favors the latter, as it's used more widely in the code (for declaring more "exotic" options), and changing those would be a larger effort than modifying this single file.
2021-06-20 14:59:51 +00:00
#define ALWAYS_INLINE __attribute__((always_inline)) inline
AK: Add ALWAYS_INLINE, NEVER_INLINE and FLATTEN macros It's tedious to write (and look at) [[gnu::always_inline]] etc. :^)
2020-04-30 09:43:25 +00:00
#ifdef NEVER_INLINE
AK: Add Endian.h header to replace NetworkOrdered.h.
2020-08-25 13:11:15 +00:00
# undef NEVER_INLINE
AK: Add ALWAYS_INLINE, NEVER_INLINE and FLATTEN macros It's tedious to write (and look at) [[gnu::always_inline]] etc. :^)
2020-04-30 09:43:25 +00:00
#endif
AK: Use `__attribute__((name))` for functions everywhere Clang enforces the ordering that attributes specified with the `[[attr_name]]` syntax must comes before those defines as `__attribute__((attr_name))`. We don't want to deal with that, so we should stick to a single syntax (for functions, at least). This commit favors the latter, as it's used more widely in the code (for declaring more "exotic" options), and changing those would be a larger effort than modifying this single file.
2021-06-20 14:59:51 +00:00
#define NEVER_INLINE __attribute__((noinline))
AK: Add ALWAYS_INLINE, NEVER_INLINE and FLATTEN macros It's tedious to write (and look at) [[gnu::always_inline]] etc. :^)
2020-04-30 09:43:25 +00:00
#ifdef FLATTEN
AK: Add Endian.h header to replace NetworkOrdered.h.
2020-08-25 13:11:15 +00:00
# undef FLATTEN
AK: Add ALWAYS_INLINE, NEVER_INLINE and FLATTEN macros It's tedious to write (and look at) [[gnu::always_inline]] etc. :^)
2020-04-30 09:43:25 +00:00
#endif
AK: Use `__attribute__((name))` for functions everywhere Clang enforces the ordering that attributes specified with the `[[attr_name]]` syntax must comes before those defines as `__attribute__((attr_name))`. We don't want to deal with that, so we should stick to a single syntax (for functions, at least). This commit favors the latter, as it's used more widely in the code (for declaring more "exotic" options), and changing those would be a larger effort than modifying this single file.
2021-06-20 14:59:51 +00:00
#define FLATTEN __attribute__((flatten))
AK: Add ALWAYS_INLINE, NEVER_INLINE and FLATTEN macros It's tedious to write (and look at) [[gnu::always_inline]] etc. :^)
2020-04-30 09:43:25 +00:00
AK: Add RETURNS_NONNULL attribute and use it This attribute tells compilers that the pointer returned by a function is never null, which lets it optimize away null checks in some places. This seems like a nice addition to `NonnullOwnPtr` and `NonnullRefPtr`. Using this attribute causes extra UBSan checks to be emitted. To offset its performance loss, some additional methods were marked ALWAYS_INLINE, which lets the compiler optimize duplicate checks
2021-06-29 13:45:24 +00:00
#ifdef RETURNS_NONNULL
# undef RETURNS_NONNULL
#endif
#define RETURNS_NONNULL __attribute__((returns_nonnull))
AK: Add #define for [[gnu::no_sanitize_address]] This lines up with other attribute global #defines
2021-05-12 11:51:02 +00:00
#ifdef NO_SANITIZE_ADDRESS
# undef NO_SANITIZE_ADDRESS
#endif
AK: Use `__attribute__((name))` for functions everywhere Clang enforces the ordering that attributes specified with the `[[attr_name]]` syntax must comes before those defines as `__attribute__((attr_name))`. We don't want to deal with that, so we should stick to a single syntax (for functions, at least). This commit favors the latter, as it's used more widely in the code (for declaring more "exotic" options), and changing those would be a larger effort than modifying this single file.
2021-06-20 14:59:51 +00:00
#define NO_SANITIZE_ADDRESS __attribute__((no_sanitize_address))
AK: Add #define for [[gnu::no_sanitize_address]] This lines up with other attribute global #defines
2021-05-12 11:51:02 +00:00
Kernel: Replace raw asm functions with naked ones
2021-07-05 13:34:07 +00:00
#ifdef NAKED
# undef NAKED
#endif
AK: Stub out the NAKED macro on AARCH64 This is almost certainly incorrect but we'll see about that once the kernel actually gets to userspace init.
2022-10-12 20:05:41 +00:00
#ifndef AK_ARCH_AARCH64
# define NAKED __attribute__((naked))
#else
# define NAKED
#endif
Kernel: Replace raw asm functions with naked ones
2021-07-05 13:34:07 +00:00
Kernel+AK: Generate compile-time error for non-sized `delete` This is a much more ergonomic option than getting a `VERIFY_NOT_REACHED()` failure at run-time. I encountered this issue with Clang, where sized deallocation is not the default due to ABI breakage concerns. Note that we can't simply just not declare these functions, because the C++ standard states: > If this function with size parameter is defined, the program shall > also define the version without the size parameter.
2021-07-15 11:50:55 +00:00
#ifdef DISALLOW
# undef DISALLOW
#endif
AK+Everywhere: Add AK_COMPILER_{GCC,CLANG} and use them most places Doesn't use them in libc headers so that those don't have to pull in AK/Platform.h. AK_COMPILER_GCC is set _only_ for gcc, not for clang too. (__GNUC__ is defined in clang builds as well.) Using AK_COMPILER_GCC simplifies things some. AK_COMPILER_CLANG isn't as much of a win, other than that it's consistent with AK_COMPILER_GCC.
2022-10-04 19:04:13 +00:00
#if defined(AK_COMPILER_CLANG)
Kernel+AK: Generate compile-time error for non-sized `delete` This is a much more ergonomic option than getting a `VERIFY_NOT_REACHED()` failure at run-time. I encountered this issue with Clang, where sized deallocation is not the default due to ABI breakage concerns. Note that we can't simply just not declare these functions, because the C++ standard states: > If this function with size parameter is defined, the program shall > also define the version without the size parameter.
2021-07-15 11:50:55 +00:00
# define DISALLOW(message) __attribute__((diagnose_if(1, message, "error")))
#else
# define DISALLOW(message) __attribute__((error(message)))
#endif
AK: Add platform macros to detect presence of AddressSanitizer The ASAN_[UN]POISON_MEMORY_REGION macros can be used to manually notify the AddressSanitizer runtime about the reachability of instrumented code accessing a memory region. This is most useful for manually managed heaps and arenas that do not go directly to malloc or alligned_alloc.
2021-05-27 23:49:40 +00:00
// GCC doesn't have __has_feature but clang does
#ifndef __has_feature
# define __has_feature(...) 0
#endif
#if __has_feature(address_sanitizer) || defined(__SANITIZE_ADDRESS__)
# define HAS_ADDRESS_SANITIZER
# define ASAN_POISON_MEMORY_REGION(addr, size) __asan_poison_memory_region(addr, size)
# define ASAN_UNPOISON_MEMORY_REGION(addr, size) __asan_unpoison_memory_region(addr, size)
#else
# define ASAN_POISON_MEMORY_REGION(addr, size)
# define ASAN_UNPOISON_MEMORY_REGION(addr, size)
#endif
Everywhere: Replace uses of __serenity__ with AK_OS_SERENITY Now that we have OS macros for essentially every supported OS, let's try to use them everywhere.
2022-10-09 21:23:23 +00:00
#ifndef AK_OS_SERENITY
AK: Don't try to include <unistd.h> on windows for PAGE_SIZE
2022-12-12 14:39:24 +00:00
# ifdef AK_OS_WINDOWS
// FIXME: No idea where to get this, but it's 4096 anyway :^)
# define PAGE_SIZE 4096
AK: On macOS host builds, wrap unistd.h with missing extern "C" During the build process on macOS, multiple versions of <unistd.h> were being included (Apple's version and GCC's version). It appears that all other places #include the version from GCC, but in Platform.h the Apple header was being used. GCC's <unistd.h> is wrapped in `extern "C"`, while Apple's is not. This causes a conflicting declaration, so we need to wrap the #include with extern "C". Issue has been observed on macOS Mojave. See https://github.com/microsoft/vcpkg/issues/11320 for a similar issue.
2021-09-01 01:23:29 +00:00
// On macOS (at least Mojave), Apple's version of this header is not wrapped
// in extern "C".
AK: Actually don't include <unistd.h> for windows in Platform.h I got the conditions wrong last time, oops :^)
2022-12-12 16:27:01 +00:00
# else
# if defined(AK_OS_MACOS)
AK: On macOS host builds, wrap unistd.h with missing extern "C" During the build process on macOS, multiple versions of <unistd.h> were being included (Apple's version and GCC's version). It appears that all other places #include the version from GCC, but in Platform.h the Apple header was being used. GCC's <unistd.h> is wrapped in `extern "C"`, while Apple's is not. This causes a conflicting declaration, so we need to wrap the #include with extern "C". Issue has been observed on macOS Mojave. See https://github.com/microsoft/vcpkg/issues/11320 for a similar issue.
2021-09-01 01:23:29 +00:00
extern "C" {
AK: Actually don't include <unistd.h> for windows in Platform.h I got the conditions wrong last time, oops :^)
2022-12-12 16:27:01 +00:00
# endif
# include <unistd.h>
# undef PAGE_SIZE
# define PAGE_SIZE sysconf(_SC_PAGESIZE)
# ifdef AK_OS_MACOS
AK: Don't try to include <unistd.h> on windows for PAGE_SIZE
2022-12-12 14:39:24 +00:00
}
AK: Actually don't include <unistd.h> for windows in Platform.h I got the conditions wrong last time, oops :^)
2022-12-12 16:27:01 +00:00
# endif
AK: On macOS host builds, wrap unistd.h with missing extern "C" During the build process on macOS, multiple versions of <unistd.h> were being included (Apple's version and GCC's version). It appears that all other places #include the version from GCC, but in Platform.h the Apple header was being used. GCC's <unistd.h> is wrapped in `extern "C"`, while Apple's is not. This causes a conflicting declaration, so we need to wrap the #include with extern "C". Issue has been observed on macOS Mojave. See https://github.com/microsoft/vcpkg/issues/11320 for a similar issue.
2021-09-01 01:23:29 +00:00
# endif
AK: Allow NonnullRefPtr::ptr() when in "unknown" typestate. Clang loses the typestate when passing NonnullRefPtr's via lambda captures. This is unfortunate, but not much we can do about it. Allowing ptr() makes it possible to use captured NonnullRefPtrs as you'd expect.
2019-07-25 09:51:24 +00:00
#endif
AK: Shim open_with_path_length() for non-Serenity builds. This is pretty ugly but I don't want to *not* use open_with_path_length() so let's just shim it.
2019-07-25 12:25:45 +00:00
Lagom: Win32 support baby steps This is the initial port of Lagom to win32. This will enable developers to use Lagom as an alternative to vanilla STL/StandardC++Library - which gives a much richer environment (think QtCore - but modern). My main incentive - is to have a native Windows Ladybird working. I am starting with AK, which does not yet fully compile (on mingw). When AK is compiling (currently fails building StringBuffer.cpp) - I will continue to LibCore and then the rest of the user space libraries (excluding the GUI, which will be another different effort). Most of the code is happily stollen from Andrew Kaster's fork - he deserves the credit. Co-authored-by: Andrew Kaster <akaster@serenityos.org>
2022-09-20 07:15:12 +00:00
#if defined(AK_OS_WINDOWS)
# define CLOCK_MONOTONIC_COARSE CLOCK_MONOTONIC
#endif
AK: Fix FreeBSD compilation for clock FreeBSD introduced CLOCK_MONOTONIC_COARSE and CLOCK_REALTIME_COARSE. This update fixes ladybird builds from FreeBSD 13.1 see clock_gettime(2) https://www.freebsd.org/cgi/man.cgi?query=clock_gettime&apropos=0&sektion=0&manpath=FreeBSD+13.1-RELEASE&arch=default&format=ascii
2022-10-13 03:10:57 +00:00
#if defined(AK_OS_BSD_GENERIC) && !defined(AK_OS_FREEBSD)
AK: Add CLOCK_*_COARSE aliases for darwin and BSD variants The coarse clocks in time.h are a linux extension that we've adopted. MacOS and the BSDs don't have it, so we need an alias in a platform header for Lagom builds.
2020-12-27 22:38:32 +00:00
# define CLOCK_MONOTONIC_COARSE CLOCK_MONOTONIC
# define CLOCK_REALTIME_COARSE CLOCK_REALTIME
#endif
AK: Allow alignment to cache line size with CACHE_ALIGNED This is particularly important to avoid false sharing, which thrashes performance when two process-shared atomics are on the same cache line.
2022-03-04 13:57:38 +00:00
AK: Prefix CACHE_ALIGNED & SYSTEM_CACHE_ALIGNMENT_SIZE
2022-07-31 20:32:33 +00:00
#ifndef AK_SYSTEM_CACHE_ALIGNMENT_SIZE
AK: Allow alignment to cache line size with CACHE_ALIGNED This is particularly important to avoid false sharing, which thrashes performance when two process-shared atomics are on the same cache line.
2022-03-04 13:57:38 +00:00
# if ARCH(AARCH64) || ARCH(x86_64)
AK: Prefix CACHE_ALIGNED & SYSTEM_CACHE_ALIGNMENT_SIZE
2022-07-31 20:32:33 +00:00
# define AK_SYSTEM_CACHE_ALIGNMENT_SIZE 64
AK: Allow alignment to cache line size with CACHE_ALIGNED This is particularly important to avoid false sharing, which thrashes performance when two process-shared atomics are on the same cache line.
2022-03-04 13:57:38 +00:00
# else
AK: Prefix CACHE_ALIGNED & SYSTEM_CACHE_ALIGNMENT_SIZE
2022-07-31 20:32:33 +00:00
# define AK_SYSTEM_CACHE_ALIGNMENT_SIZE 128
AK: Allow alignment to cache line size with CACHE_ALIGNED This is particularly important to avoid false sharing, which thrashes performance when two process-shared atomics are on the same cache line.
2022-03-04 13:57:38 +00:00
# endif
AK: Prefix CACHE_ALIGNED & SYSTEM_CACHE_ALIGNMENT_SIZE
2022-07-31 20:32:33 +00:00
#endif /* AK_SYSTEM_CACHE_ALIGNMENT_SIZE */
AK: Allow alignment to cache line size with CACHE_ALIGNED This is particularly important to avoid false sharing, which thrashes performance when two process-shared atomics are on the same cache line.
2022-03-04 13:57:38 +00:00
AK: Prefix CACHE_ALIGNED & SYSTEM_CACHE_ALIGNMENT_SIZE
2022-07-31 20:32:33 +00:00
#ifdef AK_CACHE_ALIGNED
# undef AK_CACHE_ALIGNED
AK: Allow alignment to cache line size with CACHE_ALIGNED This is particularly important to avoid false sharing, which thrashes performance when two process-shared atomics are on the same cache line.
2022-03-04 13:57:38 +00:00
#endif
AK: Prefix CACHE_ALIGNED & SYSTEM_CACHE_ALIGNMENT_SIZE
2022-07-31 20:32:33 +00:00
#define AK_CACHE_ALIGNED alignas(AK_SYSTEM_CACHE_ALIGNMENT_SIZE)
Reference in a new issue Copy permalink