26b5fdd3d0
Returned Tusage_delay to the MS/TP state machines to fix excessive Reply-Posponed messages when received packet is decoupled from immediate handling. Some BACnet routers do not handle Reply-Postponed correctly and drop the subsequent valid reply.
...
Added Slave Node FSM to linux/win32 MS/TP layer.
Added PDU queue to Linux MS/TP datalink layer.
2011-08-24 12:33:20 +00:00
2a0bf4926a
Performed standard indent script indent.sh on project files.
2010-12-15 19:02:40 +00:00
b54218939e
Added extra sub-state for MS/TP from Addendum 135-2008v. Tested with ports/bdk-atxx4-mstp/ code on the BACnet Development Kit. Clean compile on src/mstp.c, ports/at91sam7s and ports/atmega168, but untested. All other changes were not compiled or tested.
2010-10-07 15:52:45 +00:00
69fad0502a
Changed MS/TP idle state: when no-token timeout occurs, clear the receive frame flags in case a message was received and timeout occurred, which may cause MS/TP to stop working in optimized dlmstp loops which don't check for new events when either of those flags are set. Thank you, Jason!
2009-09-17 03:38:20 +00:00
4b422d6c1c
Changed modulo and division operations on 256 to bit shift using include/bytes.h macro to improve code size on architectures that don't do division natively.
2009-08-20 14:38:25 +00:00
dcd6cbed52
Updated Removed the unused state of MSTP_RECEIVE_STATE_HEADER_CRC. Hopefully this doesn't break too many implementations!
2009-07-17 16:02:31 +00:00
d9fea88227
Removed the unused state of MSTP_RECEIVE_STATE_HEADER_CRC. Hopefully this doesn't break too many implementations!
2009-07-17 15:41:37 +00:00
eb7196668c
Corrected the MS/TP TEST REQUEST state such that it now returns the data from the request per BACnet 9.3.5 which says "The data, if present, shall be that which was present in the initiating Test_Request." Thank you to Martin Labbe!
2008-03-09 13:52:48 +00:00
1477603af6
indented.
2007-11-29 16:39:48 +00:00
411d6c1b24
Indented.
2007-11-29 15:56:53 +00:00
35f2aaed00
Corrected race condition where the MS/TP Master Node FSM may miss the first packet since it may not run until after the first packet has arrived if the ReceivedValidFrame flag is set to false during MSTP_MASTER_STATE_INITIALIZE. Thank you to anand!
2007-11-16 13:27:51 +00:00
522d9f4f7d
Corrected poll station for sole master PFM transition.
2007-09-29 14:24:28 +00:00
e53f38dd40
Changed the MSTP_MASTER_STATE_ANSWER_DATA_REQUEST state to always send a REPLY_POSTPONED by
...
commenting out the section that checks for new packets to send. It was not the intent of the BACnet standard, but it makes the Master Node FSM compliant. It's either that or re-parse the message for a matching apdu and invoke id, and then defer if the next message does not match. The problem happens when an unconfirmed request (i.e. WhoIs) is followed immediated by a confirmed request (i.e. ReadProperty) before the device can respond to the unconfirmed request (i.e. the I-Am is queued in the Transmit buffer).
2007-07-14 00:09:16 +00:00
9069c939da
Fixing indent oops which removed the contents of all these files.
2007-06-14 05:33:15 +00:00
ea4863ca25
Indented.
2007-06-14 05:14:26 +00:00
5dd8c6da91
Converted to standard indent style.
2006-10-07 14:13:34 +00:00
dd9b726b5e
corrected state transition in WAIT FOR REPLY state.
2006-09-21 11:17:10 +00:00
310b4c1479
moved the mstp state machines into the ports directories since they will be custom for each port. Some ports will be single threaded while some will use an RTOS.
2006-09-16 17:48:52 +00:00
skarg