[Home]History of QueuebustersRevisited

LinuxCNCKnowledgeBase | RecentChanges | PageIndex | Preferences | LinuxCNC.org

Revision 30 . . (edit) August 24, 2012 1:02 am by MichaelHaberler
Revision 29 . . August 8, 2012 11:58 am by MichaelHaberler [*add note on expressions]
Revision 28 . . (edit) August 8, 2012 3:06 am by MichaelHaberler
Revision 27 . . (edit) August 8, 2012 2:58 am by MichaelHaberler [*note on preview accuracy]
Revision 26 . . (edit) August 8, 2012 2:52 am by MichaelHaberler [*note how to handle remapped codes wrt taintset and precondition]
Revision 25 . . (edit) August 8, 2012 2:48 am by MichaelHaberler
Revision 24 . . (edit) August 8, 2012 2:45 am by MichaelHaberler [*comment on preview, cache management task]
Revision 23 . . (edit) May 2, 2012 2:19 pm by MichaelHaberler
Revision 22 . . (edit) May 2, 2012 2:17 pm by MichaelHaberler [*fix redo label, too early]
Revision 21 . . (edit) May 2, 2012 2:14 pm by MichaelHaberler
Revision 20 . . (edit) May 2, 2012 2:13 pm by MichaelHaberler
Revision 19 . . (edit) May 2, 2012 1:58 pm by MichaelHaberler
Revision 18 . . (edit) May 2, 2012 1:53 pm by MichaelHaberler
Revision 17 . . (edit) May 2, 2012 1:48 pm by MichaelHaberler
Revision 16 . . (edit) May 2, 2012 1:47 pm by MichaelHaberler
Revision 15 . . (edit) May 2, 2012 1:43 pm by MichaelHaberler
Revision 14 . . (edit) May 2, 2012 1:37 pm by MichaelHaberler
Revision 13 . . (edit) May 2, 2012 1:18 pm by MichaelHaberler
Revision 12 . . (edit) May 2, 2012 1:17 pm by MichaelHaberler
Revision 11 . . (edit) May 2, 2012 1:10 pm by MichaelHaberler [*detail memory and processing needs]
Revision 10 . . (edit) May 1, 2012 11:35 pm by MichaelHaberler [*fix typos]
Revision 9 . . (edit) May 1, 2012 4:54 pm by KimK [fixed some minor typos (Thanks, mhaberler, for all your hard work!)]
Revision 8 . . (edit) May 1, 2012 2:26 pm by MichaelHaberler
Revision 7 . . (edit) May 1, 2012 1:28 pm by MichaelHaberler
Revision 6 . . (edit) May 1, 2012 1:09 pm by MichaelHaberler [*describe Queuebuster cache coherence method]
Revision 5 . . (edit) May 1, 2012 1:07 pm by MichaelHaberler
Revision 4 . . (edit) May 1, 2012 1:06 pm by MichaelHaberler
Revision 3 . . (edit) May 1, 2012 1:03 pm by MichaelHaberler
Revision 2 . . (edit) May 1, 2012 1:02 pm by MichaelHaberler
Revision 1 . . (edit) May 1, 2012 1:00 pm by MichaelHaberler
  

Difference (from prior major revision) (minor diff, author diff)

Changed: 27c27
** a desired property of some state does not yet hold (e.g., probe result)
** a desired property of some state does not yet hold (e.g., probe operation not yet finished, hence probe position still invalid)

Added: 313a314,318

Note added 8-8-2012




It just occurred to me that in the algorithm outline above, I implicitly assumed a strict separation of parse and execute phases for a block; during parse one would determine precondition and taint sets, sync if required, only then execute the block.

This assumed separation in the current rs274ngc code doesnt exist, in particular wrt expressions: expressions and assignments are evaluated at parse time. This implies that either syncing must happen on the fly during reference to stale parameters during expression evaluation, or expression handling is separated into a parse tree generation during read() and parse tree evaluation during execute().

LinuxCNCKnowledgeBase | RecentChanges | PageIndex | Preferences | LinuxCNC.org
Search:
Published under a Creative Commons License