[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20230315132330.450877-3-abel.vesa@linaro.org>
Date: Wed, 15 Mar 2023 15:23:27 +0200
From: Abel Vesa <abel.vesa@...aro.org>
To: "Rafael J. Wysocki" <rafael@...nel.org>,
Kevin Hilman <khilman@...nel.org>,
Ulf Hansson <ulf.hansson@...aro.org>,
Pavel Machek <pavel@....cz>, Len Brown <len.brown@...el.com>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Bjorn Andersson <andersson@...nel.org>,
Andy Gross <agross@...nel.org>,
Konrad Dybcio <konrad.dybcio@...aro.org>,
Mike Turquette <mturquette@...libre.com>,
Stephen Boyd <sboyd@...nel.org>,
Saravana Kannan <saravanak@...gle.com>
Cc: linux-pm@...r.kernel.org,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
linux-arm-msm@...r.kernel.org, linux-clk@...r.kernel.org,
Doug Anderson <dianders@...omium.org>,
Matthias Kaehlcke <mka@...omium.org>
Subject: [RFC PATCH 2/5] soc: qcom: rpmhpd: Do proper power off when state synced
Instead of aggregating different corner values on sync state callback,
call the genpd API for queuing up the power off. This will also mark the
domain as powered off in the debugfs genpd summary. Also, until sync
state has been reached, return busy on power off request, in order to
allow genpd core to know that the actual domain hasn't been powered of
from the "disable unused" late initcall.
Signed-off-by: Abel Vesa <abel.vesa@...aro.org>
---
drivers/soc/qcom/rpmhpd.c | 19 +++++++------------
1 file changed, 7 insertions(+), 12 deletions(-)
diff --git a/drivers/soc/qcom/rpmhpd.c b/drivers/soc/qcom/rpmhpd.c
index f20e2a49a669..abd999c74783 100644
--- a/drivers/soc/qcom/rpmhpd.c
+++ b/drivers/soc/qcom/rpmhpd.c
@@ -649,8 +649,12 @@ static int rpmhpd_power_off(struct generic_pm_domain *domain)
mutex_lock(&rpmhpd_lock);
ret = rpmhpd_aggregate_corner(pd, 0);
- if (!ret)
- pd->enabled = false;
+ if (!ret) {
+ if (!pd->state_synced)
+ ret = -EBUSY;
+ else
+ pd->enabled = false;
+ }
mutex_unlock(&rpmhpd_lock);
@@ -810,10 +814,8 @@ static void rpmhpd_sync_state(struct device *dev)
{
const struct rpmhpd_desc *desc = of_device_get_match_data(dev);
struct rpmhpd **rpmhpds = desc->rpmhpds;
- unsigned int corner;
struct rpmhpd *pd;
unsigned int i;
- int ret;
mutex_lock(&rpmhpd_lock);
for (i = 0; i < desc->num_pds; i++) {
@@ -822,14 +824,7 @@ static void rpmhpd_sync_state(struct device *dev)
continue;
pd->state_synced = true;
- if (pd->enabled)
- corner = max(pd->corner, pd->enable_corner);
- else
- corner = 0;
-
- ret = rpmhpd_aggregate_corner(pd, corner);
- if (ret)
- dev_err(dev, "failed to sync %s\n", pd->res_name);
+ genpd_queue_power_off_work(&pd->pd);
}
mutex_unlock(&rpmhpd_lock);
}
--
2.34.1
Powered by blists - more mailing lists