[PATCH 1/3] arm64: topology: Improve AMU-based frequency calculation

Jie Zhan zhanjie9 at hisilicon.com
Wed Nov 5 20:12:01 PST 2025



On 11/4/2025 3:55 PM, Bowen Yu wrote:
> The current approach of reverse-calculating CPU frequency from capacity
> values introduces quantization errors due to intermediate scaling of
> arch_scale_freq_capacity, which results in the calculated frequency having
> only 1/1024 resolution.
> 
> This patch:
> 1. Directly computes frequency using AMU counters in amu_scale_freq_tick():
> freq = (core_cycles_delta * timer_freq) / (const_cycles_delta * 1000)
>  - core_cycles_delta: Measured CPU cycles
>  - timer_freq: Architectural timer frequency
>  - const_cycles_delta: Reference cycles from fixed-frequency timer
> 2. Returns pre-computed avgfreq in arch_freq_get_on_cpu()
> 
> examples:
> Before change
> [root at localhost ~]# cat /sys/devices/system/cpu/cpufreq/policy*/cpuinfo_avg_freq
> 2297851
> 2297851
> 2295312
> 2297851
> 2297851
> 2295312
> 2297851
> 2295312
> 2297851
> 2297851
> 2297851
> 2295312
> 2295312
> 2297851
> 2297851
> 2297851
> 2297851
> 2300390
> 2297851
> 2297851
> 2297851
> 
> After change
> [root at localhost ~]# cat /sys/devices/system/cpu/cpufreq/policy*/cpuinfo_avg_freq
> 2299177
> 2298117
> 2299188
> 2297330
> 2296530
> 2298817
> 2298434
> 2298986
> 2298596
> 2299395
> 2299560
> 2298446
> 2299108
> 2299294
> 2298707
> 2298453
> 2298632
> 2299218
> 2297962
> 
> Signed-off-by: Bowen Yu <yubowen8 at huawei.com>
> ---
>  arch/arm64/kernel/topology.c | 20 +++++++++++---------
>  1 file changed, 11 insertions(+), 9 deletions(-)
> 
...
> @@ -288,7 +293,7 @@ int arch_freq_get_on_cpu(int cpu)
>  	unsigned int start_cpu = cpu;
>  	unsigned long last_update;
>  	unsigned int freq = 0;
> -	u64 scale;
> +	u64 delta_core_kHz;
>  
>  	if (!amu_fie_cpu_supported(cpu) || !arch_scale_freq_ref(cpu))
>  		return -EOPNOTSUPP;
> @@ -340,14 +345,11 @@ int arch_freq_get_on_cpu(int cpu)
>  			break;
>  		}
>  	}
> -	/*
> -	 * Reversed computation to the one used to determine
> -	 * the arch_freq_scale value
> -	 * (see amu_scale_freq_tick for details)
> -	 */
> -	scale = arch_scale_freq_capacity(cpu);
> -	freq = scale * arch_scale_freq_ref(cpu);
> -	freq >>= SCHED_CAPACITY_SHIFT;
> +
> +	if (check_mul_overflow(per_cpu(core_delta, cpu), arch_timer_get_cntfrq(), &delta_core_kHz))
Hi Bowen,

IIUC, the variable 'delta_core_kHz' doesn't mean its name.
'core_delta * timer_freq' is just a transitional number.
The naming is misleading.

Perhaps consider reusing 'freq'? i.e. define 'freq' as u64 and replace
'delta_core_kHz' with 'freq', then return (int)freq at the end.

Jie
> +		return -EINVAL;
> +
> +	freq = div_u64(delta_core_kHz, per_cpu(const_delta, cpu) * HZ_PER_KHZ);
>  	return freq;
>  }
>  



More information about the linux-arm-kernel mailing list