On 08/26/14 12:26, Mark Brown wrote: > On Tue, Aug 26, 2014 at 12:20:54PM -0700, Randy Dunlap wrote: >> On 08/26/14 10:25, Mark Brown wrote: > >>> index d58101e788fc..65a351d75c95 100644 >>> --- a/Documentation/video4linux/Makefile >>> +++ b/Documentation/video4linux/Makefile >>> @@ -1 +1 @@ >>> -obj-m := v4l2-pci-skeleton.o >>> +obj-$(CONFIG_VIDEO_PCI_SKELETON) := v4l2-pci-skeleton.o >>> diff --git a/drivers/media/v4l2-core/Kconfig b/drivers/media/v4l2-core/Kconfig > >>> +config VIDEO_PCI_SKELETON >>> + tristate "Skeleton PCI V4L2 driver" >>> + depends on PCI && COMPILE_TEST > >> && ?? No, don't require COMPILE_TEST. > > That's a very deliberate choice. There's no reason I can see to build > this code other than to check that it builds, it's reference code rather > than something that someone is expected to actually use in their system. > This seems like a perfect candidate for COMPILE_TEST. > >> However, PCI || COMPILE_TEST would allow it to build on arm64 >> if COMPILE_TEST is enabled, guaranteeing build errors. >> Is that what should happen? I suppose so... > > No, it's not - if it's going to depend on COMPILE_TEST at all it need to > be a hard dependency. How about just drop COMPILE_TEST? This code only builds if someone enabled BUILD_DOCSRC. That should be enough (along with PCI and some VIDEO kconfig symbols) to qualify it. -- ~Randy -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
