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>
Since all URLContexts have the same AVOptions, such AVOptions
will be applied on the outermost context only and removed from the
dict, while they probably make sense on all contexts.
This makes sure that rw_timeout gets propagated to the innermost
URLContext (to make sure it gets passed to the tcp protocol, when
opening a http connection for instance).
Alternatively, such matching options would be kept in the dict
and only removed after the ffurl_connect call.
Signed-off-by: Martin Storsjö <martin@martin.st>
If set non-zero, this limits duration of the retry_transfer_wrapper()
loop, thus affecting ffurl_read*(), ffurl_write(). As soon as
one single byte is successfully received/transmitted, the timer
restarts.
This has further changes by Michael Niedermayer and Martin Storsjö.
Signed-off-by: Martin Storsjö <martin@martin.st>
Instead of a linked list constructed at av_register_all(), store them
in a constant array of pointers.
Since no registration is necessary now, this removes some global state
from lavf. This will also allow the urlprotocol layer caller to limit
the available protocols in a simple and flexible way in the following
commits.
Note to maintainers: update tools
Note to maintainers: set a default whitelist for your protocol
If that makes no sense then consider to set "none" and thus require the user to specify a white-list
for sub-protocols to be opened
Note, testing and checking for missing changes is needed
Reviewed-by: Andreas Cadhalpun <andreas.cadhalpun@googlemail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
API allows protocol implementations to provide API that
allows to list directory content.
API is similar to POSIX opendir/readdir/closedir.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This was added in 9b07a2dc02 as an ABI hack to allow older
code built with lavf 52 to register protocols even if the size
of the URLProtocol struct was increased. Later, registering
protocols from outside of lavf was removed and this workaround
isn't needed any longer since lavf 53.
This removes an unchecked malloc and a memory leak for the cases
when this workaround actually was used - which it hasn't since
lavf 53.
Signed-off-by: Martin Storsjö <martin@martin.st>
Define positive return values as non errors and leave further meaning undefined
This allows future extensions to use these values
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
If set non-zero, limits duration of retry_transfer_wrapper() loop, thus
affects ffurl_read*(), ffurl_write()
Measured in microseconds.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This definition is in two files, since the definitions will move
to the private header at the next bump.
Signed-off-by: Martin Storsjö <martin@martin.st>
The interrupt callback has to be passed in during opening (setting it
after opening isn't enough), since a blocking open couldn't be
interrupted otherwise.
Options are passed down to procotols and also need to be available
during open() in most cases.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
This is a better io interrupt callback function, which has an
opaque parameter, which is given to the interrupt callback.
This allows callers to precisely cancel IO for one single
AVFormatContext, without interrupt other ones in the same
process.
Note, it's not needed in AVIOContext, at the moment.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
The new function is more flexible than url_exist(), as it allows to
specify which access flags to check, and does not require an explicit
open of the checked resource.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Joel Cunningham
Vittorio Giovara
Martin Storsjö
Andrey Utkin
Derek Buitenhuis
Anton Khirnov
Diego Biurrun
Michael Niedermayer
Stephan Holljes
Mariusz Szczepańczyk
Lukasz Marek
Michael Niedermayer
Luca Barbato
Jordi Ortiz
Samuel Pitoiset
Nicolas George
Stefano Sabatini