summaryrefslogtreecommitdiffstats
path: root/base/atomicops.h
Commit message (Collapse)AuthorAgeFilesLines
* Use C++11 atomicsjfb2014-10-141-20/+61
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This CL moves the Chrome code base to C++11 atomics instead of inline assembly for the toolchains that support C++11's <atomic>. The patch is constructed to modify as little code as possible, and will be followed-up with other patches to clean up the rest of the code base. This change should allow compilers to emit better code when dealing with atomics (LLVM has recently seen substantial improvements thanks to morisset's work), be more portable, eventually delete a bunch of code, and in subsequent changes atomicity will be based on memory location instead of individual accesses (making the code less error-prone, and easier to analyze). This patch stems from a fix I recently made to V8 to build under NaCl and PNaCl. This was gating V8's C++11 migration: they needed to move to the LLVM-based NaCl toolchain which didn't work with atomicops' inline assembly. V8 currently uses the __sync_* primitives for the NaCl/PNaCl build, which imply sequential consistency and are therefore suboptimal. Before doing further work I want to fix Chrome's own headers, and trickle these changes to V8. Future changes: * The atomicops headers are copied in 8 locations [0] in the Chrome code base, not all of them are up to date. Those should all be updated in their individual repositories (some are in thiry_party). * The signature of functions should be updated to take an atomic location instead of the pointer argument, and all call sites should be updated (there are current 127 such calls [1] in chromium/src in non-implementation and non-test files). * Atomic operations should be used directly. A few things worth noting: * The atomicops_internals_portable.h file contains the main implementation, and has extra notes on implementation choices. * The CL removes x86 CPU features detection from the portable implementation: - Because the headers are copied in a few places and the x86 CPU feature detection uses a static global that parses cpuid, the atomicops_internals_x86_gcc.cc file was only built once, but its struct interface was declared external and used in protobuf and tcmalloc. Removing the struct from Chrome's own base makes the linker sad because of the two uses. This has two implications: # Don't specialize atomics for SSE2 versus non SSE2. This is a non-issue since Chrome requires a minimum of SSE2. The code is therefore faster (skip a branch). # The code doesn't detect the AMD Opteron Rev E memory barrier bug from AMD errata 147 [2]. This bug affects Opterons 1xx/2xx/8xx that shipped in 2003-2004. In general compilers should work around this bug, not library code. GCC has had this workaround since 2009 [3], but it is still an open bug in LLVM [4]. See comment #27 on the code review: this CPU is used by 0.02% of Chrome users, for whom the race would have to occur and the compiler not have the workaround for the bug to manifest. This also makes the code faster by skipping a branch. * The CL moves the declaration of the x86 CPU features struct to atomicops.h, and matches its fields with the ones duplicated across the code base. This is transitional: it fixes a link failure because tcmalloc relied on the x86_gcc.{h,cc} declaration and implementation, and it fixes an ODR violation where the struct didn't always have 3 members (and the other members were sometimes accessed, uninitialized). * tsan already rewrites all atomics to its own primitives, its header is therefore now unnecessary. The implementation takes care to detect and error out if that turns out not to be true. * MemoryBarrier works around a libstdc++ bug in older versions [5]. The workaround is exactly the non-buggy code for libstdc++ (call out to the compiler's builtin). * The assembly generated by GCC 4.8 and LLVM 3.6 for x86-32/x86-64/ARM when using this portable implementation can be found in [6]. [0]: find . -name "atomicops.h" ./base/atomicops.h ./v8/src/base/atomicops.h ./native_client_sdk/src/libraries/sdk_util/atomicops.h ./third_party/webrtc/base/atomicops.h ./third_party/tcmalloc/chromium/src/base/atomicops.h ./third_party/tcmalloc/vendor/src/base/atomicops.h ./third_party/re2/util/atomicops.h ./third_party/protobuf/src/google/protobuf/stubs/atomicops.h [1]: git grep Barrier_ | grep -v atomicops | grep -v unittest | wc -l [2]: http://support.amd.com/us/Processor_TechDocs/25759.pdf [3]: https://gcc.gnu.org/ml/gcc-patches/2009-10/txt00046.txt [4]: http://llvm.org/bugs/show_bug.cgi?id=5934 [5]: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=51038 [6]: https://code.google.com/p/chromium/issues/detail?id=423074 TEST=ninja -C out/Release base_unittests TEST=trybots BUG=246514 BUG=234215 BUG=420970 BUG=423074 R= dvyukov@google.com, thakis@chromium.org, glider@chromium.org, hboehm@google.com, morisset@google.com, willchan@chromium.org Review URL: https://codereview.chromium.org/636783002 Cr-Commit-Position: refs/heads/master@{#299485}
* Adding atomic ops for MIPS64.Gordana.Cmiljanovic@imgtec.com2014-08-141-1/+2
| | | | | | | | | | | This change adds atomic operations for MIPS64. BUG=400684 Review URL: https://codereview.chromium.org/450343002 Cr-Commit-Position: refs/heads/master@{#289624} git-svn-id: svn://svn.chromium.org/chrome/trunk/src@289624 0039d316-1c4b-4281-b951-d872f2087c98
* Cleanup coding stylebyungwook.ahn@gmail.com2014-04-131-1/+1
| | | | | | | | | BUG=NONE TEST=NONE Review URL: https://codereview.chromium.org/232973002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@263554 0039d316-1c4b-4281-b951-d872f2087c98
* Add Arm64 atomicops to base.rmcilroy@chromium.org2014-03-241-1/+3
| | | | | | | | | | These are taken from the implementation in v8/src/atomicops_internals_a64_gcc.h BUG=354405 Review URL: https://codereview.chromium.org/206373005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@258985 0039d316-1c4b-4281-b951-d872f2087c98
* Atomic ops cleanup.ctruta@blackberry.com2014-02-081-2/+3
| | | | | | | | | | | | Use <stdint.h> instead of "base/basictypes.h". Put AtomicOps_Internalx86CPUFeaturesInit in the anonymous namespace. Fix formatting issues. BUG= Review URL: https://codereview.chromium.org/143273005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@249938 0039d316-1c4b-4281-b951-d872f2087c98
* Improve the implementation of atomic operations on Linux/ARM (including ↵digit@chromium.org2013-06-101-2/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Android/ARM). The previous patch at: https://chromiumcodereview.appspot.com/10831358 actually regressed the performance of atomic operations on Linux/ARM systems, because the GCC intrinsics (e.g. __sync_fetch_and_add) are very poorly implemented at the moment (and also always provide a full barrier, even when the caller doesn't need it). This replaces the implementation with a better version which: - Uses inline assembly and LDRES/STREX instructions on ARMv6+, or the old kernel helper cmpxchg implementation on ARMv5. - Still uses the kernel helper memory barrier to optimize for single core devices on ARMv6 and ARMv7, or ARMv5 binaries running on devices with a higher architecture number. - Provide truly barrier free compare-and-swap, swap and atomic-inc operations. On tested Android/ARM devices, this speeds up atomic increments by x2 to x3. This indirectly speeds up other operations relying on it (e.g. scoped_refptr<> or base::Bind()). For details, see: https://docs.google.com/a/chromium.org/spreadsheet/ccc?key=0Arp73PHrzcIQdGNUd1NGYWlfY0dKWS1EZ2V6RThhZXc&usp=sharing BUG=234215 Review URL: https://chromiumcodereview.appspot.com/16335007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@205205 0039d316-1c4b-4281-b951-d872f2087c98
* fix x32 handling of Atomic64vapier@chromium.org2013-02-071-1/+1
| | | | | | | | | | | | The x32 logic for the size of Atomic64 handles NaCL, but misses the Linux case. Check the standard __ILP32__ to handle that too. BUG=chromium-os:36866 TEST=compiled the code for x86_64 (64bit) & x86_64 (x32) Review URL: https://chromiumcodereview.appspot.com/12186005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@181114 0039d316-1c4b-4281-b951-d872f2087c98
* Workaround to make MemoryBarrier compile on Win x64scottmg@chromium.org2013-01-151-0/+9
| | | | | | | | | | | | | | windows.h #defines "MemoryBarrier" on x64 (on x86 it's an forceinline function). Because atomicops.h uses MemoryBarrier as the public api name, workaround by #undefing and then calling the implementation directly. ref: http://msdn.microsoft.com/en-us/library/windows/desktop/ms684208.aspx R=brettw@chromium.org BUG=166496 Review URL: https://chromiumcodereview.appspot.com/11888016 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@176859 0039d316-1c4b-4281-b951-d872f2087c98
* Atomics implementation for compiler-based ThreadSanitizer ↵glider@chromium.org2012-10-021-1/+3
| | | | | | | | | (http://dev.chromium.org/developers/testing/threadsanitizer-tsan-v2) BUG=128314 Review URL: https://codereview.chromium.org/10948035 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@159705 0039d316-1c4b-4281-b951-d872f2087c98
* Fix atomic ops on ARM to compile in NaCl untrusted targets.bbudge@chromium.org2012-08-221-2/+3
| | | | | | | | | | This makes the arm version of atomicops_internals use gcc intrinsics for the NaCl build. Other linux targets won't change. BUG=116317 TEST=compiles on arm bots Review URL: https://chromiumcodereview.appspot.com/10831358 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@152791 0039d316-1c4b-4281-b951-d872f2087c98
* Have iOS use the same atomicops_internal implementation that Mac uses.leng@chromium.org2012-07-131-2/+2
| | | | | | | | | | | The atomicops_internal file for Mac and iOS is not actually architecture-specific, so it is being renamd to be arch-independent. BUG=b/6754112 TEST=None Review URL: https://chromiumcodereview.appspot.com/10696193 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@146584 0039d316-1c4b-4281-b951-d872f2087c98
* Remove the rest of #pragma once in one big CL.ajwong@chromium.org2012-07-111-1/+0
| | | | | | | | | For context see this thread: https://groups.google.com/a/chromium.org/forum/?fromgroups#!topic/chromium-dev/RMcVNGjB4II TBR=thakis,pkasting,jam git-svn-id: svn://svn.chromium.org/chrome/trunk/src@146163 0039d316-1c4b-4281-b951-d872f2087c98
* [MIPS] Adding atomic ops for MIPS architecture.petarj@mips.com2012-06-181-1/+3
| | | | | | | | | | | | Initial commit for atomic operations on MIPS architecture. BUG=https://code.google.com/p/chromium/issues/detail?id=130022 TEST=make chrome Review URL: https://chromiumcodereview.appspot.com/10448043 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@142663 0039d316-1c4b-4281-b951-d872f2087c98
* windows: Remove vestigal support for /Wp64evan@chromium.org2011-07-281-5/+1
| | | | | | | | | This flag was removed from our build in 2008, so we no longer need workarounds to support it. Review URL: http://codereview.chromium.org/7477030 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@94480 0039d316-1c4b-4281-b951-d872f2087c98
* OpenBSD: make it possible to build libbase.a.phajdan.jr@chromium.org2011-04-151-0/+6
| | | | | | | | | | | | | | This still needs a little hack. I had to run gyp_chromium this way: CHROMIUM_GYP_FILE="base/base.gyp" build/gyp_chromium -DOS=openbsd \ -Duse_gnome_keyring=0 -Duse_system_libevent=1 And then run "gmake base". BUG=none Review URL: http://codereview.chromium.org/6840023 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@81709 0039d316-1c4b-4281-b951-d872f2087c98
* iwyu: Cleanup in the following files:jhawkins@chromium.org2011-04-041-2/+2
| | | | | | | | | | | | | | | | | | | | | * at_exit.cc * atomicops.h * base_paths.h * bzip2_error_handler.cc * callback_internal.h * command_line.cc * cpu.cc * environment.h * event_recorder.cc * file_descriptor_shuffle.cc * file_path.cc BUG=none TEST=none Review URL: http://codereview.chromium.org/6759017 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@80340 0039d316-1c4b-4281-b951-d872f2087c98
* Help make base compile in NaCl.abarth@chromium.org2010-12-011-0/+6
| | | | | | | | | | Turns out intptr_t isn't 64 bits wide in 64-bit NaCl. This class assumes that Atomic64 is actually 64 bits wide, so we need to use the int64_t type instead. Original patch by Eric Seidel. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@67826 0039d316-1c4b-4281-b951-d872f2087c98
* `#pragma once` for app, base, chrome, gfx, ipc, net, skia, viewsthakis@chromium.org2010-07-261-0/+1
| | | | | | | | | BUG=50273 TEST=everything still builds, build is 10% faster on windows, same speed on mac/linux TBR: erg git-svn-id: svn://svn.chromium.org/chrome/trunk/src@53716 0039d316-1c4b-4281-b951-d872f2087c98
* Some base/ progress towards building on Linux x86-64.deanm@chromium.org2009-07-271-4/+6
| | | | | | | | | | - Use long for int64 (still 64 bits) to avoid annoying nspr clash. - Fix some incorrect define guards and add some functions to atomicops. Review URL: http://codereview.chromium.org/159428 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@21709 0039d316-1c4b-4281-b951-d872f2087c98
* Add atomic operations for ARM.thestig@chromium.org2009-05-011-0/+2
| | | | | | Review URL: http://codereview.chromium.org/57031 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@15070 0039d316-1c4b-4281-b951-d872f2087c98
* NO CODE CHANGEdeanm@chromium.org2009-03-101-1/+0
| | | | | | | | | Normalize end of file newlines in base/. All files end in a single newline. Review URL: http://codereview.chromium.org/42011 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@11329 0039d316-1c4b-4281-b951-d872f2087c98
* Rename atomicops_internals_x86_mac.h back to its original name,mmentovai@google.com2008-08-271-1/+1
| | | | | | | atomicops_internals_x86_macosx.h. I didn't realize this was brought in as-is from upstream. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@1451 0039d316-1c4b-4281-b951-d872f2087c98
* Add more passing tests to base_unittests. Add a few more headers we use tommentovai@google.com2008-08-271-1/+1
| | | | | | base.xcodeproj. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@1446 0039d316-1c4b-4281-b951-d872f2087c98
* Use a more compact license header in source files.license.bot2008-08-241-28/+4
| | | | git-svn-id: svn://svn.chromium.org/chrome/trunk/src@1287 0039d316-1c4b-4281-b951-d872f2087c98
* Fix incorrect OS_APPLE to OS_MACOSX and drop COMPILER_GCC since the codedeanm@google.com2008-08-061-1/+1
| | | | | | | | uses the Apple APIs for atomic operations and not inline assembly. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@432 0039d316-1c4b-4281-b951-d872f2087c98
* Use the build_config defines in atomicops.h. This still uses the ↵deanm@google.com2008-08-051-13/+10
| | | | | | compiler/platform specific defines in the platform dependent internals headers, keeping them closer to the original implementation. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@372 0039d316-1c4b-4281-b951-d872f2087c98
* New atomic operation foundation code and interfaces. Based on the atomic ↵deanm@google.com2008-08-051-0/+165
operations in google-perftools / tcmalloc. - Makes atomic operations cross-platform (win/mac/linux). - Supports 64 bit operations on 64 bit processors. - Make atomic operations more correct on Windows (barriers). - Introduces a better / safer interface, and abstracts away most common atomic operations, reference counting and sequence numbers. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@370 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-02-14 12:41:56 +0000