[mlir][Linalg] Fix hoist padding through scf.for iter_arg

Previously, hoisting through an iter_arg would mistakenly yield the unpadded value and
cast it to the padded value.

This was incorrect and resulted in out-of-bounds accesses.
The correct formulation is to yield the padded value and extract a smaller dynamic slice
out of it.

Differential Revision: https://reviews.llvm.org/D148173