Re: [PATCH 5/6] i2c: tegra: Add runtime power-management support

[Laxman Dewangan <ldewangan-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>]

On Thursday 11 August 2016 03:46 PM, Jon Hunter wrote:
> Update the Tegra I2C driver to use runtime PM and move the code in the
> tegra_i2c_clock_enable/disable() functions to the PM runtime resume and
> suspend callbacks, respectively.
>
> Note that given that CONFIG_PM is not mandatory for Tegra, if CONFIG_PM
> is not enabled and so runtime PM is not enabled, ensure that the I2C
> clocks are turned on during probe and kept on by calling the resume
> callback directly.
>
> In the function tegra_i2c_init(), the variable 'err' does not need to be
> initialised to zero in tegra_i2c_init() because it is initialised when
> pm_runtime_get_sync() is called. Furthermore, to ensure we only return 0
> from tegra_i2c_init(), it is necessary to re-initialise 'err' to 0 after
> a successful call to pm_runtime_get_sync() because it can return a
> positive value on success. However, alternatively re-initialise 'err' by
> using the return value of the function tegra_i2c_flush_fifos() because
> it can only be 0 or -ETIMEDOUT.
>
> Signed-off-by: Jon Hunter <jonathanh-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
>
Acked-by: Laxman Dewangan <ldewangan-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>