- djm@cvs.openbsd.org 2011/05/08 12:52:01

     [PROTOCOL.mux clientloop.c clientloop.h mux.c]
     improve our behaviour when TTY allocation fails: if we are in
     RequestTTY=auto mode (the default), then do not treat at TTY
     allocation error as fatal but rather just restore the local TTY
     to cooked mode and continue. This is more graceful on devices that
     never allocate TTYs.

     If RequestTTY is set to "yes" or "force", then failure to allocate
     a TTY is fatal.

     ok markus@

1 files changed, 9 insertions, 2 deletions

diff --git a/PROTOCOL.mux b/PROTOCOL.mux
index 3da9e37ae..9ad256602 100644
--- a/PROTOCOL.mux
+++ b/PROTOCOL.mux
@@ -73,6 +73,13 @@ non-multiplexed ssh(1) connection. Two additional cases that the
 client must cope with are it receiving a signal itself and the
 server disconnecting without sending an exit message.
 
+A master may also send a MUX_S_TTY_ALLOC_FAIL before MUX_S_EXIT_MESSAGE
+if remote TTY allocation was unsuccessful. The client may use this to
+return its local tty to "cooked" mode.
+
+        uint32  MUX_S_TTY_ALLOC_FAIL
+        uint32  session id
+
 3. Health checks
 
 The client may request a health check/PID report from a server:
@@ -197,6 +204,7 @@ The MUX_S_PERMISSION_DENIED and MUX_S_FAILURE include a reason:
 #define MUX_S_ALIVE             0x80000005
 #define MUX_S_SESSION_OPENED    0x80000006
 #define MUX_S_REMOTE_PORT       0x80000007
+#define MUX_S_TTY_ALLOC_FAIL    0x80000008
 
 #define MUX_FWD_LOCAL   1
 #define MUX_FWD_REMOTE  2
@@ -208,7 +216,6 @@ XXX lock (maybe)
 XXX watch in/out traffic (pre/post crypto)
 XXX inject packet (what about replies)
 XXX server->client error/warning notifications
-XXX port0 rfwd (need custom response message)
 XXX send signals via mux
 
-$OpenBSD: PROTOCOL.mux,v 1.6 2011/05/06 22:20:10 djm Exp $
+$OpenBSD: PROTOCOL.mux,v 1.7 2011/05/08 12:52:01 djm Exp $