[PATCH] wlcore: sdio: Fix flakey SDIO runtime PM handling

[PATCH] wlcore: sdio: Fix flakey SDIO runtime PM handling

[Tony Lindgren]

We can have pm_runtime_get_sync() return 1, and we can have
pm_runtime_put_sync() return -EBUSY. See rpm_suspend() and
rpm_resume() for more information.

Fix the issue by returning 0 from wl12xx_sdio_power_on() on success.
And use pm_runtime_put() instead of pm_runtime_put_sync() for
wl12xx_sdio_power_off(), then the MMC subsystem will idle the bus
when suitable.

Otherwise wlcore can sometimes get confused and may report bogus
errors and WLAN connection can fail.

Fixes: 60f36637bbbd ("wlcore: sdio: allow pm to handle sdio power")
Signed-off-by: Tony Lindgren <tony-4v6yS6AI5VpBDgjK7y7TUQ@public.gmane.org>
---
 drivers/net/wireless/ti/wlcore/sdio.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/net/wireless/ti/wlcore/sdio.c b/drivers/net/wireless/ti/wlcore/sdio.c
--- a/drivers/net/wireless/ti/wlcore/sdio.c
+++ b/drivers/net/wireless/ti/wlcore/sdio.c
@@ -159,15 +159,15 @@ static int wl12xx_sdio_power_on(struct wl12xx_sdio_glue *glue)
 		pm_runtime_put_noidle(&card->dev);
 		dev_err(glue->dev, "%s: failed to get_sync(%d)\n",
 			__func__, ret);
-		goto out;
+
+		return ret;
 	}
 
 	sdio_claim_host(func);
 	sdio_enable_func(func);
 	sdio_release_host(func);
 
-out:
-	return ret;
+	return 0;
 }
 
 static int wl12xx_sdio_power_off(struct wl12xx_sdio_glue *glue)
@@ -180,7 +180,7 @@ static int wl12xx_sdio_power_off(struct wl12xx_sdio_glue *glue)
 	sdio_release_host(func);
 
 	/* Let runtime PM know the card is powered off */
-	return pm_runtime_put_sync(&card->dev);
+	return pm_runtime_put(&card->dev);
 }
 
 static int wl12xx_sdio_set_power(struct device *child, bool enable)
-- 
2.17.0

Re: [PATCH] wlcore: sdio: Fix flakey SDIO runtime PM handling

[Tony Lindgren]

* Tony Lindgren <tony-4v6yS6AI5VpBDgjK7y7TUQ@public.gmane.org> [180515 15:46]:
> We can have pm_runtime_get_sync() return 1, and we can have
> pm_runtime_put_sync() return -EBUSY. See rpm_suspend() and
> rpm_resume() for more information.
> 
> Fix the issue by returning 0 from wl12xx_sdio_power_on() on success.
> And use pm_runtime_put() instead of pm_runtime_put_sync() for
> wl12xx_sdio_power_off(), then the MMC subsystem will idle the bus
> when suitable.
> 
> Otherwise wlcore can sometimes get confused and may report bogus
> errors and WLAN connection can fail.
> @@ -180,7 +180,7 @@ static int wl12xx_sdio_power_off(struct wl12xx_sdio_glue *glue)
>  	sdio_release_host(func);
>  
>  	/* Let runtime PM know the card is powered off */
> -	return pm_runtime_put_sync(&card->dev);
> +	return pm_runtime_put(&card->dev);
>  }

Looks like we can still get -EBUSY here though. I think we
can just ignore the return value of pm_runtime_put() here.

Regards,

Tony