linux-next: alacrity tree build failure

linux-next: alacrity tree build failure

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 652 bytes --]

Hi Gregory,

Today's linux-next build (x86_6 _allmodconfig) failed like this:

drivers/vbus/built-in.o:(.data+0x0): multiple definition of `dev_attr_modalias'
drivers/xen/built-in.o:(.data+0x46d0): first defined here

Caused by commit 59aa8f441d27c8470764a513dafa46a77f33e953 ("vbus: add
autoprobe capability to guest").  The DEVICE_ATTR(modalias ...) should
probably be static.  I should probably be static in
drivers/xen/xenbus/xenbus_probe.c as well.

I have used the version of the alacrity tree from next-20091123 for today.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: alacrity tree build failure

[Jeremy Fitzhardinge]

On 11/23/09 23:58, Stephen Rothwell wrote:
> Today's linux-next build (x86_6 _allmodconfig) failed like this:
>
> drivers/vbus/built-in.o:(.data+0x0): multiple definition of `dev_attr_modalias'
> drivers/xen/built-in.o:(.data+0x46d0): first defined here
>
> Caused by commit 59aa8f441d27c8470764a513dafa46a77f33e953 ("vbus: add
> autoprobe capability to guest").  The DEVICE_ATTR(modalias ...) should
> probably be static.  I should probably be static in
> drivers/xen/xenbus/xenbus_probe.c as well.
>   

All those DEVICE_ATTR()s should be static in that case.  Indeed, I guess
they should be almost universally?  Should DEVICE_ATTR() include the
static?  (Hm, almost every instance in the kernel already has static; it
should probably have been part of the definition from the start, but
changing it now would cause a lot of churn.)

    J

Re: linux-next: alacrity tree build failure

[Greg KH]

On Tue, Nov 24, 2009 at 04:41:05PM -0800, Jeremy Fitzhardinge wrote:
> On 11/23/09 23:58, Stephen Rothwell wrote:
> > Today's linux-next build (x86_6 _allmodconfig) failed like this:
> >
> > drivers/vbus/built-in.o:(.data+0x0): multiple definition of `dev_attr_modalias'
> > drivers/xen/built-in.o:(.data+0x46d0): first defined here
> >
> > Caused by commit 59aa8f441d27c8470764a513dafa46a77f33e953 ("vbus: add
> > autoprobe capability to guest").  The DEVICE_ATTR(modalias ...) should
> > probably be static.  I should probably be static in
> > drivers/xen/xenbus/xenbus_probe.c as well.
> >   
> 
> All those DEVICE_ATTR()s should be static in that case.  Indeed, I guess
> they should be almost universally?  Should DEVICE_ATTR() include the
> static?  (Hm, almost every instance in the kernel already has static; it
> should probably have been part of the definition from the start, but
> changing it now would cause a lot of churn.)

There are some "non static" usages, mostly to deal with stuff within a
module that takes more than one file.  They could probably all be fixed
up if someone really is bored :)

thanks,

greg k-h