| Message ID | 20220920092227.286306-5-marcel@ziswiler.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | arm: dts: colibri-imx6: usb dual-role switching et. al. | expand |
+Greg, to get an opinion on the fixes tag. On Tue, Sep 20, 2022 at 11:22:27AM +0200, Marcel Ziswiler wrote: > From: Marcel Ziswiler <marcel.ziswiler@toradex.com> > > Remove spurious debounce property from linux,extcon-usb-gpio. > > Note that debouncing is hard-coded to 20 ms (USB_GPIO_DEBOUNCE_MS > define). > > Fixes: 0ef1969ea569 ("ARM: dts: imx7-colibri: move aliases, chosen, extcon and gpio-keys") > Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> Hello all, we did have some (internal) discussion if this patch should have the fixes tag or not. I do personally think it should not have it and should not be backported to stable tree, since this is not fixing a real bug, it's just a cleanup. On the other hand the original patch was not correct, and this change is making it right. What is the general opinion on this topic? What do the stable kernel maintainers would expect? Documentation/process/stable-kernel-rules.rst is about rules for backporting, it does not really talk about the fixes tag, but today this is used to decide if a patch should be backported or not. Francesco
On Wed, Sep 21, 2022 at 02:15:05PM +0200, Francesco Dolcini wrote: > +Greg, to get an opinion on the fixes tag. > > On Tue, Sep 20, 2022 at 11:22:27AM +0200, Marcel Ziswiler wrote: > > From: Marcel Ziswiler <marcel.ziswiler@toradex.com> > > > > Remove spurious debounce property from linux,extcon-usb-gpio. > > > > Note that debouncing is hard-coded to 20 ms (USB_GPIO_DEBOUNCE_MS > > define). > > > > Fixes: 0ef1969ea569 ("ARM: dts: imx7-colibri: move aliases, chosen, extcon and gpio-keys") > > Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> > > Hello all, > we did have some (internal) discussion if this patch should have the > fixes tag or not. > > I do personally think it should not have it and should not be backported > to stable tree, since this is not fixing a real bug, it's just a > cleanup. If it's not a real bug, why would you have a Fixes: tag on the commit? > On the other hand the original patch was not correct, and this change is > making it right. Ah, so it is a bugfix. > What is the general opinion on this topic? What do the stable kernel > maintainers would expect? It's up to you, but what is the problem with it being backported? > Documentation/process/stable-kernel-rules.rst is about rules for > backporting, it does not really talk about the fixes tag, but today this > is used to decide if a patch should be backported or not. We use Fixes: as a signal from maintainers and developers that do not normally use the cc: stable@ as documented to pick up things that look like fixes but someone forgot to ask to be backported. It's not a guarantee it will be backported, like cc: stable will be, but it is a hint to us that maybe it should be looked at. thanks, greg k-h
On Wed, Sep 21, 2022 at 02:23:17PM +0200, Greg Kroah-Hartman wrote: > On Wed, Sep 21, 2022 at 02:15:05PM +0200, Francesco Dolcini wrote: > > +Greg, to get an opinion on the fixes tag. > > > > On Tue, Sep 20, 2022 at 11:22:27AM +0200, Marcel Ziswiler wrote: > > > From: Marcel Ziswiler <marcel.ziswiler@toradex.com> > > > > > > Remove spurious debounce property from linux,extcon-usb-gpio. > > > > > > Note that debouncing is hard-coded to 20 ms (USB_GPIO_DEBOUNCE_MS > > > define). > > > > > > Fixes: 0ef1969ea569 ("ARM: dts: imx7-colibri: move aliases, chosen, extcon and gpio-keys") > > > Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> > > > > Hello all, > > we did have some (internal) discussion if this patch should have the > > fixes tag or not. > > > > I do personally think it should not have it and should not be backported > > to stable tree, since this is not fixing a real bug, it's just a > > cleanup. > > If it's not a real bug, why would you have a Fixes: tag on the commit? > > > On the other hand the original patch was not correct, and this change is > > making it right. > > Ah, so it is a bugfix. The DTS file should not contain this unknown property, however having this property present does not change the behavior of the code. It's more like an unused variable kind of error that could trigger a warning in the compiler. Francesco
diff --git a/arch/arm/boot/dts/imx7-colibri.dtsi b/arch/arm/boot/dts/imx7-colibri.dtsi index a8c31ee65623..840420f9a1cd 100644 --- a/arch/arm/boot/dts/imx7-colibri.dtsi +++ b/arch/arm/boot/dts/imx7-colibri.dtsi @@ -29,7 +29,6 @@ chosen { extcon_usbc_det: usbc-det { compatible = "linux,extcon-usb-gpio"; - debounce = <25>; id-gpio = <&gpio7 14 GPIO_ACTIVE_HIGH>; /* SODIMM 137 / USBC_DET */ pinctrl-names = "default"; pinctrl-0 = <&pinctrl_usbc_det>;