[PATCH v3] nvme: fix memory corruption for passthrough metadata

[Kanchan Joshi]

On Fri, Oct 13, 2023 at 10:08 AM Christoph Hellwig <hch at lst.de> wrote:
>
> On Fri, Oct 13, 2023 at 07:49:19AM +0530, Kanchan Joshi wrote:
> > > precedent to start doing it.
> > In my mind, this was about dealing with the specific case when the
> > kernel memory is being used for device DMA.
> > We have just two cases: (i) separate meta buffer, and (ii) bounce
> > buffer for data (+metadata).
> > I had not planned sanity checks for user inputs for anything beyond that.
> > As opposed to being preventive (in all cases), it was about failing
> > only when we are certain that DMA will take place and it will corrupt
> > kernel memory.
> >
> > In the long-term, it may be possible for the path to do away with
> > memory copies. The checks can disappear with that.
>
> As soon as the user buffer is unaligned we need to bounce buffer,
> including for the data buffer.

Yes, but that also sprinkles a bunch of checks and goes against the
theme of  and doing as minimal as possible (at least for passthrough).
Had the plain buffer (potentially unaligned) gone down, either it
would have worked or the device would not like it and user space would
have got the error anyway. No?