Unverified Commit d073bfa5 authored May 18, 2018 by Tony Lindgren Committed by Mark Brown May 21, 2018

spi: omap2-mcspi: Remove unnecessary pm_runtime_force_suspend()



Commit 5a686b2c ("spi: omap2-mcspi: Idle hardware during suspend
and resume") added calls for pm_runtime_force_suspend() and
pm_runtime_force_resume() to make sure spi is idled between
device_prepare() and device_complete().

But testing Linux next, I now noticed that we will get the following:

spi_master spi0: Failed to power device: -13

Looking at things more turns out we can just remove this non-standard
code. I was probably testing with some extra experimental patches
earlier when I thought we need pm_runtime_force_suspend() and
pm_runtime_force_resume().

Fixes: 5a686b2c ("spi: omap2-mcspi: Idle hardware during suspend
and resume")
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

parent 7e48e23a

drivers/spi/spi-omap2-mcspi.c

+0 −22

Original line number	Diff line number	Diff line
		@@ -1467,20 +1467,6 @@ MODULE_ALIAS("platform:omap2_mcspi");
		#ifdef CONFIG_SUSPEND
		static int omap2_mcspi_suspend_noirq(struct device *dev)
		{
		int error;

		/*
		* Make sure device gets idled if other drivers call SPI
		* functions between device_prepare() and device_complete()
		*/
		error = pm_runtime_force_suspend(dev);
		if (error < 0) {
		dev_err(dev, "%s: force suspend failed: %i\n",
		__func__, error);

		return error;
		}

		return pinctrl_pm_select_sleep_state(dev);
		}

		@@ -1495,14 +1481,6 @@ static int omap2_mcspi_resume_noirq(struct device *dev)
		dev_warn(mcspi->dev, "%s: failed to set pins: %i\n",
		__func__, error);

		error = pm_runtime_force_resume(mcspi->dev);
		if (error < 0) {
		dev_warn(mcspi->dev, "%s: force resume failed: %i\n",
		__func__, error);

		return error;
		}

		return 0;
		}