On Wed, May 25, 2022 at 08:08:07AM -0600, Keith Busch wrote:
> On Tue, May 24, 2022 at 08:09:21AM +0200, Christoph Hellwig wrote:
> > > bio_for_each_segment(from, *bio_orig, iter) {
> > > if (i++ < BIO_MAX_VECS)
> > > - sectors += from.bv_len >> 9;
> > > + bytes += from.bv_len;
> > > if (PageHighMem(from.bv_page))
> > > bounce = true;
> > > }
> > > if (!bounce)
> > > return;
> > >
> > > + sectors = ALIGN_DOWN(bytes, queue_logical_block_size(q)) >> 9;
> >
> > Same comment about SECTOR_SHIFT and a comment here. That being said,
> > why do we even align here? Shouldn't bytes always be setor aligned here
> > and this should be a WARN_ON or other sanity check? Probably the same
> > for the previous patch.
>
> Yes, the total bytes should be sector aligned.
>
> I'm not exactly sure why we're counting it this way, though. We could just skip
> the iterative addition and get the total from bio->bi_iter.bi_size. Unless
> bio_orig has more segments that BIO_MAX_VECS, which doesn't seem possible.
Oh, there's a comment explaining the original can have more than BIO_MAX_VECS,
so the ALIGN_DOWN is necessary to ensure a logical block sized split.
