This is a book review of the book "MLIR Foundations of Next Generation Compilers: Vol 1" by Theo McKenzie.
I think this book attempts to address a very important segment in the current compiler literature. It introduces the bits that make up MLIR and gives a good overview of what MLIR actually is.
It's target audience is largely individuals with a background in compiler design and it brings one up to speed with the internals that make up MLIR. It would be nice if in another edition the authors would get into outlining the thought process that goes into defining a dialect for a particular domain and how the advantages that perhaps defining that dialect at that level of abstraction has to offer.
This will at least give some insight into the utility of certain optimizations at that level of abstraction in contrast with representing the same operations at a lower level.
The brevity of the book is also good as it doesn't bode on unneccessarily lengthening the text for publishing sake.
Compared to "Mastering MLIR" by Oren Davis. It reads like a human wrote it. "Mastering MLIR" is i believe what they call AI slop. There is a sense in which it attempts to address what is lacking in Theo's text but it has managed to be both deep and superficial. This is an incredible feat!
But before i completely damn it it's only right if i give it the proper review it deserves.