2020-12-07 11:41:05

by Bongsu Jeon

[permalink] [raw]
Subject: [PATCH net-next] nfc: s3fwrn5: Change irqflags

From: Bongsu Jeon <[email protected]>

change irqflags from IRQF_TRIGGER_HIGH to IRQF_TRIGGER_RISING for stable
Samsung's nfc interrupt handling.

Signed-off-by: Bongsu Jeon <[email protected]>
---
drivers/nfc/s3fwrn5/i2c.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/nfc/s3fwrn5/i2c.c b/drivers/nfc/s3fwrn5/i2c.c
index e1bdde105f24..016f6b6df849 100644
--- a/drivers/nfc/s3fwrn5/i2c.c
+++ b/drivers/nfc/s3fwrn5/i2c.c
@@ -213,7 +213,7 @@ static int s3fwrn5_i2c_probe(struct i2c_client *client,
return ret;

ret = devm_request_threaded_irq(&client->dev, phy->i2c_dev->irq, NULL,
- s3fwrn5_i2c_irq_thread_fn, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+ s3fwrn5_i2c_irq_thread_fn, IRQF_TRIGGER_RISING | IRQF_ONESHOT,
S3FWRN5_I2C_DRIVER_NAME, phy);
if (ret)
s3fwrn5_remove(phy->common.ndev);
--
2.17.1


2020-12-07 11:55:53

by Krzysztof Kozlowski

[permalink] [raw]
Subject: Re: [PATCH net-next] nfc: s3fwrn5: Change irqflags

On Mon, Dec 07, 2020 at 08:38:27PM +0900, Bongsu Jeon wrote:
> From: Bongsu Jeon <[email protected]>
>
> change irqflags from IRQF_TRIGGER_HIGH to IRQF_TRIGGER_RISING for stable
> Samsung's nfc interrupt handling.

1. Describe in commit title/subject the change. Just a word "change irqflags" is
not enough.

2. Describe in commit message what you are trying to fix. Before was not
stable? The "for stable interrupt handling" is a little bit vauge.

3. This is contradictory to the bindings and current DTS. I think the
driver should not force the specific trigger type because I could
imagine some configuration that the actual interrupt to the CPU is
routed differently.

Instead, how about removing the trigger flags here and fixing the DTS
and bindings example?

Best regards,
Krzysztof

>
> Signed-off-by: Bongsu Jeon <[email protected]>
> ---
> drivers/nfc/s3fwrn5/i2c.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/nfc/s3fwrn5/i2c.c b/drivers/nfc/s3fwrn5/i2c.c
> index e1bdde105f24..016f6b6df849 100644
> --- a/drivers/nfc/s3fwrn5/i2c.c
> +++ b/drivers/nfc/s3fwrn5/i2c.c
> @@ -213,7 +213,7 @@ static int s3fwrn5_i2c_probe(struct i2c_client *client,
> return ret;
>
> ret = devm_request_threaded_irq(&client->dev, phy->i2c_dev->irq, NULL,
> - s3fwrn5_i2c_irq_thread_fn, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
> + s3fwrn5_i2c_irq_thread_fn, IRQF_TRIGGER_RISING | IRQF_ONESHOT,
> S3FWRN5_I2C_DRIVER_NAME, phy);
> if (ret)
> s3fwrn5_remove(phy->common.ndev);
> --
> 2.17.1
>

2020-12-07 13:42:51

by Bongsu Jeon

[permalink] [raw]
Subject: Re: [PATCH net-next] nfc: s3fwrn5: Change irqflags

On Mon, Dec 7, 2020 at 8:51 PM Krzysztof Kozlowski <[email protected]> wrote:
>
> On Mon, Dec 07, 2020 at 08:38:27PM +0900, Bongsu Jeon wrote:
> > From: Bongsu Jeon <[email protected]>
> >
> > change irqflags from IRQF_TRIGGER_HIGH to IRQF_TRIGGER_RISING for stable
> > Samsung's nfc interrupt handling.
>
> 1. Describe in commit title/subject the change. Just a word "change irqflags" is
> not enough.
>
Ok. I'll update it.

> 2. Describe in commit message what you are trying to fix. Before was not
> stable? The "for stable interrupt handling" is a little bit vauge.
>
Usually, Samsung's NFC Firmware sends an i2c frame as below.

1. NFC Firmware sets the gpio(interrupt pin) high when there is an i2c
frame to send.
2. If the CPU's I2C master has received the i2c frame, NFC F/W sets
the gpio low.

NFC driver's i2c interrupt handler would be called in the abnormal case
as the NFC F/W task of number 2 is delayed because of other high
priority tasks.
In that case, NFC driver will try to receive the i2c frame but there
isn't any i2c frame
to send in NFC. It would cause an I2C communication problem.
This case would hardly happen.
But, I changed the interrupt as a defense code.
If Driver uses the TRIGGER_RISING not LEVEL trigger, there would be no problem
even if the NFC F/W task is delayed.

> 3. This is contradictory to the bindings and current DTS. I think the
> driver should not force the specific trigger type because I could
> imagine some configuration that the actual interrupt to the CPU is
> routed differently.
>
> Instead, how about removing the trigger flags here and fixing the DTS
> and bindings example?
>

As I mentioned before,
I changed this code because of Samsung NFC's I2C Communication way.
So, I think that it is okay for the nfc driver to force the specific
trigger type( EDGE_RISING).

What do you think about it?

> Best regards,
> Krzysztof
>
> >
> > Signed-off-by: Bongsu Jeon <[email protected]>
> > ---
> > drivers/nfc/s3fwrn5/i2c.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/nfc/s3fwrn5/i2c.c b/drivers/nfc/s3fwrn5/i2c.c
> > index e1bdde105f24..016f6b6df849 100644
> > --- a/drivers/nfc/s3fwrn5/i2c.c
> > +++ b/drivers/nfc/s3fwrn5/i2c.c
> > @@ -213,7 +213,7 @@ static int s3fwrn5_i2c_probe(struct i2c_client *client,
> > return ret;
> >
> > ret = devm_request_threaded_irq(&client->dev, phy->i2c_dev->irq, NULL,
> > - s3fwrn5_i2c_irq_thread_fn, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
> > + s3fwrn5_i2c_irq_thread_fn, IRQF_TRIGGER_RISING | IRQF_ONESHOT,
> > S3FWRN5_I2C_DRIVER_NAME, phy);
> > if (ret)
> > s3fwrn5_remove(phy->common.ndev);
> > --
> > 2.17.1
> >

2020-12-07 14:17:58

by Krzysztof Kozlowski

[permalink] [raw]
Subject: Re: [PATCH net-next] nfc: s3fwrn5: Change irqflags

On Mon, Dec 07, 2020 at 10:39:01PM +0900, Bongsu Jeon wrote:
> On Mon, Dec 7, 2020 at 8:51 PM Krzysztof Kozlowski <[email protected]> wrote:
> >
> > On Mon, Dec 07, 2020 at 08:38:27PM +0900, Bongsu Jeon wrote:
> > > From: Bongsu Jeon <[email protected]>
> > >
> > > change irqflags from IRQF_TRIGGER_HIGH to IRQF_TRIGGER_RISING for stable
> > > Samsung's nfc interrupt handling.
> >
> > 1. Describe in commit title/subject the change. Just a word "change irqflags" is
> > not enough.
> >
> Ok. I'll update it.
>
> > 2. Describe in commit message what you are trying to fix. Before was not
> > stable? The "for stable interrupt handling" is a little bit vauge.
> >
> Usually, Samsung's NFC Firmware sends an i2c frame as below.
>
> 1. NFC Firmware sets the gpio(interrupt pin) high when there is an i2c
> frame to send.
> 2. If the CPU's I2C master has received the i2c frame, NFC F/W sets
> the gpio low.
>
> NFC driver's i2c interrupt handler would be called in the abnormal case
> as the NFC F/W task of number 2 is delayed because of other high
> priority tasks.
> In that case, NFC driver will try to receive the i2c frame but there
> isn't any i2c frame
> to send in NFC. It would cause an I2C communication problem.
> This case would hardly happen.
> But, I changed the interrupt as a defense code.
> If Driver uses the TRIGGER_RISING not LEVEL trigger, there would be no problem
> even if the NFC F/W task is delayed.

All this should be explained in commit message, not in the email.

>
> > 3. This is contradictory to the bindings and current DTS. I think the
> > driver should not force the specific trigger type because I could
> > imagine some configuration that the actual interrupt to the CPU is
> > routed differently.
> >
> > Instead, how about removing the trigger flags here and fixing the DTS
> > and bindings example?
> >
>
> As I mentioned before,
> I changed this code because of Samsung NFC's I2C Communication way.
> So, I think that it is okay for the nfc driver to force the specific
> trigger type( EDGE_RISING).
>
> What do you think about it?

Some different chip or some different hardware implementation could have
the signal inverted, e.g. edge falling, not rising. This is rather
a theoretical scenario but still such change makes the code more
generic, configurable with DTS. Therefore trigger mode should be
configured via DTS, not enforced by the driver.

Best regards,
Krzysztof

2020-12-07 14:30:28

by Bongsu Jeon

[permalink] [raw]
Subject: Re: [PATCH net-next] nfc: s3fwrn5: Change irqflags

On Mon, Dec 7, 2020 at 11:13 PM Krzysztof Kozlowski <[email protected]> wrote:
>
> On Mon, Dec 07, 2020 at 10:39:01PM +0900, Bongsu Jeon wrote:
> > On Mon, Dec 7, 2020 at 8:51 PM Krzysztof Kozlowski <[email protected]> wrote:
> > >
> > > On Mon, Dec 07, 2020 at 08:38:27PM +0900, Bongsu Jeon wrote:
> > > > From: Bongsu Jeon <[email protected]>
> > > >
> > > > change irqflags from IRQF_TRIGGER_HIGH to IRQF_TRIGGER_RISING for stable
> > > > Samsung's nfc interrupt handling.
> > >
> > > 1. Describe in commit title/subject the change. Just a word "change irqflags" is
> > > not enough.
> > >
> > Ok. I'll update it.
> >
> > > 2. Describe in commit message what you are trying to fix. Before was not
> > > stable? The "for stable interrupt handling" is a little bit vauge.
> > >
> > Usually, Samsung's NFC Firmware sends an i2c frame as below.
> >
> > 1. NFC Firmware sets the gpio(interrupt pin) high when there is an i2c
> > frame to send.
> > 2. If the CPU's I2C master has received the i2c frame, NFC F/W sets
> > the gpio low.
> >
> > NFC driver's i2c interrupt handler would be called in the abnormal case
> > as the NFC F/W task of number 2 is delayed because of other high
> > priority tasks.
> > In that case, NFC driver will try to receive the i2c frame but there
> > isn't any i2c frame
> > to send in NFC. It would cause an I2C communication problem.
> > This case would hardly happen.
> > But, I changed the interrupt as a defense code.
> > If Driver uses the TRIGGER_RISING not LEVEL trigger, there would be no problem
> > even if the NFC F/W task is delayed.
>
> All this should be explained in commit message, not in the email.
>
Okay. I will

> >
> > > 3. This is contradictory to the bindings and current DTS. I think the
> > > driver should not force the specific trigger type because I could
> > > imagine some configuration that the actual interrupt to the CPU is
> > > routed differently.
> > >
> > > Instead, how about removing the trigger flags here and fixing the DTS
> > > and bindings example?
> > >
> >
> > As I mentioned before,
> > I changed this code because of Samsung NFC's I2C Communication way.
> > So, I think that it is okay for the nfc driver to force the specific
> > trigger type( EDGE_RISING).
> >
> > What do you think about it?
>
> Some different chip or some different hardware implementation could have
> the signal inverted, e.g. edge falling, not rising. This is rather
> a theoretical scenario but still such change makes the code more
> generic, configurable with DTS. Therefore trigger mode should be
> configured via DTS, not enforced by the driver.
>
Okay. I understand it.

> Best regards,
> Krzysztof