Re: linux-next: build failure after merge of the staging tree

[Dave Penkler]

On Wed, Oct 16, 2024 at 04:15:41PM +0200, Dave Penkler wrote:
> On Wed, Oct 16, 2024 at 09:40:03AM +0200, Greg KH wrote:
> > On Tue, Oct 15, 2024 at 04:55:38PM +1100, Stephen Rothwell wrote:
> > > Hi all,
> > > 
> > > After merging the staging tree, today's linux-next build (powerpc
> > > allyesconfig) failed like this:
> > > 
> > > ld: warning: discarding dynamic section .glink
> > > ld: warning: discarding dynamic section .plt
> > > ld: linkage table error against `nec7210_board_online'
> > > ld: stubs don't match calculated size
> > > ld: can not build stubs: bad value
> [skip]
> > > 
> > > Caused by commit
> > > 
> > >   8e4841a0888c ("staging: gpib: Add Frank Mori Hess FPGA PCI GPIB driver")
> > > 
> > > I have marked that driver as BROKEN for today.
> > 
> > Thanks, I'll go take your patch from next and add it to my tree for
> > this, sorry about the build issues.
> > 
> > greg k-h
> Hi,
> All declarations and uses of `nec7210_board_online' match.
> 
> Could this be a ppc linker problem ?
> -Dave

This bug report from Red Hat would indicate that there could be a problem with the ppc linker: 
Bug 1523457 - ghc-8.2.2 build linking error on rawhide ppc64le with binutils-2.29.1
https://bugzilla.redhat.com/show_bug.cgi?id=1523457

Quoting from the bug report:
  The problem appears to be an instability in the code in the PowerPC 
  linker's backend that computes the call stubs needed to access the PLT.
  As the linker moves sections around in memory one stub switches from
  containing a backwards branch to containing a forwards branch, altering
  its size and confusing the code which had previously allocated space for
  the stub.

This seems to be the same issue we are having:
 ld: stubs don't match calculated size

It is reported to be fixed in the 2.30 FSF binutils sources.

Could we rather make GPIB_FMH depend on !PPC rather than BROKEN for now ?
-Dave