Re: 2.6.16-rc3-git7 build failure if sysfs disabled

[D. Hazelton]

On Thursday 16 February 2006 18:42, Jesper Juhl wrote:
> 2.6.16-rc3-git7 fails to build when sysfs is disabled with the
> attached .config .
>
> Without sysfs enabled :
>
>   LD      vmlinux
>
> lib/lib.a(kobject_uevent.o)(.text+0x1d5): In function `kobject_uevent':
> : undefined reference to `uevent_seqnum'
>
> lib/lib.a(kobject_uevent.o)(.text+0x1db): In function `kobject_uevent':
> : undefined reference to `uevent_seqnum'
>
> lib/lib.a(kobject_uevent.o)(.text+0x1e7): In function `kobject_uevent':
> : undefined reference to `uevent_seqnum'
>
> lib/lib.a(kobject_uevent.o)(.text+0x1ed): In function `kobject_uevent':
> : undefined reference to `uevent_seqnum'
>
> lib/lib.a(kobject_uevent.o)(.text+0x22c): In function `kobject_uevent':
> : undefined reference to `uevent_helper'
>
> lib/lib.a(kobject_uevent.o)(.text+0x2c0): In function `kobject_uevent':
> : undefined reference to `uevent_helper'
>
> lib/lib.a(kobject_uevent.o)(.text+0x2d3): In function `kobject_uevent':
> : undefined reference to `uevent_helper'
>
> make: *** [vmlinux] Error 1
>
> If sysfs is switched on the same config builds just fine.
>
> The .config is attached.

I've been looking at this for a few minutes now and wondering - since sysfs is 
an integrated part of the new module system there isn't an easy solution in 
the KConfig system. Perhaps providing a stub function that does nothing for 
builds that have sysfs turned off would be better...

Either that, or leave the function in and have its function depend, in some 
function, on whether sysfs was loaded as a module or isn't present at all. 
Truthfully, though, I don't know if that's possible.

DRH
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/