diff options
author | Sergei Petrunia <psergey@askmonty.org> | 2016-10-09 15:27:22 +0000 |
---|---|---|
committer | Sergei Petrunia <psergey@askmonty.org> | 2016-10-09 15:27:22 +0000 |
commit | 20bd26e6df3acf739ff1b2cffc41084367cd6dd0 (patch) | |
tree | 83e589f67b6e860a8ef161498eb9299ea5421e4c /include | |
parent | 4be8cae56afd7a31eb2477e9e4fe1c814748b06d (diff) | |
download | mariadb-git-20bd26e6df3acf739ff1b2cffc41084367cd6dd0.tar.gz |
Backport from facebook/mysql-5.6:
commit b5fda565d4bad66f3b5a6432236b96743ccfd323
Author: Steaphan Greene <steaphan@fb.com>
Date: Fri Feb 7 21:34:04 2014 -0800
Add basic atomic_stats type
Summary:
WebScaleSQL Feature: Basic Stats Support
This is a simple class template that implements an atomic stats variable.
The focus is on performance, not accuracy. No set operations are
guaranteed. The other operations are all atomic, so values should not
actually be lost - however, there is no attempt to be consistent in
reading multiple stats, nor any concern that different threads see any
consistent view across multiple stats.
Test Plan:
This has been tested quite a bit in isolation, and no problems have been
found. However, this has never been used in production. Also, a newer,
more sophisticated, version of this is already in development.
It also shows a small perf gain in the larger stats diffs (Table Stats,
User Stats, etc...), when compared to the built-in MySQL atomics.
Reviewers: pivanof
Reviewed By: pivanof
CC: jtolmer, MarkCallaghan, flamingcow, jeremycole, andrew-ford, inaam-rana, liang.guo.752
Differential Revision: https://reviews.facebook.net/D16029
Differential Revision: https://reviews.facebook.net/D16449
Diffstat (limited to 'include')
-rw-r--r-- | include/atomic_stat.h | 94 |
1 files changed, 94 insertions, 0 deletions
diff --git a/include/atomic_stat.h b/include/atomic_stat.h new file mode 100644 index 00000000000..04e59bd9a8a --- /dev/null +++ b/include/atomic_stat.h @@ -0,0 +1,94 @@ +/* This is an atomic integer abstract data type, for high-performance + tracking of a single stat. It intentionally permits inconsistent + atomic operations and reads, for better performance. This means + that, though no data should ever be lost by this stat, reads of it + at any time may not include all changes up to any particular point. + + So, values read from these may only be approximately correct. + + If your use-case will fail under these conditions, do not use this. + + Copyright (C) 2012 - 2014 Steaphan Greene <steaphan@gmail.com> + + This program is free software; you can redistribute it and/or + modify it under the terms of the GNU General Public License + as published by the Free Software Foundation; either version 2 + of the License, or (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the + Free Software Foundation, Inc. + 51 Franklin Street, Fifth Floor + Boston, MA 02110-1301, USA. +*/ + +#ifndef _atomic_stat_h_ +#define _atomic_stat_h_ + +#include <atomic> + +template < typename TYPE > +class atomic_stat { +public: + // Initialize value to the default for the type + atomic_stat() : value_(TYPE()) {}; + + // This enforces a strict order, as all absolute sets should + void clear() { + value_.store(TYPE(), std::memory_order_seq_cst); + }; + + // Reads can get any valid value, it doesn't matter which, exactly + TYPE load() const { + return value_.load(std::memory_order_relaxed); + }; + + // This only supplies relative arithmetic operations + // These are all done atomically, and so can show up in any order + void inc(const TYPE &other) { + value_.fetch_add(other, std::memory_order_relaxed); + }; + + void dec(const TYPE &other) { + value_.fetch_sub(other, std::memory_order_relaxed); + }; + + void inc() { + value_.fetch_add(1, std::memory_order_relaxed); + }; + + void dec() { + value_.fetch_sub(1, std::memory_order_relaxed); + }; + + // This will make one attempt to set the value to the max of + // the current value, and the passed-in value. It can fail + // for any reason, and we only try it once. + void set_max_maybe(const TYPE &new_val) { + TYPE old_val = value_; + if (new_val > old_val) { + value_.compare_exchange_weak(old_val, new_val, + std::memory_order_relaxed, + std::memory_order_relaxed); + } + }; + + // This will make one attempt to assign the value to the passed-in + // value. It can fail for any reason, and we only try it once. + void set_maybe(const TYPE &new_val) { + TYPE old_val = value_; + value_.compare_exchange_weak(old_val, new_val, + std::memory_order_relaxed, + std::memory_order_relaxed); + }; + +private: + std::atomic<TYPE> value_; +}; + +#endif // _atomic_stat_h_ |