From mboxrd@z Thu Jan 1 00:00:00 1970 From: Philippe Longepe Subject: [PATCH v1 2/2] intel_pstate: Change the setpoint for the cores Date: Tue, 3 Nov 2015 10:27:20 +0100 Message-ID: <1446542840-14982-2-git-send-email-philippe.longepe@linux.intel.com> References: <1446542840-14982-1-git-send-email-philippe.longepe@linux.intel.com> Return-path: Received: from mga02.intel.com ([134.134.136.20]:49435 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751568AbbKCJ00 (ORCPT ); Tue, 3 Nov 2015 04:26:26 -0500 In-Reply-To: <1446542840-14982-1-git-send-email-philippe.longepe@linux.intel.com> Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: linux-pm@vger.kernel.org Cc: srinivas.pandruvada@linux.intel.com, Stephane Gasparini Change the setpoint to 60 accordingly to the new core busy scaled formula. The new formaula is based on the number of cycles per seconds (average frequency) divided by the requested frequency. So, we need to chose a setpoint more aggressive to improve performance. Measured with this parameter, we noticed an improvement in Browsermark for power and perf compared to the old formula: Score without the patch: 3517 Power without the patch: 6856 mW Score with the patch: 3719 Power with the patch: 6265 mW Signed-off-by: Philippe Longepe Signed-off-by: Stephane Gasparini --- drivers/cpufreq/intel_pstate.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 421903f..43f0067 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -872,7 +872,7 @@ static struct cpu_defaults core_params = { .pid_policy = { .sample_rate_ms = 10, .deadband = 0, - .setpoint = 97, + .setpoint = 60, .p_gain_pct = 20, .d_gain_pct = 0, .i_gain_pct = 0, -- 1.9.1