Now spi_geni_grab_gpi_chan() errors are correctly reported, the
-EPROBE_DEFER error should be returned from probe in case the
GPI dma driver is built as module and/or not probed yet.

Fixes: b59c122484ec ("spi: spi-geni-qcom: Add support for GPI dma")
Fixes: 6532582c353f ("spi: spi-geni-qcom: fix error handling in spi_geni_grab_gpi_chan()")
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
---
 drivers/spi/spi-geni-qcom.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/spi/spi-geni-qcom.c b/drivers/spi/spi-geni-qcom.c
index 206cc04bb1ed..0ebcc5fe92de 100644
--- a/drivers/spi/spi-geni-qcom.c
+++ b/drivers/spi/spi-geni-qcom.c
@@ -661,7 +661,8 @@ static int spi_geni_init(struct spi_geni_master *mas)
 			geni_se_select_mode(se, GENI_GPI_DMA);
 			dev_dbg(mas->dev, "Using GPI DMA mode for SPI\n");
 			break;
-		}
+		} else if (ret == -EPROBE_DEFER)
+			goto out_pm;
 		/*
 		 * in case of failure to get gpi dma channel, we can still do the
 		 * FIFO mode, so fallthrough

---
base-commit: 925294c9aa184801cc0a451b69a18dd0fe7d847d
change-id: 20230615-topic-sm8550-upstream-fix-spi-geni-qcom-probe-9a97cb6b5ea6

Best regards,
-- 
Neil Armstrong <neil.armstrong@linaro.org>

Added the kernel-janitors mailing list to the CC.

On Thu, Jun 15, 2023 at 11:23:17AM +0200, Neil Armstrong wrote:
> Now spi_geni_grab_gpi_chan() errors are correctly reported, the
> -EPROBE_DEFER error should be returned from probe in case the
> GPI dma driver is built as module and/or not probed yet.
> 
> Fixes: b59c122484ec ("spi: spi-geni-qcom: Add support for GPI dma")
> Fixes: 6532582c353f ("spi: spi-geni-qcom: fix error handling in spi_geni_grab_gpi_chan()")
> Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>

In olden days the kernel janitors list used to have a TODO list of
random tasks that a newbie could take on.  People loved the idea of a
TODO list but no one ever maintained it so in practice it was useless
after the first year.  I've create a new KTODO tag so we can
automatically create KTODO lists.  The idea is you write a one line
summary that starts with KTODO and the subsystem prefix like so:

KTODO: static analysis: Ensure that -EPROBE_DEFER is always propogated

Then people can use lei[1] to search for the tag and get the latest
TODO list.  Here is the command to search for all the KTODO items added
in the last six months.

lei q https://lore.kernel.org/all/ -o ~/Mail/KTODO --dedupe=mid 'KTODO AND rt:6.month.ago..'

The KTODO entry should have a short summary.  People can download thread
for more context if they want.  Here is the short summary:

The -EPROBE_DEFER error code is special and needs to propogated to the
callers.  If you have code like:

	err = returns_eprobe_defer();
	if (err)
		return -EINVAL;

Then that is a bug because it's returning -EINVAL instead return err;
Use static analysis to prevent this.

regards,
dan carpenter

[1] https://lpc.events/event/11/contributions/983/attachments/759/1421/Doing%20more%20with%20lore%20and%20b4.pdf

On Thu, Jun 15, 2023 at 06:12:03PM +0300, Dan Carpenter wrote:
> Here is the command to search for all the KTODO items added
> in the last six months.
> 
> lei q https://lore.kernel.org/all/ -o ~/Mail/KTODO --dedupe=mid 'KTODO AND rt:6.month.ago..'
> 

I guess you need the -I option here.  I'm not sure what -I does.  I had
thought it might mean case insensitive search but that's not it.

lei q -I https://lore.kernel.org/all/ -o ~/Mail/KTODO --dedupe=mid 'KTODO AND rt:6.month.ago..'

Then grep ^KTODO ~/Mail/KTODO -R and cat the filename you want.

regards,
dan carpenter

On Thu, Jun 15, 2023 at 11:23:17AM +0200, Neil Armstrong wrote:

>  			dev_dbg(mas->dev, "Using GPI DMA mode for SPI\n");
>  			break;
> -		}
> +		} else if (ret == -EPROBE_DEFER)
> +			goto out_pm;
>  		/*
>  		 * in case of failure to get gpi dma channel, we can still do the

Both sides of an if statement should have braces if one side does.