Optimization for SYCL out of order queue

[winstonzhang-intel]

have to iron out a few more things, but the initial concept is this: for regular commandlist, we want to give the events device scope thus avoiding L3 flush. To achieve this, I have a list of all events with device scope being kept track on within the queue. When we get to the last one, we want that commandlist's event to have host scope and also to wait for all other device scoped events. Note to self: need to find out how to get total number of commandlists within a queue

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Comparison is base (51d7180) 15.46% compared to head (edf99c5) 15.36%. Report is 291 commits behind head on main.

:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1304      +/-   ##
==========================================
- Coverage   15.46%   15.36%   -0.10%     
==========================================
  Files         238      240       +2     
  Lines       33883    34182     +299     
  Branches     3747     3788      +41     
==========================================
+ Hits         5239     5252      +13     
- Misses      28593    28879     +286     
  Partials       51       51              

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.