Re: [PATCH] Make acpi-cpufreq more robust against BIOS freq changes behind our back

["Anthony L. Awtrey" <tony@awtrey.com>]

On 05/21/2008 09:08 AM, Thomas Renninger wrote:
> Hi,
> 
> Please consider this one for stable integration.
> 
> On Wed, 2008-05-21 at 08:51 -0400, Anthony L. Awtrey wrote:
>> On 05/20/2008 06:38 AM, Thomas Renninger wrote:
>>> It could be that you hit a bug that might get fixed with a recent
>>> commit:
>>> commit e56a727b023d40d1adf660168883f30f2e6abe0a
>>> Author: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
>>> Date:   Mon Apr 28 15:13:43 2008 -0400
>>>
>>>     [CPUFREQ] Make acpi-cpufreq more robust against BIOS freq changes
>>> behind our back.
>> Hi,
>>
>> Sorry for the public display of ignorance, but I can't seem to find this 
>> patch as a simple diff so I can test it in one of my vanilla kernel 
>> builds (either 2.6.24.2 or 2.6.25.4, please). Can you provide me a 
>> general patch or some example git commands to get a build-able source tree?
> 
> Can you reply to the stable list if your tests were successful, pls.
> 
> Even if it doesn't work for your machine it definitely does for several
> Dells and one HP laptop.
> 
> This should patch and work fine for 2.6.25. AFAIK latest Fedora is also
> 2.6.25 based, they probably want to have this.
> 
> If this one patches for you for 2.6.24.4 also, maybe it even gets in
> there.
> 
> ------------
> 
> commit e56a727b023d40d1adf660168883f30f2e6abe0a
> Author: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
> Date:   Mon Apr 28 15:13:43 2008 -0400
> 
>     [CPUFREQ] Make acpi-cpufreq more robust against BIOS freq changes behind our back.
>     
>     We checked the hardware freq with OS cached freq value in get_cur_freqon_cpu().
>     
>     Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
>     Signed-off-by: Thomas Renninger <trenn@suse.de>
>     Signed-off-by: Dave Jones <davej@redhat.com>
> 
> diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
> index e2d870d..8db8f73 100644
> --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
> +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
> @@ -339,6 +339,7 @@ static unsigned int get_cur_freq_on_cpu(unsigned int cpu)
>  {
>  	struct acpi_cpufreq_data *data = per_cpu(drv_data, cpu);
>  	unsigned int freq;
> +	unsigned int cached_freq;
>  
>  	dprintk("get_cur_freq_on_cpu (%d)\n", cpu);
>  
> @@ -347,7 +348,16 @@ static unsigned int get_cur_freq_on_cpu(unsigned int cpu)
>  		return 0;
>  	}
>  
> +	cached_freq = data->freq_table[data->acpi_data->state].frequency;
>  	freq = extract_freq(get_cur_val(&cpumask_of_cpu(cpu)), data);
> +	if (freq != cached_freq) {
> +		/*
> +		 * The dreaded BIOS frequency change behind our back.
> +		 * Force set the frequency on next target call.
> +		 */
> +		data->resume = 1;
> +	}
> +
>  	dprintk("cur freq = %u\n", freq);
>  
>  	return freq;
> 
> 

Hello,

I worked! This patch applied to 2.6.25.4 fixed the CPU stepping problem 
on the Panasonic CF-18 (Pentium M) I was having.

Tony