Boris Brezillon <boris.brezillon@xxxxxxxxxxxxx> wrote on Wed, 27 May 2020 01:03:25 +0200: > On Tue, 26 May 2020 21:56:24 +0200 > Miquel Raynal <miquel.raynal@xxxxxxxxxxx> wrote: > > > Use nand-ecc-provider as the provider DT property. Fallback to > > nand-ecc-mode if the property does not exist. > > > > Signed-off-by: Miquel Raynal <miquel.raynal@xxxxxxxxxxx> > > --- > > drivers/mtd/nand/raw/nand_base.c | 4 +++- > > 1 file changed, 3 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/mtd/nand/raw/nand_base.c b/drivers/mtd/nand/raw/nand_base.c > > index 69c1b7ab938e..7d17d52cdd34 100644 > > --- a/drivers/mtd/nand/raw/nand_base.c > > +++ b/drivers/mtd/nand/raw/nand_base.c > > @@ -5036,7 +5036,9 @@ of_get_nand_ecc_engine_type(struct device_node *np) > > const char *pm; > > int err; > > > > - err = of_property_read_string(np, "nand-ecc-mode", &pm); > > + err = of_property_read_string(np, "nand-ecc-provider", &pm); > > "nand-ecc-engine-type" > > And I'd prefer to have a different string table for that one, so we get > rid of the confusing "hw" type (which is actually "on-controller"). So this means * keeping a legacy table * supporting both tables * with both DT properties * eventually refuse the legacy entries in spi-nand Sounds like a lot of burden to me for a minor improvement. If we rename "hw" into "on-controller" we must also add "off-controller" then, but again, I find it confusing. I am a bit lost on what is the "right" direction right now.
