Re: [PATCH bpf-next 3/3] bpf: rename bpf_reg_state->off to bpf_reg_state->delta

[Eduard Zingerman <eddyz87@gmail.com>]

On Wed, 2026-02-11 at 23:08 +0000, bot+bpf-ci@kernel.org wrote:
> > diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
> > --- a/include/linux/bpf_verifier.h
> > +++ b/include/linux/bpf_verifier.h
> > @@ -40,7 +40,7 @@ struct bpf_reg_state {
> > 	/*
> > 	 * Constant delta between "linked" scalars with the same ID.
> > 	 */
> > -	s32 off;
> > +	s32 delta;
> 
> The NFP BPF offload driver still references reg->off in several places:
> 
> drivers/net/ethernet/netronome/nfp/bpf/verifier.c:
>     nfp_bpf_stack_arg_ok() {
>         off = reg->var_off.value + reg->off;
>         ...
>         old_off = old_arg->reg.var_off.value + old_arg->reg.off;
>     }
> 
>     nfp_bpf_check_stack_access() {
>         new_off = reg->off + reg->var_off.value;
>     }
> 
>     nfp_bpf_map_mark_used() {
>         off = reg->var_off.value + meta->insn.off + reg->off;
>     }
> 
> Where reg is const struct bpf_reg_state *. Was this driver intended to be
> updated as part of this rename?

This is a valid point, the driver should be updated.

> Also, the comment block inside struct bpf_reg_state describing BPF_ADD_CONST
> still references the old field name:
> 
> > 	/*
> > 	 * Upper bit of ID is used to remember relationship between "linked"
> > 	 * registers. Example:
> > 	 * r1 = r2;    both will have r1->id == r2->id == N
> > 	 * r1 += 10;   r1->id == N | BPF_ADD_CONST and r1->off == 10
> > 	 * r3 = r2;    both will have r3->id == r2->id == N
> > 	 * w3 += 10;   r3->id == N | BPF_ADD_CONST32 and r3->off == 10
> > 	 */
> 
> Should this comment also use r1->delta and r3->delta to match the new field
> name?

This is a valid point as well.

I'll wait a bit before sending v2 fixing the above issues.