Re: [PATCH v2 11/41] mtd: spi-nor: drop .parse_sfdp

[Michael Walle]

> > +	{ "gd25q256", INFO(0xc84019, 0, 64 * 1024, 0)
> >  		FLAGS(SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB | SPI_NOR_TB_SR_BIT6)
> >  		FIXUP_FLAGS(SPI_NOR_4B_OPCODES)
> >  		.fixups = &gd25q256_fixups },
>
> We might get in trouble here if SFDP advertises a wrong flash size. And
> this is because in BP, instead of relying on the info->sector_size, we
> now compute locally the sector size based on the size advertised by
> SFDP. And if the SFDP flash size is wrong, we'll break BP for this
> flash.

I'd say if the SFDP advertises a wrong flash size we are in larger
trouble than a non-working BP :)

> I'm ok taking this risk, but please update the commit message and
> inform readers about what they may hit, and what would be the fix.

sure :)

-michael

> With that feel free to add:
>
> Reviewed-by: Tudor Ambarus <tudor.ambarus@xxxxxxxxxx>