Merge branch 'nfsd-next' of git://linux-nfs.org/~bfields/linux

Pull nfsd changes from Bruce Fields:
"This includes miscellaneous bugfixes and cleanup and a performance fix
for write-heavy NFSv4 workloads.

(The most significant nfsd-relevant change this time is actually in
the delegation patches that went through Viro, fixing a long-standing
bug that can cause NFSv4 clients to miss updates made by non-nfs users
of the filesystem. Those enable some followup nfsd patches which I
have queued locally, but those can wait till 3.14)"

* 'nfsd-next' of git://linux-nfs.org/~bfields/linux: (24 commits)
nfsd: export proper maximum file size to the client
nfsd4: improve write performance with better sendspace reservations
svcrpc: remove an unnecessary assignment
sunrpc: comment typo fix
Revert "nfsd: remove_stid can be incorporated into nfs4_put_delegation"
nfsd4: fix discarded security labels on setattr
NFSD: Add support for NFS v4.2 operation checking
nfsd4: nfsd_shutdown_net needs state lock
NFSD: Combine decode operations for v4 and v4.1
nfsd: -EINVAL on invalid anonuid/gid instead of silent failure
nfsd: return better errors to exportfs
nfsd: fh_update should error out in unexpected cases
nfsd4: need to destroy revoked delegations in destroy_client
nfsd: no need to unhash_stid before free
nfsd: remove_stid can be incorporated into nfs4_put_delegation
nfsd: nfs4_open_delegation needs to remove_stid rather than unhash_stid
nfsd: nfs4_free_stid
nfsd: fix Kconfig syntax
sunrpc: trim off EC bytes in GSSAPI v2 unwrap
gss_krb5: document that we ignore sequence number
...

fs/nfsd/Kconfig

···
 	Smack policies on NFSv4 files, say N.
 
 	WARNING: there is still a chance of backwards-incompatible protocol changes.
-	For now we recommend "Y" only for developers and testers."
+	For now we recommend "Y" only for developers and testers.
 
 config NFSD_FAULT_INJECTION
 	bool "NFS server manual fault injection"

fs/nfsd/export.c

···
 		if (err)
 			goto out3;
 		exp.ex_anon_uid= make_kuid(&init_user_ns, an_int);
-		if (!uid_valid(exp.ex_anon_uid))
-			goto out3;
 
 		/* anon gid */
 		err = get_int(&mesg, &an_int);
 		if (err)
 			goto out3;
 		exp.ex_anon_gid= make_kgid(&init_user_ns, an_int);
-		if (!gid_valid(exp.ex_anon_gid))
-			goto out3;
 
 		/* fsid */
 		err = get_int(&mesg, &an_int);
···
 				   exp.ex_uuid);
 		if (err)
 			goto out4;
+		/*
+		 * No point caching this if it would immediately expire.
+		 * Also, this protects exportfs's dummy export from the
+		 * anon_uid/anon_gid checks:
+		 */
+		if (exp.h.expiry_time < seconds_since_boot())
+			goto out4;
+		/*
+		 * For some reason exportfs has been passing down an
+		 * invalid (-1) uid & gid on the "dummy" export which it
+		 * uses to test export support.  To make sure exportfs
+		 * sees errors from check_export we therefore need to
+		 * delay these checks till after check_export:
+		 */
+		err = -EINVAL;
+		if (!uid_valid(exp.ex_anon_uid))
+			goto out4;
+		if (!gid_valid(exp.ex_anon_gid))
+			goto out4;
+		err = 0;
 	}
 
 	expp = svc_export_lookup(&exp);

fs/nfsd/nfs4state.c

···
 	idr_remove(stateids, s->sc_stateid.si_opaque.so_id);
 }
 
+static void nfs4_free_stid(struct kmem_cache *slab, struct nfs4_stid *s)
+{
+	kmem_cache_free(slab, s);
+}
+
 void
 nfs4_put_delegation(struct nfs4_delegation *dp)
 {
 	if (atomic_dec_and_test(&dp->dl_count)) {
-		kmem_cache_free(deleg_slab, dp);
+		nfs4_free_stid(deleg_slab, &dp->dl_stid);
 		num_delegations--;
 	}
 }
···
 static void free_generic_stateid(struct nfs4_ol_stateid *stp)
 {
 	remove_stid(&stp->st_stid);
-	kmem_cache_free(stateid_slab, stp);
+	nfs4_free_stid(stateid_slab, &stp->st_stid);
 }
 
 static void release_lock_stateid(struct nfs4_ol_stateid *stp)
···
 static void release_open_stateid(struct nfs4_ol_stateid *stp)
 {
 	unhash_open_stateid(stp);
-	unhash_stid(&stp->st_stid);
 	free_generic_stateid(stp);
 }
 
···
 	struct nfs4_ol_stateid *s = oo->oo_last_closed_stid;
 
 	if (s) {
-		unhash_stid(&s->st_stid);
 		free_generic_stateid(s);
 		oo->oo_last_closed_stid = NULL;
 	}
···
 	while (!list_empty(&reaplist)) {
 		dp = list_entry(reaplist.next, struct nfs4_delegation, dl_recall_lru);
 		destroy_delegation(dp);
+	}
+	list_splice_init(&clp->cl_revoked, &reaplist);
+	while (!list_empty(&reaplist)) {
+		dp = list_entry(reaplist.next, struct nfs4_delegation, dl_recall_lru);
+		destroy_revoked_delegation(dp);
 	}
 	while (!list_empty(&clp->cl_openowners)) {
 		oo = list_entry(clp->cl_openowners.next, struct nfs4_openowner, oo_perclient);
···
 	open->op_delegate_type = NFS4_OPEN_DELEGATE_READ;
 	return;
 out_free:
-	unhash_stid(&dp->dl_stid);
+	remove_stid(&dp->dl_stid);
 	nfs4_put_delegation(dp);
 out_no_deleg:
 	open->op_delegate_type = NFS4_OPEN_DELEGATE_NONE;
···
 
 	nfsd4_close_open_stateid(stp);
 
-	if (cstate->minorversion) {
-		unhash_stid(&stp->st_stid);
+	if (cstate->minorversion)
 		free_generic_stateid(stp);
-	} else
+	else
 		oo->oo_last_closed_stid = stp;
 
 	if (list_empty(&oo->oo_owner.so_stateids)) {
···
 	return ret;
 }
 
-/* should be called with the state lock held */
 void
 nfs4_state_shutdown_net(struct net *net)
 {
···
 	cancel_delayed_work_sync(&nn->laundromat_work);
 	locks_end_grace(&nn->nfsd4_manager);
 
+	nfs4_lock_state();
 	INIT_LIST_HEAD(&reaplist);
 	spin_lock(&recall_lock);
 	list_for_each_safe(pos, next, &nn->del_recall_lru) {
···
 
 	nfsd4_client_tracking_exit(net);
 	nfs4_state_destroy_net(net);
+	nfs4_unlock_state();
 }
 
 void

fs/nfsd/nfs4xdr.c

···
 		label->data = kzalloc(dummy32 + 1, GFP_KERNEL);
 		if (!label->data)
 			return nfserr_jukebox;
+		label->len = dummy32;
 		defer_free(argp, kfree, label->data);
 		memcpy(label->data, buf, dummy32);
 	}
···
 nfsd4_decode_open_confirm(struct nfsd4_compoundargs *argp, struct nfsd4_open_confirm *open_conf)
 {
 	DECODE_HEAD;
-		    
+
+	if (argp->minorversion >= 1)
+		return nfserr_notsupp;
+
 	status = nfsd4_decode_stateid(argp, &open_conf->oc_req_stateid);
 	if (status)
 		return status;
 	READ_BUF(4);
 	READ32(open_conf->oc_seqid);
-						        
+
 	DECODE_TAIL;
 }
 
···
 	SAVEMEM(putfh->pf_fhval, putfh->pf_fhlen);
 
 	DECODE_TAIL;
+}
+
+static __be32
+nfsd4_decode_putpubfh(struct nfsd4_compoundargs *argp, void *p)
+{
+	if (argp->minorversion == 0)
+		return nfs_ok;
+	return nfserr_notsupp;
 }
 
 static __be32
···
 {
 	DECODE_HEAD;
 
+	if (argp->minorversion >= 1)
+		return nfserr_notsupp;
+
 	READ_BUF(sizeof(clientid_t));
 	COPYMEM(clientid, sizeof(clientid_t));
 
···
 {
 	DECODE_HEAD;
 
+	if (argp->minorversion >= 1)
+		return nfserr_notsupp;
+
 	READ_BUF(NFS4_VERIFIER_SIZE);
 	COPYMEM(setclientid->se_verf.data, NFS4_VERIFIER_SIZE);
 
···
 nfsd4_decode_setclientid_confirm(struct nfsd4_compoundargs *argp, struct nfsd4_setclientid_confirm *scd_c)
 {
 	DECODE_HEAD;
+
+	if (argp->minorversion >= 1)
+		return nfserr_notsupp;
 
 	READ_BUF(8 + NFS4_VERIFIER_SIZE);
 	COPYMEM(&scd_c->sc_clientid, 8);
···
 nfsd4_decode_release_lockowner(struct nfsd4_compoundargs *argp, struct nfsd4_release_lockowner *rlockowner)
 {
 	DECODE_HEAD;
+
+	if (argp->minorversion >= 1)
+		return nfserr_notsupp;
 
 	READ_BUF(12);
 	COPYMEM(&rlockowner->rl_clientid, sizeof(clientid_t));
···
 	[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_noop,
+	[OP_PUTPUBFH]		= (nfsd4_dec)nfsd4_decode_putpubfh,
 	[OP_PUTROOTFH]		= (nfsd4_dec)nfsd4_decode_noop,
 	[OP_READ]		= (nfsd4_dec)nfsd4_decode_read,
 	[OP_READDIR]		= (nfsd4_dec)nfsd4_decode_readdir,
···
 	[OP_VERIFY]		= (nfsd4_dec)nfsd4_decode_verify,
 	[OP_WRITE]		= (nfsd4_dec)nfsd4_decode_write,
 	[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_backchannel_ctl,
···
 	[OP_RECLAIM_COMPLETE]	= (nfsd4_dec)nfsd4_decode_reclaim_complete,
 };
 
-struct nfsd4_minorversion_ops {
-	nfsd4_dec *decoders;
-	int nops;
-};
+static inline bool
+nfsd4_opnum_in_range(struct nfsd4_compoundargs *argp, struct nfsd4_op *op)
+{
+	if (op->opnum < FIRST_NFS4_OP)
+		return false;
+	else if (argp->minorversion == 0 && op->opnum > LAST_NFS40_OP)
+		return false;
+	else if (argp->minorversion == 1 && op->opnum > LAST_NFS41_OP)
+		return false;
+	else if (argp->minorversion == 2 && op->opnum > LAST_NFS42_OP)
+		return false;
+	return true;
+}
 
-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) },
-	[2] = { nfsd41_dec_ops, ARRAY_SIZE(nfsd41_dec_ops) },
-};
+/*
+ * Return a rough estimate of the maximum possible reply size.  Note the
+ * estimate includes rpc headers so is meant to be passed to
+ * svc_reserve, not svc_reserve_auth.
+ *
+ * Also note the current compound encoding permits only one operation to
+ * use pages beyond the first one, so the maximum possible length is the
+ * maximum over these values, not the sum.
+ */
+static int nfsd4_max_reply(u32 opnum)
+{
+	switch (opnum) {
+	case OP_READLINK:
+	case OP_READDIR:
+		/*
+		 * Both of these ops take a single page for data and put
+		 * the head and tail in another page:
+		 */
+		return 2 * PAGE_SIZE;
+	case OP_READ:
+		return INT_MAX;
+	default:
+		return PAGE_SIZE;
+	}
+}
 
 static __be32
 nfsd4_decode_compound(struct nfsd4_compoundargs *argp)
 {
 	DECODE_HEAD;
 	struct nfsd4_op *op;
-	struct nfsd4_minorversion_ops *ops;
 	bool cachethis = false;
+	int max_reply = PAGE_SIZE;
 	int i;
 
 	READ_BUF(4);
···
 		}
 	}
 
-	if (argp->minorversion >= ARRAY_SIZE(nfsd4_minorversion))
+	if (argp->minorversion > NFSD_SUPPORTED_MINOR_VERSION)
 		argp->opcnt = 0;
 
-	ops = &nfsd4_minorversion[argp->minorversion];
 	for (i = 0; i < argp->opcnt; i++) {
 		op = &argp->ops[i];
 		op->replay = NULL;
···
 		READ_BUF(4);
 		READ32(op->opnum);
 
-		if (op->opnum >= FIRST_NFS4_OP && op->opnum <= LAST_NFS4_OP)
-			op->status = ops->decoders[op->opnum](argp, &op->u);
+		if (nfsd4_opnum_in_range(argp, op))
+			op->status = nfsd4_dec_ops[op->opnum](argp, &op->u);
 		else {
 			op->opnum = OP_ILLEGAL;
 			op->status = nfserr_op_illegal;
···
 		 * op in the compound wants to be cached:
 		 */
 		cachethis |= nfsd4_cache_this_op(op);
+
+		max_reply = max(max_reply, nfsd4_max_reply(op->opnum));
 	}
 	/* Sessions make the DRC unnecessary: */
 	if (argp->minorversion)
 		cachethis = false;
+	if (max_reply != INT_MAX)
+		svc_reserve(argp->rqstp, max_reply);
 	argp->rqstp->rq_cachetype = cachethis ? RC_REPLBUFF : RC_NOCACHE;
 
 	DECODE_TAIL;
···
 	if (bmval0 & FATTR4_WORD0_MAXFILESIZE) {
 		if ((buflen -= 8) < 0)
 			goto out_resource;
-		WRITE64(~(u64)0);
+		WRITE64(exp->ex_path.mnt->mnt_sb->s_maxbytes);
 	}
 	if (bmval0 & FATTR4_WORD0_MAXLINK) {
 		if ((buflen -= 4) < 0)

fs/nfsd/nfsfh.c

···
 		_fh_update_old(dentry, fhp->fh_export, &fhp->fh_handle);
 	} else {
 		if (fhp->fh_handle.fh_fileid_type != FILEID_ROOT)
-			goto out;
+			return 0;
 
 		_fh_update(fhp, fhp->fh_export, dentry);
 		if (fhp->fh_handle.fh_fileid_type == FILEID_INVALID)
 			return nfserr_opnotsupp;
 	}
-out:
 	return 0;
-
 out_bad:
 	printk(KERN_ERR "fh_update: fh not verified!\n");
-	goto out;
+	return nfserr_serverfault;
 out_negative:
 	printk(KERN_ERR "fh_update: %pd2 still negative!\n",
 		dentry);
-	goto out;
+	return nfserr_serverfault;
 }
 
 /*

include/linux/nfs4.h

···
 
 #define FIRST_NFS4_OP	OP_ACCESS
 #define LAST_NFS4_OP 	OP_RECLAIM_COMPLETE
+#define LAST_NFS40_OP	OP_RELEASE_LOCKOWNER
+#define LAST_NFS41_OP	OP_RECLAIM_COMPLETE
+#define LAST_NFS42_OP	OP_RECLAIM_COMPLETE
 
 enum nfsstat4 {
 	NFS4_OK = 0,

net/sunrpc/auth_gss/gss_krb5_unseal.c

···
 	struct xdr_netobj cksumobj = {.len = sizeof(cksumdata),
 				      .data = cksumdata};
 	s32 now;
-	u64 seqnum;
 	u8 *ptr = read_token->data;
 	u8 *cksumkey;
 	u8 flags;
···
 	if (now > ctx->endtime)
 		return GSS_S_CONTEXT_EXPIRED;
 
-	/* do sequencing checks */
-
-	seqnum = be64_to_cpup((__be64 *)ptr + 8);
+	/*
+	 * NOTE: the sequence number at ptr + 8 is skipped, rpcsec_gss
+	 * doesn't want it checked; see page 6 of rfc 2203.
+	 */
 
 	return GSS_S_COMPLETE;
 }

net/sunrpc/auth_gss/gss_krb5_wrap.c

···
 gss_unwrap_kerberos_v2(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf)
 {
 	s32		now;
-	u64		seqnum;
 	u8		*ptr;
 	u8		flags = 0x00;
 	u16		ec, rrc;
···
 	ec = be16_to_cpup((__be16 *)(ptr + 4));
 	rrc = be16_to_cpup((__be16 *)(ptr + 6));
 
-	seqnum = be64_to_cpup((__be64 *)(ptr + 8));
+	/*
+	 * NOTE: the sequence number at ptr + 8 is skipped, rpcsec_gss
+	 * doesn't want it checked; see page 6 of rfc 2203.
+	 */
 
 	if (rrc != 0)
 		rotate_left(offset + 16, buf, rrc);
···
 	buf->head[0].iov_len -= GSS_KRB5_TOK_HDR_LEN + headskip;
 	buf->len -= GSS_KRB5_TOK_HDR_LEN + headskip;
 
-	/* Trim off the checksum blob */
-	xdr_buf_trim(buf, GSS_KRB5_TOK_HDR_LEN + tailskip);
+	/* Trim off the trailing "extra count" and checksum blob */
+	xdr_buf_trim(buf, ec + GSS_KRB5_TOK_HDR_LEN + tailskip);
 	return GSS_S_COMPLETE;
 }
 

net/sunrpc/auth_gss/gss_rpc_upcall.c

···
 	if (res.context_handle) {
 		data->out_handle = rctxh.exported_context_token;
 		data->mech_oid.len = rctxh.mech.len;
-		memcpy(data->mech_oid.data, rctxh.mech.data,
+		if (rctxh.mech.data)
+			memcpy(data->mech_oid.data, rctxh.mech.data,
 						data->mech_oid.len);
 		client_name = rctxh.src_name.display_name;
 	}

net/sunrpc/auth_gss/gss_rpc_xdr.c

···
 
 	/* cred->elements */
 	err = dummy_enc_credel_array(xdr, &cred->elements);
+	if (err)
+		return err;
 
 	/* cred->cred_handle_reference */
 	err = gssx_enc_buffer(xdr, &cred->cred_handle_reference);
···
 		goto done;
 
 	/* arg->context_handle */
-	if (arg->context_handle) {
+	if (arg->context_handle)
 		err = gssx_enc_ctx(xdr, arg->context_handle);
-		if (err)
-			goto done;
-	} else {
+	else
 		err = gssx_enc_bool(xdr, 0);
-	}
+	if (err)
+		goto done;
 
 	/* arg->cred_handle */
-	if (arg->cred_handle) {
+	if (arg->cred_handle)
 		err = gssx_enc_cred(xdr, arg->cred_handle);
-		if (err)
-			goto done;
-	} else {
+	else
 		err = gssx_enc_bool(xdr, 0);
-	}
+	if (err)
+		goto done;
 
 	/* arg->input_token */
 	err = gssx_enc_in_token(xdr, &arg->input_token);
···
 		goto done;
 
 	/* arg->input_cb */
-	if (arg->input_cb) {
+	if (arg->input_cb)
 		err = gssx_enc_cb(xdr, arg->input_cb);
-		if (err)
-			goto done;
-	} else {
+	else
 		err = gssx_enc_bool(xdr, 0);
-	}
+	if (err)
+		goto done;
 
 	err = gssx_enc_bool(xdr, arg->ret_deleg_cred);
 	if (err)

net/sunrpc/auth_gss/svcauth_gss.c

···
 	if (!ud->found_creds) {
 		/* userspace seem buggy, we should always get at least a
 		 * mapping to nobody */
-		dprintk("RPC:       No creds found, marking Negative!\n");
-		set_bit(CACHE_NEGATIVE, &rsci.h.flags);
+		dprintk("RPC:       No creds found!\n");
+		goto out;
 	} else {
 
 		/* steal creds */

net/sunrpc/svc.c

···
 	rqstp->rq_vers = vers = svc_getnl(argv);	/* version number */
 	rqstp->rq_proc = proc = svc_getnl(argv);	/* procedure number */
 
-	progp = serv->sv_program;
-
 	for (progp = serv->sv_program; progp; progp = progp->pg_next)
 		if (prog == progp->pg_prog)
 			break;