crypto: tegra-se: put device into runtime suspend state when suspending

To prevent the device being accidentally put into runtime suspend state
during the whole system suspend process, pm_runtime_get_noresume() is
called upon _every_ device to increase the usage count (please refer to
drivers/base/power/main.c).

Since we don't explicitly disable the clock, pm_runtime_put_sync() in
each operation in tegra_se_suspend() will not actually call the runtime
suspend function, thus leaving the clock still enabled.

To fix this issue in a simple way, we call pm_runtime_put_sync() in the
end of tegra_se_suspend() to decrease the usage count to "0" and thus
call tegra_se_runtime_suspend() in turn to disable the clock. To pair
the usage count, we do a pm_runtime_get_noresume() in tegra_se_resume()
as we don't actually need to runtime resume the device there.

Bug 1246029

Change-Id: I64520b022b896f2867934255a55b852fafac4b63
Signed-off-by: Eric Miao <emiao@nvidia.com>
Reviewed-on: http://git-master/r/206658
GVS: Gerrit_Virtual_Submit
Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>

0 files changed, 0 insertions, 0 deletions