Re: [RESEND PATCH v4 3/3] mmc: sdhci-of-arasan: add runtime pm support
From: Shawn Lin
Date: Thu Oct 22 2015 - 23:44:31 EST
On 2015/10/22 20:08, Ulf Hansson wrote:
On 22 October 2015 at 11:06, Shawn Lin <shawn.lin@xxxxxxxxxxxxxx> wrote:
This patch add runtime_suspend and runtime_resume for
sdhci-of-arasan. Currently we also power-off phy at
runtime_suspend for power-saving.
Signed-off-by: Shawn Lin <shawn.lin@xxxxxxxxxxxxxx>
Serise-changes: 4
- remove ifdef for PM callback statement
- fix missing pm_runtime_set_active
- remove pm_runtime_dont_use_autosuspend from remove hook
- add pm_runtime_force_suspend|resume for PM callback to
deal with suspend invoked from rpm
- remove wrappers of phy ops
---
Changes in v2: None
drivers/mmc/host/sdhci-of-arasan.c | 85 ++++++++++++++++++++++++++++++++++++--
1 file changed, 82 insertions(+), 3 deletions(-)
diff --git a/drivers/mmc/host/sdhci-of-arasan.c b/drivers/mmc/host/sdhci-of-arasan.c
index 4f30716..fb1915c 100644
--- a/drivers/mmc/host/sdhci-of-arasan.c
+++ b/drivers/mmc/host/sdhci-of-arasan.c
@@ -30,6 +30,8 @@
#define CLK_CTRL_TIMEOUT_MASK (0xf << CLK_CTRL_TIMEOUT_SHIFT)
#define CLK_CTRL_TIMEOUT_MIN_EXP 13
+#define ARASAN_RPM_DELAY_MS 50
+
/**
* struct sdhci_arasan_data
* @clk_ahb: Pointer to the AHB clock
@@ -71,6 +73,46 @@ static struct sdhci_pltfm_data sdhci_arasan_pdata = {
SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN,
};
+#ifdef CONFIG_PM
+static int sdhci_arasan_runtime_suspend(struct device *dev)
+{
+ struct sdhci_host *host = dev_get_drvdata(dev);
+ struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+ struct sdhci_arasan_data *sdhci_arasan = pltfm_host->priv;
+ int ret;
+
+ ret = sdhci_runtime_suspend_host(host);
+ if (ret)
+ return ret;
+
+ if (!IS_ERR(sdhci_arasan->phy))
+ phy_power_off(sdhci_arasan->phy);
+
+ clk_disable(sdhci_arasan->clk_ahb);
+ clk_disable(pltfm_host->clk);
The above calls can be changed to clk_disable_unprepare(). Vice verse
in sdhci_arasan_runtime_resume() below.
+
+ return 0;
+}
+
+static int sdhci_arasan_runtime_resume(struct device *dev)
+{
+ struct sdhci_host *host = dev_get_drvdata(dev);
+ struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+ struct sdhci_arasan_data *sdhci_arasan = pltfm_host->priv;
+
+ clk_enable(pltfm_host->clk);
+ clk_enable(sdhci_arasan->clk_ahb);
+
+ if (!IS_ERR(sdhci_arasan->phy))
+ phy_power_on(sdhci_arasan->phy);
+
+ return sdhci_runtime_resume_host(host);
+}
+#else
+#define sdhci_arasan_runtime_suspend NULL
+#define sdhci_arasan_runtime_resume NULL
+#endif
+
#ifdef CONFIG_PM_SLEEP
/**
* sdhci_arasan_suspend - Suspend method for the driver
@@ -87,6 +129,12 @@ static int sdhci_arasan_suspend(struct device *dev)
struct sdhci_arasan_data *sdhci_arasan = pltfm_host->priv;
int ret;
+ ret = pm_runtime_force_suspend(dev);
+ if (ret) {
+ dev_err(dev, "problem force suspending\n");
+ return ret;
+ }
+
I think your are in right track, but unfortunate you still have some
work to do. :-)
You can't call sdhci_suspend_host() with a runtime suspended host. You
will for example access sdhci internal registers, even-though the
clock to the SDHCI controller is gated, which is not a good idea.
Now, this leads me to the following questions/ideas, which I am really
glad we can bring up within this context.
So, sdhci_suspend_host() and sdhci_runtime_suspend_host() performs
very similar actions, but not completely the same.
I have a feeling that we might be able to merge the above functions
into one function. The only thing that may differ between system PM
and runtime PM, is probably the wakeup settings and since that is
possible to control before the sdhci variant host driver decide to
call these functions, merging them might work.
From a wakeup point of view, does your case have different settings
between system PM and runtime PM?
What do you think?
My case doesn't have diff settings between system PM and runtime PM.
If my understanding is proper, I guess you want to remove
sdhci_suspend|resume_host and assign pm_runtime_force_suspend|resume to
SET_SYSTEM_SLEEP_PM_OPS directly by sdhci variant HCs?
If that is the case, I guess we should look more precisely into other
sdhci variant HCs to check whether they need to do diff things for
system PM and RPM or not.
ret = sdhci_suspend_host(host);
if (ret)
return ret;
@@ -140,18 +188,39 @@ static int sdhci_arasan_resume(struct device *dev)
}
}
- return sdhci_resume_host(host);
+ ret = sdhci_resume_host(host);
+ if (ret)
+ goto err_resume_host;
+
+ ret = pm_runtime_force_resume(dev);
+ if (ret) {
+ dev_err(dev, "problem force resuming\n");
+ goto err_force_resume;
+ }
+
+ return 0;
+err_force_resume:
+ sdhci_suspend_host(host);
+err_resume_host:
+ if (!IS_ERR(sdhci_arasan->phy))
+ phy_power_off(sdhci_arasan->phy);
err_phy_power:
clk_disable(pltfm_host->clk);
err_clk_en:
clk_disable(sdhci_arasan->clk_ahb);
return ret;
}
+#else
+#define sdhci_arasan_suspend NULL
+#define sdhci_arasan_resume NULL
#endif /* ! CONFIG_PM_SLEEP */
-static SIMPLE_DEV_PM_OPS(sdhci_arasan_dev_pm_ops, sdhci_arasan_suspend,
- sdhci_arasan_resume);
+static const struct dev_pm_ops sdhci_arasan_dev_pm_ops = {
+ SET_SYSTEM_SLEEP_PM_OPS(sdhci_arasan_suspend, sdhci_arasan_resume)
+ SET_RUNTIME_PM_OPS(sdhci_arasan_runtime_suspend,
+ sdhci_arasan_runtime_resume, NULL)
+};
static int sdhci_arasan_probe(struct platform_device *pdev)
{
@@ -237,6 +306,12 @@ static int sdhci_arasan_probe(struct platform_device *pdev)
}
}
+ pm_runtime_set_active(&pdev->dev);
+ pm_runtime_set_autosuspend_delay(&pdev->dev, ARASAN_RPM_DELAY_MS);
+ pm_runtime_use_autosuspend(&pdev->dev);
+ pm_runtime_enable(&pdev->dev);
+ pm_suspend_ignore_children(&pdev->dev, 1);
+
ret = sdhci_add_host(host);
if (ret)
goto err_pltfm_free;
@@ -244,6 +319,7 @@ static int sdhci_arasan_probe(struct platform_device *pdev)
return 0;
err_pltfm_free:
+ pm_runtime_disable(&pdev->dev);
sdhci_pltfm_free(pdev);
err_phy_power:
if (!IS_ERR(sdhci_arasan->phy))
@@ -265,6 +341,9 @@ static int sdhci_arasan_remove(struct platform_device *pdev)
if (!IS_ERR(sdhci_arasan->phy))
phy_exit(sdhci_arasan->phy);
+ pm_runtime_get_sync(&pdev->dev);
+ pm_runtime_disable(&pdev->dev);
+
I think a pm_runtime_put_noidle() would be good as well, since it
restores the usage counter.
clk_disable_unprepare(sdhci_arasan->clk_ahb);
return sdhci_pltfm_unregister(pdev);
--
2.3.7
Kind regards
Uffe
--
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
--
Best Regards
Shawn Lin
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/