[PATCH 0/8] [-mm] ACPI: add ACPI device sysfs interface

[PATCH 0/8] [-mm] ACPI: add ACPI device sysfs interface

[Zhang Rui]

Add sysfs interface for ACPI devices.

This patch set adds sysfs interface for ACPI EC, Fan,
Power_resource, Processor and Thermal_zone devices.

Patch 01 add some helper macros and functions.

Patch 02/03/04 add sysfs interface for ACPI EC, Fan
and Power_resource devices.

Patch 05/06/07 are for ACPI Porcessor (core/throttling/thermal).
Sysfs interfaces for ACPI Processor Power/Performance state
control are already available via cpu_freq and cpu_idle driver.

Sysfs interface for ACPI thermal_zone device is added in Patch 08.

More details are offered in each patch.
Any comments are appreciated. :)

Thanks,
Rui.

Re: [PATCH 0/8] [-mm] ACPI: add ACPI device sysfs interface

[Henrique de Moraes Holschuh]

On Tue, 20 Mar 2007, Zhang Rui wrote:
> Add sysfs interface for ACPI devices.

While doing the ibm-acpi sysfs work (not submitted yet), I have found I need
something like this to properly parse simple ulongs from sysfs:

static int parse_strtoul(const char *buf,
		unsigned long max, unsigned long *value)
{
	char *endp;

	*value = simple_strtoul(buf, &endp, 0);
	while (*endp && isspace(*endp))
		endp++;
	if (*endp || *value > max)
		return -EINVAL;

	return 0;
}

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

Re: [PATCH 0/8] [-mm] ACPI: add ACPI device sysfs interface

[Zhang Rui]

On Tue, 2007-03-20 at 12:31 -0300, Henrique de Moraes Holschuh wrote:
> On Tue, 20 Mar 2007, Zhang Rui wrote:
> > Add sysfs interface for ACPI devices.
> 
> While doing the ibm-acpi sysfs work (not submitted yet), I have found I need
> something like this to properly parse simple ulongs from sysfs:
> 
> static int parse_strtoul(const char *buf,
> 		unsigned long max, unsigned long *value)
> {
> 	char *endp;
> 
> 	*value = simple_strtoul(buf, &endp, 0);
> 	while (*endp && isspace(*endp))
> 		endp++;
> 	if (*endp || *value > max)
> 		return -EINVAL;
> 
> 	return 0;
> }
> 
Sounds nice.
But I think it can work better without the parameter "max".
Not all of the .store functions have "max" limit, e.g. the thermal
polling_freq, while some of them may need "max" and "min" limit both,
like fan state (0/3).
So why not just get the value and let the .store function judge it? :)

Thanks,
Rui.

Re: [PATCH 0/8] [-mm] ACPI: add ACPI device sysfs interface

[Henrique de Moraes Holschuh]

On Wed, 21 Mar 2007, Zhang Rui wrote:
> On Tue, 2007-03-20 at 12:31 -0300, Henrique de Moraes Holschuh wrote:
> > On Tue, 20 Mar 2007, Zhang Rui wrote:
> > > Add sysfs interface for ACPI devices.
> > 
> > While doing the ibm-acpi sysfs work (not submitted yet), I have found I need
> > something like this to properly parse simple ulongs from sysfs:
> > 
> > static int parse_strtoul(const char *buf,
> > 		unsigned long max, unsigned long *value)
> > {
> > 	char *endp;
> > 
> > 	*value = simple_strtoul(buf, &endp, 0);
> > 	while (*endp && isspace(*endp))
> > 		endp++;
> > 	if (*endp || *value > max)
> > 		return -EINVAL;
> > 
> > 	return 0;
> > }
> > 
> Sounds nice.
> But I think it can work better without the parameter "max".

Well, it is a generic parser for unsigned longs as I needed them in
ibm-acpi, where the minimum value is always zero, and there is always a
maximum value.  If you don't need "max", just remove it :-)

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh