|
|
|
/*
|
|
|
|
* This file is part of FFmpeg.
|
|
|
|
*
|
|
|
|
* FFmpeg is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
|
|
* License as published by the Free Software Foundation; either
|
|
|
|
* version 2.1 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* FFmpeg 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
|
|
|
|
* Lesser General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU Lesser General Public
|
|
|
|
* License along with FFmpeg; if not, write to the Free Software
|
|
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef AVUTIL_BUFFER_INTERNAL_H
|
|
|
|
#define AVUTIL_BUFFER_INTERNAL_H
|
|
|
|
|
|
|
|
#include <stdatomic.h>
|
|
|
|
#include <stdint.h>
|
|
|
|
|
|
|
|
#include "internal.h"
|
|
|
|
#include "buffer.h"
|
lavu: fix memory leaks by using a mutex instead of atomics
The buffer pool has to atomically add and remove entries from the linked
list of available buffers. This was done by removing the entire list
with a CAS operation, working on it, and then setting it back again
(using a retry-loop in case another thread was doing the same thing).
This could effectively cause memory leaks: while a thread was working on
the buffer list, other threads would allocate new buffers, increasing
the pool's total size. There was no real leak, but since these extra
buffers were not needed, but not free'd either (except when the buffer
pool was destroyed), this had the same effects as a real leak. For some
reason, growth was exponential, and could easily kill the process due
to OOM in real-world uses.
Fix this by using a mutex to protect the list operations. The fancy
way atomics remove the whole list to work on it is not needed anymore,
which also avoids the situation which was causing the leak.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years ago
|
|
|
#include "thread.h"
|
|
|
|
|
|
|
|
/**
|
|
|
|
* The buffer was av_realloc()ed, so it is reallocatable.
|
|
|
|
*/
|
|
|
|
#define BUFFER_FLAG_REALLOCATABLE (1 << 0)
|
|
|
|
/**
|
|
|
|
* The AVBuffer structure is part of a larger structure
|
|
|
|
* and should not be freed.
|
|
|
|
*/
|
|
|
|
#define BUFFER_FLAG_NO_FREE (1 << 1)
|
|
|
|
|
|
|
|
struct AVBuffer {
|
|
|
|
uint8_t *data; /**< data described by this buffer */
|
|
|
|
size_t size; /**< size of data in bytes */
|
|
|
|
|
|
|
|
/**
|
|
|
|
* number of existing AVBufferRef instances referring to this buffer
|
|
|
|
*/
|
|
|
|
atomic_uint refcount;
|
|
|
|
|
|
|
|
/**
|
|
|
|
* a callback for freeing the data
|
|
|
|
*/
|
|
|
|
void (*free)(void *opaque, uint8_t *data);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* an opaque pointer, to be used by the freeing callback
|
|
|
|
*/
|
|
|
|
void *opaque;
|
|
|
|
|
|
|
|
/**
|
|
|
|
* A combination of AV_BUFFER_FLAG_*
|
|
|
|
*/
|
|
|
|
int flags;
|
|
|
|
|
|
|
|
/**
|
|
|
|
* A combination of BUFFER_FLAG_*
|
|
|
|
*/
|
|
|
|
int flags_internal;
|
|
|
|
};
|
|
|
|
|
|
|
|
typedef struct BufferPoolEntry {
|
|
|
|
uint8_t *data;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Backups of the original opaque/free of the AVBuffer corresponding to
|
|
|
|
* data. They will be used to free the buffer when the pool is freed.
|
|
|
|
*/
|
|
|
|
void *opaque;
|
|
|
|
void (*free)(void *opaque, uint8_t *data);
|
|
|
|
|
|
|
|
AVBufferPool *pool;
|
lavu: fix memory leaks by using a mutex instead of atomics
The buffer pool has to atomically add and remove entries from the linked
list of available buffers. This was done by removing the entire list
with a CAS operation, working on it, and then setting it back again
(using a retry-loop in case another thread was doing the same thing).
This could effectively cause memory leaks: while a thread was working on
the buffer list, other threads would allocate new buffers, increasing
the pool's total size. There was no real leak, but since these extra
buffers were not needed, but not free'd either (except when the buffer
pool was destroyed), this had the same effects as a real leak. For some
reason, growth was exponential, and could easily kill the process due
to OOM in real-world uses.
Fix this by using a mutex to protect the list operations. The fancy
way atomics remove the whole list to work on it is not needed anymore,
which also avoids the situation which was causing the leak.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years ago
|
|
|
struct BufferPoolEntry *next;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* An AVBuffer structure to (re)use as AVBuffer for subsequent uses
|
|
|
|
* of this BufferPoolEntry.
|
|
|
|
*/
|
|
|
|
AVBuffer buffer;
|
|
|
|
} BufferPoolEntry;
|
|
|
|
|
|
|
|
struct AVBufferPool {
|
lavu: fix memory leaks by using a mutex instead of atomics
The buffer pool has to atomically add and remove entries from the linked
list of available buffers. This was done by removing the entire list
with a CAS operation, working on it, and then setting it back again
(using a retry-loop in case another thread was doing the same thing).
This could effectively cause memory leaks: while a thread was working on
the buffer list, other threads would allocate new buffers, increasing
the pool's total size. There was no real leak, but since these extra
buffers were not needed, but not free'd either (except when the buffer
pool was destroyed), this had the same effects as a real leak. For some
reason, growth was exponential, and could easily kill the process due
to OOM in real-world uses.
Fix this by using a mutex to protect the list operations. The fancy
way atomics remove the whole list to work on it is not needed anymore,
which also avoids the situation which was causing the leak.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years ago
|
|
|
AVMutex mutex;
|
|
|
|
BufferPoolEntry *pool;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This is used to track when the pool is to be freed.
|
|
|
|
* The pointer to the pool itself held by the caller is considered to
|
|
|
|
* be one reference. Each buffer requested by the caller increases refcount
|
|
|
|
* by one, returning the buffer to the pool decreases it by one.
|
|
|
|
* refcount reaches zero when the buffer has been uninited AND all the
|
|
|
|
* buffers have been released, then it's safe to free the pool and all
|
|
|
|
* the buffers in it.
|
|
|
|
*/
|
|
|
|
atomic_uint refcount;
|
|
|
|
|
|
|
|
size_t size;
|
|
|
|
void *opaque;
|
|
|
|
AVBufferRef* (*alloc)(size_t size);
|
|
|
|
AVBufferRef* (*alloc2)(void *opaque, size_t size);
|
|
|
|
void (*pool_free)(void *opaque);
|
|
|
|
};
|
|
|
|
|
|
|
|
#endif /* AVUTIL_BUFFER_INTERNAL_H */
|