From: Serge Semin <Sergey.Semin@xxxxxxxxxxxxxxxxxxxx>
Recent commit e61a41256edf ("cpufreq: dev_pm_qos_update_request() can
return 1 on success") fixed a problem when active policies traverse
was falsely stopped due to invalidly treating the non-zero return value
from freq_qos_update_request() method as an error. Yes, that function
can return positive values if the requested update actually took place.
The current problem is that the returned value is then passed to the
return cell of the cpufreq_boost_set_sw() (set_boost callback) method.
This value is then also analyzed for being non-zero, which is also
treated as having an error. As a result during the boost activation
we'll get an error returned while having the QOS frequency update
successfully performed. Fix this by returning a negative value from the
cpufreq_boost_set_sw() if actual error was encountered and zero
otherwise treating any positive values as the successful operations
completion.
Fixes: 18c49926c4bf ("cpufreq: Add QoS requests for userspace constraints")
Signed-off-by: Serge Semin <Sergey.Semin@xxxxxxxxxxxxxxxxxxxx>
Acked-by: Viresh Kumar <viresh.kumar@xxxxxxxxxx>
Cc: Alexey Malahov <Alexey.Malahov@xxxxxxxxxxxxxxxxxxxx>
Cc: Thomas Bogendoerfer <tsbogend@xxxxxxxxxxxxxxxx>
Cc: Paul Burton <paulburton@xxxxxxxxxx>
Cc: Ralf Baechle <ralf@xxxxxxxxxxxxxx>
Cc: Arnd Bergmann <arnd@xxxxxxxx>
Cc: Rob Herring <robh+dt@xxxxxxxxxx>
Cc: linux-mips@xxxxxxxxxxxxxxx
Cc: devicetree@xxxxxxxxxxxxxxx
Cc: stable@xxxxxxxxxxxxxxx
---
drivers/cpufreq/cpufreq.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index 045f9fe157ce..5870cdca88cf 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c
@@ -2554,7 +2554,7 @@ static int cpufreq_boost_set_sw(int state)
break;
}
- return ret;
+ return ret < 0 ? ret : 0;
}
int cpufreq_boost_trigger_state(int state)
--
2.25.1
