Re: [PATCH v42 06/98] hw/sd/sdcard: Do not store vendor data on block drive (CMD56)

[Philippe Mathieu-Daudé]

On 1/7/24 10:01, Luc Michel wrote:
> On 09:00 Fri 28 Jun     , Philippe Mathieu-Daudé wrote:
> > Caution: This message originated from an External Source. Use proper caution 
> > when opening attachments, clicking links, or responding.
> >
> >
> > "General command" (GEN_CMD, CMD56) is described as:
> >
> >    GEN_CMD is the same as the single block read or write
> >    commands (CMD24 or CMD17). The difference is that [...]
> >    the data block is not a memory payload data but has a
> >    vendor specific format and meaning.
> >
> > Thus this block must not be stored overwriting data block
> > on underlying storage drive. Keep it in a dedicated
> > 'vendor_data[]' array.
> >
> > Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
> > Tested-by: Cédric Le Goater <clg@redhat.com>
> > ---
> > RFC: Is it safe to reuse VMSTATE_UNUSED_V() (which happens
> > to be the same size)?
> >
> > Cc: Peter Xu <peterx@redhat.com>
> > Cc: Fabiano Rosas <farosas@suse.de>
>
> I'm not sure about this migration question.
>
> But IMHO you can simplify your implementation to avoid having to store
> and migrate this vendor_data array. After some research on this command,
> I came to the conclusion that it's used by manufacturers to return
> device health related vendor-specific data. (E.g.,
> https://images-na.ssl-images-amazon.com/images/I/91tTtUMDM3L.pdf Section
> 1.6.1). So I guess you can simply discard writes and return 0s on reads
> (or "QEMU" in ASCII or... :)).

Thanks, very interesting datasheet! Note the argument filter:

  To query the Health Status register, CMD56 with
  argument of [00 00 00 01] is used.

Since we can program this array, I'll simply add it as R/W (KISS).

> > ---
> >   hw/sd/sd.c | 17 +++++++++--------
> >   1 file changed, 9 insertions(+), 8 deletions(-)