Re: linux-next: net tree build failure

[David Miller <davem@davemloft.net>]

From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Fri, 23 Jan 2009 21:50:57 +1100

> On Fri, 23 Jan 2009 21:39:27 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > Hi again,
> > 
> > On Fri, 23 Jan 2009 21:21:28 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> > >
> > > So, this could be a compiler bug?
> > 
> > We pass -fno-common to gcc which seems to cause the (non-extern) const
> > variable to end up in the bss.
> 
>        -fno-common
>            In C, allocate even uninitialized global variables in the data
>            section of the object file, rather than generating them as common
>            blocks.  This has the effect that if the same variable is declared
>            (without "extern") in two different compilations, you will get an
>            error when you link them.  The only reason this might be useful is
>            if you wish to verify that the program will work on other systems
>            which always work this way.
> 
> I don't know why we use this, but it seems to basically make using
> "extern" on variables in header files compulsory.  But it seems to be
> moving the variable from *COM* to .bss rather than to .{ro}data as I would
> expect from the description above.
> 
> Oh, well, easy to work around.

Thanks for testing, it confirms my suspicions :-)

I'll check in the fix.