Re: [PATCH 2/2] dm-raid1: don't fail the mirror for invalid I/O errors

On Tue, Jul 07, 2026 at 03:44:54PM -0400, Mike Snitzer wrote:
> So where does dm-raid1 and dm-crypt stand relative to these DIO memory
> alignment changes?  Inferring they are pretty exposed.

dm-raid1 currently has a gap in handling badly behaved direct-io
applications. Worst case it may return success to a failed write command
that used to return an appropriate error before, so a silent corruption.
On reads, it can cause the array to be considered degraded when there's
nothing wrong with it.

There's no errors with dm-crypt today that I know of. I just want to
make it to be more aligned to the backing hardware.