curl_easy_pause.3: add multiplexed pause effects

and generally refresh and update. Remove details for ancient versions.

Reviewed-by: Jay Satiro
Closes #6360
This commit is contained in:
Daniel Stenberg 2020-12-22 16:08:43 +01:00
parent f0ba3d5e1b
commit 64e6e54f4b
No known key found for this signature in database
GPG Key ID: 5CC908FDB71E12C2
1 changed files with 23 additions and 23 deletions

View File

@ -23,10 +23,11 @@
.SH NAME .SH NAME
curl_easy_pause - pause and unpause a connection curl_easy_pause - pause and unpause a connection
.SH SYNOPSIS .SH SYNOPSIS
.nf
.B #include <curl/curl.h> .B #include <curl/curl.h>
.BI "CURLcode curl_easy_pause(CURL *"handle ", int "bitmask " );" .BI "CURLcode curl_easy_pause(CURL *"handle ", int "bitmask ");"
.fi
.SH DESCRIPTION .SH DESCRIPTION
Using this function, you can explicitly mark a running connection to get Using this function, you can explicitly mark a running connection to get
paused, and you can unpause a connection that was previously paused. paused, and you can unpause a connection that was previously paused.
@ -36,18 +37,18 @@ the write callbacks return the proper magic return code
(\fICURL_READFUNC_PAUSE\fP and \fICURL_WRITEFUNC_PAUSE\fP). A write callback (\fICURL_READFUNC_PAUSE\fP and \fICURL_WRITEFUNC_PAUSE\fP). A write callback
that returns pause signals to the library that it couldn't take care of any that returns pause signals to the library that it couldn't take care of any
data at all, and that data will then be delivered again to the callback when data at all, and that data will then be delivered again to the callback when
the writing is later unpaused. the transfer is unpaused.
While it may feel tempting, take care and notice that you cannot call this While it may feel tempting, take care and notice that you cannot call this
function from another thread. To unpause, you may for example call it from the function from another thread. To unpause, you may for example call it from the
progress callback (\fICURLOPT_PROGRESSFUNCTION(3)\fP), which gets called at progress callback (\fICURLOPT_PROGRESSFUNCTION(3)\fP), which gets called at
least once per second, even if the connection is paused. least once per second, even if the connection is paused.
When this function is called to unpause reading, the chance is high that you When this function is called to unpause receiving, the chance is high that you
will get your write callback called before this function returns. will get your write callback called before this function returns.
The \fBhandle\fP argument is of course identifying the handle that operates on The \fBhandle\fP argument identifies the transfer you want to pause or
the connection you want to pause or unpause. unpause.
A paused transfer is excluded from low speed cancels via the A paused transfer is excluded from low speed cancels via the
\fICURLOPT_LOW_SPEED_LIMIT(3)\fP option and unpausing a transfer will reset \fICURLOPT_LOW_SPEED_LIMIT(3)\fP option and unpausing a transfer will reset
@ -75,25 +76,24 @@ code means something wrong occurred after the new state was set. See the
The pausing of transfers does not work with protocols that work without The pausing of transfers does not work with protocols that work without
network connectivity, like FILE://. Trying to pause such a transfer, in any network connectivity, like FILE://. Trying to pause such a transfer, in any
direction, will cause problems in the worst case or an error in the best case. direction, will cause problems in the worst case or an error in the best case.
.SH AVAILABILITY .SH MULTIPLEXED
This function was added in libcurl 7.18.0. Before this version, there was no When a connection is used multiplexed, like for HTTP/2, and one of the
explicit support for pausing transfers. transfers over the connection is paused and the others continue flowing,
.SH "USAGE WITH THE MULTI-SOCKET INTERFACE" libcurl might end up buffering contents for the paused transfer. It has to do
Before libcurl 7.32.0, when a specific handle was unpaused with this function, this because it needs to drain the socket for the other transfers and the
there was no particular forced rechecking or similar of the socket's state, already announced window size for the paused transfer will allow the server to
which made the continuation of the transfer get delayed until next continue sending data up to that window size amount. By default, libcurl
multi-socket call invoke or even longer. Alternatively, the user could announces a 32 megabyte window size, which thus can make libcurl end up
forcibly call for example \fIcurl_multi_socket_all(3)\fP - with a rather hefty buffering 32 megabyte of data for a paused stream.
performance penalty.
Starting in libcurl 7.32.0, unpausing a transfer will schedule a timeout When such a paused stream is unpaused again, any buffered data will be
trigger for that handle 1 millisecond into the future, so that a delivered first.
curl_multi_socket_action( ... CURL_SOCKET_TIMEOUT) can be used immediately .SH AVAILABILITY
afterwards to get the transfer going again as desired. Added in libcurl 7.18.0.
.SH "MEMORY USE" .SH "MEMORY USE"
When pausing a read by returning the magic return code from a write callback, When pausing a read by returning the magic return code from a write callback,
the read data is already in libcurl's internal buffers so it'll have to keep the read data is already in libcurl's internal buffers so it'll have to keep
it in an allocated buffer until the reading is again unpaused using this it in an allocated buffer until the receiving is again unpaused using this
function. function.
If the downloaded data is compressed and is asked to get uncompressed If the downloaded data is compressed and is asked to get uncompressed
@ -101,7 +101,7 @@ automatically on download, libcurl will continue to uncompress the entire
downloaded chunk and it will cache the data uncompressed. This has the side- downloaded chunk and it will cache the data uncompressed. This has the side-
effect that if you download something that is compressed a lot, it can result effect that if you download something that is compressed a lot, it can result
in a very large data amount needing to be allocated to save the data during in a very large data amount needing to be allocated to save the data during
the pause. This said, you should probably consider not using paused reading if the pause. This said, you should probably consider not using paused receiving
you allow libcurl to uncompress data automatically. if you allow libcurl to uncompress data automatically.
.SH "SEE ALSO" .SH "SEE ALSO"
.BR curl_easy_cleanup "(3), " curl_easy_reset "(3)" .BR curl_easy_cleanup "(3), " curl_easy_reset "(3)"