| Message ID | 20240306211947.97103-2-u.kleine-koenig@pengutronix.de (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | sh: push-switch: Convert to platform remove callback returning void | expand |
On Wed, Mar 6, 2024 at 10:20 PM Uwe Kleine-König <u.kleine-koenig@pengutronix.de> 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. > > Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Gr{oetje,eeting}s, Geert
Hello, [corrected the address of the Pengutronix kernel team in Cc:] On Wed, Mar 06, 2024 at 10:19:47PM +0100, Uwe Kleine-König 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. Apart from Geert's positive review reply I didn't get any feedback on this patch and it didn't appear in next yet. As my quest to change the prototype of struct platform_driver::remove depends on these patches, it would be great if they made it in during the next merge window. Best regards Uwe
On Thu, 2024-04-11 at 09:18 +0200, Uwe Kleine-König wrote: > Apart from Geert's positive review reply I didn't get any feedback on > this patch and it didn't appear in next yet. The plan was to get everything approved before merging it for next as a partial conversion to device tree for SH would probably cause problems. Adrian
On Thu, 2024-04-11 at 09:47 +0200, John Paul Adrian Glaubitz wrote: > On Thu, 2024-04-11 at 09:18 +0200, Uwe Kleine-König wrote: > > Apart from Geert's positive review reply I didn't get any feedback on > > this patch and it didn't appear in next yet. > > The plan was to get everything approved before merging it for next > as a partial conversion to device tree for SH would probably cause > problems. Oops, sorry. I confused patches here. I'll pick it up for 6.10. Adrian
diff --git a/arch/sh/drivers/push-switch.c b/arch/sh/drivers/push-switch.c index 6ecba5f521eb..362e4860bf52 100644 --- a/arch/sh/drivers/push-switch.c +++ b/arch/sh/drivers/push-switch.c @@ -91,7 +91,7 @@ static int switch_drv_probe(struct platform_device *pdev) return ret; } -static int switch_drv_remove(struct platform_device *pdev) +static void switch_drv_remove(struct platform_device *pdev) { struct push_switch *psw = platform_get_drvdata(pdev); struct push_switch_platform_info *psw_info = pdev->dev.platform_data; @@ -106,13 +106,11 @@ static int switch_drv_remove(struct platform_device *pdev) free_irq(irq, pdev); kfree(psw); - - return 0; } static struct platform_driver switch_driver = { .probe = switch_drv_probe, - .remove = switch_drv_remove, + .remove_new = switch_drv_remove, .driver = { .name = DRV_NAME, },
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 <u.kleine-koenig@pengutronix.de> --- arch/sh/drivers/push-switch.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) base-commit: 11afac187274a6177a7ac82997f8691c0f469e41