Re: [RFC PATCH 1/3] mtd: spi-nor: sfdp: remember sfdp_size

[Michael Walle]

Am 2021-03-16 12:06, schrieb Pratyush Yadav:
> On 16/03/21 12:01PM, Michael Walle wrote:
> > Hi Pratyush,
> >
> > Am 2021-03-16 11:42, schrieb Pratyush Yadav:
> > > On 12/03/21 08:05PM, Michael Walle wrote:
> > > > Save the sftp_size in the spi_nor struct so we can use it to dump the
> > > > SFDP table without parsing the headers again.
> > > >
> > > > Signed-off-by: Michael Walle <michael@xxxxxxxx>
> > > > ---
> > > >  drivers/mtd/spi-nor/sfdp.c  | 12 ++++++++++++
> > > >  include/linux/mtd/spi-nor.h |  1 +
> > > >  2 files changed, 13 insertions(+)
> > > >
> > > > diff --git a/drivers/mtd/spi-nor/sfdp.c b/drivers/mtd/spi-nor/sfdp.c
> > > > index 25142ec4737b..b1814afefc33 100644
> > > > --- a/drivers/mtd/spi-nor/sfdp.c
> > > > +++ b/drivers/mtd/spi-nor/sfdp.c
> > > > @@ -16,6 +16,7 @@
> > > >  	(((p)->parameter_table_pointer[2] << 16) | \
> > > >  	 ((p)->parameter_table_pointer[1] <<  8) | \
> > > >  	 ((p)->parameter_table_pointer[0] <<  0))
> > > > +#define SFDP_PARAM_HEADER_PARAM_LEN(p) ((p)->length * 4)
> > > >
> > > >  #define SFDP_BFPT_ID		0xff00	/* Basic Flash Parameter Table */
> > > >  #define SFDP_SECTOR_MAP_ID	0xff81	/* Sector Map Table */
> > > > @@ -1263,6 +1264,7 @@ int spi_nor_parse_sfdp(struct spi_nor *nor,
> > > >  	struct sfdp_parameter_header *param_headers = NULL;
> > > >  	struct sfdp_header header;
> > > >  	struct device *dev = nor->dev;
> > > > +	size_t param_max_offset;
> > > >  	size_t psize;
> > > >  	int i, err;
> > > >
> > > > @@ -1285,6 +1287,9 @@ int spi_nor_parse_sfdp(struct spi_nor *nor,
> > > >  	    bfpt_header->major != SFDP_JESD216_MAJOR)
> > > >  		return -EINVAL;
> > > >
> > > > +	nor->sfdp_size = SFDP_PARAM_HEADER_PTP(bfpt_header) +
> > > > +			 SFDP_PARAM_HEADER_PARAM_LEN(bfpt_header);
> > > > +
> > > >  	/*
> > > >  	 * Allocate memory then read all parameter headers with a single
> > > >  	 * Read SFDP command. These parameter headers will actually be
> > > > parsed
> > > > @@ -1311,6 +1316,13 @@ int spi_nor_parse_sfdp(struct spi_nor *nor,
> > > >  		}
> > > >  	}
> > > >
> > > > +	for (i = 0; i < header.nph; i++) {
> > > > +		param_header = &param_headers[i];
> > > > +		param_max_offset = SFDP_PARAM_HEADER_PTP(param_header) +
> > > > +				   SFDP_PARAM_HEADER_PARAM_LEN(param_header);
> > > > +		nor->sfdp_size = max(nor->sfdp_size, param_max_offset);
> > > > +	}
> > > > +
> > >
> > > I don't see any mention in the SFDP spec (JESD216D-01) that parameter
> > > tables have to be contiguous.
> >
> > ch. 6.1, fig.10 looks like all the headers come after the initial SFDP
> > header. If that wasn't the case, how would you find the headers then?
> >
> > Also ch. 6.3
> > """All subsequent parameter headers need to be contiguous and may be
> > specified..."""
> >
> > > In fact, it says that "Parameter tables
> > > may be located anywhere in the SFDP space. They do not need to
> > > immediately follow the parameter headers".
> >
> > The tables itself, yes, but not the headers for the tables.
>
> Yes, I was talking about the tables and not the headers. There might be
> holes in the SFDP space but I don't think it would be a problem.

Ah, ok. Well basically I'm assuming that the holes will returning some
sane values. Or you could dump the space together with a mask. But that
is overkill ;)

-michael