Yang Tse [Wed, 2 Jan 2013 19:12:15 +0000 (20:12 +0100)]
build and tests: curl_10char_object_name() shell function
lib/objnames.inc provides definition of curl_10char_object_name() shell
function. The intended purpose of this function is to transliterate a
(*.c) source file name that may be longer than 10 characters, or not,
into a string with at most 10 characters which may be used as an OS/400
object name.
Test case 1221 does unit testng of this function and also verifies
that it is possible to generate distinct short object names for all
curl and libcurl *.c source file names.
lib/objnames-test.sh is the shell script used for test case 1221.
tests/runtests.pl modified to accept shell script test cases.
More details inside lib/objnames.inc and lib/objnames-test.sh
Yang Tse [Tue, 1 Jan 2013 18:24:54 +0000 (19:24 +0100)]
configure.ac: replace AM_CONFIG_HEADER with AC_CONFIG_HEADERS
automake 1.13 errors if AM_CONFIG_HEADER is used in configure script.
automake 1.13 no longer autoupdates AM_CONFIG_HEADER to
AC_CONFIG_HEADERS, thing which automake has been doing since automake
version 1.7
Given that our first automake supported version is automake 1.7,
simply replacing AM_CONFIG_HEADER usage with AC_CONFIG_HEADERS seems
enough to yet support same automake versions.
Dave Reisner reported issue with 1.13 and provided patch.
Yang Tse [Tue, 1 Jan 2013 16:26:32 +0000 (17:26 +0100)]
curl-override.m4: provide AC_CONFIG_MACRO_DIR definition conditionally
Provide a 'traceable' AC_CONFIG_MACRO_DIR definition only when using
an autoconf version that does not provide it, instead of what we were
doing up to now of providing and overriding AC_CONFIG_MACRO_DIR for
all autoconf versions.
Steve Holme [Sun, 30 Dec 2012 12:44:09 +0000 (12:44 +0000)]
imap: Code tidy up prior to adding support for the CAPABILITY command
* Changing the order of the state machine to represent the order in
which commands are sent to the server.
* Reworking the imap_endofresp() function as the FETCH response doesn't
include the command id and shouldn't be part of the length comparison
that takes into account the id string.
Steve Holme [Sat, 29 Dec 2012 19:23:10 +0000 (19:23 +0000)]
imap.c: Fixed bad state error when logging in with invalid credentials
Fixed a problem with the state machine when attempting to log in with
invalid credentials. The server would report login failure but libcurl
would not read the response due to inappropriate IMAP_STOP states being
set after the login was sent.
Nick Zitzmann [Thu, 27 Dec 2012 01:29:50 +0000 (17:29 -0800)]
darwinssl: Fixed inability to disable peer verification
... on Snow Leopard and Lion
Snow Leopard introduced the SSLSetSessionOption() function, but it
doesn't disable peer verification as expected on Snow Leopard or
Lion (it works as expected in Mountain Lion). So we now use sysctl()
to detect whether or not the user is using Snow Leopard or Lion,
and if that's the case, then we now use the deprecated
SSLSetEnableCertVerify() function instead to disable peer verification.
Marc Hoersken [Tue, 25 Dec 2012 14:15:24 +0000 (15:15 +0100)]
socklift.c: Added select_ws function to support Windows
WinSock select() does not support standard file descriptors,
it can only check SOCKETs. The following function is an attempt
to create a select() function with support for other handles.
Daniel Stenberg [Sat, 15 Dec 2012 14:31:38 +0000 (15:31 +0100)]
pop3_doing: don't call pop3_dophase_done() if already failed
... it also clobbered the 'result' return value so that it wouldn't
return the error back to the parent function properly, which broke test
809 when run with 'multi-always'.
Daniel Stenberg [Fri, 21 Dec 2012 13:38:33 +0000 (14:38 +0100)]
SCP: relative path didn't work
When prefixing a path with /~/ it is supposed to be used relative to the
user's home directory but it didn't work. Now we cut off the entire
three byte sequenct "/~/" which seems to be how OpenSSH does it.
Issue: When building a 32bit target with large file support HP-UX
<sys/socket.h> header file may simultaneously provide two different
sets of declarations for sendfile and sendpath functions, one with
static and another with external linkage. Given that we do not use
mentioned functions we really don't care which linkage is the
appropriate one, but on the other hand, the double declaration emmits
warnings when using the HP-UX compiler and errors when using modern
gcc versions resulting in fatal compilation errors.
Mentioned issue is now fixed as long as we don't use sendfile nor
sendpath functions.
Colin Watson [Fri, 7 Dec 2012 11:07:54 +0000 (12:07 +0100)]
configure: fix cross pkg-config detection
When cross-compiling, CURL_CHECK_PKGCONFIG was checking for the cross
pkg-config using ${host}-pkg-config.
The gold standard for doing this correctly is pkg-config's own macro,
PKG_PROG_PKG_CONFIG. However, on the assumption that you have a good
reason not to use that directly (reduced dependencies for maintainer
builds?), the behaviour of cURL's version should at least match.
PKG_PROG_PKG_CONFIG uses AC_PATH_TOOL, which ultimately ends up trying
${host_alias}-pkg-config; this is not quite the same as what cURL does,
and may differ because ${host} has been run through config.sub. For
instance, when cross-building to the armhf architecture on Ubuntu,
${host_alias} is arm-linux-gnueabihf while ${host} is
arm-unknown-linux-gnueabihf. This may also have been the cause of the
problem reported at http://curl.haxx.se/mail/lib-2012-04/0224.html.
AC_PATH_TOOL is significantly simpler than cURL's current code, and
dates back to well before the current minimum of Autoconf 2.57, so let's
use it instead.
Introducing a new persistent connection caching system using "bundles".
A bundle is a list of all persistent connections to the same host.
The connection cache consists of a hash of bundles, with the
hostname as the key.
The benefits may not be obvious, but they are two:
1) Faster search for connections to reuse, since the hash
lookup only finds connections to the host in question.
2) It lays out the groundworks for an upcoming patch,
which will introduce multiple HTTP pipelines.
This patch also removes the awkward list of "closure handles",
which were needed to send QUIT commands to the FTP server
when closing a connection.
Now we allocate a separate closure handle and use that
one to close all connections.
This has been tested in a live system for a few weeks, and of
course passes the test suite.
Yang Tse [Tue, 4 Dec 2012 22:30:05 +0000 (23:30 +0100)]
build: explain current role of LIBS in our Makefile.am files
BLANK_AT_MAKETIME may be used in our Makefile.am files to blank
LIBS variable used in generated makefile at makefile processing
time. Doing this functionally prevents LIBS from being used for
all link targets in given makefile.
Daniel Stenberg [Fri, 30 Nov 2012 13:52:38 +0000 (14:52 +0100)]
multi: fix re-sending request on early connection close
This handling already works with the easy-interface code. When a request
is sent on a re-used connection that gets closed by the server at the
same time as the request is sent, the situation may occur so that we can
send the request and we discover the broken connection as a RECV_ERROR
in the PERFORM state and then the request needs to be retried on a fresh
connection. Test 64 broke with 'multi-always-internally'.
Kamil Dudka [Mon, 3 Dec 2012 12:17:50 +0000 (13:17 +0100)]
nss: prevent NSS from crashing on client auth hook failure
Although it is not explicitly stated in the documentation, NSS uses
*pRetCert and *pRetKey even if the client authentication hook returns
a failure. Namely, if we destroy *pRetCert without clearing *pRetCert
afterwards, NSS destroys the certificate once again, which causes a
double free.
Fabian Keil [Sun, 25 Nov 2012 17:34:01 +0000 (18:34 +0100)]
Remove stray CRLF in chunk-encoded content-free request bodies
.. that are sent when auth-negotiating before a chunked
upload or when setting the 'Transfer-Encoding: chunked'
header and intentionally sending no content.
projects / curl / log