Subject: Re: RFC: est.c driver synced with OpenBSD.
To: Juan RP <juan@xtrarom.org>
From: Simon Burge <simonb@NetBSD.org>
List: tech-kern
Date: 08/29/2006 11:04:05
Juan RP wrote:

> A have a new laptop (Compaq Presario V4000) and I tried NetBSD
> to see how it works, so far so fine but SpeedStep didn't work.
> 
> After adding the entries to it I made it work, but after looking at the
> FreeBSD/OpenBSD code I think it's better to sync our code with them.
> 
> I synced the code with OpenBSD -current.
> 
> Full dmesg with new code:
> 
> http://www.xtrarom.org/~juan/PresarioV4000.dmesg
> 
> New est.c driver for review:
> 
> http://www.xtrarom.org/~juan/new_est.diff
> 
> New code does the following:
> 
> * Adds support for more Pentium M processors (and VIA C7 M
>    but not on this patch).
> * Reimplements the Enhanced SpeedStep CPU detection by using
>    quintuplet { vendor, MHz_hi, mV_hi, MHz_lo, mV_lo }, which matches
>    cpus more correctly that parsing the brand string.
> * It makes easier to add new models.
> 
> Can you please review or comment about it?

Do you know what needs to be done to support Core Duo CPUs?  I've got
a Dell 640m that says:

cpu0: features2 c1a9<SSE3,MONITOR,VMX,EST,TM2,xTPR>
cpu0: "Genuine Intel(R) CPU           T2400  @ 1.83GHz"
cpu0: Enhanced SpeedStep (1404 mV)  - unknown Enhanced SpeedStep CPU.
cpu1: features2 c1a9<SSE3,MONITOR,VMX,EST,TM2,xTPR>
cpu1: "Genuine Intel(R) CPU           T2400  @ 1.83GHz"
cpu1: Enhanced SpeedStep (1404 mV)  - unknown Enhanced SpeedStep CPU.

Even though there's an "EST" in the features2, est.c says "as
implemented in Pentium M processors".  Is that part of the issue?

Cheers,
Simon.