nfsd changes for 3.16

From: J. Bruce Fields
Date: Tue Jun 10 2014 - 14:16:09 EST

Please pull nfsd changes for 3.16 from

git://linux-nfs.org/~bfields/linux.git for-3.16

The largest piece is a long-overdue rewrite of the xdr code to remove
some annoying limitations: for example, there was no way to return ACLs
larger than 4K, and readdir results were returned only in 4k chunks,
limiting performance on large directories.

Also:
- part of Neil Brown's work to make NFS work reliably over the
loopback interface (so client and server can run on the same
machine without deadlocks). The rest of it is coming through
other trees.
- cleanup and bugfixes for some of the server RDMA code, from
Steve Wise.
- Various cleanup of NFSv4 state code in preparation for an
overhaul of the locking, from Jeff, Trond, and Benny.
- smaller bugfixes and cleanup from Christoph Hellwig and
Kinglong Mee.

Thanks to everyone!

This summer looks likely to be busier than usual for knfsd. Hopefully
we won't break it too badly; testing definitely welcomed.

--b.

----------------------------------------------------------------
Benny Halevy (3):
nfsd4: use recall_lock for delegation hashing
nfsd4: rename recall_lock to state_lock
nfsd4: hash deleg stateid only on successful nfs4_set_delegation

Benoit Taine (1):
nfsd: Remove assignments inside conditions

Christoph Hellwig (8):
nfsd: remove <linux/nfsd/nfsfh.h>
nfsd: move <linux/nfsd/export.h> to fs/nfsd
nfsd: move <linux/nfsd/stats.h> to fs/nfsd
nfsd: remove <linux/nfsd/debug.h>
nfsd: clean up fh_auth usage
nfsd: remove nfsd4_free_slab
nfsd: getattr for FATTR4_WORD0_FILES_AVAIL needs the statfs buffer
nfsd: remove unused function nfsd_read_file

Chuck Lever (1):
NFSD: Ignore client's source port on RDMA transports

J. Bruce Fields (59):
Merge 3.15 bugfix for 3.16
nfsd4: fix delegation cleanup on error
Merge 3.15 bugfixes for 3.16
nfsd4: READ, READDIR, etc., are idempotent
nfsd4: allow larger 4.1 session drc slots
nfsd4: read size estimate should include padding
nfsd4: fix write reply size estimate
nfsd4: decoding errors can still be cached and require space
nfsd4: embed xdr_stream in nfsd4_compoundres
nfsd4: tweak nfsd4_encode_getattr to take xdr_stream
nfsd4: move proc_compound xdr encode init to helper
nfsd4: reserve head space for krb5 integ/priv info
nfsd4: fix encoding of out-of-space replies
nfsd4: allow space for final error return
nfsd4: use xdr_reserve_space in attribute encoding
nfsd4: use xdr_stream throughout compound encoding
nfsd4: remove ADJUST_ARGS
nfsd4: no need for encode_compoundres to adjust lengths
nfsd4: keep xdr buf length updated
rpc: xdr_truncate_encode
nfsd4: use xdr_truncate_encode
nfsd4: "backfill" using write_bytes_to_xdr_buf
nfsd4: teach encoders to handle reserve_space failures
nfsd4: reserve space before inlining 0-copy pages
nfsd4: nfsd4_check_resp_size needn't recalculate length
nfsd4: remove redundant encode buffer size checking
nfsd4: size-checking cleanup
nfsd4: allow encoding across page boundaries
nfsd4: convert 4.1 replay encoding
nfsd4: don't try to encode conflicting owner if low on space
nfsd4: more precise nfsd4_max_reply
nfsd4: minor encode_read cleanup
nfsd4: nfsd4_check_resp_size should check against whole buffer
nfsd4: fix buflen calculation after read encoding
rpc: define xdr_restrict_buflen
nfsd4: adjust buflen to session channel limit
nfsd4: use session limits to release send buffer reservation
nfsd4: allow large readdirs
nfsd4: enforce rd_dircount
nfsd4: don't treat readlink like a zero-copy operation
nfsd4: better estimate of getattr response size
nfsd4: estimate sequence response size
nfsd4: turn off zero-copy-read in exotic cases
nfsd4: nfsd_vfs_read doesn't use file handle parameter
nfsd4: separate splice and readv cases
nfsd4: read encoding cleanup
nfsd4: more read encoding cleanup
nfsd4: allow exotic read compounds
nfsd4: really fix nfs4err_resource in 4.1 case
nfsd4: kill WRITE32
nfsd4: kill WRITE64
nfsd4: kill WRITEMEM
nfsd4: kill write32, write64
nfsd4: better reservation of head space for krb5
nfsd4: simplify server xdr->next_page use
nfsd4: kill READ32
nfsd4: kill READ64
nfs4: remove unused CHANGE_SECURITY_LABEL
nfsd4: fix FREE_STATEID lockowner leak

Jeff Layton (5):
nfsd: make nfsd4_encode_fattr static
nfsd: fix laundromat next-run-time calculation
nfsd: fix setting of NFS4_OO_CONFIRMED in nfsd4_open
nfsd: remove unneeded zeroing of fields in nfsd4_proc_compound
nfsd: don't halt scanning the DRC LRU list when there's an RC_INPROG entry

Kees Cook (1):
lockd: avoid warning when CONFIG_SYSCTL undefined

Kinglong Mee (16):
NFSD: cleanup unneeded including linux/export.h
SUNRPC: Fix printk that is not only for nfsd
NFSD: Use simple_read_from_buffer for coping data to userspace
NFSD: Get rid of empty function nfs4_state_init
NFSD: Don't clear SUID/SGID after root writing data
NFS4: Avoid NULL reference or double free in nfsd4_fslocs_free()
NFSD: Helper function for parsing uuid
NFSD: Adds macro EX_UUID_LEN for exports uuid's length
NFSD: remove unneeded linux/user_namespace.h include
NFSD: Cleanup unused variable in nfsd_setuser()
NFSD: Cleanup unneeded including net/ipv6.h
NFSD: Remove typedef of svc_client and svc_export in export.c
NFSD: Add missing comment of "expiry" in expkey_parse()
NFSD: Using type of uint32_t for ex_nflavors instead of int
NFSD: Error out when getting more than one fsloc/secinfo/uuid
SUNRPC/NFSD: Remove using of dprintk with KERN_WARNING

NeilBrown (2):
SUNRPC: track whether a request is coming from a loop-back interface.
nfsd: Only set PF_LESS_THROTTLE when really needed.

Steve Wise (2):
svcrdma: refactor marshalling logic
svcrdma: Fence LOCAL_INV work requests

Trond Myklebust (4):
NFSd: Remove 'inline' designation for free_client()
NFSd: Mark nfs4_free_lockowner and nfs4_free_openowner as static functions
NFSd: Clean up nfs4_preprocess_stateid_op
SUNRPC: Fix a module reference leak in svc_handle_xprt

Documentation/filesystems/nfs/nfs41-server.txt | 2 -
fs/lockd/clnt4xdr.c | 2 +
fs/lockd/clntxdr.c | 2 +
fs/lockd/svc.c | 2 +-
fs/lockd/svcsubs.c | 3 +-
fs/lockd/xdr.c | 2 +
fs/nfs/nfs4proc.c | 2 +-
fs/nfsd/acl.h | 2 +-
fs/nfsd/auth.c | 5 +-
fs/nfsd/export.c | 88 +-
{include/linux => fs}/nfsd/export.h | 14 +-
fs/nfsd/fault_inject.c | 15 +-
fs/nfsd/idmap.h | 4 +-
fs/nfsd/nfs2acl.c | 12 +-
fs/nfsd/nfs3acl.c | 6 +-
fs/nfsd/nfs3xdr.c | 27 +-
fs/nfsd/nfs4acl.c | 12 +-
fs/nfsd/nfs4idmap.c | 42 +-
fs/nfsd/nfs4proc.c | 180 ++-
fs/nfsd/nfs4state.c | 271 ++--
fs/nfsd/nfs4xdr.c | 1933 ++++++++++++++----------
fs/nfsd/nfscache.c | 17 +-
fs/nfsd/nfsctl.c | 1 -
fs/nfsd/nfsd.h | 17 +-
fs/nfsd/nfsfh.c | 25 +-
fs/nfsd/nfsfh.h | 59 +-
fs/nfsd/nfssvc.c | 6 -
fs/nfsd/nfsxdr.c | 15 +-
fs/nfsd/state.h | 5 +-
fs/nfsd/stats.c | 1 -
{include/linux => fs}/nfsd/stats.h | 8 +-
fs/nfsd/vfs.c | 155 +-
fs/nfsd/vfs.h | 10 +-
fs/nfsd/xdr4.h | 23 +-
include/linux/lockd/lockd.h | 2 +-
include/linux/nfs4.h | 2 -
include/linux/nfsd/debug.h | 19 -
include/linux/nfsd/nfsfh.h | 63 -
include/linux/sunrpc/svc.h | 13 +-
include/linux/sunrpc/svc_rdma.h | 3 +-
include/linux/sunrpc/svc_xprt.h | 2 +
include/linux/sunrpc/xdr.h | 3 +
include/uapi/linux/nfsd/nfsfh.h | 32 +-
net/sunrpc/auth_gss/svcauth_gss.c | 2 +
net/sunrpc/cache.c | 2 +-
net/sunrpc/sunrpc.h | 13 +
net/sunrpc/svc_xprt.c | 5 +-
net/sunrpc/svcauth.c | 2 +
net/sunrpc/svcsock.c | 17 +-
net/sunrpc/xdr.c | 174 ++-
net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 643 ++++----
net/sunrpc/xprtrdma/svc_rdma_sendto.c | 230 +--
net/sunrpc/xprtrdma/svc_rdma_transport.c | 69 +-
53 files changed, 2265 insertions(+), 1999 deletions(-)
rename {include/linux => fs}/nfsd/export.h (93%)
rename {include/linux => fs}/nfsd/stats.h (91%)
delete mode 100644 include/linux/nfsd/debug.h
delete mode 100644 include/linux/nfsd/nfsfh.h
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/