Re: [PATCH 0/5] I8K driver facelift

[Dmitry Torokhov <dmitry.torokhov@gmail.com>]

On Thu, 17 Mar 2005 02:37:56 -0700, Frank Sorenson <frank@tuxrocks.com> wrote:

> 1: My Inspiron 9200 (and perhaps others) doesn't seem to respond to the
> I8K_SMM_BIOS_VERSION function call, so it fails the check in i8k_probe.
> ~ The check of i8k_get_bios_version doesn't seem critical, and removing
> the return -ENODEV makes it work again for me.  That's the current
> behavior, so perhaps the printk level should just be changed to
> KERN_WARNING rather than KERN_ALERT.

You are probably right, I shoudl change that.

> 2: To compile 2.6.11 cleanly, I needed two hunks from your original
> patch 2 (perhaps you're working from a more up-to-date tree than I am?
> If so, these are probably already addressed.):
> 

Oh, sorry - when I was pereparing cumulative patch I simply missed
this bit. It is still nowhere near the official tree.

> 
> The 'temp' entries make sense, however I'm not sure about the fan_speed
> and fan_state entries.  From the perspective of how the objects are
> ordered, a fan would have 'speed' and 'state' attributes, but a
> 'fan_state' attribute wouldn't normally have a fan.  Maybe something
> along these lines would make more sense from that perspective:
> 
> ./fan/0
> ./fan/0/speed
> ./fan/0/state
> ./fan/1
> ./fan/1/speed
> ./fan/1/state
> 

Yes, as soon as I did attribute array I realized that something like
attr_array_group would reflect the structure better... We'll see what 
can be done.

Thank you for your comments and suggestions!

-- 
Dmitry