[Mlir-commits] [mlir] [mlir][Vector] Add patterns for efficient i4 -> i8 conversion emulation (PR #79494)

[Nicolas Vasilache]

https://github.com/nicolasvasilache approved this pull request.

Makes sense to have simpler patterns for the simpler aligned cases.

Out of curiosity, how hard would it be to have foldings from the existing in MLIR to get to a similar form like you have now?

If you could paste some before / after IT in the comments (or the commit message), this would also be useful.

Thanks for improving this !

https://github.com/llvm/llvm-project/pull/79494