You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

343 lines
13 KiB

/*
* 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
*/
/**
* @file
* unbuffered private I/O API
*/
#ifndef AVFORMAT_URL_H
#define AVFORMAT_URL_H
#include "avio.h"
#include "libavformat/version.h"
#include "libavutil/dict.h"
#include "libavutil/log.h"
#define URL_PROTOCOL_FLAG_NESTED_SCHEME 1 /*< The protocol name can be the first part of a nested protocol scheme */
#define URL_PROTOCOL_FLAG_NETWORK 2 /*< The protocol uses network */
extern const AVClass ffurl_context_class;
typedef struct URLContext {
const AVClass *av_class; /**< information for av_log(). Set by url_open(). */
const struct URLProtocol *prot;
void *priv_data;
char *filename; /**< specified URL */
int flags;
int max_packet_size; /**< if non zero, the stream is packetized with this max packet size */
int is_streamed; /**< true if streamed (no seek possible), default = false */
int is_connected;
AVIOInterruptCB interrupt_callback;
int64_t rw_timeout; /**< maximum time to wait for (network) read/write operation completion, in mcs */
const char *protocol_whitelist;
const char *protocol_blacklist;
} URLContext;
typedef struct URLProtocol {
const char *name;
int (*url_open)( URLContext *h, const char *url, int flags);
/**
* This callback is to be used by protocols which open further nested
* protocols. options are then to be passed to ffurl_open()/ffurl_connect()
* for those nested protocols.
*/
int (*url_open2)(URLContext *h, const char *url, int flags, AVDictionary **options);
int (*url_accept)(URLContext *s, URLContext **c);
int (*url_handshake)(URLContext *c);
/**
* Read data from the protocol.
* If data is immediately available (even less than size), EOF is
* reached or an error occurs (including EINTR), return immediately.
* Otherwise:
* In non-blocking mode, return AVERROR(EAGAIN) immediately.
* In blocking mode, wait for data/EOF/error with a short timeout (0.1s),
* and return AVERROR(EAGAIN) on timeout.
* Checking interrupt_callback, looping on EINTR and EAGAIN and until
* enough data has been read is left to the calling function; see
* retry_transfer_wrapper in avio.c.
*/
int (*url_read)( URLContext *h, unsigned char *buf, int size);
int (*url_write)(URLContext *h, const unsigned char *buf, int size);
int64_t (*url_seek)( URLContext *h, int64_t pos, int whence);
int (*url_close)(URLContext *h);
int (*url_read_pause)(URLContext *h, int pause);
int64_t (*url_read_seek)(URLContext *h, int stream_index,
int64_t timestamp, int flags);
int (*url_get_file_handle)(URLContext *h);
int (*url_get_multi_file_handle)(URLContext *h, int **handles,
int *numhandles);
HTTP: improve performance by reducing forward seeks This commit optimizes HTTP performance by reducing forward seeks, instead favoring a read-ahead and discard on the current connection (referred to as a short seek) for seeks that are within a TCP window's worth of data. This improves performance because with TCP flow control, a window's worth of data will be in the local socket buffer already or in-flight from the sender once congestion control on the sender is fully utilizing the window. Note: this approach doesn't attempt to differentiate from a newly opened connection which may not be fully utilizing the window due to congestion control vs one that is. The receiver can't get at this information, so we assume worst case; that full window is in use (we did advertise it after all) and that data could be in-flight The previous behavior of closing the connection, then opening a new with a new HTTP range value results in a massive amounts of discarded and re-sent data when large TCP windows are used. This has been observed on MacOS/iOS which starts with an initial window of 256KB and grows up to 1MB depending on the bandwidth-product delay. When seeking within a window's worth of data and we close the connection, then open a new one within the same window's worth of data, we discard from the current offset till the end of the window. Then on the new connection the server ends up re-sending the previous data from new offset till the end of old window. Example (assumes full window utilization): TCP window size: 64KB Position: 32KB Forward seek position: 40KB * (Next window) 32KB |--------------| 96KB |---------------| 160KB * 40KB |---------------| 104KB Re-sent amount: 96KB - 40KB = 56KB For a real world test example, I have MP4 file of ~25MB, which ffplay only reads ~16MB and performs 177 seeks. With current ffmpeg, this results in 177 HTTP GETs and ~73MB worth of TCP data communication. With this patch, ffmpeg issues 4 HTTP GETs and 3 seeks for a total of ~22MB of TCP data communication. To support this feature, the short seek logic in avio_seek() has been extended to call a function to get the short seek threshold value. This callback has been plumbed to the URLProtocol structure, which now has infrastructure in HTTP and TCP to get the underlying receiver window size via SO_RCVBUF. If the underlying URL and protocol don't support returning a short seek threshold, the default s->short_seek_threshold is used This feature has been tested on Windows 7 and MacOS/iOS. Windows support is slightly complicated by the fact that when TCP window auto-tuning is enabled, SO_RCVBUF doesn't report the real window size, but it does if SO_RCVBUF was manually set (disabling auto-tuning). So we can only use this optimization on Windows in the later case Signed-off-by: Joel Cunningham <joel.cunningham@me.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years ago
int (*url_get_short_seek)(URLContext *h);
int (*url_shutdown)(URLContext *h, int flags);
int priv_data_size;
const AVClass *priv_data_class;
int flags;
int (*url_check)(URLContext *h, int mask);
int (*url_open_dir)(URLContext *h);
int (*url_read_dir)(URLContext *h, AVIODirEntry **next);
int (*url_close_dir)(URLContext *h);
int (*url_delete)(URLContext *h);
int (*url_move)(URLContext *h_src, URLContext *h_dst);
const char *default_whitelist;
} URLProtocol;
/**
* Create a URLContext for accessing to the resource indicated by
* url, but do not initiate the connection yet.
*
* @param puc pointer to the location where, in case of success, the
* function puts the pointer to the created URLContext
* @param flags flags which control how the resource indicated by url
* is to be opened
* @param int_cb interrupt callback to use for the URLContext, may be
* NULL
* @return >= 0 in case of success, a negative value corresponding to an
* AVERROR code in case of failure
*/
int ffurl_alloc(URLContext **puc, const char *filename, int flags,
const AVIOInterruptCB *int_cb);
/**
* Connect an URLContext that has been allocated by ffurl_alloc
*
* @param options A dictionary filled with options for nested protocols,
* i.e. it will be passed to url_open2() for protocols implementing it.
* This parameter will be destroyed and replaced with a dict containing options
* that were not found. May be NULL.
*/
int ffurl_connect(URLContext *uc, AVDictionary **options);
/**
* Create an URLContext for accessing to the resource indicated by
* url, and open it.
*
* @param puc pointer to the location where, in case of success, the
* function puts the pointer to the created URLContext
* @param flags flags which control how the resource indicated by url
* is to be opened
* @param int_cb interrupt callback to use for the URLContext, may be
* NULL
* @param options A dictionary filled with protocol-private options. On return
* this parameter will be destroyed and replaced with a dict containing options
* that were not found. May be NULL.
* @param parent An enclosing URLContext, whose generic options should
* be applied to this URLContext as well.
* @return >= 0 in case of success, a negative value corresponding to an
* AVERROR code in case of failure
*/
int ffurl_open_whitelist(URLContext **puc, const char *filename, int flags,
const AVIOInterruptCB *int_cb, AVDictionary **options,
const char *whitelist, const char* blacklist,
URLContext *parent);
int ffurl_open(URLContext **puc, const char *filename, int flags,
const AVIOInterruptCB *int_cb, AVDictionary **options);
/**
* Accept an URLContext c on an URLContext s
*
* @param s server context
* @param c client context, must be unallocated.
* @return >= 0 on success, ff_neterrno() on failure.
*/
int ffurl_accept(URLContext *s, URLContext **c);
/**
* Perform one step of the protocol handshake to accept a new client.
* See avio_handshake() for details.
* Implementations should try to return decreasing values.
* If the protocol uses an underlying protocol, the underlying handshake is
* usually the first step, and the return value can be:
* (largest value for this protocol) + (return value from other protocol)
*
* @param c the client context
* @return >= 0 on success or a negative value corresponding
* to an AVERROR code on failure
*/
int ffurl_handshake(URLContext *c);
/**
* Read up to size bytes from the resource accessed by h, and store
* the read bytes in buf.
*
* @return The number of bytes actually read, or a negative value
* corresponding to an AVERROR code in case of error. A value of zero
* indicates that it is not possible to read more from the accessed
* resource (except if the value of the size argument is also zero).
*/
int ffurl_read(URLContext *h, unsigned char *buf, int size);
/**
* Read as many bytes as possible (up to size), calling the
* read function multiple times if necessary.
* This makes special short-read handling in applications
* unnecessary, if the return value is < size then it is
* certain there was either an error or the end of file was reached.
*/
int ffurl_read_complete(URLContext *h, unsigned char *buf, int size);
/**
* Write size bytes from buf to the resource accessed by h.
*
* @return the number of bytes actually written, or a negative value
* corresponding to an AVERROR code in case of failure
*/
int ffurl_write(URLContext *h, const unsigned char *buf, int size);
/**
* Change the position that will be used by the next read/write
* operation on the resource accessed by h.
*
* @param pos specifies the new position to set
* @param whence specifies how pos should be interpreted, it must be
* one of SEEK_SET (seek from the beginning), SEEK_CUR (seek from the
* current position), SEEK_END (seek from the end), or AVSEEK_SIZE
* (return the filesize of the requested resource, pos is ignored).
* @return a negative value corresponding to an AVERROR code in case
* of failure, or the resulting file position, measured in bytes from
* the beginning of the file. You can use this feature together with
* SEEK_CUR to read the current file position.
*/
int64_t ffurl_seek(URLContext *h, int64_t pos, int whence);
/**
* Close the resource accessed by the URLContext h, and free the
* memory used by it. Also set the URLContext pointer to NULL.
*
* @return a negative value if an error condition occurred, 0
* otherwise
*/
int ffurl_closep(URLContext **h);
int ffurl_close(URLContext *h);
/**
* Return the filesize of the resource accessed by h, AVERROR(ENOSYS)
* if the operation is not supported by h, or another negative value
* corresponding to an AVERROR error code in case of failure.
*/
int64_t ffurl_size(URLContext *h);
/**
* Return the file descriptor associated with this URL. For RTP, this
* will return only the RTP file descriptor, not the RTCP file descriptor.
*
* @return the file descriptor associated with this URL, or <0 on error.
*/
int ffurl_get_file_handle(URLContext *h);
/**
* Return the file descriptors associated with this URL.
*
* @return 0 on success or <0 on error.
*/
int ffurl_get_multi_file_handle(URLContext *h, int **handles, int *numhandles);
HTTP: improve performance by reducing forward seeks This commit optimizes HTTP performance by reducing forward seeks, instead favoring a read-ahead and discard on the current connection (referred to as a short seek) for seeks that are within a TCP window's worth of data. This improves performance because with TCP flow control, a window's worth of data will be in the local socket buffer already or in-flight from the sender once congestion control on the sender is fully utilizing the window. Note: this approach doesn't attempt to differentiate from a newly opened connection which may not be fully utilizing the window due to congestion control vs one that is. The receiver can't get at this information, so we assume worst case; that full window is in use (we did advertise it after all) and that data could be in-flight The previous behavior of closing the connection, then opening a new with a new HTTP range value results in a massive amounts of discarded and re-sent data when large TCP windows are used. This has been observed on MacOS/iOS which starts with an initial window of 256KB and grows up to 1MB depending on the bandwidth-product delay. When seeking within a window's worth of data and we close the connection, then open a new one within the same window's worth of data, we discard from the current offset till the end of the window. Then on the new connection the server ends up re-sending the previous data from new offset till the end of old window. Example (assumes full window utilization): TCP window size: 64KB Position: 32KB Forward seek position: 40KB * (Next window) 32KB |--------------| 96KB |---------------| 160KB * 40KB |---------------| 104KB Re-sent amount: 96KB - 40KB = 56KB For a real world test example, I have MP4 file of ~25MB, which ffplay only reads ~16MB and performs 177 seeks. With current ffmpeg, this results in 177 HTTP GETs and ~73MB worth of TCP data communication. With this patch, ffmpeg issues 4 HTTP GETs and 3 seeks for a total of ~22MB of TCP data communication. To support this feature, the short seek logic in avio_seek() has been extended to call a function to get the short seek threshold value. This callback has been plumbed to the URLProtocol structure, which now has infrastructure in HTTP and TCP to get the underlying receiver window size via SO_RCVBUF. If the underlying URL and protocol don't support returning a short seek threshold, the default s->short_seek_threshold is used This feature has been tested on Windows 7 and MacOS/iOS. Windows support is slightly complicated by the fact that when TCP window auto-tuning is enabled, SO_RCVBUF doesn't report the real window size, but it does if SO_RCVBUF was manually set (disabling auto-tuning). So we can only use this optimization on Windows in the later case Signed-off-by: Joel Cunningham <joel.cunningham@me.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years ago
/**
* Return the current short seek threshold value for this URL.
*
* @return threshold (>0) on success or <=0 on error.
*/
int ffurl_get_short_seek(URLContext *h);
/**
* Signal the URLContext that we are done reading or writing the stream.
*
* @param h pointer to the resource
* @param flags flags which control how the resource indicated by url
* is to be shutdown
*
* @return a negative value if an error condition occurred, 0
* otherwise
*/
int ffurl_shutdown(URLContext *h, int flags);
/**
* Check if the user has requested to interrupt a blocking function
* associated with cb.
*/
int ff_check_interrupt(AVIOInterruptCB *cb);
/* udp.c */
int ff_udp_set_remote_url(URLContext *h, const char *uri);
int ff_udp_get_local_port(URLContext *h);
/**
* Assemble a URL string from components. This is the reverse operation
* of av_url_split.
*
* Note, this requires networking to be initialized, so the caller must
* ensure ff_network_init has been called.
*
* @see av_url_split
*
* @param str the buffer to fill with the url
* @param size the size of the str buffer
* @param proto the protocol identifier, if null, the separator
* after the identifier is left out, too
* @param authorization an optional authorization string, may be null.
* An empty string is treated the same as a null string.
* @param hostname the host name string
* @param port the port number, left out from the string if negative
* @param fmt a generic format string for everything to add after the
* host/port, may be null
* @return the number of characters written to the destination buffer
*/
int ff_url_join(char *str, int size, const char *proto,
const char *authorization, const char *hostname,
int port, const char *fmt, ...) av_printf_format(7, 8);
/**
* Convert a relative url into an absolute url, given a base url.
*
* @param buf the buffer where output absolute url is written
* @param size the size of buf
* @param base the base url, may be equal to buf.
* @param rel the new url, which is interpreted relative to base
*/
void ff_make_absolute_url(char *buf, int size, const char *base,
const char *rel);
/**
* Allocate directory entry with default values.
*
* @return entry or NULL on error
*/
AVIODirEntry *ff_alloc_dir_entry(void);
const AVClass *ff_urlcontext_child_class_next(const AVClass *prev);
/**
* Construct a list of protocols matching a given whitelist and/or blacklist.
*
* @param whitelist a comma-separated list of allowed protocol names or NULL. If
* this is a non-empty string, only protocols in this list will
* be included.
* @param blacklist a comma-separated list of forbidden protocol names or NULL.
* If this is a non-empty string, all protocols in this list
* will be excluded.
*
* @return a NULL-terminated array of matching protocols. The array must be
* freed by the caller.
*/
const URLProtocol **ffurl_get_protocols(const char *whitelist,
const char *blacklist);
#endif /* AVFORMAT_URL_H */