On Sat, Aug 10, 2013 at 11:28:57AM -0700, Brian Norris wrote:
> On Wed, Aug 07, 2013 at 09:31:06AM -0300, Ezequiel Garcia wrote:
> > This driver supports NFCv1 (as found in PXA SoC) and NFCv2 (as found in
> > Armada 370/XP SoC). As both controller has a few differences, a way of
> > distinguishing between the two is needed.
> >
> > This commit introduces a new compatible string 'marvell,armada370-nand'
> > and assigns a compatible data of type enum pxa3xx_nand_variant to allow
> > such distinction.
> >
> > Signed-off-by: Ezequiel Garcia <ezequiel.garcia@xxxxxxxxxxxxxxxxxx>
> > ---
> > drivers/mtd/nand/pxa3xx_nand.c | 38 +++++++++++++++++++++++++++++++++++++-
> > 1 file changed, 37 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/mtd/nand/pxa3xx_nand.c b/drivers/mtd/nand/pxa3xx_nand.c
> > index 501e380..2582e1f 100644
> > --- a/drivers/mtd/nand/pxa3xx_nand.c
> > +++ b/drivers/mtd/nand/pxa3xx_nand.c
> > @@ -123,6 +123,11 @@ enum {
> > STATE_READY,
> > };
> >
> > +enum pxa3xx_nand_variant {
> > + PXA3XX_NAND_VARIANT_PXA,
> > + PXA3XX_NAND_VARIANT_ARMADA370,
> > +};
> > +
> > struct pxa3xx_nand_host {
> > struct nand_chip chip;
> > struct pxa3xx_nand_cmdset *cmdset;
> > @@ -171,6 +176,12 @@ struct pxa3xx_nand_info {
> > struct pxa3xx_nand_host *host[NUM_CHIP_SELECT];
> > unsigned int state;
> >
> > + /*
> > + * This driver supports NFCv1 (as found in PXA SoC)
> > + * and NFCv2 (as found in Armada 370/XP SoC).
> > + */
> > + enum pxa3xx_nand_variant variant;
> > +
> > int cs;
> > int use_ecc; /* use HW ECC ? */
> > int use_dma; /* use DMA ? */
> > @@ -1192,7 +1203,14 @@ static int pxa3xx_nand_remove(struct platform_device *pdev)
> >
> > #ifdef CONFIG_OF
> > static struct of_device_id pxa3xx_nand_dt_ids[] = {
> > - { .compatible = "marvell,pxa3xx-nand" },
> > + {
> > + .compatible = "marvell,pxa3xx-nand",
> > + .data = (void *) PXA3XX_NAND_VARIANT_PXA,
>
> It's a little more common to avoid a space between the cast and the
> rvalue. But that's a nitpick that can go either way (it's not in
> CodingStyle, for one).
>
Sure.
> > + },
> > + {
> > + .compatible = "marvell,armada370-nand",
> > + .data = (void *) PXA3XX_NAND_VARIANT_ARMADA370,
> > + },
> > {}
> > };
> > MODULE_DEVICE_TABLE(of, pxa3xx_nand_dt_ids);
> > @@ -1221,11 +1239,28 @@ static int pxa3xx_nand_probe_dt(struct platform_device *pdev)
> >
> > return 0;
> > }
> > +
> > +static enum pxa3xx_nand_variant
> > +pxa3xx_nand_get_variant(struct platform_device *pdev)
> > +{
> > + const struct of_device_id *of_id =
> > + of_match_device(pxa3xx_nand_dt_ids, &pdev->dev);
> > + if (!of_id)
> > + return PXA3XX_NAND_VARIANT_PXA;
> > + return (enum pxa3xx_nand_variant) of_id->data;
> > +}
> > #else
> > static inline int pxa3xx_nand_probe_dt(struct platform_device *pdev)
> > {
> > return 0;
> > }
>
> It looks like before this patch, you don't actually need the whole
> #ifdef/#else CONFIG_OF block. All the of_* helpers have default inline
> implementations that allow things to compile even without CONFIG_OF. So
> without CONFIG_OF, of_match_device() will just return NULL and the
> compiler can easiliy figure out that pxa3xx_nand_probe_dt() always
> should return 0.
>
> IOW, you only need a single pxa3xx_nand_probe_dt() implementation.
>
> And directly related to this patch: you don't need two
> pxa3xx_nand_get_variant() implementations either. Again,
> of_match_device() returns NULL in the !defined(CONFIG_OF) case.
>
> > +
> > +static enum pxa3xx_nand_variant
> > +pxa3xx_nand_get_variant(struct platform_device *pdev)
> > +{
> > + /* Default lefacy (non-DT) variant */
>
> s/lefacy/legacy/
>
> > + return PXA3XX_NAND_VARIANT_PXA;
> > +}
>
> Given the above comments, you won't need this version of
> pxa3xx_nand_get_variant().
>
> > #endif
> >
> > static int pxa3xx_nand_probe(struct platform_device *pdev)
> > @@ -1252,6 +1287,7 @@ static int pxa3xx_nand_probe(struct platform_device *pdev)
> > }
> >
> > info = platform_get_drvdata(pdev);
> > + info->variant = pxa3xx_nand_get_variant(pdev);
> > probe_success = 0;
> > for (cs = 0; cs < pdata->num_cs; cs++) {
> > info->cs = cs;
>
> I would recommend rewriting this patch to remove the #ifdef CONFIG_OF.
>
Good to hear this! It was my intention in the first place, but refrained
from doing so for I thought someone might complain about the CONFIG_OF removal.
So if I remember correct, I'd say this means a patch split (right?):
one patch to remove CONFIG_OF, one patch to introduce the SoC variant.
Thanks for reviewing!
--
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
