From: Niall Leonard <nl250060@ncr.com>

Use the existing shadow data register 'bgpio_data' to allow
the last written value to be returned by the read operation
when BGPIOF_NO_INPUT flag is set.

Signed-off-by: Niall Leonard <nl250060@ncr.com>
---
 drivers/gpio/gpio-mmio.c | 19 +++++++++++++++++--
 1 file changed, 17 insertions(+), 2 deletions(-)

diff --git a/drivers/gpio/gpio-mmio.c b/drivers/gpio/gpio-mmio.c
index d9dff3dc92ae..7125bd8caaa4 100644
--- a/drivers/gpio/gpio-mmio.c
+++ b/drivers/gpio/gpio-mmio.c
@@ -164,6 +164,11 @@ static int bgpio_get_set_multiple(struct gpio_chip *gc, unsigned long *mask,
 	return 0;
 }
 
+static int bgpio_get_shadow(struct gpio_chip *gc, unsigned int gpio)
+{
+	return !!(gc->bgpio_data & bgpio_line2mask(gc, gpio));
+}
+
 static int bgpio_get(struct gpio_chip *gc, unsigned int gpio)
 {
 	return !!(gc->read_reg(gc->reg_dat) & bgpio_line2mask(gc, gpio));
@@ -526,7 +531,10 @@ static int bgpio_setup_io(struct gpio_chip *gc,
 		 * reading each line individually in that fringe case.
 		 */
 	} else {
-		gc->get = bgpio_get;
+		if (flags & BGPIOF_NO_INPUT)
+			gc->get = bgpio_get_shadow;
+		else
+			gc->get = bgpio_get;
 		if (gc->be_bits)
 			gc->get_multiple = bgpio_get_multiple_be;
 		else
@@ -630,7 +638,11 @@ int bgpio_init(struct gpio_chip *gc, struct device *dev,
 	if (ret)
 		return ret;
 
-	gc->bgpio_data = gc->read_reg(gc->reg_dat);
+	if (flags & BGPIOF_NO_INPUT)
+		gc->bgpio_data = 0;
+	else
+		gc->bgpio_data = gc->read_reg(gc->reg_dat);
+
 	if (gc->set == bgpio_set_set &&
 			!(flags & BGPIOF_UNREADABLE_REG_SET))
 		gc->bgpio_data = gc->read_reg(gc->reg_set);
@@ -711,6 +723,9 @@ static struct bgpio_pdata *bgpio_parse_dt(struct platform_device *pdev,
 	if (of_property_read_bool(pdev->dev.of_node, "no-output"))
 		*flags |= BGPIOF_NO_OUTPUT;
 
+	if (of_property_read_bool(pdev->dev.of_node, "no-input"))
+		*flags |= BGPIOF_NO_INPUT;
+
 	return pdata;
 }
 #else

-- 
2.34.1

On 26/01/2023 11:17, Niall Leonard via B4 Submission Endpoint wrote:
> From: Niall Leonard <nl250060@ncr.com>
> 
> Use the existing shadow data register 'bgpio_data' to allow
> the last written value to be returned by the read operation
> when BGPIOF_NO_INPUT flag is set.
> 

(...)

>  	if (gc->set == bgpio_set_set &&
>  			!(flags & BGPIOF_UNREADABLE_REG_SET))
>  		gc->bgpio_data = gc->read_reg(gc->reg_set);
> @@ -711,6 +723,9 @@ static struct bgpio_pdata *bgpio_parse_dt(struct platform_device *pdev,
>  	if (of_property_read_bool(pdev->dev.of_node, "no-output"))
>  		*flags |= BGPIOF_NO_OUTPUT;
>  
> +	if (of_property_read_bool(pdev->dev.of_node, "no-input"))

As pointed, this brings undocumented property to two other bindings.
This needs to be fixed.

Best regards,
Krzysztof

On 29/01/2023 16:01, Krzysztof Kozlowski wrote:
> *External Message* - Use caution before opening links or attachments
> 
> On 26/01/2023 11:17, Niall Leonard via B4 Submission Endpoint wrote:
>> From: Niall Leonard <nl250060@ncr.com>
>>
>> Use the existing shadow data register 'bgpio_data' to allow
>> the last written value to be returned by the read operation
>> when BGPIOF_NO_INPUT flag is set.
>>
> 
> (...)
> 
>>   	if (gc->set == bgpio_set_set &&
>>   			!(flags & BGPIOF_UNREADABLE_REG_SET))
>>   		gc->bgpio_data = gc->read_reg(gc->reg_set);
>> @@ -711,6 +723,9 @@ static struct bgpio_pdata *bgpio_parse_dt(struct platform_device *pdev,
>>   	if (of_property_read_bool(pdev->dev.of_node, "no-output"))
>>   		*flags |= BGPIOF_NO_OUTPUT;
>>   
>> +	if (of_property_read_bool(pdev->dev.of_node, "no-input"))
> 
> As pointed, this brings undocumented property to two other bindings.
> This needs to be fixed.
> 
> Best regards,
> Krzysztof
> 
Thanks for reviewing.
I will update the other 2 bindings in the next version.

Regards,
Niall Leonard

On Thu, Jan 26, 2023 at 11:18 AM Niall Leonard via B4 Submission
Endpoint <devnull+nl250060.ncr.com@kernel.org> wrote:

> From: Niall Leonard <nl250060@ncr.com>
>
> Use the existing shadow data register 'bgpio_data' to allow
> the last written value to be returned by the read operation
> when BGPIOF_NO_INPUT flag is set.
>
> Signed-off-by: Niall Leonard <nl250060@ncr.com>

Weird hardware, but given these restrictions it makes perfect sense
to have this.
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>

Yours,
Linus Walleij