linux-pm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Konrad Dybcio <konrad.dybcio@linaro.org>
To: Andy Gross <agross@kernel.org>,
	Bjorn Andersson <andersson@kernel.org>,
	Georgi Djakov <djakov@kernel.org>,
	Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org>,
	linux-arm-msm@vger.kernel.org, linux-pm@vger.kernel.org,
	linux-kernel@vger.kernel.org,
	Konrad Dybcio <konrad.dybcio@linaro.org>,
	Marijn Suijten <marijn.suijten@somainline.org>
Subject: [PATCH v8 4/8] interconnect: qcom: rpm: Set QoS registers only once
Date: Fri, 07 Apr 2023 22:14:46 +0200	[thread overview]
Message-ID: <20230228-topic-qos-v8-4-ee696a2c15a9@linaro.org> (raw)
In-Reply-To: <20230228-topic-qos-v8-0-ee696a2c15a9@linaro.org>

The QoS registers are (or according to Qualcomm folks, on most
platforms) persistent until a full chip reboot. Move the QoS-setting
functions to the probe function so that we don't needlessly do it over
and over again.

Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
 drivers/interconnect/qcom/icc-rpm.c | 50 ++++++++++++++++---------------------
 1 file changed, 21 insertions(+), 29 deletions(-)

diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c
index 7d62c0bf2722..d79e508cb717 100644
--- a/drivers/interconnect/qcom/icc-rpm.c
+++ b/drivers/interconnect/qcom/icc-rpm.c
@@ -204,30 +204,33 @@ static int qcom_icc_qos_set(struct icc_node *node)
 	}
 }
 
-static int qcom_icc_rpm_set(int mas_rpm_id, int slv_rpm_id, u64 sum_bw)
+static int qcom_icc_rpm_set(struct qcom_icc_node *qn, u64 sum_bw)
 {
 	int ret = 0;
 
-	if (mas_rpm_id != -1) {
+	if (qn->qos.ap_owned)
+		return 0;
+
+	if (qn->mas_rpm_id != -1) {
 		ret = qcom_icc_rpm_smd_send(QCOM_SMD_RPM_ACTIVE_STATE,
 					    RPM_BUS_MASTER_REQ,
-					    mas_rpm_id,
+					    qn->mas_rpm_id,
 					    sum_bw);
 		if (ret) {
 			pr_err("qcom_icc_rpm_smd_send mas %d error %d\n",
-			       mas_rpm_id, ret);
+			       qn->mas_rpm_id, ret);
 			return ret;
 		}
 	}
 
-	if (slv_rpm_id != -1) {
+	if (qn->slv_rpm_id != -1) {
 		ret = qcom_icc_rpm_smd_send(QCOM_SMD_RPM_ACTIVE_STATE,
 					    RPM_BUS_SLAVE_REQ,
-					    slv_rpm_id,
+					    qn->slv_rpm_id,
 					    sum_bw);
 		if (ret) {
 			pr_err("qcom_icc_rpm_smd_send slv %d error %d\n",
-			       slv_rpm_id, ret);
+			       qn->slv_rpm_id, ret);
 			return ret;
 		}
 	}
@@ -235,26 +238,6 @@ static int qcom_icc_rpm_set(int mas_rpm_id, int slv_rpm_id, u64 sum_bw)
 	return ret;
 }
 
-static int __qcom_icc_set(struct icc_node *n, struct qcom_icc_node *qn,
-			  u64 sum_bw)
-{
-	int ret;
-
-	if (!qn->qos.ap_owned) {
-		/* send bandwidth request message to the RPM processor */
-		ret = qcom_icc_rpm_set(qn->mas_rpm_id, qn->slv_rpm_id, sum_bw);
-		if (ret)
-			return ret;
-	} else if (qn->qos.qos_mode != NOC_QOS_MODE_INVALID) {
-		/* set bandwidth directly from the AP */
-		ret = qcom_icc_qos_set(n, sum_bw);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
 /**
  * qcom_icc_pre_bw_aggregate - cleans up values before re-aggregate requests
  * @node: icc node to operate on
@@ -370,11 +353,12 @@ static int qcom_icc_set(struct icc_node *src, struct icc_node *dst)
 
 	sum_bw = icc_units_to_bps(max_agg_avg);
 
-	ret = __qcom_icc_set(src, src_qn, sum_bw);
+	ret = qcom_icc_rpm_set(src_qn, sum_bw);
 	if (ret)
 		return ret;
+
 	if (dst_qn) {
-		ret = __qcom_icc_set(dst, dst_qn, sum_bw);
+		ret = qcom_icc_rpm_set(dst_qn, sum_bw);
 		if (ret)
 			return ret;
 	}
@@ -528,6 +512,14 @@ int qnoc_probe(struct platform_device *pdev)
 		for (j = 0; j < qnodes[i]->num_links; j++)
 			icc_link_create(node, qnodes[i]->links[j]);
 
+		/* Set QoS registers (we only need to do it once, generally) */
+		if (qnodes[i]->qos.ap_owned &&
+		    qnodes[i]->qos.qos_mode != NOC_QOS_MODE_INVALID) {
+			ret = qcom_icc_qos_set(node);
+			if (ret)
+				return ret;
+		}
+
 		data->nodes[i] = node;
 	}
 	data->num_nodes = num_nodes;

-- 
2.40.0


  parent reply	other threads:[~2023-04-07 20:14 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-04-07 20:14 [PATCH v8 0/8] The great interconnecification fixation Konrad Dybcio
2023-04-07 20:14 ` [PATCH v8 1/8] interconnect: qcom: rpm: Rename icc desc clocks to bus_blocks Konrad Dybcio
2023-04-07 20:14 ` [PATCH v8 2/8] interconnect: qcom: rpm: Rename icc provider num_clocks to num_bus_clocks Konrad Dybcio
2023-04-07 20:14 ` [PATCH v8 3/8] interconnect: qcom: rpm: Drop unused parameters Konrad Dybcio
2023-04-07 20:14 ` Konrad Dybcio [this message]
2023-04-07 20:14 ` [PATCH v8 5/8] interconnect: qcom: rpm: Handle interface clocks Konrad Dybcio
2023-05-18 15:59   ` Georgi Djakov
2023-05-18 19:55     ` Konrad Dybcio
2023-04-07 20:14 ` [PATCH v8 6/8] interconnect: qcom: icc-rpm: Enforce 2 or 0 bus clocks Konrad Dybcio
2023-04-07 20:14 ` [PATCH v8 7/8] interconnect: qcom: rpm: Don't use clk_get_optional for bus clocks anymore Konrad Dybcio
2023-04-07 20:14 ` [PATCH v8 8/8] interconnect: qcom: msm8996: Promote to core_initcall Konrad Dybcio
2023-05-17 16:21 ` [PATCH v8 0/8] The great interconnecification fixation Konrad Dybcio

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230228-topic-qos-v8-4-ee696a2c15a9@linaro.org \
    --to=konrad.dybcio@linaro.org \
    --cc=agross@kernel.org \
    --cc=andersson@kernel.org \
    --cc=bryan.odonoghue@linaro.org \
    --cc=djakov@kernel.org \
    --cc=dmitry.baryshkov@linaro.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=marijn.suijten@somainline.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).