On Thu, Sep 12, 2019 at 4:26 PM J. Bruce Fields <bfields@xxxxxxxxxxxx> wrote: > > On Fri, Sep 06, 2019 at 04:27:32PM -0400, Olga Kornievskaia wrote: > > Trond/Anna, > > > > Please ignore this current version. With working on the server, I have > > forgotten that I had change(s) to the client side but didn't include > > them here. I'll repost with those changes. > > > > Bruce, server side changes are all here. > > There are conflicts with my 5.4 tree, mostly due to the file caching > stuff. I started trying to fix them up myself but there were more than > I expected and as I was afraid I'd screw it up. Could you take a look? I can work off your tree. > > --b. > > > > > On Fri, Sep 6, 2019 at 3:46 PM Olga Kornievskaia > > <olga.kornievskaia@xxxxxxxxx> wrote: > > > > > > As per Bruce's request submitting the client and server series together > > > as there is a common patch that's needed by both. > > > > > > No client side changes were made since the previous submission > > > > > > v6 server-side changes > > > --- removed global copy notify list and instead relying only on the idr > > > list of stateids. Laundromat now traverses that list and if copy notify > > > state wasn't referenced in a lease period, state is deleted. > > > --- removed storing parent's stid pointer in the copy notify state. > > > instead storing the parent's stateid and client id then using it to > > > lookup the stid structure and client structure during validation of > > > the stateid of the READ. > > > --- added a refcount to the copy notify state to make sure only 1 will > > > delete it (as it can be delete either by the nfs4_put_stid(), > > > laundromat, or offload_cancel op. basically all access to the copy state > > > is using just one global lock now (netd->s2s_cp_lock). > > > --- added a type to the copy_stateid_t to distinguish copy notify state > > > kept by the source server and copy state used by the destination server. > > > --- previously with a global copy notify list, the check if client has > > > state before unmounting checked if the list was empty, now the code > > > traverses the idr list and looks for anything with a matching clientid > > > (again under the global s2s_cp_lock). > > > > > > Olga Kornievskaia (19): > > > NFS NFSD: defining nl4_servers structure needed by both > > > NFS: add COPY_NOTIFY operation > > > NFS: add ca_source_server<> to COPY > > > NFS: also send OFFLOAD_CANCEL to source server > > > NFS: inter ssc open > > > NFS: skip recovery of copy open on dest server > > > NFS: for "inter" copy treat ESTALE as ENOTSUPP > > > NFS: COPY handle ERR_OFFLOAD_DENIED > > > NFS: handle source server reboot > > > NFS: replace cross device check in copy_file_range > > > NFSD fill-in netloc4 structure > > > NFSD add ca_source_server<> to COPY > > > NFSD return nfs4_stid in nfs4_preprocess_stateid_op > > > NFSD COPY_NOTIFY xdr > > > NFSD add COPY_NOTIFY operation > > > NFSD check stateids against copy stateids > > > NFSD generalize nfsd4_compound_state flag names > > > NFSD: allow inter server COPY to have a STALE source server fh > > > NFSD add nfs4 inter ssc to nfsd4_copy > > > > > > fs/nfs/nfs42.h | 15 +- > > > fs/nfs/nfs42proc.c | 193 ++++++++++++++++---- > > > fs/nfs/nfs42xdr.c | 190 +++++++++++++++++++- > > > fs/nfs/nfs4_fs.h | 11 ++ > > > fs/nfs/nfs4client.c | 2 +- > > > fs/nfs/nfs4file.c | 125 ++++++++++++- > > > fs/nfs/nfs4proc.c | 6 +- > > > fs/nfs/nfs4state.c | 29 ++- > > > fs/nfs/nfs4xdr.c | 1 + > > > fs/nfsd/Kconfig | 10 ++ > > > fs/nfsd/nfs4proc.c | 440 ++++++++++++++++++++++++++++++++++++++++++---- > > > fs/nfsd/nfs4state.c | 215 +++++++++++++++++++--- > > > fs/nfsd/nfs4xdr.c | 154 +++++++++++++++- > > > fs/nfsd/nfsd.h | 32 ++++ > > > fs/nfsd/nfsfh.h | 5 +- > > > fs/nfsd/nfssvc.c | 6 + > > > fs/nfsd/state.h | 34 +++- > > > fs/nfsd/xdr4.h | 39 +++- > > > include/linux/nfs4.h | 25 +++ > > > include/linux/nfs_fs.h | 3 +- > > > include/linux/nfs_fs_sb.h | 1 + > > > include/linux/nfs_xdr.h | 17 ++ > > > 22 files changed, 1431 insertions(+), 122 deletions(-) > > > > > > -- > > > 1.8.3.1 > > >
