Merge tag 'nfsd-5.19' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux

Pull nfsd updates from Chuck Lever:
 "We introduce 'courteous server' in this release. Previously NFSD would
  purge open and lock state for an unresponsive client after one lease
  period (typically 90 seconds). Now, after one lease period, another
  client can open and lock those files and the unresponsive client's
  lease is purged; otherwise if the unresponsive client's open and lock
  state is uncontended, the server retains that open and lock state for
  up to 24 hours, allowing the client's workload to resume after a
  lengthy network partition.

  A longstanding issue with NFSv4 file creation is also addressed.
  Previously a file creation can fail internally, returning an error to
  the client, but leave the newly created file in place as an artifact.
  The file creation code path has been reorganized so that internal
  failures and race conditions are less likely to result in an unwanted
  file creation.

  A fault injector has been added to help exercise paths that are run
  during kernel metadata cache invalidation. These caches contain
  information maintained by user space about exported filesystems. Many
  of our test workloads do not trigger cache invalidation.

  There is one patch that is needed to support PREEMPT_RT and a fix for
  an ancient 'sleep while spin-locked' splat that seems to have become
  easier to hit since v5.18-rc3"

* tag 'nfsd-5.19' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux: (36 commits)
  NFSD: nfsd_file_put() can sleep
  NFSD: Add documenting comment for nfsd4_release_lockowner()
  NFSD: Modernize nfsd4_release_lockowner()
  NFSD: Fix possible sleep during nfsd4_release_lockowner()
  nfsd: destroy percpu stats counters after reply cache shutdown
  nfsd: Fix null-ptr-deref in nfsd_fill_super()
  nfsd: Unregister the cld notifier when laundry_wq create failed
  SUNRPC: Use RMW bitops in single-threaded hot paths
  NFSD: Clean up the show_nf_flags() macro
  NFSD: Trace filecache opens
  NFSD: Move documenting comment for nfsd4_process_open2()
  NFSD: Fix whitespace
  NFSD: Remove dprintk call sites from tail of nfsd4_open()
  NFSD: Instantiate a struct file when creating a regular NFSv4 file
  NFSD: Clean up nfsd_open_verified()
  NFSD: Remove do_nfsd_create()
  NFSD: Refactor NFSv4 OPEN(CREATE)
  NFSD: Refactor NFSv3 CREATE
  NFSD: Refactor nfsd_create_setattr()
  NFSD: Avoid calling fh_drop_write() twice in do_nfsd_create()
  ...

1 files changed, 217 insertions, 47 deletions

diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index b207c76a873f..3895eb52d2b1 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -37,6 +37,8 @@
 #include <linux/falloc.h>
 #include <linux/slab.h>
 #include <linux/kthread.h>
+#include <linux/namei.h>
+
 #include <linux/sunrpc/addr.h>
 #include <linux/nfs_ssc.h>
 
@@ -235,6 +237,183 @@ static void nfsd4_set_open_owner_reply_cache(struct nfsd4_compound_state *cstate
 			&resfh->fh_handle);
 }
 
+static inline bool nfsd4_create_is_exclusive(int createmode)
+{
+	return createmode == NFS4_CREATE_EXCLUSIVE ||
+		createmode == NFS4_CREATE_EXCLUSIVE4_1;
+}
+
+static __be32
+nfsd4_vfs_create(struct svc_fh *fhp, struct dentry *child,
+		 struct nfsd4_open *open)
+{
+	struct file *filp;
+	struct path path;
+	int oflags;
+
+	oflags = O_CREAT | O_LARGEFILE;
+	switch (open->op_share_access & NFS4_SHARE_ACCESS_BOTH) {
+	case NFS4_SHARE_ACCESS_WRITE:
+		oflags |= O_WRONLY;
+		break;
+	case NFS4_SHARE_ACCESS_BOTH:
+		oflags |= O_RDWR;
+		break;
+	default:
+		oflags |= O_RDONLY;
+	}
+
+	path.mnt = fhp->fh_export->ex_path.mnt;
+	path.dentry = child;
+	filp = dentry_create(&path, oflags, open->op_iattr.ia_mode,
+			     current_cred());
+	if (IS_ERR(filp))
+		return nfserrno(PTR_ERR(filp));
+
+	open->op_filp = filp;
+	return nfs_ok;
+}
+
+/*
+ * Implement NFSv4's unchecked, guarded, and exclusive create
+ * semantics for regular files. Open state for this new file is
+ * subsequently fabricated in nfsd4_process_open2().
+ *
+ * Upon return, caller must release @fhp and @resfhp.
+ */
+static __be32
+nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
+		  struct svc_fh *resfhp, struct nfsd4_open *open)
+{
+	struct iattr *iap = &open->op_iattr;
+	struct dentry *parent, *child;
+	__u32 v_mtime, v_atime;
+	struct inode *inode;
+	__be32 status;
+	int host_err;
+
+	if (isdotent(open->op_fname, open->op_fnamelen))
+		return nfserr_exist;
+	if (!(iap->ia_valid & ATTR_MODE))
+		iap->ia_mode = 0;
+
+	status = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_EXEC);
+	if (status != nfs_ok)
+		return status;
+	parent = fhp->fh_dentry;
+	inode = d_inode(parent);
+
+	host_err = fh_want_write(fhp);
+	if (host_err)
+		return nfserrno(host_err);
+
+	fh_lock_nested(fhp, I_MUTEX_PARENT);
+
+	child = lookup_one_len(open->op_fname, parent, open->op_fnamelen);
+	if (IS_ERR(child)) {
+		status = nfserrno(PTR_ERR(child));
+		goto out;
+	}
+
+	if (d_really_is_negative(child)) {
+		status = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE);
+		if (status != nfs_ok)
+			goto out;
+	}
+
+	status = fh_compose(resfhp, fhp->fh_export, child, fhp);
+	if (status != nfs_ok)
+		goto out;
+
+	v_mtime = 0;
+	v_atime = 0;
+	if (nfsd4_create_is_exclusive(open->op_createmode)) {
+		u32 *verifier = (u32 *)open->op_verf.data;
+
+		/*
+		 * Solaris 7 gets confused (bugid 4218508) if these have
+		 * the high bit set, as do xfs filesystems without the
+		 * "bigtime" feature. So just clear the high bits. If this
+		 * is ever changed to use different attrs for storing the
+		 * verifier, then do_open_lookup() will also need to be
+		 * fixed accordingly.
+		 */
+		v_mtime = verifier[0] & 0x7fffffff;
+		v_atime = verifier[1] & 0x7fffffff;
+	}
+
+	if (d_really_is_positive(child)) {
+		status = nfs_ok;
+
+		switch (open->op_createmode) {
+		case NFS4_CREATE_UNCHECKED:
+			if (!d_is_reg(child))
+				break;
+
+			/*
+			 * In NFSv4, we don't want to truncate the file
+			 * now. This would be wrong if the OPEN fails for
+			 * some other reason. Furthermore, if the size is
+			 * nonzero, we should ignore it according to spec!
+			 */
+			open->op_truncate = (iap->ia_valid & ATTR_SIZE) &&
+						!iap->ia_size;
+			break;
+		case NFS4_CREATE_GUARDED:
+			status = nfserr_exist;
+			break;
+		case NFS4_CREATE_EXCLUSIVE:
+			if (d_inode(child)->i_mtime.tv_sec == v_mtime &&
+			    d_inode(child)->i_atime.tv_sec == v_atime &&
+			    d_inode(child)->i_size == 0) {
+				open->op_created = true;
+				break;		/* subtle */
+			}
+			status = nfserr_exist;
+			break;
+		case NFS4_CREATE_EXCLUSIVE4_1:
+			if (d_inode(child)->i_mtime.tv_sec == v_mtime &&
+			    d_inode(child)->i_atime.tv_sec == v_atime &&
+			    d_inode(child)->i_size == 0) {
+				open->op_created = true;
+				goto set_attr;	/* subtle */
+			}
+			status = nfserr_exist;
+		}
+		goto out;
+	}
+
+	if (!IS_POSIXACL(inode))
+		iap->ia_mode &= ~current_umask();
+
+	status = nfsd4_vfs_create(fhp, child, open);
+	if (status != nfs_ok)
+		goto out;
+	open->op_created = true;
+
+	/* A newly created file already has a file size of zero. */
+	if ((iap->ia_valid & ATTR_SIZE) && (iap->ia_size == 0))
+		iap->ia_valid &= ~ATTR_SIZE;
+	if (nfsd4_create_is_exclusive(open->op_createmode)) {
+		iap->ia_valid = ATTR_MTIME | ATTR_ATIME |
+				ATTR_MTIME_SET|ATTR_ATIME_SET;
+		iap->ia_mtime.tv_sec = v_mtime;
+		iap->ia_atime.tv_sec = v_atime;
+		iap->ia_mtime.tv_nsec = 0;
+		iap->ia_atime.tv_nsec = 0;
+	}
+
+set_attr:
+	status = nfsd_create_setattr(rqstp, fhp, resfhp, iap);
+
+out:
+	fh_unlock(fhp);
+	if (child && !IS_ERR(child))
+		dput(child);
+	fh_drop_write(fhp);
+	return status;
+}
+
 static __be32
 do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_open *open, struct svc_fh **resfh)
 {
@@ -264,16 +443,8 @@ do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stru
 		 * yes          | yes    | GUARDED4        | GUARDED4
 		 */
 
-		/*
-		 * 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,
-					open->op_fnamelen, &open->op_iattr,
-					*resfh, open->op_createmode,
-					(u32 *)open->op_verf.data,
-					&open->op_truncate, &open->op_created);
+		status = nfsd4_create_file(rqstp, current_fh, *resfh, open);
 		current->fs->umask = 0;
 
 		if (!status && open->op_label.len)
@@ -284,7 +455,7 @@ do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stru
 		 * use the returned bitmask to indicate which attributes
 		 * we used to store the verifier:
 		 */
-		if (nfsd_create_is_exclusive(open->op_createmode) && status == 0)
+		if (nfsd4_create_is_exclusive(open->op_createmode) && status == 0)
 			open->op_bmval[1] |= (FATTR4_WORD1_TIME_ACCESS |
 						FATTR4_WORD1_TIME_MODIFY);
 	} else
@@ -375,6 +546,8 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		(int)open->op_fnamelen, open->op_fname,
 		open->op_openowner);
 
+	open->op_filp = NULL;
+
 	/* This check required by spec. */
 	if (open->op_create && open->op_claim_type != NFS4_OPEN_CLAIM_NULL)
 		return nfserr_inval;
@@ -427,43 +600,35 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		goto out;
 
 	switch (open->op_claim_type) {
-		case NFS4_OPEN_CLAIM_DELEGATE_CUR:
-		case NFS4_OPEN_CLAIM_NULL:
-			status = do_open_lookup(rqstp, cstate, open, &resfh);
-			if (status)
-				goto out;
-			break;
-		case NFS4_OPEN_CLAIM_PREVIOUS:
-			status = nfs4_check_open_reclaim(cstate->clp);
-			if (status)
-				goto out;
-			open->op_openowner->oo_flags |= NFS4_OO_CONFIRMED;
-			reclaim = true;
-			fallthrough;
-		case NFS4_OPEN_CLAIM_FH:
-		case NFS4_OPEN_CLAIM_DELEG_CUR_FH:
-			status = do_open_fhandle(rqstp, cstate, open);
-			if (status)
-				goto out;
-			resfh = &cstate->current_fh;
-			break;
-		case NFS4_OPEN_CLAIM_DELEG_PREV_FH:
-             	case NFS4_OPEN_CLAIM_DELEGATE_PREV:
-			dprintk("NFSD: unsupported OPEN claim type %d\n",
-				open->op_claim_type);
-			status = nfserr_notsupp;
+	case NFS4_OPEN_CLAIM_DELEGATE_CUR:
+	case NFS4_OPEN_CLAIM_NULL:
+		status = do_open_lookup(rqstp, cstate, open, &resfh);
+		if (status)
 			goto out;
-		default:
-			dprintk("NFSD: Invalid OPEN claim type %d\n",
-				open->op_claim_type);
-			status = nfserr_inval;
+		break;
+	case NFS4_OPEN_CLAIM_PREVIOUS:
+		status = nfs4_check_open_reclaim(cstate->clp);
+		if (status)
+			goto out;
+		open->op_openowner->oo_flags |= NFS4_OO_CONFIRMED;
+		reclaim = true;
+		fallthrough;
+	case NFS4_OPEN_CLAIM_FH:
+	case NFS4_OPEN_CLAIM_DELEG_CUR_FH:
+		status = do_open_fhandle(rqstp, cstate, open);
+		if (status)
 			goto out;
+		resfh = &cstate->current_fh;
+		break;
+	case NFS4_OPEN_CLAIM_DELEG_PREV_FH:
+	case NFS4_OPEN_CLAIM_DELEGATE_PREV:
+		status = nfserr_notsupp;
+		goto out;
+	default:
+		status = nfserr_inval;
+		goto out;
 	}
-	/*
-	 * nfsd4_process_open2() does the actual opening of the file.  If
-	 * successful, it (1) truncates the file if open->op_truncate was
-	 * set, (2) sets open->op_stateid, (3) sets open->op_delegation.
-	 */
+
 	status = nfsd4_process_open2(rqstp, resfh, open);
 	WARN(status && open->op_created,
 	     "nfsd4_process_open2 failed to open newly-created file! status=%u\n",
@@ -471,6 +636,10 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	if (reclaim && !status)
 		nn->somebody_reclaimed = true;
 out:
+	if (open->op_filp) {
+		fput(open->op_filp);
+		open->op_filp = NULL;
+	}
 	if (resfh && resfh != &cstate->current_fh) {
 		fh_dup2(&cstate->current_fh, resfh);
 		fh_put(resfh);
@@ -801,7 +970,7 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	 * the client wants us to do more in this compound:
 	 */
 	if (!nfsd4_last_compound_op(rqstp))
-		clear_bit(RQ_SPLICE_OK, &rqstp->rq_flags);
+		__clear_bit(RQ_SPLICE_OK, &rqstp->rq_flags);
 
 	/* check stateid */
 	status = nfs4_preprocess_stateid_op(rqstp, cstate, &cstate->current_fh,
@@ -2481,11 +2650,12 @@ nfsd4_proc_compound(struct svc_rqst *rqstp)
 	cstate->minorversion = args->minorversion;
 	fh_init(current_fh, NFS4_FHSIZE);
 	fh_init(save_fh, NFS4_FHSIZE);
+
 	/*
 	 * Don't use the deferral mechanism for NFSv4; compounds make it
 	 * too hard to avoid non-idempotency problems.
 	 */
-	clear_bit(RQ_USEDEFERRAL, &rqstp->rq_flags);
+	__clear_bit(RQ_USEDEFERRAL, &rqstp->rq_flags);
 
 	/*
 	 * According to RFC3010, this takes precedence over all other errors.
@@ -2600,7 +2770,7 @@ encode_op:
 out:
 	cstate->status = status;
 	/* Reset deferral mechanism for RPC deferrals */
-	set_bit(RQ_USEDEFERRAL, &rqstp->rq_flags);
+	__set_bit(RQ_USEDEFERRAL, &rqstp->rq_flags);
 	return rpc_success;
 }