RE: [PATCH v3 2/2] mtd: spi-nor: Add support for flash reset

[Potthuri, Sai Krishna]

Hi Tudor Ambarus,

> -----Original Message-----
> From: Tudor.Ambarus@xxxxxxxxxxxxx <Tudor.Ambarus@xxxxxxxxxxxxx>
> Sent: Monday, October 3, 2022 3:13 PM
> To: Potthuri, Sai Krishna <sai.krishna.potthuri@xxxxxxx>;
> pratyush@xxxxxxxxxx; michael@xxxxxxxx; miquel.raynal@xxxxxxxxxxx;
> richard@xxxxxx; vigneshr@xxxxxx; robh+dt@xxxxxxxxxx;
> krzysztof.kozlowski+dt@xxxxxxxxxx
> Cc: devicetree@xxxxxxxxxxxxxxx; linux-mtd@xxxxxxxxxxxxxxxxxxx; linux-
> kernel@xxxxxxxxxxxxxxx; saikrishna12468@xxxxxxxxx; git (AMD-Xilinx)
> <git@xxxxxxx>
> Subject: Re: [PATCH v3 2/2] mtd: spi-nor: Add support for flash reset
> 
> On 9/8/22 09:44, Sai Krishna Potthuri wrote:
> > EXTERNAL EMAIL: Do not click links or open attachments unless you know
> > the content is safe
> >
> > Add support for spi-nor flash reset via GPIO controller by reading the
> > reset-gpio property. If there is a valid GPIO specifier then reset
> > will be performed by asserting and deasserting the GPIO using gpiod
> > APIs otherwise it will not perform any operation.
> >
> > Signed-off-by: Sai Krishna Potthuri <sai.krishna.potthuri@xxxxxxx>
> > Reviewed-by: Michael Walle <michael@xxxxxxxx>
> > ---
> >  drivers/mtd/spi-nor/core.c | 25 +++++++++++++++++++++++++
> >  1 file changed, 25 insertions(+)
> >
> > diff --git a/drivers/mtd/spi-nor/core.c b/drivers/mtd/spi-nor/core.c
> > index f2c64006f8d7..a78ab9bae2be 100644
> > --- a/drivers/mtd/spi-nor/core.c
> > +++ b/drivers/mtd/spi-nor/core.c
> > @@ -2933,6 +2933,27 @@ static void spi_nor_set_mtd_info(struct spi_nor
> *nor)
> >         mtd->_put_device = spi_nor_put_device;  }
> >
> > +static int spi_nor_hw_reset(struct spi_nor *nor) {
> > +       struct gpio_desc *reset;
> > +
> > +       reset = devm_gpiod_get_optional(nor->dev, "reset",
> GPIOD_OUT_LOW);
> > +       if (IS_ERR_OR_NULL(reset))
> > +               return PTR_ERR_OR_ZERO(reset);
> > +
> > +       /*
> > +        * Experimental delay values by looking at different flash device
> > +        * vendors datasheets.
> > +        */
> > +       usleep_range(1, 5);
> > +       gpiod_set_value_cansleep(reset, 1);
> > +       usleep_range(100, 150);
> > +       gpiod_set_value_cansleep(reset, 0);
> > +       usleep_range(1000, 1200);
> > +
> > +       return 0;
> > +}
> > +
> >  int spi_nor_scan(struct spi_nor *nor, const char *name,
> >                  const struct spi_nor_hwcaps *hwcaps)  { @@ -2965,6
> > +2986,10 @@ int spi_nor_scan(struct spi_nor *nor, const char *name,
> >         if (!nor->bouncebuf)
> >                 return -ENOMEM;
> >
> > +       ret = spi_nor_hw_reset(nor);
> 
> Should we condition the calling of this method by the absence of the
> "broken-flash-reset"
> dt prop?
This is the suggestion from Michael on top of my initial series for which
i also agreed to have this call unconditionally. If device tree is having this
reset property then we can do the reset in any case.

Regards
Sai Krishna