[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <42143388-4f2a-4b81-ac79-14851c6090e2@amd.com>
Date: Tue, 30 Jan 2024 13:54:11 -0600
From: Mario Limonciello <mario.limonciello@....com>
To: Perry Yuan <perry.yuan@....com>, rafael.j.wysocki@...el.com,
Borislav.Petkov@....com, viresh.kumar@...aro.org, Ray.Huang@....com,
gautham.shenoy@....com
Cc: Alexander.Deucher@....com, Xinmei.Huang@....com, Xiaojian.Du@....com,
Li.Meng@....com, linux-pm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH Resend 7/8] tools/power x86_energy_perf_policy: rename
some perf output strings for AMD processors
On 1/29/2024 23:56, Perry Yuan wrote:
> rename some perf strings using same definition for AMD CPPC performance
> capabilities.
I'm not sure I get why these need to deviate. The nomenclature is
clearer for everyone this way isn't it? So why not just make the same
change for everyone (IE min->lowest, max->highest, des->desired)?
>
> cpu0: [AMD HWP_REQ]: lowest 12 highest 166 desired 0 epp 128 window
> 0x131497d0 (80*10^7us) use_pkg 89
>
> cpu1: [AMD HWP_REQ]: lowest 12 highest 166 desired 0 epp 128 window
> 0x131497d0 (80*10^7us) use_pkg 89
>
> Signed-off-by: Perry Yuan <perry.yuan@....com>
> ---
> .../x86_energy_perf_policy.c | 23 ++++++++++++++-----
> 1 file changed, 17 insertions(+), 6 deletions(-)
>
> diff --git a/tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c b/tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c
> index ada1aaabad8c..845cfedab06e 100644
> --- a/tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c
> +++ b/tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c
> @@ -864,10 +864,15 @@ void print_hwp_request(int cpu, struct msr_hwp_request *h, char *str)
>
> if (str)
> printf("%s", str);
> + if (genuine_intel)
> + printf("HWP_REQ: min %d max %d des %d epp %d window 0x%x (%d*10^%dus) use_pkg %d\n",
> + h->hwp_min, h->hwp_max, h->hwp_desired, h->hwp_epp,
> + h->hwp_window, h->hwp_window & 0x7F, (h->hwp_window >> 7) & 0x7, h->hwp_use_pkg);
> + else if (authentic_amd)
> + printf("[AMD HWP_REQ]: lowest %d highest %d desired %d epp %d window 0x%x (%d*10^%dus) use_pkg %d\n",
> + h->hwp_min, h->hwp_max, h->hwp_desired, h->hwp_epp,
> + h->hwp_window, h->hwp_window & 0x7F, (h->hwp_window >> 7) & 0x7, h->hwp_use_pkg);
>
> - printf("HWP_REQ: min %d max %d des %d epp %d window 0x%x (%d*10^%dus) use_pkg %d\n",
> - h->hwp_min, h->hwp_max, h->hwp_desired, h->hwp_epp,
> - h->hwp_window, h->hwp_window & 0x7F, (h->hwp_window >> 7) & 0x7, h->hwp_use_pkg);
> }
> void print_hwp_request_pkg(int pkg, struct msr_hwp_request *h, char *str)
> {
> @@ -876,9 +881,15 @@ void print_hwp_request_pkg(int pkg, struct msr_hwp_request *h, char *str)
> if (str)
> printf("%s", str);
>
> - printf("HWP_REQ_PKG: min %d max %d des %d epp %d window 0x%x (%d*10^%dus)\n",
> - h->hwp_min, h->hwp_max, h->hwp_desired, h->hwp_epp,
> - h->hwp_window, h->hwp_window & 0x7F, (h->hwp_window >> 7) & 0x7);
> + if (genuine_intel) {
> + printf("HWP_REQ_PKG: min %d max %d des %d epp %d window 0x%x (%d*10^%dus)\n",
> + h->hwp_min, h->hwp_max, h->hwp_desired, h->hwp_epp,
> + h->hwp_window, h->hwp_window & 0x7F, (h->hwp_window >> 7) & 0x7);
> + } else if (authentic_amd) {
> + printf("AMD HWP_REQ_PKG: lowest %d highest %d desired %d epp %d window 0x%x (%d*10^%dus)\n",
> + h->hwp_min, h->hwp_max, h->hwp_desired, h->hwp_epp,
> + h->hwp_window, h->hwp_window & 0x7F, (h->hwp_window >> 7) & 0x7);
> + }
> }
> void read_hwp_request(int cpu, struct msr_hwp_request *hwp_req, unsigned int msr_offset)
> {
Powered by blists - more mailing lists