linux-kernel - Re: [PATCH 2/2] cpufreq/amd-pstate: Fix the scaling_min/max_freq setting on shared memory CPPC systems

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <be834b5a-508e-49d1-b608-62192cd7bdda@amd.com>
Date: Tue, 25 Jun 2024 10:09:51 -0500
From: Mario Limonciello <mario.limonciello@....com>
To: Dhananjay Ugwekar <Dhananjay.Ugwekar@....com>
Cc: linux-pm@...r.kernel.org, linux-kernel@...r.kernel.org,
 "Rafael J. Wysocki" <rafael@...nel.org>,
 Viresh Kumar <viresh.kumar@...aro.org>,
 Shuah Khan <skhan@...uxfoundation.org>, Huang Rui <ray.huang@....com>,
 Perry Yuan <Perry.Yuan@....com>, Meng Li <li.meng@....com>,
 "Gautham R. Shenoy" <gautham.shenoy@....com>
Subject: Re: [PATCH 2/2] cpufreq/amd-pstate: Fix the scaling_min/max_freq
 setting on shared memory CPPC systems

On 6/25/2024 08:41, Dhananjay Ugwekar wrote:
> On shared memory CPPC systems, with amd_pstate=active mode, the change
> in scaling_min/max_freq doesn't get written to the shared memory
> region. Due to this, the writes to the scaling_min/max_freq sysfs file
> don't take effect. Fix this by propagating the scaling_min/max_freq
> changes to the shared memory region.
> 
> Fixes: ffa5096a7c33 ("cpufreq: amd-pstate: implement Pstate EPP support for the AMD processors")
> Signed-off-by: Dhananjay Ugwekar <Dhananjay.Ugwekar@....com>
> ---
>   drivers/cpufreq/amd-pstate.c | 10 ++++++++++
>   1 file changed, 10 insertions(+)
> 
> diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c
> index 9ad62dbe8bfb..7c1c96abe5bd 100644
> --- a/drivers/cpufreq/amd-pstate.c
> +++ b/drivers/cpufreq/amd-pstate.c
> @@ -264,6 +264,15 @@ static int amd_pstate_set_epp(struct amd_cpudata *cpudata, u32 epp)
>   			cpudata->epp_cached = epp;
>   	} else {
>   		perf_ctrls.energy_perf = epp;
> +		perf_ctrls.max_perf = cpudata->max_limit_perf;
> +		perf_ctrls.min_perf = cpudata->min_limit_perf;
> +		perf_ctrls.desired_perf = 0U;
> +
> +		ret = cppc_set_perf(cpudata->cpu, &perf_ctrls);
> +		if (ret) {
> +			pr_debug("failed to set min max limits (%d)\n", ret);
> +			return ret;
> +		}

This feels like a handgrown implementation of amd_pstate_update_perf() 
(IE static call updated to cppc_update_perf).

Can you just call that instead?

>   		ret = cppc_set_epp_perf(cpudata->cpu, &perf_ctrls, 1);
>   		if (ret) {
>   			pr_debug("failed to set energy perf value (%d)\n", ret);
> @@ -1547,6 +1556,7 @@ static void amd_pstate_epp_update_limit(struct cpufreq_policy *policy)
>   	}
>   
>   	WRITE_ONCE(cpudata->cppc_req_cached, value);
> +

Spurious newline added here not relevant to this patch.

>   	amd_pstate_set_epp(cpudata, epp);
>   }
>