video: tegra: host: balance pm_runtime calls

pm_runtime_get_sync increments usage count first and
then carry out actual resume operation. if pm_runtime_get_sync
failed needs to decrement usage count to balance count.

Bug 1459733

Change-Id: I844c59310b992c7cac12b134c772f30857c5dec3
Signed-off-by: Shridhar Rasal <srasal@nvidia.com>
Reviewed-on: http://git-master/r/368416
Reviewed-on: http://git-master/r/411248
Reviewed-by: Sachin Nikam <snikam@nvidia.com>
Tested-by: Sachin Nikam <snikam@nvidia.com>

diff --git a/drivers/video/tegra/host/nvhost_acm.c b/drivers/video/tegra/host/nvhost_acm.c
index 3c92624fc6ae7144e81299a4714fb6498c04c15d..b85426320a77b9a625e17bf4de87dd93bbb1e7d5 100644 (file)
--- a/drivers/video/tegra/host/nvhost_acm.c
+++ b/drivers/video/tegra/host/nvhost_acm.c
@@ -169,6 +169,7 @@ int nvhost_module_busy(struct platform_device *dev)
 #ifdef CONFIG_PM_RUNTIME
        ret = pm_runtime_get_sync(&dev->dev);
        if (ret < 0) {
+               pm_runtime_put_noidle(&dev->dev);
                if (dev->dev.parent && (dev->dev.parent != &platform_bus))
                        nvhost_module_idle(nvhost_get_parent(dev));
                nvhost_err(&dev->dev, "failed to power on, err %d", ret);