Once we merge PR #65800, we’ll rename two operations in Arith. Specifically, we’re changing arith.maxf to arith.maximumf, and arith.minf to arith.minimumf.
The main goal of this PR is to align the naming of Arith operations with the LLVM intrinsics that closely correspond to their intended semantics. This is part of a larger effort outlined in this RFC to fix some issues and naming problems with the MLIR minimum and maximum operations for floating-point numbers.