The IP version choice was previously only in the UserDefined struct
within the SessionHandle, but since we sometimes alter that option
during a request we need to have it on a per-connection basis.
I also moved more "init conn" code into the allocate_conn() function
which is designed for that purpose more or less.
As this function uses return code 0 to mean that there is no timeout, it
needs to check that it doesn't return a time left value that is exactly
zero. It could lead to libcurl doing an extra 1000 ms select() call and
thus not timing out as accurately as it should.
I fell over this bug when working on the bug 3061535 but this fix does
not correct that problem alone, although this is a problem that needs to
be fixed.
Reported by: Rodric Glaser
Bug: http://curl.haxx.se/bug/view.cgi?id=3061535
The timeout is set for the connect phase already at the start of the
request so we should not add a new one, and we MUST not set expire to 0
as that will remove any other potentially existing timeouts.
Curl_getconnectinfo() is changed to return a proper curl_socket_t for
the last socket so that it'll work more portably (and cause less
compiler warnings).
When the progress callback is called during the TCP connection, an error
return would accidentally not abort the operation as intended but would
instead be counted as a failure to connect to that particular IP and
libcurl would just continue to try the next. I made singleipconnect()
and trynextip() return CURLcode properly.
Added bonus: it corrected the error code for bad --interface usages,
like tested in test 1084 and test 1085.
Reported by: Adam Light
Bug: http://curl.haxx.se/mail/lib-2010-08/0105.html
When a hostname resolves to multiple IP addresses and the first one
tried doesn't work, the socket for the second attempt may get dropped on
the floor, causing the request to eventually time out. The issue is that
when using kqueue (as on mac and bsd platforms) instead of select, the
kernel removes the first fd from kqueue when it is closed (in trynextip,
connect.c:503). Trynextip() then goes on to open a new socket, which
gets assigned the same number as the one it just closed. Later in
multi.c, socket_cb is not called because the fd is already in
multi->sockhash, so the new socket is never added to kqueue.
The correct fix is to ensure that socket_cb is called to remove the fd
when trynextip() closes the socket, and again to re-add it after
singleipsocket(). I'm not sure how to cleanly do that, but the attached
patch works around the problem in an admittedly kludgy way by delaying
the close to ensure that the newly-opened socket gets a different fd.
Daniel's added comment: I didn't spot a way to easily do a nicer fix so
I've proceeded with Ben's patch.
Bug: http://curl.haxx.se/bug/view.cgi?id=3017819
Patch by: Ben Darnell
Dirk Manske reported a regression. When connecting with the multi
interface, there were situations where libcurl wouldn't store
connect time correctly as it used to (and is documented to) do.
Using his fine sample program we could repeat it, and I wrote up
test case 573 using that code. The problem does not easily show
itself using the local test suite though.
The fix, also as suggested by Dirk, is a bit on the ugly side as
it adds yet another call to Curl_verboseconnect() and setting the
TIMER_CONNECT time. That situation is subject for some closer
inspection in the future.
Kenny To filed the bug report #2963679 with patch to fix a
problem he experienced with doing multi interface HTTP POST over
a proxy using PROXYTUNNEL. He found a case where it would connect
fine but bits.tcpconnect was not set correct so libcurl didn't
work properly.
(http://curl.haxx.se/bug/view.cgi?id=2963679)
(http://curl.haxx.se/bug/view.cgi?id=2870221) that libcurl returned an
incorrect return code from the internal trynextip() function which caused
him grief. This is a regression that was introduced in 7.19.1 and I find it
strange it hasn't hit us harder, but I won't persue into figuring out
exactly why.
SO_SNDBUF to CURL_WRITE_SIZE even if the SO_SNDBUF starts out larger. The
patch doesn't do a setsockopt if SO_SNDBUF is already greater than
CURL_WRITE_SIZE. This should help folks who have set up their computer with
large send buffers.
problem with my CURLINFO_PRIMARY_IP fix from October 7th that caused a NULL
pointer read. I also took the opportunity to clean up this logic (storing of
the connection's IP address) somewhat as we had it stored in two different
places and ways previously and they are now unified.
systems supporting getifaddrs(). Also fixed a problem where an IPv6
address could be chosen instead of an IPv4 one for --interface when it
involved a name lookup.
parser to allow numerical IPv6-addresses to be specified with the scope
given, as per RFC4007 - with a percent letter that itself needs to be URL
escaped. For example, for an address of fe80::1234%1 the HTTP URL is:
"http://[fe80::1234%251]/"
curl_easy_getinfo. It returns a pointer to a string with the most recently
used IP address. Modified test case 500 to also verify this feature. The
implementing of this feature was sponsored by Lenny Rachitsky at NeuStar.
and receive data over a connection previously setup with curl_easy_perform()
and its CURLOPT_CONNECT_ONLY option. The sendrecv.c example was added to
show how they can be used.
since libcurl used getprotobyname() and that isn't thread-safe. We now
switched to use IPPROTO_TCP unconditionally, but perhaps the proper fix is
to detect the thread-safe version of the function and use that.
http://curl.haxx.se/mail/lib-2008-05/0011.html
This happened because the tftp code always uncondionally did a bind()
without caring if one already had been done and then it failed. I wrote a
test case (1009) to verify this, but it is a bit error-prone since it will
have to pick a fixed local port number and since the tests are run on so
many different hosts in different situations I add it in disabled state.
CURLOPT_OPENSOCKETDATA to set a callback that allows an application to replace
the socket() call used by libcurl. It basically allows the app to change
address, protocol or whatever of the socket. (I also did some whitespace
indent/cleanups in lib/url.c which kind of hides some of these changes, sorry
for mixing those in.)
and allow reuse by multiple protocols. Several unused error codes were
removed. In all cases, macros were added to preserve source (and binary)
compatibility with the old names. These macros are subject to removal at
a future date, but probably not before 2009. An application can be
tested to see if it is using any obsolete code by compiling it with the
CURL_NO_OLDIES macro defined.
Documented some newer error codes in libcurl-error(3)
passed to it with curl_easy_setopt()! Previously it has always just refered
to the data, forcing the user to keep the data around until libcurl is done
with it. That is now history and libcurl will instead clone the given
strings and keep private copies.
(http://curl.haxx.se/bug/view.cgi?id=1750274) and submitted a patch for the
case where libcurl did a connect attempt to a non-listening port and didn't
provide a human readable error string back.
function that deprecates the curl_multi_socket() function. Using the new
function the application tell libcurl what action that was found in the
socket that it passes in. This gives a significant performance boost as it
allows libcurl to avoid a call to poll()/select() for every call to
curl_multi_socket*().
Check for lowercase 'bool' type at configuration stage. If not available
provide a suitable replacement with a type definition of 'unsigned char'
in setup_once.h
Move definitions of TRUE and FALSE to setup_once.h
and CURLOPT_CONNECTTIMEOUT_MS that, as their names should hint, do the
timeouts with millisecond resolution instead. The only restriction to that
is the alarm() (sometimes) used to abort name resolves as that uses full
seconds. I fixed the FTP response timeout part of the patch.
Internally we now count and keep the timeouts in milliseconds but it also
means we multiply set timeouts with 1000. The effect of this is that no
timeout can be set to more than 2^31 milliseconds (on 32 bit systems), which
equals 24.86 days. We probably couldn't before either since the code did
*1000 on the timeout values on several places already.
(http://curl.haxx.se/bug/view.cgi?id=1481217), with follow-ups by Michele Bini
and David Byron. libcurl previously wrongly used GetLastError() on windows to
get error details after socket-related function calls, when it really should
use WSAGetLastError() instead.
When changing to this, the former function Curl_ourerrno() is now instead
called Curl_sockerrno() as it is necessary to only use it to get errno from
socket-related functions as otherwise it won't work as intended on Windows.
curl tool with --local-port. Plain and simply set the range of ports to bind
the local end of connections to. Implemented on to popular demand.
Not extensively tested. Please let me know how it works.
app to retrieve the errno variable after a (connect) failure. It will make
sense to provide this for more failures in a more generic way, but let's
start like this.
all things up to work with encoded host names internally, as well as keeping
'display names' to show in debug messages. IDN resolves work for me now using
ipv6, ipv4 and ares resolving. Even cookies on IDN sites seem to do right.
sessionhandle to make the duphandle() function work as supposed. Also tried
to start document functions the doxygen way (in the headers of the functions).
Can't make it work though...
I guess it seldomly happens on linux and that's why it wasn't found before.
He used Solaris to notice it.
I took the opportunity to rewrite the Curl_connecthost() slightly to feature
less duplicate code in the two different versions (ipv4/ipv6).
hogs Windows machines when libcurl is being used multi-threaded (with > ~50
threads). Andrew Fuller helped us verify and test this.
According to a MSDN web page on connect(), it returns 0 when the connect
is done and thus we don't need the getsockopt() call anyway on Windows.
counter so that the caller needs to decrease that counter when done with
the returned data.
If compiled with MALLOCDEBUG I've added some extra checking that the counter
is decreased before a handle is closed etc.
any name resolved data in any curl handle struct. That way, we won't mind
if the cache entries are pruned for the next time we need them. We'll just
resolve them again instead.
This changes the Curl_resolv() proto. It modifies the SessionHandle struct
but perhaps most importantly, it'll make the internals somewhat dependent
on the DNS cache not being disabled as that will cripple operations somewhat.
Especially for persistant connections.
