On Wed, May 01, 2024 at 12:36:11AM -0700, Christoph Hellwig wrote:
> > @@ -377,6 +391,19 @@ int fsverity_read_merkle_tree_block(struct inode *inode,
> >
> > block->pos = pos;
> > block->size = params->block_size;
> > + block->verified = false;
> > +
> > + if (vops->read_merkle_tree_block) {
> > + struct fsverity_readmerkle req = {
> > + .inode = inode,
> > + .ra_bytes = ra_bytes,
> > + };
> > +
> > + err = vops->read_merkle_tree_block(&req, block);
> > + if (err)
> > + goto bad;
> > + return 0;
>
> I still don't understand why we're keeping two interfaces instead of
> providing a read through pagecache helper that implements the
> ->read_block interface. That makes the interface really hard to follow
> and feel rather ad-hoc. I also have vague memories of providing such a
> refactoring a long time ago.
Got a link? This is the first I've heard of this, but TBH I've been
ignoring a /lot/ of things trying to get online repair merged (thank
you!) over the past months...
--D
