The IR is designed to be easy to optimize, not easy for a real machine to execute. Among other things, it assumes it has access to an infinite number of registers so that it never needs to (and in fact is not allowed to) write a new value into a previously used register.

[–] [email protected] 1 points 1 week ago (1 children)

Thanks. It sounds like an interesting architecture to look into how the rest is abstracted within the CPU basics like ALU, timers, flags, and interrupts

[–] [email protected] 2 points 4 days ago (1 children)

It's not really an architecture that is intended to map into anything in existing hardware, but having said that, Mill Computing is working on a new extremely unconventional architecture that is a lot closer to this; you can read more about it here, and specifically the design of the register file (which resembles a convener belt) is discussed here.

[–] [email protected] 1 points 4 days ago

I was thinking of stack machines when I asked about LLVM in hardware. It is interesting to see them mentioned here. At the end of the second link to the conveyer belt description, it calls the belt a programming model. So is this actually implemented anywhere in conventional hardware. The belt and the way registers are used makes intuitive sense to me. I do not understand exactly where the ALU sits or how flags and interrupts fit in.

I get rather confused going from the basics of Ben Eater/Melvino's 8-bit processor to pipelines and out of order execution. This makes more sense in my surface understanding so far. Thanks for sharing.

[–] [email protected] 2 points 1 week ago

I don't understand your question.

[–] [email protected] 1 points 1 week ago

There was (is?) an interpreter for llvm code, but code at that level hasn't really gone through optimization, which will be specific to each compile target.