On 07/08/11 14:24, J Freyensee wrote: > On 07/08/2011 01:07 PM, Geert Uytterhoeven wrote: >> On Thu, Jul 7, 2011 at 18:32, Randy Dunlap<randy.dunlap@xxxxxxxxxx> >> wrote: >>> On Thu, 07 Jul 2011 09:29:59 -0700 J Freyensee wrote: >>>> On 07/07/2011 08:14 AM, Randy Dunlap wrote: >>>>> On Mon, 4 Jul 2011 13:45:41 -0700 Randy Dunlap wrote: >>>>>> On Mon, 4 Jul 2011 17:09:52 +1000 Stephen Rothwell wrote: >>>>>>> Changes since 20110701: >>>>>> when CONFIG_PCI is not enabled: >>>>>> >>>>>> drivers/misc/pti.c:410: error: implicit declaration of function >>>>>> 'pci_release_region' >>>>>> drivers/misc/pti.c:859: error: implicit declaration of function >>>>>> 'pci_request_region' >>>>>> >>>>>> Stubs for these 2 functions could be added to include/linux/pci.h, >>>>>> but it would >>>>>> make more sense to me just to make the driver depend on PCI since >>>>>> it seems to rely >>>>>> on so many PCI functions. or are the PCI pieces optional? >>>>> ping. still a problem in linux-next 20110706. >>>> Thanks for the ping. >>>> >>>> Is there something you need me to do or look at? >>>> >>>> For Intel-Atom architectures, the MIPI PTI module does sit on the PCI >>>> bus. It's very much dependent upon the PCI bus for working >>>> functionality and it's not optional. So yes, it makes sense for the >>>> driver to depend on PCI to be there and if PCI is not enabled on the >>>> system, don't build the pti driver. >>>> >>>> Maybe it's a Kconfig dependency? >>> Yes, please send/merge a patch that adds >>> depends on PCI >>> to "config INTEL_MID_PCI". >> http://kisskb.ellerman.id.au/kisskb/buildresult/4292241/: > I can't access that http site (with or without the ':'). Works for me (without the trailing colon). >> drivers/misc/pti.c:249: error: implicit declaration of function 'kmalloc' >> drivers/misc/pti.c:258: error: implicit declaration of function 'kfree' >> drivers/misc/pti.c:657: error: implicit declaration of function >> 'copy_from_user' >> drivers/misc/pti.c:848: error: implicit declaration of function 'kzalloc' >> >> Probably missing > >> >> #include<linux/slab.h> >> #include<linux/uaccess.h> > > I can send a patch to add these two includes. It didn't have previous > troubles compiling, so maybe one of the other includes used to be > including these? That URL (above) says: Compiler: m68k (m68k-linux-gcc (GCC) 4.4.0) and different $ARCH-es do different #inclusions. Anyway, the interfaces are used so the source file should #include the header files. -- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code *** -- To unsubscribe from this list: send the line "unsubscribe linux-next" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
