Victor A. Ying

@chsu1r I didn't know if pushing directly to swarm_backend would disrupt your work. Please let me know if you'd like me to just directly push this there.

+1 It's annoying to have a cluttered Problems pane and significantly inflated count of errors shown in the status bar, which makes it harder for me to tell at a...

@bobbrow wrote: > If we ignore the file, you will still end up with squiggles in your code because the symbols provided by that header won't be defined, right? In...

I agree it's an interesting trade-off. One thing to consider is that we're not talking about sinking memory operations or any operation with side effects: `TryToSinkInstruction` already refuses to move...

Hello, I'm looking to build an execution-driven RISC-V core performance model, to enable studies of multicore systems running interacting software threads, or studies evaluating effects of speculative execution on, for...

Thanks for the pointer to the discussion and the mention of Dramajo which I haven't looked into yet. I'm exploring options right now and might be able to allocate time...

For reference, the commit I made that fixed this in our Swarm compiler is here: https://github.mit.edu/swarm/Swarm-IR/pull/414/commits/3bed595b89c95b3d26ffb4b9d2e3a93978193c94

I'm a bit busy with non-compiler work today. Could you do it?

The extra funny thing is, if we wanted access to a list of reattaching blocks, there already was one computed earlier in the function that could be reused: https://github.com/wsmoses/Tapir-LLVM/blob/1482504e234a65bffc8c54de8de9fc877822345d/lib/Transforms/Tapir/TapirUtils.cpp#L253-L255

If I'm reading the code right, this appears to have been a problem since 285ff4617892da4132f4a0aded992dcc4c5af6d5, which is the commit that introduced `MoveStaticAllocasInClonedBlock()`. In that commit, in `extractDetachBodyToFunction()`, we had the...