[llvm] [GlobalISel] An optimizing MIR builder (PR #132282)

[Matt Arsenault via llvm-commits]

https://github.com/arsenm commented:

We don't really want to be doing a lot of optimizations in the MIRBuilder. I always found the DAG behavior confusing, and is a major contributor to how slow the DAG is

The CSEMIRBuilder is a compile time hack only. We should only add anything new to MIRBuilders if there is measured evidence for a compile time improvement by duplicating an optimization here, as opposed to only having it in a combiner 

https://github.com/llvm/llvm-project/pull/132282