Daniel Stenberg [Tue, 19 Apr 2005 23:36:21 +0000 (23:36 +0000)]
indented source to look more like other ares code,
added (somewhat ugly) typecasts to build warning-free on 64bit platforms (the
result of a (char *) - (char *) cannot be stored in an int universally)
Daniel Stenberg [Tue, 19 Apr 2005 22:12:34 +0000 (22:12 +0000)]
when --with-gnutls is used, we assume a bin/libgnutls-config file in the
given prefix. Building something with gnutls without it just is too error-
prone.
Daniel Stenberg [Mon, 18 Apr 2005 19:41:04 +0000 (19:41 +0000)]
Olivier reported that even though he used CURLOPT_PORT, libcurl clearly still
used the default port. He was right. I fixed the problem and added the test
cases 521, 522 and 523 to verify the fix.
Daniel Stenberg [Mon, 18 Apr 2005 06:57:44 +0000 (06:57 +0000)]
Modified the FTP server to use the new 'sockfilt' program to do all the socket
level stuff. The FTP server communicates with sockfilt using perl's open2().
This enables easier IPv6 support and hopefully FTP-SSL support in the future.
Added four test cases for FTP-ipv6.
Daniel Stenberg [Mon, 18 Apr 2005 05:46:10 +0000 (05:46 +0000)]
Modified to not mix ordinary print to STDOUT with a system() that prints to
stdout, since I've found cases on Solaris where the second output mixes with
the first and thus the big check-script doesn't properly find the first
string in the output stream.
Daniel Stenberg [Fri, 15 Apr 2005 08:45:35 +0000 (08:45 +0000)]
if libgnutls-config isn't found in the given path, deal with it nicer (but
it is still likely to not do very good since it can't figure out all the lib
dependencies)
Daniel Stenberg [Tue, 12 Apr 2005 07:56:57 +0000 (07:56 +0000)]
Provides an unconditional strlcat() proto even if strlcat() was found by
configure. An attempt to fix warnings when we build and the strlcat() function
is provided by one if the libs (gss or krb4) since then we have no protos
for it in a system header.
Daniel Stenberg [Thu, 7 Apr 2005 22:47:43 +0000 (22:47 +0000)]
Unfortunately, if a ca file name is set the function fails for whatever reason
(missing file, bad file, etc), gnutls will no longer handshake properly but it
just loops forever. Therefore, we must return error if we get an error when
setting the CA cert file name. This is not the same behaviour as with OpenSSL.
Question/report posted to the help-gnutls mailing list, April 8 2005.
Daniel Stenberg [Thu, 7 Apr 2005 15:27:13 +0000 (15:27 +0000)]
GnuTLS support added. There's now a "generic" SSL layer that we use all over
internally, with code provided by sslgen.c. All SSL-layer-specific code is
then written in ssluse.c (for OpenSSL) and gtls.c (for GnuTLS).
As far as possible, internals should not need to know what SSL layer that is
in use. Building with GnuTLS currently makes two test cases fail.
TODO.gnutls contains a few known outstanding issues for the GnuTLS support.
GnuTLS support is enabled with configure --with-gnutls
Daniel Stenberg [Thu, 7 Apr 2005 15:18:51 +0000 (15:18 +0000)]
ssl_version_num won't be used anymore since we will soon offer multiple SSL
layers and it won't make sense to provide a numerical version for it. I also
doubt that many people have used this for anything critical.
Daniel Stenberg [Thu, 7 Apr 2005 07:38:53 +0000 (07:38 +0000)]
Try harder to see if arpa/nameser_compat.h REALLY is a good header file to
include, as it seems at least some AIX versions don't really allow it to be
include at the same time as the original nameser.h.
Dominick Meglio [Tue, 5 Apr 2005 18:26:55 +0000 (18:26 +0000)]
Provided implementations of inet_net_pton and inet_pton from BIND for systems that do not include these functions. These will be necessary for CIDR support and IPv6 support.