[gergo_barany]

> a decent linear scan allocator which will eventually be replaced with graph coloring for optimized builds.

Before setting out to implement 1980s-style graph coloring, I would suggest considering SSA-based register allocation instead: https://compilers.cs.uni-saarland.de/projects/ssara/ , I find the slides at https://compilers.cs.uni-saarland.de/projects/ssara/hack_ssa... especially useful.

Graph coloring is a nice model for the register assignment problem. But that's a relatively easy part of overall register allocation. If your coloring fails, you need to decide what to spill and how. Graph coloring does not help you with this, you will end up having to iterate coloring and spilling until convergence, and you may spill too much as a result.

But if your program is in SSA, the special properties of SSA can be used to properly separate these subphases, do a single spilling pass first (still not easy!) and then do a coloring that is guaranteed to succeed.

I haven't looked at LLVM in a while, but 10-15 years ago it used to transform out of SSA form just before register allocation. If I had to guess, I would guess it still does so. Not destroying SSA too early could actually be a significant differentiator to LLVM's "cruft".

[gergo_barany]

Also, for a different notion of "cruft", informally it seems to me like new SSA-based compilers tend to choose an SSA representation with basic block arguments instead of the traditional phi instructions. There are probably reasons for this! I'm not aware of a Sea of Nodes compiler with (some notion corresponding to) block arguments, but it might be fun to explore this when designing a new compiler from the ground up. Might be too late for TB, though.