Re: [linux-next:master 14416/15220] fs/exportfs/expfs.c:357:5: error: expected identifier or '('

[Amir Goldstein <amir73il@gmail.com>]

On Thu, Oct 26, 2023 at 6:15 PM Amir Goldstein <amir73il@gmail.com> wrote:
>
> On Thu, Oct 26, 2023 at 5:37 PM Arnd Bergmann <arnd@arndb.de> wrote:
> >
> > On Thu, Oct 26, 2023, at 15:49, kernel test robot wrote:
> > > tree:
> > > https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git
> > > master
> > > head:   2ef7141596eed0b4b45ef18b3626f428a6b0a822
> > > commit: dfaf653dc41557548b2e75e6cd837071f7c63289 [14416/15220]
> > > exportfs: make ->encode_fh() a mandatory method for NFS export
> > > config: hexagon-randconfig-r015-20211224
> > > (https://download.01.org/0day-ci/archive/20231026/202310262151.renqMvme-lkp@intel.com/config)
> > > compiler: clang version 16.0.4
> > > (https://github.com/llvm/llvm-project.git
> > > ae42196bc493ffe877a7e3dff8be32035dea4d07)
> > > reproduce (this is a W=1 build):
> > > (https://download.01.org/0day-ci/archive/20231026/202310262151.renqMvme-lkp@intel.com/reproduce)
> > >
> > > If you fix the issue in a separate patch/commit (i.e. not just a new
> > > version of
> > > the same patch/commit), kindly add following tags
> > > | Reported-by: kernel test robot <lkp@intel.com>
> > > | Closes:
> > > https://lore.kernel.org/oe-kbuild-all/202310262151.renqMvme-lkp@intel.com/
> > >
> > > All errors (new ones prefixed by >>):
> > >
> > >>> fs/exportfs/expfs.c:357:5: error: expected identifier or '('
> > >    int generic_encode_ino32_fh(struct inode *inode, __u32 *fh, int
> > > *max_len,
> > >        ^
> > >    include/linux/exportfs.h:286:33: note: expanded from macro
> >
> > This bit is trivially fixed by allowing the protoype to
> > be visible for CONFIG_EXPORTFS=m
> >
> > --- a/include/linux/exportfs.h
> > +++ b/include/linux/exportfs.h
> > @@ -314,7 +314,7 @@ extern struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid,
> >  /*
> >   * Generic helpers for filesystems.
> >   */
> > -#ifdef CONFIG_EXPORTFS
> > +#if IS_ENABLED(CONFIG_EXPORTFS)
> >  int generic_encode_ino32_fh(struct inode *inode, __u32 *fh, int *max_len,
> >                             struct inode *parent);
> >  #else
> >
> > but then we instead get link failures when the callers are
> > built-in:
> >
> > x86_64-linux-ld: fs/ext4/super.o:(.rodata+0x11a0): undefined reference to `generic_encode_ino32_fh'
> > x86_64-linux-ld: fs/ext2/super.o:(.rodata+0x3a0): undefined reference to `generic_encode_ino32_fh'
> > x86_64-linux-ld: fs/fat/nfs.o:(.rodata+0x120): undefined reference to `generic_encode_ino32_fh'
> > x86_64-linux-ld: fs/ntfs3/super.o:(.rodata+0x2a0): undefined reference to `generic_encode_ino32_fh'
> > x86_64-linux-ld: fs/befs/linuxvfs.o:(.rodata+0x240): undefined reference to `generic_encode_ino32_fh'
> >
> > Could this function just be move to fs/libfs.c or similar?
>
> It could, but it is quite annoying to have to build the function when no code
> should be calling it.
>
> IMO, it would be better to make CONFIG_EXPORTFS a bool.
> fs/exportfs/expfs.c is quite small and CONFIG_EXPORTFS=m
> doesn't seem to be worthwhile.

Forget it. I missed the fact that the problem is that none of the
filesystem properly select EXPORTFS.

I will see if I want to fix that or give up and move the helper to fs/libfs.c.

Thanks,
Amir.