problems with longhaul cpufreq driver.

problems with longhaul cpufreq driver.

[mgillespie]

Running Redhat9 with my own kernel compilation based on 2.4.23. 

I added the cpufreq patch built in (not module).  It patched cleanlt with no errors. 

When I look at my speeds, they are all 0. 

If I check the dmesg log, I see the following on startup: 

longhaul: VIA C3 'Ezra' [C5C] CPU detected. Longhaul v2 supported. 
longhaul: MinMult=5.0x MaxMult=6.0x 
longhaul: FSB: 0MHz Lowestspeed=0MHz Highestspeed=0MHz 
longhaul: FSB:0 Mult:10.0x 

My Setup:

Via EPIA c800, and using cpufreq: cpufreq-LINUX_2_4-20031207.tar.gz 

Here is a snippet of my config (options not used removed) 

# 
# Processor type and features 
# 
CONFIG_MCYRIXIII=y 
CONFIG_X86_WP_WORKS_OK=y 
CONFIG_X86_INVLPG=y 
CONFIG_X86_CMPXCHG=y 
CONFIG_X86_XADD=y 
CONFIG_X86_BSWAP=y 
CONFIG_X86_POPAD_OK=y 
CONFIG_RWSEM_XCHGADD_ALGORITHM=y 
CONFIG_X86_L1_CACHE_SHIFT=5 
CONFIG_X86_HAS_TSC=y 
CONFIG_X86_ALIGNMENT_16=y 
CONFIG_X86_USE_3DNOW=y 
CONFIG_X86_USE_PPRO_CHECKSUM=y 
CONFIG_X86_F00F_WORKS_OK=y 
CONFIG_X86_MCE=y 

# 
# CPU Frequency scaling 
# 
CONFIG_CPU_FREQ=y 
CONFIG_CPU_FREQ_TABLE=y 
CONFIG_CPU_FREQ_PROC_INTF=y 
CONFIG_CPU_FREQ_GOV_USERSPACE=y 
CONFIG_CPU_FREQ_24_API=y 
CONFIG_X86_LONGHAUL=y 
CONFIG_X86_CPUID=m 
CONFIG_NOHIGHMEM=y 
CONFIG_MTRR=y 
CONFIG_X86_TSC=y 

Anyone else got this problem? 

Re: problems with longhaul cpufreq driver.

[Ashley Pittman]

On Mon, 2003-12-08 at 16:26, mgillespie@mgillespie.plus.com wrote:

> If I check the dmesg log, I see the following on startup: 
> 
> longhaul: VIA C3 'Ezra' [C5C] CPU detected. Longhaul v2 supported. 
> longhaul: MinMult=5.0x MaxMult=6.0x 
> longhaul: FSB: 0MHz Lowestspeed=0MHz Highestspeed=0MHz 
> longhaul: FSB:0 Mult:10.0x 

The problem here is the FSB is being reported as 0, this is wrong.

> Anyone else got this problem? 

I had a very similar problem with my c3 using a 2.6 kernel, I posted a
patch to this list a while ago that, whilst incorrect in some ways is
probably enough to allow the module to work.  There was a discussion
about a better way to do it but I don't think that anything has been
committed yet.

I would be interested to know if the patch works for you.

Ashley,

Re: problems with longhaul cpufreq driver.

[Ducrot Bruno]

On Mon, Dec 08, 2003 at 04:39:03PM +0000, Ashley Pittman wrote:
> On Mon, 2003-12-08 at 16:26, mgillespie@mgillespie.plus.com wrote:
> 
> > If I check the dmesg log, I see the following on startup: 
> > 
> > longhaul: VIA C3 'Ezra' [C5C] CPU detected. Longhaul v2 supported. 
> > longhaul: MinMult=5.0x MaxMult=6.0x 
> > longhaul: FSB: 0MHz Lowestspeed=0MHz Highestspeed=0MHz 
> > longhaul: FSB:0 Mult:10.0x 
> 
> The problem here is the FSB is being reported as 0, this is wrong.
> 
> > Anyone else got this problem? 
> 
> I had a very similar problem with my c3 using a 2.6 kernel, I posted a
> patch to this list a while ago that, whilst incorrect in some ways is
> probably enough to allow the module to work.  There was a discussion
> about a better way to do it but I don't think that anything has been
> committed yet.
> 

Ashley:

I don't know if you refer to:

http://linux-dj.bkbits.net:8080/cpufreq/cset@1.1349.1.5?nav=index.html|ChangeSet@-3w

If yes, then that bit is probably not in cvs 2.4, nor in any
linux-2.6 mainstream (I guess).


-- 
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.

Re: problems with longhaul cpufreq driver.

[Mark Gillespie]

Ducrot Bruno wrote:

> On Mon, Dec 08, 2003 at 04:39:03PM +0000, Ashley Pittman wrote:
> 
>>On Mon, 2003-12-08 at 16:26, mgillespie@mgillespie.plus.com wrote:
>>
>>
>>>If I check the dmesg log, I see the following on startup: 
>>>
>>>longhaul: VIA C3 'Ezra' [C5C] CPU detected. Longhaul v2 supported. 
>>>longhaul: MinMult=5.0x MaxMult=6.0x 
>>>longhaul: FSB: 0MHz Lowestspeed=0MHz Highestspeed=0MHz 
>>>longhaul: FSB:0 Mult:10.0x 
>>
>>The problem here is the FSB is being reported as 0, this is wrong.
>>
>>
>>>Anyone else got this problem? 
>>
>>I had a very similar problem with my c3 using a 2.6 kernel, I posted a
>>patch to this list a while ago that, whilst incorrect in some ways is
>>probably enough to allow the module to work.  There was a discussion
>>about a better way to do it but I don't think that anything has been
>>committed yet.
>>

My hack was even simpler!!  I just hacked the function guess_fsb to 
always return 133!!  (the kernel nor code is going nowhere other than 
this small eden box, so it matters not)...

My DMESG now shows:

longhaul: VIA C3 'Ezra' [C5C] CPU detected. Longhaul v2 supported.
longhaul: MinMult=5.0x MaxMult=6.0x
longhaul: FSB: 133MHz Lowestspeed=665MHz Highestspeed=798MHz


Thanks all for the pointers...