Daniel Stenberg [Wed, 31 Mar 2004 12:55:24 +0000 (12:55 +0000)]
The asynch name resolve methods now all use CURL_TIMEOUT_RESOLVE for
the specific time to wait for a resolve. The definition is at the top of
this source file.
Daniel Stenberg [Wed, 31 Mar 2004 11:55:07 +0000 (11:55 +0000)]
* Fixed a memory leak when doing repeated re-negotiations.
* Made the incoming line parser more forgiving to allow "name=contents" pairs
where the contents isn't within double quotes.
* Made the digest code return CURLDIGEST_BADALGO if a requested algorithm
isn't supported by the code.
Daniel Stenberg [Wed, 31 Mar 2004 11:50:44 +0000 (11:50 +0000)]
Added "swsbounce" magic: if this keyword is present in a <data> section it
sets the "swsbounce" magic mode. If there follows a request for the SAME
test number and the SAME part number, this mode will make the server bump
the part number internally and thus return a different <dataNUM> section
than it otherwise would.
Test case 153 uses this in case you need an example. It is pretty involved
and hard-to-use, but then the situation is pretty special over all. Enjoy.
Daniel Stenberg [Wed, 31 Mar 2004 10:59:48 +0000 (10:59 +0000)]
Moved the NI_WITHSCOPEID magic #ifdef to the top of the file and made sure
we use the NIFLAGS properly on both places in the code that use getnameinfo().
Daniel Stenberg [Wed, 31 Mar 2004 10:31:08 +0000 (10:31 +0000)]
modified the NI_WITHSCOPEID to use an AF_INET6 socket immediately and
added some more debug output to make it easier to detect failure reasons
in the autobuild logs
Daniel Stenberg [Tue, 30 Mar 2004 10:35:54 +0000 (10:35 +0000)]
Added CURL_CHECK_NI_WITHSCOPEID that checks if NI_WITHSCOPEID exists and
works. No code actually uses the HAVE_NI_WITHSCOPEID (that a positive test
results in), but this is still only for testing purposes.
Daniel Stenberg [Tue, 30 Mar 2004 06:42:12 +0000 (06:42 +0000)]
David Byron's patch was appplied to make CURLOPT_FAILONERROR work nicely
even with authentcations such as NTLM or Digest enabled. Test cases 150, 151
and 152 were added to verify the functionality.
Daniel Stenberg [Mon, 29 Mar 2004 21:29:24 +0000 (21:29 +0000)]
The select() timeout is better not static since some implementation actually
might change it. I don't *think* it does it when the timeout is 0,0 but it
is better to be sure...
Daniel Stenberg [Fri, 26 Mar 2004 13:47:46 +0000 (13:47 +0000)]
Gisle Vanem:
A patch to bypass MS' sillyness with regard to IPv6 and getaddrinfo().
The CURLDEBUG part is to avoid redefinition warning caused by memdebug.h. If
ENABLE_IPV6 isn't enabled, it doesn't matter since we never call
getaddrinfo(). Allthough we could to support weird protocols like SOCK_RDM
that Win-2K/XP has.
Daniel Stenberg [Thu, 25 Mar 2004 12:45:01 +0000 (12:45 +0000)]
so there are at least two different strerror_r() versions and our brand
new configure script detects them and now this code acts according to what
API that was detected
Daniel Stenberg [Thu, 25 Mar 2004 12:15:00 +0000 (12:15 +0000)]
strerror_r() detection changes:
1. Try with _THREAD_SAFE instead of _REENTRANT, as AIX seems to require it
and if _REENTRANT is required we should already have it set since one of
the previous tests.
2. Added API-detection for what kind of strerror_r() that is provided. The
POSIX style or the glibc style.
Tor Arntsen provided the necessary feedback these changes are based upon.
Daniel Stenberg [Thu, 25 Mar 2004 07:33:11 +0000 (07:33 +0000)]
Always include setup.h as the first header file.
Added a more verbose comment about what strerror_r() can set errno to in
case of failure.
This file still doesn't build on Solaris due to a missing 'sys_nerr' symbol.
Daniel Stenberg [Wed, 24 Mar 2004 10:52:21 +0000 (10:52 +0000)]
Avoid doing chdir .., as it breaks the ability to use symlinks properly.
chdir to absolute directory names instead. (this flaw exists in the shell
version too)
Daniel Stenberg [Tue, 23 Mar 2004 16:07:02 +0000 (16:07 +0000)]
Greg Hewgill's version of testcurl.sh rewritten in perl for greater
portability. I put it in this directory instead of the root since I think
perhaps it makes more sense.
Daniel Stenberg [Tue, 23 Mar 2004 11:43:34 +0000 (11:43 +0000)]
Fixed the time fields no never get wider than 8 letters. They can now switch
to a "days + hours" or even "just days" display if the time value is very
large. I also switched several calculations over to fixed-point instead of the
previous doubles.