public inbox for linux-omap@vger.kernel.org
 help / color / mirror / Atom feed
* watchdog pdev id set to -1
@ 2011-02-18 10:41 Koyamangalath, Abhilash
  2011-02-18 10:56 ` Felipe Balbi
  0 siblings, 1 reply; 6+ messages in thread
From: Koyamangalath, Abhilash @ 2011-02-18 10:41 UTC (permalink / raw)
  To: linux-omap@vger.kernel.org

There this boot-time message (on both omap3 and am3517 evms atleast):
[    1.904418] omap_device: omap_wdt.-1: new worst case activate latency 0: 6105

omap_init_wdt() initializes the id to -1 and then omap_device_build() is called with this value; this eventually calls platform_device_register() with a pdev having id as -1. I don't see any code that checks for this -1 and changes it to a non-erroneous value.

Perhaps there is a reason I've failed to notice?
Please comment.
Else I'm planning to submit a patch which fixes this.

- Abhilash


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: watchdog pdev id set to -1
  2011-02-18 10:41 watchdog pdev id set to -1 Koyamangalath, Abhilash
@ 2011-02-18 10:56 ` Felipe Balbi
  2011-02-18 11:09   ` Koyamangalath, Abhilash
  0 siblings, 1 reply; 6+ messages in thread
From: Felipe Balbi @ 2011-02-18 10:56 UTC (permalink / raw)
  To: Koyamangalath, Abhilash; +Cc: linux-omap@vger.kernel.org

Hi,

(your lines are over 80-characters, fix it)

On Fri, Feb 18, 2011 at 04:11:11PM +0530, Koyamangalath, Abhilash wrote:
> There this boot-time message (on both omap3 and am3517 evms atleast):
> [    1.904418] omap_device: omap_wdt.-1: new worst case activate
> latency 0: 6105
> 
> omap_init_wdt() initializes the id to -1 and then omap_device_build()
> is called with this value; this eventually calls
> platform_device_register() with a pdev having id as -1. I don't see
> any code that checks for this -1 and changes it to a non-erroneous
> value.
> 
> Perhaps there is a reason I've failed to notice?  Please comment.
> Else I'm planning to submit a patch which fixes this.

That's not an error value. If you have a -1 ID, you're only telling the
kernel "this is only omap_wdt platform_device I have", which means that
it will look in sysfs as:

/sys/devices/platform/omap_wdt/

if you have an ID >= 0, it will look as:

/sys/devices/platform/omap_wdt.0/
/sys/devices/platform/omap_wdt.1/
/sys/devices/platform/omap_wdt.2/

etc, that's all.

-- 
balbi

^ permalink raw reply	[flat|nested] 6+ messages in thread

* RE: watchdog pdev id set to -1
  2011-02-18 10:56 ` Felipe Balbi
@ 2011-02-18 11:09   ` Koyamangalath, Abhilash
  2011-02-18 11:12     ` Felipe Balbi
  0 siblings, 1 reply; 6+ messages in thread
From: Koyamangalath, Abhilash @ 2011-02-18 11:09 UTC (permalink / raw)
  To: Balbi, Felipe; +Cc: linux-omap@vger.kernel.org

I understand, but is there a subtle reason why are starting from
an erroneous-looking -1 rather than a more natural 0 (or 1 ?).


> -----Original Message-----
> From: Balbi, Felipe
> Sent: Friday, February 18, 2011 4:27 PM
> To: Koyamangalath, Abhilash
> Cc: linux-omap@vger.kernel.org
> Subject: Re: watchdog pdev id set to -1
> 
> Hi,
> 
> (your lines are over 80-characters, fix it)
> 
> On Fri, Feb 18, 2011 at 04:11:11PM +0530, Koyamangalath, Abhilash wrote:
> > There this boot-time message (on both omap3 and am3517 evms atleast):
> > [    1.904418] omap_device: omap_wdt.-1: new worst case activate
> > latency 0: 6105
> >
> > omap_init_wdt() initializes the id to -1 and then omap_device_build()
> > is called with this value; this eventually calls
> > platform_device_register() with a pdev having id as -1. I don't see
> > any code that checks for this -1 and changes it to a non-erroneous
> > value.
> >
> > Perhaps there is a reason I've failed to notice?  Please comment.
> > Else I'm planning to submit a patch which fixes this.
> 
> That's not an error value. If you have a -1 ID, you're only telling the
> kernel "this is only omap_wdt platform_device I have", which means that
> it will look in sysfs as:
> 
> /sys/devices/platform/omap_wdt/
> 
> if you have an ID >= 0, it will look as:
> 
> /sys/devices/platform/omap_wdt.0/
> /sys/devices/platform/omap_wdt.1/
> /sys/devices/platform/omap_wdt.2/
> 
> etc, that's all.
> 
> --
> balbi

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: watchdog pdev id set to -1
  2011-02-18 11:09   ` Koyamangalath, Abhilash
@ 2011-02-18 11:12     ` Felipe Balbi
  2011-02-22 21:59       ` Mark Brown
  0 siblings, 1 reply; 6+ messages in thread
From: Felipe Balbi @ 2011-02-18 11:12 UTC (permalink / raw)
  To: Koyamangalath, Abhilash; +Cc: Balbi, Felipe, linux-omap@vger.kernel.org

Hi,

(don't top-post)

On Fri, Feb 18, 2011 at 04:39:48PM +0530, Koyamangalath, Abhilash wrote:
> I understand, but is there a subtle reason why are starting from
> an erroneous-looking -1 rather than a more natural 0 (or 1 ?).

-1 means no ID. The IDs are zero-based, so we can't use zero. Just
change that value and see how things change on sysfs and dmesg then you
will know what I'm talking about. That's just an implementation decision
of using -1 to signify "no ID needed".

-- 
balbi

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: watchdog pdev id set to -1
  2011-02-18 11:12     ` Felipe Balbi
@ 2011-02-22 21:59       ` Mark Brown
  2011-03-01  0:35         ` Kevin Hilman
  0 siblings, 1 reply; 6+ messages in thread
From: Mark Brown @ 2011-02-22 21:59 UTC (permalink / raw)
  To: Felipe Balbi; +Cc: Koyamangalath, Abhilash, linux-omap@vger.kernel.org

On Fri, Feb 18, 2011 at 01:12:37PM +0200, Felipe Balbi wrote:

> On Fri, Feb 18, 2011 at 04:39:48PM +0530, Koyamangalath, Abhilash wrote:
> > I understand, but is there a subtle reason why are starting from
> > an erroneous-looking -1 rather than a more natural 0 (or 1 ?).

> -1 means no ID. The IDs are zero-based, so we can't use zero. Just
> change that value and see how things change on sysfs and dmesg then you
> will know what I'm talking about. That's just an implementation decision
> of using -1 to signify "no ID needed".

Though looking at the report it looks like the omap_device print that's
done during boot is doing the wrong thing here and displaying the -1
instead of masking it from the print which is what's expected.

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: watchdog pdev id set to -1
  2011-02-22 21:59       ` Mark Brown
@ 2011-03-01  0:35         ` Kevin Hilman
  0 siblings, 0 replies; 6+ messages in thread
From: Kevin Hilman @ 2011-03-01  0:35 UTC (permalink / raw)
  To: Mark Brown
  Cc: Felipe Balbi, Koyamangalath, Abhilash, linux-omap@vger.kernel.org

Mark Brown <broonie@opensource.wolfsonmicro.com> writes:

> On Fri, Feb 18, 2011 at 01:12:37PM +0200, Felipe Balbi wrote:
>
>> On Fri, Feb 18, 2011 at 04:39:48PM +0530, Koyamangalath, Abhilash wrote:
>> > I understand, but is there a subtle reason why are starting from
>> > an erroneous-looking -1 rather than a more natural 0 (or 1 ?).
>
>> -1 means no ID. The IDs are zero-based, so we can't use zero. Just
>> change that value and see how things change on sysfs and dmesg then you
>> will know what I'm talking about. That's just an implementation decision
>> of using -1 to signify "no ID needed".
>
> Though looking at the report it looks like the omap_device print that's
> done during boot is doing the wrong thing here and displaying the -1
> instead of masking it from the print which is what's expected.

Agreed.

Abhilash, Just send a patch to fix the omap_device printk which is the
confusing part.

Kevin

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2011-03-01  0:35 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-02-18 10:41 watchdog pdev id set to -1 Koyamangalath, Abhilash
2011-02-18 10:56 ` Felipe Balbi
2011-02-18 11:09   ` Koyamangalath, Abhilash
2011-02-18 11:12     ` Felipe Balbi
2011-02-22 21:59       ` Mark Brown
2011-03-01  0:35         ` Kevin Hilman

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox