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.

  curl that uses the new CURLOPT_FTP_SSL_CCC option in libcurl. If enabled, it
  will make libcurl shutdown SSL/TLS after the authentication is done on a
  FTP-SSL operation.

during certain conditions when GnuTLS is used.

They all now return ssize_t to Curl_write().

Unfortunately, Curl_read() is in a sorrier state but it too would benefit from
a similar cleanup.

reading the (local) CA cert file to let users easier pinpoint the actual
problem. CURLE_SSL_CACERT_BADFILE (77) is the new libcurl error code.

(http://curl.haxx.se/bug/view.cgi?id=1480821) He found and identified a
problem with how libcurl dealt with GnuTLS and a case where gnutls returned
GNUTLS_E_AGAIN indicating it would block. It would then return an unexpected
return code, making Curl_ssl_send() confuse the upper layer - causing random
28 bytes trash data to get inserted in the transfered stream.

The proper fix was to make the Curl_gtls_send() function return the proper
return codes that the callers would expect. The Curl_ossl_send() function
already did this.

client certificates! (http://curl.haxx.se/bug/view.cgi?id=1348930).

  #1334338 (http://curl.haxx.se/bug/view.cgi?id=1334338). When reading an SSL
  stream from a server and the server requests a "rehandshake", the current
  code simply returns this as an error. I have no good way to test this, but
  I've added a crude attempt of dealing with this situation slightly better -
  it makes a blocking handshake if this happens. Done like this because fixing
  this the "proper" way (that would handshake asynchronously) will require
  quite some work and I really need a good way to test this to do such a
  change.

previously failed due to GnuTLS not allowing x509 v1 CA certs by default.

Archived thread of the help-gnutls mailing list regarding this problem:

http://lists.gnu.org/archive/html/help-gnutls/2005-04/msg00000.html

(and I _am_ sorry for my confused behaviour on this problem.)

bug I thought I saw before when I changed this...!

(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.

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