Re: randconfig build error with next-20140829, in drivers/media/usb/dvb-usb/technisat-usb2.c

["Luis R. Rodriguez" <mcgrof@suse.com>]

On Fri, Aug 29, 2014 at 09:19:42AM -0700, Jim Davis wrote:
> Building with the attached random configuration file,
> 
>   LD      init/built-in.o
> drivers/built-in.o: In function `technisat_usb2_set_voltage':
> technisat-usb2.c:(.text+0x3b4919): undefined reference to `stv090x_set_gpio'
> make: *** [vmlinux] Error 1

This is because MEDIA_SUBDRV_AUTOSELECT is designed to let you
pick and choose, technically we should just have:

diff --git a/drivers/media/usb/dvb-usb/Kconfig b/drivers/media/usb/dvb-usb/Kconfig
index c5d9566..5a4e82e 100644
--- a/drivers/media/usb/dvb-usb/Kconfig
+++ b/drivers/media/usb/dvb-usb/Kconfig
@@ -313,7 +313,7 @@ config DVB_USB_AZ6027
 config DVB_USB_TECHNISAT_USB2
 	tristate "Technisat DVB-S/S2 USB2.0 support"
 	depends on DVB_USB
-	select DVB_STV090x if MEDIA_SUBDRV_AUTOSELECT
+	select DVB_STV090x
 	select DVB_STV6110x if MEDIA_SUBDRV_AUTOSELECT
 	help
 	  Say Y here to support the Technisat USB2 DVB-S/S2 device

and that would fix the issue you saw but then again if we do that
we might as well also do the same for DVB_STV6110x and a slew of
different Kconfig entries on that file.

Someone needs to make a judgement call and either fix all these
Kconfig entries or document that MEDIA_SUBDRV_AUTOSELECT will
let you shoot yourself in the foot at build time. Then what
I recommend in the meantime is simply to not trust randomconfig
builds unless you are always enabling MEDIA_SUBDRV_AUTOSELECT.

I think its fair to expect for 'make randomconfig' to give you
a configuration that lets you build things without issue so
I see this more of an issue with MEDIA_SUBDRV_AUTOSELECT and
this sloppy embedded craze.

  Luis