Fri, 20 Mar 2015 23:36:42 +0800 |
Matt Johnston |
Make main socket nonblocking. Limit writequeue size.
|
Mon, 16 Mar 2015 22:10:30 +0800 |
Matt Johnston |
Fix merge again. ses.loop_handler is required for PAM
pam
|
Mon, 16 Mar 2015 21:34:05 +0800 |
Matt Johnston |
merge pam branch up to date
pam
|
Fri, 23 Jan 2015 22:32:49 +0800 |
Matt Johnston |
resolve merge properly
pam
|
Sun, 01 Mar 2015 23:26:42 +0800 |
Matt Johnston |
reword comment for clarity
nocircbuffer
|
Sat, 28 Feb 2015 23:15:23 +0800 |
Matt Johnston |
Add cleanup
fastopen
|
Sat, 28 Feb 2015 09:06:40 +0800 |
Matt Johnston |
merge from default
fastopen
|
Tue, 24 Feb 2015 22:17:04 +0800 |
Matt Johnston |
Some additional cleanup functions
|
Tue, 24 Feb 2015 22:01:33 +0800 |
Matt Johnston |
Free memory before exiting. Based on patch from Thorsten Horstmann.
|
Tue, 24 Feb 2015 20:51:18 +0800 |
Thorsten Horstmann |
Fix for old compilers, variable declarations at beginning of functions
|
Fri, 20 Feb 2015 23:16:38 +0800 |
Matt Johnston |
Move generic network routines to netio.c
fastopen
|
Fri, 20 Feb 2015 22:13:53 +0800 |
Matt Johnston |
Update priority once the socket is open
fastopen
|
Wed, 18 Feb 2015 22:46:15 +0800 |
Matt Johnston |
async connections working
fastopen
|
Fri, 13 Feb 2015 23:47:53 +0800 |
Matt Johnston |
Add envirnonment variable for debug timestamps to roughly match
|
Fri, 13 Feb 2015 23:17:23 +0800 |
Matt Johnston |
Fix print that no longer works since we're not using fourCCs
|
Fri, 23 Jan 2015 21:38:47 +0800 |
Matt Johnston |
Merge up to date
pam
|
Tue, 19 Aug 2014 23:08:56 +0800 |
Matt Johnston |
Make keepalive handling more robust, this should now match what OpenSSH does
|
Wed, 13 Aug 2014 21:48:47 +0800 |
Matt Johnston |
Don't send SSH_MSG_UNIMPLEMENTED for keepalive responses
|
Wed, 16 Jul 2014 22:53:32 +0800 |
Matt Johnston |
Set tcp priority as follows:
|
Wed, 09 Jul 2014 22:02:22 +0800 |
Matt Johnston |
Fix auth timeout regression
|
Wed, 09 Jul 2014 00:15:20 +0800 |
Matt Johnston |
Make -K keepalive behave like OpenSSH's ServerAliveInterval
|
Sat, 15 Mar 2014 11:37:02 +0800 |
Matt Johnston |
Experiment of always writing data if available. Might waste a writev() with
|
Thu, 13 Mar 2014 23:50:09 +0800 |
Matt Johnston |
Add new monotonic_now() wrapper so that timeouts are unaffected by
|
Thu, 13 Mar 2014 23:08:47 +0800 |
Matt Johnston |
Fix typo
|
Thu, 13 Mar 2014 16:28:16 +0800 |
Yousong Zhou |
Use AUTH_TIMEOUT only before authdone != 1.
|
Wed, 12 Mar 2014 23:40:02 +0800 |
Matt Johnston |
Better PAM through recursion
pam
|
Mon, 17 Feb 2014 21:41:06 +0800 |
Matt Johnston |
Read (and enqueue) packets from interactive input even when
|
Wed, 12 Feb 2014 22:15:02 +0800 |
Matt Johnston |
cleanup before clearing keys
|
Thu, 23 Jan 2014 22:25:52 +0800 |
Matt Johnston |
requirenext doesn't need two values
|
Thu, 14 Nov 2013 22:05:47 +0800 |
Matt Johnston |
rename random.h to dbrandom.h since some OSes have a system random.h
|
Thu, 14 Nov 2013 22:03:30 +0800 |
Matt Johnston |
use oldstyle comments
|
Tue, 21 May 2013 12:09:35 +0800 |
Matt Johnston |
merge in HEAD
ecc
|
Sun, 14 Apr 2013 23:16:16 +0800 |
Matt Johnston |
requirenext fixup for firstkexfollows
|
Sun, 07 Apr 2013 01:36:42 +0800 |
Matt Johnston |
ecc kind of works, needs fixing/testing
ecc
|
Wed, 03 Apr 2013 19:23:53 +0800 |
Matt Johnston |
Just put the version string on the queue, don't use atomicio
|
Mon, 01 Apr 2013 22:26:55 +0800 |
Matt Johnston |
Run the cleanup handler also when we close due to TCP connection being closed
|
Sun, 31 Mar 2013 00:40:00 +0800 |
Matt Johnston |
send out our kexinit packet before blocking to read the SSH version string
|
Tue, 19 Mar 2013 20:15:44 +0800 |
Matt Johnston |
Android returns NULL for pw_crypt, set it to something else
|
Wed, 09 May 2012 22:37:04 +0800 |
Matt Johnston |
Return immediate success for blank passwords if allowed
|
Wed, 23 Feb 2011 15:50:30 +0000 |
Matt Johnston |
Improve capitalisation for all logged strings
|
Tue, 01 Sep 2009 16:38:26 +0000 |
Matt Johnston |
Rearrange getaddrstring() etc
|
Tue, 17 Mar 2009 22:30:25 +0000 |
Matt Johnston |
- Add the signal pipe to maxfd
|
Sun, 01 Mar 2009 16:15:57 +0000 |
Matt Johnston |
- Get rid of decryptreadbuf, just decrypt in-place with readbuf
|
Thu, 26 Feb 2009 13:20:53 +0000 |
Matt Johnston |
disapproval of revision 'a101cbd046507cf723e6362a49196dbd4b924042'
|
Thu, 26 Feb 2009 12:18:11 +0000 |
Matt Johnston |
- Try to write out as much as we can
|
Fri, 07 Nov 2008 14:11:06 +0000 |
Matt Johnston |
Idle timeout patch from Farrell Aultman. Needs testing, unsure if server
|
Mon, 29 Sep 2008 13:53:31 +0000 |
Matt Johnston |
- Add Counter Mode support
|
Mon, 22 Sep 2008 15:28:52 +0000 |
Matt Johnston |
Idle-timeout patch from Farrell Aultman, need to figure whether to only
idle-timeout
|
Mon, 15 Sep 2008 14:40:30 +0000 |
Matt Johnston |
- "-J 'nc localhost 22'" kind of works, needs fixing hostkeys, ptys etc.
|
Mon, 15 Sep 2008 14:04:55 +0000 |
Matt Johnston |
- Add run_shell_command() function to run a "sh -c" command, handling
|
Mon, 15 Sep 2008 12:51:50 +0000 |
Matt Johnston |
- Split main socket var into ses.sock_in/ses.sock_out in preparation
|
Wed, 08 Aug 2007 15:12:06 +0000 |
Matt Johnston |
- Add -K keepalive flag for dropbear and dbclient
|
Fri, 27 Jul 2007 17:13:42 +0000 |
Matt Johnston |
Prevent invalid packets being sent during key-exchange, instead queue
|
Tue, 24 Jul 2007 15:40:23 +0000 |
Matt Johnston |
Rearrange the channel buffer sizes into three neat use-editable values in
|
Fri, 09 Feb 2007 10:43:16 +0000 |
Matt Johnston |
Improve behaviour when flushing out after a process has exited.
channel-fix
|
Tue, 05 Dec 2006 13:27:59 +0000 |
Matt Johnston |
Tidy up behaviour when select() is interrupted. We follow normal
channel-fix
|
Thu, 27 Jul 2006 01:24:39 +0000 |
Matt Johnston |
Just use the normal "remote closed" handler when reading ident stings
|
Sat, 04 Mar 2006 08:22:16 +0000 |
Matt Johnston |
Exit with a message if the ssh protocol version is incompatible
|
Mon, 05 Sep 2005 15:16:52 +0000 |
Matt Johnston |
merge of da30137eaac417f50a323f9ddd29999ec21261e0
|
Mon, 05 Sep 2005 15:16:10 +0000 |
Matt Johnston |
* use own assertions which should get logged properly
|