Re: [PATCH v2] block: Fix start and length check added to iov_iter_extract_bvecs()

[Keith Busch]

On Fri, Sep 04, 2026 at 10:42:17AM +0100, David Howells wrote:
> Christoph Hellwig <hch@xxxxxx> wrote:
> 
> > Should this grow a comment explaining why for non-iov/kvec/ubuf
> > we check the alignment of every sector instead of just the
> > start alignment?
> 
> There should possibly be a comment explaining the need for an address
> alignment at all.  Is it mandatory for all devices?  Or are we being a bit
> over restrictive?  I don't think NICs capable of RDMA are so restricted, for
> example.

If a block device behind an RDMA NIC really wants to DMA a single byte
at a time, we can't describe that capability today. That would require a
0 mask, which is currently treated as "unset" and overridden with a
default 512b mask.