2016-04-08 06:59:02

by Henry Chen

Subject: [PATCH v2 1/3] mfd: mt6397: irq domain should initialize before mfd_add_devices()

Some sub driver like RTC module need irq domain from parent to create
irq mapping when driver initialize. so move mt6397_irq_init() before
mfd_add_devices().

Acked-by: John Crispin <[email protected]>
Signed-off-by: Henry Chen <[email protected]>
---
This series fixed the below warning based on "Linux kernel v4.6-rc1"
WARNING: CPU: 1 PID: 132 at kernel/mediatek/kernel/irq/irqdomain.c:471
irq_create_mapping+0xc4/0xd0

Changes in V2: Add two patch for error handle checking.
---
drivers/mfd/mt6397-core.c | 37 ++++++++++++++++++++++---------------
1 file changed, 22 insertions(+), 15 deletions(-)

diff --git a/drivers/mfd/mt6397-core.c b/drivers/mfd/mt6397-core.c
index 8e8d932..a879223 100644
--- a/drivers/mfd/mt6397-core.c
+++ b/drivers/mfd/mt6397-core.c
@@ -270,22 +270,36 @@ static int mt6397_probe(struct platform_device *pdev)
goto fail_irq;
}

+ pmic->irq = platform_get_irq(pdev, 0);
+
switch (id & 0xff) {
case MT6323_CID_CODE:
- pmic->int_con[0] = MT6323_INT_CON0;
- pmic->int_con[1] = MT6323_INT_CON1;
- pmic->int_status[0] = MT6323_INT_STATUS0;
- pmic->int_status[1] = MT6323_INT_STATUS1;
+ if (pmic->irq > 0) {
+ pmic->int_con[0] = MT6323_INT_CON0;
+ pmic->int_con[1] = MT6323_INT_CON1;
+ pmic->int_status[0] = MT6323_INT_STATUS0;
+ pmic->int_status[1] = MT6323_INT_STATUS1;
+ ret = mt6397_irq_init(pmic);
+ if (ret)
+ return ret;
+ }
+
ret = mfd_add_devices(&pdev->dev, -1, mt6323_devs,
ARRAY_SIZE(mt6323_devs), NULL, 0, NULL);
break;

case MT6397_CID_CODE:
case MT6391_CID_CODE:
- pmic->int_con[0] = MT6397_INT_CON0;
- pmic->int_con[1] = MT6397_INT_CON1;
- pmic->int_status[0] = MT6397_INT_STATUS0;
- pmic->int_status[1] = MT6397_INT_STATUS1;
+ if (pmic->irq > 0) {
+ pmic->int_con[0] = MT6397_INT_CON0;
+ pmic->int_con[1] = MT6397_INT_CON1;
+ pmic->int_status[0] = MT6397_INT_STATUS0;
+ pmic->int_status[1] = MT6397_INT_STATUS1;
+ ret = mt6397_irq_init(pmic);
+ if (ret)
+ return ret;
+ }
+
ret = mfd_add_devices(&pdev->dev, -1, mt6397_devs,
ARRAY_SIZE(mt6397_devs), NULL, 0, NULL);
break;
@@ -296,13 +310,6 @@ static int mt6397_probe(struct platform_device *pdev)
break;
}

- pmic->irq = platform_get_irq(pdev, 0);
- if (pmic->irq > 0) {
- ret = mt6397_irq_init(pmic);
- if (ret)
- return ret;
- }
-
fail_irq:
if (ret) {
irq_domain_remove(pmic->irq_domain);
--
1.8.1.1.dirty

2016-04-08 06:59:58

by Henry Chen

Subject: [PATCH v2 2/3] mfd: mt6397: Checking for null before irq_domain_remove.

It is possible that pmic->irq_domain will be NULL in fail_irq error handling.
Check before calling irq_domain_remove.

Signed-off-by: Henry Chen <[email protected]>
---
drivers/mfd/mt6397-core.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/mfd/mt6397-core.c b/drivers/mfd/mt6397-core.c
index a879223..15050cb 100644
--- a/drivers/mfd/mt6397-core.c
+++ b/drivers/mfd/mt6397-core.c
@@ -312,7 +312,8 @@ static int mt6397_probe(struct platform_device *pdev)

fail_irq:
if (ret) {
- irq_domain_remove(pmic->irq_domain);
+ if (pmic->irq_domain)
+ irq_domain_remove(pmic->irq_domain);
dev_err(&pdev->dev, "failed to add child devices: %d\n", ret);
}

--
1.8.1.1.dirty

2016-04-08 06:58:59

by Henry Chen

Subject: [PATCH v2 3/3] mfd: mt6397: check the EPROBE_DEFER from platform_get_irq.

If get the -EPROBE_DEFER from platform_get_irq, it should return back directly.

Signed-off-by: Henry Chen <[email protected]>
---
drivers/mfd/mt6397-core.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/drivers/mfd/mt6397-core.c b/drivers/mfd/mt6397-core.c
index 15050cb..b9b1e28 100644
--- a/drivers/mfd/mt6397-core.c
+++ b/drivers/mfd/mt6397-core.c
@@ -271,6 +271,8 @@ static int mt6397_probe(struct platform_device *pdev)
}

pmic->irq = platform_get_irq(pdev, 0);
+ if (pmic->irq == -EPROBE_DEFER)
+ return -EPROBE_DEFER;

switch (id & 0xff) {
case MT6323_CID_CODE:
--
1.8.1.1.dirty

2016-04-11 08:50:45

by Lee Jones

Subject: Re: [PATCH v2 3/3] mfd: mt6397: check the EPROBE_DEFER from platform_get_irq.

On Fri, 08 Apr 2016, Henry Chen wrote:

> If get the -EPROBE_DEFER from platform_get_irq, it should return back directly.
>
> Signed-off-by: Henry Chen <[email protected]>
> ---
> drivers/mfd/mt6397-core.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/mfd/mt6397-core.c b/drivers/mfd/mt6397-core.c
> index 15050cb..b9b1e28 100644
> --- a/drivers/mfd/mt6397-core.c
> +++ b/drivers/mfd/mt6397-core.c
> @@ -271,6 +271,8 @@ static int mt6397_probe(struct platform_device *pdev)
> }
>
> pmic->irq = platform_get_irq(pdev, 0);
> + if (pmic->irq == -EPROBE_DEFER)
> + return -EPROBE_DEFER;

if (ret < 0)
return ret;

... is more common, as it covers more bases, etc.

> switch (id & 0xff) {
> case MT6323_CID_CODE:

--
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog

2016-04-11 08:53:29

by Lee Jones

Subject: Re: [PATCH v2 2/3] mfd: mt6397: Checking for null before irq_domain_remove.

FAO Thomas,

On Fri, 08 Apr 2016, Henry Chen wrote:
> It is possible that pmic->irq_domain will be NULL in fail_irq error handling.
> Check before calling irq_domain_remove.
>
> Signed-off-by: Henry Chen <[email protected]>
> ---
> drivers/mfd/mt6397-core.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/mfd/mt6397-core.c b/drivers/mfd/mt6397-core.c
> index a879223..15050cb 100644
> --- a/drivers/mfd/mt6397-core.c
> +++ b/drivers/mfd/mt6397-core.c
> @@ -312,7 +312,8 @@ static int mt6397_probe(struct platform_device *pdev)
>
> fail_irq:
> if (ret) {
> - irq_domain_remove(pmic->irq_domain);
> + if (pmic->irq_domain)
> + irq_domain_remove(pmic->irq_domain);

You might be better off making this change in the framework.

Just get irq_domain_remove() to harmlessly return if !irq_domain.

> dev_err(&pdev->dev, "failed to add child devices: %d\n", ret);
> }
>

--
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog

2016-04-11 08:57:06

by Lee Jones

Subject: Re: [PATCH v2 1/3] mfd: mt6397: irq domain should initialize before mfd_add_devices()

On Fri, 08 Apr 2016, Henry Chen wrote:

> Some sub driver like RTC module need irq domain from parent to create
> irq mapping when driver initialize. so move mt6397_irq_init() before
> mfd_add_devices().
>
> Acked-by: John Crispin <[email protected]>
> Signed-off-by: Henry Chen <[email protected]>
> ---
> This series fixed the below warning based on "Linux kernel v4.6-rc1"
> WARNING: CPU: 1 PID: 132 at kernel/mediatek/kernel/irq/irqdomain.c:471
> irq_create_mapping+0xc4/0xd0
>
> Changes in V2: Add two patch for error handle checking.
> ---
> drivers/mfd/mt6397-core.c | 37 ++++++++++++++++++++++---------------
> 1 file changed, 22 insertions(+), 15 deletions(-)
>
> diff --git a/drivers/mfd/mt6397-core.c b/drivers/mfd/mt6397-core.c
> index 8e8d932..a879223 100644
> --- a/drivers/mfd/mt6397-core.c
> +++ b/drivers/mfd/mt6397-core.c
> @@ -270,22 +270,36 @@ static int mt6397_probe(struct platform_device *pdev)
> goto fail_irq;
> }
>
> + pmic->irq = platform_get_irq(pdev, 0);
> +
> switch (id & 0xff) {
> case MT6323_CID_CODE:
> - pmic->int_con[0] = MT6323_INT_CON0;
> - pmic->int_con[1] = MT6323_INT_CON1;
> - pmic->int_status[0] = MT6323_INT_STATUS0;
> - pmic->int_status[1] = MT6323_INT_STATUS1;
> + if (pmic->irq > 0) {
> + pmic->int_con[0] = MT6323_INT_CON0;
> + pmic->int_con[1] = MT6323_INT_CON1;
> + pmic->int_status[0] = MT6323_INT_STATUS0;
> + pmic->int_status[1] = MT6323_INT_STATUS1;
> + ret = mt6397_irq_init(pmic);
> + if (ret)
> + return ret;
> + }
> +

I don't think you need all these ugly indenting.

Just harmlessly return in mt6397_irq_init() if there are no IRQs
present.

> ret = mfd_add_devices(&pdev->dev, -1, mt6323_devs,
> ARRAY_SIZE(mt6323_devs), NULL, 0, NULL);
> break;
>
> case MT6397_CID_CODE:
> case MT6391_CID_CODE:
> - pmic->int_con[0] = MT6397_INT_CON0;
> - pmic->int_con[1] = MT6397_INT_CON1;
> - pmic->int_status[0] = MT6397_INT_STATUS0;
> - pmic->int_status[1] = MT6397_INT_STATUS1;
> + if (pmic->irq > 0) {
> + pmic->int_con[0] = MT6397_INT_CON0;
> + pmic->int_con[1] = MT6397_INT_CON1;
> + pmic->int_status[0] = MT6397_INT_STATUS0;
> + pmic->int_status[1] = MT6397_INT_STATUS1;
> + ret = mt6397_irq_init(pmic);
> + if (ret)
> + return ret;
> + }
> +
> ret = mfd_add_devices(&pdev->dev, -1, mt6397_devs,
> ARRAY_SIZE(mt6397_devs), NULL, 0, NULL);
> break;
> @@ -296,13 +310,6 @@ static int mt6397_probe(struct platform_device *pdev)
> break;
> }
>
> - pmic->irq = platform_get_irq(pdev, 0);
> - if (pmic->irq > 0) {
> - ret = mt6397_irq_init(pmic);
> - if (ret)
> - return ret;
> - }
> -
> fail_irq:
> if (ret) {
> irq_domain_remove(pmic->irq_domain);

--
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog

2016-04-13 05:48:30

by Henry Chen

Subject: Re: [PATCH v2 3/3] mfd: mt6397: check the EPROBE_DEFER from platform_get_irq.

On Mon, 2016-04-11 at 09:50 +0100, Lee Jones wrote:
> On Fri, 08 Apr 2016, Henry Chen wrote:
>
> > If get the -EPROBE_DEFER from platform_get_irq, it should return back directly.
> >
> > Signed-off-by: Henry Chen <[email protected]>
> > ---
> > drivers/mfd/mt6397-core.c | 2 ++
> > 1 file changed, 2 insertions(+)
> >
> > diff --git a/drivers/mfd/mt6397-core.c b/drivers/mfd/mt6397-core.c
> > index 15050cb..b9b1e28 100644
> > --- a/drivers/mfd/mt6397-core.c
> > +++ b/drivers/mfd/mt6397-core.c
> > @@ -271,6 +271,8 @@ static int mt6397_probe(struct platform_device *pdev)
> > }
> >
> > pmic->irq = platform_get_irq(pdev, 0);
> > + if (pmic->irq == -EPROBE_DEFER)
> > + return -EPROBE_DEFER;
>
> if (ret < 0)
> return ret;
>
> ... is more common, as it covers more bases, etc.

Okay, the interrupt should be a required properties, if get failed just
return back.

I will also change the binding doc.

Thanks.

>
> > switch (id & 0xff) {
> > case MT6323_CID_CODE:
>