summaryrefslogtreecommitdiff
path: root/cesar/mac/ca
AgeCommit message (Collapse)Author
2013-06-04cesar/cp[av]: set CP_TIMER_OFFSET_STA to allow beacon reception, closes #3990Nélio Laranjeiro
When there is not traffic on the medium, our CCo sends the central beacon 10ms after the beginning of the beacon period according with CA_ACCESS_MERGE_MARGIN_TCK margin.
2013-01-31cesar: check the hardware compatibility when ipmbox is activated, refs #3680Thierry Carré
Information about an assert could be collected by Linux, only after ipmbox activation.
2012-12-13cesar/mac/{ca,pbproc}: handle merged allocations transitions, closes #3500Nicolas Schodet
Previous code was always using old merge flag value in pbproc, or new merge flag value in ca.
2012-11-30cesar/mac/ca: always lock ISR in ca_mfs_update/hold, closes #3453Nicolas Schodet
This is rest of old and slow ISR locking functions.
2012-11-30cesar/mac/ca: inline ca_mfs_update and ca_mfs_hold, refs #3453Nicolas Schodet
There is no tail call optimization on our sparc compiler, reduce call depth.
2012-11-29cesar/hal/arch: add two levels for ilram, closes #3457Nélio Laranjeiro
2012-10-22cesar/mac/ca: add ca_mfs_update* function to ilram2, refs #3455Nélio Laranjeiro
2012-10-19cesar/mac/ca: use predefined values for snid_max and nek_maxMilenko Jelisavcic
2012-10-18cesar/lib/list: rename some functions, closes #3429Yacine Belkadi
To have (hopefully) more intuitive names, rename: list_shift() to list_pop_front() list_unshift() to list_push_front() list_unshift_range() to list_push_front_range() To keep things consistent after the above renaming, rename: list_pop() to list_pop_back() list_push() to list_push_back() list_push_range() to list_push_back_range()
2012-06-11cesar/mac/ca: dump ca schedules on fatal errorNicolas Schodet
2012-05-10cesar/{bsu,hal/phy/spoc,cp}: add missing function stubsNélio Laranjeiro
2011-05-25cesar/mac/pbproc: send two frame bursts, refs #362Nicolas Schodet
2011-05-19cesar/mac/ca: use claim, not assert, in local ram functionNicolas Schodet
2011-05-03cesar/mac/ca: use backoff 0 again, closes #1276Nicolas Schodet
2011-04-18cesar/mac/ca: add _date suffix in trace textNicolas Schodet
2011-04-18cesar/mac/ca: add alloc prepare traceNicolas Schodet
2011-01-19cesar/mac/{ca,pbproc}: remove AIFS between two CSMA allocations, closes #1812Nicolas Schodet
2011-01-19cesar/mac/ca/src: move state/timing check out of compute update, refs #1812Nicolas Schodet
2011-01-19cesar/mac/ca: fix define suffixNicolas Schodet
2011-01-19cesar/mac/ca: detect merge condition, handle B2BIFS better, refs #1812Nicolas Schodet
2010-12-13cesar/mac/ca: PBProc is responsible to force TX after EIFS, closes #2072Nicolas Schodet
CA can not force TX for the frame after EIFS because when PBProc calls vcs_restart, current frame has not been transmitted yet, see #1816.
2010-12-07cesar/mac/ca: do not test beacon periods update during AIFS, refs #2013Nicolas Schodet
2010-12-07cesar/mac/{ca,pbproc}: new CA interface, closes #1910Nicolas Schodet
- split VCS restart and access update in two operations, - do not reserve MFS in advance, add access get MFS / done mechanism.
2010-12-07cesar/mac/{ca,common}: use priority lists instead of heap, refs #1910Nicolas Schodet
2010-12-07cesar/mac/ca: use pointer for current beacon period, refs #1910Nicolas Schodet
2010-12-07cesar/mac/ca: remove beacon periods circular buffer, refs #1910Nicolas Schodet
2010-12-07cesar/mac/{ca,pbproc}: remove anticipation parameter, refs #1910Nicolas Schodet
2010-09-27cesar/mac/ca: remove useless codeNicolas Schodet
2010-09-07cesar/mac/ca: lock ISR while updating beacon periods, refs #1862Nicolas Schodet
2010-09-03cesar/mac/ca: fix bad macroNicolas Schodet
2010-09-03cesar/mac/ca/test/ca: fix test when activatingNicolas Schodet
2010-09-02cesar/mac: add beacon listening allocation, closes #1848Nicolas Schodet
2010-09-02cesar/mac: add beacon LID test, refs #1848Nicolas Schodet
2010-07-06cesar/mac/ca: reduce CA memory usage.laranjeiro
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@7267 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-07-01cesar/mac/ca/test/ca: test workaround, refs #1655 & #1276schodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@7264 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-07-01cesar/mac/ca: fix r7261, refs #1655 & #1276schodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@7263 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-06-29cesar/mac/ca: change backoff workaround, refs #1655 & #1276schodet
The previous workaround was modifying backoff probabilities. git-svn-id: svn+ssh://pessac/svn/cesar/trunk@7261 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-06-17cesar/mac/ca: do not use backoff 0, refs #1655 & #1276schodet
If using backoff 0, TX is programmed during the PRP. PRATIC will be desynchronised. git-svn-id: svn+ssh://pessac/svn/cesar/trunk@7237 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-04-20cesar/mac/ca: fix merge error, see r6925 & r6932schodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6940 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-04-19cesar: remove context parameter from phy_date and phy_sysdate, refs #1415schodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6931 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-04-15cesar/mac/ca: allow date in the past for AIFS after an RX FCschodet
When a FC is received just before an AIFS, the ACCESS timer is canceled. When it is reprogrammed, the date may be in the past because we need an AIFS as soon as possible. Add a test for this case so that Maximus do not byte us. git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6925 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-04-12cesar/mac/{ca,pbproc}: move beacon_period_start_date to alloc, closes #1434schodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6916 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-04-12cesar/mac/ca: stop using bit fieldsschodet
This costs too much code size. git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6915 017c9cb6-072f-447c-8318-d5b54f68fe89
2010-04-12cesar/mac/ca: move hybrid mode determination to allocation, refs #1434schodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6914 017c9cb6-072f-447c-8318-d5b54f68fe89
2009-11-26cesar/hal/arch: use HAL_DISABLE_INTERRUPTS instead of cyg_interrupt_disableschodet
The cyg_interrupt_disable function is really slow. Use the HAL_DISABLE_INTERRUPTS macro instead which is... less slow. Actually, eCos is still using a software trap to manage interruptions but this is not needed in flat mode. git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6475 017c9cb6-072f-447c-8318-d5b54f68fe89
2009-11-17cesar/mac/ca/test/fcall: use generic GPIO instead of leon onesschodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6396 017c9cb6-072f-447c-8318-d5b54f68fe89
2009-10-16cesar/mac/ca: use current date for update date, closes #646schodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6166 017c9cb6-072f-447c-8318-d5b54f68fe89
2009-10-08revert EoC changesschodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@6054 017c9cb6-072f-447c-8318-d5b54f68fe89
2009-10-07cesar/mac/ca[eoc]: port av changesschodet
git-svn-id: svn+ssh://pessac/svn/cesar/trunk@5999 017c9cb6-072f-447c-8318-d5b54f68fe89
2009-10-07cesar/mac/ca: prepare allocation parameters on allocation changeschodet
This has the side effect to include allocation end date in the allocation parameters, to be used by PB Processing. git-svn-id: svn+ssh://pessac/svn/cesar/trunk@5998 017c9cb6-072f-447c-8318-d5b54f68fe89