binutils-gdb/gdbsupport/thread-pool.h

214 lines
5.5 KiB
C
Raw Permalink Normal View History

/* Thread pool
Copyright (C) 2019-2024 Free Software Foundation, Inc.
This file is part of GDB.
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 3 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, see <http://www.gnu.org/licenses/>. */
#ifndef GDBSUPPORT_THREAD_POOL_H
#define GDBSUPPORT_THREAD_POOL_H
#include <queue>
#include <vector>
#include <functional>
#include <chrono>
#if CXX_STD_THREAD
#include <thread>
#include <mutex>
#include <condition_variable>
#include <future>
#endif
#include <optional>
namespace gdb
{
#if CXX_STD_THREAD
/* Simply use the standard future. */
template<typename T>
using future = std::future<T>;
/* ... and the standard future_status. */
using future_status = std::future_status;
#else /* CXX_STD_THREAD */
/* A compatibility enum for std::future_status. This is just the
subset needed by gdb. */
enum class future_status
{
ready,
timeout,
};
/* A compatibility wrapper for std::future. Once <thread> and
<future> are available in all GCC builds -- should that ever happen
-- this can be removed. GCC does not implement threading for
MinGW, see https://gcc.gnu.org/bugzilla/show_bug.cgi?id=93687.
Meanwhile, in this mode, there are no threads. Tasks submitted to
the thread pool are invoked immediately and their result is stored
here. The base template here simply wraps a T and provides some
std::future compatibility methods. The provided methods are chosen
based on what GDB needs presently. */
template<typename T>
class future
{
public:
explicit future (T value)
: m_value (std::move (value))
{
}
future () = default;
future (future &&other) = default;
future (const future &other) = delete;
future &operator= (future &&other) = default;
future &operator= (const future &other) = delete;
void wait () const { }
template<class Rep, class Period>
future_status wait_for (const std::chrono::duration<Rep,Period> &duration)
const
{
return future_status::ready;
}
T get () { return std::move (m_value); }
private:
T m_value;
};
/* A specialization for void. */
template<>
class future<void>
{
public:
void wait () const { }
template<class Rep, class Period>
future_status wait_for (const std::chrono::duration<Rep,Period> &duration)
const
{
return future_status::ready;
}
void get () { }
};
#endif /* CXX_STD_THREAD */
/* A thread pool.
There is a single global thread pool, see g_thread_pool. Tasks can
be submitted to the thread pool. They will be processed in worker
threads as time allows. */
class thread_pool
{
public:
/* The sole global thread pool. */
static thread_pool *g_thread_pool;
~thread_pool ();
DISABLE_COPY_AND_ASSIGN (thread_pool);
/* Set the thread count of this thread pool. By default, no threads
are created -- the thread count must be set first. */
void set_thread_count (size_t num_threads);
/* Return the number of executing threads. */
size_t thread_count () const
{
#if CXX_STD_THREAD
return m_thread_count;
#else
return 0;
#endif
}
/* Post a task to the thread pool. A future is returned, which can
be used to wait for the result. */
future<void> post_task (std::function<void ()> &&func)
{
#if CXX_STD_THREAD
std::packaged_task<void ()> task (std::move (func));
future<void> result = task.get_future ();
do_post_task (std::packaged_task<void ()> (std::move (task)));
return result;
#else
func ();
return {};
#endif /* CXX_STD_THREAD */
}
/* Post a task to the thread pool. A future is returned, which can
be used to wait for the result. */
template<typename T>
future<T> post_task (std::function<T ()> &&func)
{
#if CXX_STD_THREAD
std::packaged_task<T ()> task (std::move (func));
future<T> result = task.get_future ();
do_post_task (std::packaged_task<void ()> (std::move (task)));
return result;
#else
return future<T> (func ());
#endif /* CXX_STD_THREAD */
}
private:
thread_pool () = default;
#if CXX_STD_THREAD
/* The callback for each worker thread. */
void thread_function ();
/* Post a task to the thread pool. A future is returned, which can
be used to wait for the result. */
void do_post_task (std::packaged_task<void ()> &&func);
/* The current thread count. */
size_t m_thread_count = 0;
/* A convenience typedef for the type of a task. */
typedef std::packaged_task<void ()> task_t;
/* The tasks that have not been processed yet. An optional is used
to represent a task. If the optional is empty, then this means
that the receiving thread should terminate. If the optional is
non-empty, then it is an actual task to evaluate. */
std::queue<std::optional<task_t>> m_tasks;
/* A condition variable and mutex that are used for communication
between the main thread and the worker threads. */
std::condition_variable m_tasks_cv;
std::mutex m_tasks_mutex;
gdb: Enable early init of thread pool size This commit enables the early initialization commands (92e4e97a9f5) to modify the number of threads used by gdb's thread pool. The motivation here is to prevent gdb from spawning a detrimental number of threads on many-core systems under environments with restrictive ulimits. With gdb before this commit, the thread pool takes the following sizes: 1. Thread pool size is initialized to 0. 2. After the maintenance commands are defined, the thread pool size is set to the number of system cores (if it has not already been set). 3. Using early initialization commands, the thread pool size can be changed using "maint set worker-threads". 4. After the first prompt, the thread pool size can be changed as in the previous step. Therefore after step 2. gdb has potentially launched hundreds of threads on a many-core system. After this change, step 2 and 3 are reversed so there is an opportunity to set the required number of threads without needing to default to the number of system cores first. There does exist a configure option (added in 261b07488b9) to disable multithreading, but this does not allow for an already deployed gdb to be configured. Additionally, the default number of worker threads is clamped at eight to control the number of worker threads spawned on many-core systems. This value was chosen as testing recorded on bugzilla issue 29959 indicates that parallel efficiency drops past this point. GDB built with GCC 13. No test suite regressions detected. Compilers: GCC, ACfL, Intel, Intel LLVM, NVHPC; Platforms: x86_64, aarch64. The scenario that interests me the most involves preventing GDB from spawning any worker threads at all. This was tested by counting the number of clones observed by strace: strace -e clone,clone3 gdb/gdb -q \ --early-init-eval-command="maint set worker-threads 0" \ -ex q ./gdb/gdb |& grep --count clone The new test relies on "gdb: install CLI uiout while processing early init files" developed by Andrew Burgess. This patch will need pushing prior to this change. The clamping was tested on machines with both 16 cores and a single core. "maint show worker-threads" correctly reported eight and one respectively. Approved-By: Tom Tromey <tom@tromey.com>
2023-12-04 22:23:17 +08:00
bool m_sized_at_least_once = false;
#endif /* CXX_STD_THREAD */
};
}
#endif /* GDBSUPPORT_THREAD_POOL_H */