Граф коммитов

9989 Коммитов

Автор SHA1 Сообщение Дата
Yang Tse 8adc7038fe make comment more precise 2008-02-27 14:54:18 +00:00
Daniel Stenberg b12fef3f31 Michal Marek's cleanup of how curl_easy_setopt() is used in examples and
test code. Thanks to his curl_easy_setopt() typechecker work...
2008-02-27 09:06:15 +00:00
Gunter Knauf 6cc8df95dd added get_ver.awk since c-ares is a standalone project, and should therefore also compile when cURL is absent. 2008-02-27 01:51:40 +00:00
Gunter Knauf f105e23444 a couple of small fixes to the makefile:
fixed comments; fixed INSTDIR define, simplified rules;
changed to use get_ver.awk in current dir rather than the curl one.
2008-02-27 01:43:12 +00:00
Gunter Knauf 7513d29a48 another small change to the makefiles to simplify rules. 2008-02-27 01:36:01 +00:00
Gunter Knauf 97a41f3646 trial to fix the HP-UX breakage... 2008-02-26 23:06:15 +00:00
Gunter Knauf 84de433e62 added curl.html to install package. 2008-02-26 21:42:38 +00:00
Gunter Knauf 724ad15dad some more minor makefile changes; removed useless dist target. 2008-02-26 21:41:19 +00:00
Gunter Knauf 79aa6c841e fixed install target to create a ca-bundle.crt since we have no longer one in the project. 2008-02-26 21:24:03 +00:00
Yang Tse 058e764af8 all reads from stdin and writes to stdout will be retried until the
whole operation completes or an unrecoverable condition is detected
2008-02-26 18:13:59 +00:00
Yang Tse 0d09f342c4 refactor some code out to write_pidfile() in util.c 2008-02-26 15:06:44 +00:00
Gunter Knauf 9682c2037e Added support for server name indication (RFC 4366).
Patch submitted by Kaspar Brand.
2008-02-26 10:30:13 +00:00
Daniel Stenberg 74241e7d85 - Kaspar Brand made GnuTLS-built libcurl properly acknowledge the option that
forces it to prefer SSLv3.
2008-02-25 07:51:39 +00:00
Daniel Stenberg 3154f04fb9 now builds and runs with GnuTLS-built libcurls too 2008-02-23 23:00:24 +00:00
Daniel Stenberg 6982ed4db7 - Sam Listopad provided a patch in feature-request #1900014
http://curl.haxx.se/bug/feature.cgi?id=1900014 that makes libcurl (built to
  use OpenSSL) support a full chain of certificates in a given PKCS12
  certificate.
2008-02-23 12:27:45 +00:00
Daniel Stenberg 9dd3e4d481 - Georg Lippitsch made the src/Makefile.vc6 makefile use the same memory model
options as the lib/Makefile.vc6 already did.
2008-02-22 22:53:01 +00:00
Yang Tse 1d95109ffa Revert sockfilt.c back to revision 1.42
Changes introduced in revision 1.43 were useless
2008-02-22 09:31:18 +00:00
Daniel Stenberg e9bb7b7712 - Zmey Petroff found a crash when libcurl accessed a NULL pointer, which
happened if you set the connection cache size to 1 and for example failed to
  login to an FTP site. Bug report #1896698
  (http://curl.haxx.se/bug/view.cgi?id=1896698)
2008-02-21 17:52:16 +00:00
Gunter Knauf 5e9c564883 fixed missing header; changed bail out from exit() to return().
Mentioned on the list by Michal Marek.
2008-02-21 15:02:14 +00:00
Daniel Stenberg 3bb4602227 assert that the *connp is a non-NULL pointer when Curl_done() is called 2008-02-21 12:28:45 +00:00
Yang Tse 064eebeaf1 Avoid timeout restart when signal caught while awaiting socket and stdin events 2008-02-20 17:17:18 +00:00
Gunter Knauf 4ae644e427 reformatted comment. 2008-02-20 12:36:35 +00:00
Gunter Knauf d208e56b16 added read callback function in order to prevent crashs on Win32 when linked against DLL: 2008-02-20 12:33:45 +00:00
Daniel Stenberg e6170eb20d - Fixed test case 405 to not fail when libcurl is built with GnuTLS 2008-02-20 12:18:08 +00:00
Daniel Stenberg 2c80bcbc81 made the non-matching error code output nicer since we know it is a number
and the string contains a newline...
2008-02-20 12:14:31 +00:00
Gunter Knauf b60dbfa9e9 mention removal of SSLv2 by default. 2008-02-20 11:58:20 +00:00
Daniel Stenberg 9019fc5671 oops, fixed to build 2008-02-20 10:01:28 +00:00
Daniel Stenberg 5db0f70491 "118 - Gautam Kachroo's issue with proxies and ssl" is now in CVS 2008-02-20 09:58:42 +00:00
Daniel Stenberg 53a549000c - Based on initial work done by Gautam Kachroo to address a bug, we now keep
better control at the exact state of the connection's SSL status so that we
  know exactly when it has completed the SSL negotiation or not so that there
  won't be accidental re-uses of connections that are wrongly believed to be
  in SSL-completed-negotiate state.
2008-02-20 09:56:26 +00:00
Daniel Stenberg 55700cb01f - We no longer support setting the CURLOPT_URL option from inside a callback
such as the CURLOPT_SSL_CTX_FUNCTION one treat that as if it was a Location:
  following. The patch that introduced this feature was done for 7.11.0, but
  this code and functionality has been broken since about 7.15.4 (March 2006)
  with the introduction of non-blocking OpenSSL "connects".

  It was a hack to begin with and since it doesn't work and hasn't worked
  correctly for a long time and nobody has even noticed, I consider it a very
  suitable subject for plain removal. And so it was done.
2008-02-20 08:28:02 +00:00
Gunter Knauf f9a6062081 applied patch to disable SSLv2 by default; discussion:
http://sourceforge.net/tracker/index.php?func=detail&aid=1767276&group_id=976&atid=350976
Submitted by Kaspar Brand.
2008-02-19 23:10:07 +00:00
Dan Fandrich 0cae201044 Added test309 to test HTTP redirect to HTTPS URL 2008-02-19 21:57:41 +00:00
Yang Tse 9df37b93df juggle() actually returns bool.
Remove redundant and unreachable log message.
2008-02-19 18:51:08 +00:00
Yang Tse 7b5c86033a fix compiler warnings: 'statement is unreachable' 2008-02-19 17:25:19 +00:00
Gunter Knauf ade0890746 fix for new codestyle. 2008-02-19 16:23:03 +00:00
Gunter Knauf 7a5596bf02 made changes to work with Win32;
replaced fstat() with stat() call and bail out if local file not found.
2008-02-19 16:13:52 +00:00
Yang Tse d2125cf501 fix compiler warnings: 'enumerated type mixed with another type' 2008-02-19 15:07:50 +00:00
Yang Tse c9eb41c056 Reduce to 20 seconds the time allowed to set SO_REUSEADDR option on sockfilt listener socket.
Log some more error descriptions.
2008-02-18 20:13:30 +00:00
Daniel Stenberg 0d722204c3 https_getsock() should be static all over (and did some fixed indenting) 2008-02-18 19:53:31 +00:00
Daniel Stenberg e829d5643f the ca-bundle is now removed 2008-02-18 16:55:27 +00:00
Gunter Knauf 1093287494 added makefile for MingW32 to build most of the samples. 2008-02-18 15:43:23 +00:00
Gunter Knauf 6398f71cc4 moved sample program defines into separate Makefile.inc so that other makefiles can pick up the defines from there. 2008-02-18 15:32:34 +00:00
Gunter Knauf e2b50b203d added check symbol for linking with POSIX prelude. 2008-02-18 15:30:11 +00:00
Yang Tse ab0de23d83 fix compiler warnings:
'enumerated type mixed with another type'

and

  'variable was set but never used'
2008-02-18 13:05:46 +00:00
Daniel Stenberg ec54fbd9ed just mention in --cacert that curl normally has a default ca cert path built-in 2008-02-18 11:40:52 +00:00
Daniel Stenberg 074bd2a19b the ca-bundle is no longer shipped 2008-02-18 11:39:11 +00:00
Daniel Stenberg fb23b85770 - We're no longer providing a very old ca-bundle in the curl tarball. You can
get a fresh one downloaded and created with 'make ca-bundle' or you can get
  one from here => http://curl.haxx.se/docs/caextract.html if you want a fresh
  new one extracted from Mozilla's recent list of ca certs.

  The configure option --with-ca-bundle now lets you specify what file to use
  as default ca bundle for your build. If not specified, the configure script
  will check a few known standard places for a global ca cert to use.
2008-02-18 11:35:12 +00:00
Daniel Stenberg 3458ce9ae5 - Jerome Muffat-Meridol helped me fix Curl_done() to close the current
connection by force when it was called before the entire request is
  completed, simply because we can't know if the connection really can be
  re-used safely at that point.
2008-02-17 13:49:58 +00:00
Daniel Stenberg ba3e7a8656 rephrased comment 2008-02-17 13:49:01 +00:00
Daniel Stenberg 240bae4eb2 In Curl_done() if premature is TRUE, it means this connection was said to be
DONE before the entire request operation is complete and thus we can't know in
what state it is for re-using, so we're forced to close it. In a perfect world
we can add code that keep track of if we really must close it here or not, but
currently we have no such detail knowledge.

Jerome Muffat-Meridol helped us work this out.
2008-02-17 13:43:32 +00:00