[PATCH 1/2] dt-bindings: gpio: renesas,rcar-gpio: Document R-Car X5H (R8A78000) support

[PATCH 1/2] dt-bindings: gpio: renesas,rcar-gpio: Document R-Car X5H (R8A78000) support

[Marek Vasut]

Document support for the GPIO controller blocks in the
Renesas R-Car X5H (R8A78000) SoC.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
---
Cc: Bartosz Golaszewski <brgl@kernel.org>
Cc: Conor Dooley <conor+dt@kernel.org>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Krzysztof Kozlowski <krzk+dt@kernel.org>
Cc: Linus Walleij <linusw@kernel.org>
Cc: Rob Herring <robh@kernel.org>
Cc: devicetree@vger.kernel.org
Cc: linux-gpio@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-renesas-soc@vger.kernel.org
---
 .../devicetree/bindings/gpio/renesas,rcar-gpio.yaml          | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/Documentation/devicetree/bindings/gpio/renesas,rcar-gpio.yaml b/Documentation/devicetree/bindings/gpio/renesas,rcar-gpio.yaml
index d32e103a64aac..95f5175c7a9da 100644
--- a/Documentation/devicetree/bindings/gpio/renesas,rcar-gpio.yaml
+++ b/Documentation/devicetree/bindings/gpio/renesas,rcar-gpio.yaml
@@ -56,6 +56,11 @@ properties:
               - renesas,gpio-r8a779h0     # R-Car V4M
           - const: renesas,rcar-gen4-gpio # R-Car Gen4
 
+      - items:
+          - enum:
+              - renesas,gpio-r8a78000     # R-Car X5H
+          - const: renesas,rcar-gen5-gpio # R-Car Gen5
+
   reg:
     maxItems: 1
 
-- 
2.53.0

[PATCH 2/2] gpio: rcar: Add R-Car X5H (R8A78000) support

[Marek Vasut]

R-Car X5H (R8A78000) is the first member of the R-Car Gen5 family.
Add support for R-Car X5H, which has slightly different GPIO block
register layout compared to previous generations. Introduce offset
remap function which performs 1:1 remap for R-Car Gen1..4 and a bit
more complex remap for R-Car Gen5.

The GPIO block register offsets on R-Car Gen5 changed and the change
can be divided into five groups, registers which remained at the
same offset, INDT register shifted by +0x10, OUTDTSEL register
shifted by -0x34, INEN register shifted by -0x38 and the rest of
the registers used by the driver shifted by +0x70 .

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
---
Cc: Bartosz Golaszewski <brgl@kernel.org>
Cc: Conor Dooley <conor+dt@kernel.org>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Krzysztof Kozlowski <krzk+dt@kernel.org>
Cc: Linus Walleij <linusw@kernel.org>
Cc: Rob Herring <robh@kernel.org>
Cc: devicetree@vger.kernel.org
Cc: linux-gpio@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-renesas-soc@vger.kernel.org
---
 drivers/gpio/gpio-rcar.c | 61 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 61 insertions(+)

diff --git a/drivers/gpio/gpio-rcar.c b/drivers/gpio/gpio-rcar.c
index 09bebde5c4260..a22112d9dce0f 100644
--- a/drivers/gpio/gpio-rcar.c
+++ b/drivers/gpio/gpio-rcar.c
@@ -36,6 +36,7 @@ struct gpio_rcar_info {
 	bool has_both_edge_trigger;
 	bool has_always_in;
 	bool has_inen;
+	bool has_layout_gen5;
 };
 
 struct gpio_rcar_priv {
@@ -65,14 +66,59 @@ struct gpio_rcar_priv {
 
 #define RCAR_MAX_GPIO_PER_BANK		32
 
+static inline int gpio_rcar_remap_offset(struct gpio_rcar_priv *p, int *offs)
+{
+	/* R-Car Gen4 and older do not need any offset remap. */
+	if (!p->info.has_layout_gen5)
+		return 0;
+
+	/*
+	 * R-Car Gen5 register layout is slightly different and the offsets
+	 * that have to be added to or subtracted from each register offset
+	 * can be divided into five groups, listed below.
+	 */
+	switch (*offs) {
+	case IOINTSEL...OUTDT:
+		return 0;
+	case INDT:
+		*offs += 0x10;
+		return 0;
+	case INTDT...EDGLEVEL:
+		fallthrough;
+	case BOTHEDGE:
+		*offs += 0x70;
+		return 0;
+	case OUTDTSEL:
+		*offs -= 0x34;
+		return 0;
+	case INEN:
+		*offs -= 0x38;
+		return 0;
+	default:
+		/*
+		 * This here must never be reached, if this is reached, that
+		 * means there is a catastrophic failure in the driver. Skip
+		 * any IO read/write to prevent further damage.
+		 */
+		WARN_ON(1);
+		return -EINVAL;
+	}
+}
+
 static inline u32 gpio_rcar_read(struct gpio_rcar_priv *p, int offs)
 {
+	if (gpio_rcar_remap_offset(p, &offs))
+		return 0;
+
 	return ioread32(p->base + offs);
 }
 
 static inline void gpio_rcar_write(struct gpio_rcar_priv *p, int offs,
 				   u32 value)
 {
+	if (gpio_rcar_remap_offset(p, &offs))
+		return;
+
 	iowrite32(value, p->base + offs);
 }
 
@@ -399,6 +445,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen1 = {
 	.has_both_edge_trigger = false,
 	.has_always_in = false,
 	.has_inen = false,
+	.has_layout_gen5 = false,
 };
 
 static const struct gpio_rcar_info gpio_rcar_info_gen2 = {
@@ -406,6 +453,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen2 = {
 	.has_both_edge_trigger = true,
 	.has_always_in = false,
 	.has_inen = false,
+	.has_layout_gen5 = false,
 };
 
 static const struct gpio_rcar_info gpio_rcar_info_gen3 = {
@@ -413,6 +461,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen3 = {
 	.has_both_edge_trigger = true,
 	.has_always_in = true,
 	.has_inen = false,
+	.has_layout_gen5 = false,
 };
 
 static const struct gpio_rcar_info gpio_rcar_info_gen4 = {
@@ -420,6 +469,15 @@ static const struct gpio_rcar_info gpio_rcar_info_gen4 = {
 	.has_both_edge_trigger = true,
 	.has_always_in = true,
 	.has_inen = true,
+	.has_layout_gen5 = false,
+};
+
+static const struct gpio_rcar_info gpio_rcar_info_gen5 = {
+	.has_outdtsel = true,
+	.has_both_edge_trigger = true,
+	.has_always_in = true,
+	.has_inen = true,
+	.has_layout_gen5 = true,
 };
 
 static const struct of_device_id gpio_rcar_of_table[] = {
@@ -438,6 +496,9 @@ static const struct of_device_id gpio_rcar_of_table[] = {
 	}, {
 		.compatible = "renesas,rcar-gen4-gpio",
 		.data = &gpio_rcar_info_gen4,
+	}, {
+		.compatible = "renesas,rcar-gen5-gpio",
+		.data = &gpio_rcar_info_gen5,
 	}, {
 		.compatible = "renesas,gpio-rcar",
 		.data = &gpio_rcar_info_gen1,
-- 
2.53.0

Re: [PATCH 2/2] gpio: rcar: Add R-Car X5H (R8A78000) support

[Bartosz Golaszewski]

On Sat, 4 Jul 2026 17:13:47 +0200, Marek Vasut
<marek.vasut+renesas@mailbox.org> said:
> R-Car X5H (R8A78000) is the first member of the R-Car Gen5 family.
> Add support for R-Car X5H, which has slightly different GPIO block
> register layout compared to previous generations. Introduce offset
> remap function which performs 1:1 remap for R-Car Gen1..4 and a bit
> more complex remap for R-Car Gen5.
>
> The GPIO block register offsets on R-Car Gen5 changed and the change
> can be divided into five groups, registers which remained at the
> same offset, INDT register shifted by +0x10, OUTDTSEL register
> shifted by -0x34, INEN register shifted by -0x38 and the rest of
> the registers used by the driver shifted by +0x70 .
>
> Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
> ---
> Cc: Bartosz Golaszewski <brgl@kernel.org>
> Cc: Conor Dooley <conor+dt@kernel.org>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Krzysztof Kozlowski <krzk+dt@kernel.org>
> Cc: Linus Walleij <linusw@kernel.org>
> Cc: Rob Herring <robh@kernel.org>
> Cc: devicetree@vger.kernel.org
> Cc: linux-gpio@vger.kernel.org
> Cc: linux-kernel@vger.kernel.org
> Cc: linux-renesas-soc@vger.kernel.org
> ---
>  drivers/gpio/gpio-rcar.c | 61 ++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 61 insertions(+)
>
> diff --git a/drivers/gpio/gpio-rcar.c b/drivers/gpio/gpio-rcar.c
> index 09bebde5c4260..a22112d9dce0f 100644
> --- a/drivers/gpio/gpio-rcar.c
> +++ b/drivers/gpio/gpio-rcar.c
> @@ -36,6 +36,7 @@ struct gpio_rcar_info {
>  	bool has_both_edge_trigger;
>  	bool has_always_in;
>  	bool has_inen;
> +	bool has_layout_gen5;
>  };
>
>  struct gpio_rcar_priv {
> @@ -65,14 +66,59 @@ struct gpio_rcar_priv {
>
>  #define RCAR_MAX_GPIO_PER_BANK		32
>
> +static inline int gpio_rcar_remap_offset(struct gpio_rcar_priv *p, int *offs)
> +{
> +	/* R-Car Gen4 and older do not need any offset remap. */
> +	if (!p->info.has_layout_gen5)
> +		return 0;
> +
> +	/*
> +	 * R-Car Gen5 register layout is slightly different and the offsets
> +	 * that have to be added to or subtracted from each register offset
> +	 * can be divided into five groups, listed below.
> +	 */
> +	switch (*offs) {
> +	case IOINTSEL...OUTDT:
> +		return 0;
> +	case INDT:
> +		*offs += 0x10;
> +		return 0;
> +	case INTDT...EDGLEVEL:
> +		fallthrough;
> +	case BOTHEDGE:
> +		*offs += 0x70;
> +		return 0;
> +	case OUTDTSEL:
> +		*offs -= 0x34;
> +		return 0;
> +	case INEN:
> +		*offs -= 0x38;
> +		return 0;
> +	default:
> +		/*
> +		 * This here must never be reached, if this is reached, that
> +		 * means there is a catastrophic failure in the driver. Skip
> +		 * any IO read/write to prevent further damage.
> +		 */
> +		WARN_ON(1);
> +		return -EINVAL;
> +	}
> +}
> +
>  static inline u32 gpio_rcar_read(struct gpio_rcar_priv *p, int offs)
>  {
> +	if (gpio_rcar_remap_offset(p, &offs))
> +		return 0;
> +
>  	return ioread32(p->base + offs);
>  }
>
>  static inline void gpio_rcar_write(struct gpio_rcar_priv *p, int offs,
>  				   u32 value)
>  {
> +	if (gpio_rcar_remap_offset(p, &offs))
> +		return;
> +
>  	iowrite32(value, p->base + offs);
>  }
>
> @@ -399,6 +445,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen1 = {
>  	.has_both_edge_trigger = false,
>  	.has_always_in = false,
>  	.has_inen = false,
> +	.has_layout_gen5 = false,
>  };
>
>  static const struct gpio_rcar_info gpio_rcar_info_gen2 = {
> @@ -406,6 +453,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen2 = {
>  	.has_both_edge_trigger = true,
>  	.has_always_in = false,
>  	.has_inen = false,
> +	.has_layout_gen5 = false,
>  };
>
>  static const struct gpio_rcar_info gpio_rcar_info_gen3 = {
> @@ -413,6 +461,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen3 = {
>  	.has_both_edge_trigger = true,
>  	.has_always_in = true,
>  	.has_inen = false,
> +	.has_layout_gen5 = false,
>  };
>
>  static const struct gpio_rcar_info gpio_rcar_info_gen4 = {
> @@ -420,6 +469,15 @@ static const struct gpio_rcar_info gpio_rcar_info_gen4 = {
>  	.has_both_edge_trigger = true,
>  	.has_always_in = true,
>  	.has_inen = true,
> +	.has_layout_gen5 = false,

This looks good but do we really need to change these lines if it's zeroes
anyway?

Bart

> +};
> +
> +static const struct gpio_rcar_info gpio_rcar_info_gen5 = {
> +	.has_outdtsel = true,
> +	.has_both_edge_trigger = true,
> +	.has_always_in = true,
> +	.has_inen = true,
> +	.has_layout_gen5 = true,
>  };
>
>  static const struct of_device_id gpio_rcar_of_table[] = {
> @@ -438,6 +496,9 @@ static const struct of_device_id gpio_rcar_of_table[] = {
>  	}, {
>  		.compatible = "renesas,rcar-gen4-gpio",
>  		.data = &gpio_rcar_info_gen4,
> +	}, {
> +		.compatible = "renesas,rcar-gen5-gpio",
> +		.data = &gpio_rcar_info_gen5,
>  	}, {
>  		.compatible = "renesas,gpio-rcar",
>  		.data = &gpio_rcar_info_gen1,
> --
> 2.53.0
>
>

Re: [PATCH 2/2] gpio: rcar: Add R-Car X5H (R8A78000) support

[Marek Vasut]

On 7/6/26 11:19 AM, Bartosz Golaszewski wrote:

Hello Bartosz,

>> +static inline int gpio_rcar_remap_offset(struct gpio_rcar_priv *p, int *offs)
>> +{

I am hoping to get some input on this remap function.

>> +	/* R-Car Gen4 and older do not need any offset remap. */
>> +	if (!p->info.has_layout_gen5)
>> +		return 0;
>> +
>> +	/*
>> +	 * R-Car Gen5 register layout is slightly different and the offsets
>> +	 * that have to be added to or subtracted from each register offset
>> +	 * can be divided into five groups, listed below.
>> +	 */
>> +	switch (*offs) {
>> +	case IOINTSEL...OUTDT:
>> +		return 0;
>> +	case INDT:
>> +		*offs += 0x10;
>> +		return 0;
>> +	case INTDT...EDGLEVEL:
>> +		fallthrough;
>> +	case BOTHEDGE:
>> +		*offs += 0x70;
>> +		return 0;
>> +	case OUTDTSEL:
>> +		*offs -= 0x34;
>> +		return 0;
>> +	case INEN:
>> +		*offs -= 0x38;
>> +		return 0;
>> +	default:
>> +		/*
>> +		 * This here must never be reached, if this is reached, that
>> +		 * means there is a catastrophic failure in the driver. Skip
>> +		 * any IO read/write to prevent further damage.
>> +		 */
>> +		WARN_ON(1);
>> +		return -EINVAL;
>> +	}
>> +}
>> +

[...]

>> @@ -399,6 +445,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen1 = {
>>   	.has_both_edge_trigger = false,
>>   	.has_always_in = false,
>>   	.has_inen = false,
>> +	.has_layout_gen5 = false,
>>   };
>>
>>   static const struct gpio_rcar_info gpio_rcar_info_gen2 = {
>> @@ -406,6 +453,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen2 = {
>>   	.has_both_edge_trigger = true,
>>   	.has_always_in = false,
>>   	.has_inen = false,
>> +	.has_layout_gen5 = false,
>>   };
>>
>>   static const struct gpio_rcar_info gpio_rcar_info_gen3 = {
>> @@ -413,6 +461,7 @@ static const struct gpio_rcar_info gpio_rcar_info_gen3 = {
>>   	.has_both_edge_trigger = true,
>>   	.has_always_in = true,
>>   	.has_inen = false,
>> +	.has_layout_gen5 = false,
>>   };
>>
>>   static const struct gpio_rcar_info gpio_rcar_info_gen4 = {
>> @@ -420,6 +469,15 @@ static const struct gpio_rcar_info gpio_rcar_info_gen4 = {
>>   	.has_both_edge_trigger = true,
>>   	.has_always_in = true,
>>   	.has_inen = true,
>> +	.has_layout_gen5 = false,
> 
> This looks good but do we really need to change these lines if it's zeroes
> anyway?

I'm just following the pattern in the driver, which does full assignment 
of all fields for all generations of the controller. See .has_inen = 
false in Gen1 for example. We don't have to do the assignment, but maybe 
doing so is more comprehensible ?

-- 
Best regards,
Marek Vasut

Re: [PATCH 2/2] gpio: rcar: Add R-Car X5H (R8A78000) support

[Geert Uytterhoeven]

Hi Marek,

Thanks for your patch!

On Mon, 6 Jul 2026 at 21:44, Marek Vasut <marek.vasut@mailbox.org> wrote:
> On 7/6/26 11:19 AM, Bartosz Golaszewski wrote:
> >> +static inline int gpio_rcar_remap_offset(struct gpio_rcar_priv *p, int *offs)
> >> +{
>
> I am hoping to get some input on this remap function.

I haven't looked at your patch in detail yet, but the remap function
was the first thing that struck my eyes.  This might impact performance
of bit-banging and of the sloppy logic analyzer.
Have you looked at the code generated by the compiler?
Perhaps it would be better to use a table, like sci_port_params.regs[]
in the sh-sci driver, and riic_of_data.regs in the riic driver?

> >> +    /* R-Car Gen4 and older do not need any offset remap. */
> >> +    if (!p->info.has_layout_gen5)
> >> +            return 0;
> >> +
> >> +    /*
> >> +     * R-Car Gen5 register layout is slightly different and the offsets
> >> +     * that have to be added to or subtracted from each register offset
> >> +     * can be divided into five groups, listed below.
> >> +     */
> >> +    switch (*offs) {
> >> +    case IOINTSEL...OUTDT:
> >> +            return 0;
> >> +    case INDT:
> >> +            *offs += 0x10;
> >> +            return 0;
> >> +    case INTDT...EDGLEVEL:
> >> +            fallthrough;
> >> +    case BOTHEDGE:
> >> +            *offs += 0x70;
> >> +            return 0;
> >> +    case OUTDTSEL:
> >> +            *offs -= 0x34;
> >> +            return 0;
> >> +    case INEN:
> >> +            *offs -= 0x38;
> >> +            return 0;
> >> +    default:
> >> +            /*
> >> +             * This here must never be reached, if this is reached, that
> >> +             * means there is a catastrophic failure in the driver. Skip
> >> +             * any IO read/write to prevent further damage.
> >> +             */
> >> +            WARN_ON(1);
> >> +            return -EINVAL;
> >> +    }
> >> +}
> >> +

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 2/2] gpio: rcar: Add R-Car X5H (R8A78000) support

[Marek Vasut]

On 7/7/26 8:52 AM, Geert Uytterhoeven wrote:

Hello Geert,

>> On 7/6/26 11:19 AM, Bartosz Golaszewski wrote:
>>>> +static inline int gpio_rcar_remap_offset(struct gpio_rcar_priv *p, int *offs)
>>>> +{
>>
>> I am hoping to get some input on this remap function.
> 
> I haven't looked at your patch in detail yet, but the remap function
> was the first thing that struck my eyes.  This might impact performance
> of bit-banging and of the sloppy logic analyzer.

Regarding the performance impact -- yes, this will impact performance. 
Both the remap function and register table look up will, but at least 
the remap function is small and (subset of it) ends up inlined, the 
indirect look up table is bigger and unlikely to be inlined. That is why 
I opted for this over a table look up.

> Have you looked at the code generated by the compiler?

I did, but there is a lot of inlining going on.

> Perhaps it would be better to use a table, like sci_port_params.regs[]
> in the sh-sci driver, and riic_of_data.regs in the riic driver?

Please see above.

[...]

-- 
Best regards,
Marek Vasut

Re: [PATCH 2/2] gpio: rcar: Add R-Car X5H (R8A78000) support

[Bartosz Golaszewski]

On Mon, 6 Jul 2026 15:06:01 +0200, Marek Vasut <marek.vasut@mailbox.org> said:
> On 7/6/26 11:19 AM, Bartosz Golaszewski wrote:
>
> Hello Bartosz,
>
>>> +static inline int gpio_rcar_remap_offset(struct gpio_rcar_priv *p, int *offs)
>>> +{
>
> I am hoping to get some input on this remap function.
>

You mean its potential impact on performance or the implementation? In any
case, as Geert said: a table of function pointers could help to improve it.
I don't mind it in its current form if it works for you.

Bart

Re: [PATCH 2/2] gpio: rcar: Add R-Car X5H (R8A78000) support

[Geert Uytterhoeven]

Hi Bartosz,

On Tue, 7 Jul 2026 at 15:48, Bartosz Golaszewski <brgl@kernel.org> wrote:
> On Mon, 6 Jul 2026 15:06:01 +0200, Marek Vasut <marek.vasut@mailbox.org> said:
> > On 7/6/26 11:19 AM, Bartosz Golaszewski wrote:
> >>> +static inline int gpio_rcar_remap_offset(struct gpio_rcar_priv *p, int *offs)
> >>> +{
> >
> > I am hoping to get some input on this remap function.
>
> You mean its potential impact on performance or the implementation? In any
> case, as Geert said: a table of function pointers could help to improve it.
> I don't mind it in its current form if it works for you.

A table of register offsets.

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 1/2] dt-bindings: gpio: renesas,rcar-gpio: Document R-Car X5H (R8A78000) support

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 75 bytes --]

Acked-by: Conor Dooley <conor.dooley@microchip.com>
pw-bot: not-applicable

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]