Re: [PATCH v4 bpf-next 1/2] bpf: improve the general precision of tnum_mul

[Eduard Zingerman <eddyz87@xxxxxxxxx>]

On Mon, 2025-08-25 at 09:46 +0530, Nandakumar Edamana wrote:
> Status as of now:
> 
> DECIDED:
> 
> 1. Replace the current outer comment for the new tnum_mul() with a
>     cleaner explanation and the example from the README of the test
>     program.
> 
> 2. (Related to PATCH 2/2) Drop the trivial tests.
> 
> UNDECIDED:
> 
> Instead of just doing tnum_mul(a, b),
> 
> a) whether to do best(tnum_mul(a, b), tnum_mul(b, a))
> b) whether to do best(best(tnum_mul(a, b), tnum_mul(b, a)),
>                        best(tnum_mul_old(a, b), tnum_mul_old(b, a)))

I'd drop both undecided points.