Mason Protter

If you're using a macro for the surface syntax, why do you need `Some` at all? Just make the macro directly construct the `Fix` object.

Hm, it seems that julia is being a bit overly cautious about avoiding us mutating a region of stack memory that belonged to an immutable variable, so a `freeze`/`thaw` is...

Okay, I'll see what I can do. What is your appetite here for a breaking version? I'm picturing replacing what is currently `StaticStrideArray` with a `StrideArray` wrapping a `Ref{

> How would this be different from a `SizedArray`? It wouldn't be backed by a `Vector` storage, but instead a type which can be stack allocated.

One thing to be aware of is that `malloc`/`free` is *way* faster on Linux than MacOS for whatever reason. I've often found that `malloc`/`free` was barely slower than `alloc` (at...

Interestingly, a large part of the overhead can be reduced by using union splitting: ```julia @inline function smallarray3(::Type{T}, ::StaticInt{SpillSize}, size::Integer...) where {T, SpillSize} @assert isbitstype(T) N = prod(size) if N

@Zentrik can I see how you added the `unsafe_alloca` intrinsic?

Oh that's very interesting! I expected this to have the same problem that adding it with `llvmcall` would have, which is that even if the `llvmcall` is inlined, there is...

My usecase is that I want to make a large statically sized vector, then take a small dynamic view of that vector then reshape it to a matrix.

Nice, sounds good