janhauer [Thu, 11 Mar 2010 09:42:25 +0000 (09:42 +0000)]
bugfix (by Roman Lim): reference voltage was turned off immediately when RefVolt_XXXV.stop() was called, but there should be a small delay; also, the internal state of the component did not get back to GENERATOR_OFF
klueska [Mon, 15 Feb 2010 03:16:14 +0000 (03:16 +0000)]
Changed the default to be that all platforms #include stdio.h unless specifically set to include something else, and added support for imote2, hopefully...
mmaroti [Sat, 13 Feb 2010 22:50:09 +0000 (22:50 +0000)]
fix uin16_t alignment bug for the msp430
Committed on the Free edition of March Hare Software CVSNT Server.
Upgrade to CVS Suite for more features and support:
http://march-hare.com/cvsnt/
mmaroti [Fri, 12 Feb 2010 21:52:45 +0000 (21:52 +0000)]
Fix the time stamping correction code
Committed on the Free edition of March Hare Software CVSNT Server.
Upgrade to CVS Suite for more features and support:
http://march-hare.com/cvsnt/
janhauer [Thu, 4 Feb 2010 16:31:15 +0000 (16:31 +0000)]
- bugfix: empty data frame sent in response to a datarequest (somtimes) had wrong header size
- updated signature of dummy component NoPromiscuousModeP
gnawali [Thu, 4 Feb 2010 07:31:46 +0000 (07:31 +0000)]
get rid of +5 for rounding, change arg name in functions in CtpInfo to match the arg names in the implementation, fix bug reported by Faisal Aslam - VERY_LARGE_EETX was set to 0xff
andreaskoepke [Wed, 27 Jan 2010 14:42:10 +0000 (14:42 +0000)]
in newer mspgcc/nesc(?) sizeof(message_header_t) > sizeof(message_radio_header_t),
leading to an off by one error. In addition, the message_header_t is aligned on the LSBs.
This fixes all the issues.
andreaskoepke [Sun, 24 Jan 2010 23:01:32 +0000 (23:01 +0000)]
use the same data structure to calculate lengths in transmit and in receive
(removes a bug with newer msp-gcc versions)
check truth in C styls
resource allocation is initialization
razvanm [Sat, 16 Jan 2010 10:02:29 +0000 (10:02 +0000)]
Workaround the erase problems with AT45DB
The previous commit did not take in consideration that STM45P
requires the erase operation. The new strategy is to make the
FlashVolumeManager refuse the erase for AT45DB, and make the
tos-deluge attempt to overwrite of the metadata only when the
erase fails.
ayer1 [Wed, 13 Jan 2010 15:08:56 +0000 (15:08 +0000)]
adjusting wait time for powercycle based upon wider testing of sd card
brands for successful transition between spi and raw data modes; many
fail at 20ms.
janhauer [Tue, 5 Jan 2010 16:41:16 +0000 (16:41 +0000)]
- fixed an issue with the beacon sequence number being set incorrectly
- fixed an issue with indirect transmissions (frame was not allocated properly)
- added some debug info
scipio [Mon, 28 Dec 2009 23:12:13 +0000 (23:12 +0000)]
Fixed three bugs:
1) Peter Bigot @ Rincon noted that the changed flag was incorrectly cleared
and set, such that elapsed time would not be considered in trickle timers.
This caused trickle timers to fire late.
2) The possibility of a subtraction wraparound could cause timers to
have very large values and fire very late.
3) A repetition of subtracting the elapsed time (in both the fired
event and in adjustTime) would cause timers to fire early.
mmaroti [Wed, 23 Dec 2009 04:35:42 +0000 (04:35 +0000)]
revert changes because of RC testing
Committed on the Free edition of March Hare Software CVSNT Server.
Upgrade to CVS Suite for more features and support:
http://march-hare.com/cvsnt/
mmaroti [Mon, 21 Dec 2009 10:12:29 +0000 (10:12 +0000)]
define a NOASSERT macro
Committed on the Free edition of March Hare Software CVSNT Server.
Upgrade to CVS Suite for more features and support:
http://march-hare.com/cvsnt/