Re: [PATCH bpf-next v4 3/3] bpf: Avoid redundant min()/max() in cnum signed bounds

> Does the new precondition match how these functions are actually called?

Yes, it matches, currently no input cnum can be empty in the verifier. The
previous result was also not meaningful for empty cnum's. We left the empty
check for the consideration of performance and to match the old implementation.

> Should the precondition be enforced at the call sites, or should
> smin()/smax() keep returning an ordered pair for EMPTY?

The precondition is enforced by the caller as it was before.