chiark / gitweb /
ian [Sat, 17 May 2008 18:19:13 +0000 (18:19 +0000)]
do not always use current position when predicting; do do so when reporting initially
ian [Sat, 17 May 2008 17:02:59 +0000 (17:02 +0000)]
bigger margins
ian [Sat, 17 May 2008 16:57:06 +0000 (16:57 +0000)]
set memory limits
ian [Sat, 17 May 2008 16:49:37 +0000 (16:49 +0000)]
rtprio exec properly
ian [Sat, 17 May 2008 16:40:45 +0000 (16:40 +0000)]
use suid helper for sched_setscheduler
ian [Sat, 17 May 2008 15:46:32 +0000 (15:46 +0000)]
set scheduler priority
ian [Sat, 17 May 2008 15:35:41 +0000 (15:35 +0000)]
remove useless test
ian [Sat, 17 May 2008 14:59:24 +0000 (14:59 +0000)]
found implementation of SIGXCPU etc.
ian [Sat, 17 May 2008 14:50:01 +0000 (14:50 +0000)]
support PICs change of X from PING to PONG; also leave various useless stuff out of the simulation log
ian [Sat, 17 May 2008 14:45:46 +0000 (14:45 +0000)]
ping/pong spots loopback
ian [Sat, 17 May 2008 14:21:52 +0000 (14:21 +0000)]
handle watchdog
ian [Sat, 17 May 2008 13:59:44 +0000 (13:59 +0000)]
new watchdog and short circuit arrangements; also add _ to end of p0_booster_userfault_ to indicate inverted polarity
ian [Sat, 17 May 2008 13:59:28 +0000 (13:59 +0000)]
add _ to end of p0_booster_userfault_ to indicate inverted polarity
ian [Sat, 17 May 2008 13:58:37 +0000 (13:58 +0000)]
do not bother messing with cducharging on panic
ian [Sat, 17 May 2008 13:58:06 +0000 (13:58 +0000)]
extend comment re pin[lat]_... macros
ian [Fri, 16 May 2008 19:03:45 +0000 (19:03 +0000)]
rounding errors
ian [Fri, 16 May 2008 18:33:59 +0000 (18:33 +0000)]
fix train not on layout speed request death; hope to fix polarisation; turn off on death
ian [Fri, 16 May 2008 18:17:44 +0000 (18:17 +0000)]
non-autoretrying short circuit cutout
ian [Thu, 15 May 2008 23:00:33 +0000 (23:00 +0000)]
sort out comments, and elapsed time calculation, for new regime
ian [Wed, 14 May 2008 21:27:38 +0000 (21:27 +0000)]
redid safety prediction's handling of uncertainty
ian [Tue, 13 May 2008 18:48:40 +0000 (18:48 +0000)]
on detection, when stopping, do not include an extra lump of uncertainty in the stopping distance
ian [Mon, 12 May 2008 22:56:38 +0000 (22:56 +0000)]
remove obsolete file
ian [Sun, 11 May 2008 23:24:31 +0000 (23:24 +0000)]
layout can do crossover labels as implied by ours.redacted.segjoins.manual but we cannot generate that file atm
ian [Sun, 11 May 2008 22:40:38 +0000 (22:40 +0000)]
manual nmra does not coredump right away but still does not work
ian [Sun, 11 May 2008 22:27:34 +0000 (22:27 +0000)]
get polarity right at the crossover
ian [Sun, 11 May 2008 22:04:32 +0000 (22:04 +0000)]
do not crash if we go out of Run with reserved movements - tell safety we're quitting
ian [Sun, 11 May 2008 22:04:06 +0000 (22:04 +0000)]
do not loop too much in waggle_settle_*
ian [Sun, 11 May 2008 21:45:32 +0000 (21:45 +0000)]
do not wire #2 RD4 twice!
ian [Sun, 11 May 2008 21:20:08 +0000 (21:20 +0000)]
detect duplicate pin use
ian [Sun, 11 May 2008 19:59:45 +0000 (19:59 +0000)]
join X2 and X1 together into X2 with the new relay (waggler)
ian [Sun, 11 May 2008 19:59:13 +0000 (19:59 +0000)]
get waggler encoding right
ian [Sun, 11 May 2008 19:51:04 +0000 (19:51 +0000)]
first part of waggle encode bugfix
ian [Sun, 11 May 2008 19:50:46 +0000 (19:50 +0000)]
fix waggling and some of the messages
ian [Sun, 11 May 2008 19:50:29 +0000 (19:50 +0000)]
fix argument parsing
ian [Sun, 11 May 2008 18:34:24 +0000 (18:34 +0000)]
allow comments on the same line as non-comment
ian [Sun, 11 May 2008 14:30:18 +0000 (14:30 +0000)]
handle NULs more sanely
ian [Sun, 11 May 2008 11:37:41 +0000 (11:37 +0000)]
New manual mode
ian [Sun, 11 May 2008 11:17:19 +0000 (11:17 +0000)]
remove obsolete cdumgr.c
ian [Sun, 11 May 2008 11:15:35 +0000 (11:15 +0000)]
wagglers as a movfeatkind
ian [Sun, 11 May 2008 10:29:19 +0000 (10:29 +0000)]
split points into mfq and points core, so we can have relays
ian [Sun, 11 May 2008 09:16:31 +0000 (09:16 +0000)]
indirect movpos method table
ian [Sat, 10 May 2008 23:27:14 +0000 (23:27 +0000)]
better error message for seg in wiring but not in topology
ian [Sat, 10 May 2008 20:49:20 +0000 (20:49 +0000)]
wagglers in hostside
ian [Sat, 10 May 2008 20:39:35 +0000 (20:39 +0000)]
wagglers work - removed debug
ian [Sat, 10 May 2008 20:38:29 +0000 (20:38 +0000)]
wagglers work - before remove debug
ian [Sat, 10 May 2008 11:43:12 +0000 (11:43 +0000)]
Implement waggle
ian [Sat, 10 May 2008 11:25:23 +0000 (11:25 +0000)]
explicit inversion command
ian [Sat, 10 May 2008 11:24:45 +0000 (11:24 +0000)]
merge branch-deadend-2008-04-13-fiddlesegs after all (cvs up -j branchpoint-deadend-2008-04-13-fiddlesegs -j branch-deadend-2008-04-13-fiddlesegs) - we want to be able to mess with individual movfeat segments after all when we wire with relays. Discard corresponding change to ours.m4 which was erroneous
ian [Sat, 10 May 2008 10:11:07 +0000 (10:11 +0000)]
wip new things
ian [Fri, 9 May 2008 18:49:59 +0000 (18:49 +0000)]
wip new things
ian [Wed, 7 May 2008 23:32:28 +0000 (23:32 +0000)]
New "waggle" object kind for relays - processing from wiring etc.
ian [Tue, 6 May 2008 23:37:56 +0000 (23:37 +0000)]
make points_local_init core code into subroutine, for future wagglers
ian [Tue, 6 May 2008 23:36:31 +0000 (23:36 +0000)]
show TABLAT and TBLPTR*
ian [Mon, 5 May 2008 13:21:20 +0000 (13:21 +0000)]
do not regard a train as satisfying detection at a location if the movposcomb is not known
ian [Mon, 5 May 2008 13:04:11 +0000 (13:04 +0000)]
one bug fixed; other is just consequence of tail-eating
ian [Mon, 5 May 2008 12:58:42 +0000 (12:58 +0000)]
when train is stopping, do not advance to next segment detection first, and remain within stopping distance of next detected segment
ian [Mon, 5 May 2008 11:49:46 +0000 (11:49 +0000)]
cope with X1/X2 opposing interference
ian [Mon, 5 May 2008 11:23:49 +0000 (11:23 +0000)]
really set polarity of interferers during resolution
ian [Mon, 5 May 2008 11:12:57 +0000 (11:12 +0000)]
set polarity of interferers during resolution (involves breaking out resmain_getmovpos)
ian [Mon, 5 May 2008 10:57:27 +0000 (10:57 +0000)]
polarise the interferer too
ian [Mon, 5 May 2008 10:45:15 +0000 (10:45 +0000)]
fix simulation consistency of flush junk
ian [Mon, 5 May 2008 10:42:59 +0000 (10:42 +0000)]
bugfixes to polarity and reporting
ian [Mon, 5 May 2008 10:36:25 +0000 (10:36 +0000)]
extensive trackloc change (calls nextseg early, etc.) and bugfixes
ian [Mon, 5 May 2008 10:36:05 +0000 (10:36 +0000)]
do not duplicate ouhexi messages
ian [Sun, 4 May 2008 23:50:30 +0000 (23:50 +0000)]
sensible mispositioning messages
ian [Sun, 4 May 2008 23:46:56 +0000 (23:46 +0000)]
optimistic point setting not yet tested as current home shinkansen placement does not work; also speed 60 anticlockwise seemed not to work
ian [Sun, 4 May 2008 22:52:11 +0000 (22:52 +0000)]
abandoning accel is not accel
ian [Sun, 4 May 2008 22:38:16 +0000 (22:38 +0000)]
tolerate SignallingHorizonReached in new detection
ian [Sun, 4 May 2008 22:11:43 +0000 (22:11 +0000)]
no really proper tail-eating
ian [Sun, 4 May 2008 21:50:56 +0000 (21:50 +0000)]
proper handling of tail-eating ownership
ian [Sun, 4 May 2008 21:49:59 +0000 (21:49 +0000)]
cope with 0-length messages in oprintf, and do tryflush just before bombing due to buffer overflow
ian [Sun, 4 May 2008 21:27:23 +0000 (21:27 +0000)]
preserve inversions
ian [Sun, 4 May 2008 21:27:14 +0000 (21:27 +0000)]
sane error message when trying to change point under train
ian [Sun, 4 May 2008 21:12:06 +0000 (21:12 +0000)]
do not panic on interference detection
ian [Sun, 4 May 2008 21:06:18 +0000 (21:06 +0000)]
correct argument order for change_needed
ian [Sun, 4 May 2008 21:00:30 +0000 (21:00 +0000)]
seems to be able to polarise
ian [Sun, 4 May 2008 20:35:34 +0000 (20:35 +0000)]
get position right during safety detection-caused stop; much safety debugging output
ian [Sun, 4 May 2008 19:48:48 +0000 (19:48 +0000)]
really prefer existing point position
ian [Sun, 4 May 2008 19:39:39 +0000 (19:39 +0000)]
saner safety error messages
ian [Sun, 4 May 2008 19:32:17 +0000 (19:32 +0000)]
prevent execution of speed and movfeat commands in < Sta_Run
ian [Sun, 4 May 2008 19:21:31 +0000 (19:21 +0000)]
properly set ownership of nose during resolution
ian [Sun, 4 May 2008 18:25:21 +0000 (18:25 +0000)]
suppress boring detections by default
ian [Sun, 4 May 2008 18:25:10 +0000 (18:25 +0000)]
prefer +persist.*.old when simulating
ian [Sun, 4 May 2008 18:24:55 +0000 (18:24 +0000)]
fix uninitialised reference
ian [Sun, 4 May 2008 18:24:41 +0000 (18:24 +0000)]
rename movfeat command to movpos
ian [Sun, 4 May 2008 17:28:37 +0000 (17:28 +0000)]
simulation tested and seems reproduceable
ian [Sun, 4 May 2008 16:26:40 +0000 (16:26 +0000)]
initial simulation tests OK
ian [Sun, 4 May 2008 16:09:31 +0000 (16:09 +0000)]
better filenames
ian [Sun, 4 May 2008 16:03:00 +0000 (16:03 +0000)]
do not install persist during simulation
ian [Sun, 4 May 2008 16:01:24 +0000 (16:01 +0000)]
remove obsolete stuff
ian [Sun, 4 May 2008 16:00:03 +0000 (16:00 +0000)]
log copy feature - not yet tested
ian [Sun, 4 May 2008 15:42:30 +0000 (15:42 +0000)]
simulation - compiles but not yet tested
ian [Thu, 1 May 2008 23:16:54 +0000 (23:16 +0000)]
missing newlines
ian [Thu, 1 May 2008 23:10:26 +0000 (23:10 +0000)]
fixes
ian [Thu, 1 May 2008 23:02:34 +0000 (23:02 +0000)]
correct tail positioning, and detection processing, during resolution; rename predict_confirm to predict
ian [Tue, 29 Apr 2008 18:03:05 +0000 (18:03 +0000)]
fix multi movpos
ian [Sun, 27 Apr 2008 23:33:28 +0000 (23:33 +0000)]
how to fit ours.neato.redacted on a sheet of A4 right now
ian [Sun, 27 Apr 2008 23:31:06 +0000 (23:31 +0000)]
parsing and other bugfixes
ian [Sun, 27 Apr 2008 23:11:33 +0000 (23:11 +0000)]
when toggling booster output pin, use LATC not PORTC to change and check previous value etc.
ian [Sun, 27 Apr 2008 23:07:43 +0000 (23:07 +0000)]
fixed shinkansen resolution problem