[OpenAFS-devel] 2004-06-15 snap on FC2: no good still (sys_call_table, Error -1 in basic initialization)

[Rainer Schöpf]

On Tue, 15 Jun 2004, Matthew Miller wrote:

 > On Tue, Jun 15, 2004 at 10:38:57PM +0200, Rainer Schöpf wrote:
 > > regparm(0). Diffs of my changes are attached.
 >
 > Well, progress. With all of your patches applied (except the inodesecurity
 > one -- that doesn't apply cleanly and when modified to apply, makes the
 > module not load complaining about security_inode_alloc, so I removed that
 > and just left in the "ip->i_security = NULL;" change), I can do ls in /afs
 > and see the dynroot properly. But changing to bu.edu and doing an ls there
 > gives a familiar-looking oops:

OK, try the following (untested) patch.


*** src/fsint/Kvice.xdr.c~      Thu Jun 10 13:32:43 2004
--- src/fsint/Kvice.xdr.c       Tue Jun 15 23:52:15 2004
***************
*** 9,15 ****
  #ifndef FSINT_COMMON_XG


! bool_t
  xdr_AFSFid(XDR *xdrs, AFSFid *objp)
  {
        if (!xdr_afs_uint32(xdrs, &objp->Volume)) {
--- 9,15 ----
  #ifndef FSINT_COMMON_XG


! bool_t LINUX_REGPARM_0
  xdr_AFSFid(XDR *xdrs, AFSFid *objp)
  {
        if (!xdr_afs_uint32(xdrs, &objp->Volume)) {
***************
*** 27,33 ****



! bool_t
  xdr_AFSCallBack(XDR *xdrs, AFSCallBack *objp)
  {
        if (!xdr_afs_uint32(xdrs, &objp->CallBackVersion)) {
--- 27,33 ----



! bool_t LINUX_REGPARM_0
  xdr_AFSCallBack(XDR *xdrs, AFSCallBack *objp)
  {
        if (!xdr_afs_uint32(xdrs, &objp->CallBackVersion)) {