[RESEND v12 0/3] mmc: mediatek: add support for SDIO async IRQ

[RESEND v12 0/3] mmc: mediatek: add support for SDIO async IRQ

[Axe Yang]

Changes in v12:
- assign NULL to pins_eint directly instead of using kfree()

Changes in v11:
- remove '_irq' suffix in interrupts-names property
- fix yaml example build error
- refactor msdc_enable_sdio_irq(), free pins_eint if async irq is not supported

Changes in v10:
- add sample node for SDIO host which support wakeup interrupt in yaml
- skip MMC_PM_WAKE_SDIO_IRQ check before enable SDIO async interrupt
- add MMC_PM_KEEP_POWER check before SDIO eint pinstate parsing
- use dev_pm_set_dedicated_wake_irq_reverse() to correct irq control sequence
- set dedicated irq in msdc_enable_sdio_irq() rather than msdc_drv_probe()
- remove unnecessary wake irq control, rpm/dpm system shall manage that
- move wake irq/msdc irq control back to system suspend phase, use rpm_suspend
  and rpm_resume to ensure irq control sequence:
     disable msdc irq -> enable wake irq -> disable wake irq -> enable msdc irq
- simplify variables, check pins_eint to know whether wakeup settings are managed

Changes in v9:
- remove pinctrl "state_dat1"

Changes in v8:
- remove maxItems property under pinctrl-names property

Changes in v7:
- add device_init_wakeup() to register SDIO host as wakeup source

Changes in v6:
- abandon cap-sdio-async-irq flag, use wakeup-source flag instead
- extend interrupts and pinctrls in mediatek mmc host controller DT documents
- add mmc_card_enable_async_irq() to access enable_async_irq flag
- simplify wakeup irq implementation with dedicate wake up irq related interface

Changes in v5:
- resort variables to reversed xmas tree order
- restore old copyright year range and add current year back

Changes in v4:
- add MMC_CAP2_SDIO_ASYNC_IRQ judge before lookup eint pinctrl
- replace spin_lock_irqsave() variant with spin_lock() in eint irq handler

Changes in v3:
- correct abbreviations with capital letters in commit message
- replace copyright year with 2022 in mtk-sd.c
- remove unnessary pointer casting
- adjust variable order to reversed xmas tree
- remove a redundant blank line
- refine if statement, following standard pattern

Changes in v2:
- change flag name from 'cap-sdio-async-int' to 'cap-sdio-async-irq'
- change corresponding macro names from xxx_INT to xxx_IRQ
- resort new member in msdc_host structure
- refine function msdc_request_dat1_eint_irq()
- rename msdc_{suspend,resume} function names, add suffix '_noirq'
- add MMC_CAP2_NO_SDIO judgement before parse eint related pin setting

Axe Yang (3):
  dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties
  mmc: core: Add support for SDIO wakeup interrupt
  mmc: mediatek: add support for SDIO eint wakup IRQ

 .../devicetree/bindings/mmc/mtk-sd.yaml       | 50 +++++++++++-
 drivers/mmc/core/sdio.c                       | 14 ++++
 drivers/mmc/host/mtk-sd.c                     | 80 +++++++++++++++++--
 include/linux/mmc/card.h                      |  8 +-
 include/linux/mmc/sdio.h                      |  5 ++
 5 files changed, 149 insertions(+), 8 deletions(-)

-- 
2.25.1

[RESEND v12 1/3] dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties

[Axe Yang]

Extend interrupts and pinctrls for SDIO wakeup interrupt feature.
This feature allow SDIO devices alarm asynchronous interrupt to host
even when host stop providing clock to SDIO card. An extra wakeup
interrupt and pinctrl states for SDIO DAT1 pin state switching are
required in this scenario.

Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Axe Yang <axe.yang@mediatek.com>
---
 .../devicetree/bindings/mmc/mtk-sd.yaml       | 50 ++++++++++++++++++-
 1 file changed, 49 insertions(+), 1 deletion(-)

diff --git a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
index 2a2e9fa8c188..e83bf10281d6 100644
--- a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
+++ b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
@@ -72,12 +72,27 @@ properties:
       - const: ahb_cg
 
   interrupts:
-    maxItems: 1
+    description:
+      Should at least contain MSDC GIC interrupt. To support SDIO in-band wakeup, an extended
+      interrupt is required and be configured as wakeup source irq.
+    minItems: 1
+    maxItems: 2
+
+  interrupt-names:
+    items:
+      - const: msdc
+      - const: sdio_wakeup
 
   pinctrl-names:
+    description:
+      Should at least contain default and state_uhs. To support SDIO in-band wakeup, dat1 pin
+      will be switched between GPIO mode and SDIO DAT1 mode, state_eint and state_dat1 are
+      mandatory in this scenarios.
+    minItems: 2
     items:
       - const: default
       - const: state_uhs
+      - const: state_eint
 
   pinctrl-0:
     description:
@@ -89,6 +104,11 @@ properties:
       should contain uhs mode pin ctrl.
     maxItems: 1
 
+  pinctrl-2:
+    description:
+      should switch dat1 pin to GPIO mode.
+    maxItems: 1
+
   assigned-clocks:
     description:
       PLL of the source clock.
@@ -208,4 +228,32 @@ examples:
         mediatek,hs400-cmd-resp-sel-rising;
     };
 
+    mmc3: mmc@11260000 {
+        compatible = "mediatek,mt8173-mmc";
+        reg = <0x11260000 0x1000>;
+        clock-names = "source", "hclk";
+        clocks = <&pericfg CLK_PERI_MSDC30_3>,
+                 <&topckgen CLK_TOP_MSDC50_2_H_SEL>;
+        interrupt-names = "msdc", "sdio_wakeup";
+        interrupts-extended = <&gic GIC_SPI 74 IRQ_TYPE_LEVEL_LOW 0>,
+                     <&pio 23 IRQ_TYPE_LEVEL_LOW>;
+        pinctrl-names = "default", "state_uhs", "state_eint";
+        pinctrl-0 = <&mmc2_pins_default>;
+        pinctrl-1 = <&mmc2_pins_uhs>;
+        pinctrl-2 = <&mmc2_pins_eint>;
+        bus-width = <4>;
+        max-frequency = <200000000>;
+        cap-sd-highspeed;
+        sd-uhs-sdr104;
+        keep-power-in-suspend;
+        wakeup-source;
+        cap-sdio-irq;
+        no-mmc;
+        no-sd;
+        non-removable;
+        vmmc-supply = <&sdio_fixed_3v3>;
+        vqmmc-supply = <&mt6397_vgp3_reg>;
+        mmc-pwrseq = <&wifi_pwrseq>;
+    };
+
 ...
-- 
2.25.1

Re: [RESEND v12 1/3] dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties

[Rob Herring]

On Wed, 25 May 2022 09:51:38 +0800, Axe Yang wrote:
> Extend interrupts and pinctrls for SDIO wakeup interrupt feature.
> This feature allow SDIO devices alarm asynchronous interrupt to host
> even when host stop providing clock to SDIO card. An extra wakeup
> interrupt and pinctrl states for SDIO DAT1 pin state switching are
> required in this scenario.
> 
> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
> Signed-off-by: Axe Yang <axe.yang@mediatek.com>
> ---
>  .../devicetree/bindings/mmc/mtk-sd.yaml       | 50 ++++++++++++++++++-
>  1 file changed, 49 insertions(+), 1 deletion(-)
> 

Reviewed-by: Rob Herring <robh@kernel.org>

Re: [RESEND v12 1/3] dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties

[Ulf Hansson]

On Wed, 25 May 2022 at 03:51, Axe Yang <axe.yang@mediatek.com> wrote:
>
> Extend interrupts and pinctrls for SDIO wakeup interrupt feature.
> This feature allow SDIO devices alarm asynchronous interrupt to host
> even when host stop providing clock to SDIO card. An extra wakeup
> interrupt and pinctrl states for SDIO DAT1 pin state switching are
> required in this scenario.
>
> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
> Signed-off-by: Axe Yang <axe.yang@mediatek.com>
> ---
>  .../devicetree/bindings/mmc/mtk-sd.yaml       | 50 ++++++++++++++++++-
>  1 file changed, 49 insertions(+), 1 deletion(-)
>
> diff --git a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> index 2a2e9fa8c188..e83bf10281d6 100644
> --- a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> +++ b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> @@ -72,12 +72,27 @@ properties:
>        - const: ahb_cg
>
>    interrupts:
> -    maxItems: 1
> +    description:
> +      Should at least contain MSDC GIC interrupt. To support SDIO in-band wakeup, an extended
> +      interrupt is required and be configured as wakeup source irq.
> +    minItems: 1
> +    maxItems: 2
> +
> +  interrupt-names:
> +    items:
> +      - const: msdc
> +      - const: sdio_wakeup
>
>    pinctrl-names:
> +    description:
> +      Should at least contain default and state_uhs. To support SDIO in-band wakeup, dat1 pin
> +      will be switched between GPIO mode and SDIO DAT1 mode, state_eint and state_dat1 are
> +      mandatory in this scenarios.
> +    minItems: 2
>      items:
>        - const: default
>        - const: state_uhs
> +      - const: state_eint

Don't you need something along the lines of the below instead? I mean
the "state_eint" isn't always needed, right?

oneOf:
  - items:
      - const: default
      - const: state_uhs
    - items:
        - const: default
        - const: state_uhs
        - const: state_eint

>
>    pinctrl-0:
>      description:
> @@ -89,6 +104,11 @@ properties:
>        should contain uhs mode pin ctrl.
>      maxItems: 1
>
> +  pinctrl-2:
> +    description:
> +      should switch dat1 pin to GPIO mode.
> +    maxItems: 1
> +
>    assigned-clocks:
>      description:
>        PLL of the source clock.
> @@ -208,4 +228,32 @@ examples:
>          mediatek,hs400-cmd-resp-sel-rising;
>      };
>
> +    mmc3: mmc@11260000 {
> +        compatible = "mediatek,mt8173-mmc";
> +        reg = <0x11260000 0x1000>;
> +        clock-names = "source", "hclk";
> +        clocks = <&pericfg CLK_PERI_MSDC30_3>,
> +                 <&topckgen CLK_TOP_MSDC50_2_H_SEL>;
> +        interrupt-names = "msdc", "sdio_wakeup";
> +        interrupts-extended = <&gic GIC_SPI 74 IRQ_TYPE_LEVEL_LOW 0>,
> +                     <&pio 23 IRQ_TYPE_LEVEL_LOW>;
> +        pinctrl-names = "default", "state_uhs", "state_eint";
> +        pinctrl-0 = <&mmc2_pins_default>;
> +        pinctrl-1 = <&mmc2_pins_uhs>;
> +        pinctrl-2 = <&mmc2_pins_eint>;
> +        bus-width = <4>;
> +        max-frequency = <200000000>;
> +        cap-sd-highspeed;
> +        sd-uhs-sdr104;
> +        keep-power-in-suspend;
> +        wakeup-source;
> +        cap-sdio-irq;
> +        no-mmc;
> +        no-sd;
> +        non-removable;
> +        vmmc-supply = <&sdio_fixed_3v3>;
> +        vqmmc-supply = <&mt6397_vgp3_reg>;
> +        mmc-pwrseq = <&wifi_pwrseq>;
> +    };
> +
>  ...
> --
> 2.25.1
>

Kind regards
Uffe

Re: [RESEND v12 1/3] dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties

[Rob Herring]

On Fri, Jun 03, 2022 at 09:28:37AM +0200, Ulf Hansson wrote:
> On Wed, 25 May 2022 at 03:51, Axe Yang <axe.yang@mediatek.com> wrote:
> >
> > Extend interrupts and pinctrls for SDIO wakeup interrupt feature.
> > This feature allow SDIO devices alarm asynchronous interrupt to host
> > even when host stop providing clock to SDIO card. An extra wakeup
> > interrupt and pinctrl states for SDIO DAT1 pin state switching are
> > required in this scenario.
> >
> > Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
> > Signed-off-by: Axe Yang <axe.yang@mediatek.com>
> > ---
> >  .../devicetree/bindings/mmc/mtk-sd.yaml       | 50 ++++++++++++++++++-
> >  1 file changed, 49 insertions(+), 1 deletion(-)
> >
> > diff --git a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > index 2a2e9fa8c188..e83bf10281d6 100644
> > --- a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > +++ b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > @@ -72,12 +72,27 @@ properties:
> >        - const: ahb_cg
> >
> >    interrupts:
> > -    maxItems: 1
> > +    description:
> > +      Should at least contain MSDC GIC interrupt. To support SDIO in-band wakeup, an extended
> > +      interrupt is required and be configured as wakeup source irq.
> > +    minItems: 1
> > +    maxItems: 2
> > +
> > +  interrupt-names:
> > +    items:
> > +      - const: msdc
> > +      - const: sdio_wakeup
> >
> >    pinctrl-names:
> > +    description:
> > +      Should at least contain default and state_uhs. To support SDIO in-band wakeup, dat1 pin
> > +      will be switched between GPIO mode and SDIO DAT1 mode, state_eint and state_dat1 are
> > +      mandatory in this scenarios.
> > +    minItems: 2
> >      items:
> >        - const: default
> >        - const: state_uhs
> > +      - const: state_eint
> 
> Don't you need something along the lines of the below instead? I mean
> the "state_eint" isn't always needed, right?
> 
> oneOf:
>   - items:
>       - const: default
>       - const: state_uhs
>     - items:
>         - const: default
>         - const: state_uhs
>         - const: state_eint

This is equivalent to what was done. The 'minItems: 2' makes the 3rd 
item optional.

Rob

Re: [RESEND v12 1/3] dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties

[Axe Yang]

On Fri, 2022-06-03 at 09:28 +0200, Ulf Hansson wrote:
> On Wed, 25 May 2022 at 03:51, Axe Yang <axe.yang@mediatek.com> wrote:
> > 
> > Extend interrupts and pinctrls for SDIO wakeup interrupt feature.
> > This feature allow SDIO devices alarm asynchronous interrupt to
> > host
> > even when host stop providing clock to SDIO card. An extra wakeup
> > interrupt and pinctrl states for SDIO DAT1 pin state switching are
> > required in this scenario.
> > 
> > Reviewed-by: AngeloGioacchino Del Regno <
> > angelogioacchino.delregno@collabora.com>
> > Signed-off-by: Axe Yang <axe.yang@mediatek.com>
> > ---
> >  .../devicetree/bindings/mmc/mtk-sd.yaml       | 50
> > ++++++++++++++++++-
> >  1 file changed, 49 insertions(+), 1 deletion(-)
> > 
> > diff --git a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > index 2a2e9fa8c188..e83bf10281d6 100644
> > --- a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > +++ b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > @@ -72,12 +72,27 @@ properties:
> >        - const: ahb_cg
> > 
> >    interrupts:
> > -    maxItems: 1
> > +    description:
> > +      Should at least contain MSDC GIC interrupt. To support SDIO
> > in-band wakeup, an extended
> > +      interrupt is required and be configured as wakeup source
> > irq.
> > +    minItems: 1
> > +    maxItems: 2
> > +
> > +  interrupt-names:
> > +    items:
> > +      - const: msdc
> > +      - const: sdio_wakeup
> > 
> >    pinctrl-names:
> > +    description:
> > +      Should at least contain default and state_uhs. To support
> > SDIO in-band wakeup, dat1 pin
> > +      will be switched between GPIO mode and SDIO DAT1 mode,
> > state_eint and state_dat1 are
> > +      mandatory in this scenarios.
> > +    minItems: 2
> >      items:
> >        - const: default
> >        - const: state_uhs
> > +      - const: state_eint
> 
> Don't you need something along the lines of the below instead? I mean
> the "state_eint" isn't always needed, right?
> 
> oneOf:
>   - items:
>       - const: default
>       - const: state_uhs
>     - items:
>         - const: default
>         - const: state_uhs
>         - const: state_eint
> 
No, it is not always needed. 
As Rob said, the 'minItems: 2' makes the 3rd item optional.
Combine 'minItems' and 'description' fields, it is easy for others to
understand how to set pinctrl related properities.

Anyway, If you insist 'oneOf' is the better way, I can update that in
next version. What do you think?

And thanks to your comment, I found a mistake in 'description', I
should remove descriptions related to 'state_dat1', and I will update
that in next version.

And do you have any comment on patch 2/3 and 3/3?


Regards,
Axe

Re: [RESEND v12 1/3] dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties

[Ulf Hansson]

On Tue, 7 Jun 2022 at 10:24, Axe Yang <axe.yang@mediatek.com> wrote:
>
> On Fri, 2022-06-03 at 09:28 +0200, Ulf Hansson wrote:
> > On Wed, 25 May 2022 at 03:51, Axe Yang <axe.yang@mediatek.com> wrote:
> > >
> > > Extend interrupts and pinctrls for SDIO wakeup interrupt feature.
> > > This feature allow SDIO devices alarm asynchronous interrupt to
> > > host
> > > even when host stop providing clock to SDIO card. An extra wakeup
> > > interrupt and pinctrl states for SDIO DAT1 pin state switching are
> > > required in this scenario.
> > >
> > > Reviewed-by: AngeloGioacchino Del Regno <
> > > angelogioacchino.delregno@collabora.com>
> > > Signed-off-by: Axe Yang <axe.yang@mediatek.com>
> > > ---
> > >  .../devicetree/bindings/mmc/mtk-sd.yaml       | 50
> > > ++++++++++++++++++-
> > >  1 file changed, 49 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > > b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > > index 2a2e9fa8c188..e83bf10281d6 100644
> > > --- a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > > +++ b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > > @@ -72,12 +72,27 @@ properties:
> > >        - const: ahb_cg
> > >
> > >    interrupts:
> > > -    maxItems: 1
> > > +    description:
> > > +      Should at least contain MSDC GIC interrupt. To support SDIO
> > > in-band wakeup, an extended
> > > +      interrupt is required and be configured as wakeup source
> > > irq.
> > > +    minItems: 1
> > > +    maxItems: 2
> > > +
> > > +  interrupt-names:
> > > +    items:
> > > +      - const: msdc
> > > +      - const: sdio_wakeup
> > >
> > >    pinctrl-names:
> > > +    description:
> > > +      Should at least contain default and state_uhs. To support
> > > SDIO in-band wakeup, dat1 pin
> > > +      will be switched between GPIO mode and SDIO DAT1 mode,
> > > state_eint and state_dat1 are
> > > +      mandatory in this scenarios.
> > > +    minItems: 2
> > >      items:
> > >        - const: default
> > >        - const: state_uhs
> > > +      - const: state_eint
> >
> > Don't you need something along the lines of the below instead? I mean
> > the "state_eint" isn't always needed, right?
> >
> > oneOf:
> >   - items:
> >       - const: default
> >       - const: state_uhs
> >     - items:
> >         - const: default
> >         - const: state_uhs
> >         - const: state_eint
> >
> No, it is not always needed.
> As Rob said, the 'minItems: 2' makes the 3rd item optional.
> Combine 'minItems' and 'description' fields, it is easy for others to
> understand how to set pinctrl related properities.

Yes, I agree.

>
> Anyway, If you insist 'oneOf' is the better way, I can update that in
> next version. What do you think?

I am fine with it as is, sorry for the noise.

>
> And thanks to your comment, I found a mistake in 'description', I
> should remove descriptions related to 'state_dat1', and I will update
> that in next version.

Okay.

>
> And do you have any comment on patch 2/3 and 3/3?

Sorry for the delay, I will have a look asap.

>
>
> Regards,
> Axe
>
>

Kind regards
Uffe

[RESEND v12 2/3] mmc: core: Add support for SDIO wakeup interrupt

[Axe Yang]

If wakeup-source flag is set in host dts node, parse EAI information
from SDIO CCCR interrupt externsion segment for in-band wakeup. If
async interrupt is supported by SDIO card then enable it and set
enable_async_irq flag in sdio_cccr structure to 1. The parse flow is
implemented in sdio_read_cccr().

Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Axe Yang <axe.yang@mediatek.com>
---
 drivers/mmc/core/sdio.c  | 14 ++++++++++++++
 include/linux/mmc/card.h |  8 +++++++-
 include/linux/mmc/sdio.h |  5 +++++
 3 files changed, 26 insertions(+), 1 deletion(-)

diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c
index 25799accf8a0..165d8b3c3a96 100644
--- a/drivers/mmc/core/sdio.c
+++ b/drivers/mmc/core/sdio.c
@@ -226,6 +226,20 @@ static int sdio_read_cccr(struct mmc_card *card, u32 ocr)
 				card->sw_caps.sd3_drv_type |= SD_DRIVER_TYPE_C;
 			if (data & SDIO_DRIVE_SDTD)
 				card->sw_caps.sd3_drv_type |= SD_DRIVER_TYPE_D;
+
+			ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_INTERRUPT_EXT, 0, &data);
+			if (ret)
+				goto out;
+
+			if (data & SDIO_INTERRUPT_EXT_SAI) {
+				data |= SDIO_INTERRUPT_EXT_EAI;
+				ret = mmc_io_rw_direct(card, 1, 0, SDIO_CCCR_INTERRUPT_EXT,
+						       data, NULL);
+				if (ret)
+					goto out;
+
+				card->cccr.enable_async_irq = 1;
+			}
 		}
 
 		/* if no uhs mode ensure we check for high speed */
diff --git a/include/linux/mmc/card.h b/include/linux/mmc/card.h
index 37f975875102..968650c0b345 100644
--- a/include/linux/mmc/card.h
+++ b/include/linux/mmc/card.h
@@ -219,7 +219,8 @@ struct sdio_cccr {
 				wide_bus:1,
 				high_power:1,
 				high_speed:1,
-				disable_cd:1;
+				disable_cd:1,
+				enable_async_irq:1;
 };
 
 struct sdio_cis {
@@ -343,6 +344,11 @@ static inline bool mmc_large_sector(struct mmc_card *card)
 	return card->ext_csd.data_sector_size == 4096;
 }
 
+static inline int mmc_card_enable_async_irq(struct mmc_card *card)
+{
+	return card->cccr.enable_async_irq;
+}
+
 bool mmc_card_is_blockaddr(struct mmc_card *card);
 
 #define mmc_card_mmc(c)		((c)->type == MMC_TYPE_MMC)
diff --git a/include/linux/mmc/sdio.h b/include/linux/mmc/sdio.h
index 2a05d1ac4f0e..1ef400f28642 100644
--- a/include/linux/mmc/sdio.h
+++ b/include/linux/mmc/sdio.h
@@ -159,6 +159,11 @@
 #define  SDIO_DTSx_SET_TYPE_A	(1 << SDIO_DRIVE_DTSx_SHIFT)
 #define  SDIO_DTSx_SET_TYPE_C	(2 << SDIO_DRIVE_DTSx_SHIFT)
 #define  SDIO_DTSx_SET_TYPE_D	(3 << SDIO_DRIVE_DTSx_SHIFT)
+
+#define SDIO_CCCR_INTERRUPT_EXT	0x16
+#define SDIO_INTERRUPT_EXT_SAI	(1 << 0)
+#define SDIO_INTERRUPT_EXT_EAI	(1 << 1)
+
 /*
  * Function Basic Registers (FBR)
  */
-- 
2.25.1

[RESEND v12 3/3] mmc: mediatek: add support for SDIO eint wakup IRQ

[Axe Yang]

Add support for eint IRQ when MSDC is used as an SDIO host. This
feature requires SDIO device support async IRQ function. With this
feature, SDIO host can be awakened by SDIO card in suspend state,
without additional pin.

MSDC driver will time-share the SDIO DAT1 pin. During suspend, MSDC
turn off clock and switch SDIO DAT1 pin to GPIO mode. And during
resume, switch GPIO function back to DAT1 mode then turn on clock.

Some device tree property should be added or modified in MSDC node
to support SDIO eint IRQ. Pinctrls "state_eint" is mandatory. Since
this feature depends on asynchronous interrupts, "wakeup-source",
"keep-power-in-suspend" and "cap-sdio-irq" flags are necessary, and
the interrupts list should be extended(the interrupt named with
sdio_wakeup):
        &mmcX {
		...
		interrupt-names = "msdc", "sdio_wakeup";
		interrupts-extended = <...>,
                              	      <&pio xxx IRQ_TYPE_LEVEL_LOW>;
                ...
                pinctrl-names = "default", "state_uhs", "state_eint";
                ...
                pinctrl-2 = <&mmc2_pins_eint>;
                ...
                cap-sdio-irq;
		keep-power-in-suspend;
		wakeup-source;
                ...
        };
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Co-developed-by: Yong Mao <yong.mao@mediatek.com>
Signed-off-by: Yong Mao <yong.mao@mediatek.com>
Signed-off-by: Axe Yang <axe.yang@mediatek.com>
---
 drivers/mmc/host/mtk-sd.c | 80 ++++++++++++++++++++++++++++++++++++---
 1 file changed, 74 insertions(+), 6 deletions(-)

diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
index 195dc897188b..2d5b23616df6 100644
--- a/drivers/mmc/host/mtk-sd.c
+++ b/drivers/mmc/host/mtk-sd.c
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0-only
 /*
- * Copyright (c) 2014-2015 MediaTek Inc.
+ * Copyright (c) 2014-2015, 2022 MediaTek Inc.
  * Author: Chaotian.Jing <chaotian.jing@mediatek.com>
  */
 
@@ -20,6 +20,7 @@
 #include <linux/platform_device.h>
 #include <linux/pm.h>
 #include <linux/pm_runtime.h>
+#include <linux/pm_wakeirq.h>
 #include <linux/regulator/consumer.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>
@@ -440,8 +441,10 @@ struct msdc_host {
 	struct pinctrl *pinctrl;
 	struct pinctrl_state *pins_default;
 	struct pinctrl_state *pins_uhs;
+	struct pinctrl_state *pins_eint;
 	struct delayed_work req_timeout;
 	int irq;		/* host interrupt */
+	int eint_irq;		/* interrupt from sdio device for waking up system */
 	struct reset_control *reset;
 
 	struct clk *src_clk;	/* msdc source clock */
@@ -1520,17 +1523,41 @@ static void __msdc_enable_sdio_irq(struct msdc_host *host, int enb)
 
 static void msdc_enable_sdio_irq(struct mmc_host *mmc, int enb)
 {
-	unsigned long flags;
 	struct msdc_host *host = mmc_priv(mmc);
+	unsigned long flags;
+	int ret;
 
 	spin_lock_irqsave(&host->lock, flags);
 	__msdc_enable_sdio_irq(host, enb);
 	spin_unlock_irqrestore(&host->lock, flags);
 
-	if (enb)
-		pm_runtime_get_noresume(host->dev);
-	else
-		pm_runtime_put_noidle(host->dev);
+	if (mmc_card_enable_async_irq(mmc->card) && host->pins_eint) {
+		if (enb) {
+			pinctrl_select_state(host->pinctrl, host->pins_eint);
+			ret = dev_pm_set_dedicated_wake_irq_reverse(host->dev, host->eint_irq);
+
+			if (ret) {
+				dev_err(host->dev, "Failed to register SDIO wakeup irq!\n");
+				host->pins_eint = NULL;
+				pm_runtime_get_noresume(host->dev);
+			} else {
+				dev_info(host->dev, "SDIO eint irq: %d!\n", host->eint_irq);
+				device_init_wakeup(host->dev, true);
+			}
+
+			pinctrl_select_state(host->pinctrl, host->pins_uhs);
+		} else {
+			dev_pm_clear_wake_irq(host->dev);
+		}
+	} else {
+		if (enb) {
+			/* Ensure host->pins_eint is NULL */
+			host->pins_eint = NULL;
+			pm_runtime_get_noresume(host->dev);
+		} else {
+			pm_runtime_put_noidle(host->dev);
+		}
+	}
 }
 
 static irqreturn_t msdc_cmdq_irq(struct msdc_host *host, u32 intsts)
@@ -2631,6 +2658,19 @@ static int msdc_drv_probe(struct platform_device *pdev)
 		goto host_free;
 	}
 
+	/* Support for SDIO eint irq ? */
+	if ((mmc->pm_caps & MMC_PM_WAKE_SDIO_IRQ) && (mmc->pm_caps & MMC_PM_KEEP_POWER)) {
+		host->eint_irq = platform_get_irq_byname(pdev, "sdio_wakeup");
+		if (host->eint_irq > 0) {
+			host->pins_eint = pinctrl_lookup_state(host->pinctrl, "state_eint");
+			if (IS_ERR(host->pins_eint)) {
+				dev_err_probe(&pdev->dev, PTR_ERR(host->pins_eint),
+					      "Cannot find pinctrl eint!\n");
+				host->pins_eint = NULL;
+			}
+		}
+	}
+
 	msdc_of_property_parse(pdev, host);
 
 	host->dev = &pdev->dev;
@@ -2845,6 +2885,12 @@ static int __maybe_unused msdc_runtime_suspend(struct device *dev)
 	struct msdc_host *host = mmc_priv(mmc);
 
 	msdc_save_reg(host);
+
+	if (sdio_irq_claimed(mmc) && host->pins_eint) {
+		disable_irq(host->irq);
+		pinctrl_select_state(host->pinctrl, host->pins_eint);
+		sdr_clr_bits(host->base + SDC_CFG, SDC_CFG_SDIOIDE);
+	}
 	msdc_gate_clock(host);
 	return 0;
 }
@@ -2860,12 +2906,19 @@ static int __maybe_unused msdc_runtime_resume(struct device *dev)
 		return ret;
 
 	msdc_restore_reg(host);
+
+	if (sdio_irq_claimed(mmc) && host->pins_eint) {
+		sdr_set_bits(host->base + SDC_CFG, SDC_CFG_SDIOIDE);
+		pinctrl_select_state(host->pinctrl, host->pins_uhs);
+		enable_irq(host->irq);
+	}
 	return 0;
 }
 
 static int __maybe_unused msdc_suspend(struct device *dev)
 {
 	struct mmc_host *mmc = dev_get_drvdata(dev);
+	struct msdc_host *host = mmc_priv(mmc);
 	int ret;
 
 	if (mmc->caps2 & MMC_CAP2_CQE) {
@@ -2874,11 +2927,26 @@ static int __maybe_unused msdc_suspend(struct device *dev)
 			return ret;
 	}
 
+	if (sdio_irq_claimed(mmc) && host->pins_eint) {
+		pm_runtime_put_sync_suspend(dev);
+
+		return 0;
+	}
+
 	return pm_runtime_force_suspend(dev);
 }
 
 static int __maybe_unused msdc_resume(struct device *dev)
 {
+	struct mmc_host *mmc = dev_get_drvdata(dev);
+	struct msdc_host *host = mmc_priv(mmc);
+
+	if (sdio_irq_claimed(mmc) && host->pins_eint) {
+		pm_runtime_get_sync(dev);
+
+		return 0;
+	}
+
 	return pm_runtime_force_resume(dev);
 }
 
-- 
2.25.1

Re: [RESEND v12 3/3] mmc: mediatek: add support for SDIO eint wakup IRQ

[Chaotian Jing]

On Wed, 2022-05-25 at 09:51 +0800, Axe Yang wrote:
> Add support for eint IRQ when MSDC is used as an SDIO host. This
> feature requires SDIO device support async IRQ function. With this
> feature, SDIO host can be awakened by SDIO card in suspend state,
> without additional pin.
> 
> MSDC driver will time-share the SDIO DAT1 pin. During suspend, MSDC
> turn off clock and switch SDIO DAT1 pin to GPIO mode. And during
> resume, switch GPIO function back to DAT1 mode then turn on clock.
> 
> Some device tree property should be added or modified in MSDC node
> to support SDIO eint IRQ. Pinctrls "state_eint" is mandatory. Since
> this feature depends on asynchronous interrupts, "wakeup-source",
> "keep-power-in-suspend" and "cap-sdio-irq" flags are necessary, and
> the interrupts list should be extended(the interrupt named with
> sdio_wakeup):
>         &mmcX {
> 		...
> 		interrupt-names = "msdc", "sdio_wakeup";
> 		interrupts-extended = <...>,
>                               	      <&pio xxx
> IRQ_TYPE_LEVEL_LOW>;
>                 ...
>                 pinctrl-names = "default", "state_uhs", "state_eint";
>                 ...
>                 pinctrl-2 = <&mmc2_pins_eint>;
>                 ...
>                 cap-sdio-irq;
> 		keep-power-in-suspend;
> 		wakeup-source;
>                 ...
>         };
> Reviewed-by: AngeloGioacchino Del Regno <
> angelogioacchino.delregno@collabora.com>
> Co-developed-by: Yong Mao <yong.mao@mediatek.com>
> Signed-off-by: Yong Mao <yong.mao@mediatek.com>
> Signed-off-by: Axe Yang <axe.yang@mediatek.com>
Reviewed-by: Chaotian Jing <chaotian.jing@mediatek.com>
> ---
>  drivers/mmc/host/mtk-sd.c | 80 ++++++++++++++++++++++++++++++++++++-
> --
>  1 file changed, 74 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
> index 195dc897188b..2d5b23616df6 100644
> --- a/drivers/mmc/host/mtk-sd.c
> +++ b/drivers/mmc/host/mtk-sd.c
> @@ -1,6 +1,6 @@
>  // SPDX-License-Identifier: GPL-2.0-only
>  /*
> - * Copyright (c) 2014-2015 MediaTek Inc.
> + * Copyright (c) 2014-2015, 2022 MediaTek Inc.
>   * Author: Chaotian.Jing <chaotian.jing@mediatek.com>
>   */
>  
> @@ -20,6 +20,7 @@
>  #include <linux/platform_device.h>
>  #include <linux/pm.h>
>  #include <linux/pm_runtime.h>
> +#include <linux/pm_wakeirq.h>
>  #include <linux/regulator/consumer.h>
>  #include <linux/slab.h>
>  #include <linux/spinlock.h>
> @@ -440,8 +441,10 @@ struct msdc_host {
>  	struct pinctrl *pinctrl;
>  	struct pinctrl_state *pins_default;
>  	struct pinctrl_state *pins_uhs;
> +	struct pinctrl_state *pins_eint;
>  	struct delayed_work req_timeout;
>  	int irq;		/* host interrupt */
> +	int eint_irq;		/* interrupt from sdio device for
> waking up system */
>  	struct reset_control *reset;
>  
>  	struct clk *src_clk;	/* msdc source clock */
> @@ -1520,17 +1523,41 @@ static void __msdc_enable_sdio_irq(struct
> msdc_host *host, int enb)
>  
>  static void msdc_enable_sdio_irq(struct mmc_host *mmc, int enb)
>  {
> -	unsigned long flags;
>  	struct msdc_host *host = mmc_priv(mmc);
> +	unsigned long flags;
> +	int ret;
>  
>  	spin_lock_irqsave(&host->lock, flags);
>  	__msdc_enable_sdio_irq(host, enb);
>  	spin_unlock_irqrestore(&host->lock, flags);
>  
> -	if (enb)
> -		pm_runtime_get_noresume(host->dev);
> -	else
> -		pm_runtime_put_noidle(host->dev);
> +	if (mmc_card_enable_async_irq(mmc->card) && host->pins_eint) {
> +		if (enb) {
> +			pinctrl_select_state(host->pinctrl, host-
> >pins_eint);
> +			ret =
> dev_pm_set_dedicated_wake_irq_reverse(host->dev, host->eint_irq);
> +
> +			if (ret) {
> +				dev_err(host->dev, "Failed to register
> SDIO wakeup irq!\n");
> +				host->pins_eint = NULL;
> +				pm_runtime_get_noresume(host->dev);
> +			} else {
> +				dev_info(host->dev, "SDIO eint irq:
> %d!\n", host->eint_irq);
> +				device_init_wakeup(host->dev, true);
> +			}
> +
> +			pinctrl_select_state(host->pinctrl, host-
> >pins_uhs);
> +		} else {
> +			dev_pm_clear_wake_irq(host->dev);
> +		}
> +	} else {
> +		if (enb) {
> +			/* Ensure host->pins_eint is NULL */
> +			host->pins_eint = NULL;
> +			pm_runtime_get_noresume(host->dev);
> +		} else {
> +			pm_runtime_put_noidle(host->dev);
> +		}
> +	}
>  }
>  
>  static irqreturn_t msdc_cmdq_irq(struct msdc_host *host, u32 intsts)
> @@ -2631,6 +2658,19 @@ static int msdc_drv_probe(struct
> platform_device *pdev)
>  		goto host_free;
>  	}
>  
> +	/* Support for SDIO eint irq ? */
> +	if ((mmc->pm_caps & MMC_PM_WAKE_SDIO_IRQ) && (mmc->pm_caps &
> MMC_PM_KEEP_POWER)) {
> +		host->eint_irq = platform_get_irq_byname(pdev,
> "sdio_wakeup");
> +		if (host->eint_irq > 0) {
> +			host->pins_eint = pinctrl_lookup_state(host-
> >pinctrl, "state_eint");
> +			if (IS_ERR(host->pins_eint)) {
> +				dev_err_probe(&pdev->dev, PTR_ERR(host-
> >pins_eint),
> +					      "Cannot find pinctrl
> eint!\n");
> +				host->pins_eint = NULL;
> +			}
> +		}
> +	}
> +
>  	msdc_of_property_parse(pdev, host);
>  
>  	host->dev = &pdev->dev;
> @@ -2845,6 +2885,12 @@ static int __maybe_unused
> msdc_runtime_suspend(struct device *dev)
>  	struct msdc_host *host = mmc_priv(mmc);
>  
>  	msdc_save_reg(host);
> +
> +	if (sdio_irq_claimed(mmc) && host->pins_eint) {
> +		disable_irq(host->irq);
> +		pinctrl_select_state(host->pinctrl, host->pins_eint);
> +		sdr_clr_bits(host->base + SDC_CFG, SDC_CFG_SDIOIDE);
> +	}
>  	msdc_gate_clock(host);
>  	return 0;
>  }
> @@ -2860,12 +2906,19 @@ static int __maybe_unused
> msdc_runtime_resume(struct device *dev)
>  		return ret;
>  
>  	msdc_restore_reg(host);
> +
> +	if (sdio_irq_claimed(mmc) && host->pins_eint) {
> +		sdr_set_bits(host->base + SDC_CFG, SDC_CFG_SDIOIDE);
> +		pinctrl_select_state(host->pinctrl, host->pins_uhs);
> +		enable_irq(host->irq);
> +	}
>  	return 0;
>  }
>  
>  static int __maybe_unused msdc_suspend(struct device *dev)
>  {
>  	struct mmc_host *mmc = dev_get_drvdata(dev);
> +	struct msdc_host *host = mmc_priv(mmc);
>  	int ret;
>  
>  	if (mmc->caps2 & MMC_CAP2_CQE) {
> @@ -2874,11 +2927,26 @@ static int __maybe_unused msdc_suspend(struct
> device *dev)
>  			return ret;
>  	}
>  
> +	if (sdio_irq_claimed(mmc) && host->pins_eint) {
> +		pm_runtime_put_sync_suspend(dev);
> +
> +		return 0;
> +	}
> +
>  	return pm_runtime_force_suspend(dev);
>  }
>  
>  static int __maybe_unused msdc_resume(struct device *dev)
>  {
> +	struct mmc_host *mmc = dev_get_drvdata(dev);
> +	struct msdc_host *host = mmc_priv(mmc);
> +
> +	if (sdio_irq_claimed(mmc) && host->pins_eint) {
> +		pm_runtime_get_sync(dev);
> +
> +		return 0;
> +	}
> +
>  	return pm_runtime_force_resume(dev);
>  }
>  

Re: [RESEND v12 3/3] mmc: mediatek: add support for SDIO eint wakup IRQ

[Ulf Hansson]

On Wed, 25 May 2022 at 03:51, Axe Yang <axe.yang@mediatek.com> wrote:
>
> Add support for eint IRQ when MSDC is used as an SDIO host. This
> feature requires SDIO device support async IRQ function. With this
> feature, SDIO host can be awakened by SDIO card in suspend state,
> without additional pin.
>
> MSDC driver will time-share the SDIO DAT1 pin. During suspend, MSDC
> turn off clock and switch SDIO DAT1 pin to GPIO mode. And during
> resume, switch GPIO function back to DAT1 mode then turn on clock.
>
> Some device tree property should be added or modified in MSDC node
> to support SDIO eint IRQ. Pinctrls "state_eint" is mandatory. Since
> this feature depends on asynchronous interrupts, "wakeup-source",
> "keep-power-in-suspend" and "cap-sdio-irq" flags are necessary, and
> the interrupts list should be extended(the interrupt named with
> sdio_wakeup):
>         &mmcX {
>                 ...
>                 interrupt-names = "msdc", "sdio_wakeup";
>                 interrupts-extended = <...>,
>                                       <&pio xxx IRQ_TYPE_LEVEL_LOW>;
>                 ...
>                 pinctrl-names = "default", "state_uhs", "state_eint";
>                 ...
>                 pinctrl-2 = <&mmc2_pins_eint>;
>                 ...
>                 cap-sdio-irq;
>                 keep-power-in-suspend;
>                 wakeup-source;
>                 ...
>         };
> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
> Co-developed-by: Yong Mao <yong.mao@mediatek.com>
> Signed-off-by: Yong Mao <yong.mao@mediatek.com>
> Signed-off-by: Axe Yang <axe.yang@mediatek.com>
> ---
>  drivers/mmc/host/mtk-sd.c | 80 ++++++++++++++++++++++++++++++++++++---
>  1 file changed, 74 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
> index 195dc897188b..2d5b23616df6 100644
> --- a/drivers/mmc/host/mtk-sd.c
> +++ b/drivers/mmc/host/mtk-sd.c
> @@ -1,6 +1,6 @@
>  // SPDX-License-Identifier: GPL-2.0-only
>  /*
> - * Copyright (c) 2014-2015 MediaTek Inc.
> + * Copyright (c) 2014-2015, 2022 MediaTek Inc.
>   * Author: Chaotian.Jing <chaotian.jing@mediatek.com>
>   */
>
> @@ -20,6 +20,7 @@
>  #include <linux/platform_device.h>
>  #include <linux/pm.h>
>  #include <linux/pm_runtime.h>
> +#include <linux/pm_wakeirq.h>
>  #include <linux/regulator/consumer.h>
>  #include <linux/slab.h>
>  #include <linux/spinlock.h>
> @@ -440,8 +441,10 @@ struct msdc_host {
>         struct pinctrl *pinctrl;
>         struct pinctrl_state *pins_default;
>         struct pinctrl_state *pins_uhs;
> +       struct pinctrl_state *pins_eint;
>         struct delayed_work req_timeout;
>         int irq;                /* host interrupt */
> +       int eint_irq;           /* interrupt from sdio device for waking up system */
>         struct reset_control *reset;
>
>         struct clk *src_clk;    /* msdc source clock */
> @@ -1520,17 +1523,41 @@ static void __msdc_enable_sdio_irq(struct msdc_host *host, int enb)
>
>  static void msdc_enable_sdio_irq(struct mmc_host *mmc, int enb)
>  {
> -       unsigned long flags;
>         struct msdc_host *host = mmc_priv(mmc);
> +       unsigned long flags;
> +       int ret;
>
>         spin_lock_irqsave(&host->lock, flags);
>         __msdc_enable_sdio_irq(host, enb);
>         spin_unlock_irqrestore(&host->lock, flags);
>
> -       if (enb)
> -               pm_runtime_get_noresume(host->dev);
> -       else
> -               pm_runtime_put_noidle(host->dev);
> +       if (mmc_card_enable_async_irq(mmc->card) && host->pins_eint) {
> +               if (enb) {
> +                       pinctrl_select_state(host->pinctrl, host->pins_eint);

This looks a bit odd to me. The pins are not supposed to be configured
for wakeirq at this point, right?

As I understand it, the pin state for wakeirq should be set from the
->runtime_suspend() callback, no?

> +                       ret = dev_pm_set_dedicated_wake_irq_reverse(host->dev, host->eint_irq);
> +
> +                       if (ret) {
> +                               dev_err(host->dev, "Failed to register SDIO wakeup irq!\n");
> +                               host->pins_eint = NULL;
> +                               pm_runtime_get_noresume(host->dev);
> +                       } else {
> +                               dev_info(host->dev, "SDIO eint irq: %d!\n", host->eint_irq);

If you want to log a message, please use a dev_dbg for this instead.

> +                               device_init_wakeup(host->dev, true);

To me, it looks like this is better called from ->probe(), once.

> +                       }
> +
> +                       pinctrl_select_state(host->pinctrl, host->pins_uhs);

Assuming that we can drop the earlier call to pinctrl_select_state()
to set "host->pins_eint", this call can be dropped too.

> +               } else {
> +                       dev_pm_clear_wake_irq(host->dev);
> +               }
> +       } else {
> +               if (enb) {
> +                       /* Ensure host->pins_eint is NULL */
> +                       host->pins_eint = NULL;
> +                       pm_runtime_get_noresume(host->dev);
> +               } else {
> +                       pm_runtime_put_noidle(host->dev);
> +               }
> +       }
>  }
>
>  static irqreturn_t msdc_cmdq_irq(struct msdc_host *host, u32 intsts)
> @@ -2631,6 +2658,19 @@ static int msdc_drv_probe(struct platform_device *pdev)
>                 goto host_free;
>         }
>
> +       /* Support for SDIO eint irq ? */
> +       if ((mmc->pm_caps & MMC_PM_WAKE_SDIO_IRQ) && (mmc->pm_caps & MMC_PM_KEEP_POWER)) {
> +               host->eint_irq = platform_get_irq_byname(pdev, "sdio_wakeup");
> +               if (host->eint_irq > 0) {
> +                       host->pins_eint = pinctrl_lookup_state(host->pinctrl, "state_eint");
> +                       if (IS_ERR(host->pins_eint)) {
> +                               dev_err_probe(&pdev->dev, PTR_ERR(host->pins_eint),
> +                                             "Cannot find pinctrl eint!\n");

We can probably use dev_err() instead of dev_err_probe() as
pinctrl_lookup_state() should never return -EPROBE_DEFER, I think.

> +                               host->pins_eint = NULL;
> +                       }
> +               }
> +       }
> +
>         msdc_of_property_parse(pdev, host);
>
>         host->dev = &pdev->dev;
> @@ -2845,6 +2885,12 @@ static int __maybe_unused msdc_runtime_suspend(struct device *dev)
>         struct msdc_host *host = mmc_priv(mmc);
>
>         msdc_save_reg(host);
> +
> +       if (sdio_irq_claimed(mmc) && host->pins_eint) {
> +               disable_irq(host->irq);
> +               pinctrl_select_state(host->pinctrl, host->pins_eint);
> +               sdr_clr_bits(host->base + SDC_CFG, SDC_CFG_SDIOIDE);

__msdc_enable_sdio_irq() also calls "sdr_clr_bits(host->base +
MSDC_INTEN, MSDC_INTEN_SDIOIRQ);"

Perhaps we should call __msdc_enable_sdio_irq() here instead? To be consistent.

> +       }
>         msdc_gate_clock(host);
>         return 0;
>  }
> @@ -2860,12 +2906,19 @@ static int __maybe_unused msdc_runtime_resume(struct device *dev)
>                 return ret;
>
>         msdc_restore_reg(host);
> +
> +       if (sdio_irq_claimed(mmc) && host->pins_eint) {

msdc_restore_reg() already calls __msdc_enable_sdio_irq(), but based
only upon whether sdio_irq_claimed() returns true.

It looks like we should align the code in
msdc_runtime_resume|suspend(). Perhaps sdio_irq_claimed() should
indicate in both cases that __msdc_enable_sdio_irq() needs to be
called, while "host->pins_eint" means that we have also additional
wakeup configurations (pins and irqs) to handle.

> +               sdr_set_bits(host->base + SDC_CFG, SDC_CFG_SDIOIDE);
> +               pinctrl_select_state(host->pinctrl, host->pins_uhs);
> +               enable_irq(host->irq);
> +       }
>         return 0;
>  }
>
>  static int __maybe_unused msdc_suspend(struct device *dev)
>  {
>         struct mmc_host *mmc = dev_get_drvdata(dev);
> +       struct msdc_host *host = mmc_priv(mmc);
>         int ret;
>
>         if (mmc->caps2 & MMC_CAP2_CQE) {
> @@ -2874,11 +2927,26 @@ static int __maybe_unused msdc_suspend(struct device *dev)
>                         return ret;
>         }
>
> +       if (sdio_irq_claimed(mmc) && host->pins_eint) {
> +               pm_runtime_put_sync_suspend(dev);
> +
> +               return 0;
> +       }
> +

I assume the point with the above is to trigger rpm_suspend() to be
called for the device, so that the wakeirq can be enabled, correctly?

However, this isn't the correct way to do it (for various reasons I
can explain, if you want). Instead I think there are two options going
forward:
1. Deal with the wakeirq from the system suspend/resume callbacks,
locally in the driver.
2. Extend pm_runtime_force_suspend|resume() to let it deal with the
wakeirq for us. Similar to what rpm_suspend|resume() do.

I am inclined to try with option 2) first, as this would prevent the
boilerplate code that otherwise gets introduced by option 1). To help
out, I have prepared a patch that I am about to send, I will keep you
posted.

>         return pm_runtime_force_suspend(dev);
>  }
>
>  static int __maybe_unused msdc_resume(struct device *dev)
>  {
> +       struct mmc_host *mmc = dev_get_drvdata(dev);
> +       struct msdc_host *host = mmc_priv(mmc);
> +
> +       if (sdio_irq_claimed(mmc) && host->pins_eint) {
> +               pm_runtime_get_sync(dev);

Similar comments apply to this as for msdc_suspend().

> +
> +               return 0;
> +       }
> +
>         return pm_runtime_force_resume(dev);
>  }
>

Kind regards
Uffe

Re: [RESEND v12 3/3] mmc: mediatek: add support for SDIO eint wakup IRQ

[Axe Yang]

On Wed, 2022-06-08 at 16:13 +0200, Ulf Hansson wrote:
> On Wed, 25 May 2022 at 03:51, Axe Yang <axe.yang@mediatek.com> wrote:
> > 
> > Add support for eint IRQ when MSDC is used as an SDIO host. This
> > feature requires SDIO device support async IRQ function. With this
> > feature, SDIO host can be awakened by SDIO card in suspend state,
> > without additional pin.
> > 
> > MSDC driver will time-share the SDIO DAT1 pin. During suspend, MSDC
> > turn off clock and switch SDIO DAT1 pin to GPIO mode. And during
> > resume, switch GPIO function back to DAT1 mode then turn on clock.
> > 
> > Some device tree property should be added or modified in MSDC node
> > to support SDIO eint IRQ. Pinctrls "state_eint" is mandatory. Since
> > this feature depends on asynchronous interrupts, "wakeup-source",
> > "keep-power-in-suspend" and "cap-sdio-irq" flags are necessary, and
> > the interrupts list should be extended(the interrupt named with
> > sdio_wakeup):
> >         &mmcX {
> >                 ...
> >                 interrupt-names = "msdc", "sdio_wakeup";
> >                 interrupts-extended = <...>,
> >                                       <&pio xxx
> > IRQ_TYPE_LEVEL_LOW>;
> >                 ...
> >                 pinctrl-names = "default", "state_uhs",
> > "state_eint";
> >                 ...
> >                 pinctrl-2 = <&mmc2_pins_eint>;
> >                 ...
> >                 cap-sdio-irq;
> >                 keep-power-in-suspend;
> >                 wakeup-source;
> >                 ...
> >         };
> > Reviewed-by: AngeloGioacchino Del Regno <
> > angelogioacchino.delregno@collabora.com>
> > Co-developed-by: Yong Mao <yong.mao@mediatek.com>
> > Signed-off-by: Yong Mao <yong.mao@mediatek.com>
> > Signed-off-by: Axe Yang <axe.yang@mediatek.com>
> > ---
> >  drivers/mmc/host/mtk-sd.c | 80
> > ++++++++++++++++++++++++++++++++++++---
> >  1 file changed, 74 insertions(+), 6 deletions(-)
> > 
> > diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
> > index 195dc897188b..2d5b23616df6 100644
> > --- a/drivers/mmc/host/mtk-sd.c
> > +++ b/drivers/mmc/host/mtk-sd.c
> > @@ -1,6 +1,6 @@
> >  // SPDX-License-Identifier: GPL-2.0-only
> >  /*
> > - * Copyright (c) 2014-2015 MediaTek Inc.
> > + * Copyright (c) 2014-2015, 2022 MediaTek Inc.
> >   * Author: Chaotian.Jing <chaotian.jing@mediatek.com>
> >   */
> > 
> > @@ -20,6 +20,7 @@
> >  #include <linux/platform_device.h>
> >  #include <linux/pm.h>
> >  #include <linux/pm_runtime.h>
> > +#include <linux/pm_wakeirq.h>
> >  #include <linux/regulator/consumer.h>
> >  #include <linux/slab.h>
> >  #include <linux/spinlock.h>
> > @@ -440,8 +441,10 @@ struct msdc_host {
> >         struct pinctrl *pinctrl;
> >         struct pinctrl_state *pins_default;
> >         struct pinctrl_state *pins_uhs;
> > +       struct pinctrl_state *pins_eint;
> >         struct delayed_work req_timeout;
> >         int irq;                /* host interrupt */
> > +       int eint_irq;           /* interrupt from sdio device for
> > waking up system */
> >         struct reset_control *reset;
> > 
> >         struct clk *src_clk;    /* msdc source clock */
> > @@ -1520,17 +1523,41 @@ static void __msdc_enable_sdio_irq(struct
> > msdc_host *host, int enb)
> > 
> >  static void msdc_enable_sdio_irq(struct mmc_host *mmc, int enb)
> >  {
> > -       unsigned long flags;
> >         struct msdc_host *host = mmc_priv(mmc);
> > +       unsigned long flags;
> > +       int ret;
> > 
> >         spin_lock_irqsave(&host->lock, flags);
> >         __msdc_enable_sdio_irq(host, enb);
> >         spin_unlock_irqrestore(&host->lock, flags);
> > 
> > -       if (enb)
> > -               pm_runtime_get_noresume(host->dev);
> > -       else
> > -               pm_runtime_put_noidle(host->dev);
> > +       if (mmc_card_enable_async_irq(mmc->card) && host-
> > >pins_eint) {
> > +               if (enb) {
> > +                       pinctrl_select_state(host->pinctrl, host-
> > >pins_eint);
> 
> This looks a bit odd to me. The pins are not supposed to be
> configured
> for wakeirq at this point, right?
> 
> As I understand it, the pin state for wakeirq should be set from the
> ->runtime_suspend() callback, no?

Yes, it is odd, but necessary.
In dev_pm_set_dedicated_wake_irq_reverse() -> ... ->
request_threaded_irq() -> __setup_irq() -> irq_request_resources() ->
mtk_eint_irq_request_resources(), the SDIO DAT1 pin will be force reset
to GPIO mode.
I have to call pinctrl_select_state() in pairs to restore DAT1 pin
state to MSDC mode.

> 
> > +                       ret =
> > dev_pm_set_dedicated_wake_irq_reverse(host->dev, host->eint_irq);
> > +
> > +                       if (ret) {
> > +                               dev_err(host->dev, "Failed to
> > register SDIO wakeup irq!\n");
> > +                               host->pins_eint = NULL;
> > +                               pm_runtime_get_noresume(host->dev);
> > +                       } else {
> > +                               dev_info(host->dev, "SDIO eint irq:
> > %d!\n", host->eint_irq);
> 
> If you want to log a message, please use a dev_dbg for this instead.

Sure, will change to dev_dbg in next version.

> 
> > +                               device_init_wakeup(host->dev,
> > true);
> 
> To me, it looks like this is better called from ->probe(), once.

Will move it to probe() in next version.

> 
> > +                       }
> > +
> > +                       pinctrl_select_state(host->pinctrl, host-
> > >pins_uhs);
> 
> Assuming that we can drop the earlier call to pinctrl_select_state()
> to set "host->pins_eint", this call can be dropped too.

Can not drop this call, it is for restore DAT1 pinmux to MSDC mode.

> 
> > +               } else {
> > +                       dev_pm_clear_wake_irq(host->dev);
> > +               }
> > +       } else {
> > +               if (enb) {
> > +                       /* Ensure host->pins_eint is NULL */
> > +                       host->pins_eint = NULL;
> > +                       pm_runtime_get_noresume(host->dev);
> > +               } else {
> > +                       pm_runtime_put_noidle(host->dev);
> > +               }
> > +       }
> >  }
> > 
> >  static irqreturn_t msdc_cmdq_irq(struct msdc_host *host, u32
> > intsts)
> > @@ -2631,6 +2658,19 @@ static int msdc_drv_probe(struct
> > platform_device *pdev)
> >                 goto host_free;
> >         }
> > 
> > +       /* Support for SDIO eint irq ? */
> > +       if ((mmc->pm_caps & MMC_PM_WAKE_SDIO_IRQ) && (mmc->pm_caps
> > & MMC_PM_KEEP_POWER)) {
> > +               host->eint_irq = platform_get_irq_byname(pdev,
> > "sdio_wakeup");
> > +               if (host->eint_irq > 0) {
> > +                       host->pins_eint =
> > pinctrl_lookup_state(host->pinctrl, "state_eint");
> > +                       if (IS_ERR(host->pins_eint)) {
> > +                               dev_err_probe(&pdev->dev,
> > PTR_ERR(host->pins_eint),
> > +                                             "Cannot find pinctrl
> > eint!\n");
> 
> We can probably use dev_err() instead of dev_err_probe() as
> pinctrl_lookup_state() should never return -EPROBE_DEFER, I think.

Yes. will update that in next version.

> 
> > +                               host->pins_eint = NULL;
> > +                       }
> > +               }
> > +       }
> > +
> >         msdc_of_property_parse(pdev, host);
> > 
> >         host->dev = &pdev->dev;
> > @@ -2845,6 +2885,12 @@ static int __maybe_unused
> > msdc_runtime_suspend(struct device *dev)
> >         struct msdc_host *host = mmc_priv(mmc);
> > 
> >         msdc_save_reg(host);
> > +
> > +       if (sdio_irq_claimed(mmc) && host->pins_eint) {
> > +               disable_irq(host->irq);
> > +               pinctrl_select_state(host->pinctrl, host-
> > >pins_eint);
> > +               sdr_clr_bits(host->base + SDC_CFG,
> > SDC_CFG_SDIOIDE);
> 
> __msdc_enable_sdio_irq() also calls "sdr_clr_bits(host->base +
> MSDC_INTEN, MSDC_INTEN_SDIOIRQ);"
> 
> Perhaps we should call __msdc_enable_sdio_irq() here instead? To be
> consistent.

Yes, we can call __msdc_enable_sdio_irq() instead. Will update that in
next verison.

> 
> > +       }
> >         msdc_gate_clock(host);
> >         return 0;
> >  }
> > @@ -2860,12 +2906,19 @@ static int __maybe_unused
> > msdc_runtime_resume(struct device *dev)
> >                 return ret;
> > 
> >         msdc_restore_reg(host);
> > +
> > +       if (sdio_irq_claimed(mmc) && host->pins_eint) {
> 
> msdc_restore_reg() already calls __msdc_enable_sdio_irq(), but based
> only upon whether sdio_irq_claimed() returns true.
> 
Will remove redundancy MSDC_INTEN_SDIOIRQ control in
msdc_runtion_resume() in next version.

> It looks like we should align the code in
> msdc_runtime_resume|suspend(). Perhaps sdio_irq_claimed() should
> indicate in both cases that __msdc_enable_sdio_irq() needs to be
> called, while "host->pins_eint" means that we have also additional
> wakeup configurations (pins and irqs) to handle.

sdr_set_bits(..., SDC_CFG_SDIOIDE) is only needed when wake up irq is
enabled. So, maybe I can refactor this part as below:
   if (host->pins_eint) {
   	...
   	if (sdio_irq_claimed()) 
   		__msdc_enable_sdio_irq(host, 0);
   	...
   }
in msdc_runtime_suspend() ?

in msdc_runtime_resume(), __msdc_enable_sdio_irq() is called in
msdc_restore_reg(). 
While in msdc_runtime_suspend(), it is called only if  host->pins_eint
return true.
In this case, msdc_runtime_resume|suspend() are not prefectly aligned.
Any better way to do that?

> 
> > +               sdr_set_bits(host->base + SDC_CFG,
> > SDC_CFG_SDIOIDE);
> > +               pinctrl_select_state(host->pinctrl, host-
> > >pins_uhs);
> > +               enable_irq(host->irq);
> > +       }
> >         return 0;
> >  }
> > 
> >  static int __maybe_unused msdc_suspend(struct device *dev)
> >  {
> >         struct mmc_host *mmc = dev_get_drvdata(dev);
> > +       struct msdc_host *host = mmc_priv(mmc);
> >         int ret;
> > 
> >         if (mmc->caps2 & MMC_CAP2_CQE) {
> > @@ -2874,11 +2927,26 @@ static int __maybe_unused
> > msdc_suspend(struct device *dev)
> >                         return ret;
> >         }
> > 
> > +       if (sdio_irq_claimed(mmc) && host->pins_eint) {
> > +               pm_runtime_put_sync_suspend(dev);
> > +
> > +               return 0;
> > +       }
> > +
> 
> I assume the point with the above is to trigger rpm_suspend() to be
> called for the device, so that the wakeirq can be enabled, correctly?

Yes, the intention is to trigger rpm_suspend|resume(), with these two
functions all operations on wakeirq can be performed completely.
But with pm_rumtime_force_resume(), dev->power.needs_force_resume is
false, the function goto out directly everytime.

> 
> However, this isn't the correct way to do it (for various reasons I
> can explain, if you want). Instead I think there are two options
> going
> forward:
> 1. Deal with the wakeirq from the system suspend/resume callbacks,
> locally in the driver.
> 2. Extend pm_runtime_force_suspend|resume() to let it deal with the
> wakeirq for us. Similar to what rpm_suspend|resume() do.
> 
> I am inclined to try with option 2) first, as this would prevent the
> boilerplate code that otherwise gets introduced by option 1). To help
> out, I have prepared a patch that I am about to send, I will keep you
> posted.

Thanks for the patch. And I still need to use pm_runtime_get_noresume()
to bump up runtime PM usage counter to ensure pm_rumtime_force_resume()
can be fully executed.

> 
> >         return pm_runtime_force_suspend(dev);
> >  }
> > 
> >  static int __maybe_unused msdc_resume(struct device *dev)
> >  {
> > +       struct mmc_host *mmc = dev_get_drvdata(dev);
> > +       struct msdc_host *host = mmc_priv(mmc);
> > +
> > +       if (sdio_irq_claimed(mmc) && host->pins_eint) {
> > +               pm_runtime_get_sync(dev);
> 
> Similar comments apply to this as for msdc_suspend().

pm_runtime_put_noidle() will be added here to decrement runtime PM
usage counter. Do you have any comments on this?

> 
> > +
> > +               return 0;
> > +       }
> > +
> >         return pm_runtime_force_resume(dev);
> >  }
> > 
> 
> Kind regards
> Uffe

Regards,
Axe

Re: [RESEND v12 0/3] mmc: mediatek: add support for SDIO async IRQ

[Axe Yang]

Hi,

Gentle ping for this set.

On Wed, 2022-05-25 at 09:51 +0800, Axe Yang wrote:
> Changes in v12:
> - assign NULL to pins_eint directly instead of using kfree()
> 
> Changes in v11:
> - remove '_irq' suffix in interrupts-names property
> - fix yaml example build error
> - refactor msdc_enable_sdio_irq(), free pins_eint if async irq is not
> supported
> 
> Changes in v10:
> - add sample node for SDIO host which support wakeup interrupt in
> yaml
> - skip MMC_PM_WAKE_SDIO_IRQ check before enable SDIO async interrupt
> - add MMC_PM_KEEP_POWER check before SDIO eint pinstate parsing
> - use dev_pm_set_dedicated_wake_irq_reverse() to correct irq control
> sequence
> - set dedicated irq in msdc_enable_sdio_irq() rather than
> msdc_drv_probe()
> - remove unnecessary wake irq control, rpm/dpm system shall manage
> that
> - move wake irq/msdc irq control back to system suspend phase, use
> rpm_suspend
>   and rpm_resume to ensure irq control sequence:
>      disable msdc irq -> enable wake irq -> disable wake irq ->
> enable msdc irq
> - simplify variables, check pins_eint to know whether wakeup settings
> are managed
> 
> Changes in v9:
> - remove pinctrl "state_dat1"
> 
> Changes in v8:
> - remove maxItems property under pinctrl-names property
> 
> Changes in v7:
> - add device_init_wakeup() to register SDIO host as wakeup source
> 
> Changes in v6:
> - abandon cap-sdio-async-irq flag, use wakeup-source flag instead
> - extend interrupts and pinctrls in mediatek mmc host controller DT
> documents
> - add mmc_card_enable_async_irq() to access enable_async_irq flag
> - simplify wakeup irq implementation with dedicate wake up irq
> related interface
> 
> Changes in v5:
> - resort variables to reversed xmas tree order
> - restore old copyright year range and add current year back
> 
> Changes in v4:
> - add MMC_CAP2_SDIO_ASYNC_IRQ judge before lookup eint pinctrl
> - replace spin_lock_irqsave() variant with spin_lock() in eint irq
> handler
> 
> Changes in v3:
> - correct abbreviations with capital letters in commit message
> - replace copyright year with 2022 in mtk-sd.c
> - remove unnessary pointer casting
> - adjust variable order to reversed xmas tree
> - remove a redundant blank line
> - refine if statement, following standard pattern
> 
> Changes in v2:
> - change flag name from 'cap-sdio-async-int' to 'cap-sdio-async-irq'
> - change corresponding macro names from xxx_INT to xxx_IRQ
> - resort new member in msdc_host structure
> - refine function msdc_request_dat1_eint_irq()
> - rename msdc_{suspend,resume} function names, add suffix '_noirq'
> - add MMC_CAP2_NO_SDIO judgement before parse eint related pin
> setting
> 
> Axe Yang (3):
>   dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties
>   mmc: core: Add support for SDIO wakeup interrupt
>   mmc: mediatek: add support for SDIO eint wakup IRQ
> 
>  .../devicetree/bindings/mmc/mtk-sd.yaml       | 50 +++++++++++-
>  drivers/mmc/core/sdio.c                       | 14 ++++
>  drivers/mmc/host/mtk-sd.c                     | 80
> +++++++++++++++++--
>  include/linux/mmc/card.h                      |  8 +-
>  include/linux/mmc/sdio.h                      |  5 ++
>  5 files changed, 149 insertions(+), 8 deletions(-)
> 

Re: [RESEND v12 0/3] mmc: mediatek: add support for SDIO async IRQ

[Andy Shevchenko]

On Thu, Jun 02, 2022 at 02:07:04PM +0800, Axe Yang wrote:
> Hi,
> 
> Gentle ping for this set.

It's a merge window, so your code perhaps will be considered next week,
or later.

-- 
With Best Regards,
Andy Shevchenko

Re: [RESEND v12 0/3] mmc: mediatek: add support for SDIO async IRQ

[Axe Yang]

On Thu, 2022-06-02 at 14:10 +0300, Andy Shevchenko wrote:
> On Thu, Jun 02, 2022 at 02:07:04PM +0800, Axe Yang wrote:
> > Hi,
> > 
> > Gentle ping for this set.
> 
> It's a merge window, so your code perhaps will be considered next
> week,
> or later.
> 

Got it, thanks.

Regards,
Axe