Re: [PATCH 2/3] blk-crypto: make blk_crypto_evict_key() more robust

[Eric Biggers <ebiggers@kernel.org>]

On Fri, Mar 03, 2023 at 11:45:00AM -0800, Nathan Huckleberry wrote:
> >  int __blk_crypto_evict_key(struct blk_crypto_profile *profile,
> >                            const struct blk_crypto_key *key)
> > @@ -389,22 +377,22 @@ int __blk_crypto_evict_key(struct blk_crypto_profile *profile,
> >
> >         blk_crypto_hw_enter(profile);
> >         slot = blk_crypto_find_keyslot(profile, key);
> > -       if (!slot)
> > -               goto out_unlock;
> > -
> > -       if (WARN_ON_ONCE(atomic_read(&slot->slot_refs) != 0)) {
> > -               err = -EBUSY;
> > -               goto out_unlock;
> > +       if (slot) {
> > +               if (WARN_ON_ONCE(atomic_read(&slot->slot_refs) != 0)) {
> > +                       /* BUG: key is still in use by I/O */
> > +                       err = -EBUSY;
> > +               } else {
> > +                       err = profile->ll_ops.keyslot_evict(
> > +                                       profile, key,
> > +                                       blk_crypto_keyslot_index(slot));
> > +               }
> > +               /*
> > +                * Callers may free the key even on error, so unlink the key
> > +                * from the hash table and clear slot->key even on error.
> > +                */
> > +               hlist_del(&slot->hash_node);
> > +               slot->key = NULL;
> >         }
> 
> The !slot case still needs to be handled. If profile->num_slots != 0
> and !slot, we'll get an invalid index from blk_crypto_keyslot_index.
> 
> With that change,
> Reviewed-by: Nathan Huckleberry <nhuck@google.com>
> 
> Thanks,
> Huck
> 
> > -       err = profile->ll_ops.keyslot_evict(profile, key,
> > -                                           blk_crypto_keyslot_index(slot));
> > -       if (err)
> > -               goto out_unlock;
> > -
> > -       hlist_del(&slot->hash_node);
> > -       slot->key = NULL;
> > -       err = 0;
> > -out_unlock:
> >         blk_crypto_hw_exit(profile);
> >         return err;
> >  }

I'm not sure what you're referring to.  The !slot case is handled correctly, and
it's the same as before.

- Eric