Compile sync without sync reg

[paili0628]

Fixed the pass compile-sync-without-sync-reg
The original problem was simply that the clear group did not have a done condition, if I ran validate sooner, this could be immediately discovered.
Compiles @sync without use of std_sync_reg
Upon encountering @sync, it first instantiates a std_reg(1) for each thread(bar) and a std_reg(1) for each barrier (s)
It then continuously assigns the value of (s.in) to 1'd1 guarded by the
expression that all values of bar for threads under the barrier are
set to 1'd1
Then it replaces the @sync control operator with

seq {
     barrier;
     clear;
 }

barrier simply sets the value of bar to 1'd1 and then waits for s.out to be up
clear resets the value of bar to 1'd0 for reuse of barrier
Using this method, each thread only incurs 3 cycles of latency overhead for the barrier, and we theoretically won't have a limit for number of threads under one barrier

[rachitnigam]

Seems like this PR still has conflicts with master branch. Can you fix them and let me know afterwards? The diff is currently too big and I'm assuming that's because the branch hasn't been merged with master yet.

[paili0628]

@rachitnigam fixed merging conflict. Please take a look.

[rachitnigam]

Couple of thoughts on reorganizing some code. High-level suggestions:

1. Make BarrierMap a newtype and define methods on it instead of doing the trait trick
2. One of the big pitches for this was to support multiple threads (more than 2), so let's make sure that this PR can actually do that

I think this will make a cool case study of the paper but we need to make sure that the baseline implementation is good enough for a reasonable comparison.

[rachitnigam]

Changes look good! Looks like two tests are failing because #1422 improved the cycle count. Merge at your discretion! Great job! We should make this the default compile-sync pass in the pipeline and deprecate the old one once you're convinced it works for general programs