I recommend to replace the word “pm_runtime” by the
alternative “PM run time system” in the patch subject.
> Calling pm_runtime_get_sync increments the counter even in case of
> failure, causing incorrect ref count if pm_runtime_put is not called in
> error handling paths.
Should the term “reference count” be used here?
> Call pm_runtime_put if pm_runtime_get_sync fails.
The diff hunks show an other function name.
…
> +++ b/drivers/gpio/gpio-arizona.c
> @@ -64,6 +64,7 @@ static int arizona_gpio_get(struct gpio_chip *chip,
> unsigned offset)
> ret = pm_runtime_get_sync(chip->parent);
> if (ret < 0) {
> dev_err(chip->parent, "Failed to resume: %d\n", ret);
> + pm_runtime_put_autosuspend(chip->parent);
> return ret;
> }
You propose to use identical statements in three if branches.
Please add a corresponding jump target for better exception handling.
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/coding-style.rst?id=435faf5c218a47fd6258187f62d9bb1009717896#n455
Would you like to add the tag “Fixes” to the commit message?
I find it amazing how many questionable implementation details
you pointed out recently.
Were these contributions triggered by an evolving source code analysis
tool like CheQ?
https://github.com/umnsec/cheq/
Regards,
Markus
