| Message ID | 20180705204703.GA12352@embeddedor.com (mailing list archive) |
|---|---|
| State | Changes Requested, archived |
| Delegated to: | Darren Hart |
| Headers | show |
On Thu, Jul 05, 2018 at 03:47:03PM -0500, Gustavo A. R. Silva wrote: > In preparation to enabling -Wimplicit-fallthrough, mark switch cases > where we are expecting to fall through. > > Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> > --- > drivers/platform/x86/sony-laptop.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c > index b205b03..e614cb7 100644 > --- a/drivers/platform/x86/sony-laptop.c > +++ b/drivers/platform/x86/sony-laptop.c > @@ -4427,6 +4427,7 @@ sony_pic_read_possible_resource(struct acpi_resource *resource, void *context) > default: > dprintk("Resource %d isn't an IRQ nor an IO port\n", > resource->type); > + /* fall through */ Here too, I wonder if this is intentional. Either way, from what I can see, the final line in the function: return AE_CTRL_TERMINATE; Is unreachable as there are no "break" statements in the switch, and the default falls through to return AE_OK. Something doesn't seem right here.
On Sat, Jul 7, 2018 at 2:24 AM, Darren Hart <dvhart@infradead.org> wrote: > return AE_CTRL_TERMINATE; > > Is unreachable as there are no "break" statements in the switch, and the default > falls through to return AE_OK. Something doesn't seem right here. Darren, I didn't look at the code in question, though just to clarify such error codes. If the function used in acpi_walk_namespace() or alike, AE_OK means "To continue", while AE_CTRL_TERMINATE to "stop, whatever we looked for is found right now".
On Fri, Jul 06, 2018 at 04:24:27PM -0700, Darren Hart wrote: > On Thu, Jul 05, 2018 at 03:47:03PM -0500, Gustavo A. R. Silva wrote: > > In preparation to enabling -Wimplicit-fallthrough, mark switch cases > > where we are expecting to fall through. > > > > Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> > > --- > > drivers/platform/x86/sony-laptop.c | 1 + > > 1 file changed, 1 insertion(+) > > > > diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c > > index b205b03..e614cb7 100644 > > --- a/drivers/platform/x86/sony-laptop.c > > +++ b/drivers/platform/x86/sony-laptop.c > > @@ -4427,6 +4427,7 @@ sony_pic_read_possible_resource(struct acpi_resource *resource, void *context) > > default: > > dprintk("Resource %d isn't an IRQ nor an IO port\n", > > resource->type); > > + /* fall through */ > > Here too, I wonder if this is intentional. Either way, from what I can see, the > final line in the function: > > return AE_CTRL_TERMINATE; > > Is unreachable as there are no "break" statements in the switch, and the default > falls through to return AE_OK. Something doesn't seem right here. I think so too. Looks to me that the default case was meant to return AE_CTRL_TERMINATE (i.e. swapping it to be last would do). Having written that code aeons ago, I'm not sure if there is a good reason to be like that, though it still looks like a bug (that existed ever since...). Want to fix it as part of this patch series? Or I can send a patch myself, let me know.
On Sun, Jul 08, 2018 at 03:56:17AM +0900, Mattia Dongili wrote: > On Fri, Jul 06, 2018 at 04:24:27PM -0700, Darren Hart wrote: > > On Thu, Jul 05, 2018 at 03:47:03PM -0500, Gustavo A. R. Silva wrote: > > > In preparation to enabling -Wimplicit-fallthrough, mark switch cases > > > where we are expecting to fall through. > > > > > > Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> > > > --- > > > drivers/platform/x86/sony-laptop.c | 1 + > > > 1 file changed, 1 insertion(+) > > > > > > diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c > > > index b205b03..e614cb7 100644 > > > --- a/drivers/platform/x86/sony-laptop.c > > > +++ b/drivers/platform/x86/sony-laptop.c > > > @@ -4427,6 +4427,7 @@ sony_pic_read_possible_resource(struct acpi_resource *resource, void *context) > > > default: > > > dprintk("Resource %d isn't an IRQ nor an IO port\n", > > > resource->type); > > > + /* fall through */ > > > > Here too, I wonder if this is intentional. Either way, from what I can see, the > > final line in the function: > > > > return AE_CTRL_TERMINATE; > > > > Is unreachable as there are no "break" statements in the switch, and the default > > falls through to return AE_OK. Something doesn't seem right here. > > I think so too. Looks to me that the default case was meant to return > AE_CTRL_TERMINATE (i.e. swapping it to be last would do). > > Having written that code aeons ago, I'm not sure if there is a good > reason to be like that, though it still looks like a bug (that > existed ever since...). > > Want to fix it as part of this patch series? Or I can send a patch > myself, let me know. Mattia, I'd prefer to decouple the fix from the implicit fall through discussion. Please send a separate fix.
diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index b205b03..e614cb7 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -4427,6 +4427,7 @@ sony_pic_read_possible_resource(struct acpi_resource *resource, void *context) default: dprintk("Resource %d isn't an IRQ nor an IO port\n", resource->type); + /* fall through */ case ACPI_RESOURCE_TYPE_END_TAG: return AE_OK;
In preparation to enabling -Wimplicit-fallthrough, mark switch cases where we are expecting to fall through. Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> --- drivers/platform/x86/sony-laptop.c | 1 + 1 file changed, 1 insertion(+)