On 20/06/16 18:53, Kevin Hilman wrote:
Sudeep Holla <sudeep.ho...@arm.com> writes:
This patch hooks up the support for device power domain provided by
SCPI using the Linux generic power domain infrastructure.
Cc: "Rafael J. Wysocki" <r...@rjwysocki.net>
Cc: Kevin Hilman <khil...@kernel.org>
Cc: Ulf Hansson <ulf.hans...@linaro.org>
Cc: linux...@vger.kernel.org
Signed-off-by: Sudeep Holla <sudeep.ho...@arm.com>
[...]
+static int scpi_pd_power(struct scpi_pm_domain *pd, bool power_on)
+{
+ int ret;
+ enum scpi_power_domain_state state;
+
+ if (power_on)
+ state = SCPI_PD_STATE_ON;
+ else
+ state = SCPI_PD_STATE_OFF;
+
+ ret = pd->ops->device_set_power_state(pd->domain, state);
There should probably be some sanity checks here that these function
pointers are non-NULL.
Yes I agree. Since with the current scpi driver, if scpi_ops is
populated, it's all non-NULL I skipped the sanity check as the probe
will check for non-NULL scpi_ops.
However, with extension work by Neil I agree, we may have to revisit all
such callers.
--
Regards,
Sudeep
