| Message ID | 0330B85B-A297-482B-80A9-E01F236D33D2@martin.sperl.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Sun, Mar 29, 2015 at 04:03:26PM +0200, Martin Sperl wrote: > Could get replaced by something similar to this. > if (!gpio_is_valid(spi->gpio_cs)) { > if (spi->chip_select == 0) { > spi->gpio_cs = 8; > gpio_set_output(spi->gpio_cs); > } > if (spi->chip_select == 1) { > spi->gpio_cs = 7; > gpio_set_output(spi->gpio_cs); > } > } > But I do not know what is the correct call to use. I don't know what you mean by "correct call" here, sorry. If you're talking about working out which GPIO maps to the /CS pin then can the pinctrl API help?
> On 30.03.2015, at 06:22, Mark Brown <broonie@kernel.org> wrote: >> >> But I do not know what is the correct call to use. > > I don't know what you mean by "correct call" here, sorry. If you're > talking about working out which GPIO maps to the /CS pin then can the > pinctrl API help? Well - how do I change the mode of a GPIO to output (which is typically ALT0)? something like this: >> gpio_set_output(spi->gpio_cs); I was hoping someone would have an explicit pointer... Martin -- To unsubscribe from this list: send the line "unsubscribe linux-spi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 03/30/2015 01:20 PM, Martin Sperl wrote: > >> On 30.03.2015, at 06:22, Mark Brown <broonie@kernel.org> wrote: >>> >>> But I do not know what is the correct call to use. >> >> I don't know what you mean by "correct call" here, sorry. If you're >> talking about working out which GPIO maps to the /CS pin then can the >> pinctrl API help? > > Well - how do I change the mode of a GPIO to output (which is typically ALT0)? > something like this: > >>> gpio_set_output(spi->gpio_cs); > > I was hoping someone would have an explicit pointer... Presumably the following DT bindings would work: Documentation/devicetree/bindings/pinctrl/pinctrl-bindings.txt Documentation/devicetree/bindings/pinctrl/brcm,bcm2835-gpio.txt -- To unsubscribe from this list: send the line "unsubscribe linux-spi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 03/29/2015 08:03 AM, Martin Sperl wrote: > Signed-off-by: Martin Sperl <kernel@martin.sperl.org> > Tested-by: Martin Sperl <kernel@martin.sperl.org> It would be useful to have a commit description saying why such a warning was useful. > diff --git a/drivers/spi/spi-bcm2835.c b/drivers/spi/spi-bcm2835.c > @@ -82,6 +82,7 @@ struct bcm2835_spi { > u8 *rx_buf; > int tx_len; > int rx_len; > + bool native_cs_use_warning_done; > }; ... > @@ -294,6 +298,14 @@ static int bcm2835_spi_setup(struct spi_device *spi) ... > + /* we are in the native chipselect case now, > + * so warn about the fact that some things may not work as well What things and why? (I'm asking mainly because that should be in the commit description and perhaps this comment). > + */ > + if (!bs->native_cs_use_warning_done) { > + dev_warn(&spi->dev, > + "setup: native chipselect is used - some driver functions/optimizations are not applied\n"); > + bs->native_cs_use_warning_done = 1; > + } Isn't there a warn_once() that would remove the need to check/set native_cs_use_warning_done? -- To unsubscribe from this list: send the line "unsubscribe linux-spi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/spi/spi-bcm2835.c b/drivers/spi/spi-bcm2835.c index adf157b..601fc5e 100644 --- a/drivers/spi/spi-bcm2835.c +++ b/drivers/spi/spi-bcm2835.c @@ -82,6 +82,7 @@ struct bcm2835_spi { u8 *rx_buf; int tx_len; int rx_len; + bool native_cs_use_warning_done; }; static inline u32 bcm2835_rd(struct bcm2835_spi *bs, unsigned reg) @@ -287,6 +288,9 @@ static void bcm2835_spi_set_cs(struct spi_device *spi, bool gpio_level) static int bcm2835_spi_setup(struct spi_device *spi) { + struct spi_master *master = spi->master; + struct bcm2835_spi *bs = spi_master_get_devdata(master); + /* * sanity checking the native-chipselects */ @@ -294,6 +298,14 @@ static int bcm2835_spi_setup(struct spi_device *spi) return 0; if (gpio_is_valid(spi->cs_gpio)) return 0; + /* we are in the native chipselect case now, + * so warn about the fact that some things may not work as well + */ + if (!bs->native_cs_use_warning_done) { + dev_warn(&spi->dev, + "setup: native chipselect is used - some driver functions/optimizations are not applied\n"); + bs->native_cs_use_warning_done = 1; + } if (spi->chip_select < 3) return 0;