summaryrefslogtreecommitdiff
path: root/lib/input_handlers.cpp
AgeCommit message (Collapse)Author
2013-06-20gras: work on goddamn done logicJosh Blum
How blocks mark themselves done has been one of the most annoying things in this development. This done logic is only valuable for QA tests, it doesnt even exist in the practical use case. How it works now: * blocks mark done when sync inputs are done or all inputs are done * removed the force_done input config, its no longer needed * the wait() implementation gives blocks a grace period between an input becoming done and the block itself becoming done. After the grace period, the block is forced done.
2013-06-06gras: moved block ptr into dataJosh Blum
2013-06-06gras: changes for actor in worker Apology commitJosh Blum
2013-06-05gras: compiles now with separate data structJosh Blum
2013-04-14Merge branch 'benchmark_work'Josh Blum
2013-04-14gras: move code into component filesJosh Blum
2013-04-14gras: added more GRAS_UNLIKELY in handlersJosh Blum
2013-04-06gras: cleaner way for msgs to override item reserveJosh Blum
2013-03-31gras: fix #60 by checking done logic after handlerJosh Blum
2013-03-27gras: also release all tags and msgs on block stopJosh Blum
2013-03-18gras: msg port reserve tweak convenienceJosh Blum
2013-03-16gras: added num_read count for pop msg to fix #53Josh Blum
2013-03-11gras: separate queue for messages - #54Josh Blum
2013-03-04stats: sweet pie charts for per block statsJosh Blum
2013-02-25gras: work on new port config APIJosh Blum
combined item size config into port config removed set/get, its just a set by reference, this simplified a lot of client code Commit call will apply the changes at runtime.
2012-12-22use sbuffer config for alloc apiJosh Blum
2012-12-16various fixes for unit tests passJosh Blum
2012-12-15created input and output update per indexJosh Blum
2012-12-15work on the buffer queue apiJosh Blum
2012-12-01input_tags_changed logic to avoid sorting msgsJosh Blum
2012-11-24created common function is_work_allowed and input checkJosh Blum
at least one port needs a tag or an item, this helps the reserve items == 0 case
2012-11-21implement buffer stitching on input, output always flushesJosh Blum
2012-11-18correct comment typo, its fail states notJosh Blum
2012-11-18improvements to done logic (helps reserve 0 case)Josh Blum
2012-11-09switch to boost software licenseJosh Blum
2012-10-28work on stand-alone library buildJosh Blum
2012-10-13simplfied input queues, WIP, but workingJosh Blum
2012-10-13lots of mini tweaks for QA passing, WIP...Josh Blum
2012-10-12working on qa passing, temp input accumulation logicJosh Blum
2012-10-04created any_inputs_done logic for done logicJosh Blum
2012-10-02dammit... find and replace did it againJosh Blum
2012-09-30created a bitset wrapper w/ all() routineJosh Blum
as it turns out ~ was making a new bitset and a malloc just make some common code to avoid the reuse
2012-09-29create IO subscriber bitset for tracking doneJosh Blum
2012-09-29fix for the tokens, were going the wrong directionJosh Blum
2012-09-29ported input and output port handlers to apologyJosh Blum