The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Signed-off-by: Uwe Kleine-König <[email protected]>
---
drivers/mcb/mcb-lpc.c | 6 ++----
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/mcb/mcb-lpc.c b/drivers/mcb/mcb-lpc.c
index a851e0236464..2bec2086ee17 100644
--- a/drivers/mcb/mcb-lpc.c
+++ b/drivers/mcb/mcb-lpc.c
@@ -97,13 +97,11 @@ static int mcb_lpc_probe(struct platform_device *pdev)
return ret;
}
-static int mcb_lpc_remove(struct platform_device *pdev)
+static void mcb_lpc_remove(struct platform_device *pdev)
{
struct priv *priv = platform_get_drvdata(pdev);
mcb_release_bus(priv->bus);
-
- return 0;
}
static struct platform_device *mcb_lpc_pdev;
@@ -140,7 +138,7 @@ static struct platform_driver mcb_lpc_driver = {
.name = "mcb-lpc",
},
.probe = mcb_lpc_probe,
- .remove = mcb_lpc_remove,
+ .remove_new = mcb_lpc_remove,
};
static const struct dmi_system_id mcb_lpc_dmi_table[] = {
base-commit: 11afac187274a6177a7ac82997f8691c0f469e41
--
2.43.0
Hello,
On Thu, Mar 07, 2024 at 06:59:03PM +0100, Uwe Kleine-K?nig wrote:
> The .remove() callback for a platform driver returns an int which makes
> many driver authors wrongly assume it's possible to do error handling by
> returning an error code. However the value returned is ignored (apart
> from emitting a warning) and this typically results in resource leaks.
>
> To improve here there is a quest to make the remove callback return
> void. In the first step of this quest all drivers are converted to
> .remove_new(), which already returns void. Eventually after all drivers
> are converted, .remove_new() will be renamed to .remove().
>
> Trivially convert this driver from always returning zero in the remove
> callback to the void returning variant.
I didn't get any feedback in reply to this patch. My quest to change the
prototype of struct platform_driver::remove depends on that one, so it
would be great if it could make it into the mainline during the next
merge window.
I added Greg to Cc: who's tree the last changes to this driver went in
through. If I don't get feedback until v6.9, I'd ask him to take this
patch without further maintainer feedback.
Best regards
Uwe
--
Pengutronix e.K. | Uwe Kleine-K?nig |
Industrial Linux Solutions | https://www.pengutronix.de/ |