Add global lock around the caching mechanism

[c42f]

Should fix #8

I had some confusing deadlocks trying out ReentrantLock, so I opted to use SpinLock here. Generally that's not advisable, but I think the inside of the locked regions here should be minimal enough that it's safe.

As to the cause of the deadlock, I don't really know what that's about. Perhaps it's causing us to accidentally task switch when running the @generated function body... I could imagine that interacting badly with the compiler somehow.

[codecov]

Codecov Report

Merging #11 into master will increase coverage by 0.92%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master      #11      +/-   ##
==========================================
+ Coverage   91.17%   92.10%   +0.92%     
==========================================
  Files           1        1              
  Lines          34       38       +4     
==========================================
+ Hits           31       35       +4     
  Misses          3        3              

Impacted Files	Coverage Δ
src/RuntimeGeneratedFunctions.jl	92.10% <100.00%> (+0.92%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 77a2637...0fab5cb. Read the comment docs.