A couple years ago I wrote a bit about how our codegen for the lock statement could sometimes lead to situations in which an unoptimized build had different potential deadlocks than an optimized build of the same source code. This is unfortunate, so we've
Read More...
