Re: [PATCH] block: reuse original bio_vec array for integrity during clone

[Christoph Hellwig <hch@lst.de>]

On Wed, Jul 03, 2024 at 11:08:49AM +0800, Ming Lei wrote:
> > -       bip = bio_integrity_alloc(bio, gfp_mask, bip_src->bip_vcnt);
> > +       bip = bio_integrity_alloc(bio, gfp_mask, 0);
> >         if (IS_ERR(bip))
> >                 return PTR_ERR(bip);
> >
> > -       memcpy(bip->bip_vec, bip_src->bip_vec,
> > -              bip_src->bip_vcnt * sizeof(struct bio_vec));
> > -
> > -       bip->bip_vcnt = bip_src->bip_vcnt;
> > +       bip->bip_vec = bip_src->bip_vec;
> >         bip->bip_iter = bip_src->bip_iter;
> >         bip->bip_flags = bip_src->bip_flags & ~BIP_BLOCK_INTEGRITY;
> 
> I am curious how the patch avoids double free? Given nothing changes
> in bip free code path and source bip_vec is associated with this bip.

bvec_free only frees the bvec array if nr_vecs is > BIO_INLINE_VECS.
bio_integrity_clone now passes 0 as nr_vecs to bio_integrity_alloc
so it won't ever free the bvec array.  This matches what we are
doing for the data bvec array in the bio.