Re: [PATCH v4 2/2] cpufreq: add virtual-cpufreq driver

[Hongyan Xia]

On 11/11/2023 01:49, David Dai wrote:
> [...]
> diff --git a/drivers/cpufreq/Makefile b/drivers/cpufreq/Makefile
> index 8d141c71b016..eb72ecdc24db 100644
> --- a/drivers/cpufreq/Makefile
> +++ b/drivers/cpufreq/Makefile
> @@ -16,6 +16,7 @@ obj-$(CONFIG_CPU_FREQ_GOV_ATTR_SET)	+= cpufreq_governor_attr_set.o
>   
>   obj-$(CONFIG_CPUFREQ_DT)		+= cpufreq-dt.o
>   obj-$(CONFIG_CPUFREQ_DT_PLATDEV)	+= cpufreq-dt-platdev.o
> +obj-$(CONFIG_CPUFREQ_VIRT)		+= virtual-cpufreq.o
>   
>   # Traces
>   CFLAGS_amd-pstate-trace.o               := -I$(src)
> diff --git a/drivers/cpufreq/virtual-cpufreq.c b/drivers/cpufreq/virtual-cpufreq.c
> new file mode 100644
> index 000000000000..f828d3345a68
> --- /dev/null
> +++ b/drivers/cpufreq/virtual-cpufreq.c
> @@ -0,0 +1,201 @@
> +// SPDX-License-Identifier: GPL-2.0-only
> +/*
> + * Copyright (C) 2023 Google LLC
> + */
> +
> +#include <linux/arch_topology.h>
> +#include <linux/cpufreq.h>
> +#include <linux/init.h>
> +#include <linux/sched.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/of_address.h>
> +#include <linux/of_platform.h>
> +#include <linux/pm_opp.h>
> +#include <linux/slab.h>
> +
> +#define REG_CUR_FREQ_KHZ_OFFSET 0x0
> +#define REG_SET_FREQ_KHZ_OFFSET 0x4
> +#define PER_CPU_OFFSET 0x8
> +
> +static void __iomem *base;
> +
> +static void virt_scale_freq_tick(void)
> +{
> +	int cpu = smp_processor_id();
> +	u32 max_freq = (u32)cpufreq_get_hw_max_freq(cpu);
> +	u64 cur_freq;
> +	unsigned long scale;
> +
> +	cur_freq = (u64)readl_relaxed(base + cpu * PER_CPU_OFFSET
> +			+ REG_CUR_FREQ_KHZ_OFFSET);
> +
> +	cur_freq <<= SCHED_CAPACITY_SHIFT;
> +	scale = (unsigned long)div_u64(cur_freq, max_freq);
> +	scale = min(scale, SCHED_CAPACITY_SCALE);
> +
> +	this_cpu_write(arch_freq_scale, scale);
> +}

Here we update the scaling factor in the guest, but is there any way to 
let the guest know when the host dequeues the vCPU so that the guest 
PELT signal doesn't appear larger than it actually is? Is this a known 
limitation and is there a way to mitigate it?

> [...]