Re: [PATCH bpf-next 2/4] bpf: track find_equal_scalars history on per-instruction level

[Andrii Nakryiko <andrii.nakryiko@gmail.com>]

On Wed, Feb 28, 2024 at 1:16 PM Eduard Zingerman <eddyz87@gmail.com> wrote:
>
> On Wed, 2024-02-28 at 11:58 -0800, Andrii Nakryiko wrote:
> [...]
>
> > > diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
> > > index cbfb235984c8..26e32555711c 100644
> > > --- a/include/linux/bpf_verifier.h
> > > +++ b/include/linux/bpf_verifier.h
> > > @@ -361,6 +361,7 @@ struct bpf_jmp_history_entry {
> > >         u32 prev_idx : 22;
> > >         /* special flags, e.g., whether insn is doing register stack spill/load */
> > >         u32 flags : 10;
> > > +       u64 equal_scalars;
> >
> > nit: should we call this concept as a bit more generic "linked
> > registers" instead of "equal scalars"?
>
> It's a historical name for the feature and it is present in a few commit and tests.
> Agree that "linked_registers" is better in current context.
> A bit reluctant but can change it here.

I'd start with calling this specific field either "linked_regs" or
"linked_set". It's a superset of "equal scalars", so we don't strictly
need to rename all the existing mentions of "equal_scalars" in
existing code.

>
> [...]
>

[...]