[dm-devel] [PATCH 0/3] fix direct io errors on dm-crypt

[Keith Busch]

On Thu, Nov 10, 2022 at 06:24:03PM +0000, Eric Biggers wrote:
> On Thu, Nov 03, 2022 at 08:25:56AM -0700, Keith Busch wrote:
> > From: Keith Busch <kbusch at kernel.org>
> > 
> > The 6.0 kernel made some changes to the direct io interface to allow
> > offsets in user addresses. This based on the hardware's capabilities
> > reported in the request_queue's dma_alignment attribute.
> > 
> > dm-crypt requires direct io be aligned to the block size. Since it was
> > only ever using the default 511 dma mask, this requirement may fail if
> > formatted to something larger, like 4k, which will result in unexpected
> > behavior with direct-io.
> > 
> > There are two parts to fixing this:
> > 
> >   First, the attribute needs to be moved to the queue_limit so that it
> >   can properly stack with device mappers.
> > 
> >   Second, dm-crypt provides its minimum required limit to match the
> >   logical block size.
> > 
> > Keith Busch (3):
> >   block: make dma_alignment a stacking queue_limit
> >   dm-crypt: provide dma_alignment limit in io_hints
> >   block: make blk_set_default_limits() private
> 
> Hi Keith, can you send out an updated version of this patch series that
> addresses the feedback?
> 
> I'd really like for this bug to be fixed before 6.1 is released, so that there
> isn't a known bug in STATX_DIOALIGN already upon release.

Sorry for the delay, v2 sent.