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 (mostly) ignored
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.
wlcore_remove() returned zero unconditionally. With that converted to
return void instead, the wl12xx and wl18xx driver can be converted to
.remove_new trivially.
Signed-off-by: Uwe Kleine-König <[email protected]>
---
drivers/net/wireless/ti/wl12xx/main.c | 6 +++---
drivers/net/wireless/ti/wl18xx/main.c | 2 +-
drivers/net/wireless/ti/wlcore/main.c | 6 ++----
drivers/net/wireless/ti/wlcore/wlcore.h | 2 +-
4 files changed, 7 insertions(+), 9 deletions(-)
diff --git a/drivers/net/wireless/ti/wl12xx/main.c b/drivers/net/wireless/ti/wl12xx/main.c
index d06a2c419447..de045fe4ca1e 100644
--- a/drivers/net/wireless/ti/wl12xx/main.c
+++ b/drivers/net/wireless/ti/wl12xx/main.c
@@ -1919,7 +1919,7 @@ static int wl12xx_probe(struct platform_device *pdev)
return ret;
}
-static int wl12xx_remove(struct platform_device *pdev)
+static void wl12xx_remove(struct platform_device *pdev)
{
struct wl1271 *wl = platform_get_drvdata(pdev);
struct wl12xx_priv *priv;
@@ -1928,7 +1928,7 @@ static int wl12xx_remove(struct platform_device *pdev)
kfree(priv->rx_mem_addr);
- return wlcore_remove(pdev);
+ wlcore_remove(pdev);
}
static const struct platform_device_id wl12xx_id_table[] = {
@@ -1939,7 +1939,7 @@ MODULE_DEVICE_TABLE(platform, wl12xx_id_table);
static struct platform_driver wl12xx_driver = {
.probe = wl12xx_probe,
- .remove = wl12xx_remove,
+ .remove_new = wl12xx_remove,
.id_table = wl12xx_id_table,
.driver = {
.name = "wl12xx_driver",
diff --git a/drivers/net/wireless/ti/wl18xx/main.c b/drivers/net/wireless/ti/wl18xx/main.c
index 0b3cf8477c6c..d4a89401f2c4 100644
--- a/drivers/net/wireless/ti/wl18xx/main.c
+++ b/drivers/net/wireless/ti/wl18xx/main.c
@@ -2033,7 +2033,7 @@ MODULE_DEVICE_TABLE(platform, wl18xx_id_table);
static struct platform_driver wl18xx_driver = {
.probe = wl18xx_probe,
- .remove = wlcore_remove,
+ .remove_new = wlcore_remove,
.id_table = wl18xx_id_table,
.driver = {
.name = "wl18xx_driver",
diff --git a/drivers/net/wireless/ti/wlcore/main.c b/drivers/net/wireless/ti/wlcore/main.c
index bf21611872a3..9dfd832b1d06 100644
--- a/drivers/net/wireless/ti/wlcore/main.c
+++ b/drivers/net/wireless/ti/wlcore/main.c
@@ -6737,7 +6737,7 @@ int wlcore_probe(struct wl1271 *wl, struct platform_device *pdev)
}
EXPORT_SYMBOL_GPL(wlcore_probe);
-int wlcore_remove(struct platform_device *pdev)
+void wlcore_remove(struct platform_device *pdev)
{
struct wlcore_platdev_data *pdev_data = dev_get_platdata(&pdev->dev);
struct wl1271 *wl = platform_get_drvdata(pdev);
@@ -6752,7 +6752,7 @@ int wlcore_remove(struct platform_device *pdev)
if (pdev_data->family && pdev_data->family->nvs_name)
wait_for_completion(&wl->nvs_loading_complete);
if (!wl->initialized)
- return 0;
+ return;
if (wl->wakeirq >= 0) {
dev_pm_clear_wake_irq(wl->dev);
@@ -6772,8 +6772,6 @@ int wlcore_remove(struct platform_device *pdev)
free_irq(wl->irq, wl);
wlcore_free_hw(wl);
-
- return 0;
}
EXPORT_SYMBOL_GPL(wlcore_remove);
diff --git a/drivers/net/wireless/ti/wlcore/wlcore.h b/drivers/net/wireless/ti/wlcore/wlcore.h
index 81c94d390623..1f8511bf9bb3 100644
--- a/drivers/net/wireless/ti/wlcore/wlcore.h
+++ b/drivers/net/wireless/ti/wlcore/wlcore.h
@@ -497,7 +497,7 @@ struct wl1271 {
};
int wlcore_probe(struct wl1271 *wl, struct platform_device *pdev);
-int wlcore_remove(struct platform_device *pdev);
+void wlcore_remove(struct platform_device *pdev);
struct ieee80211_hw *wlcore_alloc_hw(size_t priv_size, u32 aggr_buf_size,
u32 mbox_size);
int wlcore_free_hw(struct wl1271 *wl);
base-commit: 0bb80ecc33a8fb5a682236443c1e740d5c917d1d
--
2.40.1
Uwe Kleine-König <[email protected]> writes:
> 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 (mostly) ignored
> 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.
>
> wlcore_remove() returned zero unconditionally. With that converted to
> return void instead, the wl12xx and wl18xx driver can be converted to
> .remove_new trivially.
>
> Signed-off-by: Uwe Kleine-König <[email protected]>
> ---
> drivers/net/wireless/ti/wl12xx/main.c | 6 +++---
> drivers/net/wireless/ti/wl18xx/main.c | 2 +-
> drivers/net/wireless/ti/wlcore/main.c | 6 ++----
> drivers/net/wireless/ti/wlcore/wlcore.h | 2 +-
> 4 files changed, 7 insertions(+), 9 deletions(-)
wireless patches go to wireless-next, not net-next. But no need to
resend because of this.
--
https://patchwork.kernel.org/project/linux-wireless/list/
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
Hello,
On Tue, Sep 12, 2023 at 08:23:18PM +0300, Kalle Valo wrote:
> Uwe Kleine-K?nig <[email protected]> writes:
>
> > 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 (mostly) ignored
> > 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.
> >
> > wlcore_remove() returned zero unconditionally. With that converted to
> > return void instead, the wl12xx and wl18xx driver can be converted to
> > .remove_new trivially.
> >
> > Signed-off-by: Uwe Kleine-K?nig <[email protected]>
> > ---
> > drivers/net/wireless/ti/wl12xx/main.c | 6 +++---
> > drivers/net/wireless/ti/wl18xx/main.c | 2 +-
> > drivers/net/wireless/ti/wlcore/main.c | 6 ++----
> > drivers/net/wireless/ti/wlcore/wlcore.h | 2 +-
> > 4 files changed, 7 insertions(+), 9 deletions(-)
>
> wireless patches go to wireless-next, not net-next. But no need to
> resend because of this.
So for the next patch to drivers/net/wireless: I should write "[PATCH
wireless-next]" in the Subject? Do the other special rules for net-next
apply to wireless-next, too? (E.g. that I must not send patches for
-next during the merge window and the rules about comments.)
Best regards
Uwe
--
Pengutronix e.K. | Uwe Kleine-K?nig |
Industrial Linux Solutions | https://www.pengutronix.de/ |
Uwe Kleine-König <[email protected]> writes:
> Hello,
>
> On Tue, Sep 12, 2023 at 08:23:18PM +0300, Kalle Valo wrote:
>> Uwe Kleine-König <[email protected]> writes:
>>
>> > 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 (mostly) ignored
>> > 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.
>> >
>> > wlcore_remove() returned zero unconditionally. With that converted to
>> > return void instead, the wl12xx and wl18xx driver can be converted to
>> > .remove_new trivially.
>> >
>> > Signed-off-by: Uwe Kleine-König <[email protected]>
>> > ---
>> > drivers/net/wireless/ti/wl12xx/main.c | 6 +++---
>> > drivers/net/wireless/ti/wl18xx/main.c | 2 +-
>> > drivers/net/wireless/ti/wlcore/main.c | 6 ++----
>> > drivers/net/wireless/ti/wlcore/wlcore.h | 2 +-
>> > 4 files changed, 7 insertions(+), 9 deletions(-)
>>
>> wireless patches go to wireless-next, not net-next. But no need to
>> resend because of this.
>
> So for the next patch to drivers/net/wireless: I should write "[PATCH
> wireless-next]" in the Subject?
That's not a requirement from our side but feel free to add that if you
want. Usually we do so that if the patch has "[PATCH wireless]" or
"[PATCH v6.6]" we queue the patch to wireless tree, otherwise it's
queued for wireless-next.
> Do the other special rules for net-next apply to wireless-next, too?
> (E.g. that I must not send patches for -next during the merge window
> and the rules about comments.)
Yeah, we do work a bit different compared to net-next. For example,
wireless-next remains open during merge windows so you can submit
patches anytime. We have tried to document our practises in the wiki
link below in my signature.
--
https://patchwork.kernel.org/project/linux-wireless/list/
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
Uwe Kleine-König <[email protected]> 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 (mostly) ignored
> 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.
>
> wlcore_remove() returned zero unconditionally. With that converted to
> return void instead, the wl12xx and wl18xx driver can be converted to
> .remove_new trivially.
>
> Signed-off-by: Uwe Kleine-König <[email protected]>
Patch applied to wireless-next.git, thanks.
f00928012886 wifi: wlcore: Convert to platform remove callback returning void
--
https://patchwork.kernel.org/project/linux-wireless/patch/[email protected]/
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches