From 361cad3c89070aeb37560860ea8bfc092d545adc Mon Sep 17 00:00:00 2001 From: Andrew Elble Date: Wed, 2 Dec 2015 09:20:57 -0500 Subject: nfs: Fix race in __update_open_stateid() We've seen this in a packet capture - I've intermixed what I think was going on. The fix here is to grab the so_lock sooner. 1964379 -> #1 open (for write) reply seqid=1 1964393 -> #2 open (for read) reply seqid=2 __nfs4_close(), state->n_wronly-- nfs4_state_set_mode_locked(), changes state->state = [R] state->flags is [RW] state->state is [R], state->n_wronly == 0, state->n_rdonly == 1 1964398 -> #3 open (for write) call -> because close is already running 1964399 -> downgrade (to read) call seqid=2 (close of #1) 1964402 -> #3 open (for write) reply seqid=3 __update_open_stateid() nfs_set_open_stateid_locked(), changes state->flags state->flags is [RW] state->state is [R], state->n_wronly == 0, state->n_rdonly == 1 new sequence number is exposed now via nfs4_stateid_copy() next step would be update_open_stateflags(), pending so_lock 1964403 -> downgrade reply seqid=2, fails with OLD_STATEID (close of #1) nfs4_close_prepare() gets so_lock and recalcs flags -> send close 1964405 -> downgrade (to read) call seqid=3 (close of #1 retry) __update_open_stateid() gets so_lock * update_open_stateflags() updates state->n_wronly. nfs4_state_set_mode_locked() updates state->state state->flags is [RW] state->state is [RW], state->n_wronly == 1, state->n_rdonly == 1 * should have suppressed the preceding nfs4_close_prepare() from sending open_downgrade 1964406 -> write call 1964408 -> downgrade (to read) reply seqid=4 (close of #1 retry) nfs_clear_open_stateid_locked() state->flags is [R] state->state is [RW], state->n_wronly == 1, state->n_rdonly == 1 1964409 -> write reply (fails, openmode) Signed-off-by: Andrew Elble Cc: stable@vger,kernel.org Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 8981803..2f1714f 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1385,6 +1385,7 @@ static void __update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_s * Protect the call to nfs4_state_set_mode_locked and * serialise the stateid update */ + spin_lock(&state->owner->so_lock); write_seqlock(&state->seqlock); if (deleg_stateid != NULL) { nfs4_stateid_copy(&state->stateid, deleg_stateid); @@ -1393,7 +1394,6 @@ static void __update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_s if (open_stateid != NULL) nfs_set_open_stateid_locked(state, open_stateid, fmode); write_sequnlock(&state->seqlock); - spin_lock(&state->owner->so_lock); update_open_stateflags(state, fmode); spin_unlock(&state->owner->so_lock); } -- cgit v1.1 From 95864c9154c1385c33e6782f4467c8f1422c1c90 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 26 Dec 2015 15:06:03 -0500 Subject: NFS: Allow the combination pNFS and labeled NFS Fix the nfs4_pnfs_open_bitmap so that it also allows for labeled NFS. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 2f1714f..1028d6d 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -208,6 +208,9 @@ static const u32 nfs4_pnfs_open_bitmap[3] = { | FATTR4_WORD1_TIME_METADATA | FATTR4_WORD1_TIME_MODIFY, FATTR4_WORD2_MDSTHRESHOLD +#ifdef CONFIG_NFS_V4_SECURITY_LABEL + | FATTR4_WORD2_SECURITY_LABEL +#endif }; static const u32 nfs4_open_noattr_bitmap[3] = { -- cgit v1.1 From 48c9579a1afe4315be4576ec4c0b24c2715da0e0 Mon Sep 17 00:00:00 2001 From: Olga Kornievskaia Date: Tue, 24 Nov 2015 13:29:41 -0500 Subject: Adding stateid information to tracepoints Operations to which stateid information is added: close, delegreturn, open, read, setattr, layoutget, layoutcommit, test_stateid, write, lock, locku, lockt Format is "stateid=:", also "openstateid=", "layoutstateid=", and "lockstateid=" for open_file, layoutget, set_lock tracepoints. New function is added to internal.h, nfs_stateid_hash(), to compute the hash trace_nfs4_setattr() is moved from nfs4_do_setattr() to _nfs4_do_setattr() to get access to stateid. trace_nfs4_setattr and trace_nfs4_delegreturn are changed from INODE_EVENT to new event type, INODE_STATEID_EVENT which is same as INODE_EVENT but adds stateid information for locking tracepoints, moved trace_nfs4_set_lock() into _nfs4_do_setlk() to get access to stateid information, and removed trace_nfs4_lock_reclaim(), trace_nfs4_lock_expired() as they call into _nfs4_do_setlk() and both were previously same LOCK_EVENT type. Signed-off-by: Olga Kornievskaia Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 8981803..492d303 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2703,6 +2703,7 @@ static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred, status = nfs4_call_sync(server->client, server, &msg, &arg.seq_args, &res.seq_res, 1); if (status == 0 && state != NULL) renew_lease(server, timestamp); + trace_nfs4_setattr(inode, &arg.stateid, status); return status; } @@ -2719,7 +2720,6 @@ static int nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred, int err; do { err = _nfs4_do_setattr(inode, cred, fattr, sattr, state, ilabel, olabel); - trace_nfs4_setattr(inode, err); switch (err) { case -NFS4ERR_OPENMODE: if (!(sattr->ia_valid & ATTR_SIZE)) { @@ -5426,7 +5426,7 @@ int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4 int err; do { err = _nfs4_proc_delegreturn(inode, cred, stateid, issync); - trace_nfs4_delegreturn(inode, err); + trace_nfs4_delegreturn(inode, stateid, err); switch (err) { case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: @@ -5936,6 +5936,7 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *f data->cancelled = 1; rpc_put_task(task); dprintk("%s: done, ret = %d!\n", __func__, ret); + trace_nfs4_set_lock(fl, state, &data->res.stateid, cmd, ret); return ret; } @@ -5952,7 +5953,6 @@ static int nfs4_lock_reclaim(struct nfs4_state *state, struct file_lock *request if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0) return 0; err = _nfs4_do_setlk(state, F_SETLK, request, NFS_LOCK_RECLAIM); - trace_nfs4_lock_reclaim(request, state, F_SETLK, err); if (err != -NFS4ERR_DELAY) break; nfs4_handle_exception(server, err, &exception); @@ -5979,7 +5979,6 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0) return 0; err = _nfs4_do_setlk(state, F_SETLK, request, NFS_LOCK_EXPIRED); - trace_nfs4_lock_expired(request, state, F_SETLK, err); switch (err) { default: goto out; @@ -6087,7 +6086,6 @@ static int nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock * do { err = _nfs4_proc_setlk(state, cmd, request); - trace_nfs4_set_lock(request, state, cmd, err); if (err == -NFS4ERR_DENIED) err = -EAGAIN; err = nfs4_handle_exception(NFS_SERVER(state->inode), @@ -7994,6 +7992,7 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags) trace_nfs4_layoutget(lgp->args.ctx, &lgp->args.range, &lgp->res.range, + &lgp->res.stateid, status); /* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */ if (status == 0 && lgp->res.layoutp->len) @@ -8101,7 +8100,7 @@ int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync) return PTR_ERR(task); if (sync) status = task->tk_status; - trace_nfs4_layoutreturn(lrp->args.inode, status); + trace_nfs4_layoutreturn(lrp->args.inode, &lrp->args.stateid, status); dprintk("<-- %s status=%d\n", __func__, status); rpc_put_task(task); return status; @@ -8249,7 +8248,7 @@ nfs4_proc_layoutcommit(struct nfs4_layoutcommit_data *data, bool sync) return PTR_ERR(task); if (sync) status = task->tk_status; - trace_nfs4_layoutcommit(data->args.inode, status); + trace_nfs4_layoutcommit(data->args.inode, &data->args.stateid, status); dprintk("%s: status %d\n", __func__, status); rpc_put_task(task); return status; -- cgit v1.1 From 9759b0fb1d202e6bc9bdab10612189abc9389df8 Mon Sep 17 00:00:00 2001 From: Olga Kornievskaia Date: Tue, 24 Nov 2015 13:29:42 -0500 Subject: Adding tracepoint to cached open Signed-off-by: Olga Kornievskaia Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 492d303..e941261 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1598,6 +1598,7 @@ _nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data) if (!data->rpc_done) { state = nfs4_try_open_cached(data); + trace_nfs4_cached_open(data->state); goto out; } @@ -2015,6 +2016,7 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata) } return; unlock_no_action: + trace_nfs4_cached_open(data->state); rcu_read_unlock(); out_no_action: task->tk_action = NULL; -- cgit v1.1 From f2dd436edb2b38b555f3ff7257308bb56fa4eea5 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 8 Oct 2015 11:33:17 -0400 Subject: NFSv4: Fix unused variable warnings in nfs4_init_*_client_string() Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index e941261..18d862d 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -5050,7 +5050,6 @@ static void nfs4_init_boot_verifier(const struct nfs_client *clp, static int nfs4_init_nonuniform_client_string(struct nfs_client *clp) { - int result; size_t len; char *str; @@ -5078,7 +5077,7 @@ nfs4_init_nonuniform_client_string(struct nfs_client *clp) return -ENOMEM; rcu_read_lock(); - result = scnprintf(str, len, "Linux NFSv4.0 %s/%s %s", + scnprintf(str, len, "Linux NFSv4.0 %s/%s %s", clp->cl_ipaddr, rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR), rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_PROTO)); @@ -5091,7 +5090,6 @@ nfs4_init_nonuniform_client_string(struct nfs_client *clp) static int nfs4_init_uniquifier_client_string(struct nfs_client *clp) { - int result; size_t len; char *str; @@ -5111,7 +5109,7 @@ nfs4_init_uniquifier_client_string(struct nfs_client *clp) if (!str) return -ENOMEM; - result = scnprintf(str, len, "Linux NFSv%u.%u %s/%s", + scnprintf(str, len, "Linux NFSv%u.%u %s/%s", clp->rpc_ops->version, clp->cl_minorversion, nfs4_client_id_uniquifier, clp->cl_rpcclient->cl_nodename); @@ -5122,7 +5120,6 @@ nfs4_init_uniquifier_client_string(struct nfs_client *clp) static int nfs4_init_uniform_client_string(struct nfs_client *clp) { - int result; size_t len; char *str; @@ -5147,7 +5144,7 @@ nfs4_init_uniform_client_string(struct nfs_client *clp) if (!str) return -ENOMEM; - result = scnprintf(str, len, "Linux NFSv%u.%u %s", + scnprintf(str, len, "Linux NFSv%u.%u %s", clp->rpc_ops->version, clp->cl_minorversion, clp->cl_rpcclient->cl_nodename); clp->cl_owner_id = str; -- cgit v1.1 From 99ade3c71b1e40e7174d6527709399a87f3d05e0 Mon Sep 17 00:00:00 2001 From: Andrew Elble Date: Wed, 2 Dec 2015 09:39:51 -0500 Subject: nfs: machine credential support for additional operations Allow LAYOUTRETURN and DELEGRETURN to use machine credentials if the server supports it. Add request for OPEN_DOWNGRADE as the close path also uses that. Signed-off-by: Andrew Elble Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 18d862d..a7d564a 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -5383,6 +5383,11 @@ static int _nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, co if (data == NULL) return -ENOMEM; nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1); + + nfs4_state_protect(server->nfs_client, + NFS_SP4_MACH_CRED_CLEANUP, + &task_setup_data.rpc_client, &msg); + data->args.fhandle = &data->fh; data->args.stateid = &data->stateid; data->args.bitmask = server->cache_consistency_bitmask; @@ -6859,10 +6864,13 @@ static const struct nfs41_state_protection nfs4_sp4_mach_cred_request = { }, .allow.u.words = { [0] = 1 << (OP_CLOSE) | + 1 << (OP_OPEN_DOWNGRADE) | 1 << (OP_LOCKU) | + 1 << (OP_DELEGRETURN) | 1 << (OP_COMMIT), [1] = 1 << (OP_SECINFO - 32) | 1 << (OP_SECINFO_NO_NAME - 32) | + 1 << (OP_LAYOUTRETURN - 32) | 1 << (OP_TEST_STATEID - 32) | 1 << (OP_FREE_STATEID - 32) | 1 << (OP_WRITE - 32) @@ -6927,11 +6935,19 @@ static int nfs4_sp4_select_mode(struct nfs_client *clp, } if (test_bit(OP_CLOSE, sp->allow.u.longs) && + test_bit(OP_OPEN_DOWNGRADE, sp->allow.u.longs) && + test_bit(OP_DELEGRETURN, sp->allow.u.longs) && test_bit(OP_LOCKU, sp->allow.u.longs)) { dfprintk(MOUNT, " cleanup mode enabled\n"); set_bit(NFS_SP4_MACH_CRED_CLEANUP, &clp->cl_sp4_flags); } + if (test_bit(OP_LAYOUTRETURN, sp->allow.u.longs)) { + dfprintk(MOUNT, " pnfs cleanup mode enabled\n"); + set_bit(NFS_SP4_MACH_CRED_PNFS_CLEANUP, + &clp->cl_sp4_flags); + } + if (test_bit(OP_SECINFO, sp->allow.u.longs) && test_bit(OP_SECINFO_NO_NAME, sp->allow.u.longs)) { dfprintk(MOUNT, " secinfo mode enabled\n"); @@ -8084,6 +8100,10 @@ int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync) }; int status = 0; + nfs4_state_protect(NFS_SERVER(lrp->args.inode)->nfs_client, + NFS_SP4_MACH_CRED_PNFS_CLEANUP, + &task_setup_data.rpc_client, &msg); + dprintk("--> %s\n", __func__); if (!sync) { lrp->inode = nfs_igrab_and_active(lrp->args.inode); -- cgit v1.1 From 1a093ceb053832c25b92f3cf26b957543c7baf9b Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 28 Dec 2015 11:27:15 -0500 Subject: NFSv4.1/pnfs: Fixup an lo->plh_block_lgets imbalance in layoutreturn Since commit 2d8ae84fbc32, nothing is bumping lo->plh_block_lgets in the layoutreturn path, so it should not be touched in nfs4_layoutreturn_release either. Fixes: 2d8ae84fbc32 ("NFSv4.1/pnfs: Remove redundant lo->plh_block_lgets...") Cc: stable@vger.kernel.org # 4.3+ Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 1028d6d..5dd2cde 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -8057,7 +8057,6 @@ static void nfs4_layoutreturn_release(void *calldata) pnfs_set_layout_stateid(lo, &lrp->res.stateid, true); pnfs_mark_matching_lsegs_invalid(lo, &freeme, &lrp->args.range); pnfs_clear_layoutreturn_waitbit(lo); - lo->plh_block_lgets--; spin_unlock(&lo->plh_inode->i_lock); pnfs_free_lseg_list(&freeme); pnfs_put_layout_hdr(lrp->args.layout); -- cgit v1.1 From 7c1e6e58e2aa8b95d1b9888db1945eb0a388efb9 Mon Sep 17 00:00:00 2001 From: Peng Tao Date: Sat, 5 Dec 2015 17:01:01 +0800 Subject: NFS41: map NFS4ERR_LAYOUTUNAVAILABLE to ENODATA Instead of mapping it to EIO that is a fatal error and fails application. We'll go inband after getting NFS4ERR_LAYOUTUNAVAILABLE. Signed-off-by: Peng Tao Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 18d862d..0065bbd 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -7795,6 +7795,15 @@ static void nfs4_layoutget_done(struct rpc_task *task, void *calldata) switch (task->tk_status) { case 0: goto out; + + /* + * NFS4ERR_LAYOUTUNAVAILABLE means we are not supposed to use pnfs + * on the file. set tk_status to -ENODATA to tell upper layer to + * retry go inband. + */ + case -NFS4ERR_LAYOUTUNAVAILABLE: + task->tk_status = -ENODATA; + goto out; /* * NFS4ERR_BADLAYOUT means the MDS cannot return a layout of * length lgp->args.minlength != 0 (see RFC5661 section 18.43.3). -- cgit v1.1 From ab7d763e477c5be33ac9cffc68e808bbd69371f7 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 28 Dec 2015 10:32:11 -0500 Subject: pNFS: Ensure nfs4_layoutget_prepare returns the correct error If we're unable to perform the layoutget due to an invalid open stateid or a bulk recall, ensure that we return the error so that the caller can decide on an appropriate action. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 18d862d..fcd7a90 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -7760,6 +7760,7 @@ nfs4_layoutget_prepare(struct rpc_task *task, void *calldata) struct nfs4_layoutget *lgp = calldata; struct nfs_server *server = NFS_SERVER(lgp->args.inode); struct nfs4_session *session = nfs4_get_session(server); + int ret; dprintk("--> %s\n", __func__); /* Note the is a race here, where a CB_LAYOUTRECALL can come in @@ -7770,12 +7771,12 @@ nfs4_layoutget_prepare(struct rpc_task *task, void *calldata) if (nfs41_setup_sequence(session, &lgp->args.seq_args, &lgp->res.seq_res, task)) return; - if (pnfs_choose_layoutget_stateid(&lgp->args.stateid, + ret = pnfs_choose_layoutget_stateid(&lgp->args.stateid, NFS_I(lgp->args.inode)->layout, &lgp->args.range, - lgp->args.ctx->state)) { - rpc_exit(task, NFS4_OK); - } + lgp->args.ctx->state); + if (ret < 0) + rpc_exit(task, ret); } static void nfs4_layoutget_done(struct rpc_task *task, void *calldata) -- cgit v1.1 From 0654cc726fc6eed6dca915fb65ba7975716ea080 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 28 Dec 2015 11:48:14 -0500 Subject: NFSv4.1/pNFS: Add a helper to mark the layout as returned This ensures that we don't reuse the stateid if a layout return or implied layout return means that we've returned all layout segments Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs/nfs/nfs4proc.c') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index fcd7a90..883da29 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -8049,9 +8049,10 @@ static void nfs4_layoutreturn_release(void *calldata) dprintk("--> %s\n", __func__); spin_lock(&lo->plh_inode->i_lock); + pnfs_mark_matching_lsegs_invalid(lo, &freeme, &lrp->args.range); + pnfs_mark_layout_returned_if_empty(lo); if (lrp->res.lrs_present) pnfs_set_layout_stateid(lo, &lrp->res.stateid, true); - pnfs_mark_matching_lsegs_invalid(lo, &freeme, &lrp->args.range); pnfs_clear_layoutreturn_waitbit(lo); lo->plh_block_lgets--; spin_unlock(&lo->plh_inode->i_lock); -- cgit v1.1