[OpenAFS] Still Compiling on FC6T2 (try again)

William Murray W.J.Murray@rl.ac.uk
Sat, 19 Aug 2006 09:09:18 +0100


  Thanks Chas,
             That fixes the first error...but then we get:

/usr/local/src/openafs-1.5.7/src/libafs/MODLOAD-2.6.17-1.2571.fc6-MP/osi_vf=
sops.c:87: error: too few arguments to function =E2=80=98get_sb_nodev=E2=80=
=99

The symbol
GET_SB_HAS_STRUCT_VFSMOUNT
is not defined...looks like maybe it should be?

  Bill

On Fri, 2006-08-18 at 11:44 -0400, chas williams - CONTRACTOR wrote:
> apply this.  i was thinking about making the whole mutex api generic
> which would tend to solve this problem.  something like osi_mutex_t,
> osi_mutex_lock(), osi_mutex_unlock() which each arch would need to=20
> define (this is already done in rx so it would just get moved to the
> the afs side).
>=20
> Index: src/afs/LINUX/osi_pag_module.c
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> RCS file: /cvs/openafs/src/afs/LINUX/osi_pag_module.c,v
> retrieving revision 1.1.2.4
> diff -u -u -r1.1.2.4 osi_pag_module.c
> --- src/afs/LINUX/osi_pag_module.c	13 Aug 2006 04:53:27 -0000	1.1.2.4
> +++ src/afs/LINUX/osi_pag_module.c	18 Aug 2006 15:42:50 -0000
> @@ -55,8 +55,12 @@
>  MODULE_PARM_DESC(this_cell, "Local cell name");
> =20
>  #if defined(AFS_LINUX24_ENV)
> -DECLARE_MUTEX(afs_global_lock);
>  struct proc_dir_entry *openafs_procfs;
> +#endif
> +#if LINUX_VERSION_CODE >=3D KERNEL_VERSION(2,6,16)
> +DEFINE_MUTEX(afs_global_lock);
> +#elif LINUX_VERSION_CODE >=3D KERNEL_VERSION(2,4,0)
> +DECLARE_MUTEX(afs_global_lock);
>  #else
>  struct semaphore afs_global_lock =3D MUTEX;
>  #endif
>=20
> In message <1155915200.3004.19.camel@BillMurray>,William Murray writes:
> >  OK, 1.5.7 does get further.
> >
> >/usr/local/src/openafs-1.5.7/src/libafs/MODLOAD-2.6.17-1.2571.fc6-MP/osi=
_pag_m
> >odule.c:58: error: conflicting types for 'afs_global_lock'
> >/usr/local/src/openafs-1.5.7/src/afs/LINUX/osi_machdep.h:197: error:
> >previous declaration of 'afs_global_lock' was here
> >
> >Does anyone know which is a better bet?
> >  Bill
> >
> >
> >
> >
> >On Fri, 2006-08-18 at 16:00 +0200, Stephan Wiesand wrote:
> >> On Fri, 18 Aug 2006, Derrick J Brashear wrote:
> >>=20
> >> > The CVS head probably still isn't what you want. 1.5.7 is a few hour=
s old.
> >=20
> >> > Try it.
> >>=20
> >> Maybe more people would if someone released the volume ;-)
> >>=20
> >> > Derrick
> >> > On Fri, 18 Aug 2006, William John Murray wrote:
> >> >
> >> >> Hi guys,
> >> >> I watched all the discussion about FC6T2 at the weekend,
> >> >> but it did not seem to apply to me yet. I am stuck (with the latest=
 CVS)
> >> >> at:
> >> >> /usr/local/src/openafs-snap-2006-08-14/src/libafs/MODLOAD-2.6.17-1.=
2564.f
> >c6-MP/rx.c:2397:=20
> >> >> error: implicit declaration of function =18IN6_IS_ADDR_V4COMPAT=19
> >> >>=20
> >> >> I don't understand where the in3t6 definitions are supposed to come=
 from:
> >=20
> >> >> there
> >> >> arer no results for
> >> >> find /usr -name inet\*
> >> >> or
> >> >> find /usr -name netinet\*
> >> >> on my system.
> >> >> Does anyone know what happened here?
> >> >> Thanks,
> >> >> Bill
> >> >>=20
> >> >> _______________________________________________
> >> >> OpenAFS-info mailing list
> >> >> OpenAFS-info@openafs.org
> >> >> https://lists.openafs.org/mailman/listinfo/openafs-info
> >> >>=20
> >> > _______________________________________________
> >> > OpenAFS-info mailing list
> >> > OpenAFS-info@openafs.org
> >> > https://lists.openafs.org/mailman/listinfo/openafs-info
> >> >
> >>=20
> >_______________________________________________
> >OpenAFS-info mailing list
> >OpenAFS-info@openafs.org
> >https://lists.openafs.org/mailman/listinfo/openafs-info
> >