Re: [greybus-dev] [PATCH v2] staging: greybus: arche: wrap over-length lines

[Joe Perches]

On Mon, 2017-07-10 at 15:01 +0530, Viresh Kumar wrote:
> On 10-07-17, 11:30, Frans Klaver wrote:
> > On Mon, Jul 10, 2017 at 6:46 AM, Viresh Kumar  
> > wrote:
> > > Hi Mitchell,
> > > 
> > > On 09-07-17, 20:25, Mitchell Tasman wrote:
> > > > Adjust formatting of various statements to keep line length within
> > > > the 80 column limit preferred by the Linux kernel coding style.
> > > 
> > > We try to follow that most of the time, but the end result should be 
> > > easily
> > > readable.  If it isn't, then we just ignore the rule.
> > > 
> > > > Signed-off-by: Mitchell Tasman 
> > > > ---
> > > > Changes in v2: Add back a missing space in a comment
> > > > 
> > > >  drivers/staging/greybus/arche-platform.c | 29 
> > > > +++--
> > > >  1 file changed, 19 insertions(+), 10 deletions(-)
> > > > 
> > > > diff --git a/drivers/staging/greybus/arche-platform.c 
> > > > b/drivers/staging/greybus/arche-platform.c
> > > > index eced2d2..eedd239 100644
> > > > --- a/drivers/staging/greybus/arche-platform.c
> > > > +++ b/drivers/staging/greybus/arche-platform.c
> > > > @@ -172,15 +172,21 @@ static irqreturn_t arche_platform_wd_irq(int irq, 
> > > > void *devid)
> > > >   if (arche_pdata->wake_detect_state == WD_STATE_BOOT_INIT) 
> > > > {
> > > >   if (time_before(jiffies,
> > > >   arche_pdata->wake_detect_start +
> > > > - 
> > > > msecs_to_jiffies(WD_COLDBOOT_PULSE_WIDTH_MS))) {
> > > > - 
> > > > arche_platform_set_wake_detect_state(arche_pdata,
> > > > -  
> > > > WD_STATE_IDLE);
> > > > + msecs_to_jiffies(
> > > > + 
> > > > WD_COLDBOOT_PULSE_WIDTH_MS))) {
> > > > + arche_platform_set_wake_detect_state(
> > > 
> > > We don't break the lines like this in kernel to satisfy the 80 column 
> > > width
> > > rule. Surely, some places would have similar code, but in general this 
> > > isn't
> > > recommended. And that's why we never bothered about 80 column rule in 
> > > this and
> > > below cases.
> > 
> > In cases like this, one could argue that you should start looking at
> > the level of indentation, rather than the line length per se. After
> > all, the documentation states that "if you need more than 3 levels of
> > indentation, you're screwed anyway, and you should fix your program".
> 
> And I completely agree to that :)

Another thing to consider is to s/arche_platform_//
as the identifier lengths are just way too long.

Re: [greybus-dev] [PATCH v2] staging: greybus: arche: wrap over-length lines

[Joe Perches]

On Mon, 2017-07-10 at 15:01 +0530, Viresh Kumar wrote:
> On 10-07-17, 11:30, Frans Klaver wrote:
> > On Mon, Jul 10, 2017 at 6:46 AM, Viresh Kumar  
> > wrote:
> > > Hi Mitchell,
> > > 
> > > On 09-07-17, 20:25, Mitchell Tasman wrote:
> > > > Adjust formatting of various statements to keep line length within
> > > > the 80 column limit preferred by the Linux kernel coding style.
> > > 
> > > We try to follow that most of the time, but the end result should be 
> > > easily
> > > readable.  If it isn't, then we just ignore the rule.
> > > 
> > > > Signed-off-by: Mitchell Tasman 
> > > > ---
> > > > Changes in v2: Add back a missing space in a comment
> > > > 
> > > >  drivers/staging/greybus/arche-platform.c | 29 
> > > > +++--
> > > >  1 file changed, 19 insertions(+), 10 deletions(-)
> > > > 
> > > > diff --git a/drivers/staging/greybus/arche-platform.c 
> > > > b/drivers/staging/greybus/arche-platform.c
> > > > index eced2d2..eedd239 100644
> > > > --- a/drivers/staging/greybus/arche-platform.c
> > > > +++ b/drivers/staging/greybus/arche-platform.c
> > > > @@ -172,15 +172,21 @@ static irqreturn_t arche_platform_wd_irq(int irq, 
> > > > void *devid)
> > > >   if (arche_pdata->wake_detect_state == WD_STATE_BOOT_INIT) 
> > > > {
> > > >   if (time_before(jiffies,
> > > >   arche_pdata->wake_detect_start +
> > > > - 
> > > > msecs_to_jiffies(WD_COLDBOOT_PULSE_WIDTH_MS))) {
> > > > - 
> > > > arche_platform_set_wake_detect_state(arche_pdata,
> > > > -  
> > > > WD_STATE_IDLE);
> > > > + msecs_to_jiffies(
> > > > + 
> > > > WD_COLDBOOT_PULSE_WIDTH_MS))) {
> > > > + arche_platform_set_wake_detect_state(
> > > 
> > > We don't break the lines like this in kernel to satisfy the 80 column 
> > > width
> > > rule. Surely, some places would have similar code, but in general this 
> > > isn't
> > > recommended. And that's why we never bothered about 80 column rule in 
> > > this and
> > > below cases.
> > 
> > In cases like this, one could argue that you should start looking at
> > the level of indentation, rather than the line length per se. After
> > all, the documentation states that "if you need more than 3 levels of
> > indentation, you're screwed anyway, and you should fix your program".
> 
> And I completely agree to that :)

Another thing to consider is to s/arche_platform_//
as the identifier lengths are just way too long.

Re: [greybus-dev] [PATCH v2] staging: greybus: arche: wrap over-length lines

[Viresh Kumar]

On 10-07-17, 11:30, Frans Klaver wrote:
> On Mon, Jul 10, 2017 at 6:46 AM, Viresh Kumar  wrote:
> > Hi Mitchell,
> >
> > On 09-07-17, 20:25, Mitchell Tasman wrote:
> >> Adjust formatting of various statements to keep line length within
> >> the 80 column limit preferred by the Linux kernel coding style.
> >
> > We try to follow that most of the time, but the end result should be easily
> > readable.  If it isn't, then we just ignore the rule.
> >
> >> Signed-off-by: Mitchell Tasman 
> >> ---
> >> Changes in v2: Add back a missing space in a comment
> >>
> >>  drivers/staging/greybus/arche-platform.c | 29 
> >> +++--
> >>  1 file changed, 19 insertions(+), 10 deletions(-)
> >>
> >> diff --git a/drivers/staging/greybus/arche-platform.c 
> >> b/drivers/staging/greybus/arche-platform.c
> >> index eced2d2..eedd239 100644
> >> --- a/drivers/staging/greybus/arche-platform.c
> >> +++ b/drivers/staging/greybus/arche-platform.c
> >> @@ -172,15 +172,21 @@ static irqreturn_t arche_platform_wd_irq(int irq, 
> >> void *devid)
> >>   if (arche_pdata->wake_detect_state == WD_STATE_BOOT_INIT) {
> >>   if (time_before(jiffies,
> >>   arche_pdata->wake_detect_start +
> >> - 
> >> msecs_to_jiffies(WD_COLDBOOT_PULSE_WIDTH_MS))) {
> >> - 
> >> arche_platform_set_wake_detect_state(arche_pdata,
> >> -  
> >> WD_STATE_IDLE);
> >> + msecs_to_jiffies(
> >> + 
> >> WD_COLDBOOT_PULSE_WIDTH_MS))) {
> >> + arche_platform_set_wake_detect_state(
> >
> > We don't break the lines like this in kernel to satisfy the 80 column width
> > rule. Surely, some places would have similar code, but in general this isn't
> > recommended. And that's why we never bothered about 80 column rule in this 
> > and
> > below cases.
> 
> In cases like this, one could argue that you should start looking at
> the level of indentation, rather than the line length per se. After
> all, the documentation states that "if you need more than 3 levels of
> indentation, you're screwed anyway, and you should fix your program".

And I completely agree to that :)

-- 
viresh

Re: [greybus-dev] [PATCH v2] staging: greybus: arche: wrap over-length lines

[Viresh Kumar]

On 10-07-17, 11:30, Frans Klaver wrote:
> On Mon, Jul 10, 2017 at 6:46 AM, Viresh Kumar  wrote:
> > Hi Mitchell,
> >
> > On 09-07-17, 20:25, Mitchell Tasman wrote:
> >> Adjust formatting of various statements to keep line length within
> >> the 80 column limit preferred by the Linux kernel coding style.
> >
> > We try to follow that most of the time, but the end result should be easily
> > readable.  If it isn't, then we just ignore the rule.
> >
> >> Signed-off-by: Mitchell Tasman 
> >> ---
> >> Changes in v2: Add back a missing space in a comment
> >>
> >>  drivers/staging/greybus/arche-platform.c | 29 
> >> +++--
> >>  1 file changed, 19 insertions(+), 10 deletions(-)
> >>
> >> diff --git a/drivers/staging/greybus/arche-platform.c 
> >> b/drivers/staging/greybus/arche-platform.c
> >> index eced2d2..eedd239 100644
> >> --- a/drivers/staging/greybus/arche-platform.c
> >> +++ b/drivers/staging/greybus/arche-platform.c
> >> @@ -172,15 +172,21 @@ static irqreturn_t arche_platform_wd_irq(int irq, 
> >> void *devid)
> >>   if (arche_pdata->wake_detect_state == WD_STATE_BOOT_INIT) {
> >>   if (time_before(jiffies,
> >>   arche_pdata->wake_detect_start +
> >> - 
> >> msecs_to_jiffies(WD_COLDBOOT_PULSE_WIDTH_MS))) {
> >> - 
> >> arche_platform_set_wake_detect_state(arche_pdata,
> >> -  
> >> WD_STATE_IDLE);
> >> + msecs_to_jiffies(
> >> + 
> >> WD_COLDBOOT_PULSE_WIDTH_MS))) {
> >> + arche_platform_set_wake_detect_state(
> >
> > We don't break the lines like this in kernel to satisfy the 80 column width
> > rule. Surely, some places would have similar code, but in general this isn't
> > recommended. And that's why we never bothered about 80 column rule in this 
> > and
> > below cases.
> 
> In cases like this, one could argue that you should start looking at
> the level of indentation, rather than the line length per se. After
> all, the documentation states that "if you need more than 3 levels of
> indentation, you're screwed anyway, and you should fix your program".

And I completely agree to that :)

-- 
viresh

Re: [greybus-dev] [PATCH v2] staging: greybus: arche: wrap over-length lines

[Frans Klaver]

On Mon, Jul 10, 2017 at 6:46 AM, Viresh Kumar  wrote:
> Hi Mitchell,
>
> On 09-07-17, 20:25, Mitchell Tasman wrote:
>> Adjust formatting of various statements to keep line length within
>> the 80 column limit preferred by the Linux kernel coding style.
>
> We try to follow that most of the time, but the end result should be easily
> readable.  If it isn't, then we just ignore the rule.
>
>> Signed-off-by: Mitchell Tasman 
>> ---
>> Changes in v2: Add back a missing space in a comment
>>
>>  drivers/staging/greybus/arche-platform.c | 29 +++--
>>  1 file changed, 19 insertions(+), 10 deletions(-)
>>
>> diff --git a/drivers/staging/greybus/arche-platform.c 
>> b/drivers/staging/greybus/arche-platform.c
>> index eced2d2..eedd239 100644
>> --- a/drivers/staging/greybus/arche-platform.c
>> +++ b/drivers/staging/greybus/arche-platform.c
>> @@ -172,15 +172,21 @@ static irqreturn_t arche_platform_wd_irq(int irq, void 
>> *devid)
>>   if (arche_pdata->wake_detect_state == WD_STATE_BOOT_INIT) {
>>   if (time_before(jiffies,
>>   arche_pdata->wake_detect_start +
>> - 
>> msecs_to_jiffies(WD_COLDBOOT_PULSE_WIDTH_MS))) {
>> - 
>> arche_platform_set_wake_detect_state(arche_pdata,
>> -  
>> WD_STATE_IDLE);
>> + msecs_to_jiffies(
>> + WD_COLDBOOT_PULSE_WIDTH_MS))) {
>> + arche_platform_set_wake_detect_state(
>
> We don't break the lines like this in kernel to satisfy the 80 column width
> rule. Surely, some places would have similar code, but in general this isn't
> recommended. And that's why we never bothered about 80 column rule in this and
> below cases.

In cases like this, one could argue that you should start looking at
the level of indentation, rather than the line length per se. After
all, the documentation states that "if you need more than 3 levels of
indentation, you're screwed anyway, and you should fix your program".

Re: [greybus-dev] [PATCH v2] staging: greybus: arche: wrap over-length lines

[Frans Klaver]

On Mon, Jul 10, 2017 at 6:46 AM, Viresh Kumar  wrote:
> Hi Mitchell,
>
> On 09-07-17, 20:25, Mitchell Tasman wrote:
>> Adjust formatting of various statements to keep line length within
>> the 80 column limit preferred by the Linux kernel coding style.
>
> We try to follow that most of the time, but the end result should be easily
> readable.  If it isn't, then we just ignore the rule.
>
>> Signed-off-by: Mitchell Tasman 
>> ---
>> Changes in v2: Add back a missing space in a comment
>>
>>  drivers/staging/greybus/arche-platform.c | 29 +++--
>>  1 file changed, 19 insertions(+), 10 deletions(-)
>>
>> diff --git a/drivers/staging/greybus/arche-platform.c 
>> b/drivers/staging/greybus/arche-platform.c
>> index eced2d2..eedd239 100644
>> --- a/drivers/staging/greybus/arche-platform.c
>> +++ b/drivers/staging/greybus/arche-platform.c
>> @@ -172,15 +172,21 @@ static irqreturn_t arche_platform_wd_irq(int irq, void 
>> *devid)
>>   if (arche_pdata->wake_detect_state == WD_STATE_BOOT_INIT) {
>>   if (time_before(jiffies,
>>   arche_pdata->wake_detect_start +
>> - 
>> msecs_to_jiffies(WD_COLDBOOT_PULSE_WIDTH_MS))) {
>> - 
>> arche_platform_set_wake_detect_state(arche_pdata,
>> -  
>> WD_STATE_IDLE);
>> + msecs_to_jiffies(
>> + WD_COLDBOOT_PULSE_WIDTH_MS))) {
>> + arche_platform_set_wake_detect_state(
>
> We don't break the lines like this in kernel to satisfy the 80 column width
> rule. Surely, some places would have similar code, but in general this isn't
> recommended. And that's why we never bothered about 80 column rule in this and
> below cases.

In cases like this, one could argue that you should start looking at
the level of indentation, rather than the line length per se. After
all, the documentation states that "if you need more than 3 levels of
indentation, you're screwed anyway, and you should fix your program".

Re: [greybus-dev] [PATCH v2] staging: greybus: arche: wrap over-length lines

[Viresh Kumar]

Hi Mitchell,

On 09-07-17, 20:25, Mitchell Tasman wrote:
> Adjust formatting of various statements to keep line length within
> the 80 column limit preferred by the Linux kernel coding style.

We try to follow that most of the time, but the end result should be easily
readable.  If it isn't, then we just ignore the rule.

> Signed-off-by: Mitchell Tasman 
> ---
> Changes in v2: Add back a missing space in a comment
> 
>  drivers/staging/greybus/arche-platform.c | 29 +++--
>  1 file changed, 19 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/staging/greybus/arche-platform.c 
> b/drivers/staging/greybus/arche-platform.c
> index eced2d2..eedd239 100644
> --- a/drivers/staging/greybus/arche-platform.c
> +++ b/drivers/staging/greybus/arche-platform.c
> @@ -172,15 +172,21 @@ static irqreturn_t arche_platform_wd_irq(int irq, void 
> *devid)
>   if (arche_pdata->wake_detect_state == WD_STATE_BOOT_INIT) {
>   if (time_before(jiffies,
>   arche_pdata->wake_detect_start +
> - 
> msecs_to_jiffies(WD_COLDBOOT_PULSE_WIDTH_MS))) {
> - 
> arche_platform_set_wake_detect_state(arche_pdata,
> -  
> WD_STATE_IDLE);
> + msecs_to_jiffies(
> + WD_COLDBOOT_PULSE_WIDTH_MS))) {
> + arche_platform_set_wake_detect_state(

We don't break the lines like this in kernel to satisfy the 80 column width
rule. Surely, some places would have similar code, but in general this isn't
recommended. And that's why we never bothered about 80 column rule in this and
below cases.

> + arche_pdata,
> + WD_STATE_IDLE);
>   } else {
> - /* Check we are not in middle of irq thread 
> already */
> + /*
> +  * Check we are not in middle of irq thread
> +  * already
> +  */

Yes, such changes would be fine.

>   if (arche_pdata->wake_detect_state !=
>   WD_STATE_COLDBOOT_START) {
> - 
> arche_platform_set_wake_detect_state(arche_pdata,
> -  
> WD_STATE_COLDBOOT_TRIG);
> + arche_platform_set_wake_detect_state(

Not this.

> + arche_pdata,
> + WD_STATE_COLDBOOT_TRIG);
>   spin_unlock_irqrestore(
>   _pdata->wake_lock,
>   flags);
> @@ -199,8 +205,9 @@ static irqreturn_t arche_platform_wd_irq(int irq, void 
> *devid)
>* beyond 30msec, then it is coldboot else fallback
>* to standby boot.
>*/
> - arche_platform_set_wake_detect_state(arche_pdata,
> -  
> WD_STATE_BOOT_INIT);
> + arche_platform_set_wake_detect_state(

Not this.

> + arche_pdata,
> + WD_STATE_BOOT_INIT);
>   }
>   }
>  
> @@ -657,12 +664,14 @@ static SIMPLE_DEV_PM_OPS(arche_platform_pm_ops,
>   arche_platform_resume);
>  
>  static const struct of_device_id arche_platform_of_match[] = {
> - { .compatible = "google,arche-platform", }, /* Use PID/VID of SVC 
> device */
> + /* Use PID/VID of SVC device */
> + { .compatible = "google,arche-platform", },

This would be fine though.

>   { },
>  };
>  
>  static const struct of_device_id arche_combined_id[] = {
> - { .compatible = "google,arche-platform", }, /* Use PID/VID of SVC 
> device */
> + /* Use PID/VID of SVC device */
> + { .compatible = "google,arche-platform", },
>   { .compatible = "usb,2", },
>   { },
>  };

-- 
viresh

Re: [greybus-dev] [PATCH v2] staging: greybus: arche: wrap over-length lines

[Viresh Kumar]

Hi Mitchell,

On 09-07-17, 20:25, Mitchell Tasman wrote:
> Adjust formatting of various statements to keep line length within
> the 80 column limit preferred by the Linux kernel coding style.

We try to follow that most of the time, but the end result should be easily
readable.  If it isn't, then we just ignore the rule.

> Signed-off-by: Mitchell Tasman 
> ---
> Changes in v2: Add back a missing space in a comment
> 
>  drivers/staging/greybus/arche-platform.c | 29 +++--
>  1 file changed, 19 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/staging/greybus/arche-platform.c 
> b/drivers/staging/greybus/arche-platform.c
> index eced2d2..eedd239 100644
> --- a/drivers/staging/greybus/arche-platform.c
> +++ b/drivers/staging/greybus/arche-platform.c
> @@ -172,15 +172,21 @@ static irqreturn_t arche_platform_wd_irq(int irq, void 
> *devid)
>   if (arche_pdata->wake_detect_state == WD_STATE_BOOT_INIT) {
>   if (time_before(jiffies,
>   arche_pdata->wake_detect_start +
> - 
> msecs_to_jiffies(WD_COLDBOOT_PULSE_WIDTH_MS))) {
> - 
> arche_platform_set_wake_detect_state(arche_pdata,
> -  
> WD_STATE_IDLE);
> + msecs_to_jiffies(
> + WD_COLDBOOT_PULSE_WIDTH_MS))) {
> + arche_platform_set_wake_detect_state(

We don't break the lines like this in kernel to satisfy the 80 column width
rule. Surely, some places would have similar code, but in general this isn't
recommended. And that's why we never bothered about 80 column rule in this and
below cases.

> + arche_pdata,
> + WD_STATE_IDLE);
>   } else {
> - /* Check we are not in middle of irq thread 
> already */
> + /*
> +  * Check we are not in middle of irq thread
> +  * already
> +  */

Yes, such changes would be fine.

>   if (arche_pdata->wake_detect_state !=
>   WD_STATE_COLDBOOT_START) {
> - 
> arche_platform_set_wake_detect_state(arche_pdata,
> -  
> WD_STATE_COLDBOOT_TRIG);
> + arche_platform_set_wake_detect_state(

Not this.

> + arche_pdata,
> + WD_STATE_COLDBOOT_TRIG);
>   spin_unlock_irqrestore(
>   _pdata->wake_lock,
>   flags);
> @@ -199,8 +205,9 @@ static irqreturn_t arche_platform_wd_irq(int irq, void 
> *devid)
>* beyond 30msec, then it is coldboot else fallback
>* to standby boot.
>*/
> - arche_platform_set_wake_detect_state(arche_pdata,
> -  
> WD_STATE_BOOT_INIT);
> + arche_platform_set_wake_detect_state(

Not this.

> + arche_pdata,
> + WD_STATE_BOOT_INIT);
>   }
>   }
>  
> @@ -657,12 +664,14 @@ static SIMPLE_DEV_PM_OPS(arche_platform_pm_ops,
>   arche_platform_resume);
>  
>  static const struct of_device_id arche_platform_of_match[] = {
> - { .compatible = "google,arche-platform", }, /* Use PID/VID of SVC 
> device */
> + /* Use PID/VID of SVC device */
> + { .compatible = "google,arche-platform", },

This would be fine though.

>   { },
>  };
>  
>  static const struct of_device_id arche_combined_id[] = {
> - { .compatible = "google,arche-platform", }, /* Use PID/VID of SVC 
> device */
> + /* Use PID/VID of SVC device */
> + { .compatible = "google,arche-platform", },
>   { .compatible = "usb,2", },
>   { },
>  };

-- 
viresh