[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ZxCmHErZZmY27H3M@BLRRASHENOY1.amd.com>
Date: Thu, 17 Oct 2024 11:22:28 +0530
From: "Gautham R. Shenoy" <gautham.shenoy@....com>
To: Mario Limonciello <mario.limonciello@....com>
Cc: Borislav Petkov <bp@...en8.de>, Hans de Goede <hdegoede@...hat.com>,
Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>,
x86@...nel.org, Perry Yuan <perry.yuan@....com>,
linux-kernel@...r.kernel.org, linux-doc@...r.kernel.org,
linux-pm@...r.kernel.org, platform-driver-x86@...r.kernel.org,
Shyam Sundar S K <Shyam-sundar.S-k@....com>
Subject: Re: [PATCH v3 07/14] platform/x86: hfi: init per-cpu scores for each
class
On Tue, Oct 15, 2024 at 04:36:38PM -0500, Mario Limonciello wrote:
> From: Perry Yuan <Perry.Yuan@....com>
>
> Initialize per cpu score `amd_hfi_ipcc_scores` which store energy score
> and performance score data for each class.
>
> `Classic core` and `Dense core` are ranked according to those values as
> energy efficiency capability or performance capability.
> OS scheduler will pick cores from the ranking list on each class ID for
> the thread which provide the class id got from hardware feedback
> interface.
>
> Signed-off-by: Perry Yuan <Perry.Yuan@....com>
> Co-developed-by: Mario Limonciello <mario.limonciello@....com>
> Signed-off-by: Mario Limonciello <mario.limonciello@....com>
> ---
> v2->v3:
> * Drop jump label (Ilpo)
> ---
> drivers/platform/x86/amd/hfi/hfi.c | 31 ++++++++++++++++++++++++++++++
> 1 file changed, 31 insertions(+)
>
> diff --git a/drivers/platform/x86/amd/hfi/hfi.c b/drivers/platform/x86/amd/hfi/hfi.c
> index d2952233442b..63e66ab60655 100644
> --- a/drivers/platform/x86/amd/hfi/hfi.c
> +++ b/drivers/platform/x86/amd/hfi/hfi.c
> @@ -118,6 +118,8 @@ struct amd_hfi_cpuinfo {
>
> static DEFINE_PER_CPU(struct amd_hfi_cpuinfo, amd_hfi_cpuinfo) = {.class_index = -1};
>
> +static DEFINE_MUTEX(hfi_cpuinfo_lock);
> +
> static int find_cpu_index_by_apicid(unsigned int target_apicid)
> {
> int cpu_index;
> @@ -238,6 +240,31 @@ static void amd_hfi_remove(struct platform_device *pdev)
> mutex_destroy(&dev->lock);
> }
>
> +static int amd_set_hfi_ipcc_score(struct amd_hfi_cpuinfo *hfi_cpuinfo, int cpu)
> +{
> + for (int i = 0; i < hfi_cpuinfo->nr_class; i++)
> + WRITE_ONCE(hfi_cpuinfo->ipcc_scores[i],
> + hfi_cpuinfo->amd_hfi_classes[i].perf);
@cpu is not used here. But I suppose it will be used in a subsequent
patch to inform the scheduler about the changed score for this cpu.
Otherwise, looks good to me.
Reviewed-by: Gautham R. Shenoy <gautham.shenoy@....com>
--
Thanks and Regards
gautham.
> +
> + return 0;
> +}
> +
> +static int update_hfi_ipcc_scores(void)
> +{
> + int cpu;
> + int ret;
> +
> + for_each_present_cpu(cpu) {
> + struct amd_hfi_cpuinfo *hfi_cpuinfo = per_cpu_ptr(&amd_hfi_cpuinfo, cpu);
> +
> + ret = amd_set_hfi_ipcc_score(hfi_cpuinfo, cpu);
> + if (ret)
> + return ret;
> + }
> +
> + return 0;
> +}
> +
> static int amd_hfi_metadata_parser(struct platform_device *pdev,
> struct amd_hfi_data *amd_hfi_data)
> {
> @@ -321,6 +348,10 @@ static int amd_hfi_probe(struct platform_device *pdev)
> if (ret)
> return ret;
>
> + ret = update_hfi_ipcc_scores();
> + if (ret)
> + return ret;
> +
> return 0;
> }
>
> --
> 2.43.0
>
Powered by blists - more mailing lists