[Mlir-commits] [mlir] Add arith expansion of f8E8M0 type for extf/trunc ops (PR #140332)

[Benoit Jacob]

bjacob wrote:

As discussed on the IREE issue, I believe that there is a longstanding tradition that if a conversion returns a finite value then it should be the nearest representable value, that is corroborated by table 3 in the OCP spec discussing the "overflow or saturate" semantics, that is broken by taking the absolute value.

Concretely here, if the input value is the result of a calculation that accidentally yielded a tiny negative value, clamping to the nearest representable value seems more useful than bouncing back towards larger magnitudes.

https://github.com/llvm/llvm-project/pull/140332