From 2283963f27fdd56b185e49a964c290130c7c95ab Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 11 Jan 2009 14:27:17 -0500 Subject: nfsd4: split lockstateid/openstateid release logic The flags here attempt to make the code more general, but I find it actually just adds confusion. I think it's clearer to separate the logic for the open and lock cases entirely. And eventually we may want to separate the stateowner and stateid types as well, as many of the fields aren't shared between the lock and open cases. Also move to eliminate forward references. Start with the stateid's. Signed-off-by: J. Bruce Fields Reviewed-by: Benny Halevy --- fs/nfsd/nfs4state.c | 55 +++++++++++++++++++++++++++++++---------------------- 1 file changed, 32 insertions(+), 23 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index b6f60f48e94b..e83b3c865aa3 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -119,7 +119,6 @@ opaque_hashval(const void *ptr, int nbytes) /* forward declarations */ static void release_stateowner(struct nfs4_stateowner *sop); -static void release_stateid(struct nfs4_stateid *stp, int flags); /* * Delegation state @@ -311,6 +310,34 @@ static struct list_head unconf_id_hashtbl[CLIENT_HASH_SIZE]; static struct list_head client_lru; static struct list_head close_lru; +static void unhash_generic_stateid(struct nfs4_stateid *stp) +{ + list_del(&stp->st_hash); + list_del(&stp->st_perfile); + list_del(&stp->st_perstateowner); +} + +static void free_generic_stateid(struct nfs4_stateid *stp) +{ + put_nfs4_file(stp->st_file); + kmem_cache_free(stateid_slab, stp); +} + +static void release_lock_stateid(struct nfs4_stateid *stp) +{ + unhash_generic_stateid(stp); + locks_remove_posix(stp->st_vfs_file, (fl_owner_t)stp->st_stateowner); + free_generic_stateid(stp); +} + +static void release_open_stateid(struct nfs4_stateid *stp) +{ + unhash_generic_stateid(stp); + release_stateid_lockowners(stp); + nfsd_close(stp->st_vfs_file); + free_generic_stateid(stp); +} + static inline void renew_client(struct nfs4_client *clp) { @@ -1065,9 +1092,9 @@ unhash_stateowner(struct nfs4_stateowner *sop) stp = list_entry(sop->so_stateids.next, struct nfs4_stateid, st_perstateowner); if (sop->so_is_open_owner) - release_stateid(stp, OPEN_STATE); + release_open_stateid(stp); else - release_stateid(stp, LOCK_STATE); + release_lock_stateid(stp); } } @@ -1105,24 +1132,6 @@ init_stateid(struct nfs4_stateid *stp, struct nfs4_file *fp, struct nfsd4_open * stp->st_openstp = NULL; } -static void -release_stateid(struct nfs4_stateid *stp, int flags) -{ - struct file *filp = stp->st_vfs_file; - - list_del(&stp->st_hash); - list_del(&stp->st_perfile); - list_del(&stp->st_perstateowner); - if (flags & OPEN_STATE) { - release_stateid_lockowners(stp); - stp->st_vfs_file = NULL; - nfsd_close(filp); - } else if (flags & LOCK_STATE) - locks_remove_posix(filp, (fl_owner_t) stp->st_stateowner); - put_nfs4_file(stp->st_file); - kmem_cache_free(stateid_slab, stp); -} - static void move_to_close_lru(struct nfs4_stateowner *sop) { @@ -1764,7 +1773,7 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf init_stateid(stp, fp, open); status = nfsd4_truncate(rqstp, current_fh, open); if (status) { - release_stateid(stp, OPEN_STATE); + release_open_stateid(stp); goto out; } } @@ -2373,7 +2382,7 @@ nfsd4_close(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, memcpy(&close->cl_stateid, &stp->st_stateid, sizeof(stateid_t)); /* release_stateid() calls nfsd_close() if needed */ - release_stateid(stp, OPEN_STATE); + release_open_stateid(stp); /* place unused nfs4_stateowners on so_close_lru list to be * released by the laundromat service after the lease period -- cgit v1.2.3 From f1d110caf7d759eae2c02c84343f63d83db9b9be Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 11 Jan 2009 14:37:31 -0500 Subject: nfsd4: remove a forward declaration Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 85 ++++++++++++++++++++++++++--------------------------- 1 file changed, 42 insertions(+), 43 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index e83b3c865aa3..6ab30772496e 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -75,7 +75,6 @@ static stateid_t onestateid; /* bits all 1 */ /* forward declarations */ static struct nfs4_stateid * find_stateid(stateid_t *stid, int flags); static struct nfs4_delegation * find_delegation_stateid(struct inode *ino, stateid_t *stid); -static void release_stateid_lockowners(struct nfs4_stateid *open_stp); static char user_recovery_dirname[PATH_MAX] = "/var/lib/nfs/v4recovery"; static void nfs4_set_recdir(char *recdir); @@ -330,6 +329,20 @@ static void release_lock_stateid(struct nfs4_stateid *stp) free_generic_stateid(stp); } +static void +release_stateid_lockowners(struct nfs4_stateid *open_stp) +{ + struct nfs4_stateowner *lock_sop; + + while (!list_empty(&open_stp->st_lockowners)) { + lock_sop = list_entry(open_stp->st_lockowners.next, + struct nfs4_stateowner, so_perstateid); + /* list_del(&open_stp->st_lockowners); */ + BUG_ON(lock_sop->so_is_open_owner); + release_stateowner(lock_sop); + } +} + static void release_open_stateid(struct nfs4_stateid *stp) { unhash_generic_stateid(stp); @@ -338,6 +351,34 @@ static void release_open_stateid(struct nfs4_stateid *stp) free_generic_stateid(stp); } +static void +unhash_stateowner(struct nfs4_stateowner *sop) +{ + struct nfs4_stateid *stp; + + list_del(&sop->so_idhash); + list_del(&sop->so_strhash); + if (sop->so_is_open_owner) + list_del(&sop->so_perclient); + list_del(&sop->so_perstateid); + while (!list_empty(&sop->so_stateids)) { + stp = list_entry(sop->so_stateids.next, + struct nfs4_stateid, st_perstateowner); + if (sop->so_is_open_owner) + release_open_stateid(stp); + else + release_lock_stateid(stp); + } +} + +static void +release_stateowner(struct nfs4_stateowner *sop) +{ + unhash_stateowner(sop); + list_del(&sop->so_close_lru); + nfs4_put_stateowner(sop); +} + static inline void renew_client(struct nfs4_client *clp) { @@ -1064,48 +1105,6 @@ alloc_init_open_stateowner(unsigned int strhashval, struct nfs4_client *clp, str return sop; } -static void -release_stateid_lockowners(struct nfs4_stateid *open_stp) -{ - struct nfs4_stateowner *lock_sop; - - while (!list_empty(&open_stp->st_lockowners)) { - lock_sop = list_entry(open_stp->st_lockowners.next, - struct nfs4_stateowner, so_perstateid); - /* list_del(&open_stp->st_lockowners); */ - BUG_ON(lock_sop->so_is_open_owner); - release_stateowner(lock_sop); - } -} - -static void -unhash_stateowner(struct nfs4_stateowner *sop) -{ - struct nfs4_stateid *stp; - - list_del(&sop->so_idhash); - list_del(&sop->so_strhash); - if (sop->so_is_open_owner) - list_del(&sop->so_perclient); - list_del(&sop->so_perstateid); - while (!list_empty(&sop->so_stateids)) { - stp = list_entry(sop->so_stateids.next, - struct nfs4_stateid, st_perstateowner); - if (sop->so_is_open_owner) - release_open_stateid(stp); - else - release_lock_stateid(stp); - } -} - -static void -release_stateowner(struct nfs4_stateowner *sop) -{ - unhash_stateowner(sop); - list_del(&sop->so_close_lru); - nfs4_put_stateowner(sop); -} - static inline void init_stateid(struct nfs4_stateid *stp, struct nfs4_file *fp, struct nfsd4_open *open) { struct nfs4_stateowner *sop = open->op_stateowner; -- cgit v1.2.3 From f044ff830f1afe91e4388320f0c7b6e08d2e05f8 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 11 Jan 2009 15:24:04 -0500 Subject: nfsd4: split open/lockowner release code The caller always knows specifically whether it's releasing a lockowner or an openowner, and the code is simpler if we use separate functions (and the apparent recursion is gone). Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 57 ++++++++++++++++++++++++++++++++--------------------- 1 file changed, 34 insertions(+), 23 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 6ab30772496e..41a3590ef2cc 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -116,9 +116,6 @@ opaque_hashval(const void *ptr, int nbytes) return x; } -/* forward declarations */ -static void release_stateowner(struct nfs4_stateowner *sop); - /* * Delegation state */ @@ -329,6 +326,26 @@ static void release_lock_stateid(struct nfs4_stateid *stp) free_generic_stateid(stp); } +static void unhash_lockowner(struct nfs4_stateowner *sop) +{ + struct nfs4_stateid *stp; + + list_del(&sop->so_idhash); + list_del(&sop->so_strhash); + list_del(&sop->so_perstateid); + while (!list_empty(&sop->so_stateids)) { + stp = list_first_entry(&sop->so_stateids, + struct nfs4_stateid, st_perstateowner); + release_lock_stateid(stp); + } +} + +static void release_lockowner(struct nfs4_stateowner *sop) +{ + unhash_lockowner(sop); + nfs4_put_stateowner(sop); +} + static void release_stateid_lockowners(struct nfs4_stateid *open_stp) { @@ -339,7 +356,7 @@ release_stateid_lockowners(struct nfs4_stateid *open_stp) struct nfs4_stateowner, so_perstateid); /* list_del(&open_stp->st_lockowners); */ BUG_ON(lock_sop->so_is_open_owner); - release_stateowner(lock_sop); + release_lockowner(lock_sop); } } @@ -351,30 +368,24 @@ static void release_open_stateid(struct nfs4_stateid *stp) free_generic_stateid(stp); } -static void -unhash_stateowner(struct nfs4_stateowner *sop) +static void unhash_openowner(struct nfs4_stateowner *sop) { struct nfs4_stateid *stp; list_del(&sop->so_idhash); list_del(&sop->so_strhash); - if (sop->so_is_open_owner) - list_del(&sop->so_perclient); - list_del(&sop->so_perstateid); + list_del(&sop->so_perclient); + list_del(&sop->so_perstateid); /* XXX: necessary? */ while (!list_empty(&sop->so_stateids)) { - stp = list_entry(sop->so_stateids.next, - struct nfs4_stateid, st_perstateowner); - if (sop->so_is_open_owner) - release_open_stateid(stp); - else - release_lock_stateid(stp); + stp = list_first_entry(&sop->so_stateids, + struct nfs4_stateid, st_perstateowner); + release_open_stateid(stp); } } -static void -release_stateowner(struct nfs4_stateowner *sop) +static void release_openowner(struct nfs4_stateowner *sop) { - unhash_stateowner(sop); + unhash_openowner(sop); list_del(&sop->so_close_lru); nfs4_put_stateowner(sop); } @@ -488,7 +499,7 @@ expire_client(struct nfs4_client *clp) list_del(&clp->cl_lru); while (!list_empty(&clp->cl_openowners)) { sop = list_entry(clp->cl_openowners.next, struct nfs4_stateowner, so_perclient); - release_stateowner(sop); + release_openowner(sop); } put_nfs4_client(clp); } @@ -1443,7 +1454,7 @@ nfsd4_process_open1(struct nfsd4_open *open) if (!sop->so_confirmed) { /* Replace unconfirmed owners without checking for replay. */ clp = sop->so_client; - release_stateowner(sop); + release_openowner(sop); open->op_stateowner = NULL; goto renew; } @@ -1906,7 +1917,7 @@ nfs4_laundromat(void) } dprintk("NFSD: purging unused open stateowner (so_id %d)\n", sop->so_id); - release_stateowner(sop); + release_openowner(sop); } if (clientid_val < NFSD_LAUNDROMAT_MINTIMEOUT) clientid_val = NFSD_LAUNDROMAT_MINTIMEOUT; @@ -2796,7 +2807,7 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, } out: if (status && lock->lk_is_new && lock_sop) - release_stateowner(lock_sop); + release_lockowner(lock_sop); if (lock->lk_replay_owner) { nfs4_get_stateowner(lock->lk_replay_owner); cstate->replay_owner = lock->lk_replay_owner; @@ -3045,7 +3056,7 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp, /* unhash_stateowner deletes so_perclient only * for openowners. */ list_del(&sop->so_perclient); - release_stateowner(sop); + release_lockowner(sop); } out: nfs4_unlock_state(); -- cgit v1.2.3 From 12214cb78147d42da28bc0d1b2917584e0a8efcd Mon Sep 17 00:00:00 2001 From: Qinghuang Feng Date: Mon, 12 Jan 2009 03:13:53 +0800 Subject: NFSD: cleanup for nfs3proc.c MSDOS_SUPER_MAGIC is defined in , so use MSDOS_SUPER_MAGIC directly. Signed-off-by: Qinghuang Feng Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs3proc.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c index 9dbd2eb91281..579ce8c69daa 100644 --- a/fs/nfsd/nfs3proc.c +++ b/fs/nfsd/nfs3proc.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include @@ -569,7 +570,7 @@ nfsd3_proc_fsinfo(struct svc_rqst * rqstp, struct nfsd_fhandle *argp, struct super_block *sb = argp->fh.fh_dentry->d_inode->i_sb; /* Note that we don't care for remote fs's here */ - if (sb->s_magic == 0x4d44 /* MSDOS_SUPER_MAGIC */) { + if (sb->s_magic == MSDOS_SUPER_MAGIC) { resp->f_properties = NFS3_FSF_BILLYBOY; } resp->f_maxfilesize = sb->s_maxbytes; @@ -610,7 +611,7 @@ nfsd3_proc_pathconf(struct svc_rqst * rqstp, struct nfsd_fhandle *argp, resp->p_link_max = EXT2_LINK_MAX; resp->p_name_max = EXT2_NAME_LEN; break; - case 0x4d44: /* MSDOS_SUPER_MAGIC */ + case MSDOS_SUPER_MAGIC: resp->p_case_insensitive = 1; resp->p_case_preserving = 0; break; -- cgit v1.2.3 From 686665619e4424b4f80c3a49e3e1229f27ea565c Mon Sep 17 00:00:00 2001 From: Manish Katiyar Date: Thu, 29 Jan 2009 20:53:57 +0530 Subject: nfsd : Define NFSD only when FILE_LOCKING is enabled Enable NFSD only when FILE_LOCKING is enabled, since we don't want to support NFSD without FILE_LOCKING. Signed-off-by: Manish Katiyar Signed-off-by: J. Bruce Fields --- fs/nfsd/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig index 44d7d04dab95..503b9da159a3 100644 --- a/fs/nfsd/Kconfig +++ b/fs/nfsd/Kconfig @@ -1,6 +1,7 @@ config NFSD tristate "NFS server support" depends on INET + depends on FILE_LOCKING select LOCKD select SUNRPC select EXPORTFS -- cgit v1.2.3 From 99f88726381f676bba6e7dcf74b7412857d7946a Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 2 Feb 2009 15:12:27 -0500 Subject: nfsd: clarify exclusive create bitmask result. The use of |= is confusing--the bitmask is always initialized to zero in this case, so we're effectively just doing an assignment here. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 9fa60a3ad48c..85b9a084177a 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -103,11 +103,13 @@ do_open_lookup(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_o (u32 *)open->op_verf.data, &open->op_truncate, &created); - /* If we ever decide to use different attrs to store the - * verifier in nfsd_create_v3, then we'll need to change this + /* + * Following rfc 3530 14.2.16, use the returned bitmask + * to indicate which attributes we used to store the + * verifier: */ if (open->op_createmode == NFS4_CREATE_EXCLUSIVE && status == 0) - open->op_bmval[1] |= (FATTR4_WORD1_TIME_ACCESS | + open->op_bmval[1] = (FATTR4_WORD1_TIME_ACCESS | FATTR4_WORD1_TIME_MODIFY); } else { status = nfsd_lookup(rqstp, current_fh, -- cgit v1.2.3 From 13024b7b4097d33fe6bba34b1c83602e88753270 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 2 Feb 2009 17:04:03 -0500 Subject: nfsd4: fix misplaced comment Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 85b9a084177a..326506272258 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -120,9 +120,9 @@ do_open_lookup(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_o goto out; set_change_info(&open->op_cinfo, current_fh); + fh_dup2(current_fh, &resfh); /* set reply cache */ - fh_dup2(current_fh, &resfh); open->op_stateowner->so_replay.rp_openfh_len = resfh.fh_handle.fh_size; memcpy(open->op_stateowner->so_replay.rp_openfh, &resfh.fh_handle.fh_base, resfh.fh_handle.fh_size); -- cgit v1.2.3 From a4773c08f2872626cb923433284488fbe8acb0ae Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 2 Feb 2009 17:23:10 -0500 Subject: nfsd4: use helper for copying filehandles for replay Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 17 ++++++----------- 1 file changed, 6 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 326506272258..af66073ed423 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -123,10 +123,8 @@ do_open_lookup(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_o fh_dup2(current_fh, &resfh); /* set reply cache */ - open->op_stateowner->so_replay.rp_openfh_len = resfh.fh_handle.fh_size; - memcpy(open->op_stateowner->so_replay.rp_openfh, - &resfh.fh_handle.fh_base, resfh.fh_handle.fh_size); - + fh_copy_shallow(&open->op_stateowner->so_replay.rp_openfh, + &resfh.fh_handle); if (!created) status = do_open_permission(rqstp, current_fh, open, NFSD_MAY_NOP); @@ -152,10 +150,8 @@ do_open_fhandle(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_ memset(&open->op_cinfo, 0, sizeof(struct nfsd4_change_info)); /* set replay cache */ - open->op_stateowner->so_replay.rp_openfh_len = current_fh->fh_handle.fh_size; - memcpy(open->op_stateowner->so_replay.rp_openfh, - ¤t_fh->fh_handle.fh_base, - current_fh->fh_handle.fh_size); + fh_copy_shallow(&open->op_stateowner->so_replay.rp_openfh, + ¤t_fh->fh_handle); open->op_truncate = (open->op_iattr.ia_valid & ATTR_SIZE) && (open->op_iattr.ia_size == 0); @@ -187,9 +183,8 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (status == nfserr_replay_me) { struct nfs4_replay *rp = &open->op_stateowner->so_replay; fh_put(&cstate->current_fh); - cstate->current_fh.fh_handle.fh_size = rp->rp_openfh_len; - memcpy(&cstate->current_fh.fh_handle.fh_base, rp->rp_openfh, - rp->rp_openfh_len); + fh_copy_shallow(&cstate->current_fh.fh_handle, + &rp->rp_openfh); status = fh_verify(rqstp, &cstate->current_fh, 0, NFSD_MAY_NOP); if (status) dprintk("nfsd4_open: replay failed" -- cgit v1.2.3 From 6c02eaa1d1e53b9b2cc27d0c6fff3e57da4b611f Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 2 Feb 2009 17:30:51 -0500 Subject: nfsd4: use helper for copying delegation filehandle Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 4 ++-- fs/nfsd/nfs4state.c | 4 +--- 2 files changed, 3 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index c464181b5994..c6804db33c17 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -218,7 +218,7 @@ static int encode_cb_recall(struct xdr_stream *xdr, struct nfs4_cb_recall *cb_rec) { __be32 *p; - int len = cb_rec->cbr_fhlen; + int len = cb_rec->cbr_fh.fh_size; RESERVE_SPACE(12+sizeof(cb_rec->cbr_stateid) + len); WRITE32(OP_CB_RECALL); @@ -226,7 +226,7 @@ encode_cb_recall(struct xdr_stream *xdr, struct nfs4_cb_recall *cb_rec) WRITEMEM(&cb_rec->cbr_stateid.si_opaque, sizeof(stateid_opaque_t)); WRITE32(cb_rec->cbr_trunc); WRITE32(len); - WRITEMEM(cb_rec->cbr_fhval, len); + WRITEMEM(&cb_rec->cbr_fh.fh_base, len); return 0; } diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 41a3590ef2cc..7f616e928a57 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -215,9 +215,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f dp->dl_stateid.si_stateownerid = current_delegid++; dp->dl_stateid.si_fileid = 0; dp->dl_stateid.si_generation = 0; - dp->dl_fhlen = current_fh->fh_handle.fh_size; - memcpy(dp->dl_fhval, ¤t_fh->fh_handle.fh_base, - current_fh->fh_handle.fh_size); + fh_copy_shallow(&dp->dl_fh, ¤t_fh->fh_handle); dp->dl_time = 0; atomic_set(&dp->dl_count, 1); list_add(&dp->dl_perfile, &fp->fi_delegations); -- cgit v1.2.3 From e37da04ed145d45c2a698d7cb373a7e1191fbe86 Mon Sep 17 00:00:00 2001 From: Alexandros Batsakis Date: Thu, 18 Dec 2008 19:55:16 -0800 Subject: nfsd: lock state around put client and delegation in nfsd4_cb_recall not having the state locked before putting the client/delegation causes a bug. Also removed the comment from the function header about the state being already locked Signed-off-by: Alexandros Batsakis Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index c6804db33c17..3ddc9fb2e358 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -451,7 +451,6 @@ nfsd4_probe_callback(struct nfs4_client *clp) /* * called with dp->dl_count inc'ed. - * nfs4_lock_state() may or may not have been called. */ void nfsd4_cb_recall(struct nfs4_delegation *dp) @@ -491,7 +490,9 @@ out_put_cred: * Success or failure, now we're either waiting for lease expiration * or deleg_return. */ + nfs4_lock_state(); put_nfs4_client(clp); nfs4_put_delegation(dp); + nfs4_unlock_state(); return; } -- cgit v1.2.3 From e33d1ea60c3a17b8b5c2910b1eef4c1faf0ac450 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Mon, 9 Feb 2009 12:30:43 -0500 Subject: lockd: clean up blocking lock cases of nlsmvc_lock() No change in behavior, just rearranging the switch so that we break out of the switch if and only if we're in the wait case. Signed-off-by: Miklos Szeredi Signed-off-by: J. Bruce Fields --- fs/lockd/svclock.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index 763b78a6e9de..83ee34203bd7 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -426,8 +426,15 @@ nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file, ret = nlm_granted; goto out; case -EAGAIN: + /* + * If this is a blocking request for an + * already pending lock request then we need + * to put it back on lockd's block list + */ + if (wait) + break; ret = nlm_lck_denied; - break; + goto out; case FILE_LOCK_DEFERRED: if (wait) break; @@ -443,10 +450,6 @@ nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file, goto out; } - ret = nlm_lck_denied; - if (!wait) - goto out; - ret = nlm_lck_blocked; /* Append to list of blocked */ -- cgit v1.2.3 From 4ac35c2f794503d3acda20d98e89cf63f6e94332 Mon Sep 17 00:00:00 2001 From: wengang wang Date: Tue, 10 Feb 2009 11:27:51 +0800 Subject: nfsd(v2/v3): fix the failure of creation from HPUX client sometimes HPUX nfs client sends a create request to linux nfs server(v2/v3). the dump of the request is like: obj_attributes mode: value follows set_it: value follows (1) mode: 00 uid: no value set_it: no value (0) gid: value follows set_it: value follows (1) gid: 8030 size: value follows set_it: value follows (1) size: 0 atime: don't change set_it: don't change (0) mtime: don't change set_it: don't change (0) note that mode is 00(havs no rwx privilege even for the owner) and it requires to set size to 0. as current nfsd(v2/v3) implementation, the server does mainly 2 steps: 1) creates the file in mode specified by calling vfs_create(). 2) sets attributes for the file by calling nfsd_setattr(). at step 2), it finally calls file system specific setattr() function which may fail when checking permission because changing size needs WRITE privilege but it has none since mode is 000. for this case, a new file created, we may simply ignore the request of setting size to 0, so that WRITE privilege is not needed and the open succeeds. Signed-off-by: Wengang Wang -- vfs.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 6e50aaa56ca2..0c076293155d 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -1176,6 +1176,21 @@ nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *resfhp, return 0; } +/* HPUX client sometimes creates a file in mode 000, and sets size to 0. + * setting size to 0 may fail for some specific file systems by the permission + * checking which requires WRITE permission but the mode is 000. + * we ignore the resizing(to 0) on the just new created file, since the size is + * 0 after file created. + * + * call this only after vfs_create() is called. + * */ +static void +nfsd_check_ignore_resizing(struct iattr *iap) +{ + if ((iap->ia_valid & ATTR_SIZE) && (iap->ia_size == 0)) + iap->ia_valid &= ~ATTR_SIZE; +} + /* * Create a file (regular, directory, device, fifo); UNIX sockets * not yet implemented. @@ -1271,6 +1286,8 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, switch (type) { case S_IFREG: host_err = vfs_create(dirp, dchild, iap->ia_mode, NULL); + if (!host_err) + nfsd_check_ignore_resizing(iap); break; case S_IFDIR: host_err = vfs_mkdir(dirp, dchild, iap->ia_mode); @@ -1424,6 +1441,8 @@ nfsd_create_v3(struct svc_rqst *rqstp, struct svc_fh *fhp, /* setattr will sync the child (or not) */ } + nfsd_check_ignore_resizing(iap); + if (createmode == NFS3_CREATE_EXCLUSIVE) { /* Cram the verifier into atime/mtime */ iap->ia_valid = ATTR_MTIME|ATTR_ATIME -- cgit v1.2.3 From a4455be0850009f5da9a3b82523079922cd4b26e Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 10:40:22 -0800 Subject: nfsd4: trivial preprocess_stateid_op cleanup Remove a couple redundant comments, adjust style; no change in behavior. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 7f616e928a57..b7e2f251ea95 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2072,21 +2072,21 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl if (ZERO_STATEID(stateid) || ONE_STATEID(stateid)) return check_special_stateids(current_fh, stateid, flags); - /* STALE STATEID */ status = nfserr_stale_stateid; if (STALE_STATEID(stateid)) goto out; - /* BAD STATEID */ status = nfserr_bad_stateid; if (!stateid->si_fileid) { /* delegation stateid */ - if(!(dp = find_delegation_stateid(ino, stateid))) { + dp = find_delegation_stateid(ino, stateid); + if (!dp) { dprintk("NFSD: delegation stateid not found\n"); goto out; } stidp = &dp->dl_stateid; } else { /* open or lock stateid */ - if (!(stp = find_stateid(stateid, flags))) { + stp = find_stateid(stateid, flags); + if (!stp) { dprintk("NFSD: open or lock stateid not found\n"); goto out; } @@ -2100,13 +2100,15 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl if (status) goto out; if (stp) { - if ((status = nfs4_check_openmode(stp,flags))) + status = nfs4_check_openmode(stp, flags); + if (status) goto out; renew_client(stp->st_stateowner->so_client); if (filpp) *filpp = stp->st_vfs_file; } else { - if ((status = nfs4_check_delegmode(dp, flags))) + status = nfs4_check_delegmode(dp, flags); + if (status) goto out; renew_client(dp->dl_client); if (flags & DELEG_RET) -- cgit v1.2.3 From 0c2a498fa6d33d8ca9c8a0c29039c41e1734cb9e Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 11:11:50 -0800 Subject: nfsd4: move check_stateid_generation check No change in behavior. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index b7e2f251ea95..d6ca2be306dc 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2084,6 +2084,9 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl goto out; } stidp = &dp->dl_stateid; + status = check_stateid_generation(stateid, stidp); + if (status) + goto out; } else { /* open or lock stateid */ stp = find_stateid(stateid, flags); if (!stp) { @@ -2095,10 +2098,10 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl if (!stp->st_stateowner->so_confirmed) goto out; stidp = &stp->st_stateid; + status = check_stateid_generation(stateid, stidp); + if (status) + goto out; } - status = check_stateid_generation(stateid, stidp); - if (status) - goto out; if (stp) { status = nfs4_check_openmode(stp, flags); if (status) -- cgit v1.2.3 From dc9bf700ed2fc3eab50f31000b13fda781e7c9f1 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 11:14:43 -0800 Subject: nfsd4: remove redundant "if" in nfs4_preprocess_stateid_op Note that we exit this first big "if" with stp == NULL if and only if we took the first branch; therefore, the second "if" is redundant, and we can just combine the two, simplifying the logic. Reviewed-by: Yang Hongyang Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 19 ++++++++----------- 1 file changed, 8 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index d6ca2be306dc..16fcb656f47f 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2087,6 +2087,14 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl status = check_stateid_generation(stateid, stidp); if (status) goto out; + status = nfs4_check_delegmode(dp, flags); + if (status) + goto out; + renew_client(dp->dl_client); + if (flags & DELEG_RET) + unhash_delegation(dp); + if (filpp) + *filpp = dp->dl_vfs_file; } else { /* open or lock stateid */ stp = find_stateid(stateid, flags); if (!stp) { @@ -2101,23 +2109,12 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl status = check_stateid_generation(stateid, stidp); if (status) goto out; - } - if (stp) { status = nfs4_check_openmode(stp, flags); if (status) goto out; renew_client(stp->st_stateowner->so_client); if (filpp) *filpp = stp->st_vfs_file; - } else { - status = nfs4_check_delegmode(dp, flags); - if (status) - goto out; - renew_client(dp->dl_client); - if (flags & DELEG_RET) - unhash_delegation(dp); - if (filpp) - *filpp = dp->dl_vfs_file; } status = nfs_ok; out: -- cgit v1.2.3 From fd03b09906c32aea7b47f1275c9cd6034141159d Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 11:27:30 -0800 Subject: nfsd4: remove unneeded local variable We no longer need stidp. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 16fcb656f47f..099938ebc9d3 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2056,7 +2056,6 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl { struct nfs4_stateid *stp = NULL; struct nfs4_delegation *dp = NULL; - stateid_t *stidp; struct inode *ino = current_fh->fh_dentry->d_inode; __be32 status; @@ -2083,8 +2082,7 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl dprintk("NFSD: delegation stateid not found\n"); goto out; } - stidp = &dp->dl_stateid; - status = check_stateid_generation(stateid, stidp); + status = check_stateid_generation(stateid, &dp->dl_stateid); if (status) goto out; status = nfs4_check_delegmode(dp, flags); @@ -2105,8 +2103,7 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl goto out; if (!stp->st_stateowner->so_confirmed) goto out; - stidp = &stp->st_stateid; - status = check_stateid_generation(stateid, stidp); + status = check_stateid_generation(stateid, &stp->st_stateid); if (status) goto out; status = nfs4_check_openmode(stp, flags); -- cgit v1.2.3 From 819a8f539acf7838d62fec20e88401ff53303cd1 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 12:13:24 -0800 Subject: nfsd4: remove some dprintk's I can't recall ever seeing these printk's used to debug a problem. I'll happily put them back if we see a case where they'd be useful. (Though if we do that the find_XXX() errors would probably be better reported in find_XXX() functions themselves.) Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 11 ++--------- 1 file changed, 2 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 099938ebc9d3..909a7a493688 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2059,9 +2059,6 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl struct inode *ino = current_fh->fh_dentry->d_inode; __be32 status; - dprintk("NFSD: preprocess_stateid_op: stateid = (%08x/%08x/%08x/%08x)\n", - stateid->si_boot, stateid->si_stateownerid, - stateid->si_fileid, stateid->si_generation); if (filpp) *filpp = NULL; @@ -2078,10 +2075,8 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl status = nfserr_bad_stateid; if (!stateid->si_fileid) { /* delegation stateid */ dp = find_delegation_stateid(ino, stateid); - if (!dp) { - dprintk("NFSD: delegation stateid not found\n"); + if (!dp) goto out; - } status = check_stateid_generation(stateid, &dp->dl_stateid); if (status) goto out; @@ -2095,10 +2090,8 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl *filpp = dp->dl_vfs_file; } else { /* open or lock stateid */ stp = find_stateid(stateid, flags); - if (!stp) { - dprintk("NFSD: open or lock stateid not found\n"); + if (!stp) goto out; - } if ((flags & CHECK_FH) && nfs4_check_fh(current_fh, stp)) goto out; if (!stp->st_stateowner->so_confirmed) -- cgit v1.2.3 From 3e633079e377d2b527a8390f63ceb887b5cabfbf Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 13:17:19 -0800 Subject: nfsd4: add a helper function to decide if stateid is delegation Make this check self-documenting. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 909a7a493688..d5555850cb64 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2048,6 +2048,11 @@ static int check_stateid_generation(stateid_t *in, stateid_t *ref) return nfs_ok; } +static int is_delegation_stateid(stateid_t *stateid) +{ + return stateid->si_fileid == 0; +} + /* * Checks for stateid operations */ @@ -2073,7 +2078,7 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl goto out; status = nfserr_bad_stateid; - if (!stateid->si_fileid) { /* delegation stateid */ + if (is_delegation_stateid(stateid)) { dp = find_delegation_stateid(ino, stateid); if (!dp) goto out; -- cgit v1.2.3 From 203a8c8e66278a5936a230edaac29017e50c88fb Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 13:29:14 -0800 Subject: nfsd4: separate delegreturn case from preprocess_stateid_op Delegreturn is enough a special case for preprocess_stateid_op to warrant just open-coding it in delegreturn. There should be no change in behavior here; we're just reshuffling code. Thanks to Yang Hongyang for catching a critical typo. Reviewed-by: Yang Hongyang Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 40 ++++++++++++++++++++++++++++------------ 1 file changed, 28 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index d5555850cb64..3570a0d1133f 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2000,10 +2000,7 @@ out: static inline __be32 check_special_stateids(svc_fh *current_fh, stateid_t *stateid, int flags) { - /* Trying to call delegreturn with a special stateid? Yuch: */ - if (!(flags & (RD_STATE | WR_STATE))) - return nfserr_bad_stateid; - else if (ONE_STATEID(stateid) && (flags & RD_STATE)) + if (ONE_STATEID(stateid) && (flags & RD_STATE)) return nfs_ok; else if (locks_in_grace()) { /* Answer in remaining cases depends on existance of @@ -2024,8 +2021,7 @@ check_special_stateids(svc_fh *current_fh, stateid_t *stateid, int flags) static inline int io_during_grace_disallowed(struct inode *inode, int flags) { - return locks_in_grace() && (flags & (RD_STATE | WR_STATE)) - && mandatory_lock(inode); + return locks_in_grace() && mandatory_lock(inode); } static int check_stateid_generation(stateid_t *in, stateid_t *ref) @@ -2089,8 +2085,6 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl if (status) goto out; renew_client(dp->dl_client); - if (flags & DELEG_RET) - unhash_delegation(dp); if (filpp) *filpp = dp->dl_vfs_file; } else { /* open or lock stateid */ @@ -2408,16 +2402,38 @@ __be32 nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_delegreturn *dr) { + struct nfs4_delegation *dp; + stateid_t *stateid = &dr->dr_stateid; + struct inode *inode; __be32 status; if ((status = fh_verify(rqstp, &cstate->current_fh, S_IFREG, 0))) - goto out; + return status; + inode = cstate->current_fh.fh_dentry->d_inode; nfs4_lock_state(); - status = nfs4_preprocess_stateid_op(&cstate->current_fh, - &dr->dr_stateid, DELEG_RET, NULL); - nfs4_unlock_state(); + status = nfserr_bad_stateid; + if (ZERO_STATEID(stateid) || ONE_STATEID(stateid)) + goto out; + status = nfserr_stale_stateid; + if (STALE_STATEID(stateid)) + goto out; + status = nfs_ok; + if (!is_delegation_stateid(stateid)) + goto out; + status = nfserr_bad_stateid; + dp = find_delegation_stateid(inode, stateid); + if (!dp) + goto out; + status = check_stateid_generation(stateid, &dp->dl_stateid); + if (status) + goto out; + renew_client(dp->dl_client); + + unhash_delegation(dp); out: + nfs4_unlock_state(); + return status; } -- cgit v1.2.3 From 7e0f7cf582abd6c85232331dfe726a4e4b0fd98e Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 13:32:28 -0800 Subject: nfsd4: fail when delegreturn gets a non-delegation stateid Previous cleanup reveals an obvious (though harmless) bug: when delegreturn gets a stateid that isn't for a delegation, it should return an error rather than doing nothing. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 3570a0d1133f..6ae28e606afc 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2418,10 +2418,9 @@ nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, status = nfserr_stale_stateid; if (STALE_STATEID(stateid)) goto out; - status = nfs_ok; + status = nfserr_bad_stateid; if (!is_delegation_stateid(stateid)) goto out; - status = nfserr_bad_stateid; dp = find_delegation_stateid(inode, stateid); if (!dp) goto out; -- cgit v1.2.3 From 6150ef0dc7f734366d297e2eb5697ae458a1ea19 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 13:36:16 -0800 Subject: nfsd4: remove unused CHECK_FH flag All users now pass this, so it's meaningless. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 6 +++--- fs/nfsd/nfs4state.c | 2 +- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index af66073ed423..77f584f69dfe 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -519,7 +519,7 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, /* check stateid */ if ((status = nfs4_preprocess_stateid_op(&cstate->current_fh, &read->rd_stateid, - CHECK_FH | RD_STATE, &read->rd_filp))) { + RD_STATE, &read->rd_filp))) { dprintk("NFSD: nfsd4_read: couldn't process stateid!\n"); goto out; } @@ -651,7 +651,7 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (setattr->sa_iattr.ia_valid & ATTR_SIZE) { nfs4_lock_state(); status = nfs4_preprocess_stateid_op(&cstate->current_fh, - &setattr->sa_stateid, CHECK_FH | WR_STATE, NULL); + &setattr->sa_stateid, WR_STATE, NULL); nfs4_unlock_state(); if (status) { dprintk("NFSD: nfsd4_setattr: couldn't process stateid!\n"); @@ -690,7 +690,7 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nfs4_lock_state(); status = nfs4_preprocess_stateid_op(&cstate->current_fh, stateid, - CHECK_FH | WR_STATE, &filp); + WR_STATE, &filp); if (filp) get_file(filp); nfs4_unlock_state(); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 6ae28e606afc..5957f7766bdc 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2091,7 +2091,7 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl stp = find_stateid(stateid, flags); if (!stp) goto out; - if ((flags & CHECK_FH) && nfs4_check_fh(current_fh, stp)) + if (nfs4_check_fh(current_fh, stp)) goto out; if (!stp->st_stateowner->so_confirmed) goto out; -- cgit v1.2.3 From 18f82731b7784ba81ee9b1ed6a8179b577fa898b Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 15:23:01 -0800 Subject: nfsd4: rename io_during_grace_disallowed Use a slightly clearer, more concise name. Also removed unused argument. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 5957f7766bdc..89e575e7daea 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2019,7 +2019,7 @@ check_special_stateids(svc_fh *current_fh, stateid_t *stateid, int flags) * that are not able to provide mandatory locking. */ static inline int -io_during_grace_disallowed(struct inode *inode, int flags) +grace_disallows_io(struct inode *inode) { return locks_in_grace() && mandatory_lock(inode); } @@ -2063,7 +2063,7 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl if (filpp) *filpp = NULL; - if (io_during_grace_disallowed(ino, flags)) + if (grace_disallows_io(ino)) return nfserr_grace; if (ZERO_STATEID(stateid) || ONE_STATEID(stateid)) -- cgit v1.2.3 From d7fdcfe0aaaf6dffca6fa857bab374182fe7ca8b Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 21 Feb 2009 15:39:54 -0800 Subject: nfsd4: put_nfs4_client does not require state lock Since free_client() is guaranteed to only be called once, and to only touch the client structure itself (not any common data structures), it has no need for the state lock. Signed-off-by: J. Bruce Fields Cc: Alexandros Batsakis --- fs/nfsd/nfs4callback.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 3ddc9fb2e358..3fd7136321ca 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -490,8 +490,8 @@ out_put_cred: * Success or failure, now we're either waiting for lease expiration * or deleg_return. */ - nfs4_lock_state(); put_nfs4_client(clp); + nfs4_lock_state(); nfs4_put_delegation(dp); nfs4_unlock_state(); return; -- cgit v1.2.3 From 8b671b80707e4fc76adfe4387df07b3be1007c1e Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 22 Feb 2009 14:51:34 -0800 Subject: nfsd4: remove use of mutex for file_hashtable As part of reducing the scope of the client_mutex, and in order to remove the need for mutexes from the callback code (so that callbacks can be done as asynchronous rpc calls), move manipulations of the file_hashtable under the recall_lock. Update the relevant comments while we're here. Signed-off-by: J. Bruce Fields Cc: Alexandros Batsakis Reviewed-by: Benny Halevy --- fs/nfsd/nfs4callback.c | 2 -- fs/nfsd/nfs4state.c | 47 +++++++++++++++++++++++------------------------ 2 files changed, 23 insertions(+), 26 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 3fd7136321ca..5dcd38e5f138 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -491,8 +491,6 @@ out_put_cred: * or deleg_return. */ put_nfs4_client(clp); - nfs4_lock_state(); nfs4_put_delegation(dp); - nfs4_unlock_state(); return; } diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 89e575e7daea..54651aa45790 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -78,14 +78,18 @@ static struct nfs4_delegation * find_delegation_stateid(struct inode *ino, state static char user_recovery_dirname[PATH_MAX] = "/var/lib/nfs/v4recovery"; static void nfs4_set_recdir(char *recdir); -/* Locking: - * - * client_mutex: - * protects clientid_hashtbl[], clientstr_hashtbl[], - * unconfstr_hashtbl[], uncofid_hashtbl[]. - */ +/* Locking: */ + +/* Currently used for almost all code touching nfsv4 state: */ static DEFINE_MUTEX(client_mutex); +/* + * Currently used for the del_recall_lru and file hash table. In an + * effort to decrease the scope of the client_mutex, this spinlock may + * eventually cover more: + */ +static DEFINE_SPINLOCK(recall_lock); + static struct kmem_cache *stateowner_slab = NULL; static struct kmem_cache *file_slab = NULL; static struct kmem_cache *stateid_slab = NULL; @@ -116,33 +120,23 @@ opaque_hashval(const void *ptr, int nbytes) return x; } -/* - * Delegation state - */ - -/* recall_lock protects the del_recall_lru */ -static DEFINE_SPINLOCK(recall_lock); static struct list_head del_recall_lru; -static void -free_nfs4_file(struct kref *kref) -{ - struct nfs4_file *fp = container_of(kref, struct nfs4_file, fi_ref); - list_del(&fp->fi_hash); - iput(fp->fi_inode); - kmem_cache_free(file_slab, fp); -} - static inline void put_nfs4_file(struct nfs4_file *fi) { - kref_put(&fi->fi_ref, free_nfs4_file); + if (atomic_dec_and_lock(&fi->fi_ref, &recall_lock)) { + list_del(&fi->fi_hash); + spin_unlock(&recall_lock); + iput(fi->fi_inode); + kmem_cache_free(file_slab, fi); + } } static inline void get_nfs4_file(struct nfs4_file *fi) { - kref_get(&fi->fi_ref); + atomic_inc(&fi->fi_ref); } static int num_delegations; @@ -1000,11 +994,13 @@ alloc_init_file(struct inode *ino) fp = kmem_cache_alloc(file_slab, GFP_KERNEL); if (fp) { - kref_init(&fp->fi_ref); + atomic_set(&fp->fi_ref, 1); INIT_LIST_HEAD(&fp->fi_hash); INIT_LIST_HEAD(&fp->fi_stateids); INIT_LIST_HEAD(&fp->fi_delegations); + spin_lock(&recall_lock); list_add(&fp->fi_hash, &file_hashtbl[hashval]); + spin_unlock(&recall_lock); fp->fi_inode = igrab(ino); fp->fi_id = current_fileid++; fp->fi_had_conflict = false; @@ -1177,12 +1173,15 @@ find_file(struct inode *ino) unsigned int hashval = file_hashval(ino); struct nfs4_file *fp; + spin_lock(&recall_lock); list_for_each_entry(fp, &file_hashtbl[hashval], fi_hash) { if (fp->fi_inode == ino) { get_nfs4_file(fp); + spin_unlock(&recall_lock); return fp; } } + spin_unlock(&recall_lock); return NULL; } -- cgit v1.2.3 From 418cd20aa19b54554cab383e2fd0d1cb8c4732ee Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 22 Feb 2009 15:52:13 -0800 Subject: nfsd4: fix do_probe_callback errors The errors returned aren't used. Just return 0 and make them available to a dprintk(). Also, consistently use -ERRNO errors instead of nfs errors. Signed-off-by: J. Bruce Fields Reviewed-by: Benny Halevy --- fs/nfsd/nfs4callback.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 5dcd38e5f138..8d55f5047503 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -389,12 +389,10 @@ static int do_probe_callback(void *data) .rpc_argp = clp, }; struct rpc_clnt *client; - int status; + int status = -EINVAL; - if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5)) { - status = nfserr_cb_path_down; + if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5)) goto out_err; - } /* Initialize address */ memset(&addr, 0, sizeof(addr)); @@ -405,8 +403,9 @@ static int do_probe_callback(void *data) /* Create RPC client */ client = rpc_create(&args); if (IS_ERR(client)) { - dprintk("NFSD: couldn't create callback client\n"); status = PTR_ERR(client); + dprintk("NFSD: couldn't create callback client: %d\n", + status); goto out_err; } @@ -422,10 +421,10 @@ static int do_probe_callback(void *data) out_release_client: rpc_shutdown_client(client); out_err: - dprintk("NFSD: warning: no callback path to client %.*s\n", - (int)clp->cl_name.len, clp->cl_name.data); + dprintk("NFSD: warning: no callback path to client %.*s: error %d\n", + (int)clp->cl_name.len, clp->cl_name.data, status); put_nfs4_client(clp); - return status; + return 0; } /* -- cgit v1.2.3 From a601caeda21c0e94c153dbd146ec0899cc5f324f Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 22 Feb 2009 16:43:45 -0800 Subject: nfsd4: move rpc_client setup to a separate function Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 29 +++++++++++++++++++++-------- 1 file changed, 21 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 8d55f5047503..290289bd44f7 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -361,9 +361,8 @@ static struct rpc_program cb_program = { /* Reference counting, callback cleanup, etc., all look racy as heck. * And why is cb_set an atomic? */ -static int do_probe_callback(void *data) +static struct rpc_clnt *setup_callback_client(struct nfs4_client *clp) { - struct nfs4_client *clp = data; struct sockaddr_in addr; struct nfs4_callback *cb = &clp->cl_callback; struct rpc_timeout timeparms = { @@ -384,15 +383,10 @@ static int do_probe_callback(void *data) .flags = (RPC_CLNT_CREATE_NOPING | RPC_CLNT_CREATE_QUIET), .client_name = clp->cl_principal, }; - struct rpc_message msg = { - .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], - .rpc_argp = clp, - }; struct rpc_clnt *client; - int status = -EINVAL; if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5)) - goto out_err; + return ERR_PTR(-EINVAL); /* Initialize address */ memset(&addr, 0, sizeof(addr)); @@ -402,6 +396,25 @@ static int do_probe_callback(void *data) /* Create RPC client */ client = rpc_create(&args); + if (IS_ERR(client)) + dprintk("NFSD: couldn't create callback client: %ld\n", + PTR_ERR(client)); + return client; + +} + +static int do_probe_callback(void *data) +{ + struct nfs4_client *clp = data; + struct nfs4_callback *cb = &clp->cl_callback; + struct rpc_message msg = { + .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], + .rpc_argp = clp, + }; + struct rpc_clnt *client; + int status; + + client = setup_callback_client(clp); if (IS_ERR(client)) { status = PTR_ERR(client); dprintk("NFSD: couldn't create callback client: %d\n", -- cgit v1.2.3 From 1e685ec270cb97680be4eb8cf6b615f5f7f1403a Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 4 Mar 2009 23:06:06 +0200 Subject: NFSD: return nfsv4 error code nfserr_notsupp rather than nfsv[23]'s nfserr_opnotsupp Thanks for Bill Baker at sun.com for catching this at Connectathon 2009. This bug was introduced in 2.6.27 Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 9250067943d8..aaf5f234df25 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1005,7 +1005,7 @@ nfsd4_decode_noop(struct nfsd4_compoundargs *argp, void *p) static __be32 nfsd4_decode_notsupp(struct nfsd4_compoundargs *argp, void *p) { - return nfserr_opnotsupp; + return nfserr_notsupp; } typedef __be32(*nfsd4_dec)(struct nfsd4_compoundargs *argp, void *); -- cgit v1.2.3 From 31dec2538e45e9fff2007ea1f4c6bae9f78db724 Mon Sep 17 00:00:00 2001 From: David Shaw Date: Thu, 5 Mar 2009 20:16:14 -0500 Subject: Short write in nfsd becomes a full write to the client If a filesystem being written to via NFS returns a short write count (as opposed to an error) to nfsd, nfsd treats that as a success for the entire write, rather than the short count that actually succeeded. For example, given a 8192 byte write, if the underlying filesystem only writes 4096 bytes, nfsd will ack back to the nfs client that all 8192 bytes were written. The nfs client does have retry logic for short writes, but this is never called as the client is told the complete write succeeded. There are probably other ways it could happen, but in my case it happened with a fuse (filesystem in userspace) filesystem which can rather easily have a partial write. Here is a patch to properly return the short write count to the client. Signed-off-by: David Shaw Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs3proc.c | 5 +++-- fs/nfsd/nfs4proc.c | 7 +++++-- fs/nfsd/nfsproc.c | 3 ++- fs/nfsd/vfs.c | 13 +++++++------ 4 files changed, 17 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c index 579ce8c69daa..7c9fe838f038 100644 --- a/fs/nfsd/nfs3proc.c +++ b/fs/nfsd/nfs3proc.c @@ -203,6 +203,7 @@ nfsd3_proc_write(struct svc_rqst *rqstp, struct nfsd3_writeargs *argp, struct nfsd3_writeres *resp) { __be32 nfserr; + unsigned long cnt = argp->len; dprintk("nfsd: WRITE(3) %s %d bytes at %ld%s\n", SVCFH_fmt(&argp->fh), @@ -215,9 +216,9 @@ nfsd3_proc_write(struct svc_rqst *rqstp, struct nfsd3_writeargs *argp, nfserr = nfsd_write(rqstp, &resp->fh, NULL, argp->offset, rqstp->rq_vec, argp->vlen, - argp->len, + &cnt, &resp->committed); - resp->count = argp->count; + resp->count = cnt; RETURN_STATUS(nfserr); } diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 77f584f69dfe..283d77a47120 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -682,6 +682,7 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct file *filp = NULL; u32 *p; __be32 status = nfs_ok; + unsigned long cnt; /* no need to check permission - this will be done in nfsd_write() */ @@ -700,7 +701,7 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, return status; } - write->wr_bytes_written = write->wr_buflen; + cnt = write->wr_buflen; write->wr_how_written = write->wr_stable_how; p = (u32 *)write->wr_verifier.data; *p++ = nfssvc_boot.tv_sec; @@ -708,10 +709,12 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, status = nfsd_write(rqstp, &cstate->current_fh, filp, write->wr_offset, rqstp->rq_vec, write->wr_vlen, - write->wr_buflen, &write->wr_how_written); + &cnt, &write->wr_how_written); if (filp) fput(filp); + write->wr_bytes_written = cnt; + if (status == nfserr_symlink) status = nfserr_inval; return status; diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c index 6f7f26351227..e298e260b5f1 100644 --- a/fs/nfsd/nfsproc.c +++ b/fs/nfsd/nfsproc.c @@ -180,6 +180,7 @@ nfsd_proc_write(struct svc_rqst *rqstp, struct nfsd_writeargs *argp, { __be32 nfserr; int stable = 1; + unsigned long cnt = argp->len; dprintk("nfsd: WRITE %s %d bytes at %d\n", SVCFH_fmt(&argp->fh), @@ -188,7 +189,7 @@ nfsd_proc_write(struct svc_rqst *rqstp, struct nfsd_writeargs *argp, nfserr = nfsd_write(rqstp, fh_copy(&resp->fh, &argp->fh), NULL, argp->offset, rqstp->rq_vec, argp->vlen, - argp->len, + &cnt, &stable); return nfsd_return_attrs(nfserr, resp); } diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 0c076293155d..54404d730809 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -960,7 +960,7 @@ static void kill_suid(struct dentry *dentry) static __be32 nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, loff_t offset, struct kvec *vec, int vlen, - unsigned long cnt, int *stablep) + unsigned long *cnt, int *stablep) { struct svc_export *exp; struct dentry *dentry; @@ -974,7 +974,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, err = nfserr_perm; if ((fhp->fh_export->ex_flags & NFSEXP_MSNFS) && - (!lock_may_write(file->f_path.dentry->d_inode, offset, cnt))) + (!lock_may_write(file->f_path.dentry->d_inode, offset, *cnt))) goto out; #endif @@ -1006,7 +1006,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, host_err = vfs_writev(file, (struct iovec __user *)vec, vlen, &offset); set_fs(oldfs); if (host_err >= 0) { - nfsdstats.io_write += cnt; + nfsdstats.io_write += host_err; fsnotify_modify(file->f_path.dentry); } @@ -1051,9 +1051,10 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, } dprintk("nfsd: write complete host_err=%d\n", host_err); - if (host_err >= 0) + if (host_err >= 0) { err = 0; - else + *cnt = host_err; + } else err = nfserrno(host_err); out: return err; @@ -1095,7 +1096,7 @@ out: */ __be32 nfsd_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, - loff_t offset, struct kvec *vec, int vlen, unsigned long cnt, + loff_t offset, struct kvec *vec, int vlen, unsigned long *cnt, int *stablep) { __be32 err = 0; -- cgit v1.2.3 From a1c8c4d1ff54c6c86930ee3c4c73c69eeb9ede61 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 9 Mar 2009 12:17:29 -0400 Subject: nfsd4: support putpubfh operation Currently putpubfh returns NFSERR_OPNOTSUPP, which isn't actually allowed for v4. The right error is probably NFSERR_NOTSUPP. But let's just implement it; though rarely seen, it can be used by Solaris (with a special mount option), is mandated by the rfc, and is trivial for us to support. Thanks to Yang Hongyang for pointing out the original problem, and to Mike Eisler, Tom Talpey, Trond Myklebust, and Dave Noveck for further argument.... Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 2 +- fs/nfsd/nfs4xdr.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 283d77a47120..36d74cda87a8 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1045,7 +1045,7 @@ static struct nfsd4_operation nfsd4_ops[OP_RELEASE_LOCKOWNER+1] = { .op_name = "OP_PUTFH", }, [OP_PUTPUBFH] = { - /* unsupported, just for future reference: */ + .op_func = (nfsd4op_func)nfsd4_putrootfh, .op_flags = ALLOWED_WITHOUT_FH | ALLOWED_ON_ABSENT_FS, .op_name = "OP_PUTPUBFH", }, diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index aaf5f234df25..76a0b2a8d69b 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1031,7 +1031,7 @@ static nfsd4_dec nfsd4_dec_ops[] = { [OP_OPEN_CONFIRM] = (nfsd4_dec)nfsd4_decode_open_confirm, [OP_OPEN_DOWNGRADE] = (nfsd4_dec)nfsd4_decode_open_downgrade, [OP_PUTFH] = (nfsd4_dec)nfsd4_decode_putfh, - [OP_PUTPUBFH] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_PUTPUBFH] = (nfsd4_dec)nfsd4_decode_noop, [OP_PUTROOTFH] = (nfsd4_dec)nfsd4_decode_noop, [OP_READ] = (nfsd4_dec)nfsd4_decode_read, [OP_READDIR] = (nfsd4_dec)nfsd4_decode_readdir, -- cgit v1.2.3 From 05f4f678b0511a24795a017b5332455077be3b1c Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 13 Mar 2009 16:02:59 -0400 Subject: nfsd4: don't do lookup within readdir in recovery code The main nfsd code was recently modified to no longer do lookups from withing the readdir callback, to avoid locking problems on certain filesystems. This (rather hacky, and overdue for replacement) NFSv4 recovery code has the same problem. Fix it to build up a list of names (instead of dentries) and do the lookups afterwards. Reported symptoms were a deadlock in the xfs code (called from nfsd4_recdir_load), with /var/lib/nfs on xfs. Signed-off-by: J. Bruce Fields Reported-by: David Warren --- fs/nfsd/nfs4recover.c | 71 ++++++++++++++++++++++----------------------------- 1 file changed, 31 insertions(+), 40 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index 74f7b67567fd..b11cf8d34280 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -182,36 +182,26 @@ out_unlock: typedef int (recdir_func)(struct dentry *, struct dentry *); -struct dentry_list { - struct dentry *dentry; +struct name_list { + char name[HEXDIR_LEN]; struct list_head list; }; -struct dentry_list_arg { - struct list_head dentries; - struct dentry *parent; -}; - static int -nfsd4_build_dentrylist(void *arg, const char *name, int namlen, +nfsd4_build_namelist(void *arg, const char *name, int namlen, loff_t offset, u64 ino, unsigned int d_type) { - struct dentry_list_arg *dla = arg; - struct list_head *dentries = &dla->dentries; - struct dentry *parent = dla->parent; - struct dentry *dentry; - struct dentry_list *child; + struct list_head *names = arg; + struct name_list *entry; - if (name && isdotent(name, namlen)) + if (namlen != HEXDIR_LEN - 1) return 0; - dentry = lookup_one_len(name, parent, namlen); - if (IS_ERR(dentry)) - return PTR_ERR(dentry); - child = kmalloc(sizeof(*child), GFP_KERNEL); - if (child == NULL) + entry = kmalloc(sizeof(struct name_list), GFP_KERNEL); + if (entry == NULL) return -ENOMEM; - child->dentry = dentry; - list_add(&child->list, dentries); + memcpy(entry->name, name, HEXDIR_LEN - 1); + entry->name[HEXDIR_LEN - 1] = '\0'; + list_add(&entry->list, names); return 0; } @@ -220,11 +210,9 @@ nfsd4_list_rec_dir(struct dentry *dir, recdir_func *f) { const struct cred *original_cred; struct file *filp; - struct dentry_list_arg dla = { - .parent = dir, - }; - struct list_head *dentries = &dla.dentries; - struct dentry_list *child; + LIST_HEAD(names); + struct name_list *entry; + struct dentry *dentry; int status; if (!rec_dir_init) @@ -233,31 +221,34 @@ nfsd4_list_rec_dir(struct dentry *dir, recdir_func *f) status = nfs4_save_creds(&original_cred); if (status < 0) return status; - INIT_LIST_HEAD(dentries); filp = dentry_open(dget(dir), mntget(rec_dir.mnt), O_RDONLY, current_cred()); status = PTR_ERR(filp); if (IS_ERR(filp)) goto out; - INIT_LIST_HEAD(dentries); - status = vfs_readdir(filp, nfsd4_build_dentrylist, &dla); + status = vfs_readdir(filp, nfsd4_build_namelist, &names); fput(filp); - while (!list_empty(dentries)) { - child = list_entry(dentries->next, struct dentry_list, list); - status = f(dir, child->dentry); + while (!list_empty(&names)) { + entry = list_entry(names.next, struct name_list, list); + + dentry = lookup_one_len(entry->name, dir, HEXDIR_LEN-1); + if (IS_ERR(dentry)) { + status = PTR_ERR(dentry); + goto out; + } + status = f(dir, dentry); + dput(dentry); if (status) goto out; - list_del(&child->list); - dput(child->dentry); - kfree(child); + list_del(&entry->list); + kfree(entry); } out: - while (!list_empty(dentries)) { - child = list_entry(dentries->next, struct dentry_list, list); - list_del(&child->list); - dput(child->dentry); - kfree(child); + while (!list_empty(&names)) { + entry = list_entry(names.next, struct name_list, list); + list_del(&entry->list); + kfree(entry); } nfs4_reset_creds(original_cred); return status; -- cgit v1.2.3 From 5cb031b0afddad73ea4191c9f0b76d20ca447dc0 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 14 Mar 2009 16:38:41 -0400 Subject: nfsd4: remove redundant check from nfsd4_open Note that we already checked for this invalid case at the top of this function. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 4 ---- 1 file changed, 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 36d74cda87a8..f156b85b4129 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -206,10 +206,6 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, switch (open->op_claim_type) { case NFS4_OPEN_CLAIM_DELEGATE_CUR: - status = nfserr_inval; - if (open->op_create) - goto out; - /* fall through */ case NFS4_OPEN_CLAIM_NULL: /* * (1) set CURRENT_FH to the file being opened, -- cgit v1.2.3 From 8bbfa9f3889b643fc7de82c0c761ef17097f8faf Mon Sep 17 00:00:00 2001 From: Greg Banks Date: Tue, 13 Jan 2009 21:26:34 +1100 Subject: knfsd: remove the nfsd thread busy histogram Stop gathering the data that feeds the 'th' line in /proc/net/rpc/nfsd because the questionable data provided is not worth the scalability impact of calculating it. Instead, always report zeroes. The current approach suffers from three major issues: 1. update_thread_usage() increments buckets by call service time or call arrival time...in jiffies. On lightly loaded machines, call service times are usually < 1 jiffy; on heavily loaded machines call arrival times will be << 1 jiffy. So a large portion of the updates to the buckets are rounded down to zero, and the histogram is undercounting. 2. As seen previously on the nfs mailing list, the format in which the histogram is presented is cryptic, difficult to explain, and difficult to use. 3. Updating the histogram requires taking a global spinlock and dirtying the global variables nfsd_last_call, nfsd_busy, and nfsdstats *twice* on every RPC call, which is a significant scaling limitation. Testing on a 4 CPU 4 NIC Altix using 4 IRIX clients each doing 1K streaming reads at full line rate, shows the stats update code (inlined into nfsd()) takes about 1.7% of each CPU. This patch drops the contribution from nfsd() into the profile noise. This patch is a forward-ported version of knfsd-remove-nfsd-threadstats which has been shipping in the SGI "Enhanced NFS" product since 2006. In that time, exactly one customer has noticed that the threadstats were missing. It has been previously posted: http://article.gmane.org/gmane.linux.nfs/10376 and more recently requested to be posted again. Signed-off-by: Greg Banks Signed-off-by: J. Bruce Fields --- fs/nfsd/nfssvc.c | 28 ---------------------------- 1 file changed, 28 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 07e4f5d7baa8..c3eb0759fd57 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -40,9 +40,6 @@ extern struct svc_program nfsd_program; static int nfsd(void *vrqstp); struct timeval nfssvc_boot; -static atomic_t nfsd_busy; -static unsigned long nfsd_last_call; -static DEFINE_SPINLOCK(nfsd_call_lock); /* * nfsd_mutex protects nfsd_serv -- both the pointer itself and the members @@ -227,7 +224,6 @@ int nfsd_create_serv(void) nfsd_max_blksize /= 2; } - atomic_set(&nfsd_busy, 0); nfsd_serv = svc_create_pooled(&nfsd_program, nfsd_max_blksize, AF_INET, nfsd_last_thread, nfsd, THIS_MODULE); @@ -376,26 +372,6 @@ nfsd_svc(unsigned short port, int nrservs) return error; } -static inline void -update_thread_usage(int busy_threads) -{ - unsigned long prev_call; - unsigned long diff; - int decile; - - spin_lock(&nfsd_call_lock); - prev_call = nfsd_last_call; - nfsd_last_call = jiffies; - decile = busy_threads*10/nfsdstats.th_cnt; - if (decile>0 && decile <= 10) { - diff = nfsd_last_call - prev_call; - if ( (nfsdstats.th_usage[decile-1] += diff) >= NFSD_USAGE_WRAP) - nfsdstats.th_usage[decile-1] -= NFSD_USAGE_WRAP; - if (decile == 10) - nfsdstats.th_fullcnt++; - } - spin_unlock(&nfsd_call_lock); -} /* * This is the NFS server kernel thread @@ -464,8 +440,6 @@ nfsd(void *vrqstp) continue; } - update_thread_usage(atomic_read(&nfsd_busy)); - atomic_inc(&nfsd_busy); /* Lock the export hash tables for reading. */ exp_readlock(); @@ -474,8 +448,6 @@ nfsd(void *vrqstp) /* Unlock export hash tables */ exp_readunlock(); - update_thread_usage(atomic_read(&nfsd_busy)); - atomic_dec(&nfsd_busy); } /* Clear signals before calling svc_exit_thread() */ -- cgit v1.2.3 From 03cf6c9f49a8fea953d38648d016e3f46e814991 Mon Sep 17 00:00:00 2001 From: Greg Banks Date: Tue, 13 Jan 2009 21:26:36 +1100 Subject: knfsd: add file to export stats about nfsd pools Add /proc/fs/nfsd/pool_stats to export to userspace various statistics about the operation of rpc server thread pools. This patch is based on a forward-ported version of knfsd-add-pool-thread-stats which has been shipping in the SGI "Enhanced NFS" product since 2006 and which was previously posted: http://article.gmane.org/gmane.linux.nfs/10375 It has also been updated thus: * moved EXPORT_SYMBOL() to near the function it exports * made the new struct struct seq_operations const * used SEQ_START_TOKEN instead of ((void *)1) * merged fix from SGI PV 990526 "sunrpc: use dprintk instead of printk in svc_pool_stats_*()" by Harshula Jayasuriya. * merged fix from SGI PV 964001 "Crash reading pool_stats before nfsds are started". Signed-off-by: Greg Banks Signed-off-by: Harshula Jayasuriya Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 12 ++++++++++++ fs/nfsd/nfssvc.c | 7 +++++++ 2 files changed, 19 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 3d93b2064ce5..4adebb6312c4 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -60,6 +60,7 @@ enum { NFSD_FO_UnlockFS, NFSD_Threads, NFSD_Pool_Threads, + NFSD_Pool_Stats, NFSD_Versions, NFSD_Ports, NFSD_MaxBlkSize, @@ -172,6 +173,16 @@ static const struct file_operations exports_operations = { .owner = THIS_MODULE, }; +extern int nfsd_pool_stats_open(struct inode *inode, struct file *file); + +static struct file_operations pool_stats_operations = { + .open = nfsd_pool_stats_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, + .owner = THIS_MODULE, +}; + /*----------------------------------------------------------------------------*/ /* * payload - write methods @@ -1246,6 +1257,7 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent) [NFSD_Fh] = {"filehandle", &transaction_ops, S_IWUSR|S_IRUSR}, [NFSD_Threads] = {"threads", &transaction_ops, S_IWUSR|S_IRUSR}, [NFSD_Pool_Threads] = {"pool_threads", &transaction_ops, S_IWUSR|S_IRUSR}, + [NFSD_Pool_Stats] = {"pool_stats", &pool_stats_operations, S_IRUGO}, [NFSD_Versions] = {"versions", &transaction_ops, S_IWUSR|S_IRUSR}, [NFSD_Ports] = {"portlist", &transaction_ops, S_IWUSR|S_IRUGO}, [NFSD_MaxBlkSize] = {"max_block_size", &transaction_ops, S_IWUSR|S_IRUGO}, diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index c3eb0759fd57..ef0a3686639d 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -546,3 +546,10 @@ nfsd_dispatch(struct svc_rqst *rqstp, __be32 *statp) nfsd_cache_update(rqstp, proc->pc_cachetype, statp + 1); return 1; } + +int nfsd_pool_stats_open(struct inode *inode, struct file *file) +{ + if (nfsd_serv == NULL) + return -ENODEV; + return svc_pool_stats_open(nfsd_serv, file); +} -- cgit v1.2.3 From 026722c25e6eb018eab8b9a3c198c258f5b7a2e7 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Wed, 18 Mar 2009 15:06:26 -0400 Subject: nfsd4: don't check ip address in setclientid The spec allows clients to change ip address, so we shouldn't be requiring that setclientid always come from the same address. For example, a client could reboot and get a new dhcpd address, but still present the same clientid to the server. In that case the server should revoke the client's previous state and allow it to continue, instead of (as it currently does) returning a CLID_INUSE error. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 54651aa45790..070e9e5c0452 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -791,10 +791,9 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (conf) { /* RFC 3530 14.2.33 CASE 0: */ status = nfserr_clid_inuse; - if (!same_creds(&conf->cl_cred, &rqstp->rq_cred) - || conf->cl_addr != sin->sin_addr.s_addr) { - dprintk("NFSD: setclientid: string in use by clientat %pI4\n", - &conf->cl_addr); + if (!same_creds(&conf->cl_cred, &rqstp->rq_cred)) { + dprintk("NFSD: setclientid: string in use by client" + " at %pI4\n", &conf->cl_addr); goto out; } } -- cgit v1.2.3 From 0953e620de0538cbd081f1b45126f6098112a598 Mon Sep 17 00:00:00 2001 From: "Sachin S. Prabhu" Date: Mon, 23 Feb 2009 16:22:03 +0000 Subject: Inconsistent setattr behaviour There is an inconsistency seen in the behaviour of nfs compared to other local filesystems on linux when changing owner or group of a directory. If the directory has SUID/SGID flags set, on changing owner or group on the directory, the flags are stripped off on nfs. These flags are maintained on other filesystems such as ext3. To reproduce on a nfs share or local filesystem, run the following commands mkdir test; chmod +s+g test; chown user1 test; ls -ld test On the nfs share, the flags are stripped and the output seen is drwxr-xr-x 2 user1 root 4096 Feb 23 2009 test On other local filesystems(ex: ext3), the flags are not stripped and the output seen is drwsr-sr-x 2 user1 root 4096 Feb 23 13:57 test chown_common() called from sys_chown() will only strip the flags if the inode is not a directory. static int chown_common(struct dentry * dentry, uid_t user, gid_t group) { .. if (!S_ISDIR(inode->i_mode)) newattrs.ia_valid |= ATTR_KILL_SUID | ATTR_KILL_SGID | ATTR_KILL_PRIV; .. } See: http://www.opengroup.org/onlinepubs/7990989775/xsh/chown.html "If the path argument refers to a regular file, the set-user-ID (S_ISUID) and set-group-ID (S_ISGID) bits of the file mode are cleared upon successful return from chown(), unless the call is made by a process with appropriate privileges, in which case it is implementation-dependent whether these bits are altered. If chown() is successfully invoked on a file that is not a regular file, these bits may be cleared. These bits are defined in ." The behaviour as it stands does not appear to violate POSIX. However the actions performed are inconsistent when comparing ext3 and nfs. Signed-off-by: Sachin Prabhu Acked-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 54404d730809..8790571b30fd 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -366,8 +366,9 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap, } /* Revoke setuid/setgid on chown */ - if (((iap->ia_valid & ATTR_UID) && iap->ia_uid != inode->i_uid) || - ((iap->ia_valid & ATTR_GID) && iap->ia_gid != inode->i_gid)) { + if (!S_ISDIR(inode->i_mode) && + (((iap->ia_valid & ATTR_UID) && iap->ia_uid != inode->i_uid) || + ((iap->ia_valid & ATTR_GID) && iap->ia_gid != inode->i_gid))) { iap->ia_valid |= ATTR_KILL_PRIV; if (iap->ia_valid & ATTR_MODE) { /* we're setting mode too, just clear the s*id bits */ -- cgit v1.2.3 From e354d571bb481f1d71f2c3004b9ff570b32e83bd Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Sat, 28 Mar 2009 11:30:52 +0300 Subject: nfsd: embed nfsd4_current_state in nfsd4_compoundres Remove the allocation of struct nfsd4_compound_state. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 40 +++++++++------------------------------- 1 file changed, 9 insertions(+), 31 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index f156b85b4129..7bb8cb321dfe 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -809,29 +809,6 @@ static inline void nfsd4_increment_op_stats(u32 opnum) nfsdstats.nfs4_opcount[opnum]++; } -static void cstate_free(struct nfsd4_compound_state *cstate) -{ - if (cstate == NULL) - return; - fh_put(&cstate->current_fh); - fh_put(&cstate->save_fh); - BUG_ON(cstate->replay_owner); - kfree(cstate); -} - -static struct nfsd4_compound_state *cstate_alloc(void) -{ - struct nfsd4_compound_state *cstate; - - cstate = kmalloc(sizeof(struct nfsd4_compound_state), GFP_KERNEL); - if (cstate == NULL) - return NULL; - fh_init(&cstate->current_fh, NFS4_FHSIZE); - fh_init(&cstate->save_fh, NFS4_FHSIZE); - cstate->replay_owner = NULL; - return cstate; -} - typedef __be32(*nfsd4op_func)(struct svc_rqst *, struct nfsd4_compound_state *, void *); @@ -859,12 +836,13 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, { struct nfsd4_op *op; struct nfsd4_operation *opdesc; - struct nfsd4_compound_state *cstate = NULL; + struct nfsd4_compound_state *cstate = &resp->cstate; int slack_bytes; __be32 status; resp->xbuf = &rqstp->rq_res; - resp->p = rqstp->rq_res.head[0].iov_base + rqstp->rq_res.head[0].iov_len; + resp->p = rqstp->rq_res.head[0].iov_base + + rqstp->rq_res.head[0].iov_len; resp->tagp = resp->p; /* reserve space for: taglen, tag, and opcnt */ resp->p += 2 + XDR_QUADLEN(args->taglen); @@ -873,6 +851,9 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, resp->tag = args->tag; resp->opcnt = 0; resp->rqstp = rqstp; + resp->cstate.replay_owner = NULL; + fh_init(&resp->cstate.current_fh, NFS4_FHSIZE); + fh_init(&resp->cstate.save_fh, NFS4_FHSIZE); /* * According to RFC3010, this takes precedence over all other errors. @@ -881,11 +862,6 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, if (args->minorversion > NFSD_SUPPORTED_MINOR_VERSION) goto out; - status = nfserr_resource; - cstate = cstate_alloc(); - if (cstate == NULL) - goto out; - status = nfs_ok; while (!status && resp->opcnt < args->opcnt) { op = &args->ops[resp->opcnt++]; @@ -958,7 +934,9 @@ encode_op: nfsd4_increment_op_stats(op->opnum); } - cstate_free(cstate); + fh_put(&resp->cstate.current_fh); + fh_put(&resp->cstate.save_fh); + BUG_ON(resp->cstate.replay_owner); out: nfsd4_release_compoundargs(args); dprintk("nfsv4 compound returned %d\n", ntohl(status)); -- cgit v1.2.3 From 20766016329eb4985c2c8b2a1b2333e0f865fdf9 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Sat, 28 Mar 2009 11:32:05 +0300 Subject: nfsd: remove nfsd4_ops array size There's no need for it. Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 7bb8cb321dfe..249dad987a16 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -943,7 +943,7 @@ out: return status; } -static struct nfsd4_operation nfsd4_ops[OP_RELEASE_LOCKOWNER+1] = { +static struct nfsd4_operation nfsd4_ops[] = { [OP_ACCESS] = { .op_func = (nfsd4op_func)nfsd4_access, .op_name = "OP_ACCESS", -- cgit v1.2.3 From 2f425878b6a71571341dcd3f9e9d1a6f6355da9c Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:27:32 +0300 Subject: nfsd: don't use the deferral service, return NFS4ERR_DELAY On an NFSv4.1 server cache miss that causes an upcall, NFS4ERR_DELAY will be returned. It is up to the NFSv4.1 client to resend only the operations that have not been processed. Initialize rq_usedeferral to 1 in svc_process(). It sill be turned off in nfsd4_proc_compound() only when NFSv4.1 Sessions are used. Note: this isn't an adequate solution on its own. It's acceptable as a way to get some minimal 4.1 up and working, but we're going to have to find a way to avoid returning DELAY in all common cases before 4.1 can really be considered ready. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfsd41: reverse rq_nodeferral negative logic] Signed-off-by: Benny Halevy [sunrpc: initialize rq_usedeferral] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 249dad987a16..ded469ff08b3 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -854,6 +854,8 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, resp->cstate.replay_owner = NULL; fh_init(&resp->cstate.current_fh, NFS4_FHSIZE); fh_init(&resp->cstate.save_fh, NFS4_FHSIZE); + /* Use the deferral mechanism only for NFSv4.0 compounds */ + rqstp->rq_usedeferral = (args->minorversion == 0); /* * According to RFC3010, this takes precedence over all other errors. @@ -933,12 +935,18 @@ encode_op: nfsd4_increment_op_stats(op->opnum); } + if (!rqstp->rq_usedeferral && status == nfserr_dropit) { + dprintk("%s Dropit - send NFS4ERR_DELAY\n", __func__); + status = nfserr_jukebox; + } fh_put(&resp->cstate.current_fh); fh_put(&resp->cstate.save_fh); BUG_ON(resp->cstate.replay_owner); out: nfsd4_release_compoundargs(args); + /* Reset deferral mechanism for RPC deferrals */ + rqstp->rq_usedeferral = 1; dprintk("nfsv4 compound returned %d\n", ntohl(status)); return status; } -- cgit v1.2.3 From 7116ed6b9973021ff43edeb10f4cb834db94000f Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:27:43 +0300 Subject: nfsd41: sessions basic data types This patch provides basic data structures representing the nfs41 sessions and slots, plus helpers for keeping a reference count on the session and freeing it. Note that our server only support a headerpadsz of 0 and it ignores backchannel attributes at the moment. Signed-off-by: Benny Halevy [nfsd41: remove headerpadsz from channel attributes] [nfsd41: embed nfsd4_channel in nfsd4_session] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfsd41: use bool inuse for slot state] Signed-off-by: Benny Halevy [nfsd41 remove sl_session from nfsd4_slot] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 070e9e5c0452..8c70f12159b6 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -382,6 +382,24 @@ static void release_openowner(struct nfs4_stateowner *sop) nfs4_put_stateowner(sop); } +static void +release_session(struct nfsd4_session *ses) +{ + list_del(&ses->se_hash); + list_del(&ses->se_perclnt); + nfsd4_put_session(ses); +} + +void +free_session(struct kref *kref) +{ + struct nfsd4_session *ses; + + ses = container_of(kref, struct nfsd4_session, se_ref); + kfree(ses->se_slots); + kfree(ses); +} + static inline void renew_client(struct nfs4_client *clp) { -- cgit v1.2.3 From 9fb870702d02c05f9410423bfff3f63e46e26180 Mon Sep 17 00:00:00 2001 From: Marc Eshel Date: Fri, 3 Apr 2009 08:27:46 +0300 Subject: nfsd41: introduce nfs4_client cl_sessions list [get rid of CONFIG_NFSD_V4_1] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 8c70f12159b6..b2611999600a 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -528,6 +528,7 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir) INIT_LIST_HEAD(&clp->cl_strhash); INIT_LIST_HEAD(&clp->cl_openowners); INIT_LIST_HEAD(&clp->cl_delegations); + INIT_LIST_HEAD(&clp->cl_sessions); INIT_LIST_HEAD(&clp->cl_lru); return clp; } -- cgit v1.2.3 From c4bf7868064ce8b9c75d8049d077e593c20602b3 Mon Sep 17 00:00:00 2001 From: Marc Eshel Date: Fri, 3 Apr 2009 08:27:49 +0300 Subject: nfsd41: release_session when client is expired Signed-off-by: Benny Halevy [add CONFIG_NFSD_V4_1 to fix v4.0 regression bug] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index b2611999600a..f23385b13065 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -511,6 +511,12 @@ expire_client(struct nfs4_client *clp) sop = list_entry(clp->cl_openowners.next, struct nfs4_stateowner, so_perclient); release_openowner(sop); } + while (!list_empty(&clp->cl_sessions)) { + struct nfsd4_session *ses; + ses = list_entry(clp->cl_sessions.next, struct nfsd4_session, + se_perclnt); + release_session(ses); + } put_nfs4_client(clp); } -- cgit v1.2.3 From 5282fd724b667b7d65f2e41e405a825e58a78813 Mon Sep 17 00:00:00 2001 From: Marc Eshel Date: Fri, 3 Apr 2009 08:27:52 +0300 Subject: nfsd41: sessionid hashing Simple sessionid hashing using its monotonically increasing sequence number. Locking considerations: sessionid_hashtbl access is controlled by the sessionid_lock spin lock. It must be taken for insert, delete, and lookup. nfsd4_sequence looks up the session id and if the session is found, it calls nfsd4_get_session (still under the sessionid_lock). nfsd4_destroy_session calls nfsd4_put_session after unhashing it, so when the session's kref reaches zero it's going to get freed. Signed-off-by: Benny Halevy [we don't use a prime for sessionid hash table size] [use sessionid_lock spin lock] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 55 ++++++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 54 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index f23385b13065..fc20e1f38d75 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -382,11 +382,62 @@ static void release_openowner(struct nfs4_stateowner *sop) nfs4_put_stateowner(sop); } +static DEFINE_SPINLOCK(sessionid_lock); +#define SESSION_HASH_SIZE 512 +static struct list_head sessionid_hashtbl[SESSION_HASH_SIZE]; + +static inline int +hash_sessionid(struct nfs4_sessionid *sessionid) +{ + struct nfsd4_sessionid *sid = (struct nfsd4_sessionid *)sessionid; + + return sid->sequence % SESSION_HASH_SIZE; +} + +static inline void +dump_sessionid(const char *fn, struct nfs4_sessionid *sessionid) +{ + u32 *ptr = (u32 *)(&sessionid->data[0]); + dprintk("%s: %u:%u:%u:%u\n", fn, ptr[0], ptr[1], ptr[2], ptr[3]); +} + +/* caller must hold sessionid_lock */ +static struct nfsd4_session * +find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid) +{ + struct nfsd4_session *elem; + int idx; + + dump_sessionid(__func__, sessionid); + idx = hash_sessionid(sessionid); + dprintk("%s: idx is %d\n", __func__, idx); + /* Search in the appropriate list */ + list_for_each_entry(elem, &sessionid_hashtbl[idx], se_hash) { + dump_sessionid("list traversal", &elem->se_sessionid); + if (!memcmp(elem->se_sessionid.data, sessionid->data, + NFS4_MAX_SESSIONID_LEN)) { + return elem; + } + } + + dprintk("%s: session not found\n", __func__); + return NULL; +} + +/* caller must hold sessionid_lock */ static void -release_session(struct nfsd4_session *ses) +unhash_session(struct nfsd4_session *ses) { list_del(&ses->se_hash); list_del(&ses->se_perclnt); +} + +static void +release_session(struct nfsd4_session *ses) +{ + spin_lock(&sessionid_lock); + unhash_session(ses); + spin_unlock(&sessionid_lock); nfsd4_put_session(ses); } @@ -3205,6 +3256,8 @@ nfs4_state_init(void) INIT_LIST_HEAD(&unconf_str_hashtbl[i]); INIT_LIST_HEAD(&unconf_id_hashtbl[i]); } + for (i = 0; i < SESSION_HASH_SIZE; i++) + INIT_LIST_HEAD(&sessionid_hashtbl[i]); for (i = 0; i < FILE_HASH_SIZE; i++) { INIT_LIST_HEAD(&file_hashtbl[i]); } -- cgit v1.2.3 From 2db134eb3b39faefc7fbfb200156d175edba2f68 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:27:55 +0300 Subject: nfsd41: xdr infrastructure Define nfsd41_dec_ops vector and add it to nfsd4_minorversion for minorversion 1. Note: nfsd4_enc_ops vector is shared for v4.0 and v4.1 since we don't need to filter out obsolete ops as this is done in the decoding phase. exchange_id, create_session, destroy_session, and sequence ops are implemented as stubs returning nfserr_opnotsupp at this stage. [was nfsd41: xdr stubs] [get rid of CONFIG_NFSD_V4_1] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 148 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 148 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 76a0b2a8d69b..6973d61bedea 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -996,6 +996,34 @@ nfsd4_decode_release_lockowner(struct nfsd4_compoundargs *argp, struct nfsd4_rel DECODE_TAIL; } +static __be32 +nfsd4_decode_exchange_id(struct nfsd4_compoundargs *argp, + struct nfsd4_exchange_id *clid) +{ + return nfserr_opnotsupp; /* stub */ +} + +static __be32 +nfsd4_decode_create_session(struct nfsd4_compoundargs *argp, + struct nfsd4_create_session *sess) +{ + return nfserr_opnotsupp; /* stub */ +} + +static __be32 +nfsd4_decode_destroy_session(struct nfsd4_compoundargs *argp, + struct nfsd4_destroy_session *destroy_session) +{ + return nfserr_opnotsupp; /* stub */ +} + +static __be32 +nfsd4_decode_sequence(struct nfsd4_compoundargs *argp, + struct nfsd4_sequence *seq) +{ + return nfserr_opnotsupp; /* stub */ +} + static __be32 nfsd4_decode_noop(struct nfsd4_compoundargs *argp, void *p) { @@ -1050,6 +1078,67 @@ static nfsd4_dec nfsd4_dec_ops[] = { [OP_RELEASE_LOCKOWNER] = (nfsd4_dec)nfsd4_decode_release_lockowner, }; +static nfsd4_dec nfsd41_dec_ops[] = { + [OP_ACCESS] (nfsd4_dec)nfsd4_decode_access, + [OP_CLOSE] (nfsd4_dec)nfsd4_decode_close, + [OP_COMMIT] (nfsd4_dec)nfsd4_decode_commit, + [OP_CREATE] (nfsd4_dec)nfsd4_decode_create, + [OP_DELEGPURGE] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_DELEGRETURN] (nfsd4_dec)nfsd4_decode_delegreturn, + [OP_GETATTR] (nfsd4_dec)nfsd4_decode_getattr, + [OP_GETFH] (nfsd4_dec)nfsd4_decode_noop, + [OP_LINK] (nfsd4_dec)nfsd4_decode_link, + [OP_LOCK] (nfsd4_dec)nfsd4_decode_lock, + [OP_LOCKT] (nfsd4_dec)nfsd4_decode_lockt, + [OP_LOCKU] (nfsd4_dec)nfsd4_decode_locku, + [OP_LOOKUP] (nfsd4_dec)nfsd4_decode_lookup, + [OP_LOOKUPP] (nfsd4_dec)nfsd4_decode_noop, + [OP_NVERIFY] (nfsd4_dec)nfsd4_decode_verify, + [OP_OPEN] (nfsd4_dec)nfsd4_decode_open, + [OP_OPENATTR] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_OPEN_CONFIRM] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_OPEN_DOWNGRADE] (nfsd4_dec)nfsd4_decode_open_downgrade, + [OP_PUTFH] (nfsd4_dec)nfsd4_decode_putfh, + [OP_PUTPUBFH] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_PUTROOTFH] (nfsd4_dec)nfsd4_decode_noop, + [OP_READ] (nfsd4_dec)nfsd4_decode_read, + [OP_READDIR] (nfsd4_dec)nfsd4_decode_readdir, + [OP_READLINK] (nfsd4_dec)nfsd4_decode_noop, + [OP_REMOVE] (nfsd4_dec)nfsd4_decode_remove, + [OP_RENAME] (nfsd4_dec)nfsd4_decode_rename, + [OP_RENEW] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_RESTOREFH] (nfsd4_dec)nfsd4_decode_noop, + [OP_SAVEFH] (nfsd4_dec)nfsd4_decode_noop, + [OP_SECINFO] (nfsd4_dec)nfsd4_decode_secinfo, + [OP_SETATTR] (nfsd4_dec)nfsd4_decode_setattr, + [OP_SETCLIENTID] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_SETCLIENTID_CONFIRM](nfsd4_dec)nfsd4_decode_notsupp, + [OP_VERIFY] (nfsd4_dec)nfsd4_decode_verify, + [OP_WRITE] (nfsd4_dec)nfsd4_decode_write, + [OP_RELEASE_LOCKOWNER] (nfsd4_dec)nfsd4_decode_notsupp, + + /* new operations for NFSv4.1 */ + [OP_BACKCHANNEL_CTL] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_BIND_CONN_TO_SESSION](nfsd4_dec)nfsd4_decode_notsupp, + [OP_EXCHANGE_ID] (nfsd4_dec)nfsd4_decode_exchange_id, + [OP_CREATE_SESSION] (nfsd4_dec)nfsd4_decode_create_session, + [OP_DESTROY_SESSION] (nfsd4_dec)nfsd4_decode_destroy_session, + [OP_FREE_STATEID] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_GET_DIR_DELEGATION] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_GETDEVICEINFO] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_GETDEVICELIST] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_LAYOUTCOMMIT] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_LAYOUTGET] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_LAYOUTRETURN] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_SECINFO_NO_NAME] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_SEQUENCE] (nfsd4_dec)nfsd4_decode_sequence, + [OP_SET_SSV] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_TEST_STATEID] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_WANT_DELEGATION] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_DESTROY_CLIENTID] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_RECLAIM_COMPLETE] (nfsd4_dec)nfsd4_decode_notsupp, +}; + struct nfsd4_minorversion_ops { nfsd4_dec *decoders; int nops; @@ -1057,6 +1146,7 @@ struct nfsd4_minorversion_ops { static struct nfsd4_minorversion_ops nfsd4_minorversion[] = { [0] = { nfsd4_dec_ops, ARRAY_SIZE(nfsd4_dec_ops) }, + [1] = { nfsd41_dec_ops, ARRAY_SIZE(nfsd41_dec_ops) }, }; static __be32 @@ -2571,6 +2661,38 @@ nfsd4_encode_write(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_w return nfserr; } +static __be32 +nfsd4_encode_exchange_id(struct nfsd4_compoundres *resp, int nfserr, + struct nfsd4_exchange_id *exid) +{ + /* stub */ + return nfserr; +} + +static __be32 +nfsd4_encode_create_session(struct nfsd4_compoundres *resp, int nfserr, + struct nfsd4_create_session *sess) +{ + /* stub */ + return nfserr; +} + +static __be32 +nfsd4_encode_destroy_session(struct nfsd4_compoundres *resp, int nfserr, + struct nfsd4_destroy_session *destroy_session) +{ + /* stub */ + return nfserr; +} + +static __be32 +nfsd4_encode_sequence(struct nfsd4_compoundres *resp, int nfserr, + struct nfsd4_sequence *seq) +{ + /* stub */ + return nfserr; +} + static __be32 nfsd4_encode_noop(struct nfsd4_compoundres *resp, __be32 nfserr, void *p) { @@ -2579,6 +2701,11 @@ nfsd4_encode_noop(struct nfsd4_compoundres *resp, __be32 nfserr, void *p) typedef __be32(* nfsd4_enc)(struct nfsd4_compoundres *, __be32, void *); +/* + * Note: nfsd4_enc_ops vector is shared for v4.0 and v4.1 + * since we don't need to filter out obsolete ops as this is + * done in the decoding phase. + */ static nfsd4_enc nfsd4_enc_ops[] = { [OP_ACCESS] = (nfsd4_enc)nfsd4_encode_access, [OP_CLOSE] = (nfsd4_enc)nfsd4_encode_close, @@ -2617,6 +2744,27 @@ static nfsd4_enc nfsd4_enc_ops[] = { [OP_VERIFY] = (nfsd4_enc)nfsd4_encode_noop, [OP_WRITE] = (nfsd4_enc)nfsd4_encode_write, [OP_RELEASE_LOCKOWNER] = (nfsd4_enc)nfsd4_encode_noop, + + /* NFSv4.1 operations */ + [OP_BACKCHANNEL_CTL] = (nfsd4_enc)nfsd4_encode_noop, + [OP_BIND_CONN_TO_SESSION] = (nfsd4_enc)nfsd4_encode_noop, + [OP_EXCHANGE_ID] = (nfsd4_enc)nfsd4_encode_exchange_id, + [OP_CREATE_SESSION] = (nfsd4_enc)nfsd4_encode_create_session, + [OP_DESTROY_SESSION] = (nfsd4_enc)nfsd4_encode_destroy_session, + [OP_FREE_STATEID] = (nfsd4_enc)nfsd4_encode_noop, + [OP_GET_DIR_DELEGATION] = (nfsd4_enc)nfsd4_encode_noop, + [OP_GETDEVICEINFO] = (nfsd4_enc)nfsd4_encode_noop, + [OP_GETDEVICELIST] = (nfsd4_enc)nfsd4_encode_noop, + [OP_LAYOUTCOMMIT] = (nfsd4_enc)nfsd4_encode_noop, + [OP_LAYOUTGET] = (nfsd4_enc)nfsd4_encode_noop, + [OP_LAYOUTRETURN] = (nfsd4_enc)nfsd4_encode_noop, + [OP_SECINFO_NO_NAME] = (nfsd4_enc)nfsd4_encode_noop, + [OP_SEQUENCE] = (nfsd4_enc)nfsd4_encode_sequence, + [OP_SET_SSV] = (nfsd4_enc)nfsd4_encode_noop, + [OP_TEST_STATEID] = (nfsd4_enc)nfsd4_encode_noop, + [OP_WANT_DELEGATION] = (nfsd4_enc)nfsd4_encode_noop, + [OP_DESTROY_CLIENTID] = (nfsd4_enc)nfsd4_encode_noop, + [OP_RECLAIM_COMPLETE] = (nfsd4_enc)nfsd4_encode_noop, }; void -- cgit v1.2.3 From 069b6ad4bb20abf175ea7875e82e8002154773af Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:27:58 +0300 Subject: nfsd41: proc stubs Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 22 ++++++++++++++++++++++ fs/nfsd/nfs4state.c | 32 ++++++++++++++++++++++++++++++++ 2 files changed, 54 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index ded469ff08b3..f7be7fabe62c 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1101,6 +1101,28 @@ static struct nfsd4_operation nfsd4_ops[] = { .op_flags = ALLOWED_WITHOUT_FH | ALLOWED_ON_ABSENT_FS, .op_name = "OP_RELEASE_LOCKOWNER", }, + + /* NFSv4.1 operations */ + [OP_EXCHANGE_ID] = { + .op_func = (nfsd4op_func)nfsd4_exchange_id, + .op_flags = ALLOWED_WITHOUT_FH, + .op_name = "OP_EXCHANGE_ID", + }, + [OP_CREATE_SESSION] = { + .op_func = (nfsd4op_func)nfsd4_create_session, + .op_flags = ALLOWED_WITHOUT_FH, + .op_name = "OP_CREATE_SESSION", + }, + [OP_DESTROY_SESSION] = { + .op_func = (nfsd4op_func)nfsd4_destroy_session, + .op_flags = ALLOWED_WITHOUT_FH, + .op_name = "OP_DESTROY_SESSION", + }, + [OP_SEQUENCE] = { + .op_func = (nfsd4op_func)nfsd4_sequence, + .op_flags = ALLOWED_WITHOUT_FH, + .op_name = "OP_SEQUENCE", + }, }; static const char *nfsd4_op_name(unsigned opnum) diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index fc20e1f38d75..cfc01f415d15 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -832,6 +832,38 @@ out_err: return; } +__be32 +nfsd4_exchange_id(struct svc_rqst *rqstp, + struct nfsd4_compound_state *cstate, + struct nfsd4_exchange_id *exid) +{ + return -1; /* stub */ +} + +__be32 +nfsd4_create_session(struct svc_rqst *rqstp, + struct nfsd4_compound_state *cstate, + struct nfsd4_create_session *cr_ses) +{ + return -1; /* stub */ +} + +__be32 +nfsd4_destroy_session(struct svc_rqst *r, + struct nfsd4_compound_state *cstate, + struct nfsd4_destroy_session *sessionid) +{ + return -1; /* stub */ +} + +__be32 +nfsd4_sequence(struct svc_rqst *r, + struct nfsd4_compound_state *cstate, + struct nfsd4_sequence *seq) +{ + return -1; /* stub */ +} + __be32 nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_setclientid *setclid) -- cgit v1.2.3 From 0733d21338747483985a5964e852af160d88e429 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:01 +0300 Subject: nfsd41: exchange_id operation Implement the exchange_id operation confoming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-28 Based on the client provided name, hash a client id. If a confirmed one is found, compare the op's creds and verifier. If the creds match and the verifier is different then expire the old client (client re-incarnated), otherwise, if both match, assume it's a replay and ignore it. If an unconfirmed client is found, then copy the new creds and verifer if need update, otherwise assume replay. The client is moved to a confirmed state on create_session. In the nfs41 branch set the exchange_id flags to EXCHGID4_FLAG_USE_NON_PNFS | EXCHGID4_FLAG_SUPP_MOVED_REFER (pNFS is not supported, Referrals are supported, Migration is not.). Address various scenarios from section 18.35 of the spec: 1. Check for EXCHGID4_FLAG_UPD_CONFIRMED_REC_A and set EXCHGID4_FLAG_CONFIRMED_R as appropriate. 2. Return error codes per 18.35.4 scenarios. 3. Update client records or generate new client ids depending on scenario. Note: 18.35.4 case 3 probably still needs revisiting. The handling seems not quite right. Signed-off-by: Benny Halevy Signed-off-by: Andy Adamosn Signed-off-by: Benny Halevy [nfsd41: use utsname for major_id (and copy to server_scope)] [nfsd41: fix handling of various exchange id scenarios] Signed-off-by: Mike Sager Signed-off-by: Benny Halevy [nfsd41: reverse use of EXCHGID4_INVAL_FLAG_MASK_A] [simplify nfsd4_encode_exchange_id error handling] [nfsd41: embed an xdr_netobj in nfsd4_exchange_id] [nfsd41: return nfserr_serverfault for spa_how == SP4_MACH_CRED] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 142 +++++++++++++++++++++++++++++++++++++++++++++++++- fs/nfsd/nfs4xdr.c | 147 ++++++++++++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 284 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index cfc01f415d15..4f963e902972 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -832,12 +832,152 @@ out_err: return; } +/* + * Set the exchange_id flags returned by the server. + */ +static void +nfsd4_set_ex_flags(struct nfs4_client *new, struct nfsd4_exchange_id *clid) +{ + /* pNFS is not supported */ + new->cl_exchange_flags |= EXCHGID4_FLAG_USE_NON_PNFS; + + /* Referrals are supported, Migration is not. */ + new->cl_exchange_flags |= EXCHGID4_FLAG_SUPP_MOVED_REFER; + + /* set the wire flags to return to client. */ + clid->flags = new->cl_exchange_flags; +} + __be32 nfsd4_exchange_id(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_exchange_id *exid) { - return -1; /* stub */ + struct nfs4_client *unconf, *conf, *new; + int status; + unsigned int strhashval; + char dname[HEXDIR_LEN]; + nfs4_verifier verf = exid->verifier; + u32 ip_addr = svc_addr_in(rqstp)->sin_addr.s_addr; + + dprintk("%s rqstp=%p exid=%p clname.len=%u clname.data=%p " + " ip_addr=%u flags %x, spa_how %d\n", + __func__, rqstp, exid, exid->clname.len, exid->clname.data, + ip_addr, exid->flags, exid->spa_how); + + if (!check_name(exid->clname) || (exid->flags & ~EXCHGID4_FLAG_MASK_A)) + return nfserr_inval; + + /* Currently only support SP4_NONE */ + switch (exid->spa_how) { + case SP4_NONE: + break; + case SP4_SSV: + return nfserr_encr_alg_unsupp; + default: + BUG(); /* checked by xdr code */ + case SP4_MACH_CRED: + return nfserr_serverfault; /* no excuse :-/ */ + } + + status = nfs4_make_rec_clidname(dname, &exid->clname); + + if (status) + goto error; + + strhashval = clientstr_hashval(dname); + + nfs4_lock_state(); + status = nfs_ok; + + conf = find_confirmed_client_by_str(dname, strhashval); + if (conf) { + if (!same_verf(&verf, &conf->cl_verifier)) { + /* 18.35.4 case 8 */ + if (exid->flags & EXCHGID4_FLAG_UPD_CONFIRMED_REC_A) { + status = nfserr_not_same; + goto out; + } + /* Client reboot: destroy old state */ + expire_client(conf); + goto out_new; + } + if (!same_creds(&conf->cl_cred, &rqstp->rq_cred)) { + /* 18.35.4 case 9 */ + if (exid->flags & EXCHGID4_FLAG_UPD_CONFIRMED_REC_A) { + status = nfserr_perm; + goto out; + } + expire_client(conf); + goto out_new; + } + if (ip_addr != conf->cl_addr && + !(exid->flags & EXCHGID4_FLAG_UPD_CONFIRMED_REC_A)) { + /* Client collision. 18.35.4 case 3 */ + status = nfserr_clid_inuse; + goto out; + } + /* + * Set bit when the owner id and verifier map to an already + * confirmed client id (18.35.3). + */ + exid->flags |= EXCHGID4_FLAG_CONFIRMED_R; + + /* + * Falling into 18.35.4 case 2, possible router replay. + * Leave confirmed record intact and return same result. + */ + copy_verf(conf, &verf); + new = conf; + goto out_copy; + } else { + /* 18.35.4 case 7 */ + if (exid->flags & EXCHGID4_FLAG_UPD_CONFIRMED_REC_A) { + status = nfserr_noent; + goto out; + } + } + + unconf = find_unconfirmed_client_by_str(dname, strhashval); + if (unconf) { + /* + * Possible retry or client restart. Per 18.35.4 case 4, + * a new unconfirmed record should be generated regardless + * of whether any properties have changed. + */ + expire_client(unconf); + } + +out_new: + /* Normal case */ + new = create_client(exid->clname, dname); + if (new == NULL) { + status = nfserr_resource; + goto out; + } + + copy_verf(new, &verf); + copy_cred(&new->cl_cred, &rqstp->rq_cred); + new->cl_addr = ip_addr; + gen_clid(new); + gen_confirm(new); + add_to_unconfirmed(new, strhashval); +out_copy: + exid->clientid.cl_boot = new->cl_clientid.cl_boot; + exid->clientid.cl_id = new->cl_clientid.cl_id; + + new->cl_seqid = exid->seqid = 1; + nfsd4_set_ex_flags(new, exid); + + dprintk("nfsd4_exchange_id seqid %d flags %x\n", + new->cl_seqid, new->cl_exchange_flags); + status = nfs_ok; + +out: + nfs4_unlock_state(); +error: + dprintk("nfsd4_exchange_id returns %d\n", ntohl(status)); + return status; } __be32 diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 6973d61bedea..bebf6d249069 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -45,6 +45,7 @@ #include #include #include +#include #include #include #include @@ -998,9 +999,100 @@ nfsd4_decode_release_lockowner(struct nfsd4_compoundargs *argp, struct nfsd4_rel static __be32 nfsd4_decode_exchange_id(struct nfsd4_compoundargs *argp, - struct nfsd4_exchange_id *clid) + struct nfsd4_exchange_id *exid) { - return nfserr_opnotsupp; /* stub */ + int dummy; + DECODE_HEAD; + + READ_BUF(NFS4_VERIFIER_SIZE); + COPYMEM(exid->verifier.data, NFS4_VERIFIER_SIZE); + + READ_BUF(4); + READ32(exid->clname.len); + + READ_BUF(exid->clname.len); + SAVEMEM(exid->clname.data, exid->clname.len); + + READ_BUF(4); + READ32(exid->flags); + + /* Ignore state_protect4_a */ + READ_BUF(4); + READ32(exid->spa_how); + switch (exid->spa_how) { + case SP4_NONE: + break; + case SP4_MACH_CRED: + /* spo_must_enforce */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy * 4); + p += dummy; + + /* spo_must_allow */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy * 4); + p += dummy; + break; + case SP4_SSV: + /* ssp_ops */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy * 4); + p += dummy; + + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy * 4); + p += dummy; + + /* ssp_hash_algs<> */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy); + p += XDR_QUADLEN(dummy); + + /* ssp_encr_algs<> */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy); + p += XDR_QUADLEN(dummy); + + /* ssp_window and ssp_num_gss_handles */ + READ_BUF(8); + READ32(dummy); + READ32(dummy); + break; + default: + goto xdr_error; + } + + /* Ignore Implementation ID */ + READ_BUF(4); /* nfs_impl_id4 array length */ + READ32(dummy); + + if (dummy > 1) + goto xdr_error; + + if (dummy == 1) { + /* nii_domain */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy); + p += XDR_QUADLEN(dummy); + + /* nii_name */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy); + p += XDR_QUADLEN(dummy); + + /* nii_date */ + READ_BUF(12); + p += 3; + } + DECODE_TAIL; } static __be32 @@ -2665,8 +2757,55 @@ static __be32 nfsd4_encode_exchange_id(struct nfsd4_compoundres *resp, int nfserr, struct nfsd4_exchange_id *exid) { - /* stub */ - return nfserr; + ENCODE_HEAD; + char *major_id; + char *server_scope; + int major_id_sz; + int server_scope_sz; + uint64_t minor_id = 0; + + if (nfserr) + return nfserr; + + major_id = utsname()->nodename; + major_id_sz = strlen(major_id); + server_scope = utsname()->nodename; + server_scope_sz = strlen(server_scope); + + RESERVE_SPACE( + 8 /* eir_clientid */ + + 4 /* eir_sequenceid */ + + 4 /* eir_flags */ + + 4 /* spr_how (SP4_NONE) */ + + 8 /* so_minor_id */ + + 4 /* so_major_id.len */ + + (XDR_QUADLEN(major_id_sz) * 4) + + 4 /* eir_server_scope.len */ + + (XDR_QUADLEN(server_scope_sz) * 4) + + 4 /* eir_server_impl_id.count (0) */); + + WRITEMEM(&exid->clientid, 8); + WRITE32(exid->seqid); + WRITE32(exid->flags); + + /* state_protect4_r. Currently only support SP4_NONE */ + BUG_ON(exid->spa_how != SP4_NONE); + WRITE32(exid->spa_how); + + /* The server_owner struct */ + WRITE64(minor_id); /* Minor id */ + /* major id */ + WRITE32(major_id_sz); + WRITEMEM(major_id, major_id_sz); + + /* Server scope */ + WRITE32(server_scope_sz); + WRITEMEM(server_scope, server_scope_sz); + + /* Implementation id */ + WRITE32(0); /* zero length nfs_impl_id4 array */ + ADJUST_ARGS(); + return 0; } static __be32 -- cgit v1.2.3 From a1bcecd29cdf1670df6908a620add4211c0abb7a Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:05 +0300 Subject: nfsd41: match clientid establishment method We need to distinguish between client names provided by NFSv4.0 clients SETCLIENTID and those provided by NFSv4.1 via EXCHANGE_ID when looking up the clientid by string. Signed-off-by: Benny Halevy Signed-off-by: Andy Adamson [nfsd41: use boolean values for use_exchange_id argument] Signed-off-by: Benny Halevy [nfsd41: simplify match_clientid_establishment logic] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4recover.c | 3 ++- fs/nfsd/nfs4state.c | 42 +++++++++++++++++++++++++++++++----------- 2 files changed, 33 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index b11cf8d34280..3444c0052a87 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -344,7 +344,8 @@ purge_old(struct dentry *parent, struct dentry *child) { int status; - if (nfs4_has_reclaimed_state(child->d_name.name)) + /* note: we currently use this path only for minorversion 0 */ + if (nfs4_has_reclaimed_state(child->d_name.name, false)) return 0; status = nfsd4_clear_clid_dir(parent, child); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 4f963e902972..dfc6d946cdfe 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -715,25 +715,45 @@ find_unconfirmed_client(clientid_t *clid) return NULL; } +/* + * Return 1 iff clp's clientid establishment method matches the use_exchange_id + * parameter. Matching is based on the fact the at least one of the + * EXCHGID4_FLAG_USE_{NON_PNFS,PNFS_MDS,PNFS_DS} flags must be set for v4.1 + * + * FIXME: we need to unify the clientid namespaces for nfsv4.x + * and correctly deal with client upgrade/downgrade in EXCHANGE_ID + * and SET_CLIENTID{,_CONFIRM} + */ +static inline int +match_clientid_establishment(struct nfs4_client *clp, bool use_exchange_id) +{ + bool has_exchange_flags = (clp->cl_exchange_flags != 0); + return use_exchange_id == has_exchange_flags; +} + static struct nfs4_client * -find_confirmed_client_by_str(const char *dname, unsigned int hashval) +find_confirmed_client_by_str(const char *dname, unsigned int hashval, + bool use_exchange_id) { struct nfs4_client *clp; list_for_each_entry(clp, &conf_str_hashtbl[hashval], cl_strhash) { - if (same_name(clp->cl_recdir, dname)) + if (same_name(clp->cl_recdir, dname) && + match_clientid_establishment(clp, use_exchange_id)) return clp; } return NULL; } static struct nfs4_client * -find_unconfirmed_client_by_str(const char *dname, unsigned int hashval) +find_unconfirmed_client_by_str(const char *dname, unsigned int hashval, + bool use_exchange_id) { struct nfs4_client *clp; list_for_each_entry(clp, &unconf_str_hashtbl[hashval], cl_strhash) { - if (same_name(clp->cl_recdir, dname)) + if (same_name(clp->cl_recdir, dname) && + match_clientid_establishment(clp, use_exchange_id)) return clp; } return NULL; @@ -890,7 +910,7 @@ nfsd4_exchange_id(struct svc_rqst *rqstp, nfs4_lock_state(); status = nfs_ok; - conf = find_confirmed_client_by_str(dname, strhashval); + conf = find_confirmed_client_by_str(dname, strhashval, true); if (conf) { if (!same_verf(&verf, &conf->cl_verifier)) { /* 18.35.4 case 8 */ @@ -938,7 +958,7 @@ nfsd4_exchange_id(struct svc_rqst *rqstp, } } - unconf = find_unconfirmed_client_by_str(dname, strhashval); + unconf = find_unconfirmed_client_by_str(dname, strhashval, true); if (unconf) { /* * Possible retry or client restart. Per 18.35.4 case 4, @@ -1035,7 +1055,7 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, strhashval = clientstr_hashval(dname); nfs4_lock_state(); - conf = find_confirmed_client_by_str(dname, strhashval); + conf = find_confirmed_client_by_str(dname, strhashval, false); if (conf) { /* RFC 3530 14.2.33 CASE 0: */ status = nfserr_clid_inuse; @@ -1050,7 +1070,7 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, * has a description of SETCLIENTID request processing consisting * of 5 bullet points, labeled as CASE0 - CASE4 below. */ - unconf = find_unconfirmed_client_by_str(dname, strhashval); + unconf = find_unconfirmed_client_by_str(dname, strhashval, false); status = nfserr_resource; if (!conf) { /* @@ -1205,7 +1225,7 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp, unsigned int hash = clientstr_hashval(unconf->cl_recdir); conf = find_confirmed_client_by_str(unconf->cl_recdir, - hash); + hash, false); if (conf) { nfsd4_remove_clid_dir(conf); expire_client(conf); @@ -3326,12 +3346,12 @@ alloc_reclaim(void) } int -nfs4_has_reclaimed_state(const char *name) +nfs4_has_reclaimed_state(const char *name, bool use_exchange_id) { unsigned int strhashval = clientstr_hashval(name); struct nfs4_client *clp; - clp = find_confirmed_client_by_str(name, strhashval); + clp = find_confirmed_client_by_str(name, strhashval, use_exchange_id); return clp ? 1 : 0; } -- cgit v1.2.3 From b85d4c01b76f6969a085d07a767fa45225cb14be Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 3 Apr 2009 08:28:08 +0300 Subject: nfsd41: sequence operation Implement the sequence operation conforming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 Check for stale clientid (as derived from the sessionid). Enforce slotid range and exactly-once semantics using the slotid and seqid. If everything went well renew the client lease and mark the slot INPROGRESS. Add a struct nfsd4_slot pointer to struct nfsd4_compound_state. To be used for sessions DRC replay. [nfsd41: rename sequence catchthis to cachethis] Signed-off-by: Andy Adamson [pulled some code to set cstate->slot from "nfsd DRC logic"] [use sessionid_lock spin lock] [nfsd41: use bool inuse for slot state] Signed-off-by: Benny Halevy [nfsd: add a struct nfsd4_slot pointer to struct nfsd4_compound_state] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfsd41: add nfsd4_session pointer to nfsd4_compound_state] [nfsd41: set cstate session] [nfsd41: use cstate session in nfsd4_sequence] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [simplify nfsd4_encode_sequence error handling] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 73 +++++++++++++++++++++++++++++++++++++++++++++++++++-- fs/nfsd/nfs4xdr.c | 33 +++++++++++++++++++++--- 2 files changed, 101 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index dfc6d946cdfe..24b6e0593184 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1000,6 +1000,32 @@ error: return status; } +static int +check_slot_seqid(u32 seqid, struct nfsd4_slot *slot) +{ + dprintk("%s enter. seqid %d slot->sl_seqid %d\n", __func__, seqid, + slot->sl_seqid); + + /* The slot is in use, and no response has been sent. */ + if (slot->sl_inuse) { + if (seqid == slot->sl_seqid) + return nfserr_jukebox; + else + return nfserr_seq_misordered; + } + /* Normal */ + if (likely(seqid == slot->sl_seqid + 1)) + return nfs_ok; + /* Replay */ + if (seqid == slot->sl_seqid) + return nfserr_replay_cache; + /* Wraparound */ + if (seqid == 1 && (slot->sl_seqid + 1) == 0) + return nfs_ok; + /* Misordered replay or misordered new request */ + return nfserr_seq_misordered; +} + __be32 nfsd4_create_session(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, @@ -1017,11 +1043,54 @@ nfsd4_destroy_session(struct svc_rqst *r, } __be32 -nfsd4_sequence(struct svc_rqst *r, +nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_sequence *seq) { - return -1; /* stub */ + struct nfsd4_session *session; + struct nfsd4_slot *slot; + int status; + + spin_lock(&sessionid_lock); + status = nfserr_badsession; + session = find_in_sessionid_hashtbl(&seq->sessionid); + if (!session) + goto out; + + status = nfserr_badslot; + if (seq->slotid >= session->se_fnumslots) + goto out; + + slot = &session->se_slots[seq->slotid]; + dprintk("%s: slotid %d\n", __func__, seq->slotid); + + status = check_slot_seqid(seq->seqid, slot); + if (status == nfserr_replay_cache) { + cstate->slot = slot; + cstate->session = session; + goto replay_cache; + } + if (status) + goto out; + + /* Success! bump slot seqid */ + slot->sl_inuse = true; + slot->sl_seqid = seq->seqid; + + cstate->slot = slot; + cstate->session = session; + +replay_cache: + /* Renew the clientid on success and on replay. + * Hold a session reference until done processing the compound: + * nfsd4_put_session called only if the cstate slot is set. + */ + renew_client(session->se_client); + nfsd4_get_session(session); +out: + spin_unlock(&sessionid_lock); + dprintk("%s: return %d\n", __func__, ntohl(status)); + return status; } __be32 diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index bebf6d249069..8556575480ce 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1113,7 +1113,16 @@ static __be32 nfsd4_decode_sequence(struct nfsd4_compoundargs *argp, struct nfsd4_sequence *seq) { - return nfserr_opnotsupp; /* stub */ + DECODE_HEAD; + + READ_BUF(NFS4_MAX_SESSIONID_LEN + 16); + COPYMEM(seq->sessionid.data, NFS4_MAX_SESSIONID_LEN); + READ32(seq->seqid); + READ32(seq->slotid); + READ32(seq->maxslots); + READ32(seq->cachethis); + + DECODE_TAIL; } static __be32 @@ -2828,8 +2837,26 @@ static __be32 nfsd4_encode_sequence(struct nfsd4_compoundres *resp, int nfserr, struct nfsd4_sequence *seq) { - /* stub */ - return nfserr; + ENCODE_HEAD; + + if (nfserr) + return nfserr; + + RESERVE_SPACE(NFS4_MAX_SESSIONID_LEN + 20); + WRITEMEM(seq->sessionid.data, NFS4_MAX_SESSIONID_LEN); + WRITE32(seq->seqid); + WRITE32(seq->slotid); + WRITE32(seq->maxslots); + /* + * FIXME: for now: + * target_maxslots = maxslots + * status_flags = 0 + */ + WRITE32(seq->maxslots); + WRITE32(0); + + ADJUST_ARGS(); + return 0; } static __be32 -- cgit v1.2.3 From f9bb94c4c60f6e1d1717077bfddb614f03a607d1 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:12 +0300 Subject: nfsd41: enforce NFS4ERR_SEQUENCE_POS operation order rules for minorversion != 0 only. Signed-off-by: Andy Adamson [nfsd41: do not verify nfserr_sequence_pos for minorversion 0] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 40 ++++++++++++++++++++++++++++++++-------- fs/nfsd/nfs4state.c | 4 ++++ 2 files changed, 36 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index f7be7fabe62c..ec51936d2ce2 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -811,14 +811,15 @@ static inline void nfsd4_increment_op_stats(u32 opnum) typedef __be32(*nfsd4op_func)(struct svc_rqst *, struct nfsd4_compound_state *, void *); +enum nfsd4_op_flags { + ALLOWED_WITHOUT_FH = 1 << 0, /* No current filehandle required */ + ALLOWED_ON_ABSENT_FS = 2 << 0, /* ops processed on absent fs */ + ALLOWED_AS_FIRST_OP = 3 << 0, /* ops reqired first in compound */ +}; struct nfsd4_operation { nfsd4op_func op_func; u32 op_flags; -/* Most ops require a valid current filehandle; a few don't: */ -#define ALLOWED_WITHOUT_FH 1 -/* GETATTR and ops not listed as returning NFS4ERR_MOVED: */ -#define ALLOWED_ON_ABSENT_FS 2 char *op_name; }; @@ -826,6 +827,23 @@ static struct nfsd4_operation nfsd4_ops[]; static const char *nfsd4_op_name(unsigned opnum); +/* + * Enforce NFSv4.1 COMPOUND ordering rules. + * + * TODO: + * - enforce NFS4ERR_NOT_ONLY_OP, + * - DESTROY_SESSION MUST be the final operation in the COMPOUND request. + */ +static bool nfs41_op_ordering_ok(struct nfsd4_compoundargs *args) +{ + if (args->minorversion && args->opcnt > 0) { + struct nfsd4_op *op = &args->ops[0]; + return (op->status == nfserr_op_illegal) || + (nfsd4_ops[op->opnum].op_flags & ALLOWED_AS_FIRST_OP); + } + return true; +} + /* * COMPOUND call. */ @@ -864,6 +882,12 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, if (args->minorversion > NFSD_SUPPORTED_MINOR_VERSION) goto out; + if (!nfs41_op_ordering_ok(args)) { + op = &args->ops[0]; + op->status = nfserr_sequence_pos; + goto encode_op; + } + status = nfs_ok; while (!status && resp->opcnt < args->opcnt) { op = &args->ops[resp->opcnt++]; @@ -1105,22 +1129,22 @@ static struct nfsd4_operation nfsd4_ops[] = { /* NFSv4.1 operations */ [OP_EXCHANGE_ID] = { .op_func = (nfsd4op_func)nfsd4_exchange_id, - .op_flags = ALLOWED_WITHOUT_FH, + .op_flags = ALLOWED_WITHOUT_FH | ALLOWED_AS_FIRST_OP, .op_name = "OP_EXCHANGE_ID", }, [OP_CREATE_SESSION] = { .op_func = (nfsd4op_func)nfsd4_create_session, - .op_flags = ALLOWED_WITHOUT_FH, + .op_flags = ALLOWED_WITHOUT_FH | ALLOWED_AS_FIRST_OP, .op_name = "OP_CREATE_SESSION", }, [OP_DESTROY_SESSION] = { .op_func = (nfsd4op_func)nfsd4_destroy_session, - .op_flags = ALLOWED_WITHOUT_FH, + .op_flags = ALLOWED_WITHOUT_FH | ALLOWED_AS_FIRST_OP, .op_name = "OP_DESTROY_SESSION", }, [OP_SEQUENCE] = { .op_func = (nfsd4op_func)nfsd4_sequence, - .op_flags = ALLOWED_WITHOUT_FH, + .op_flags = ALLOWED_WITHOUT_FH | ALLOWED_AS_FIRST_OP, .op_name = "OP_SEQUENCE", }, }; diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 24b6e0593184..9243dca3576c 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1047,10 +1047,14 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_sequence *seq) { + struct nfsd4_compoundres *resp = rqstp->rq_resp; struct nfsd4_session *session; struct nfsd4_slot *slot; int status; + if (resp->opcnt != 1) + return nfserr_sequence_pos; + spin_lock(&sessionid_lock); status = nfserr_badsession; session = find_in_sessionid_hashtbl(&seq->sessionid); -- cgit v1.2.3 From 074fe897536f095309c5aaffcf912952882ab2cb Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:15 +0300 Subject: nfsd41: DRC save, restore, and clear functions Cache all the result pages, including the rpc header in rq_respages[0], for a request in the slot table cache entry. Cache the statp pointer from nfsd_dispatch which points into rq_respages[0] just past the rpc header. When setting a cache entry, calculate and save the length of the nfs data minus the rpc header for rq_respages[0]. When replaying a cache entry, replace the cached rpc header with the replayed request rpc result header, unless there is not enough room in the cached results first page. In that case, use the cached rpc header. The sessions fore channel maxresponse size cached is set to NFSD_PAGES_PER_SLOT * PAGE_SIZE. For compounds we are cacheing with operations such as READDIR that use the xdr_buf->pages to hold data, we choose to cache the extra page of data rather than copying data from xdr_buf->pages into the xdr_buf->head page. [nfsd41: limit cache to maxresponsesize_cached] [nfsd41: mv nfsd4_set_statp under CONFIG_NFSD_V4_1] [nfsd41: rename nfsd4_move_pages] [nfsd41: rename page_no variable] [nfsd41: rename nfsd4_set_cache_entry] [nfsd41: fix nfsd41_copy_replay_data comment] [nfsd41: add to nfsd4_set_cache_entry] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 142 ++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/nfsd/nfssvc.c | 4 ++ 2 files changed, 146 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 9243dca3576c..a37b91dab1bf 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -852,6 +852,148 @@ out_err: return; } +void +nfsd4_set_statp(struct svc_rqst *rqstp, __be32 *statp) +{ + struct nfsd4_compoundres *resp = rqstp->rq_resp; + + resp->cstate.statp = statp; +} + +/* + * Dereference the result pages. + */ +static void +nfsd4_release_respages(struct page **respages, short resused) +{ + int i; + + dprintk("--> %s\n", __func__); + for (i = 0; i < resused; i++) { + if (!respages[i]) + continue; + put_page(respages[i]); + respages[i] = NULL; + } +} + +static void +nfsd4_copy_pages(struct page **topages, struct page **frompages, short count) +{ + int i; + + for (i = 0; i < count; i++) { + topages[i] = frompages[i]; + if (!topages[i]) + continue; + get_page(topages[i]); + } +} + +/* + * Cache the reply pages up to NFSD_PAGES_PER_SLOT + 1, clearing the previous + * pages. We add a page to NFSD_PAGES_PER_SLOT for the case where the total + * length of the XDR response is less than se_fmaxresp_cached + * (NFSD_PAGES_PER_SLOT * PAGE_SIZE) but the xdr_buf pages is used for a + * of the reply (e.g. readdir). + * + * Store the base and length of the rq_req.head[0] page + * of the NFSv4.1 data, just past the rpc header. + */ +void +nfsd4_store_cache_entry(struct nfsd4_compoundres *resp) +{ + struct nfsd4_cache_entry *entry = &resp->cstate.slot->sl_cache_entry; + struct svc_rqst *rqstp = resp->rqstp; + struct nfsd4_compoundargs *args = rqstp->rq_argp; + struct nfsd4_op *op = &args->ops[resp->opcnt]; + struct kvec *resv = &rqstp->rq_res.head[0]; + + dprintk("--> %s entry %p\n", __func__, entry); + + /* Don't cache a failed OP_SEQUENCE. */ + if (resp->opcnt == 1 && op->opnum == OP_SEQUENCE && resp->cstate.status) + return; + nfsd4_release_respages(entry->ce_respages, entry->ce_resused); + entry->ce_resused = rqstp->rq_resused; + if (entry->ce_resused > NFSD_PAGES_PER_SLOT + 1) + entry->ce_resused = NFSD_PAGES_PER_SLOT + 1; + nfsd4_copy_pages(entry->ce_respages, rqstp->rq_respages, + entry->ce_resused); + entry->ce_status = resp->cstate.status; + entry->ce_datav.iov_base = resp->cstate.statp; + entry->ce_datav.iov_len = resv->iov_len - ((char *)resp->cstate.statp - + (char *)page_address(rqstp->rq_respages[0])); + entry->ce_opcnt = resp->opcnt; + /* Current request rpc header length*/ + entry->ce_rpchdrlen = (char *)resp->cstate.statp - + (char *)page_address(rqstp->rq_respages[0]); +} + +/* + * We keep the rpc header, but take the nfs reply from the replycache. + */ +static int +nfsd41_copy_replay_data(struct nfsd4_compoundres *resp, + struct nfsd4_cache_entry *entry) +{ + struct svc_rqst *rqstp = resp->rqstp; + struct kvec *resv = &resp->rqstp->rq_res.head[0]; + int len; + + /* Current request rpc header length*/ + len = (char *)resp->cstate.statp - + (char *)page_address(rqstp->rq_respages[0]); + if (entry->ce_datav.iov_len + len > PAGE_SIZE) { + dprintk("%s v41 cached reply too large (%Zd).\n", __func__, + entry->ce_datav.iov_len); + return 0; + } + /* copy the cached reply nfsd data past the current rpc header */ + memcpy((char *)resv->iov_base + len, entry->ce_datav.iov_base, + entry->ce_datav.iov_len); + resv->iov_len = len + entry->ce_datav.iov_len; + return 1; +} + +/* + * Keep the first page of the replay. Copy the NFSv4.1 data from the first + * cached page. Replace any futher replay pages from the cache. + */ +__be32 +nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp) +{ + struct nfsd4_cache_entry *entry = &resp->cstate.slot->sl_cache_entry; + __be32 status; + + dprintk("--> %s entry %p\n", __func__, entry); + + + if (!nfsd41_copy_replay_data(resp, entry)) { + /* + * Not enough room to use the replay rpc header, send the + * cached header. Release all the allocated result pages. + */ + svc_free_res_pages(resp->rqstp); + nfsd4_copy_pages(resp->rqstp->rq_respages, entry->ce_respages, + entry->ce_resused); + } else { + /* Release all but the first allocated result page */ + + resp->rqstp->rq_resused--; + svc_free_res_pages(resp->rqstp); + + nfsd4_copy_pages(&resp->rqstp->rq_respages[1], + &entry->ce_respages[1], + entry->ce_resused - 1); + } + + resp->rqstp->rq_resused = entry->ce_resused; + status = entry->ce_status; + + return status; +} + /* * Set the exchange_id flags returned by the server. */ diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index ef0a3686639d..b5168d1898ec 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -515,6 +515,10 @@ nfsd_dispatch(struct svc_rqst *rqstp, __be32 *statp) + rqstp->rq_res.head[0].iov_len; rqstp->rq_res.head[0].iov_len += sizeof(__be32); + /* NFSv4.1 DRC requires statp */ + if (rqstp->rq_vers == 4) + nfsd4_set_statp(rqstp, statp); + /* Now call the procedure handler, and encode NFS status. */ nfserr = proc->pc_func(rqstp, rqstp->rq_argp, rqstp->rq_resp); nfserr = map_new_errors(rqstp->rq_vers, nfserr); -- cgit v1.2.3 From c3d06f9ce8544fecfe13e377d1e2c2e47fe18dbc Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:18 +0300 Subject: nfsd41: hard page limit for DRC Use no more than 1/128th of the number of free pages at nfsd startup for the v4.1 DRC. This is an arbitrary default which should probably end up under the control of an administrator. Signed-off-by: Andy Adamson [moved added fields in struct svc_serv under CONFIG_NFSD_V4_1] Signed-off-by: Benny Halevy [fix set_max_drc calculation of sv_drc_max_pages] [moved NFSD_DRC_SIZE_SHIFT's declaration up in header file] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfssvc.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index b5168d1898ec..b53a098e97a4 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include @@ -197,6 +198,26 @@ void nfsd_reset_versions(void) } } +/* + * Each session guarantees a negotiated per slot memory cache for replies + * which in turn consumes memory beyond the v2/v3/v4.0 server. A dedicated + * NFSv4.1 server might want to use more memory for a DRC than a machine + * with mutiple services. + * + * Impose a hard limit on the number of pages for the DRC which varies + * according to the machines free pages. This is of course only a default. + * + * For now this is a #defined shift which could be under admin control + * in the future. + */ +static void set_max_drc(void) +{ + nfsd_serv->sv_drc_max_pages = nr_free_buffer_pages() + >> NFSD_DRC_SIZE_SHIFT; + nfsd_serv->sv_drc_pages_used = 0; + dprintk("%s svc_drc_max_pages %u\n", __func__, + nfsd_serv->sv_drc_max_pages); +} int nfsd_create_serv(void) { @@ -229,6 +250,8 @@ int nfsd_create_serv(void) nfsd_last_thread, nfsd, THIS_MODULE); if (nfsd_serv == NULL) err = -ENOMEM; + else + set_max_drc(); do_gettimeofday(&nfssvc_boot); /* record boot time */ return err; -- cgit v1.2.3 From da3846a2866ddf239311766ff434a82e7b4ac701 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:22 +0300 Subject: nfsd41: nfsd DRC logic Replay a request in nfsd4_sequence. Add a minorversion to struct nfsd4_compound_state. Pass the current slot to nfs4svc_encode_compound res via struct nfsd4_compoundres to set an NFSv4.1 DRC entry. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfsd41: use bool inuse for slot state] Signed-off-by: Benny Halevy [nfsd41: use cstate session in nfs4svc_encode_compoundres] [nfsd41 replace nfsd4_set_cache_entry] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 7 +++++++ fs/nfsd/nfs4state.c | 6 ++++++ fs/nfsd/nfs4xdr.c | 11 +++++++++++ 3 files changed, 24 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index ec51936d2ce2..9e2ee75e0f7c 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -936,6 +936,12 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, BUG_ON(op->status == nfs_ok); encode_op: + /* Only from SEQUENCE or CREATE_SESSION */ + if (resp->cstate.status == nfserr_replay_cache) { + dprintk("%s NFS4.1 replay from cache\n", __func__); + status = op->status; + goto out; + } if (op->status == nfserr_replay_me) { op->replay = &cstate->replay_owner->so_replay; nfsd4_encode_replay(resp, op); @@ -964,6 +970,7 @@ encode_op: status = nfserr_jukebox; } + resp->cstate.status = status; fh_put(&resp->cstate.current_fh); fh_put(&resp->cstate.save_fh); BUG_ON(resp->cstate.replay_owner); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index a37b91dab1bf..e20d4345040f 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -989,6 +989,8 @@ nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp) } resp->rqstp->rq_resused = entry->ce_resused; + resp->opcnt = entry->ce_opcnt; + resp->cstate.iovlen = entry->ce_datav.iov_len + entry->ce_rpchdrlen; status = entry->ce_status; return status; @@ -1214,6 +1216,10 @@ nfsd4_sequence(struct svc_rqst *rqstp, if (status == nfserr_replay_cache) { cstate->slot = slot; cstate->session = session; + /* Return the cached reply status and set cstate->status + * for nfsd4_svc_encode_compoundres processing*/ + status = nfsd4_replay_cache_entry(resp); + cstate->status = nfserr_replay_cache; goto replay_cache; } if (status) diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 8556575480ce..09415bcf078e 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3049,6 +3049,17 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo iov = &rqstp->rq_res.head[0]; iov->iov_len = ((char*)resp->p) - (char*)iov->iov_base; BUG_ON(iov->iov_len > PAGE_SIZE); + if (resp->cstate.slot != NULL) { + if (resp->cstate.status == nfserr_replay_cache) { + iov->iov_len = resp->cstate.iovlen; + } else { + nfsd4_store_cache_entry(resp); + dprintk("%s: SET SLOT STATE TO AVAILABLE\n", __func__); + resp->cstate.slot->sl_inuse = 0; + } + if (resp->cstate.session) + nfsd4_put_session(resp->cstate.session); + } return 1; } -- cgit v1.2.3 From 14778a133e3be332be77d981552a79260a61ee17 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:25 +0300 Subject: nfsd41: clear DRC cache on free_session Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index e20d4345040f..f25a7d2a9fac 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -441,12 +441,19 @@ release_session(struct nfsd4_session *ses) nfsd4_put_session(ses); } +static void nfsd4_release_respages(struct page **respages, short resused); + void free_session(struct kref *kref) { struct nfsd4_session *ses; + int i; ses = container_of(kref, struct nfsd4_session, se_ref); + for (i = 0; i < ses->se_fnumslots; i++) { + struct nfsd4_cache_entry *e = &ses->se_slots[i].sl_cache_entry; + nfsd4_release_respages(e->ce_respages, e->ce_resused); + } kfree(ses->se_slots); kfree(ses); } -- cgit v1.2.3 From ec6b5d7b5064fde27aee798b81107ea3a830de85 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:28 +0300 Subject: nfsd41: create_session operation Implement the create_session operation confoming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 Look up the client id (generated by the server on exchange_id, given by the client on create_session). If neither a confirmed or unconfirmed client is found then the client id is stale If a confirmed cilent is found (i.e. we already received create_session for it) then compare the sequence id to determine if it's a replay or possibly a mis-ordered rpc. If the seqid is in order, update the confirmed client seqid and procedd with updating the session parameters. If an unconfirmed client_id is found then verify the creds and seqid. If both match move the client id to confirmed state and proceed with processing the create_session. Currently, we do not support persistent sessions, and RDMA. alloc_init_session generates a new sessionid and creates a session structure. NFSD_PAGES_PER_SLOT is used for the max response cached calculation, and for the counting of DRC pages using the hard limits set in struct srv_serv. A note on NFSD_PAGES_PER_SLOT: Other patches in this series allow for NFSD_PAGES_PER_SLOT + 1 pages to be cached in a DRC slot when the response size is less than NFSD_PAGES_PER_SLOT * PAGE_SIZE but xdr_buf pages are used. e.g. a READDIR operation will encode a small amount of data in the xdr_buf head, and then the READDIR in the xdr_buf pages. So, the hard limit calculation use of pages by a session is underestimated by the number of cached operations using the xdr_buf pages. Yet another patch caches no pages for the solo sequence operation, or any compound where cache_this is False. So the hard limit calculation use of pages by a session is overestimated by the number of these operations in the cache. TODO: improve resource pre-allocation and negotiate session parameters accordingly. Respect and possibly adjust backchannel attributes. Signed-off-by: Marc Eshel Signed-off-by: Dean Hildebrand [nfsd41: remove headerpadsz from channel attributes] Our client and server only support a headerpadsz of 0. [nfsd41: use DRC limits in fore channel init] [nfsd41: do not change CREATE_SESSION back channel attrs] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [use sessionid_lock spin lock] [nfsd41: use bool inuse for slot state] Signed-off-by: Benny Halevy [nfsd41 remove sl_session from alloc_init_session] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [simplify nfsd4_encode_create_session error handling] [nfsd41: fix comment style in init_forechannel_attrs] [nfsd41: allocate struct nfsd4_session and slot table in one piece] [nfsd41: no need to INIT_LIST_HEAD in alloc_init_session just prior to list_add] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 188 +++++++++++++++++++++++++++++++++++++++++++++++++++- fs/nfsd/nfs4xdr.c | 148 ++++++++++++++++++++++++++++++++++++++++- 2 files changed, 332 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index f25a7d2a9fac..463ae39df148 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -68,6 +68,7 @@ static u32 current_delegid = 1; static u32 nfs4_init; static stateid_t zerostateid; /* bits all 0 */ static stateid_t onestateid; /* bits all 1 */ +static u64 current_sessionid = 1; #define ZERO_STATEID(stateid) (!memcmp((stateid), &zerostateid, sizeof(stateid_t))) #define ONE_STATEID(stateid) (!memcmp((stateid), &onestateid, sizeof(stateid_t))) @@ -401,6 +402,131 @@ dump_sessionid(const char *fn, struct nfs4_sessionid *sessionid) dprintk("%s: %u:%u:%u:%u\n", fn, ptr[0], ptr[1], ptr[2], ptr[3]); } +static void +gen_sessionid(struct nfsd4_session *ses) +{ + struct nfs4_client *clp = ses->se_client; + struct nfsd4_sessionid *sid; + + sid = (struct nfsd4_sessionid *)ses->se_sessionid.data; + sid->clientid = clp->cl_clientid; + sid->sequence = current_sessionid++; + sid->reserved = 0; +} + +/* + * Give the client the number of slots it requests bound by + * NFSD_MAX_SLOTS_PER_SESSION and by sv_drc_max_pages. + * + * If we run out of pages (sv_drc_pages_used == sv_drc_max_pages) we + * should (up to a point) re-negotiate active sessions and reduce their + * slot usage to make rooom for new connections. For now we just fail the + * create session. + */ +static int set_forechannel_maxreqs(struct nfsd4_channel_attrs *fchan) +{ + int status = 0, np = fchan->maxreqs * NFSD_PAGES_PER_SLOT; + + spin_lock(&nfsd_serv->sv_lock); + if (np + nfsd_serv->sv_drc_pages_used > nfsd_serv->sv_drc_max_pages) + np = nfsd_serv->sv_drc_max_pages - nfsd_serv->sv_drc_pages_used; + nfsd_serv->sv_drc_pages_used += np; + spin_unlock(&nfsd_serv->sv_lock); + + if (np <= 0) { + status = nfserr_resource; + fchan->maxreqs = 0; + } else + fchan->maxreqs = np / NFSD_PAGES_PER_SLOT; + + return status; +} + +/* + * fchan holds the client values on input, and the server values on output + */ +static int init_forechannel_attrs(struct svc_rqst *rqstp, + struct nfsd4_session *session, + struct nfsd4_channel_attrs *fchan) +{ + int status = 0; + __u32 maxcount = svc_max_payload(rqstp); + + /* headerpadsz set to zero in encode routine */ + + /* Use the client's max request and max response size if possible */ + if (fchan->maxreq_sz > maxcount) + fchan->maxreq_sz = maxcount; + session->se_fmaxreq_sz = fchan->maxreq_sz; + + if (fchan->maxresp_sz > maxcount) + fchan->maxresp_sz = maxcount; + session->se_fmaxresp_sz = fchan->maxresp_sz; + + /* Set the max response cached size our default which is + * a multiple of PAGE_SIZE and small */ + session->se_fmaxresp_cached = NFSD_PAGES_PER_SLOT * PAGE_SIZE; + fchan->maxresp_cached = session->se_fmaxresp_cached; + + /* Use the client's maxops if possible */ + if (fchan->maxops > NFSD_MAX_OPS_PER_COMPOUND) + fchan->maxops = NFSD_MAX_OPS_PER_COMPOUND; + session->se_fmaxops = fchan->maxops; + + /* try to use the client requested number of slots */ + if (fchan->maxreqs > NFSD_MAX_SLOTS_PER_SESSION) + fchan->maxreqs = NFSD_MAX_SLOTS_PER_SESSION; + + /* FIXME: Error means no more DRC pages so the server should + * recover pages from existing sessions. For now fail session + * creation. + */ + status = set_forechannel_maxreqs(fchan); + + session->se_fnumslots = fchan->maxreqs; + return status; +} + +static int +alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp, + struct nfsd4_create_session *cses) +{ + struct nfsd4_session *new, tmp; + int idx, status = nfserr_resource, slotsize; + + memset(&tmp, 0, sizeof(tmp)); + + /* FIXME: For now, we just accept the client back channel attributes. */ + status = init_forechannel_attrs(rqstp, &tmp, &cses->fore_channel); + if (status) + goto out; + + /* allocate struct nfsd4_session and slot table in one piece */ + slotsize = tmp.se_fnumslots * sizeof(struct nfsd4_slot); + new = kzalloc(sizeof(*new) + slotsize, GFP_KERNEL); + if (!new) + goto out; + + memcpy(new, &tmp, sizeof(*new)); + + new->se_client = clp; + gen_sessionid(new); + idx = hash_sessionid(&new->se_sessionid); + memcpy(clp->cl_sessionid.data, new->se_sessionid.data, + NFS4_MAX_SESSIONID_LEN); + + new->se_flags = cses->flags; + kref_init(&new->se_ref); + spin_lock(&sessionid_lock); + list_add(&new->se_hash, &sessionid_hashtbl[idx]); + list_add(&new->se_perclnt, &clp->cl_sessions); + spin_unlock(&sessionid_lock); + + status = nfs_ok; +out: + return status; +} + /* caller must hold sessionid_lock */ static struct nfsd4_session * find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid) @@ -1182,7 +1308,67 @@ nfsd4_create_session(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_create_session *cr_ses) { - return -1; /* stub */ + u32 ip_addr = svc_addr_in(rqstp)->sin_addr.s_addr; + struct nfs4_client *conf, *unconf; + int status = 0; + + nfs4_lock_state(); + unconf = find_unconfirmed_client(&cr_ses->clientid); + conf = find_confirmed_client(&cr_ses->clientid); + + if (conf) { + status = nfs_ok; + if (conf->cl_seqid == cr_ses->seqid) { + dprintk("Got a create_session replay! seqid= %d\n", + conf->cl_seqid); + goto out_replay; + } else if (cr_ses->seqid != conf->cl_seqid + 1) { + status = nfserr_seq_misordered; + dprintk("Sequence misordered!\n"); + dprintk("Expected seqid= %d but got seqid= %d\n", + conf->cl_seqid, cr_ses->seqid); + goto out; + } + conf->cl_seqid++; + } else if (unconf) { + if (!same_creds(&unconf->cl_cred, &rqstp->rq_cred) || + (ip_addr != unconf->cl_addr)) { + status = nfserr_clid_inuse; + goto out; + } + + if (unconf->cl_seqid != cr_ses->seqid) { + status = nfserr_seq_misordered; + goto out; + } + + move_to_confirmed(unconf); + + /* + * We do not support RDMA or persistent sessions + */ + cr_ses->flags &= ~SESSION4_PERSIST; + cr_ses->flags &= ~SESSION4_RDMA; + + conf = unconf; + } else { + status = nfserr_stale_clientid; + goto out; + } + + status = alloc_init_session(rqstp, conf, cr_ses); + if (status) + goto out; + +out_replay: + memcpy(cr_ses->sessionid.data, conf->cl_sessionid.data, + NFS4_MAX_SESSIONID_LEN); + cr_ses->seqid = conf->cl_seqid; + +out: + nfs4_unlock_state(); + dprintk("%s returns %d\n", __func__, ntohl(status)); + return status; } __be32 diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 09415bcf078e..671f9b96429b 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1099,7 +1099,108 @@ static __be32 nfsd4_decode_create_session(struct nfsd4_compoundargs *argp, struct nfsd4_create_session *sess) { - return nfserr_opnotsupp; /* stub */ + DECODE_HEAD; + + u32 dummy; + char *machine_name; + int i; + int nr_secflavs; + + READ_BUF(16); + COPYMEM(&sess->clientid, 8); + READ32(sess->seqid); + READ32(sess->flags); + + /* Fore channel attrs */ + READ_BUF(28); + READ32(dummy); /* headerpadsz is always 0 */ + READ32(sess->fore_channel.maxreq_sz); + READ32(sess->fore_channel.maxresp_sz); + READ32(sess->fore_channel.maxresp_cached); + READ32(sess->fore_channel.maxops); + READ32(sess->fore_channel.maxreqs); + READ32(sess->fore_channel.nr_rdma_attrs); + if (sess->fore_channel.nr_rdma_attrs == 1) { + READ_BUF(4); + READ32(sess->fore_channel.rdma_attrs); + } else if (sess->fore_channel.nr_rdma_attrs > 1) { + dprintk("Too many fore channel attr bitmaps!\n"); + goto xdr_error; + } + + /* Back channel attrs */ + READ_BUF(28); + READ32(dummy); /* headerpadsz is always 0 */ + READ32(sess->back_channel.maxreq_sz); + READ32(sess->back_channel.maxresp_sz); + READ32(sess->back_channel.maxresp_cached); + READ32(sess->back_channel.maxops); + READ32(sess->back_channel.maxreqs); + READ32(sess->back_channel.nr_rdma_attrs); + if (sess->back_channel.nr_rdma_attrs == 1) { + READ_BUF(4); + READ32(sess->back_channel.rdma_attrs); + } else if (sess->back_channel.nr_rdma_attrs > 1) { + dprintk("Too many back channel attr bitmaps!\n"); + goto xdr_error; + } + + READ_BUF(8); + READ32(sess->callback_prog); + + /* callback_sec_params4 */ + READ32(nr_secflavs); + for (i = 0; i < nr_secflavs; ++i) { + READ_BUF(4); + READ32(dummy); + switch (dummy) { + case RPC_AUTH_NULL: + /* Nothing to read */ + break; + case RPC_AUTH_UNIX: + READ_BUF(8); + /* stamp */ + READ32(dummy); + + /* machine name */ + READ32(dummy); + READ_BUF(dummy); + SAVEMEM(machine_name, dummy); + + /* uid, gid */ + READ_BUF(8); + READ32(sess->uid); + READ32(sess->gid); + + /* more gids */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy * 4); + for (i = 0; i < dummy; ++i) + READ32(dummy); + break; + case RPC_AUTH_GSS: + dprintk("RPC_AUTH_GSS callback secflavor " + "not supported!\n"); + READ_BUF(8); + /* gcbp_service */ + READ32(dummy); + /* gcbp_handle_from_server */ + READ32(dummy); + READ_BUF(dummy); + p += XDR_QUADLEN(dummy); + /* gcbp_handle_from_client */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy); + p += XDR_QUADLEN(dummy); + break; + default: + dprintk("Illegal callback secflavor\n"); + return nfserr_inval; + } + } + DECODE_TAIL; } static __be32 @@ -2821,8 +2922,49 @@ static __be32 nfsd4_encode_create_session(struct nfsd4_compoundres *resp, int nfserr, struct nfsd4_create_session *sess) { - /* stub */ - return nfserr; + ENCODE_HEAD; + + if (nfserr) + return nfserr; + + RESERVE_SPACE(24); + WRITEMEM(sess->sessionid.data, NFS4_MAX_SESSIONID_LEN); + WRITE32(sess->seqid); + WRITE32(sess->flags); + ADJUST_ARGS(); + + RESERVE_SPACE(28); + WRITE32(0); /* headerpadsz */ + WRITE32(sess->fore_channel.maxreq_sz); + WRITE32(sess->fore_channel.maxresp_sz); + WRITE32(sess->fore_channel.maxresp_cached); + WRITE32(sess->fore_channel.maxops); + WRITE32(sess->fore_channel.maxreqs); + WRITE32(sess->fore_channel.nr_rdma_attrs); + ADJUST_ARGS(); + + if (sess->fore_channel.nr_rdma_attrs) { + RESERVE_SPACE(4); + WRITE32(sess->fore_channel.rdma_attrs); + ADJUST_ARGS(); + } + + RESERVE_SPACE(28); + WRITE32(0); /* headerpadsz */ + WRITE32(sess->back_channel.maxreq_sz); + WRITE32(sess->back_channel.maxresp_sz); + WRITE32(sess->back_channel.maxresp_cached); + WRITE32(sess->back_channel.maxops); + WRITE32(sess->back_channel.maxreqs); + WRITE32(sess->back_channel.nr_rdma_attrs); + ADJUST_ARGS(); + + if (sess->back_channel.nr_rdma_attrs) { + RESERVE_SPACE(4); + WRITE32(sess->back_channel.rdma_attrs); + ADJUST_ARGS(); + } + return 0; } static __be32 -- cgit v1.2.3 From 38eb76a54d803e6792816623651b1a9cb85f8d01 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:32 +0300 Subject: nfsd41: Add a create session replay cache Replace the nfs4_client cl_seqid field with a single struct nfs41_slot used for the create session replay cache. The CREATE_SESSION slot sets the sl_session pointer to NULL. Otherwise, the slot and it's replay cache are used just like the session slots. Fix unconfirmed create_session replay response by initializing the create_session slot sequence id to 0. A future patch will set the CREATE_SESSION cache when a SEQUENCE operation preceeds the CREATE_SESSION operation. This compound is currently only cached in the session slot table. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfsd41: use bool inuse for slot state] Signed-off-by: Benny Halevy [nfsd41: revert portion of nfsd4_set_cache_entry] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 39 +++++++++++++++++++++++++++------------ 1 file changed, 27 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 463ae39df148..58f9797eb09e 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -649,6 +649,8 @@ static inline void free_client(struct nfs4_client *clp) { shutdown_callback_client(clp); + nfsd4_release_respages(clp->cl_slot.sl_cache_entry.ce_respages, + clp->cl_slot.sl_cache_entry.ce_resused); if (clp->cl_cred.cr_group_info) put_group_info(clp->cl_cred.cr_group_info); kfree(clp->cl_principal); @@ -1263,11 +1265,12 @@ out_copy: exid->clientid.cl_boot = new->cl_clientid.cl_boot; exid->clientid.cl_id = new->cl_clientid.cl_id; - new->cl_seqid = exid->seqid = 1; + new->cl_slot.sl_seqid = 0; + exid->seqid = 1; nfsd4_set_ex_flags(new, exid); dprintk("nfsd4_exchange_id seqid %d flags %x\n", - new->cl_seqid, new->cl_exchange_flags); + new->cl_slot.sl_seqid, new->cl_exchange_flags); status = nfs_ok; out: @@ -1309,7 +1312,9 @@ nfsd4_create_session(struct svc_rqst *rqstp, struct nfsd4_create_session *cr_ses) { u32 ip_addr = svc_addr_in(rqstp)->sin_addr.s_addr; + struct nfsd4_compoundres *resp = rqstp->rq_resp; struct nfs4_client *conf, *unconf; + struct nfsd4_slot *slot = NULL; int status = 0; nfs4_lock_state(); @@ -1317,19 +1322,24 @@ nfsd4_create_session(struct svc_rqst *rqstp, conf = find_confirmed_client(&cr_ses->clientid); if (conf) { - status = nfs_ok; - if (conf->cl_seqid == cr_ses->seqid) { + slot = &conf->cl_slot; + status = check_slot_seqid(cr_ses->seqid, slot); + if (status == nfserr_replay_cache) { dprintk("Got a create_session replay! seqid= %d\n", - conf->cl_seqid); - goto out_replay; - } else if (cr_ses->seqid != conf->cl_seqid + 1) { + slot->sl_seqid); + cstate->slot = slot; + cstate->status = status; + /* Return the cached reply status */ + status = nfsd4_replay_cache_entry(resp); + goto out; + } else if (cr_ses->seqid != conf->cl_slot.sl_seqid + 1) { status = nfserr_seq_misordered; dprintk("Sequence misordered!\n"); dprintk("Expected seqid= %d but got seqid= %d\n", - conf->cl_seqid, cr_ses->seqid); + slot->sl_seqid, cr_ses->seqid); goto out; } - conf->cl_seqid++; + conf->cl_slot.sl_seqid++; } else if (unconf) { if (!same_creds(&unconf->cl_cred, &rqstp->rq_cred) || (ip_addr != unconf->cl_addr)) { @@ -1337,11 +1347,15 @@ nfsd4_create_session(struct svc_rqst *rqstp, goto out; } - if (unconf->cl_seqid != cr_ses->seqid) { + slot = &unconf->cl_slot; + status = check_slot_seqid(cr_ses->seqid, slot); + if (status) { + /* an unconfirmed replay returns misordered */ status = nfserr_seq_misordered; goto out; } + slot->sl_seqid++; /* from 0 to 1 */ move_to_confirmed(unconf); /* @@ -1360,11 +1374,12 @@ nfsd4_create_session(struct svc_rqst *rqstp, if (status) goto out; -out_replay: memcpy(cr_ses->sessionid.data, conf->cl_sessionid.data, NFS4_MAX_SESSIONID_LEN); - cr_ses->seqid = conf->cl_seqid; + cr_ses->seqid = slot->sl_seqid; + slot->sl_inuse = true; + cstate->slot = slot; out: nfs4_unlock_state(); dprintk("%s returns %d\n", __func__, ntohl(status)); -- cgit v1.2.3 From bf864a31d50e3e94d6e76537b97d664913906ff8 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:35 +0300 Subject: nfsd41: non-page DRC for solo sequence responses A session inactivity time compound (lease renewal) or a compound where the sequence operation has sa_cachethis set to FALSE do not require any pages to be held in the v4.1 DRC. This is because struct nfsd4_slot is already caching the session information. Add logic to the nfs41 server to not cache response pages for solo sequence responses. Return nfserr_replay_uncached_rep on the operation following the sequence operation when sa_cachethis is FALSE. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfsd41: use cstate session in nfsd4_replay_cache_entry] [nfsd41: rename nfsd4_no_page_in_cache] [nfsd41 rename nfsd4_enc_no_page_replay] [nfsd41 nfsd4_is_solo_sequence] [nfsd41 change nfsd4_not_cached return] Signed-off-by: Andy Adamson [changed return type to bool] Signed-off-by: Benny Halevy [nfsd41 drop parens in nfsd4_is_solo_sequence call] Signed-off-by: Andy Adamson [changed "== 0" to "!"] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 34 ++++++++++++++++++++++++++++++++-- fs/nfsd/nfs4state.c | 47 +++++++++++++++++++++++++++++++++++++++++------ fs/nfsd/nfs4xdr.c | 5 +++-- 3 files changed, 76 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 9e2ee75e0f7c..ae21a4efe36c 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -827,6 +827,34 @@ static struct nfsd4_operation nfsd4_ops[]; static const char *nfsd4_op_name(unsigned opnum); +/* + * This is a replay of a compound for which no cache entry pages + * were used. Encode the sequence operation, and if cachethis is FALSE + * encode the uncache rep error on the next operation. + */ +static __be32 +nfsd4_enc_uncached_replay(struct nfsd4_compoundargs *args, + struct nfsd4_compoundres *resp) +{ + struct nfsd4_op *op; + + dprintk("--> %s resp->opcnt %d ce_cachethis %u \n", __func__, + resp->opcnt, resp->cstate.slot->sl_cache_entry.ce_cachethis); + + /* Encode the replayed sequence operation */ + BUG_ON(resp->opcnt != 1); + op = &args->ops[resp->opcnt - 1]; + nfsd4_encode_operation(resp, op); + + /*return nfserr_retry_uncached_rep in next operation. */ + if (resp->cstate.slot->sl_cache_entry.ce_cachethis == 0) { + op = &args->ops[resp->opcnt++]; + op->status = nfserr_retry_uncached_rep; + nfsd4_encode_operation(resp, op); + } + return op->status; +} + /* * Enforce NFSv4.1 COMPOUND ordering rules. * @@ -895,7 +923,6 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, dprintk("nfsv4 compound op #%d/%d: %d (%s)\n", resp->opcnt, args->opcnt, op->opnum, nfsd4_op_name(op->opnum)); - /* * The XDR decode routines may have pre-set op->status; * for example, if there is a miscellaneous XDR error @@ -939,7 +966,10 @@ encode_op: /* Only from SEQUENCE or CREATE_SESSION */ if (resp->cstate.status == nfserr_replay_cache) { dprintk("%s NFS4.1 replay from cache\n", __func__); - status = op->status; + if (nfsd4_not_cached(resp)) + status = nfsd4_enc_uncached_replay(args, resp); + else + status = op->status; goto out; } if (op->status == nfserr_replay_me) { diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 58f9797eb09e..04a395fb5dce 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1049,17 +1049,31 @@ nfsd4_store_cache_entry(struct nfsd4_compoundres *resp) /* Don't cache a failed OP_SEQUENCE. */ if (resp->opcnt == 1 && op->opnum == OP_SEQUENCE && resp->cstate.status) return; + nfsd4_release_respages(entry->ce_respages, entry->ce_resused); + entry->ce_opcnt = resp->opcnt; + entry->ce_status = resp->cstate.status; + + /* + * Don't need a page to cache just the sequence operation - the slot + * does this for us! + */ + + if (nfsd4_not_cached(resp)) { + entry->ce_resused = 0; + entry->ce_rpchdrlen = 0; + dprintk("%s Just cache SEQUENCE. ce_cachethis %d\n", __func__, + resp->cstate.slot->sl_cache_entry.ce_cachethis); + return; + } entry->ce_resused = rqstp->rq_resused; if (entry->ce_resused > NFSD_PAGES_PER_SLOT + 1) entry->ce_resused = NFSD_PAGES_PER_SLOT + 1; nfsd4_copy_pages(entry->ce_respages, rqstp->rq_respages, entry->ce_resused); - entry->ce_status = resp->cstate.status; entry->ce_datav.iov_base = resp->cstate.statp; entry->ce_datav.iov_len = resv->iov_len - ((char *)resp->cstate.statp - (char *)page_address(rqstp->rq_respages[0])); - entry->ce_opcnt = resp->opcnt; /* Current request rpc header length*/ entry->ce_rpchdrlen = (char *)resp->cstate.statp - (char *)page_address(rqstp->rq_respages[0]); @@ -1096,13 +1110,28 @@ nfsd41_copy_replay_data(struct nfsd4_compoundres *resp, * cached page. Replace any futher replay pages from the cache. */ __be32 -nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp) +nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp, + struct nfsd4_sequence *seq) { struct nfsd4_cache_entry *entry = &resp->cstate.slot->sl_cache_entry; __be32 status; dprintk("--> %s entry %p\n", __func__, entry); + /* + * If this is just the sequence operation, we did not keep + * a page in the cache entry because we can just use the + * slot info stored in struct nfsd4_sequence that was checked + * against the slot in nfsd4_sequence(). + * + * This occurs when seq->cachethis is FALSE, or when the client + * session inactivity timer fires and a solo sequence operation + * is sent (lease renewal). + */ + if (seq && nfsd4_not_cached(resp)) { + seq->maxslots = resp->cstate.session->se_fnumslots; + return nfs_ok; + } if (!nfsd41_copy_replay_data(resp, entry)) { /* @@ -1330,7 +1359,7 @@ nfsd4_create_session(struct svc_rqst *rqstp, cstate->slot = slot; cstate->status = status; /* Return the cached reply status */ - status = nfsd4_replay_cache_entry(resp); + status = nfsd4_replay_cache_entry(resp, NULL); goto out; } else if (cr_ses->seqid != conf->cl_slot.sl_seqid + 1) { status = nfserr_seq_misordered; @@ -1380,6 +1409,8 @@ nfsd4_create_session(struct svc_rqst *rqstp, slot->sl_inuse = true; cstate->slot = slot; + /* Ensure a page is used for the cache */ + slot->sl_cache_entry.ce_cachethis = 1; out: nfs4_unlock_state(); dprintk("%s returns %d\n", __func__, ntohl(status)); @@ -1425,8 +1456,8 @@ nfsd4_sequence(struct svc_rqst *rqstp, cstate->slot = slot; cstate->session = session; /* Return the cached reply status and set cstate->status - * for nfsd4_svc_encode_compoundres processing*/ - status = nfsd4_replay_cache_entry(resp); + * for nfsd4_svc_encode_compoundres processing */ + status = nfsd4_replay_cache_entry(resp, seq); cstate->status = nfserr_replay_cache; goto replay_cache; } @@ -1436,6 +1467,10 @@ nfsd4_sequence(struct svc_rqst *rqstp, /* Success! bump slot seqid */ slot->sl_inuse = true; slot->sl_seqid = seq->seqid; + slot->sl_cache_entry.ce_cachethis = seq->cachethis; + /* Always set the cache entry cachethis for solo sequence */ + if (nfsd4_is_solo_sequence(resp)) + slot->sl_cache_entry.ce_cachethis = 1; cstate->slot = slot; cstate->session = session; diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 671f9b96429b..64bc2150a6fa 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -2975,7 +2975,7 @@ nfsd4_encode_destroy_session(struct nfsd4_compoundres *resp, int nfserr, return nfserr; } -static __be32 +__be32 nfsd4_encode_sequence(struct nfsd4_compoundres *resp, int nfserr, struct nfsd4_sequence *seq) { @@ -3192,7 +3192,8 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo iov->iov_len = ((char*)resp->p) - (char*)iov->iov_base; BUG_ON(iov->iov_len > PAGE_SIZE); if (resp->cstate.slot != NULL) { - if (resp->cstate.status == nfserr_replay_cache) { + if (resp->cstate.status == nfserr_replay_cache && + !nfsd4_not_cached(resp)) { iov->iov_len = resp->cstate.iovlen; } else { nfsd4_store_cache_entry(resp); -- cgit v1.2.3 From e10e0cfc2f27364c73b28adbd3c8688d97049e73 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 3 Apr 2009 08:28:38 +0300 Subject: nfsd41: destroy_session operation Implement the destory_session operation confoming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 [use sessionid_lock spin lock] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 29 ++++++++++++++++++++++++++++- fs/nfsd/nfs4xdr.c | 7 +++++-- 2 files changed, 33 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 04a395fb5dce..9192e5b35f42 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1422,7 +1422,34 @@ nfsd4_destroy_session(struct svc_rqst *r, struct nfsd4_compound_state *cstate, struct nfsd4_destroy_session *sessionid) { - return -1; /* stub */ + struct nfsd4_session *ses; + u32 status = nfserr_badsession; + + /* Notes: + * - The confirmed nfs4_client->cl_sessionid holds destroyed sessinid + * - Should we return nfserr_back_chan_busy if waiting for + * callbacks on to-be-destroyed session? + * - Do we need to clear any callback info from previous session? + */ + + dump_sessionid(__func__, &sessionid->sessionid); + spin_lock(&sessionid_lock); + ses = find_in_sessionid_hashtbl(&sessionid->sessionid); + if (!ses) { + spin_unlock(&sessionid_lock); + goto out; + } + + unhash_session(ses); + spin_unlock(&sessionid_lock); + + /* wait for callbacks */ + shutdown_callback_client(ses->se_client); + nfsd4_put_session(ses); + status = nfs_ok; +out: + dprintk("%s returns %d\n", __func__, ntohl(status)); + return status; } __be32 diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 64bc2150a6fa..c6a726d04efd 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1207,7 +1207,11 @@ static __be32 nfsd4_decode_destroy_session(struct nfsd4_compoundargs *argp, struct nfsd4_destroy_session *destroy_session) { - return nfserr_opnotsupp; /* stub */ + DECODE_HEAD; + READ_BUF(NFS4_MAX_SESSIONID_LEN); + COPYMEM(destroy_session->sessionid.data, NFS4_MAX_SESSIONID_LEN); + + DECODE_TAIL; } static __be32 @@ -2971,7 +2975,6 @@ static __be32 nfsd4_encode_destroy_session(struct nfsd4_compoundres *resp, int nfserr, struct nfsd4_destroy_session *destroy_session) { - /* stub */ return nfserr; } -- cgit v1.2.3 From dd453dfd70538cadc02cb47ff8d8cfd0cb8cf435 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 3 Apr 2009 08:28:41 +0300 Subject: nfsd: pass nfsd4_compound_state* to nfs4_preprocess_{state,seq}id_op Currently we only use cstate->current_fh, will also be used by nfsd41 code. Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 10 ++++------ fs/nfsd/nfs4state.c | 22 ++++++++++++++-------- 2 files changed, 18 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index ae21a4efe36c..e6a0f314fdbe 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -513,9 +513,8 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nfs4_lock_state(); /* check stateid */ - if ((status = nfs4_preprocess_stateid_op(&cstate->current_fh, - &read->rd_stateid, - RD_STATE, &read->rd_filp))) { + if ((status = nfs4_preprocess_stateid_op(cstate, &read->rd_stateid, + RD_STATE, &read->rd_filp))) { dprintk("NFSD: nfsd4_read: couldn't process stateid!\n"); goto out; } @@ -646,7 +645,7 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (setattr->sa_iattr.ia_valid & ATTR_SIZE) { nfs4_lock_state(); - status = nfs4_preprocess_stateid_op(&cstate->current_fh, + status = nfs4_preprocess_stateid_op(cstate, &setattr->sa_stateid, WR_STATE, NULL); nfs4_unlock_state(); if (status) { @@ -686,8 +685,7 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, return nfserr_inval; nfs4_lock_state(); - status = nfs4_preprocess_stateid_op(&cstate->current_fh, stateid, - WR_STATE, &filp); + status = nfs4_preprocess_stateid_op(cstate, stateid, WR_STATE, &filp); if (filp) get_file(filp); nfs4_unlock_state(); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 9192e5b35f42..f49f305096ec 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2810,10 +2810,12 @@ static int is_delegation_stateid(stateid_t *stateid) * Checks for stateid operations */ __be32 -nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int flags, struct file **filpp) +nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, + stateid_t *stateid, int flags, struct file **filpp) { struct nfs4_stateid *stp = NULL; struct nfs4_delegation *dp = NULL; + struct svc_fh *current_fh = &cstate->current_fh; struct inode *ino = current_fh->fh_dentry->d_inode; __be32 status; @@ -2878,10 +2880,14 @@ setlkflg (int type) * Checks for sequence id mutating operations. */ static __be32 -nfs4_preprocess_seqid_op(struct svc_fh *current_fh, u32 seqid, stateid_t *stateid, int flags, struct nfs4_stateowner **sopp, struct nfs4_stateid **stpp, struct nfsd4_lock *lock) +nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid, + stateid_t *stateid, int flags, + struct nfs4_stateowner **sopp, + struct nfs4_stateid **stpp, struct nfsd4_lock *lock) { struct nfs4_stateid *stp; struct nfs4_stateowner *sop; + struct svc_fh *current_fh = &cstate->current_fh; __be32 status; dprintk("NFSD: preprocess_seqid_op: seqid=%d " @@ -3004,7 +3010,7 @@ nfsd4_open_confirm(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nfs4_lock_state(); - if ((status = nfs4_preprocess_seqid_op(&cstate->current_fh, + if ((status = nfs4_preprocess_seqid_op(cstate, oc->oc_seqid, &oc->oc_req_stateid, CONFIRM | OPEN_STATE, &oc->oc_stateowner, &stp, NULL))) @@ -3074,7 +3080,7 @@ nfsd4_open_downgrade(struct svc_rqst *rqstp, return nfserr_inval; nfs4_lock_state(); - if ((status = nfs4_preprocess_seqid_op(&cstate->current_fh, + if ((status = nfs4_preprocess_seqid_op(cstate, od->od_seqid, &od->od_stateid, OPEN_STATE, @@ -3127,7 +3133,7 @@ nfsd4_close(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nfs4_lock_state(); /* check close_lru for replay */ - if ((status = nfs4_preprocess_seqid_op(&cstate->current_fh, + if ((status = nfs4_preprocess_seqid_op(cstate, close->cl_seqid, &close->cl_stateid, OPEN_STATE | CLOSE_STATE, @@ -3474,7 +3480,7 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, goto out; /* validate and update open stateid and open seqid */ - status = nfs4_preprocess_seqid_op(&cstate->current_fh, + status = nfs4_preprocess_seqid_op(cstate, lock->lk_new_open_seqid, &lock->lk_new_open_stateid, OPEN_STATE, @@ -3501,7 +3507,7 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, goto out; } else { /* lock (lock owner + lock stateid) already exists */ - status = nfs4_preprocess_seqid_op(&cstate->current_fh, + status = nfs4_preprocess_seqid_op(cstate, lock->lk_old_lock_seqid, &lock->lk_old_lock_stateid, LOCK_STATE, @@ -3697,7 +3703,7 @@ nfsd4_locku(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nfs4_lock_state(); - if ((status = nfs4_preprocess_seqid_op(&cstate->current_fh, + if ((status = nfs4_preprocess_seqid_op(cstate, locku->lu_seqid, &locku->lu_stateid, LOCK_STATE, -- cgit v1.2.3 From 6668958fac1d05f55420de702f3678d46c1e93a5 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:45 +0300 Subject: nfsd41: stateid handling When sessions are used, stateful operation sequenceid and stateid handling are not used. When sessions are used, on the first open set the seqid to 1, mark state confirmed and skip seqid processing. When sessionas are used the stateid generation number is ignored when it is zero whereas without sessions bad_stateid or stale stateid is returned. Add flags to propagate session use to all stateful ops and down to check_stateid_generation. Signed-off-by: Benny Halevy Signed-off-by: Andy Adamson [nfsd4_has_session should return a boolean, not u32] Signed-off-by: Benny Halevy [nfsd41: pass nfsd4_compoundres * to nfsd4_process_open1] [nfsd41: calculate HAS_SESSION in nfs4_preprocess_stateid_op] [nfsd41: calculate HAS_SESSION in nfs4_preprocess_seqid_op] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 5 ++++- fs/nfsd/nfs4state.c | 47 +++++++++++++++++++++++++++++++++++++++-------- fs/nfsd/nfs4xdr.c | 2 +- 3 files changed, 44 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index e6a0f314fdbe..aeb4888bdfad 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -168,6 +168,8 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_open *open) { __be32 status; + struct nfsd4_compoundres *resp; + dprintk("NFSD: nfsd4_open filename %.*s op_stateowner %p\n", (int)open->op_fname.len, open->op_fname.data, open->op_stateowner); @@ -179,7 +181,8 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nfs4_lock_state(); /* check seqid for replay. set nfs4_owner */ - status = nfsd4_process_open1(open); + resp = rqstp->rq_resp; + status = nfsd4_process_open1(&resp->cstate, open); if (status == nfserr_replay_me) { struct nfs4_replay *rp = &open->op_stateowner->so_replay; fh_put(&cstate->current_fh); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index f49f305096ec..bbaf3c9bfe5a 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2183,7 +2183,8 @@ static struct lock_manager_operations nfsd_lease_mng_ops = { __be32 -nfsd4_process_open1(struct nfsd4_open *open) +nfsd4_process_open1(struct nfsd4_compound_state *cstate, + struct nfsd4_open *open) { clientid_t *clientid = &open->op_clientid; struct nfs4_client *clp = NULL; @@ -2206,6 +2207,9 @@ nfsd4_process_open1(struct nfsd4_open *open) return nfserr_expired; goto renew; } + /* When sessions are used, skip open sequenceid processing */ + if (nfsd4_has_session(cstate)) + goto renew; if (!sop->so_confirmed) { /* Replace unconfirmed owners without checking for replay. */ clp = sop->so_client; @@ -2483,6 +2487,7 @@ out: __be32 nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_open *open) { + struct nfsd4_compoundres *resp = rqstp->rq_resp; struct nfs4_file *fp = NULL; struct inode *ino = current_fh->fh_dentry->d_inode; struct nfs4_stateid *stp = NULL; @@ -2541,9 +2546,14 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf release_open_stateid(stp); goto out; } + if (nfsd4_has_session(&resp->cstate)) + update_stateid(&stp->st_stateid); } memcpy(&open->op_stateid, &stp->st_stateid, sizeof(stateid_t)); + if (nfsd4_has_session(&resp->cstate)) + open->op_stateowner->so_confirmed = 1; + /* * Attempt to hand out a delegation. No error return, because the * OPEN succeeds even if we fail. @@ -2564,7 +2574,8 @@ out: * To finish the open response, we just need to set the rflags. */ open->op_rflags = NFS4_OPEN_RESULT_LOCKTYPE_POSIX; - if (!open->op_stateowner->so_confirmed) + if (!open->op_stateowner->so_confirmed && + !nfsd4_has_session(&resp->cstate)) open->op_rflags |= NFS4_OPEN_RESULT_CONFIRM; return status; @@ -2781,8 +2792,15 @@ grace_disallows_io(struct inode *inode) return locks_in_grace() && mandatory_lock(inode); } -static int check_stateid_generation(stateid_t *in, stateid_t *ref) +static int check_stateid_generation(stateid_t *in, stateid_t *ref, int flags) { + /* + * When sessions are used the stateid generation number is ignored + * when it is zero. + */ + if ((flags & HAS_SESSION) && in->si_generation == 0) + goto out; + /* If the client sends us a stateid from the future, it's buggy: */ if (in->si_generation > ref->si_generation) return nfserr_bad_stateid; @@ -2798,6 +2816,7 @@ static int check_stateid_generation(stateid_t *in, stateid_t *ref) */ if (in->si_generation < ref->si_generation) return nfserr_old_stateid; +out: return nfs_ok; } @@ -2825,6 +2844,9 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, if (grace_disallows_io(ino)) return nfserr_grace; + if (nfsd4_has_session(cstate)) + flags |= HAS_SESSION; + if (ZERO_STATEID(stateid) || ONE_STATEID(stateid)) return check_special_stateids(current_fh, stateid, flags); @@ -2837,7 +2859,8 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, dp = find_delegation_stateid(ino, stateid); if (!dp) goto out; - status = check_stateid_generation(stateid, &dp->dl_stateid); + status = check_stateid_generation(stateid, &dp->dl_stateid, + flags); if (status) goto out; status = nfs4_check_delegmode(dp, flags); @@ -2854,7 +2877,8 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, goto out; if (!stp->st_stateowner->so_confirmed) goto out; - status = check_stateid_generation(stateid, &stp->st_stateid); + status = check_stateid_generation(stateid, &stp->st_stateid, + flags); if (status) goto out; status = nfs4_check_openmode(stp, flags); @@ -2905,6 +2929,10 @@ nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid, if (STALE_STATEID(stateid)) return nfserr_stale_stateid; + + if (nfsd4_has_session(cstate)) + flags |= HAS_SESSION; + /* * We return BAD_STATEID if filehandle doesn't match stateid, * the confirmed flag is incorrecly set, or the generation @@ -2961,7 +2989,7 @@ nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid, * For the moment, we ignore the possibility of * generation number wraparound. */ - if (seqid != sop->so_seqid) + if (!(flags & HAS_SESSION) && seqid != sop->so_seqid) goto check_replay; if (sop->so_confirmed && flags & CONFIRM) { @@ -2974,7 +3002,7 @@ nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid, " confirmed yet!\n"); return nfserr_bad_stateid; } - status = check_stateid_generation(stateid, &stp->st_stateid); + status = check_stateid_generation(stateid, &stp->st_stateid, flags); if (status) return status; renew_client(sop->so_client); @@ -3169,11 +3197,14 @@ nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stateid_t *stateid = &dr->dr_stateid; struct inode *inode; __be32 status; + int flags = 0; if ((status = fh_verify(rqstp, &cstate->current_fh, S_IFREG, 0))) return status; inode = cstate->current_fh.fh_dentry->d_inode; + if (nfsd4_has_session(cstate)) + flags |= HAS_SESSION; nfs4_lock_state(); status = nfserr_bad_stateid; if (ZERO_STATEID(stateid) || ONE_STATEID(stateid)) @@ -3187,7 +3218,7 @@ nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, dp = find_delegation_stateid(inode, stateid); if (!dp) goto out; - status = check_stateid_generation(stateid, &dp->dl_stateid); + status = check_stateid_generation(stateid, &dp->dl_stateid, flags); if (status) goto out; renew_client(dp->dl_client); diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index c6a726d04efd..dd81ac1a1c6c 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3194,7 +3194,7 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo iov = &rqstp->rq_res.head[0]; iov->iov_len = ((char*)resp->p) - (char*)iov->iov_base; BUG_ON(iov->iov_len > PAGE_SIZE); - if (resp->cstate.slot != NULL) { + if (nfsd4_has_session(&resp->cstate)) { if (resp->cstate.status == nfserr_replay_cache && !nfsd4_not_cached(resp)) { iov->iov_len = resp->cstate.iovlen; -- cgit v1.2.3 From 496c262cf01106a546ffb7df6fea84b8b881ee19 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:48 +0300 Subject: nfsd41: check encode size for sessions maxresponse cached Calculate the space the compound response has taken after encoding the current operation. pad: add on 8 bytes for the next operation's op_code and status so that there is room to cache a failure on the next operation. Compare this length to the session se_fmaxresp_cached and return nfserr_rep_too_big_to_cache if the length is too large. Our se_fmaxresp_cached will always be a multiple of PAGE_SIZE, and so will be at least a page and will therefore hold the xdr_buf head. Signed-off-by: Andy Adamson [nfsd41: non-page DRC for solo sequence responses] [fixed nfsd4_check_drc_limit cosmetics] Signed-off-by: Benny Halevy [nfsd41: use cstate session in nfsd4_check_drc_limit] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 51 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 51 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index dd81ac1a1c6c..7fdee828f44e 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3078,6 +3078,54 @@ static nfsd4_enc nfsd4_enc_ops[] = { [OP_RECLAIM_COMPLETE] = (nfsd4_enc)nfsd4_encode_noop, }; +/* + * Calculate the total amount of memory that the compound response has taken + * after encoding the current operation. + * + * pad: add on 8 bytes for the next operation's op_code and status so that + * there is room to cache a failure on the next operation. + * + * Compare this length to the session se_fmaxresp_cached. + * + * Our se_fmaxresp_cached will always be a multiple of PAGE_SIZE, and so + * will be at least a page and will therefore hold the xdr_buf head. + */ +static int nfsd4_check_drc_limit(struct nfsd4_compoundres *resp) +{ + int status = 0; + struct xdr_buf *xb = &resp->rqstp->rq_res; + struct nfsd4_compoundargs *args = resp->rqstp->rq_argp; + struct nfsd4_session *session = NULL; + struct nfsd4_slot *slot = resp->cstate.slot; + u32 length, tlen = 0, pad = 8; + + if (!nfsd4_has_session(&resp->cstate)) + return status; + + session = resp->cstate.session; + if (session == NULL || slot->sl_cache_entry.ce_cachethis == 0) + return status; + + if (resp->opcnt >= args->opcnt) + pad = 0; /* this is the last operation */ + + if (xb->page_len == 0) { + length = (char *)resp->p - (char *)xb->head[0].iov_base + pad; + } else { + if (xb->tail[0].iov_base && xb->tail[0].iov_len > 0) + tlen = (char *)resp->p - (char *)xb->tail[0].iov_base; + + length = xb->head[0].iov_len + xb->page_len + tlen + pad; + } + dprintk("%s length %u, xb->page_len %u tlen %u pad %u\n", __func__, + length, xb->page_len, tlen, pad); + + if (length <= session->se_fmaxresp_cached) + return status; + else + return nfserr_rep_too_big_to_cache; +} + void nfsd4_encode_operation(struct nfsd4_compoundres *resp, struct nfsd4_op *op) { @@ -3094,6 +3142,9 @@ nfsd4_encode_operation(struct nfsd4_compoundres *resp, struct nfsd4_op *op) BUG_ON(op->opnum < 0 || op->opnum >= ARRAY_SIZE(nfsd4_enc_ops) || !nfsd4_enc_ops[op->opnum]); op->status = nfsd4_enc_ops[op->opnum](resp, op->status, &op->u); + /* nfsd4_check_drc_limit guarantees enough room for error status */ + if (!op->status && nfsd4_check_drc_limit(resp)) + op->status = nfserr_rep_too_big_to_cache; status: /* * Note: We write the status directly, instead of using WRITE32(), -- cgit v1.2.3 From 60adfc50de3855628dea8f8896a65f471f51301c Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:50 +0300 Subject: nfsd41: clientid handling Extract the clientid from sessionid to set the op_clientid on open. Verify that the clid for other stateful ops is zero for minorversion != 0 Do all other checks for stateful ops without sessions. Signed-off-by: Benny Halevy Signed-off-by: Andy Adamson [fixed whitespace indent] Signed-off-by: Benny Halevy [nfsd41 remove sl_session from nfsd4_open] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 12 ++++++++++++ fs/nfsd/nfs4state.c | 14 ++++++++------ fs/nfsd/nfs4xdr.c | 9 +++++++++ 3 files changed, 29 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index aeb4888bdfad..3fdcca53212a 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -162,6 +162,15 @@ do_open_fhandle(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_ return status; } +static void +copy_clientid(clientid_t *clid, struct nfsd4_session *session) +{ + struct nfsd4_sessionid *sid = + (struct nfsd4_sessionid *)session->se_sessionid.data; + + clid->cl_boot = sid->clientid.cl_boot; + clid->cl_id = sid->clientid.cl_id; +} static __be32 nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, @@ -178,6 +187,9 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (open->op_create && open->op_claim_type != NFS4_OPEN_CLAIM_NULL) return nfserr_inval; + if (nfsd4_has_session(cstate)) + copy_clientid(&open->op_clientid, cstate->session); + nfs4_lock_state(); /* check seqid for replay. set nfs4_owner */ diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index bbaf3c9bfe5a..d227f85b5ed2 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -603,8 +603,8 @@ STALE_CLIENTID(clientid_t *clid) { if (clid->cl_boot == boot_time) return 0; - dprintk("NFSD stale clientid (%08x/%08x)\n", - clid->cl_boot, clid->cl_id); + dprintk("NFSD stale clientid (%08x/%08x) boot_time %08lx\n", + clid->cl_boot, clid->cl_id, boot_time); return 1; } @@ -2965,8 +2965,9 @@ nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid, if (lock->lk_is_new) { if (!sop->so_is_open_owner) return nfserr_bad_stateid; - if (!same_clid(&clp->cl_clientid, lockclid)) - return nfserr_bad_stateid; + if (!(flags & HAS_SESSION) && + !same_clid(&clp->cl_clientid, lockclid)) + return nfserr_bad_stateid; /* stp is the open stateid */ status = nfs4_check_openmode(stp, lkflg); if (status) @@ -3507,7 +3508,8 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfs4_file *fp; status = nfserr_stale_clientid; - if (STALE_CLIENTID(&lock->lk_new_clientid)) + if (!nfsd4_has_session(cstate) && + STALE_CLIENTID(&lock->lk_new_clientid)) goto out; /* validate and update open stateid and open seqid */ @@ -3661,7 +3663,7 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nfs4_lock_state(); status = nfserr_stale_clientid; - if (STALE_CLIENTID(&lockt->lt_clientid)) + if (!nfsd4_has_session(cstate) && STALE_CLIENTID(&lockt->lt_clientid)) goto out; if ((status = fh_verify(rqstp, &cstate->current_fh, S_IFREG, 0))) { diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 7fdee828f44e..059d4aad39b8 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -189,6 +189,11 @@ static __be32 *read_buf(struct nfsd4_compoundargs *argp, u32 nbytes) return p; } +static int zero_clientid(clientid_t *clid) +{ + return (clid->cl_boot == 0) && (clid->cl_id == 0); +} + static int defer_free(struct nfsd4_compoundargs *argp, void (*release)(const void *), void *p) @@ -584,6 +589,8 @@ nfsd4_decode_lockt(struct nfsd4_compoundargs *argp, struct nfsd4_lockt *lockt) READ_BUF(lockt->lt_owner.len); READMEM(lockt->lt_owner.data, lockt->lt_owner.len); + if (argp->minorversion && !zero_clientid(&lockt->lt_clientid)) + return nfserr_inval; DECODE_TAIL; } @@ -994,6 +1001,8 @@ nfsd4_decode_release_lockowner(struct nfsd4_compoundargs *argp, struct nfsd4_rel READ_BUF(rlockowner->rl_owner.len); READMEM(rlockowner->rl_owner.data, rlockowner->rl_owner.len); + if (argp->minorversion && !zero_clientid(&rlockowner->rl_clientid)) + return nfserr_inval; DECODE_TAIL; } -- cgit v1.2.3 From d87a8ade95288f28c729e076cd74929f3f199b6c Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:53 +0300 Subject: nfsd41: access_valid For nfs41, the open share flags are used also for delegation "wants" and "signals". Check that they are valid. Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 1 + fs/nfsd/nfs4state.c | 20 +++++++++++++++----- 2 files changed, 16 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 3fdcca53212a..db208dd8fdca 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -910,6 +910,7 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, resp->tag = args->tag; resp->opcnt = 0; resp->rqstp = rqstp; + resp->cstate.minorversion = args->minorversion; resp->cstate.replay_owner = NULL; fh_init(&resp->cstate.current_fh, NFS4_FHSIZE); fh_init(&resp->cstate.save_fh, NFS4_FHSIZE); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index d227f85b5ed2..374aaf3808e8 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1943,11 +1943,21 @@ find_file(struct inode *ino) return NULL; } -static inline int access_valid(u32 x) +static inline int access_valid(u32 x, u32 minorversion) { - if (x < NFS4_SHARE_ACCESS_READ) + if ((x & NFS4_SHARE_ACCESS_MASK) < NFS4_SHARE_ACCESS_READ) return 0; - if (x > NFS4_SHARE_ACCESS_BOTH) + if ((x & NFS4_SHARE_ACCESS_MASK) > NFS4_SHARE_ACCESS_BOTH) + return 0; + x &= ~NFS4_SHARE_ACCESS_MASK; + if (minorversion && x) { + if ((x & NFS4_SHARE_WANT_MASK) > NFS4_SHARE_WANT_CANCEL) + return 0; + if ((x & NFS4_SHARE_WHEN_MASK) > NFS4_SHARE_PUSH_DELEG_WHEN_UNCONTENDED) + return 0; + x &= ~(NFS4_SHARE_WANT_MASK | NFS4_SHARE_WHEN_MASK); + } + if (x) return 0; return 1; } @@ -2495,7 +2505,7 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf __be32 status; status = nfserr_inval; - if (!access_valid(open->op_share_access) + if (!access_valid(open->op_share_access, resp->cstate.minorversion) || !deny_valid(open->op_share_deny)) goto out; /* @@ -3104,7 +3114,7 @@ nfsd4_open_downgrade(struct svc_rqst *rqstp, (int)cstate->current_fh.fh_dentry->d_name.len, cstate->current_fh.fh_dentry->d_name.name); - if (!access_valid(od->od_share_access) + if (!access_valid(od->od_share_access, cstate->minorversion) || !deny_valid(od->od_share_deny)) return nfserr_inval; -- cgit v1.2.3 From 84459a1162801fb84734e5f9e6dc5194f791d69b Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:28:56 +0300 Subject: nfsd41: add OPEN4_SHARE_ACCESS_WANT nfs4_stateid bmap Separate the access bits from the want bits and enable __set_bit to work correctly with st_access_bmap. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 374aaf3808e8..c65a27b76a9d 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1889,7 +1889,8 @@ init_stateid(struct nfs4_stateid *stp, struct nfs4_file *fp, struct nfsd4_open * stp->st_stateid.si_generation = 0; stp->st_access_bmap = 0; stp->st_deny_bmap = 0; - __set_bit(open->op_share_access, &stp->st_access_bmap); + __set_bit(open->op_share_access & ~NFS4_SHARE_WANT_MASK, + &stp->st_access_bmap); __set_bit(open->op_share_deny, &stp->st_deny_bmap); stp->st_openstp = NULL; } -- cgit v1.2.3 From 8daf220a6a83c47b9648c28bb819c14c60bad7f9 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 3 Apr 2009 08:28:59 +0300 Subject: nfsd41: control nfsv4.1 svc via /proc/fs/nfsd/versions Support enabling and disabling nfsv4.1 via /proc/fs/nfsd/versions by writing the strings "+4.1" or "-4.1" correspondingly. Use user mode nfs-utils (rpc.nfsd option) to enable. This will allow us to get rid of CONFIG_NFSD_V4_1 [nfsd41: disable support for minorversion by default] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 2 +- fs/nfsd/nfsctl.c | 26 +++++++++++++++++++++++--- fs/nfsd/nfssvc.c | 24 ++++++++++++++++++++++++ 3 files changed, 48 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index db208dd8fdca..52bb14debb30 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -921,7 +921,7 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, * According to RFC3010, this takes precedence over all other errors. */ status = nfserr_minor_vers_mismatch; - if (args->minorversion > NFSD_SUPPORTED_MINOR_VERSION) + if (args->minorversion > nfsd_supported_minorversion) goto out; if (!nfs41_op_ordering_ok(args)) { diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 4adebb6312c4..a9b8c75bf0bf 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -792,8 +792,9 @@ out_free: static ssize_t __write_versions(struct file *file, char *buf, size_t size) { char *mesg = buf; - char *vers, sign; + char *vers, *minorp, sign; int len, num; + unsigned minor; ssize_t tlen = 0; char *sep; @@ -814,9 +815,20 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size) do { sign = *vers; if (sign == '+' || sign == '-') - num = simple_strtol((vers+1), NULL, 0); + num = simple_strtol((vers+1), &minorp, 0); else - num = simple_strtol(vers, NULL, 0); + num = simple_strtol(vers, &minorp, 0); + if (*minorp == '.') { + if (num < 4) + return -EINVAL; + minor = simple_strtoul(minorp+1, NULL, 0); + if (minor == 0) + return -EINVAL; + if (nfsd_minorversion(minor, sign == '-' ? + NFSD_CLEAR : NFSD_SET) < 0) + return -EINVAL; + goto next; + } switch(num) { case 2: case 3: @@ -826,6 +838,7 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size) default: return -EINVAL; } + next: vers += len + 1; tlen += len; } while ((len = qword_get(&mesg, vers, size)) > 0); @@ -844,6 +857,13 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size) num); sep = " "; } + if (nfsd_vers(4, NFSD_AVAIL)) + for (minor = 1; minor <= NFSD_SUPPORTED_MINOR_VERSION; minor++) + len += sprintf(buf+len, " %c4.%u", + (nfsd_vers(4, NFSD_TEST) && + nfsd_minorversion(minor, NFSD_TEST)) ? + '+' : '-', + minor); len += sprintf(buf+len, "\n"); return len; } diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index b53a098e97a4..e9d57734a348 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -121,6 +121,8 @@ struct svc_program nfsd_program = { }; +u32 nfsd_supported_minorversion; + int nfsd_vers(int vers, enum vers_op change) { if (vers < NFSD_MINVERS || vers >= NFSD_NRVERS) @@ -147,6 +149,28 @@ int nfsd_vers(int vers, enum vers_op change) } return 0; } + +int nfsd_minorversion(u32 minorversion, enum vers_op change) +{ + if (minorversion > NFSD_SUPPORTED_MINOR_VERSION) + return -1; + switch(change) { + case NFSD_SET: + nfsd_supported_minorversion = minorversion; + break; + case NFSD_CLEAR: + if (minorversion == 0) + return -1; + nfsd_supported_minorversion = minorversion - 1; + break; + case NFSD_TEST: + return minorversion <= nfsd_supported_minorversion; + case NFSD_AVAIL: + return minorversion <= NFSD_SUPPORTED_MINOR_VERSION; + } + return 0; +} + /* * Maximum number of nfsd processes */ -- cgit v1.2.3 From c0d6fc8a2d55a8235c301aeb6d5254d5992895d1 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 3 Apr 2009 08:29:05 +0300 Subject: nfsd41: pass writable attrs mask to nfsd4_decode_fattr In preparation for EXCLUSIVE4_1 Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 059d4aad39b8..42f9fb661006 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -251,9 +251,14 @@ nfsd4_decode_bitmap(struct nfsd4_compoundargs *argp, u32 *bmval) DECODE_TAIL; } +static u32 nfsd_attrmask[] = { + NFSD_WRITEABLE_ATTRS_WORD0, + NFSD_WRITEABLE_ATTRS_WORD1 +}; + static __be32 -nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, struct iattr *iattr, - struct nfs4_acl **acl) +nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, u32 *writable, + struct iattr *iattr, struct nfs4_acl **acl) { int expected_len, len = 0; u32 dummy32; @@ -271,7 +276,7 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, struct iattr *ia */ if ((bmval[0] & ~NFSD_SUPPORTED_ATTRS_WORD0) || (bmval[1] & ~NFSD_SUPPORTED_ATTRS_WORD1)) return nfserr_attrnotsupp; - if ((bmval[0] & ~NFSD_WRITEABLE_ATTRS_WORD0) || (bmval[1] & ~NFSD_WRITEABLE_ATTRS_WORD1)) + if ((bmval[0] & ~writable[0]) || (bmval[1] & ~writable[1])) return nfserr_inval; READ_BUF(4); @@ -499,7 +504,9 @@ nfsd4_decode_create(struct nfsd4_compoundargs *argp, struct nfsd4_create *create if ((status = check_filename(create->cr_name, create->cr_namelen, nfserr_inval))) return status; - if ((status = nfsd4_decode_fattr(argp, create->cr_bmval, &create->cr_iattr, &create->cr_acl))) + status = nfsd4_decode_fattr(argp, create->cr_bmval, nfsd_attrmask, + &create->cr_iattr, &create->cr_acl); + if (status) goto out; DECODE_TAIL; @@ -660,7 +667,9 @@ nfsd4_decode_open(struct nfsd4_compoundargs *argp, struct nfsd4_open *open) switch (open->op_createmode) { case NFS4_CREATE_UNCHECKED: case NFS4_CREATE_GUARDED: - if ((status = nfsd4_decode_fattr(argp, open->op_bmval, &open->op_iattr, &open->op_acl))) + status = nfsd4_decode_fattr(argp, open->op_bmval, + nfsd_attrmask, &open->op_iattr, &open->op_acl); + if (status) goto out; break; case NFS4_CREATE_EXCLUSIVE: @@ -859,7 +868,7 @@ nfsd4_decode_setattr(struct nfsd4_compoundargs *argp, struct nfsd4_setattr *seta status = nfsd4_decode_stateid(argp, &setattr->sa_stateid); if (status) return status; - return nfsd4_decode_fattr(argp, setattr->sa_bmval, + return nfsd4_decode_fattr(argp, setattr->sa_bmval, nfsd_attrmask, &setattr->sa_iattr, &setattr->sa_acl); } -- cgit v1.2.3 From 95ec28cda323104bbff64fc7ec8ee4c9042e51fa Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 3 Apr 2009 08:29:08 +0300 Subject: nfsd: dynamically skip encoded fattr bitmap in _nfsd4_verify _nfsd4_verify currently skips 3 words from the encoded buffer begining. With support for 3-word attr bitmaps in nfsd41, nfsd4_encode_fattr may encode 1, 2, or 3 words, and not always 2 as it used to be, hence we need to find out where to skip using the encoded bitmap length. Note: This patch may be applied over pre-nfsd41 nfsd. Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 52bb14debb30..937853f62834 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -775,7 +775,8 @@ _nfsd4_verify(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (status) goto out_kfree; - p = buf + 3; + /* skip bitmap */ + p = buf + 1 + ntohl(buf[0]); status = nfserr_not_same; if (ntohl(*p++) != verify->ve_attrlen) goto out_kfree; -- cgit v1.2.3 From 7e70570647827345352cf6c17461c9fa166f570a Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 3 Apr 2009 08:29:11 +0300 Subject: nfsd41: support for 3-word long attribute bitmask Also, use client minorversion to generate supported attrs Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 15 ++++++++------ fs/nfsd/nfs4xdr.c | 59 +++++++++++++++++++++++++++++++++++++++++++----------- 2 files changed, 56 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 937853f62834..e206053433e3 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -463,8 +463,9 @@ nfsd4_getattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (getattr->ga_bmval[1] & NFSD_WRITEONLY_ATTRS_WORD1) return nfserr_inval; - getattr->ga_bmval[0] &= NFSD_SUPPORTED_ATTRS_WORD0; - getattr->ga_bmval[1] &= NFSD_SUPPORTED_ATTRS_WORD1; + getattr->ga_bmval[0] &= nfsd_suppattrs0(cstate->minorversion); + getattr->ga_bmval[1] &= nfsd_suppattrs1(cstate->minorversion); + getattr->ga_bmval[2] &= nfsd_suppattrs2(cstate->minorversion); getattr->ga_fhp = &cstate->current_fh; return nfs_ok; @@ -555,8 +556,9 @@ nfsd4_readdir(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (readdir->rd_bmval[1] & NFSD_WRITEONLY_ATTRS_WORD1) return nfserr_inval; - readdir->rd_bmval[0] &= NFSD_SUPPORTED_ATTRS_WORD0; - readdir->rd_bmval[1] &= NFSD_SUPPORTED_ATTRS_WORD1; + readdir->rd_bmval[0] &= nfsd_suppattrs0(cstate->minorversion); + readdir->rd_bmval[1] &= nfsd_suppattrs1(cstate->minorversion); + readdir->rd_bmval[2] &= nfsd_suppattrs2(cstate->minorversion); if ((cookie > ~(u32)0) || (cookie == 1) || (cookie == 2) || (cookie == 0 && memcmp(readdir->rd_verf.data, zeroverf.data, NFS4_VERIFIER_SIZE))) @@ -746,8 +748,9 @@ _nfsd4_verify(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (status) return status; - if ((verify->ve_bmval[0] & ~NFSD_SUPPORTED_ATTRS_WORD0) - || (verify->ve_bmval[1] & ~NFSD_SUPPORTED_ATTRS_WORD1)) + if ((verify->ve_bmval[0] & ~nfsd_suppattrs0(cstate->minorversion)) + || (verify->ve_bmval[1] & ~nfsd_suppattrs1(cstate->minorversion)) + || (verify->ve_bmval[2] & ~nfsd_suppattrs2(cstate->minorversion))) return nfserr_attrnotsupp; if ((verify->ve_bmval[0] & FATTR4_WORD0_RDATTR_ERROR) || (verify->ve_bmval[1] & NFSD_WRITEONLY_ATTRS_WORD1)) diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 42f9fb661006..70296fedee41 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -236,6 +236,7 @@ nfsd4_decode_bitmap(struct nfsd4_compoundargs *argp, u32 *bmval) bmval[0] = 0; bmval[1] = 0; + bmval[2] = 0; READ_BUF(4); READ32(bmlen); @@ -247,13 +248,16 @@ nfsd4_decode_bitmap(struct nfsd4_compoundargs *argp, u32 *bmval) READ32(bmval[0]); if (bmlen > 1) READ32(bmval[1]); + if (bmlen > 2) + READ32(bmval[2]); DECODE_TAIL; } static u32 nfsd_attrmask[] = { NFSD_WRITEABLE_ATTRS_WORD0, - NFSD_WRITEABLE_ATTRS_WORD1 + NFSD_WRITEABLE_ATTRS_WORD1, + NFSD_WRITEABLE_ATTRS_WORD2 }; static __be32 @@ -274,9 +278,12 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, u32 *writable, * According to spec, unsupported attributes return ERR_ATTRNOTSUPP; * read-only attributes return ERR_INVAL. */ - if ((bmval[0] & ~NFSD_SUPPORTED_ATTRS_WORD0) || (bmval[1] & ~NFSD_SUPPORTED_ATTRS_WORD1)) + if ((bmval[0] & ~nfsd_suppattrs0(argp->minorversion)) || + (bmval[1] & ~nfsd_suppattrs1(argp->minorversion)) || + (bmval[2] & ~nfsd_suppattrs2(argp->minorversion))) return nfserr_attrnotsupp; - if ((bmval[0] & ~writable[0]) || (bmval[1] & ~writable[1])) + if ((bmval[0] & ~writable[0]) || (bmval[1] & ~writable[1]) || + (bmval[2] & ~writable[2])) return nfserr_inval; READ_BUF(4); @@ -411,6 +418,7 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, u32 *writable, goto xdr_error; } } + BUG_ON(bmval[2]); /* no such writeable attr supported yet */ if (len != expected_len) goto xdr_error; @@ -1726,6 +1734,7 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, { u32 bmval0 = bmval[0]; u32 bmval1 = bmval[1]; + u32 bmval2 = bmval[2]; struct kstat stat; struct svc_fh tempfh; struct kstatfs statfs; @@ -1739,12 +1748,16 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, int err; int aclsupport = 0; struct nfs4_acl *acl = NULL; + struct nfsd4_compoundres *resp = rqstp->rq_resp; + u32 minorversion = resp->cstate.minorversion; BUG_ON(bmval1 & NFSD_WRITEONLY_ATTRS_WORD1); - BUG_ON(bmval0 & ~NFSD_SUPPORTED_ATTRS_WORD0); - BUG_ON(bmval1 & ~NFSD_SUPPORTED_ATTRS_WORD1); + BUG_ON(bmval0 & ~nfsd_suppattrs0(minorversion)); + BUG_ON(bmval1 & ~nfsd_suppattrs1(minorversion)); + BUG_ON(bmval2 & ~nfsd_suppattrs2(minorversion)); if (exp->ex_fslocs.migrated) { + BUG_ON(bmval[2]); status = fattr_handle_absent_fs(&bmval0, &bmval1, &rdattr_err); if (status) goto out; @@ -1790,22 +1803,42 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, if ((buflen -= 16) < 0) goto out_resource; - WRITE32(2); - WRITE32(bmval0); - WRITE32(bmval1); + if (unlikely(bmval2)) { + WRITE32(3); + WRITE32(bmval0); + WRITE32(bmval1); + WRITE32(bmval2); + } else if (likely(bmval1)) { + WRITE32(2); + WRITE32(bmval0); + WRITE32(bmval1); + } else { + WRITE32(1); + WRITE32(bmval0); + } attrlenp = p++; /* to be backfilled later */ if (bmval0 & FATTR4_WORD0_SUPPORTED_ATTRS) { - u32 word0 = NFSD_SUPPORTED_ATTRS_WORD0; + u32 word0 = nfsd_suppattrs0(minorversion); + u32 word1 = nfsd_suppattrs1(minorversion); + u32 word2 = nfsd_suppattrs2(minorversion); + if ((buflen -= 12) < 0) goto out_resource; if (!aclsupport) word0 &= ~FATTR4_WORD0_ACL; if (!exp->ex_fslocs.locations) word0 &= ~FATTR4_WORD0_FS_LOCATIONS; - WRITE32(2); - WRITE32(word0); - WRITE32(NFSD_SUPPORTED_ATTRS_WORD1); + if (!word2) { + WRITE32(2); + WRITE32(word0); + WRITE32(word1); + } else { + WRITE32(3); + WRITE32(word0); + WRITE32(word1); + WRITE32(word2); + } } if (bmval0 & FATTR4_WORD0_TYPE) { if ((buflen -= 4) < 0) @@ -2115,6 +2148,8 @@ out_acl: } WRITE64(stat.ino); } + BUG_ON(bmval2); /* FIXME: not implemented yet */ + *attrlenp = htonl((char *)p - (char *)attrlenp - 4); *countp = p - buffer; status = nfs_ok; -- cgit v1.2.3 From 8c18f2052e756e7d5dea712fc6e7ed70c00e8a39 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 3 Apr 2009 08:29:14 +0300 Subject: nfsd41: SUPPATTR_EXCLCREAT attribute Return bitmask for supported EXCLUSIVE4_1 create attributes. Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 70296fedee41..533d14fec99e 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -2148,7 +2148,12 @@ out_acl: } WRITE64(stat.ino); } - BUG_ON(bmval2); /* FIXME: not implemented yet */ + if (bmval2 & FATTR4_WORD2_SUPPATTR_EXCLCREAT) { + WRITE32(3); + WRITE32(NFSD_SUPPATTR_EXCLCREAT_WORD0); + WRITE32(NFSD_SUPPATTR_EXCLCREAT_WORD1); + WRITE32(NFSD_SUPPATTR_EXCLCREAT_WORD2); + } *attrlenp = htonl((char *)p - (char *)attrlenp - 4); *countp = p - buffer; -- cgit v1.2.3 From 79fb54abd285b442e1f30f851902f3ddf58e7704 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Fri, 3 Apr 2009 08:29:17 +0300 Subject: nfsd41: CREATE_EXCLUSIVE4_1 Implement the CREATE_EXCLUSIVE4_1 open mode conforming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 This mode allows the client to atomically create a file if it doesn't exist while setting some of its attributes. It must be implemented if the server supports persistent reply cache and/or pnfs. Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 15 +++++++++++++++ fs/nfsd/nfs4xdr.c | 17 +++++++++++++++++ 2 files changed, 32 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index e206053433e3..b2883e9c6381 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -93,6 +93,21 @@ do_open_lookup(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_o open->op_truncate = 0; if (open->op_create) { + /* FIXME: check session persistence and pnfs flags. + * The nfsv4.1 spec requires the following semantics: + * + * Persistent | pNFS | Server REQUIRED | Client Allowed + * Reply Cache | server | | + * -------------+--------+-----------------+-------------------- + * no | no | EXCLUSIVE4_1 | EXCLUSIVE4_1 + * | | | (SHOULD) + * | | and EXCLUSIVE4 | or EXCLUSIVE4 + * | | | (SHOULD NOT) + * no | yes | EXCLUSIVE4_1 | EXCLUSIVE4_1 + * yes | no | GUARDED4 | GUARDED4 + * yes | yes | GUARDED4 | GUARDED4 + */ + /* * Note: create modes (UNCHECKED,GUARDED...) are the same * in NFSv4 as in v3. diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 533d14fec99e..b820c311931c 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -260,6 +260,12 @@ static u32 nfsd_attrmask[] = { NFSD_WRITEABLE_ATTRS_WORD2 }; +static u32 nfsd41_ex_attrmask[] = { + NFSD_SUPPATTR_EXCLCREAT_WORD0, + NFSD_SUPPATTR_EXCLCREAT_WORD1, + NFSD_SUPPATTR_EXCLCREAT_WORD2 +}; + static __be32 nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, u32 *writable, struct iattr *iattr, struct nfs4_acl **acl) @@ -684,6 +690,17 @@ nfsd4_decode_open(struct nfsd4_compoundargs *argp, struct nfsd4_open *open) READ_BUF(8); COPYMEM(open->op_verf.data, 8); break; + case NFS4_CREATE_EXCLUSIVE4_1: + if (argp->minorversion < 1) + goto xdr_error; + READ_BUF(8); + COPYMEM(open->op_verf.data, 8); + status = nfsd4_decode_fattr(argp, open->op_bmval, + nfsd41_ex_attrmask, &open->op_iattr, + &open->op_acl); + if (status) + goto out; + break; default: goto xdr_error; } -- cgit v1.2.3 From f0ad670d7061efad138df19aefe569263c4ea37b Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Mon, 6 Apr 2009 12:00:36 +0300 Subject: nfsd41: define NFSD_DRC_SIZE_SHIFT in set_max_drc Fixes the following compiler error: fs/nfsd/nfssvc.c: In function 'set_max_drc': fs/nfsd/nfssvc.c:240: error: 'NFSD_DRC_SIZE_SHIFT' undeclared CONFIG_NFSD_V4 is not set Reported-by: Alexander Beregalov Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfssvc.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index e9d57734a348..469c931cca95 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -236,6 +236,8 @@ void nfsd_reset_versions(void) */ static void set_max_drc(void) { + /* The percent of nr_free_buffer_pages used by the V4.1 server DRC */ + #define NFSD_DRC_SIZE_SHIFT 7 nfsd_serv->sv_drc_max_pages = nr_free_buffer_pages() >> NFSD_DRC_SIZE_SHIFT; nfsd_serv->sv_drc_pages_used = 0; -- cgit v1.2.3