Re: [PATCH bpf-next v2 6/9] selftests: bpf: fix test_kfunc_dynptr_param

[Eduard Zingerman <eddyz87@gmail.com>]

On Wed, 2025-12-31 at 19:39 +0000, Puranjay Mohan wrote:
> On Wed, Dec 31, 2025 at 7:29 PM Eduard Zingerman <eddyz87@gmail.com> wrote:
> > 
> > On Wed, 2025-12-31 at 09:08 -0800, Puranjay Mohan wrote:
> > > As verifier now assumes that all kfuncs only takes trusted pointer
> > > arguments, passing 0 (NULL) to a kfunc that doesn't mark the argument as
> > > __nullable or __opt will be rejected with a failure message of: Possibly
> > > NULL pointer passed to trusted arg<n>
> > > 
> > > Pass a non-null value to the kfunc to test the expected failure mode.
> > > 
> > > Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
> > > ---
> > 
> > Acked-by: Eduard Zingerman <eddyz87@gmail.com>
> > 
> > Unrelated to this patch-set:
> > what do you think about merging __nullable and __opt?
> 
> 
> Yes, I think we should only have __nullable, because that is how
> programmers are used to think about functions, just pass NULL if it is
> unused. But I will see what the verifier expects differently from
> these two and send a patch to merge them.

Ack, thank you.
After cursory examination seems doable to me.
__nullable was introduced [1] after __opt, but mailing list discussion
does not mention __opt at all.

[1] https://lore.kernel.org/all/20231018061746.111364-7-zhouchuyi@bytedance.com/T/#u