[PATCH v1 0/2] rtc: Enhance RTC driver with BBPU bit definitions and shutdown handling

[PATCH v1 0/2] rtc: Enhance RTC driver with BBPU bit definitions and shutdown handling

[ot_shunxi.zhang]

From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>

(1) Add new bits definitions for RTC_BBPU register, 
(2) Add new functions to reset the BBPU alarm status in mt6397 RTC driver
(3) Add mtk_rtc_shutdown function to handle RTC shutdown events.

Shunxi Zhang (2):
  mfd: mt6397: Add new bit definitions for RTC_BBPU register
  rtc: mt6397: Add BBPU alarm status reset and shutdown handling

 drivers/rtc/rtc-mt6397.c       | 37 ++++++++++++++++++++++++++++++++++
 include/linux/mfd/mt6397/rtc.h |  3 +++
 2 files changed, 40 insertions(+)

-- 
2.45.2

[PATCH v1 1/2] mfd: mt6397: Add new bit definitions for RTC_BBPU register

[ot_shunxi.zhang]

From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>

This patch adds new bit definitions for the RTC_BBPU register in the
mt6397 RTC header file. The following bit definitions are introduced:

Signed-off-by: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
---
 include/linux/mfd/mt6397/rtc.h | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/include/linux/mfd/mt6397/rtc.h b/include/linux/mfd/mt6397/rtc.h
index 27883af44f87..001cef6b7302 100644
--- a/include/linux/mfd/mt6397/rtc.h
+++ b/include/linux/mfd/mt6397/rtc.h
@@ -15,8 +15,11 @@
 #include <linux/rtc.h>
 
 #define RTC_BBPU               0x0000
+#define RTC_BBPU_PWREN         BIT(0)
+#define RTC_BBPU_CLR           BIT(1)
+#define RTC_BBPU_RESET_AL      BIT(3)
 #define RTC_BBPU_CBUSY         BIT(6)
-#define RTC_BBPU_KEY            (0x43 << 8)
+#define RTC_BBPU_KEY           (0x43 << 8)
 
 #define RTC_WRTGR_MT6358       0x003a
 #define RTC_WRTGR_MT6397       0x003c
-- 
2.46.0

Re: [PATCH v1 1/2] mfd: mt6397: Add new bit definitions for RTC_BBPU register

[Krzysztof Kozlowski]

On 11/08/2025 10:15, ot_shunxi.zhang@mediatek.com wrote:
> From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> 
> This patch adds new bit definitions for the RTC_BBPU register in the

Why? There is no user of these. Don't add useless defines.

> mt6397 RTC header file. The following bit definitions are introduced:

Hm?

> 
> Signed-off-by: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> ---
>  include/linux/mfd/mt6397/rtc.h | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/include/linux/mfd/mt6397/rtc.h b/include/linux/mfd/mt6397/rtc.h
> index 27883af44f87..001cef6b7302 100644
> --- a/include/linux/mfd/mt6397/rtc.h
> +++ b/include/linux/mfd/mt6397/rtc.h
> @@ -15,8 +15,11 @@
>  #include <linux/rtc.h>
>  
>  #define RTC_BBPU               0x0000
> +#define RTC_BBPU_PWREN         BIT(0)
> +#define RTC_BBPU_CLR           BIT(1)
> +#define RTC_BBPU_RESET_AL      BIT(3)
>  #define RTC_BBPU_CBUSY         BIT(6)
> -#define RTC_BBPU_KEY            (0x43 << 8)
> +#define RTC_BBPU_KEY           (0x43 << 8)


Why?



Best regards,
Krzysztof

Re: [PATCH v1 1/2] mfd: mt6397: Add new bit definitions for RTC_BBPU register

[Giorgi Tchankvetadze]

Cc: Krzysztof Kozlowski <krzk@kernel.org>,
    linux-mediatek@lists.infradead.org,
    linux-rtc@vger.kernel.org,
    linux-kernel@vger.kernel.org

Shunxi,

Can you confirm whether `RTC_BBPU_PWREN` (bit 0),
`RTC_BBPU_CLR` (bit 1) and `RTC_BBPU_RESET_AL` (bit 3) are documented
in the MT6397 datasheet (please cite section/page)? They look like
standard RTC controls (power enable, clear/reset, alarm reset) and
might be useful to include, but I agree with Krzysztof that adding
definitions with no users can accumulate technical debt.

Suggestion: either
- add the definitions when a driver actually needs them, or
- keep them now but add a short rationale in the commit message
  (datasheet reference + intended use) so future reviewers understand
  why they exist.

Also: please split cosmetic whitespace fixes (RTC_BBPU_KEY) into a
separate patch to make review/merge easier.

Thanks for the patch; I’m following the thread.

— Giorgi


On Mon, Aug 11, 2025 at 3:03 PM Krzysztof Kozlowski <krzk@kernel.org> wrote:
>
> On 11/08/2025 10:15, ot_shunxi.zhang@mediatek.com wrote:
> > From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> >
> > This patch adds new bit definitions for the RTC_BBPU register in the
>
> Why? There is no user of these. Don't add useless defines.
>
> > mt6397 RTC header file. The following bit definitions are introduced:
>
> Hm?
>
> >
> > Signed-off-by: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> > ---
> >  include/linux/mfd/mt6397/rtc.h | 5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/include/linux/mfd/mt6397/rtc.h b/include/linux/mfd/mt6397/rtc.h
> > index 27883af44f87..001cef6b7302 100644
> > --- a/include/linux/mfd/mt6397/rtc.h
> > +++ b/include/linux/mfd/mt6397/rtc.h
> > @@ -15,8 +15,11 @@
> >  #include <linux/rtc.h>
> >
> >  #define RTC_BBPU               0x0000
> > +#define RTC_BBPU_PWREN         BIT(0)
> > +#define RTC_BBPU_CLR           BIT(1)
> > +#define RTC_BBPU_RESET_AL      BIT(3)
> >  #define RTC_BBPU_CBUSY         BIT(6)
> > -#define RTC_BBPU_KEY            (0x43 << 8)
> > +#define RTC_BBPU_KEY           (0x43 << 8)
>
>
> Why?
>
>
>
> Best regards,
> Krzysztof
>

Re: [PATCH v1 1/2] mfd: mt6397: Add new bit definitions for RTC_BBPU register

[Shunxi Zhang (章顺喜)]

On Mon, 2025-08-11 at 15:21 +0400, Giorgi Tchankvetadze wrote:
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
> 
> 
> Cc: Krzysztof Kozlowski <krzk@kernel.org>,
>     linux-mediatek@lists.infradead.org,
>     linux-rtc@vger.kernel.org,
>     linux-kernel@vger.kernel.org
> 
> Shunxi,
> 
> Can you confirm whether `RTC_BBPU_PWREN` (bit 0),
> `RTC_BBPU_CLR` (bit 1) and `RTC_BBPU_RESET_AL` (bit 3) are documented
> in the MT6397 datasheet (please cite section/page)? They look like
> standard RTC controls (power enable, clear/reset, alarm reset) and
> might be useful to include, but I agree with Krzysztof that adding
> definitions with no users can accumulate technical debt.
> 
> Suggestion: either
> - add the definitions when a driver actually needs them, or
> - keep them now but add a short rationale in the commit message
>   (datasheet reference + intended use) so future reviewers understand
>   why they exist.
> 
> Also: please split cosmetic whitespace fixes (RTC_BBPU_KEY) into a
> separate patch to make review/merge easier.
> 
> Thanks for the patch; I’m following the thread.
> 
> — Giorgi

Dear sir,
The MT6397 is an integration of several ICs and does not have a
separate IC specification. I will check the relevant IC datasheets
again. I will remove the useless define in next version.

Thanks for your commnets.

Best regards
Shunxi Zhangus
> 
> 
> On Mon, Aug 11, 2025 at 3:03 PM Krzysztof Kozlowski <krzk@kernel.org>
> wrote:
> > 
> > On 11/08/2025 10:15, ot_shunxi.zhang@mediatek.com wrote:
> > > From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> > > 
> > > This patch adds new bit definitions for the RTC_BBPU register in
> > > the
> > 
> > Why? There is no user of these. Don't add useless defines.
> > 
> > > mt6397 RTC header file. The following bit definitions are
> > > introduced:
> > 
> > Hm?
> > 
> > > 
> > > Signed-off-by: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> > > ---
> > >  include/linux/mfd/mt6397/rtc.h | 5 ++++-
> > >  1 file changed, 4 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/include/linux/mfd/mt6397/rtc.h
> > > b/include/linux/mfd/mt6397/rtc.h
> > > index 27883af44f87..001cef6b7302 100644
> > > --- a/include/linux/mfd/mt6397/rtc.h
> > > +++ b/include/linux/mfd/mt6397/rtc.h
> > > @@ -15,8 +15,11 @@
> > >  #include <linux/rtc.h>
> > > 
> > >  #define RTC_BBPU               0x0000
> > > +#define RTC_BBPU_PWREN         BIT(0)
> > > +#define RTC_BBPU_CLR           BIT(1)
> > > +#define RTC_BBPU_RESET_AL      BIT(3)
> > >  #define RTC_BBPU_CBUSY         BIT(6)
> > > -#define RTC_BBPU_KEY            (0x43 << 8)
> > > +#define RTC_BBPU_KEY           (0x43 << 8)
> > 
> > 
> > Why?
> > 
> > 
> > 
> > Best regards,
> > Krzysztof
> > 

Re: [PATCH v1 1/2] mfd: mt6397: Add new bit definitions for RTC_BBPU register

[Shunxi Zhang (章顺喜)]

On Mon, 2025-08-11 at 13:02 +0200, Krzysztof Kozlowski wrote:
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
> 
> 
> On 11/08/2025 10:15, ot_shunxi.zhang@mediatek.com wrote:
> > From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> > 
> > This patch adds new bit definitions for the RTC_BBPU register in
> > the
> 
> Why? There is no user of these. Don't add useless defines.
> 
> > mt6397 RTC header file. The following bit definitions are
> > introduced:
> 
> Hm?
> 
> > 
> > Signed-off-by: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> > ---
> >  include/linux/mfd/mt6397/rtc.h | 5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> > 
> > diff --git a/include/linux/mfd/mt6397/rtc.h
> > b/include/linux/mfd/mt6397/rtc.h
> > index 27883af44f87..001cef6b7302 100644
> > --- a/include/linux/mfd/mt6397/rtc.h
> > +++ b/include/linux/mfd/mt6397/rtc.h
> > @@ -15,8 +15,11 @@
> >  #include <linux/rtc.h>
> > 
> >  #define RTC_BBPU               0x0000
> > +#define RTC_BBPU_PWREN         BIT(0)
> > +#define RTC_BBPU_CLR           BIT(1)
> > +#define RTC_BBPU_RESET_AL      BIT(3)
> >  #define RTC_BBPU_CBUSY         BIT(6)
> > -#define RTC_BBPU_KEY            (0x43 << 8)
> > +#define RTC_BBPU_KEY           (0x43 << 8)
> 
> 
> Why?
> 
> 
> 
> Best regards,
> Krzysztof

Dear sir,
The MT6397 is an integration of several ICs and does not have a
separate IC specification. I will check the relevant IC datasheets
again. I will remove the useless define in next version.

Thanks for your commnets.

Best regards
Shunxi Zhangus

Re: [PATCH v1 1/2] mfd: mt6397: Add new bit definitions for RTC_BBPU register

[Krzysztof Kozlowski]

On 19/08/2025 09:54, Shunxi Zhang (章顺喜) wrote:
>>>  #define RTC_BBPU               0x0000
>>> +#define RTC_BBPU_PWREN         BIT(0)
>>> +#define RTC_BBPU_CLR           BIT(1)
>>> +#define RTC_BBPU_RESET_AL      BIT(3)
>>>  #define RTC_BBPU_CBUSY         BIT(6)
>>> -#define RTC_BBPU_KEY            (0x43 << 8)
>>> +#define RTC_BBPU_KEY           (0x43 << 8)
>>
>>
>> Why?
>>
>>
>>
>> Best regards,
>> Krzysztof
> 
> Dear sir,
> The MT6397 is an integration of several ICs and does not have a
> separate IC specification. I will check the relevant IC datasheets
> again. I will remove the useless define in next version.


I understand nothing from this reply. I do not see it related to my
questions at all. Do you want to say that you change indentation,
because MT6397 is integration? That makes absolutely NO SENSE!

Best regards,
Krzysztof

Re: [PATCH v1 1/2] mfd: mt6397: Add new bit definitions for RTC_BBPU register

[Shunxi Zhang (章顺喜)]

On Tue, 2025-08-19 at 10:16 +0200, Krzysztof Kozlowski wrote:
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
> 
> 
> On 19/08/2025 09:54, Shunxi Zhang (章顺喜) wrote:
> > > >  #define RTC_BBPU               0x0000
> > > > +#define RTC_BBPU_PWREN         BIT(0)
> > > > +#define RTC_BBPU_CLR           BIT(1)
> > > > +#define RTC_BBPU_RESET_AL      BIT(3)
> > > >  #define RTC_BBPU_CBUSY         BIT(6)
> > > > -#define RTC_BBPU_KEY            (0x43 << 8)
> > > > +#define RTC_BBPU_KEY           (0x43 << 8)
> > > 
> > > 
> > > Why?
> > > 
> > > 
> > > 
> > > Best regards,
> > > Krzysztof
> > 
> > Dear sir,
> > The MT6397 is an integration of several ICs and does not have a
> > separate IC specification. I will check the relevant IC datasheets
> > again. I will remove the useless define in next version.
> 
> 
> I understand nothing from this reply. I do not see it related to my
> questions at all. Do you want to say that you change indentation,
> because MT6397 is integration? That makes absolutely NO SENSE!
> 
> Best regards,
> Krzysztof

Dear sir,
   I will submit a separate patch for the format modification. The
reply to the previous email was not about this format issue. I am very
sorry for causing your misunderstanding.

Best Regards
Shunxi Zhang

[PATCH v1 2/2] rtc: mt6397: Add BBPU alarm status reset and shutdown handling

[ot_shunxi.zhang]

From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>

This patch introduces a new function, mtk_rtc_reset_bbpu_alarm_status,
to reset the BBPU alarm status in the MT6397 RTC driver. This function
writes the necessary bits to the RTC_BBPU register to clear the alarm
status and ensure proper operation.

Additionally, the mtk_rtc_shutdown function is added to handle RTC
shutdown events. It resets the BBPU alarm status and updates the
RTC_IRQ_EN register to disable the one-shot alarm interrupt,
ensuring a clean shutdown process.

Signed-off-by: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
---
 drivers/rtc/rtc-mt6397.c | 36 +++++++++++++++++++++++++++++++++++-
 1 file changed, 35 insertions(+), 1 deletion(-)

diff --git a/drivers/rtc/rtc-mt6397.c b/drivers/rtc/rtc-mt6397.c
index 692c00ff544b..063bd399de8c 100644
--- a/drivers/rtc/rtc-mt6397.c
+++ b/drivers/rtc/rtc-mt6397.c
@@ -37,6 +37,21 @@ static int mtk_rtc_write_trigger(struct mt6397_rtc *rtc)
 	return ret;
 }
 
+static void mtk_rtc_reset_bbpu_alarm_status(struct mt6397_rtc *rtc)
+{
+	u32 bbpu = RTC_BBPU_KEY | RTC_BBPU_PWREN | RTC_BBPU_RESET_AL;
+	int ret;
+
+	ret = regmap_write(rtc->regmap, rtc->addr_base + RTC_BBPU, bbpu);
+	if (ret < 0) {
+		dev_err(rtc->rtc_dev->dev.parent, "%s: write rtc bbpu error\n",
+			__func__);
+		return;
+	}
+
+	mtk_rtc_write_trigger(rtc);
+}
+
 static irqreturn_t mtk_rtc_irq_handler_thread(int irq, void *data)
 {
 	struct mt6397_rtc *rtc = data;
@@ -51,6 +66,8 @@ static irqreturn_t mtk_rtc_irq_handler_thread(int irq, void *data)
 		if (regmap_write(rtc->regmap, rtc->addr_base + RTC_IRQ_EN,
 				 irqen) == 0)
 			mtk_rtc_write_trigger(rtc);
+
+		mtk_rtc_reset_bbpu_alarm_status(rtc);
 		mutex_unlock(&rtc->lock);
 
 		return IRQ_HANDLED;
@@ -297,6 +314,22 @@ static int mtk_rtc_probe(struct platform_device *pdev)
 	return devm_rtc_register_device(rtc->rtc_dev);
 }
 
+static void mtk_rtc_shutdown(struct platform_device *pdev)
+{
+	struct mt6397_rtc *rtc = platform_get_drvdata(pdev);
+	int ret = 0;
+
+	mtk_rtc_reset_bbpu_alarm_status(rtc);
+
+	ret = regmap_update_bits(rtc->regmap,
+				 rtc->addr_base + RTC_IRQ_EN,
+				 RTC_IRQ_EN_ONESHOT_AL, 0);
+	if (ret < 0)
+		return;
+
+	mtk_rtc_write_trigger(rtc);
+}
+
 #ifdef CONFIG_PM_SLEEP
 static int mt6397_rtc_suspend(struct device *dev)
 {
@@ -345,7 +378,8 @@ static struct platform_driver mtk_rtc_driver = {
 		.of_match_table = mt6397_rtc_of_match,
 		.pm = &mt6397_pm_ops,
 	},
-	.probe	= mtk_rtc_probe,
+	.probe = mtk_rtc_probe,
+	.shutdown = mtk_rtc_shutdown,
 };
 
 module_platform_driver(mtk_rtc_driver);
-- 
2.46.0

Re: [PATCH v1 2/2] rtc: mt6397: Add BBPU alarm status reset and shutdown handling

[Alexandre Belloni]

On 11/08/2025 16:15:34+0800, ot_shunxi.zhang@mediatek.com wrote:
> From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> 
> This patch introduces a new function, mtk_rtc_reset_bbpu_alarm_status,
> to reset the BBPU alarm status in the MT6397 RTC driver. This function
> writes the necessary bits to the RTC_BBPU register to clear the alarm
> status and ensure proper operation.
> 
> Additionally, the mtk_rtc_shutdown function is added to handle RTC
> shutdown events. It resets the BBPU alarm status and updates the
> RTC_IRQ_EN register to disable the one-shot alarm interrupt,
> ensuring a clean shutdown process.
> 
> Signed-off-by: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> ---
>  drivers/rtc/rtc-mt6397.c | 36 +++++++++++++++++++++++++++++++++++-
>  1 file changed, 35 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/rtc/rtc-mt6397.c b/drivers/rtc/rtc-mt6397.c
> index 692c00ff544b..063bd399de8c 100644
> --- a/drivers/rtc/rtc-mt6397.c
> +++ b/drivers/rtc/rtc-mt6397.c
> @@ -37,6 +37,21 @@ static int mtk_rtc_write_trigger(struct mt6397_rtc *rtc)
>  	return ret;
>  }
>  
> +static void mtk_rtc_reset_bbpu_alarm_status(struct mt6397_rtc *rtc)
> +{
> +	u32 bbpu = RTC_BBPU_KEY | RTC_BBPU_PWREN | RTC_BBPU_RESET_AL;
> +	int ret;
> +
> +	ret = regmap_write(rtc->regmap, rtc->addr_base + RTC_BBPU, bbpu);
> +	if (ret < 0) {
> +		dev_err(rtc->rtc_dev->dev.parent, "%s: write rtc bbpu error\n",
> +			__func__);
> +		return;
> +	}
> +
> +	mtk_rtc_write_trigger(rtc);
> +}
> +
>  static irqreturn_t mtk_rtc_irq_handler_thread(int irq, void *data)
>  {
>  	struct mt6397_rtc *rtc = data;
> @@ -51,6 +66,8 @@ static irqreturn_t mtk_rtc_irq_handler_thread(int irq, void *data)
>  		if (regmap_write(rtc->regmap, rtc->addr_base + RTC_IRQ_EN,
>  				 irqen) == 0)
>  			mtk_rtc_write_trigger(rtc);
> +
> +		mtk_rtc_reset_bbpu_alarm_status(rtc);
>  		mutex_unlock(&rtc->lock);
>  
>  		return IRQ_HANDLED;
> @@ -297,6 +314,22 @@ static int mtk_rtc_probe(struct platform_device *pdev)
>  	return devm_rtc_register_device(rtc->rtc_dev);
>  }
>  
> +static void mtk_rtc_shutdown(struct platform_device *pdev)
> +{
> +	struct mt6397_rtc *rtc = platform_get_drvdata(pdev);
> +	int ret = 0;
> +
> +	mtk_rtc_reset_bbpu_alarm_status(rtc);
> +
> +	ret = regmap_update_bits(rtc->regmap,
> +				 rtc->addr_base + RTC_IRQ_EN,
> +				 RTC_IRQ_EN_ONESHOT_AL, 0);
> +	if (ret < 0)
> +		return;
> +
> +	mtk_rtc_write_trigger(rtc);

The whole goal of the RTC is to wakeup the system, why would you disable
the alarm on shutdown?

> +}
> +
>  #ifdef CONFIG_PM_SLEEP
>  static int mt6397_rtc_suspend(struct device *dev)
>  {
> @@ -345,7 +378,8 @@ static struct platform_driver mtk_rtc_driver = {
>  		.of_match_table = mt6397_rtc_of_match,
>  		.pm = &mt6397_pm_ops,
>  	},
> -	.probe	= mtk_rtc_probe,
> +	.probe = mtk_rtc_probe,
> +	.shutdown = mtk_rtc_shutdown,
>  };
>  
>  module_platform_driver(mtk_rtc_driver);
> -- 
> 2.46.0
> 

-- 
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

Re: [PATCH v1 2/2] rtc: mt6397: Add BBPU alarm status reset and shutdown handling

[Shunxi Zhang (章顺喜)]

On Mon, 2025-08-11 at 15:09 +0200, Alexandre Belloni wrote:
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
> 
> 
> On 11/08/2025 16:15:34+0800, ot_shunxi.zhang@mediatek.com wrote:
> > From: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> > 
> > This patch introduces a new function,
> > mtk_rtc_reset_bbpu_alarm_status,
> > to reset the BBPU alarm status in the MT6397 RTC driver. This
> > function
> > writes the necessary bits to the RTC_BBPU register to clear the
> > alarm
> > status and ensure proper operation.
> > 
> > Additionally, the mtk_rtc_shutdown function is added to handle RTC
> > shutdown events. It resets the BBPU alarm status and updates the
> > RTC_IRQ_EN register to disable the one-shot alarm interrupt,
> > ensuring a clean shutdown process.
> > 
> > Signed-off-by: Shunxi Zhang <ot_shunxi.zhang@mediatek.com>
> > ---
> >  drivers/rtc/rtc-mt6397.c | 36 +++++++++++++++++++++++++++++++++++-
> >  1 file changed, 35 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/rtc/rtc-mt6397.c b/drivers/rtc/rtc-mt6397.c
> > index 692c00ff544b..063bd399de8c 100644
> > --- a/drivers/rtc/rtc-mt6397.c
> > +++ b/drivers/rtc/rtc-mt6397.c
> > @@ -37,6 +37,21 @@ static int mtk_rtc_write_trigger(struct
> > mt6397_rtc *rtc)
> >       return ret;
> >  }
> > 
> > +static void mtk_rtc_reset_bbpu_alarm_status(struct mt6397_rtc
> > *rtc)
> > +{
> > +     u32 bbpu = RTC_BBPU_KEY | RTC_BBPU_PWREN | RTC_BBPU_RESET_AL;
> > +     int ret;
> > +
> > +     ret = regmap_write(rtc->regmap, rtc->addr_base + RTC_BBPU,
> > bbpu);
> > +     if (ret < 0) {
> > +             dev_err(rtc->rtc_dev->dev.parent, "%s: write rtc bbpu
> > error\n",
> > +                     __func__);
> > +             return;
> > +     }
> > +
> > +     mtk_rtc_write_trigger(rtc);
> > +}
> > +
> >  static irqreturn_t mtk_rtc_irq_handler_thread(int irq, void *data)
> >  {
> >       struct mt6397_rtc *rtc = data;
> > @@ -51,6 +66,8 @@ static irqreturn_t mtk_rtc_irq_handler_thread(int
> > irq, void *data)
> >               if (regmap_write(rtc->regmap, rtc->addr_base +
> > RTC_IRQ_EN,
> >                                irqen) == 0)
> >                       mtk_rtc_write_trigger(rtc);
> > +
> > +             mtk_rtc_reset_bbpu_alarm_status(rtc);
> >               mutex_unlock(&rtc->lock);
> > 
> >               return IRQ_HANDLED;
> > @@ -297,6 +314,22 @@ static int mtk_rtc_probe(struct
> > platform_device *pdev)
> >       return devm_rtc_register_device(rtc->rtc_dev);
> >  }
> > 
> > +static void mtk_rtc_shutdown(struct platform_device *pdev)
> > +{
> > +     struct mt6397_rtc *rtc = platform_get_drvdata(pdev);
> > +     int ret = 0;
> > +
> > +     mtk_rtc_reset_bbpu_alarm_status(rtc);
> > +
> > +     ret = regmap_update_bits(rtc->regmap,
> > +                              rtc->addr_base + RTC_IRQ_EN,
> > +                              RTC_IRQ_EN_ONESHOT_AL, 0);
> > +     if (ret < 0)
> > +             return;
> > +
> > +     mtk_rtc_write_trigger(rtc);
> 
> The whole goal of the RTC is to wakeup the system, why would you
> disable
> the alarm on shutdown?

Dear sir,
  I will remove the flow of "disable alarm shutdown"
in next version. thanks for your comments.  

Best regards
shunxi zhang

> 
> > +}
> > +
> >  #ifdef CONFIG_PM_SLEEP
> >  static int mt6397_rtc_suspend(struct device *dev)
> >  {
> > @@ -345,7 +378,8 @@ static struct platform_driver mtk_rtc_driver =
> > {
> >               .of_match_table = mt6397_rtc_of_match,
> >               .pm = &mt6397_pm_ops,
> >       },
> > -     .probe  = mtk_rtc_probe,
> > +     .probe = mtk_rtc_probe,
> > +     .shutdown = mtk_rtc_shutdown,
> >  };
> > 
> >  module_platform_driver(mtk_rtc_driver);
> > --
> > 2.46.0
> > 
> 
> --
> Alexandre Belloni, co-owner and COO, Bootlin
> Embedded Linux and Kernel engineering
> 
https://urldefense.com/v3/__https://bootlin.com__;!!CTRNKA9wMg0ARbw!hGMitEEUU-sbH-rcMCMQ4Vlsn7NQeHdUrg9nEKYPyDME5fkbgsVciZd8SURxkvWA9Z1qX3oyNVZEWvVatYExZ8D7cFvdiyS_3Q_6$