[patch] Fix various bugs when using the gnutls async API

[Sjoerd Simons]

Hi,

Our XMPP library (wocky) uses gnutls in a quite heavily async way, which
unfortunately seems to have triggered some bugs in the lastest
versions :/. Some patches for git HEAD attached, which seems to fix
things nicely for me.

Some extra comments for some of the patches:

* 0002-Check-whether-the-error-is-fatal-in-more-cases.patch

  I don't know the code well enough, but i'm not sure why the handshake 
  hash buffers are cleared on non-fatal errors in some cases. I went for
  the more conservative approach here and just change that around when 
  it caused things to break for me...

* 0003-Remember-the-amount-of-user-data-we-re-sending-out.patch
  I guess It could be argued that this information should be stored
  inside the record_send_buffer instead of in the session internal
  struct directly. I'm happy to change the patch around to do just
  that :)

-- 
Sjoerd Simons <address@hidden>
Collabora Ltd.

0001-Add-state-for-flushing-the-handshake-buffer.patch
Description: Text Data

0002-Check-whether-the-error-is-fatal-in-more-cases.patch
Description: Text Data

0003-Remember-the-amount-of-user-data-we-re-sending-out.patch
Description: Text Data