[ltp] [PATCH] ibm_acpi - Allow fan to be set to full speed

[Thomas Renninger]

Bret Comstock Waldow wrote:
> On Mon, 22 Aug 2005 00:34, honey@gneek.com wrote:
> 
>>Why don't our thinkpad fans go that fast on their own when near
>>critical temperature?  It seems to be a solution to me.
> 
> Do we know that the fan can operate at these speeds without significantly 
> shortening it's lifespan?
> 
> Does the fan ever run this fast under Windows?  If not, and Windows 
> nevertheless doesn't go into thermal shutdown, then we are missing some other 
> part of the power/thermal management picture, and trying to push the fan into 
> making up for our lack of knowledge.

I fear you are right.
I expect these Thinkpad models rely on a proper OS thermal (not fan!) management, which is currently
broken for the ondemand (and probably also for the conservative) governor.
It was a pleasure to see thermal managemnt working with the userspace governor, people would never
notice that their machine sometime run with 1200 MHz instead of 1700 after 20 Mins of permanent load.

It's hard to believe that IBM designed these machines to work on 92 C permanently,
OS system has to set freq back to 1200Mhz (or whatever lower speed), but it looks like so.

Yes, it would be very interessting to see how the machines behave under windows, especially
concerning max freq and max temp.

> 
> The easy answer is easy - do we know this is the right one?  I'd like to hear 
> some discussion.
I am also curious about any information, one finds out.

Thanks,

       Thomas
> 
> I don't mean "well, if it goes fast, it will cool better, won't it?" - too 
> many ifs.  I mean something like "it goes this fast in Windows routinely, so 
> we know we aren't abandoning all the engineering work IBM put into the 
> design".
>