Merge tag 'nfsd-4.17' of git://linux-nfs.org/~bfields/linux

Pull nfsd updates from Bruce Fields:
 "Chuck Lever did a bunch of work on nfsd tracepoints, on RDMA, and on
  server xdr decoding (with an eye towards eliminating a data copy in
  the RDMA case).

  I did some refactoring of the delegation code in preparation for
  eliminating some delegation self-conflicts and implementing write
  delegations"

* tag 'nfsd-4.17' of git://linux-nfs.org/~bfields/linux: (40 commits)
  nfsd: fix incorrect umasks
  sunrpc: remove incorrect HMAC request initialization
  NFSD: Clean up legacy NFS SYMLINK argument XDR decoders
  NFSD: Clean up legacy NFS WRITE argument XDR decoders
  nfsd: Trace NFSv4 COMPOUND execution
  nfsd: Add I/O trace points in the NFSv4 read proc
  nfsd: Add I/O trace points in the NFSv4 write path
  nfsd: Add "nfsd_" to trace point names
  nfsd: Record request byte count, not count of vectors
  nfsd: Fix NFSD trace points
  svc: Report xprt dequeue latency
  sunrpc: Report per-RPC execution stats
  sunrpc: Re-purpose trace_svc_process
  sunrpc: Save remote presentation address in svc_xprt for trace events
  sunrpc: Simplify trace_svc_recv
  sunrpc: Simplify do_enqueue tracing
  sunrpc: Move trace_svc_xprt_dequeue()
  sunrpc: Update show_svc_xprt_flags() to include recently added flags
  svc: Simplify ->xpo_secure_port
  sunrpc: Remove unneeded pointer dereference
  ...

Author: torvalds

fs/lockd/svc.c

@@ -57,8 +57,8 @@ static struct task_struct	*nlmsvc_task;
 static struct svc_rqst		*nlmsvc_rqst;
 unsigned long			nlmsvc_timeout;
 
-atomic_t nlm_ntf_refcnt = ATOMIC_INIT(0);
-DECLARE_WAIT_QUEUE_HEAD(nlm_ntf_wq);
+static atomic_t nlm_ntf_refcnt = ATOMIC_INIT(0);
+static DECLARE_WAIT_QUEUE_HEAD(nlm_ntf_wq);
 
 unsigned int lockd_net_id;
 

fs/nfsd/nfs3proc.c

@@ -192,6 +192,7 @@ nfsd3_proc_write(struct svc_rqst *rqstp)
 	struct nfsd3_writeres *resp = rqstp->rq_resp;
 	__be32	nfserr;
 	unsigned long cnt = argp->len;
+	unsigned int nvecs;
 
 	dprintk("nfsd: WRITE(3)    %s %d bytes at %Lu%s\n",
 				SVCFH_fmt(&argp->fh),
@@ -201,9 +202,12 @@ nfsd3_proc_write(struct svc_rqst *rqstp)
 
 	fh_copy(&resp->fh, &argp->fh);
 	resp->committed = argp->stable;
+	nvecs = svc_fill_write_vector(rqstp, &argp->first, cnt);
+	if (!nvecs)
+		RETURN_STATUS(nfserr_io);
 	nfserr = nfsd_write(rqstp, &resp->fh, argp->offset,
-				rqstp->rq_vec, argp->vlen,
-				&cnt, resp->committed);
+			    rqstp->rq_vec, nvecs, &cnt,
+			    resp->committed);
 	resp->count = cnt;
 	RETURN_STATUS(nfserr);
 }
@@ -279,6 +283,16 @@ nfsd3_proc_symlink(struct svc_rqst *rqstp)
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
 	__be32	nfserr;
 
+	if (argp->tlen == 0)
+		RETURN_STATUS(nfserr_inval);
+	if (argp->tlen > NFS3_MAXPATHLEN)
+		RETURN_STATUS(nfserr_nametoolong);
+
+	argp->tname = svc_fill_symlink_pathname(rqstp, &argp->first,
+						argp->tlen);
+	if (IS_ERR(argp->tname))
+		RETURN_STATUS(nfserrno(PTR_ERR(argp->tname)));
+
 	dprintk("nfsd: SYMLINK(3)  %s %.*s -> %.*s\n",
 				SVCFH_fmt(&argp->ffh),
 				argp->flen, argp->fname,

fs/nfsd/nfs3xdr.c

@@ -391,7 +391,7 @@ int
 nfs3svc_decode_writeargs(struct svc_rqst *rqstp, __be32 *p)
 {
 	struct nfsd3_writeargs *args = rqstp->rq_argp;
-	unsigned int len, v, hdr, dlen;
+	unsigned int len, hdr, dlen;
 	u32 max_blocksize = svc_max_payload(rqstp);
 	struct kvec *head = rqstp->rq_arg.head;
 	struct kvec *tail = rqstp->rq_arg.tail;
@@ -433,17 +433,9 @@ nfs3svc_decode_writeargs(struct svc_rqst *rqstp, __be32 *p)
 		args->count = max_blocksize;
 		len = args->len = max_blocksize;
 	}
-	rqstp->rq_vec[0].iov_base = (void*)p;
-	rqstp->rq_vec[0].iov_len = head->iov_len - hdr;
-	v = 0;
-	while (len > rqstp->rq_vec[v].iov_len) {
-		len -= rqstp->rq_vec[v].iov_len;
-		v++;
-		rqstp->rq_vec[v].iov_base = page_address(rqstp->rq_pages[v]);
-		rqstp->rq_vec[v].iov_len = PAGE_SIZE;
-	}
-	rqstp->rq_vec[v].iov_len = len;
-	args->vlen = v + 1;
+
+	args->first.iov_base = (void *)p;
+	args->first.iov_len = head->iov_len - hdr;
 	return 1;
 }
 
@@ -489,51 +481,24 @@ int
 nfs3svc_decode_symlinkargs(struct svc_rqst *rqstp, __be32 *p)
 {
 	struct nfsd3_symlinkargs *args = rqstp->rq_argp;
-	unsigned int len, avail;
-	char *old, *new;
-	struct kvec *vec;
+	char *base = (char *)p;
+	size_t dlen;
 
 	if (!(p = decode_fh(p, &args->ffh)) ||
-	    !(p = decode_filename(p, &args->fname, &args->flen))
-		)
+	    !(p = decode_filename(p, &args->fname, &args->flen)))
 		return 0;
 	p = decode_sattr3(p, &args->attrs);
 
-	/* now decode the pathname, which might be larger than the first page.
-	 * As we have to check for nul's anyway, we copy it into a new page
-	 * This page appears in the rq_res.pages list, but as pages_len is always
-	 * 0, it won't get in the way
-	 */
-	len = ntohl(*p++);
-	if (len == 0 || len > NFS3_MAXPATHLEN || len >= PAGE_SIZE)
-		return 0;
-	args->tname = new = page_address(*(rqstp->rq_next_page++));
-	args->tlen = len;
-	/* first copy and check from the first page */
-	old = (char*)p;
-	vec = &rqstp->rq_arg.head[0];
-	if ((void *)old > vec->iov_base + vec->iov_len)
-		return 0;
-	avail = vec->iov_len - (old - (char*)vec->iov_base);
-	while (len && avail && *old) {
-		*new++ = *old++;
-		len--;
-		avail--;
-	}
-	/* now copy next page if there is one */
-	if (len && !avail && rqstp->rq_arg.page_len) {
-		avail = min_t(unsigned int, rqstp->rq_arg.page_len, PAGE_SIZE);
-		old = page_address(rqstp->rq_arg.pages[0]);
-	}
-	while (len && avail && *old) {
-		*new++ = *old++;
-		len--;
-		avail--;
-	}
-	*new = '\0';
-	if (len)
-		return 0;
+	args->tlen = ntohl(*p++);
+
+	args->first.iov_base = p;
+	args->first.iov_len = rqstp->rq_arg.head[0].iov_len;
+	args->first.iov_len -= (char *)p - base;
 
+	dlen = args->first.iov_len + rqstp->rq_arg.page_len +
+	       rqstp->rq_arg.tail[0].iov_len;
+	if (dlen < XDR_QUADLEN(args->tlen) << 2)
+		return 0;
 	return 1;
 }
 

fs/nfsd/nfs4callback.c

@@ -223,8 +223,8 @@ static int nfs_cb_stat_to_errno(int status)
 	return -status;
 }
 
-static int decode_cb_op_status(struct xdr_stream *xdr, enum nfs_opnum4 expected,
-			       int *status)
+static int decode_cb_op_status(struct xdr_stream *xdr,
+			       enum nfs_cb_opnum4 expected, int *status)
 {
 	__be32 *p;
 	u32 op;

fs/nfsd/nfs4layouts.c

@@ -165,7 +165,7 @@ nfsd4_free_layout_stateid(struct nfs4_stid *stid)
 	struct nfs4_client *clp = ls->ls_stid.sc_client;
 	struct nfs4_file *fp = ls->ls_stid.sc_file;
 
-	trace_layoutstate_free(&ls->ls_stid.sc_stateid);
+	trace_nfsd_layoutstate_free(&ls->ls_stid.sc_stateid);
 
 	spin_lock(&clp->cl_lock);
 	list_del_init(&ls->ls_perclnt);
@@ -264,7 +264,7 @@ nfsd4_alloc_layout_stateid(struct nfsd4_compound_state *cstate,
 	list_add(&ls->ls_perfile, &fp->fi_lo_states);
 	spin_unlock(&fp->fi_lock);
 
-	trace_layoutstate_alloc(&ls->ls_stid.sc_stateid);
+	trace_nfsd_layoutstate_alloc(&ls->ls_stid.sc_stateid);
 	return ls;
 }
 
@@ -334,7 +334,7 @@ nfsd4_recall_file_layout(struct nfs4_layout_stateid *ls)
 	if (list_empty(&ls->ls_layouts))
 		goto out_unlock;
 
-	trace_layout_recall(&ls->ls_stid.sc_stateid);
+	trace_nfsd_layout_recall(&ls->ls_stid.sc_stateid);
 
 	refcount_inc(&ls->ls_stid.sc_count);
 	nfsd4_run_cb(&ls->ls_recall);
@@ -507,7 +507,7 @@ nfsd4_return_file_layouts(struct svc_rqst *rqstp,
 						false, lrp->lr_layout_type,
 						&ls);
 	if (nfserr) {
-		trace_layout_return_lookup_fail(&lrp->lr_sid);
+		trace_nfsd_layout_return_lookup_fail(&lrp->lr_sid);
 		return nfserr;
 	}
 
@@ -523,7 +523,7 @@ nfsd4_return_file_layouts(struct svc_rqst *rqstp,
 			nfs4_inc_and_copy_stateid(&lrp->lr_sid, &ls->ls_stid);
 		lrp->lrs_present = 1;
 	} else {
-		trace_layoutstate_unhash(&ls->ls_stid.sc_stateid);
+		trace_nfsd_layoutstate_unhash(&ls->ls_stid.sc_stateid);
 		nfs4_unhash_stid(&ls->ls_stid);
 		lrp->lrs_present = 0;
 	}
@@ -694,7 +694,7 @@ nfsd4_cb_layout_done(struct nfsd4_callback *cb, struct rpc_task *task)
 		/*
 		 * Unknown error or non-responding client, we'll need to fence.
 		 */
-		trace_layout_recall_fail(&ls->ls_stid.sc_stateid);
+		trace_nfsd_layout_recall_fail(&ls->ls_stid.sc_stateid);
 
 		ops = nfsd4_layout_ops[ls->ls_layout_type];
 		if (ops->fence_client)
@@ -703,7 +703,7 @@ nfsd4_cb_layout_done(struct nfsd4_callback *cb, struct rpc_task *task)
 			nfsd4_cb_layout_fail(ls);
 		return -1;
 	case -NFS4ERR_NOMATCHING_LAYOUT:
-		trace_layout_recall_done(&ls->ls_stid.sc_stateid);
+		trace_nfsd_layout_recall_done(&ls->ls_stid.sc_stateid);
 		task->tk_status = 0;
 		return 1;
 	}
@@ -716,7 +716,7 @@ nfsd4_cb_layout_release(struct nfsd4_callback *cb)
 		container_of(cb, struct nfs4_layout_stateid, ls_recall);
 	LIST_HEAD(reaplist);
 
-	trace_layout_recall_release(&ls->ls_stid.sc_stateid);
+	trace_nfsd_layout_recall_release(&ls->ls_stid.sc_stateid);
 
 	nfsd4_return_all_layouts(ls, &reaplist);
 	nfsd4_free_layouts(&reaplist);

fs/nfsd/nfs4proc.c

@@ -32,6 +32,7 @@
  *  NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
  *  SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  */
+#include <linux/fs_struct.h>
 #include <linux/file.h>
 #include <linux/falloc.h>
 #include <linux/slab.h>
@@ -252,11 +253,13 @@ do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stru
 		 * Note: create modes (UNCHECKED,GUARDED...) are the same
 		 * in NFSv4 as in v3 except EXCLUSIVE4_1.
 		 */
+		current->fs->umask = open->op_umask;
 		status = do_nfsd_create(rqstp, current_fh, open->op_fname.data,
 					open->op_fname.len, &open->op_iattr,
 					*resfh, open->op_createmode,
 					(u32 *)open->op_verf.data,
 					&open->op_truncate, &open->op_created);
+		current->fs->umask = 0;
 
 		if (!status && open->op_label.len)
 			nfsd4_security_inode_setsecctx(*resfh, &open->op_label, open->op_bmval);
@@ -603,6 +606,7 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	if (status)
 		return status;
 
+	current->fs->umask = create->cr_umask;
 	switch (create->cr_type) {
 	case NF4LNK:
 		status = nfsd_symlink(rqstp, &cstate->current_fh,
@@ -611,20 +615,22 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		break;
 
 	case NF4BLK:
+		status = nfserr_inval;
 		rdev = MKDEV(create->cr_specdata1, create->cr_specdata2);
 		if (MAJOR(rdev) != create->cr_specdata1 ||
 		    MINOR(rdev) != create->cr_specdata2)
-			return nfserr_inval;
+			goto out_umask;
 		status = nfsd_create(rqstp, &cstate->current_fh,
 				     create->cr_name, create->cr_namelen,
 				     &create->cr_iattr, S_IFBLK, rdev, &resfh);
 		break;
 
 	case NF4CHR:
+		status = nfserr_inval;
 		rdev = MKDEV(create->cr_specdata1, create->cr_specdata2);
 		if (MAJOR(rdev) != create->cr_specdata1 ||
 		    MINOR(rdev) != create->cr_specdata2)
-			return nfserr_inval;
+			goto out_umask;
 		status = nfsd_create(rqstp, &cstate->current_fh,
 				     create->cr_name, create->cr_namelen,
 				     &create->cr_iattr,S_IFCHR, rdev, &resfh);
@@ -668,6 +674,8 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	fh_dup2(&cstate->current_fh, &resfh);
 out:
 	fh_put(&resfh);
+out_umask:
+	current->fs->umask = 0;
 	return status;
 }
 
@@ -751,6 +759,9 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	if (read->rd_offset >= OFFSET_MAX)
 		return nfserr_inval;
 
+	trace_nfsd_read_start(rqstp, &cstate->current_fh,
+			      read->rd_offset, read->rd_length);
+
 	/*
 	 * If we do a zero copy read, then a client will see read data
 	 * that reflects the state of the file *after* performing the
@@ -783,6 +794,8 @@ nfsd4_read_release(union nfsd4_op_u *u)
 {
 	if (u->read.rd_filp)
 		fput(u->read.rd_filp);
+	trace_nfsd_read_done(u->read.rd_rqstp, u->read.rd_fhp,
+			     u->read.rd_offset, u->read.rd_length);
 }
 
 static __be32
@@ -1001,14 +1014,16 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	if (write->wr_offset >= OFFSET_MAX)
 		return nfserr_inval;
 
+	cnt = write->wr_buflen;
+	trace_nfsd_write_start(rqstp, &cstate->current_fh,
+			       write->wr_offset, cnt);
 	status = nfs4_preprocess_stateid_op(rqstp, cstate, &cstate->current_fh,
 						stateid, WR_STATE, &filp, NULL);
 	if (status) {
 		dprintk("NFSD: nfsd4_write: couldn't process stateid!\n");
 		return status;
 	}
 
-	cnt = write->wr_buflen;
 	write->wr_how_written = write->wr_stable_how;
 	gen_boot_verifier(&write->wr_verifier, SVC_NET(rqstp));
 
@@ -1021,7 +1036,8 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	fput(filp);
 
 	write->wr_bytes_written = cnt;
-
+	trace_nfsd_write_done(rqstp, &cstate->current_fh,
+			      write->wr_offset, cnt);
 	return status;
 }
 
@@ -1106,7 +1122,6 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	else {
 		copy->cp_res.wr_bytes_written = bytes;
 		copy->cp_res.wr_stable_how = NFS_UNSTABLE;
-		copy->cp_consecutive = 1;
 		copy->cp_synchronous = 1;
 		gen_boot_verifier(&copy->cp_res.wr_verifier, SVC_NET(rqstp));
 		status = nfs_ok;
@@ -1412,7 +1427,7 @@ nfsd4_layoutget(struct svc_rqst *rqstp,
 	nfserr = nfsd4_preprocess_layout_stateid(rqstp, cstate, &lgp->lg_sid,
 						true, lgp->lg_layout_type, &ls);
 	if (nfserr) {
-		trace_layout_get_lookup_fail(&lgp->lg_sid);
+		trace_nfsd_layout_get_lookup_fail(&lgp->lg_sid);
 		goto out;
 	}
 
@@ -1481,7 +1496,7 @@ nfsd4_layoutcommit(struct svc_rqst *rqstp,
 						false, lcp->lc_layout_type,
 						&ls);
 	if (nfserr) {
-		trace_layout_commit_lookup_fail(&lcp->lc_sid);
+		trace_nfsd_layout_commit_lookup_fail(&lcp->lc_sid);
 		/* fixup error code as per RFC5661 */
 		if (nfserr == nfserr_bad_stateid)
 			nfserr = nfserr_badlayout;
@@ -1714,12 +1729,10 @@ nfsd4_proc_compound(struct svc_rqst *rqstp)
 		goto encode_op;
 	}
 
+	trace_nfsd_compound(rqstp, args->opcnt);
 	while (!status && resp->opcnt < args->opcnt) {
 		op = &args->ops[resp->opcnt++];
 
-		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
@@ -1793,9 +1806,8 @@ nfsd4_proc_compound(struct svc_rqst *rqstp)
 			status = op->status;
 		}
 
-		dprintk("nfsv4 compound op %p opcnt %d #%d: %d: status %d\n",
-			args->ops, args->opcnt, resp->opcnt, op->opnum,
-			be32_to_cpu(status));
+		trace_nfsd_compound_status(args->opcnt, resp->opcnt, status,
+					   nfsd4_op_name(op->opnum));
 
 		nfsd4_cstate_clear_replay(cstate);
 		nfsd4_increment_op_stats(op->opnum);