Skip to contentSkip to conversation

5.15.154-rc1 review

PATCH In review Greg Kroah-Hartman 8 participants 702 messages Apr 8, 2024 – Apr 10, 2024 Page 4 of 7

Subsystems (8)
lore
Share & tools
lore
author maintainer
#317 lore raw

PATCH 317/690 NFSD: Refactor NFSv3 CREATE

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit df9606abddfb01090d5ece7dcc2441d848f690f0 ]

The NFSv3 CREATE and NFSv4 OPEN(CREATE) use cases are about to diverge such that it makes sense to split do_nfsd_create() into one version for NFSv3 and one for NFSv4.

As a first step, copy do_nfsd_create() to nfs3proc.c and remove NFSv4-specific logic.

One immediate legibility benefit is that the logic for handling NFSv3 createhow is now quite straightforward. NFSv4 createhow has some subtleties that IMO do not belong in generic code.

---
 fs/nfsd/nfs3proc.c | 127 ++++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 121 insertions(+), 6 deletions(-)
fs/nfsd/nfs3proc.c +121 -6
index 86163ecbb015d..57854ca022d18 100644
@@ -8,6 +8,7 @@
 #include <linux/fs.h>
 #include <linux/ext2_fs.h>
 #include <linux/magic.h>
+#include <linux/namei.h>
 
 #include "cache.h"
 #include "xdr3.h"
@@ -220,10 +221,126 @@ nfsd3_proc_write(struct svc_rqst *rqstp)
 }
 
 /*
- * With NFSv3, CREATE processing is a lot easier than with NFSv2.
- * At least in theory; we'll see how it fares in practice when the
- * first reports about SunOS compatibility problems start to pour in...
+ * Implement NFSv3's unchecked, guarded, and exclusive CREATE
+ * semantics for regular files. Except for the created file,
+ * this operation is stateless on the server.
+ *
+ * Upon return, caller must release @fhp and @resfhp.
  */
+static __be32
+nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
+		  struct svc_fh *resfhp, struct nfsd3_createargs *argp)
+{
+	struct iattr *iap = &argp->attrs;
+	struct dentry *parent, *child;
+	__u32 v_mtime, v_atime;
+	struct inode *inode;
+	__be32 status;
+	int host_err;
+
+	if (isdotent(argp->name, argp->len))
+		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(argp->name, parent, argp->len);
+	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 (argp->createmode == NFS3_CREATE_EXCLUSIVE) {
+		u32 *verifier = (u32 *)argp->verf;
+
+		/*
+		 * 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.
+		 */
+		v_mtime = verifier[0] & 0x7fffffff;
+		v_atime = verifier[1] & 0x7fffffff;
+	}
+
+	if (d_really_is_positive(child)) {
+		status = nfs_ok;
+
+		switch (argp->createmode) {
+		case NFS3_CREATE_UNCHECKED:
+			if (!d_is_reg(child))
+				break;
+			iap->ia_valid &= ATTR_SIZE;
+			goto set_attr;
+		case NFS3_CREATE_GUARDED:
+			status = nfserr_exist;
+			break;
+		case NFS3_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) {
+				break;
+			}
+			status = nfserr_exist;
+		}
+		goto out;
+	}
+
+	if (!IS_POSIXACL(inode))
+		iap->ia_mode &= ~current_umask();
+
+	host_err = vfs_create(&init_user_ns, inode, child, iap->ia_mode, true);
+	if (host_err < 0) {
+		status = nfserrno(host_err);
+		goto out;
+	}
+
+	/* 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 (argp->createmode == NFS3_CREATE_EXCLUSIVE) {
+		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
 nfsd3_proc_create(struct svc_rqst *rqstp)
 {
@@ -239,9 +356,7 @@ nfsd3_proc_create(struct svc_rqst *rqstp)
 	dirfhp = fh_copy(&resp->dirfh, &argp->fh);
 	newfhp = fh_init(&resp->fh, NFS3_FHSIZE);
 
-	resp->status = do_nfsd_create(rqstp, dirfhp, argp->name, argp->len,
-				      &argp->attrs, newfhp, argp->createmode,
-				      (u32 *)argp->verf, NULL, NULL);
+	resp->status = nfsd3_create_file(rqstp, dirfhp, newfhp, argp);
 	return rpc_success;
 }
 
Signature · 1 line
2.43.0
author maintainer
#318 lore raw

PATCH 318/690 NFSD: Refactor NFSv4 OPEN(CREATE)

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 254454a5aa4a9f696d6bae080c08d5863e650f49 ]

Copy do_nfsd_create() to nfs4proc.c and remove NFSv3-specific logic.

---
 fs/nfsd/nfs4proc.c | 162 ++++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 152 insertions(+), 10 deletions(-)
fs/nfsd/nfs4proc.c +152 -10
index 73c62561580a1..489cdcd8f8c9a 100644
@@ -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 +237,154 @@ 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;
+}
+
+/*
+ * 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();
+
+	host_err = vfs_create(&init_user_ns, inode, child, iap->ia_mode, true);
+	if (host_err < 0) {
+		status = nfserrno(host_err);
+		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 +414,8 @@ 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 +426,7 @@ 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
Signature · 1 line
2.43.0
author maintainer
#319 lore raw

PATCH 319/690 NFSD: Remove do_nfsd_create()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 1c388f27759c5d9271d4fca081f7ee138986eb7d ]

Now that its two callers have their own version-specific instance of this function, do_nfsd_create() is no longer used.

---
 fs/nfsd/vfs.c | 150 --------------------------------------------------
 fs/nfsd/vfs.h |  10 ----
 2 files changed, 160 deletions(-)
fs/nfsd/vfs.c +0 -150
index e4f100a43ce52..9dd14c0eaebd1 100644
@@ -1395,156 +1395,6 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 					rdev, resfhp);
 }
 
-/*
- * NFSv3 and NFSv4 version of nfsd_create
- */
-__be32
-do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		char *fname, int flen, struct iattr *iap,
-		struct svc_fh *resfhp, int createmode, u32 *verifier,
-	        bool *truncp, bool *created)
-{
-	struct dentry	*dentry, *dchild = NULL;
-	struct inode	*dirp;
-	__be32		err;
-	int		host_err;
-	__u32		v_mtime=0, v_atime=0;
-
-	err = nfserr_perm;
-	if (!flen)
-		goto out;
-	err = nfserr_exist;
-	if (isdotent(fname, flen))
-		goto out;
-	if (!(iap->ia_valid & ATTR_MODE))
-		iap->ia_mode = 0;
-	err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_EXEC);
-	if (err)
-		goto out;
-
-	dentry = fhp->fh_dentry;
-	dirp = d_inode(dentry);
-
-	host_err = fh_want_write(fhp);
-	if (host_err)
-		goto out_nfserr;
-
-	fh_lock_nested(fhp, I_MUTEX_PARENT);
-
-	/*
-	 * Compose the response file handle.
-	 */
-	dchild = lookup_one_len(fname, dentry, flen);
-	host_err = PTR_ERR(dchild);
-	if (IS_ERR(dchild))
-		goto out_nfserr;
-
-	/* If file doesn't exist, check for permissions to create one */
-	if (d_really_is_negative(dchild)) {
-		err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE);
-		if (err)
-			goto out;
-	}
-
-	err = fh_compose(resfhp, fhp->fh_export, dchild, fhp);
-	if (err)
-		goto out;
-
-	if (nfsd_create_is_exclusive(createmode)) {
-		/* solaris7 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(dchild)) {
-		err = 0;
-
-		switch (createmode) {
-		case NFS3_CREATE_UNCHECKED:
-			if (! d_is_reg(dchild))
-				goto out;
-			else if (truncp) {
-				/* in nfsv4, we need to treat this case a little
-				 * differently.  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!
-				 */
-				*truncp = (iap->ia_valid & ATTR_SIZE) && !iap->ia_size;
-			}
-			else {
-				iap->ia_valid &= ATTR_SIZE;
-				goto set_attr;
-			}
-			break;
-		case NFS3_CREATE_EXCLUSIVE:
-			if (   d_inode(dchild)->i_mtime.tv_sec == v_mtime
-			    && d_inode(dchild)->i_atime.tv_sec == v_atime
-			    && d_inode(dchild)->i_size  == 0 ) {
-				if (created)
-					*created = true;
-				break;
-			}
-			fallthrough;
-		case NFS4_CREATE_EXCLUSIVE4_1:
-			if (   d_inode(dchild)->i_mtime.tv_sec == v_mtime
-			    && d_inode(dchild)->i_atime.tv_sec == v_atime
-			    && d_inode(dchild)->i_size  == 0 ) {
-				if (created)
-					*created = true;
-				goto set_attr;
-			}
-			fallthrough;
-		case NFS3_CREATE_GUARDED:
-			err = nfserr_exist;
-		}
-		goto out;
-	}
-
-	if (!IS_POSIXACL(dirp))
-		iap->ia_mode &= ~current_umask();
-
-	host_err = vfs_create(&init_user_ns, dirp, dchild, iap->ia_mode, true);
-	if (host_err < 0)
-		goto out_nfserr;
-	if (created)
-		*created = true;
-
-	nfsd_check_ignore_resizing(iap);
-
-	if (nfsd_create_is_exclusive(createmode)) {
-		/* Cram the verifier into atime/mtime */
-		iap->ia_valid = ATTR_MTIME|ATTR_ATIME
-			| ATTR_MTIME_SET|ATTR_ATIME_SET;
-		/* XXX someone who knows this better please fix it for nsec */ 
-		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:
-	err = nfsd_create_setattr(rqstp, fhp, resfhp, iap);
-
- out:
-	fh_unlock(fhp);
-	if (dchild && !IS_ERR(dchild))
-		dput(dchild);
-	fh_drop_write(fhp);
- 	return err;
- 
- out_nfserr:
-	err = nfserrno(host_err);
-	goto out;
-}
-
 /*
  * Read a symlink. On entry, *lenp must contain the maximum path length that
  * fits into the buffer. On return, it contains the true length.
fs/nfsd/vfs.h +0 -10
index 1f32a83456b03..f99794b033a55 100644
@@ -71,10 +71,6 @@ __be32		nfsd_create(struct svc_rqst *, struct svc_fh *,
 __be32		nfsd_access(struct svc_rqst *, struct svc_fh *, u32 *, u32 *);
 __be32		nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 				struct svc_fh *resfhp, struct iattr *iap);
-__be32		do_nfsd_create(struct svc_rqst *, struct svc_fh *,
-				char *name, int len, struct iattr *attrs,
-				struct svc_fh *res, int createmode,
-				u32 *verifier, bool *truncp, bool *created);
 __be32		nfsd_commit(struct svc_rqst *rqst, struct svc_fh *fhp,
 				u64 offset, u32 count, __be32 *verf);
 #ifdef CONFIG_NFSD_V4
@@ -161,10 +157,4 @@ static inline __be32 fh_getattr(const struct svc_fh *fh, struct kstat *stat)
 				    AT_STATX_SYNC_AS_STAT));
 }
 
-static inline int nfsd_create_is_exclusive(int createmode)
-{
-	return createmode == NFS3_CREATE_EXCLUSIVE
-	       || createmode == NFS4_CREATE_EXCLUSIVE4_1;
-}
-
 #endif /* LINUX_NFSD_VFS_H */
Signature · 1 line
2.43.0
author maintainer
#320 lore raw

PATCH 320/690 NFSD: Clean up nfsd_open_verified()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit f4d84c52643ae1d63a8e73e2585464470e7944d1 ]

Its only caller always passes S_IFREG as the @type parameter. As an additional clean-up, add a kerneldoc comment.

---
 fs/nfsd/filecache.c |  4 ++--
 fs/nfsd/vfs.c       | 15 ++++++++++++---
 fs/nfsd/vfs.h       |  2 +-
 3 files changed, 15 insertions(+), 6 deletions(-)
fs/nfsd/filecache.c +2 -2
index b1afe6db589f2..0f6553b316f58 100644
@@ -976,8 +976,8 @@ nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 	nf->nf_mark = nfsd_file_mark_find_or_create(nf);
 	if (nf->nf_mark)
-		status = nfsd_open_verified(rqstp, fhp, S_IFREG,
-				may_flags, &nf->nf_file);
+		status = nfsd_open_verified(rqstp, fhp, may_flags,
+					    &nf->nf_file);
 	else
 		status = nfserr_jukebox;
 	/*
fs/nfsd/vfs.c +12 -3
index 9dd14c0eaebd1..6689ad5bb790d 100644
@@ -834,14 +834,23 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type,
 	return err;
 }
 
+/**
+ * nfsd_open_verified - Open a regular file for the filecache
+ * @rqstp: RPC request
+ * @fhp: NFS filehandle of the file to open
+ * @may_flags: internal permission flags
+ * @filp: OUT: open "struct file *"
+ *
+ * Returns an nfsstat value in network byte order.
+ */
 __be32
-nfsd_open_verified(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type,
-		int may_flags, struct file **filp)
+nfsd_open_verified(struct svc_rqst *rqstp, struct svc_fh *fhp, int may_flags,
+		   struct file **filp)
 {
 	__be32 err;
 
 	validate_process_creds();
-	err = __nfsd_open(rqstp, fhp, type, may_flags, filp);
+	err = __nfsd_open(rqstp, fhp, S_IFREG, may_flags, filp);
 	validate_process_creds();
 	return err;
 }
fs/nfsd/vfs.h +1 -1
index f99794b033a55..26347d76f44a0 100644
@@ -86,7 +86,7 @@ __be32		nfsd_setxattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 int 		nfsd_open_break_lease(struct inode *, int);
 __be32		nfsd_open(struct svc_rqst *, struct svc_fh *, umode_t,
 				int, struct file **);
-__be32		nfsd_open_verified(struct svc_rqst *, struct svc_fh *, umode_t,
+__be32		nfsd_open_verified(struct svc_rqst *, struct svc_fh *,
 				int, struct file **);
 __be32		nfsd_splice_read(struct svc_rqst *rqstp, struct svc_fh *fhp,
 				struct file *file, loff_t offset,
Signature · 1 line
2.43.0
author maintainer
#321 lore raw

PATCH 321/690 NFSD: Instantiate a struct file when creating a regular NFSv4 file

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit fb70bf124b051d4ded4ce57511dfec6d3ebf2b43 ]

There have been reports of races that cause NFSv4 OPEN(CREATE) to return an error even though the requested file was created. NFSv4 does not provide a status code for this case.

To mitigate some of these problems, reorganize the NFSv4 OPEN(CREATE) logic to allocate resources before the file is actually created, and open the new file while the parent directory is still locked.

Two new APIs are added:

+ Add an API that works like nfsd_file_acquire() but does not open the underlying file. The OPEN(CREATE) path can use this API when it already has an open file.

+ Add an API that is kin to dentry_open(). NFSD needs to create a file and grab an open "struct file *" atomically. The alloc_empty_file() has to be done before the inode create. If it fails (for example, because the NFS server has exceeded its max_files limit), we avoid creating the file and can still return an error to the NFS client.

---
 fs/nfsd/filecache.c | 51 ++++++++++++++++++++++++++++++++++++++-------
 fs/nfsd/filecache.h |  2 ++
 fs/nfsd/nfs4proc.c  | 43 ++++++++++++++++++++++++++++++++++----
 fs/nfsd/nfs4state.c | 16 +++++++++++---
 fs/nfsd/xdr4.h      |  1 +
 fs/open.c           | 42 +++++++++++++++++++++++++++++++++++++
 include/linux/fs.h  |  2 ++
 7 files changed, 143 insertions(+), 14 deletions(-)
fs/nfsd/filecache.c +44 -7
index 0f6553b316f58..045f5a869ddc7 100644
@@ -892,9 +892,9 @@ nfsd_file_is_cached(struct inode *inode)
 	return ret;
 }
 
-__be32
-nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		  unsigned int may_flags, struct nfsd_file **pnf)
+static __be32
+nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
+		     unsigned int may_flags, struct nfsd_file **pnf, bool open)
 {
 	__be32	status;
 	struct net *net = SVC_NET(rqstp);
@@ -975,10 +975,13 @@ nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		nfsd_file_gc();
 
 	nf->nf_mark = nfsd_file_mark_find_or_create(nf);
-	if (nf->nf_mark)
-		status = nfsd_open_verified(rqstp, fhp, may_flags,
-					    &nf->nf_file);
-	else
+	if (nf->nf_mark) {
+		if (open)
+			status = nfsd_open_verified(rqstp, fhp, may_flags,
+						    &nf->nf_file);
+		else
+			status = nfs_ok;
+	} else
 		status = nfserr_jukebox;
 	/*
 	 * If construction failed, or we raced with a call to unlink()
@@ -998,6 +1001,40 @@ nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	goto out;
 }
 
+/**
+ * nfsd_file_acquire - Get a struct nfsd_file with an open file
+ * @rqstp: the RPC transaction being executed
+ * @fhp: the NFS filehandle of the file to be opened
+ * @may_flags: NFSD_MAY_ settings for the file
+ * @pnf: OUT: new or found "struct nfsd_file" object
+ *
+ * Returns nfs_ok and sets @pnf on success; otherwise an nfsstat in
+ * network byte order is returned.
+ */
+__be32
+nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
+		  unsigned int may_flags, struct nfsd_file **pnf)
+{
+	return nfsd_do_file_acquire(rqstp, fhp, may_flags, pnf, true);
+}
+
+/**
+ * nfsd_file_create - Get a struct nfsd_file, do not open
+ * @rqstp: the RPC transaction being executed
+ * @fhp: the NFS filehandle of the file just created
+ * @may_flags: NFSD_MAY_ settings for the file
+ * @pnf: OUT: new or found "struct nfsd_file" object
+ *
+ * Returns nfs_ok and sets @pnf on success; otherwise an nfsstat in
+ * network byte order is returned.
+ */
+__be32
+nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
+		 unsigned int may_flags, struct nfsd_file **pnf)
+{
+	return nfsd_do_file_acquire(rqstp, fhp, may_flags, pnf, false);
+}
+
 /*
  * Note that fields may be added, removed or reordered in the future. Programs
  * scraping this file for info should test the labels to ensure they're
fs/nfsd/filecache.h +2 -0
index 63104be2865c5..c9e3c6eb4776e 100644
@@ -57,5 +57,7 @@ void nfsd_file_close_inode_sync(struct inode *inode);
 bool nfsd_file_is_cached(struct inode *inode);
 __be32 nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		  unsigned int may_flags, struct nfsd_file **nfp);
+__be32 nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
+		  unsigned int may_flags, struct nfsd_file **nfp);
 int	nfsd_file_cache_stats_open(struct inode *, struct file *);
 #endif /* _FS_NFSD_FILECACHE_H */
fs/nfsd/nfs4proc.c +39 -4

68 lines not on this page · raw message

fs/nfsd/nfs4state.c +13 -3
index 5bbf769b688bc..f9681a4d116ad 100644
@@ -5110,9 +5110,19 @@ static __be32 nfs4_get_vfs_file(struct svc_rqst *rqstp, struct nfs4_file *fp,
 
 	if (!fp->fi_fds[oflag]) {
 		spin_unlock(&fp->fi_lock);
-		status = nfsd_file_acquire(rqstp, cur_fh, access, &nf);
-		if (status)
-			goto out_put_access;
+
+		if (!open->op_filp) {
+			status = nfsd_file_acquire(rqstp, cur_fh, access, &nf);
+			if (status != nfs_ok)
+				goto out_put_access;
+		} else {
+			status = nfsd_file_create(rqstp, cur_fh, access, &nf);
+			if (status != nfs_ok)
+				goto out_put_access;
+			nf->nf_file = open->op_filp;
+			open->op_filp = NULL;
+		}
+
 		spin_lock(&fp->fi_lock);
 		if (!fp->fi_fds[oflag]) {
 			fp->fi_fds[oflag] = nf;
fs/nfsd/xdr4.h +1 -0
index 4f1090c32c29b..77286e8c9ab02 100644
@@ -273,6 +273,7 @@ struct nfsd4_open {
 	bool		op_truncate;        /* used during processing */
 	bool		op_created;         /* used during processing */
 	struct nfs4_openowner *op_openowner; /* used during processing */
+	struct file	*op_filp;           /* used during processing */
 	struct nfs4_file *op_file;          /* used during processing */
 	struct nfs4_ol_stateid *op_stp;	    /* used during processing */
 	struct nfs4_clnt_odstate *op_odstate; /* used during processing */
fs/open.c +42 -0

48 lines not on this page · raw message

include/linux/fs.h +2 -0
index 371d67c9221c5..ef5a04d626953 100644
@@ -2755,6 +2755,8 @@ static inline struct file *file_open_root_mnt(struct vfsmount *mnt,
 			      name, flags, mode);
 }
 extern struct file * dentry_open(const struct path *, int, const struct cred *);
+extern struct file *dentry_create(const struct path *path, int flags,
+				  umode_t mode, const struct cred *cred);
 extern struct file * open_with_fake_path(const struct path *, int,
 					 struct inode*, const struct cred *);
 static inline struct file *file_clone_open(struct file *file)
Signature · 1 line
2.43.0
author maintainer
#322 lore raw

PATCH 322/690 NFSD: Remove dprintk call sites from tail of nfsd4_open()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit f67a16b147045815b6aaafeef8663e5faeb6d569 ]

Clean up: These relics are not likely to benefit server administrators.

---
 fs/nfsd/nfs4proc.c | 4 ----
 1 file changed, 4 deletions(-)
fs/nfsd/nfs4proc.c +0 -4
index f0cb92466da84..611aedeab406b 100644
@@ -622,13 +622,9 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 			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;
 			goto out;
 		default:
-			dprintk("NFSD: Invalid OPEN claim type %d\n",
-				open->op_claim_type);
 			status = nfserr_inval;
 			goto out;
 	}
Signature · 1 line
2.43.0
author maintainer
#323 lore raw

PATCH 323/690 NFSD: Fix whitespace

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 26320d7e317c37404c811603d50d811132aef78c ]

Clean up: Pull case arms back one tab stop to conform every other switch statement in fs/nfsd/nfs4proc.c.

---
 fs/nfsd/nfs4proc.c | 50 +++++++++++++++++++++++-----------------------
 1 file changed, 25 insertions(+), 25 deletions(-)
fs/nfsd/nfs4proc.c +25 -25
index 611aedeab406b..6a9c7e6a23ba5 100644
@@ -600,33 +600,33 @@ 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:
-			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:
-			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
Signature · 1 line
2.43.0
author maintainer
#324 lore raw

PATCH 324/690 NFSD: Move documenting comment for nfsd4_process_open2()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 7e2ce0cc15a509b859199235a2bad9cece00f67a ]

Clean up nfsd4_open() by converting a large comment at the only call site for nfsd4_process_open2() to a kerneldoc comment in front of that function.

---
 fs/nfsd/nfs4proc.c  |  6 +-----
 fs/nfsd/nfs4state.c | 12 ++++++++++++
 2 files changed, 13 insertions(+), 5 deletions(-)
fs/nfsd/nfs4proc.c +1 -5
index 6a9c7e6a23ba5..3ac2978c596ae 100644
@@ -628,11 +628,7 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		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",
fs/nfsd/nfs4state.c +12 -0
index f9681a4d116ad..d79b736019d49 100644
@@ -5465,6 +5465,18 @@ static void nfsd4_deleg_xgrade_none_ext(struct nfsd4_open *open,
 	 */
 }
 
+/**
+ * nfsd4_process_open2 - finish open processing
+ * @rqstp: the RPC transaction being executed
+ * @current_fh: NFSv4 COMPOUND's current filehandle
+ * @open: OPEN arguments
+ *
+ * If successful, (1) truncate the file if open->op_truncate was
+ * set, (2) set open->op_stateid, (3) set open->op_delegation.
+ *
+ * Returns %nfs_ok on success; otherwise an nfs4stat value in
+ * network byte order is returned.
+ */
 __be32
 nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_open *open)
 {
Signature · 1 line
2.43.0
author maintainer
#325 lore raw

PATCH 325/690 NFSD: Trace filecache opens

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 0122e882119ddbd9efa6edfeeac3f5c704a7aeea ]

Instrument calls to nfsd_open_verified() to get a sense of the filecache hit rate.

---
 fs/nfsd/filecache.c |  5 +++--
 fs/nfsd/trace.h     | 28 ++++++++++++++++++++++++++++
 2 files changed, 31 insertions(+), 2 deletions(-)
fs/nfsd/filecache.c +3 -2
index 045f5a869ddc7..0863bf5050935 100644
@@ -976,10 +976,11 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 	nf->nf_mark = nfsd_file_mark_find_or_create(nf);
 	if (nf->nf_mark) {
-		if (open)
+		if (open) {
 			status = nfsd_open_verified(rqstp, fhp, may_flags,
 						    &nf->nf_file);
-		else
+			trace_nfsd_file_open(nf, status);
+		} else
 			status = nfs_ok;
 	} else
 		status = nfserr_jukebox;
fs/nfsd/trace.h +28 -0
index 7f3f40f6c0ff3..3cff3ada00a85 100644
@@ -775,6 +775,34 @@ TRACE_EVENT(nfsd_file_acquire,
 			__entry->nf_file, __entry->status)
 );
 
+TRACE_EVENT(nfsd_file_open,
+	TP_PROTO(struct nfsd_file *nf, __be32 status),
+	TP_ARGS(nf, status),
+	TP_STRUCT__entry(
+		__field(unsigned int, nf_hashval)
+		__field(void *, nf_inode)	/* cannot be dereferenced */
+		__field(int, nf_ref)
+		__field(unsigned long, nf_flags)
+		__field(unsigned long, nf_may)
+		__field(void *, nf_file)	/* cannot be dereferenced */
+	),
+	TP_fast_assign(
+		__entry->nf_hashval = nf->nf_hashval;
+		__entry->nf_inode = nf->nf_inode;
+		__entry->nf_ref = refcount_read(&nf->nf_ref);
+		__entry->nf_flags = nf->nf_flags;
+		__entry->nf_may = nf->nf_may;
+		__entry->nf_file = nf->nf_file;
+	),
+	TP_printk("hash=0x%x inode=%p ref=%d flags=%s may=%s file=%p",
+		__entry->nf_hashval,
+		__entry->nf_inode,
+		__entry->nf_ref,
+		show_nf_flags(__entry->nf_flags),
+		show_nfsd_may_flags(__entry->nf_may),
+		__entry->nf_file)
+)
+
 DECLARE_EVENT_CLASS(nfsd_file_search_class,
 	TP_PROTO(struct inode *inode, unsigned int hash, int found),
 	TP_ARGS(inode, hash, found),
Signature · 1 line
2.43.0
author maintainer
#326 lore raw

PATCH 326/690 SUNRPC: Use RMW bitops in single-threaded hot paths

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 28df0988815f63e2af5e6718193c9f68681ad7ff ]

I noticed CPU pipeline stalls while using perf.

Once an svc thread is scheduled and executing an RPC, no other processes will touch svc_rqst::rq_flags. Thus bus-locked atomics are not needed outside the svc thread scheduler.

---
 fs/nfsd/nfs4proc.c                       | 7 ++++---
 fs/nfsd/nfs4xdr.c                        | 2 +-
 net/sunrpc/auth_gss/svcauth_gss.c        | 4 ++--
 net/sunrpc/svc.c                         | 6 +++---
 net/sunrpc/svc_xprt.c                    | 2 +-
 net/sunrpc/svcsock.c                     | 8 ++++----
 net/sunrpc/xprtrdma/svc_rdma_transport.c | 2 +-
 7 files changed, 16 insertions(+), 15 deletions(-)
fs/nfsd/nfs4proc.c +4 -3
index 3ac2978c596ae..5b56877c7fb57 100644
@@ -970,7 +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,
@@ -2644,11 +2644,12 @@ 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.
@@ -2770,7 +2771,7 @@ nfsd4_proc_compound(struct svc_rqst *rqstp)
 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;
 }
 
fs/nfsd/nfs4xdr.c +1 -1
index 771d3057577ef..96d41b1cc2d17 100644
@@ -2408,7 +2408,7 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp)
 	argp->rqstp->rq_cachetype = cachethis ? RC_REPLBUFF : RC_NOCACHE;
 
 	if (readcount > 1 || max_reply > PAGE_SIZE - auth_slack)
-		clear_bit(RQ_SPLICE_OK, &argp->rqstp->rq_flags);
+		__clear_bit(RQ_SPLICE_OK, &argp->rqstp->rq_flags);
 
 	return true;
 }
net/sunrpc/auth_gss/svcauth_gss.c +2 -2
index 48b608cb5f5ec..2381c5d1b0710 100644
@@ -900,7 +900,7 @@ unwrap_integ_data(struct svc_rqst *rqstp, struct xdr_buf *buf, u32 seq, struct g
 	 * rejecting the server-computed MIC in this somewhat rare case,
 	 * do not use splice with the GSS integrity service.
 	 */
-	clear_bit(RQ_SPLICE_OK, &rqstp->rq_flags);
+	__clear_bit(RQ_SPLICE_OK, &rqstp->rq_flags);
 
 	/* Did we already verify the signature on the original pass through? */
 	if (rqstp->rq_deferred)
@@ -972,7 +972,7 @@ unwrap_priv_data(struct svc_rqst *rqstp, struct xdr_buf *buf, u32 seq, struct gs
 	int pad, remaining_len, offset;
 	u32 rseqno;
 
-	clear_bit(RQ_SPLICE_OK, &rqstp->rq_flags);
+	__clear_bit(RQ_SPLICE_OK, &rqstp->rq_flags);
 
 	priv_len = svc_getnl(&buf->head[0]);
 	if (rqstp->rq_deferred) {
net/sunrpc/svc.c +3 -3
index 87da3ff46ce9a..f2a8c1ee8530e 100644
@@ -1281,10 +1281,10 @@ svc_process_common(struct svc_rqst *rqstp, struct kvec *argv, struct kvec *resv)
 		goto err_short_len;
 
 	/* Will be turned off by GSS integrity and privacy services */
-	set_bit(RQ_SPLICE_OK, &rqstp->rq_flags);
+	__set_bit(RQ_SPLICE_OK, &rqstp->rq_flags);
 	/* Will be turned off only when NFSv4 Sessions are used */
-	set_bit(RQ_USEDEFERRAL, &rqstp->rq_flags);
-	clear_bit(RQ_DROPME, &rqstp->rq_flags);
+	__set_bit(RQ_USEDEFERRAL, &rqstp->rq_flags);
+	__clear_bit(RQ_DROPME, &rqstp->rq_flags);
 
 	svc_putu32(resv, rqstp->rq_xid);
 
net/sunrpc/svc_xprt.c +1 -1
index 67ccf1a6459ae..39acef5134f5c 100644
@@ -1251,7 +1251,7 @@ static struct cache_deferred_req *svc_defer(struct cache_req *req)
 	trace_svc_defer(rqstp);
 	svc_xprt_get(rqstp->rq_xprt);
 	dr->xprt = rqstp->rq_xprt;
-	set_bit(RQ_DROPME, &rqstp->rq_flags);
+	__set_bit(RQ_DROPME, &rqstp->rq_flags);
 
 	dr->handle.revisit = svc_revisit;
 	return &dr->handle;
net/sunrpc/svcsock.c +4 -4
index be7081284a098..46cea0e413aeb 100644
@@ -298,9 +298,9 @@ static void svc_sock_setbufsize(struct svc_sock *svsk, unsigned int nreqs)
 static void svc_sock_secure_port(struct svc_rqst *rqstp)
 {
 	if (svc_port_is_privileged(svc_addr(rqstp)))
-		set_bit(RQ_SECURE, &rqstp->rq_flags);
+		__set_bit(RQ_SECURE, &rqstp->rq_flags);
 	else
-		clear_bit(RQ_SECURE, &rqstp->rq_flags);
+		__clear_bit(RQ_SECURE, &rqstp->rq_flags);
 }
 
 /*
@@ -1005,9 +1005,9 @@ static int svc_tcp_recvfrom(struct svc_rqst *rqstp)
 	rqstp->rq_xprt_ctxt   = NULL;
 	rqstp->rq_prot	      = IPPROTO_TCP;
 	if (test_bit(XPT_LOCAL, &svsk->sk_xprt.xpt_flags))
-		set_bit(RQ_LOCAL, &rqstp->rq_flags);
+		__set_bit(RQ_LOCAL, &rqstp->rq_flags);
 	else
-		clear_bit(RQ_LOCAL, &rqstp->rq_flags);
+		__clear_bit(RQ_LOCAL, &rqstp->rq_flags);
 
 	p = (__be32 *)rqstp->rq_arg.head[0].iov_base;
 	calldir = p[1];
net/sunrpc/xprtrdma/svc_rdma_transport.c +1 -1
index f776f0cb471f0..ac147304fb0e9 100644
@@ -602,7 +602,7 @@ static int svc_rdma_has_wspace(struct svc_xprt *xprt)
 
 static void svc_rdma_secure_port(struct svc_rqst *rqstp)
 {
-	set_bit(RQ_SECURE, &rqstp->rq_flags);
+	__set_bit(RQ_SECURE, &rqstp->rq_flags);
 }
 
 static void svc_rdma_kill_temp_xprt(struct svc_xprt *xprt)
Signature · 1 line
2.43.0
author maintainer
#327 lore raw

PATCH 327/690 nfsd: Unregister the cld notifier when laundry_wq create failed

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Zhang Xiaoxu <[email protected]>

[ Upstream commit 62fdb65edb6c43306c774939001f3a00974832aa ]

If laundry_wq create failed, the cld notifier should be unregistered.

---
 fs/nfsd/nfsctl.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
fs/nfsd/nfsctl.c +3 -1
index 322a208878f2c..55949e60897d5 100644
@@ -1543,12 +1543,14 @@ static int __init init_nfsd(void)
 		goto out_free_filesystem;
 	retval = register_cld_notifier();
 	if (retval)
-		goto out_free_all;
+		goto out_free_subsys;
 	retval = nfsd4_create_laundry_wq();
 	if (retval)
 		goto out_free_all;
 	return 0;
 out_free_all:
+	unregister_cld_notifier();
+out_free_subsys:
 	unregister_pernet_subsys(&nfsd_net_ops);
 out_free_filesystem:
 	unregister_filesystem(&nfsd_fs_type);
Signature · 1 line
2.43.0
author maintainer
#328 lore raw

PATCH 328/690 nfsd: Fix null-ptr-deref in nfsd_fill_super()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Zhang Xiaoxu <[email protected]>

[ Upstream commit 6f6f84aa215f7b6665ccbb937db50860f9ec2989 ]

KASAN report null-ptr-deref as follows:

  BUG: KASAN: null-ptr-deref in nfsd_fill_super+0xc6/0xe0 [nfsd]
  Write of size 8 at addr 000000000000005d by task a.out/852

  CPU: 7 PID: 852 Comm: a.out Not tainted 5.18.0-rc7-dirty #66
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-1.fc33 04/01/2014
  Call Trace:
   <TASK>
   dump_stack_lvl+0x34/0x44
   kasan_report+0xab/0x120
   ? nfsd_mkdir+0x71/0x1c0 [nfsd]
   ? nfsd_fill_super+0xc6/0xe0 [nfsd]
   nfsd_fill_super+0xc6/0xe0 [nfsd]
   ? nfsd_mkdir+0x1c0/0x1c0 [nfsd]
   get_tree_keyed+0x8e/0x100
   vfs_get_tree+0x41/0xf0
   __do_sys_fsconfig+0x590/0x670
   ? fscontext_read+0x180/0x180
   ? anon_inode_getfd+0x4f/0x70
   do_syscall_64+0x35/0x80
   entry_SYSCALL_64_after_hwframe+0x44/0xae

This can be reproduce by concurrent operations:
1. fsopen(nfsd)/fsconfig
2. insmod/rmmod nfsd

Since the nfsd file system is registered before than nfsd_net allocated, the caller may get the file_system_type and use the nfsd_net before it allocated, then null-ptr-deref occurred.

So init_nfsd() should call register_filesystem() last.

Fixes bd5ae9288d64 ("nfsd: register pernet ops last, unregister first")
---
 fs/nfsd/nfsctl.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
fs/nfsd/nfsctl.c +7 -7
index 55949e60897d5..0621c2faf2424 100644
@@ -1535,25 +1535,25 @@ static int __init init_nfsd(void)
 	retval = create_proc_exports_entry();
 	if (retval)
 		goto out_free_lockd;
-	retval = register_filesystem(&nfsd_fs_type);
-	if (retval)
-		goto out_free_exports;
 	retval = register_pernet_subsys(&nfsd_net_ops);
 	if (retval < 0)
-		goto out_free_filesystem;
+		goto out_free_exports;
 	retval = register_cld_notifier();
 	if (retval)
 		goto out_free_subsys;
 	retval = nfsd4_create_laundry_wq();
+	if (retval)
+		goto out_free_cld;
+	retval = register_filesystem(&nfsd_fs_type);
 	if (retval)
 		goto out_free_all;
 	return 0;
 out_free_all:
+	nfsd4_destroy_laundry_wq();
+out_free_cld:
 	unregister_cld_notifier();
 out_free_subsys:
 	unregister_pernet_subsys(&nfsd_net_ops);
-out_free_filesystem:
-	unregister_filesystem(&nfsd_fs_type);
 out_free_exports:
 	remove_proc_entry("fs/nfs/exports", NULL);
 	remove_proc_entry("fs/nfs", NULL);
@@ -1571,6 +1571,7 @@ static int __init init_nfsd(void)
 
 static void __exit exit_nfsd(void)
 {
+	unregister_filesystem(&nfsd_fs_type);
 	nfsd4_destroy_laundry_wq();
 	unregister_cld_notifier();
 	unregister_pernet_subsys(&nfsd_net_ops);
@@ -1581,7 +1582,6 @@ static void __exit exit_nfsd(void)
 	nfsd_lockd_shutdown();
 	nfsd4_free_slabs();
 	nfsd4_exit_pnfs();
-	unregister_filesystem(&nfsd_fs_type);
 }
 
 MODULE_AUTHOR("Olaf Kirch <[email protected]>");
Signature · 1 line
2.43.0
author maintainer
#329 lore raw

PATCH 329/690 NFSD: Modernize nfsd4_release_lockowner()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit bd8fdb6e545f950f4654a9a10d7e819ad48146e5 ]

Refactor: Use existing helpers that other lock operations use. This change removes several automatic variables, so re-organize the variable declarations for readability.

---
 fs/nfsd/nfs4state.c | 36 +++++++++++-------------------------
 1 file changed, 11 insertions(+), 25 deletions(-)
fs/nfsd/nfs4state.c +11 -25
index d79b736019d49..2d52656095340 100644
@@ -7548,16 +7548,13 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp,
 			union nfsd4_op_u *u)
 {
 	struct nfsd4_release_lockowner *rlockowner = &u->release_lockowner;
+	struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id);
 	clientid_t *clid = &rlockowner->rl_clientid;
-	struct nfs4_stateowner *sop;
-	struct nfs4_lockowner *lo = NULL;
 	struct nfs4_ol_stateid *stp;
-	struct xdr_netobj *owner = &rlockowner->rl_owner;
-	unsigned int hashval = ownerstr_hashval(owner);
-	__be32 status;
-	struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id);
+	struct nfs4_lockowner *lo;
 	struct nfs4_client *clp;
-	LIST_HEAD (reaplist);
+	LIST_HEAD(reaplist);
+	__be32 status;
 
 	dprintk("nfsd4_release_lockowner clientid: (%08x/%08x):\n",
 		clid->cl_boot, clid->cl_id);
@@ -7565,30 +7562,19 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp,
 	status = set_client(clid, cstate, nn);
 	if (status)
 		return status;
-
 	clp = cstate->clp;
-	/* Find the matching lock stateowner */
-	spin_lock(&clp->cl_lock);
-	list_for_each_entry(sop, &clp->cl_ownerstr_hashtbl[hashval],
-			    so_strhash) {
 
-		if (sop->so_is_open_owner || !same_owner_str(sop, owner))
-			continue;
-
-		if (atomic_read(&sop->so_count) != 1) {
-			spin_unlock(&clp->cl_lock);
-			return nfserr_locks_held;
-		}
-
-		lo = lockowner(sop);
-		nfs4_get_stateowner(sop);
-		break;
-	}
+	spin_lock(&clp->cl_lock);
+	lo = find_lockowner_str_locked(clp, &rlockowner->rl_owner);
 	if (!lo) {
 		spin_unlock(&clp->cl_lock);
 		return status;
 	}
-
+	if (atomic_read(&lo->lo_owner.so_count) != 2) {
+		spin_unlock(&clp->cl_lock);
+		nfs4_put_stateowner(&lo->lo_owner);
+		return nfserr_locks_held;
+	}
 	unhash_lockowner_locked(lo);
 	while (!list_empty(&lo->lo_owner.so_stateids)) {
 		stp = list_first_entry(&lo->lo_owner.so_stateids,
Signature · 1 line
2.43.0
author maintainer
#330 lore raw

PATCH 330/690 NFSD: Add documenting comment for nfsd4_release_lockowner()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 043862b09cc00273e35e6c3a6389957953a34207 ]

And return explicit nfserr values that match what is documented in the new comment / API contract.

---
 fs/nfsd/nfs4state.c | 23 ++++++++++++++++++++---
 1 file changed, 20 insertions(+), 3 deletions(-)
fs/nfsd/nfs4state.c +20 -3
index 2d52656095340..08700b6acba31 100644
@@ -7542,6 +7542,23 @@ check_for_locks(struct nfs4_file *fp, struct nfs4_lockowner *lowner)
 	return status;
 }
 
+/**
+ * nfsd4_release_lockowner - process NFSv4.0 RELEASE_LOCKOWNER operations
+ * @rqstp: RPC transaction
+ * @cstate: NFSv4 COMPOUND state
+ * @u: RELEASE_LOCKOWNER arguments
+ *
+ * The lockowner's so_count is bumped when a lock record is added
+ * or when copying a conflicting lock. The latter case is brief,
+ * but can lead to fleeting false positives when looking for
+ * locks-in-use.
+ *
+ * Return values:
+ *   %nfs_ok: lockowner released or not found
+ *   %nfserr_locks_held: lockowner still in use
+ *   %nfserr_stale_clientid: clientid no longer active
+ *   %nfserr_expired: clientid not recognized
+ */
 __be32
 nfsd4_release_lockowner(struct svc_rqst *rqstp,
 			struct nfsd4_compound_state *cstate,
@@ -7568,7 +7585,7 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp,
 	lo = find_lockowner_str_locked(clp, &rlockowner->rl_owner);
 	if (!lo) {
 		spin_unlock(&clp->cl_lock);
-		return status;
+		return nfs_ok;
 	}
 	if (atomic_read(&lo->lo_owner.so_count) != 2) {
 		spin_unlock(&clp->cl_lock);
@@ -7584,11 +7601,11 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp,
 		put_ol_stateid_locked(stp, &reaplist);
 	}
 	spin_unlock(&clp->cl_lock);
+
 	free_ol_stateid_reaplist(&reaplist);
 	remove_blocked_locks(lo);
 	nfs4_put_stateowner(&lo->lo_owner);
-
-	return status;
+	return nfs_ok;
 }
 
 static inline struct nfs4_client_reclaim *
Signature · 1 line
2.43.0
author maintainer
#331 lore raw

PATCH 331/690 NFSD: nfsd_file_put() can sleep

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 08af54b3e5729bc1d56ad3190af811301bdc37a1 ]

Now that there are no more callers of nfsd_file_put() that might hold a spin lock, ensure the lockdep infrastructure can catch newly introduced calls to nfsd_file_put() made while a spinlock is held.

---
 fs/nfsd/filecache.c | 2 ++
 1 file changed, 2 insertions(+)
fs/nfsd/filecache.c +2 -0
index 0863bf5050935..27952e2f3aa14 100644
@@ -296,6 +296,8 @@ nfsd_file_put_noref(struct nfsd_file *nf)
 void
 nfsd_file_put(struct nfsd_file *nf)
 {
+	might_sleep();
+
 	set_bit(NFSD_FILE_REFERENCED, &nf->nf_flags);
 	if (test_bit(NFSD_FILE_HASHED, &nf->nf_flags) == 0) {
 		nfsd_file_flush(nf);
Signature · 1 line
2.43.0
author maintainer
#332 lore raw

PATCH 332/690 NFSD: Fix potential use-after-free in nfsd_file_put()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit b6c71c66b0ad8f2b59d9bc08c7a5079b110bec01 ]

nfsd_file_put_noref() can free @nf, so don't dereference @nf immediately upon return from nfsd_file_put_noref().

Fixes 999397926ab3 ("nfsd: Clean up nfsd_file_put()")
---
 fs/nfsd/filecache.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
fs/nfsd/filecache.c +5 -4
index 27952e2f3aa14..1d3d13b78be0e 100644
@@ -302,11 +302,12 @@ nfsd_file_put(struct nfsd_file *nf)
 	if (test_bit(NFSD_FILE_HASHED, &nf->nf_flags) == 0) {
 		nfsd_file_flush(nf);
 		nfsd_file_put_noref(nf);
-	} else {
+	} else if (nf->nf_file) {
 		nfsd_file_put_noref(nf);
-		if (nf->nf_file)
-			nfsd_file_schedule_laundrette();
-	}
+		nfsd_file_schedule_laundrette();
+	} else
+		nfsd_file_put_noref(nf);
+
 	if (atomic_long_read(&nfsd_filecache_count) >= NFSD_FILE_LRU_LIMIT)
 		nfsd_file_gc();
 }
Signature · 1 line
2.43.0
author maintainer
#333 lore raw

PATCH 333/690 NFS: restore module put when manager exits.

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit 080abad71e99d2becf38c978572982130b927a28 ]

Commit f49169c97fce ("NFSD: Remove svc_serv_ops::svo_module") removed calls to module_put_and_kthread_exit() from threads that acted as SUNRPC servers and had a related svc_serv_ops structure. This was correct.

It ALSO removed the module_put_and_kthread_exit() call from nfs4_run_state_manager() which is NOT a SUNRPC service.

Consequently every time the NFSv4 state manager runs the module count increments and won't be decremented. So the nfsv4 module cannot be unloaded.

So restore the module_put_and_kthread_exit() call.

Fixes f49169c97fce ("NFSD: Remove svc_serv_ops::svo_module")
---
 fs/nfs/nfs4state.c | 1 +
 1 file changed, 1 insertion(+)
fs/nfs/nfs4state.c +1 -0
index 61050ffac93ef..d7868cc527805 100644
@@ -2766,5 +2766,6 @@ static int nfs4_run_state_manager(void *ptr)
 		goto again;
 
 	nfs_put_client(clp);
+	module_put_and_kthread_exit(0);
 	return 0;
 }
Signature · 1 line
2.43.0
author maintainer
#334 lore raw

PATCH 334/690 fanotify: refine the validation checks on non-dir inode mask

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Amir Goldstein <[email protected]>

[ Upstream commit 8698e3bab4dd7968666e84e111d0bfd17c040e77 ]

Commit ceaf69f8eadc ("fanotify: do not allow setting dirent events in mask of non-dir") added restrictions about setting dirent events in the mask of a non-dir inode mark, which does not make any sense.

For backward compatibility, these restictions were added only to new (v5.17+) APIs.

It also does not make any sense to set the flags FAN_EVENT_ON_CHILD or FAN_ONDIR in the mask of a non-dir inode. Add these flags to the dir-only restriction of the new APIs as well.

Move the check of the dir-only flags for new APIs into the helper fanotify_events_supported(), which is only called for FAN_MARK_ADD, because there is no need to error on an attempt to remove the dir-only flags from non-dir inode.

Fixes ceaf69f8eadc ("fanotify: do not allow setting dirent events in mask of non-dir")
[ cel: adjusted to apply on v5.15.y ]
---
 fs/notify/fanotify/fanotify_user.c | 30 +++++++++++++++---------------
 include/linux/fanotify.h           |  4 ++++
 2 files changed, 19 insertions(+), 15 deletions(-)
fs/notify/fanotify/fanotify_user.c +15 -15
index 6db5a0b03a78d..433c89fdcf0cd 100644
@@ -1509,10 +1509,14 @@ static int fanotify_test_fid(struct dentry *dentry)
 	return 0;
 }
 
-static int fanotify_events_supported(struct path *path, __u64 mask,
+static int fanotify_events_supported(struct fsnotify_group *group,
+				     struct path *path, __u64 mask,
 				     unsigned int flags)
 {
 	unsigned int mark_type = flags & FANOTIFY_MARK_TYPE_BITS;
+	/* Strict validation of events in non-dir inode mask with v5.17+ APIs */
+	bool strict_dir_events = FAN_GROUP_FLAG(group, FAN_REPORT_TARGET_FID) ||
+				 (mask & FAN_RENAME);
 
 	/*
 	 * Some filesystems such as 'proc' acquire unusual locks when opening
@@ -1540,6 +1544,15 @@ static int fanotify_events_supported(struct path *path, __u64 mask,
 	    path->mnt->mnt_sb->s_flags & SB_NOUSER)
 		return -EINVAL;
 
+	/*
+	 * We shouldn't have allowed setting dirent events and the directory
+	 * flags FAN_ONDIR and FAN_EVENT_ON_CHILD in mask of non-dir inode,
+	 * but because we always allowed it, error only when using new APIs.
+	 */
+	if (strict_dir_events && mark_type == FAN_MARK_INODE &&
+	    !d_is_dir(path->dentry) && (mask & FANOTIFY_DIRONLY_EVENT_BITS))
+		return -ENOTDIR;
+
 	return 0;
 }
 
@@ -1686,7 +1699,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 		goto fput_and_out;
 
 	if (flags & FAN_MARK_ADD) {
-		ret = fanotify_events_supported(&path, mask, flags);
+		ret = fanotify_events_supported(group, &path, mask, flags);
 		if (ret)
 			goto path_put_and_out;
 	}
@@ -1709,19 +1722,6 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	else
 		mnt = path.mnt;
 
-	/*
-	 * FAN_RENAME is not allowed on non-dir (for now).
-	 * We shouldn't have allowed setting any dirent events in mask of
-	 * non-dir, but because we always allowed it, error only if group
-	 * was initialized with the new flag FAN_REPORT_TARGET_FID.
-	 */
-	ret = -ENOTDIR;
-	if (inode && !S_ISDIR(inode->i_mode) &&
-	    ((mask & FAN_RENAME) ||
-	     ((mask & FANOTIFY_DIRENT_EVENTS) &&
-	      FAN_GROUP_FLAG(group, FAN_REPORT_TARGET_FID))))
-		goto path_put_and_out;
-
 	/* Mask out FAN_EVENT_ON_CHILD flag for sb/mount/non-dir marks */
 	if (mnt || !S_ISDIR(inode->i_mode)) {
 		mask &= ~FAN_EVENT_ON_CHILD;
include/linux/fanotify.h +4 -0
index 81f45061c1b18..4f6cbe6c6e235 100644
@@ -113,6 +113,10 @@ extern struct ctl_table fanotify_table[]; /* for sysctl */
 					 FANOTIFY_PERM_EVENTS | \
 					 FAN_Q_OVERFLOW | FAN_ONDIR)
 
+/* Events and flags relevant only for directories */
+#define FANOTIFY_DIRONLY_EVENT_BITS	(FANOTIFY_DIRENT_EVENTS | \
+					 FAN_EVENT_ON_CHILD | FAN_ONDIR)
+
 #define ALL_FANOTIFY_EVENT_BITS		(FANOTIFY_OUTGOING_EVENTS | \
 					 FANOTIFY_EVENT_FLAGS)
 
Signature · 1 line
2.43.0
author maintainer
#335 lore raw

PATCH 335/690 NFSD: Decode NFSv4 birth time attribute

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 5b2f3e0777da2a5dd62824bbe2fdab1d12caaf8f ]

NFSD has advertised support for the NFSv4 time_create attribute since commit e377a3e698fb ("nfsd: Add support for the birth time attribute").

Igor Mammedov reports that Mac OS clients attempt to set the NFSv4 birth time attribute via OPEN(CREATE) and SETATTR if the server indicates that it supports it, but since the above commit was merged, those attempts now fail.

Table 5 in RFC 8881 lists the time_create attribute as one that can be both set and retrieved, but the above commit did not add server support for clients to provide a time_create attribute. IMO that's a bug in our implementation of the NFSv4 protocol, which this commit addresses.

Whether NFSD silently ignores the new birth time or actually sets it is another matter. I haven't found another filesystem service in the Linux kernel that enables users or clients to modify a file's birth time attribute.

This commit reflects my (perhaps incorrect) understanding of whether Linux users can set a file's birth time. NFSD will now recognize a time_create attribute but it ignores its value. It clears the time_create bit in the returned attribute bitmask to indicate that the value was not used.

Fixes e377a3e698fb ("nfsd: Add support for the birth time attribute")
---
 fs/nfsd/nfs4xdr.c | 9 +++++++++
 fs/nfsd/nfsd.h    | 3 ++-
 2 files changed, 11 insertions(+), 1 deletion(-)
fs/nfsd/nfs4xdr.c +9 -0
index 96d41b1cc2d17..07f891d7fa0ae 100644
@@ -470,6 +470,15 @@ nfsd4_decode_fattr4(struct nfsd4_compoundargs *argp, u32 *bmval, u32 bmlen,
 			return nfserr_bad_xdr;
 		}
 	}
+	if (bmval[1] & FATTR4_WORD1_TIME_CREATE) {
+		struct timespec64 ts;
+
+		/* No Linux filesystem supports setting this attribute. */
+		bmval[1] &= ~FATTR4_WORD1_TIME_CREATE;
+		status = nfsd4_decode_nfstime4(argp, &ts);
+		if (status)
+			return status;
+	}
 	if (bmval[1] & FATTR4_WORD1_TIME_MODIFY_SET) {
 		u32 set_it;
 
fs/nfsd/nfsd.h +2 -1
index 847b482155ae9..9a8b09afc1733 100644
@@ -465,7 +465,8 @@ static inline bool nfsd_attrs_supported(u32 minorversion, const u32 *bmval)
 	(FATTR4_WORD0_SIZE | FATTR4_WORD0_ACL)
 #define NFSD_WRITEABLE_ATTRS_WORD1 \
 	(FATTR4_WORD1_MODE | FATTR4_WORD1_OWNER | FATTR4_WORD1_OWNER_GROUP \
-	| FATTR4_WORD1_TIME_ACCESS_SET | FATTR4_WORD1_TIME_MODIFY_SET)
+	| FATTR4_WORD1_TIME_ACCESS_SET | FATTR4_WORD1_TIME_CREATE \
+	| FATTR4_WORD1_TIME_MODIFY_SET)
 #ifdef CONFIG_NFSD_V4_SECURITY_LABEL
 #define MAYBE_FATTR4_WORD2_SECURITY_LABEL \
 	FATTR4_WORD2_SECURITY_LABEL
Signature · 1 line
2.43.0
author maintainer
#336 lore raw

PATCH 336/690 fs: inotify: Fix typo in inotify comment

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Oliver Ford <[email protected]>

Correct spelling in comment.

---
 fs/notify/inotify/inotify_user.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/notify/inotify/inotify_user.c +1 -1
index 266b1302290ba..131938986e54f 100644
@@ -128,7 +128,7 @@ static inline u32 inotify_mask_to_arg(__u32 mask)
 		       IN_Q_OVERFLOW);
 }
 
-/* intofiy userspace file descriptor functions */
+/* inotify userspace file descriptor functions */
 static __poll_t inotify_poll(struct file *file, poll_table *wait)
 {
 	struct fsnotify_group *group = file->private_data;
Signature · 1 line
2.43.0
author maintainer
#337 lore raw

PATCH 337/690 fanotify: prepare for setting event flags in ignore mask

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Amir Goldstein <[email protected]>

[ Upstream commit 31a371e419c885e0f137ce70395356ba8639dc52 ]

Setting flags FAN_ONDIR FAN_EVENT_ON_CHILD in ignore mask has no effect. The FAN_EVENT_ON_CHILD flag in mask implicitly applies to ignore mask and ignore mask is always implicitly applied to events on directories.

Define a mark flag that replaces this legacy behavior with logic of applying the ignore mask according to event flags in ignore mask.

Implement the new logic to prepare for supporting an ignore mask that ignores events on children and ignore mask that does not ignore events on directories.

To emphasize the change in terminology, also rename ignored_mask mark member to ignore_mask and use accessors to get only the effective ignored events or the ignored events and flags.

This change in terminology finally aligns with the "ignore mask" language in man pages and in most of the comments.

---
 fs/notify/fanotify/fanotify.c      | 19 ++++---
 fs/notify/fanotify/fanotify_user.c | 21 ++++---
 fs/notify/fdinfo.c                 |  6 +-
 fs/notify/fsnotify.c               | 21 ++++---
 include/linux/fsnotify_backend.h   | 89 ++++++++++++++++++++++++++++--
 5 files changed, 121 insertions(+), 35 deletions(-)
fs/notify/fanotify/fanotify.c +11 -8
index 4f897e1095470..cd7d09a569fff 100644
@@ -295,12 +295,13 @@ static u32 fanotify_group_event_mask(struct fsnotify_group *group,
 				     const void *data, int data_type,
 				     struct inode *dir)
 {
-	__u32 marks_mask = 0, marks_ignored_mask = 0;
+	__u32 marks_mask = 0, marks_ignore_mask = 0;
 	__u32 test_mask, user_mask = FANOTIFY_OUTGOING_EVENTS |
 				     FANOTIFY_EVENT_FLAGS;
 	const struct path *path = fsnotify_data_path(data, data_type);
 	unsigned int fid_mode = FAN_GROUP_FLAG(group, FANOTIFY_FID_BITS);
 	struct fsnotify_mark *mark;
+	bool ondir = event_mask & FAN_ONDIR;
 	int type;
 
 	pr_debug("%s: report_mask=%x mask=%x data=%p data_type=%d\n",
@@ -315,19 +316,21 @@ static u32 fanotify_group_event_mask(struct fsnotify_group *group,
 			return 0;
 	} else if (!(fid_mode & FAN_REPORT_FID)) {
 		/* Do we have a directory inode to report? */
-		if (!dir && !(event_mask & FS_ISDIR))
+		if (!dir && !ondir)
 			return 0;
 	}
 
 	fsnotify_foreach_iter_mark_type(iter_info, mark, type) {
-		/* Apply ignore mask regardless of mark's ISDIR flag */
-		marks_ignored_mask |= mark->ignored_mask;
+		/*
+		 * Apply ignore mask depending on event flags in ignore mask.
+		 */
+		marks_ignore_mask |=
+			fsnotify_effective_ignore_mask(mark, ondir, type);
 
 		/*
-		 * If the event is on dir and this mark doesn't care about
-		 * events on dir, don't send it!
+		 * Send the event depending on event flags in mark mask.
 		 */
-		if (event_mask & FS_ISDIR && !(mark->mask & FS_ISDIR))
+		if (!fsnotify_mask_applicable(mark->mask, ondir, type))
 			continue;
 
 		marks_mask |= mark->mask;
@@ -336,7 +339,7 @@ static u32 fanotify_group_event_mask(struct fsnotify_group *group,
 		*match_mask |= 1U << type;
 	}
 
-	test_mask = event_mask & marks_mask & ~marks_ignored_mask;
+	test_mask = event_mask & marks_mask & ~marks_ignore_mask;
 
 	/*
 	 * For dirent modification events (create/delete/move) that do not carry
fs/notify/fanotify/fanotify_user.c +12 -9
index 433c89fdcf0cd..9367216263cab 100644
@@ -1008,7 +1008,7 @@ static __u32 fanotify_mark_remove_from_mask(struct fsnotify_mark *fsn_mark,
 	if (!(flags & FAN_MARK_IGNORED_MASK)) {
 		fsn_mark->mask &= ~mask;
 	} else {
-		fsn_mark->ignored_mask &= ~mask;
+		fsn_mark->ignore_mask &= ~mask;
 	}
 	newmask = fsnotify_calc_mask(fsn_mark);
 	/*
@@ -1017,7 +1017,7 @@ static __u32 fanotify_mark_remove_from_mask(struct fsnotify_mark *fsn_mark,
 	 * changes to the mask.
 	 * Destroy mark when only umask bits remain.
 	 */
-	*destroy = !((fsn_mark->mask | fsn_mark->ignored_mask) & ~umask);
+	*destroy = !((fsn_mark->mask | fsn_mark->ignore_mask) & ~umask);
 	spin_unlock(&fsn_mark->lock);
 
 	return oldmask & ~newmask;
@@ -1086,7 +1086,7 @@ static bool fanotify_mark_update_flags(struct fsnotify_mark *fsn_mark,
 	/*
 	 * Setting FAN_MARK_IGNORED_SURV_MODIFY for the first time may lead to
 	 * the removal of the FS_MODIFY bit in calculated mask if it was set
-	 * because of an ignored mask that is now going to survive FS_MODIFY.
+	 * because of an ignore mask that is now going to survive FS_MODIFY.
 	 */
 	if ((fan_flags & FAN_MARK_IGNORED_MASK) &&
 	    (fan_flags & FAN_MARK_IGNORED_SURV_MODIFY) &&
@@ -1119,7 +1119,7 @@ static bool fanotify_mark_add_to_mask(struct fsnotify_mark *fsn_mark,
 	if (!(fan_flags & FAN_MARK_IGNORED_MASK))
 		fsn_mark->mask |= mask;
 	else
-		fsn_mark->ignored_mask |= mask;
+		fsn_mark->ignore_mask |= mask;
 
 	recalc = fsnotify_calc_mask(fsn_mark) &
 		~fsnotify_conn_mask(fsn_mark->connector);
@@ -1257,7 +1257,7 @@ static int fanotify_add_inode_mark(struct fsnotify_group *group,
 
 	/*
 	 * If some other task has this inode open for write we should not add
-	 * an ignored mark, unless that ignored mark is supposed to survive
+	 * an ignore mask, unless that ignore mask is supposed to survive
 	 * modification changes anyway.
 	 */
 	if ((flags & FAN_MARK_IGNORED_MASK) &&
@@ -1567,7 +1567,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	__kernel_fsid_t __fsid, *fsid = NULL;
 	u32 valid_mask = FANOTIFY_EVENTS | FANOTIFY_EVENT_FLAGS;
 	unsigned int mark_type = flags & FANOTIFY_MARK_TYPE_BITS;
-	bool ignored = flags & FAN_MARK_IGNORED_MASK;
+	bool ignore = flags & FAN_MARK_IGNORED_MASK;
 	unsigned int obj_type, fid_mode;
 	u32 umask = 0;
 	int ret;
@@ -1616,8 +1616,11 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	if (mask & ~valid_mask)
 		return -EINVAL;
 
-	/* Event flags (ONDIR, ON_CHILD) are meaningless in ignored mask */
-	if (ignored)
+	/*
+	 * Event flags (FAN_ONDIR, FAN_EVENT_ON_CHILD) have no effect with
+	 * FAN_MARK_IGNORED_MASK.
+	 */
+	if (ignore)
 		mask &= ~FANOTIFY_EVENT_FLAGS;
 
 	f = fdget(fanotify_fd);
@@ -1731,7 +1734,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 		 * events with parent/name info for non-directory.
 		 */
 		if ((fid_mode & FAN_REPORT_DIR_FID) &&
-		    (flags & FAN_MARK_ADD) && !ignored)
+		    (flags & FAN_MARK_ADD) && !ignore)
 			mask |= FAN_EVENT_ON_CHILD;
 	}
 
fs/notify/fdinfo.c +3 -3
index 59fb40abe33d3..55081ae3a6ec0 100644
@@ -113,7 +113,7 @@ static void fanotify_fdinfo(struct seq_file *m, struct fsnotify_mark *mark)
 			return;
 		seq_printf(m, "fanotify ino:%lx sdev:%x mflags:%x mask:%x ignored_mask:%x ",
 			   inode->i_ino, inode->i_sb->s_dev,
-			   mflags, mark->mask, mark->ignored_mask);
+			   mflags, mark->mask, mark->ignore_mask);
 		show_mark_fhandle(m, inode);
 		seq_putc(m, '\n');
 		iput(inode);
@@ -121,12 +121,12 @@ static void fanotify_fdinfo(struct seq_file *m, struct fsnotify_mark *mark)
 		struct mount *mnt = fsnotify_conn_mount(mark->connector);
 
 		seq_printf(m, "fanotify mnt_id:%x mflags:%x mask:%x ignored_mask:%x\n",
-			   mnt->mnt_id, mflags, mark->mask, mark->ignored_mask);
+			   mnt->mnt_id, mflags, mark->mask, mark->ignore_mask);
 	} else if (mark->connector->type == FSNOTIFY_OBJ_TYPE_SB) {
 		struct super_block *sb = fsnotify_conn_sb(mark->connector);
 
 		seq_printf(m, "fanotify sdev:%x mflags:%x mask:%x ignored_mask:%x\n",
-			   sb->s_dev, mflags, mark->mask, mark->ignored_mask);
+			   sb->s_dev, mflags, mark->mask, mark->ignore_mask);
 	}
 }
 
fs/notify/fsnotify.c +12 -9
index 0b3e74935cb4f..8687562df2e37 100644
@@ -324,7 +324,8 @@ static int send_to_group(__u32 mask, const void *data, int data_type,
 	struct fsnotify_group *group = NULL;
 	__u32 test_mask = (mask & ALL_FSNOTIFY_EVENTS);
 	__u32 marks_mask = 0;
-	__u32 marks_ignored_mask = 0;
+	__u32 marks_ignore_mask = 0;
+	bool is_dir = mask & FS_ISDIR;
 	struct fsnotify_mark *mark;
 	int type;
 
@@ -336,7 +337,7 @@ static int send_to_group(__u32 mask, const void *data, int data_type,
 		fsnotify_foreach_iter_mark_type(iter_info, mark, type) {
 			if (!(mark->flags &
 			      FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY))
-				mark->ignored_mask = 0;
+				mark->ignore_mask = 0;
 		}
 	}
 
@@ -344,14 +345,15 @@ static int send_to_group(__u32 mask, const void *data, int data_type,
 	fsnotify_foreach_iter_mark_type(iter_info, mark, type) {
 		group = mark->group;
 		marks_mask |= mark->mask;
-		marks_ignored_mask |= mark->ignored_mask;
+		marks_ignore_mask |=
+			fsnotify_effective_ignore_mask(mark, is_dir, type);
 	}
 
-	pr_debug("%s: group=%p mask=%x marks_mask=%x marks_ignored_mask=%x data=%p data_type=%d dir=%p cookie=%d\n",
-		 __func__, group, mask, marks_mask, marks_ignored_mask,
+	pr_debug("%s: group=%p mask=%x marks_mask=%x marks_ignore_mask=%x data=%p data_type=%d dir=%p cookie=%d\n",
+		 __func__, group, mask, marks_mask, marks_ignore_mask,
 		 data, data_type, dir, cookie);
 
-	if (!(test_mask & marks_mask & ~marks_ignored_mask))
+	if (!(test_mask & marks_mask & ~marks_ignore_mask))
 		return 0;
 
 	if (group->ops->handle_event) {
@@ -423,7 +425,8 @@ static bool fsnotify_iter_select_report_types(
 			 * But is *this mark* watching children?
 			 */
 			if (type == FSNOTIFY_ITER_TYPE_PARENT &&
-			    !(mark->mask & FS_EVENT_ON_CHILD))
+			    !(mark->mask & FS_EVENT_ON_CHILD) &&
+			    !(fsnotify_ignore_mask(mark) & FS_EVENT_ON_CHILD))
 				continue;
 
 			fsnotify_iter_set_report_type(iter_info, type);
@@ -532,8 +535,8 @@ int fsnotify(__u32 mask, const void *data, int data_type, struct inode *dir,
 
 
 	/*
-	 * If this is a modify event we may need to clear some ignored masks.
-	 * In that case, the object with ignored masks will have the FS_MODIFY
+	 * If this is a modify event we may need to clear some ignore masks.
+	 * In that case, the object with ignore masks will have the FS_MODIFY
 	 * event in its mask.
 	 * Otherwise, return if none of the marks care about this type of event.
 	 */
include/linux/fsnotify_backend.h +83 -6

119 lines not on this page · raw message

Signature · 1 line
2.43.0
author maintainer
#338 lore raw

PATCH 338/690 fanotify: cleanups for fanotify_mark() input validations

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Amir Goldstein <[email protected]>

[ Upstream commit 8afd7215aa97f8868d033f6e1d01a276ab2d29c0 ]

Create helper fanotify_may_update_existing_mark() for checking for conflicts between existing mark flags and fanotify_mark() flags.

Use variable mark_cmd to make the checks for mark command bits cleaner.

---
 fs/notify/fanotify/fanotify_user.c | 30 +++++++++++++++++++++---------
 include/linux/fanotify.h           |  9 +++++----
 2 files changed, 26 insertions(+), 13 deletions(-)
fs/notify/fanotify/fanotify_user.c +21 -9
index 9367216263cab..870db0f361f4c 100644
@@ -1183,6 +1183,19 @@ static int fanotify_group_init_error_pool(struct fsnotify_group *group)
 					 sizeof(struct fanotify_error_event));
 }
 
+static int fanotify_may_update_existing_mark(struct fsnotify_mark *fsn_mark,
+					      unsigned int fan_flags)
+{
+	/*
+	 * Non evictable mark cannot be downgraded to evictable mark.
+	 */
+	if (fan_flags & FAN_MARK_EVICTABLE &&
+	    !(fsn_mark->flags & FSNOTIFY_MARK_FLAG_NO_IREF))
+		return -EEXIST;
+
+	return 0;
+}
+
 static int fanotify_add_mark(struct fsnotify_group *group,
 			     fsnotify_connp_t *connp, unsigned int obj_type,
 			     __u32 mask, unsigned int fan_flags,
@@ -1204,13 +1217,11 @@ static int fanotify_add_mark(struct fsnotify_group *group,
 	}
 
 	/*
-	 * Non evictable mark cannot be downgraded to evictable mark.
+	 * Check if requested mark flags conflict with an existing mark flags.
 	 */
-	if (fan_flags & FAN_MARK_EVICTABLE &&
-	    !(fsn_mark->flags & FSNOTIFY_MARK_FLAG_NO_IREF)) {
-		ret = -EEXIST;
+	ret = fanotify_may_update_existing_mark(fsn_mark, fan_flags);
+	if (ret)
 		goto out;
-	}
 
 	/*
 	 * Error events are pre-allocated per group, only if strictly
@@ -1567,6 +1578,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	__kernel_fsid_t __fsid, *fsid = NULL;
 	u32 valid_mask = FANOTIFY_EVENTS | FANOTIFY_EVENT_FLAGS;
 	unsigned int mark_type = flags & FANOTIFY_MARK_TYPE_BITS;
+	unsigned int mark_cmd = flags & FANOTIFY_MARK_CMD_BITS;
 	bool ignore = flags & FAN_MARK_IGNORED_MASK;
 	unsigned int obj_type, fid_mode;
 	u32 umask = 0;
@@ -1596,7 +1608,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 		return -EINVAL;
 	}
 
-	switch (flags & (FAN_MARK_ADD | FAN_MARK_REMOVE | FAN_MARK_FLUSH)) {
+	switch (mark_cmd) {
 	case FAN_MARK_ADD:
 	case FAN_MARK_REMOVE:
 		if (!mask)
@@ -1685,7 +1697,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	if (mask & FAN_RENAME && !(fid_mode & FAN_REPORT_NAME))
 		goto fput_and_out;
 
-	if (flags & FAN_MARK_FLUSH) {
+	if (mark_cmd == FAN_MARK_FLUSH) {
 		ret = 0;
 		if (mark_type == FAN_MARK_MOUNT)
 			fsnotify_clear_vfsmount_marks_by_group(group);
@@ -1701,7 +1713,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	if (ret)
 		goto fput_and_out;
 
-	if (flags & FAN_MARK_ADD) {
+	if (mark_cmd == FAN_MARK_ADD) {
 		ret = fanotify_events_supported(group, &path, mask, flags);
 		if (ret)
 			goto path_put_and_out;
@@ -1739,7 +1751,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	}
 
 	/* create/update an inode mark */
-	switch (flags & (FAN_MARK_ADD | FAN_MARK_REMOVE)) {
+	switch (mark_cmd) {
 	case FAN_MARK_ADD:
 		if (mark_type == FAN_MARK_MOUNT)
 			ret = fanotify_add_vfsmount_mark(group, mnt, mask,
include/linux/fanotify.h +5 -4
index 4f6cbe6c6e235..c9e185407ebcb 100644
@@ -61,15 +61,16 @@ extern struct ctl_table fanotify_table[]; /* for sysctl */
 #define FANOTIFY_MARK_TYPE_BITS	(FAN_MARK_INODE | FAN_MARK_MOUNT | \
 				 FAN_MARK_FILESYSTEM)
 
+#define FANOTIFY_MARK_CMD_BITS	(FAN_MARK_ADD | FAN_MARK_REMOVE | \
+				 FAN_MARK_FLUSH)
+
 #define FANOTIFY_MARK_FLAGS	(FANOTIFY_MARK_TYPE_BITS | \
-				 FAN_MARK_ADD | \
-				 FAN_MARK_REMOVE | \
+				 FANOTIFY_MARK_CMD_BITS | \
 				 FAN_MARK_DONT_FOLLOW | \
 				 FAN_MARK_ONLYDIR | \
 				 FAN_MARK_IGNORED_MASK | \
 				 FAN_MARK_IGNORED_SURV_MODIFY | \
-				 FAN_MARK_EVICTABLE | \
-				 FAN_MARK_FLUSH)
+				 FAN_MARK_EVICTABLE)
 
 /*
  * Events that can be reported with data type FSNOTIFY_EVENT_PATH.
Signature · 1 line
2.43.0
author maintainer
#339 lore raw

PATCH 339/690 fanotify: introduce FAN_MARK_IGNORE

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Amir Goldstein <[email protected]>

[ Upstream commit e252f2ed1c8c6c3884ab5dd34e003ed21f1fe6e0 ]

This flag is a new way to configure ignore mask which allows adding and removing the event flags FAN_ONDIR and FAN_EVENT_ON_CHILD in ignore mask.

The legacy FAN_MARK_IGNORED_MASK flag would always ignore events on directories and would ignore events on children depending on whether the FAN_EVENT_ON_CHILD flag was set in the (non ignored) mask.

FAN_MARK_IGNORE can be used to ignore events on children without setting FAN_EVENT_ON_CHILD in the mark's mask and will not ignore events on directories unconditionally, only when FAN_ONDIR is set in ignore mask.

The new behavior is non-downgradable. After calling fanotify_mark() with FAN_MARK_IGNORE once, calling fanotify_mark() with FAN_MARK_IGNORED_MASK on the same object will return EEXIST error.

Setting the event flags with FAN_MARK_IGNORE on a non-dir inode mark has no meaning and will return ENOTDIR error.

The meaning of FAN_MARK_IGNORED_SURV_MODIFY is preserved with the new FAN_MARK_IGNORE flag, but with a few semantic differences:

1. FAN_MARK_IGNORED_SURV_MODIFY is required for filesystem and mount marks and on an inode mark on a directory. Omitting this flag will return EINVAL or EISDIR error.

2. An ignore mask on a non-directory inode that survives modify could never be downgraded to an ignore mask that does not survive modify. With new FAN_MARK_IGNORE semantics we make that rule explicit - trying to update a surviving ignore mask without the flag FAN_MARK_IGNORED_SURV_MODIFY will return EEXIST error.

The conveniene macro FAN_MARK_IGNORE_SURV is added for (FAN_MARK_IGNORE | FAN_MARK_IGNORED_SURV_MODIFY), because the common case should use short constant names.

---
 fs/notify/fanotify/fanotify.h      |  2 +
 fs/notify/fanotify/fanotify_user.c | 63 +++++++++++++++++++++++++-----
 include/linux/fanotify.h           |  5 ++-
 include/uapi/linux/fanotify.h      |  8 ++++
 4 files changed, 67 insertions(+), 11 deletions(-)
fs/notify/fanotify/fanotify.h +2 -0
index 80e0ec95b1131..1d9f11255c64f 100644
@@ -499,6 +499,8 @@ static inline unsigned int fanotify_mark_user_flags(struct fsnotify_mark *mark)
 		mflags |= FAN_MARK_IGNORED_SURV_MODIFY;
 	if (mark->flags & FSNOTIFY_MARK_FLAG_NO_IREF)
 		mflags |= FAN_MARK_EVICTABLE;
+	if (mark->flags & FSNOTIFY_MARK_FLAG_HAS_IGNORE_FLAGS)
+		mflags |= FAN_MARK_IGNORE;
 
 	return mflags;
 }
fs/notify/fanotify/fanotify_user.c +53 -10

141 lines not on this page · raw message

include/linux/fanotify.h +4 -1
index c9e185407ebcb..558844c8d2598 100644
@@ -64,11 +64,14 @@ extern struct ctl_table fanotify_table[]; /* for sysctl */
 #define FANOTIFY_MARK_CMD_BITS	(FAN_MARK_ADD | FAN_MARK_REMOVE | \
 				 FAN_MARK_FLUSH)
 
+#define FANOTIFY_MARK_IGNORE_BITS (FAN_MARK_IGNORED_MASK | \
+				   FAN_MARK_IGNORE)
+
 #define FANOTIFY_MARK_FLAGS	(FANOTIFY_MARK_TYPE_BITS | \
 				 FANOTIFY_MARK_CMD_BITS | \
+				 FANOTIFY_MARK_IGNORE_BITS | \
 				 FAN_MARK_DONT_FOLLOW | \
 				 FAN_MARK_ONLYDIR | \
-				 FAN_MARK_IGNORED_MASK | \
 				 FAN_MARK_IGNORED_SURV_MODIFY | \
 				 FAN_MARK_EVICTABLE)
 
include/uapi/linux/fanotify.h +8 -0
index f1f89132d60e2..d8536d77fea1c 100644
@@ -83,12 +83,20 @@
 #define FAN_MARK_FLUSH		0x00000080
 /* FAN_MARK_FILESYSTEM is	0x00000100 */
 #define FAN_MARK_EVICTABLE	0x00000200
+/* This bit is mutually exclusive with FAN_MARK_IGNORED_MASK bit */
+#define FAN_MARK_IGNORE		0x00000400
 
 /* These are NOT bitwise flags.  Both bits can be used togther.  */
 #define FAN_MARK_INODE		0x00000000
 #define FAN_MARK_MOUNT		0x00000010
 #define FAN_MARK_FILESYSTEM	0x00000100
 
+/*
+ * Convenience macro - FAN_MARK_IGNORE requires FAN_MARK_IGNORED_SURV_MODIFY
+ * for non-inode mark types.
+ */
+#define FAN_MARK_IGNORE_SURV	(FAN_MARK_IGNORE | FAN_MARK_IGNORED_SURV_MODIFY)
+
 /* Deprecated - do not use this in programs and do not add new flags here! */
 #define FAN_ALL_MARK_FLAGS	(FAN_MARK_ADD |\
 				 FAN_MARK_REMOVE |\
Signature · 1 line
2.43.0
author maintainer
#340 lore raw

PATCH 340/690 fsnotify: Fix comment typo

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Xin Gao <[email protected]>

[ Upstream commit feee1ce45a5666bbdb08c5bb2f5f394047b1915b ]

The double `if' is duplicated in line 104, remove one.

---
 fs/notify/fsnotify.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/notify/fsnotify.c +1 -1
index 8687562df2e37..7974e91ffe134 100644
@@ -100,7 +100,7 @@ void fsnotify_sb_delete(struct super_block *sb)
  * Given an inode, first check if we care what happens to our children.  Inotify
  * and dnotify both tell their parents about events.  If we care about any event
  * on a child we run all of our children and set a dentry flag saying that the
- * parent cares.  Thus when an event happens on a child it can quickly tell if
+ * parent cares.  Thus when an event happens on a child it can quickly tell
  * if there is a need to find a parent and send the event to the parent.
  */
 void __fsnotify_update_child_dentry_flags(struct inode *inode)
Signature · 1 line
2.43.0
author maintainer
#341 lore raw

PATCH 341/690 NLM: Defend against file_lock changes after vfs_test_lock()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Benjamin Coddington <[email protected]>

[ Upstream commit 184cefbe62627730c30282df12bcff9aae4816ea ]

Instead of trusting that struct file_lock returns completely unchanged after vfs_test_lock() when there's no conflicting lock, stash away our nlm_lockowner reference so we can properly release it for all cases.

This defends against another file_lock implementation overwriting fl_owner when the return type is F_UNLCK.

---
 fs/lockd/svc4proc.c         |  4 +++-
 fs/lockd/svclock.c          | 10 +---------
 fs/lockd/svcproc.c          |  5 ++++-
 include/linux/lockd/lockd.h |  1 +
 4 files changed, 9 insertions(+), 11 deletions(-)
fs/lockd/svc4proc.c +3 -1
index 1c9214801e69e..930e90f21b151 100644
@@ -95,6 +95,7 @@ __nlm4svc_proc_test(struct svc_rqst *rqstp, struct nlm_res *resp)
 	struct nlm_args *argp = rqstp->rq_argp;
 	struct nlm_host	*host;
 	struct nlm_file	*file;
+	struct nlm_lockowner *test_owner;
 	__be32 rc = rpc_success;
 
 	dprintk("lockd: TEST4        called\n");
@@ -104,6 +105,7 @@ __nlm4svc_proc_test(struct svc_rqst *rqstp, struct nlm_res *resp)
 	if ((resp->status = nlm4svc_retrieve_args(rqstp, argp, &host, &file)))
 		return resp->status == nlm_drop_reply ? rpc_drop_reply :rpc_success;
 
+	test_owner = argp->lock.fl.fl_owner;
 	/* Now check for conflicting locks */
 	resp->status = nlmsvc_testlock(rqstp, file, host, &argp->lock, &resp->lock, &resp->cookie);
 	if (resp->status == nlm_drop_reply)
@@ -111,7 +113,7 @@ __nlm4svc_proc_test(struct svc_rqst *rqstp, struct nlm_res *resp)
 	else
 		dprintk("lockd: TEST4        status %d\n", ntohl(resp->status));
 
-	nlmsvc_release_lockowner(&argp->lock);
+	nlmsvc_put_lockowner(test_owner);
 	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rc;
fs/lockd/svclock.c +1 -9
index cb3658ab9b7ae..9c1aa75441e1c 100644
@@ -340,7 +340,7 @@ nlmsvc_get_lockowner(struct nlm_lockowner *lockowner)
 	return lockowner;
 }
 
-static void nlmsvc_put_lockowner(struct nlm_lockowner *lockowner)
+void nlmsvc_put_lockowner(struct nlm_lockowner *lockowner)
 {
 	if (!refcount_dec_and_lock(&lockowner->count, &lockowner->host->h_lock))
 		return;
@@ -590,7 +590,6 @@ nlmsvc_testlock(struct svc_rqst *rqstp, struct nlm_file *file,
 	int			error;
 	int			mode;
 	__be32			ret;
-	struct nlm_lockowner	*test_owner;
 
 	dprintk("lockd: nlmsvc_testlock(%s/%ld, ty=%d, %Ld-%Ld)\n",
 				nlmsvc_file_inode(file)->i_sb->s_id,
@@ -604,9 +603,6 @@ nlmsvc_testlock(struct svc_rqst *rqstp, struct nlm_file *file,
 		goto out;
 	}
 
-	/* If there's a conflicting lock, remember to clean up the test lock */
-	test_owner = (struct nlm_lockowner *)lock->fl.fl_owner;
-
 	mode = lock_to_openmode(&lock->fl);
 	error = vfs_test_lock(file->f_file[mode], &lock->fl);
 	if (error) {
@@ -635,10 +631,6 @@ nlmsvc_testlock(struct svc_rqst *rqstp, struct nlm_file *file,
 	conflock->fl.fl_end = lock->fl.fl_end;
 	locks_release_private(&lock->fl);
 
-	/* Clean up the test lock */
-	lock->fl.fl_owner = NULL;
-	nlmsvc_put_lockowner(test_owner);
-
 	ret = nlm_lck_denied;
 out:
 	return ret;
fs/lockd/svcproc.c +4 -1
index 99696d3f6dd66..c215a4599d5c8 100644
@@ -116,6 +116,7 @@ __nlmsvc_proc_test(struct svc_rqst *rqstp, struct nlm_res *resp)
 	struct nlm_args *argp = rqstp->rq_argp;
 	struct nlm_host	*host;
 	struct nlm_file	*file;
+	struct nlm_lockowner *test_owner;
 	__be32 rc = rpc_success;
 
 	dprintk("lockd: TEST          called\n");
@@ -125,6 +126,8 @@ __nlmsvc_proc_test(struct svc_rqst *rqstp, struct nlm_res *resp)
 	if ((resp->status = nlmsvc_retrieve_args(rqstp, argp, &host, &file)))
 		return resp->status == nlm_drop_reply ? rpc_drop_reply :rpc_success;
 
+	test_owner = argp->lock.fl.fl_owner;
+
 	/* Now check for conflicting locks */
 	resp->status = cast_status(nlmsvc_testlock(rqstp, file, host, &argp->lock, &resp->lock, &resp->cookie));
 	if (resp->status == nlm_drop_reply)
@@ -133,7 +136,7 @@ __nlmsvc_proc_test(struct svc_rqst *rqstp, struct nlm_res *resp)
 		dprintk("lockd: TEST          status %d vers %d\n",
 			ntohl(resp->status), rqstp->rq_vers);
 
-	nlmsvc_release_lockowner(&argp->lock);
+	nlmsvc_put_lockowner(test_owner);
 	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rc;
include/linux/lockd/lockd.h +1 -0
index fcef192e5e45e..70ce419e27093 100644
@@ -292,6 +292,7 @@ void		  nlmsvc_locks_init_private(struct file_lock *, struct nlm_host *, pid_t);
 __be32		  nlm_lookup_file(struct svc_rqst *, struct nlm_file **,
 					struct nlm_lock *);
 void		  nlm_release_file(struct nlm_file *);
+void		  nlmsvc_put_lockowner(struct nlm_lockowner *);
 void		  nlmsvc_release_lockowner(struct nlm_lock *);
 void		  nlmsvc_mark_resources(struct net *);
 void		  nlmsvc_free_host_resources(struct nlm_host *);
Signature · 1 line
2.43.0
author maintainer
#342 lore raw

PATCH 342/690 NFSD: Instrument fh_verify()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 051382885552e12541cc0ebf82092be374a9ed2a ]

Capture file handles and how they map to local inodes. In particular, NFSv4 PUTFH uses fh_verify() so we can now observe which file handles are the target of OPEN, LOOKUP, RENAME, and so on.

---
 fs/nfsd/nfsfh.c |  5 +++--
 fs/nfsd/trace.h | 46 ++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 49 insertions(+), 2 deletions(-)
fs/nfsd/nfsfh.c +3 -2
index c29baa03dfafd..5e2ed4b2a925c 100644
@@ -331,8 +331,6 @@ fh_verify(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, int access)
 	struct dentry	*dentry;
 	__be32		error;
 
-	dprintk("nfsd: fh_verify(%s)\n", SVCFH_fmt(fhp));
-
 	if (!fhp->fh_dentry) {
 		error = nfsd_set_fh_dentry(rqstp, fhp);
 		if (error)
@@ -340,6 +338,9 @@ fh_verify(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, int access)
 	}
 	dentry = fhp->fh_dentry;
 	exp = fhp->fh_export;
+
+	trace_nfsd_fh_verify(rqstp, fhp, type, access);
+
 	/*
 	 * We still have to do all these permission checks, even when
 	 * fh_dentry is already set:
fs/nfsd/trace.h +46 -0

52 lines not on this page · raw message

Signature · 1 line
2.43.0
author maintainer
#343 lore raw

PATCH 343/690 NFSD: Fix space and spelling mistake

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Zhang Jiaming <[email protected]>

[ Upstream commit f532c9ff103897be0e2a787c0876683c3dc39ed3 ]

Add a blank space after ','.
Change 'succesful' to 'successful'.

---
 fs/nfsd/nfs4proc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
fs/nfsd/nfs4proc.c +2 -2
index 5b56877c7fb57..d70c4e78f0b3f 100644
@@ -828,7 +828,7 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 			goto out_umask;
 		status = nfsd_create(rqstp, &cstate->current_fh,
 				     create->cr_name, create->cr_namelen,
-				     &create->cr_iattr,S_IFCHR, rdev, &resfh);
+				     &create->cr_iattr, S_IFCHR, rdev, &resfh);
 		break;
 
 	case NF4SOCK:
@@ -2712,7 +2712,7 @@ nfsd4_proc_compound(struct svc_rqst *rqstp)
 		if (op->opdesc->op_flags & OP_MODIFIES_SOMETHING) {
 			/*
 			 * Don't execute this op if we couldn't encode a
-			 * succesful reply:
+			 * successful reply:
 			 */
 			u32 plen = op->opdesc->op_rsize_bop(rqstp, op);
 			/*
Signature · 1 line
2.43.0
author maintainer
#344 lore raw

PATCH 344/690 nfsd: remove redundant assignment to variable len

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Colin Ian King <[email protected]>

[ Upstream commit 842e00ac3aa3b4a4f7f750c8ab54f8578fc875d3 ]

Variable len is being assigned a value zero and this is never read, it is being re-assigned later. The assignment is redundant and can be removed.

Cleans up clang scan-build warning:
fs/nfsd/nfsctl.c:636:2: warning: Value stored to 'len' is never read

---
 fs/nfsd/nfsctl.c | 1 -
 1 file changed, 1 deletion(-)
fs/nfsd/nfsctl.c +0 -1
index 0621c2faf2424..66c352bf61b1d 100644
@@ -633,7 +633,6 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size)
 	}
 
 	/* Now write current state into reply buffer */
-	len = 0;
 	sep = "";
 	remaining = SIMPLE_TRANSACTION_LIMIT;
 	for (num=2 ; num <= 4 ; num++) {
Signature · 1 line
2.43.0
author maintainer
#345 lore raw

PATCH 345/690 NFSD: Demote a WARN to a pr_warn()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit ca3f9acb6d3faf78da2b63324f7c737dbddf7f69 ]

The call trace doesn't add much value, but it sure is noisy.

---
 fs/nfsd/nfs4proc.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
fs/nfsd/nfs4proc.c +3 -3
index d70c4e78f0b3f..15991eb9b8d8c 100644
@@ -630,9 +630,9 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	}
 
 	status = nfsd4_process_open2(rqstp, resfh, open);
-	WARN(status && open->op_created,
-	     "nfsd4_process_open2 failed to open newly-created file! status=%u\n",
-	     be32_to_cpu(status));
+	if (status && open->op_created)
+		pr_warn("nfsd4_process_open2 failed to open newly-created file: status=%u\n",
+			be32_to_cpu(status));
 	if (reclaim && !status)
 		nn->somebody_reclaimed = true;
 out:
Signature · 1 line
2.43.0
author maintainer
#346 lore raw

PATCH 346/690 NFSD: Report filecache LRU size

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 0fd244c115f0321fc5e34ad2291f2a572508e3f7 ]

Surface the NFSD filecache's LRU list length to help field troubleshooters monitor filecache issues.

---
 fs/nfsd/filecache.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
fs/nfsd/filecache.c +3 -1
index 1d3d13b78be0e..377d8211200ff 100644
@@ -1047,7 +1047,7 @@ nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 {
 	unsigned int i, count = 0, longest = 0;
-	unsigned long hits = 0;
+	unsigned long lru = 0, hits = 0;
 
 	/*
 	 * No need for spinlocks here since we're not terribly interested in
@@ -1060,6 +1060,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 			count += nfsd_file_hashtbl[i].nfb_count;
 			longest = max(longest, nfsd_file_hashtbl[i].nfb_count);
 		}
+		lru = list_lru_count(&nfsd_file_lru);
 	}
 	mutex_unlock(&nfsd_mutex);
 
@@ -1068,6 +1069,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 
 	seq_printf(m, "total entries: %u\n", count);
 	seq_printf(m, "longest chain: %u\n", longest);
+	seq_printf(m, "lru entries:   %lu\n", lru);
 	seq_printf(m, "cache hits:    %lu\n", hits);
 	return 0;
 }
Signature · 1 line
2.43.0
author maintainer
#347 lore raw

PATCH 347/690 NFSD: Report count of calls to nfsd_file_acquire()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 29d4bdbbb910f33d6058d2c51278f00f656df325 ]

Count the number of successful acquisitions that did not create a file (ie, acquisitions that do not result in a compulsory cache miss). This count can be compared directly with the reported hit count to compute a hit ratio.

---
 fs/nfsd/filecache.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
fs/nfsd/filecache.c +9 -2
index 377d8211200ff..5a09b76ae25a8 100644
@@ -42,6 +42,7 @@ struct nfsd_fcache_bucket {
 };
 
 static DEFINE_PER_CPU(unsigned long, nfsd_file_cache_hits);
+static DEFINE_PER_CPU(unsigned long, nfsd_file_acquisitions);
 
 struct nfsd_fcache_disposal {
 	struct work_struct work;
@@ -954,6 +955,8 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	status = nfserrno(nfsd_open_break_lease(file_inode(nf->nf_file), may_flags));
 out:
 	if (status == nfs_ok) {
+		if (open)
+			this_cpu_inc(nfsd_file_acquisitions);
 		*pnf = nf;
 	} else {
 		nfsd_file_put(nf);
@@ -1046,8 +1049,9 @@ nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
  */
 static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 {
+	unsigned long hits = 0, acquisitions = 0;
 	unsigned int i, count = 0, longest = 0;
-	unsigned long lru = 0, hits = 0;
+	unsigned long lru = 0;
 
 	/*
 	 * No need for spinlocks here since we're not terribly interested in
@@ -1064,13 +1068,16 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 	}
 	mutex_unlock(&nfsd_mutex);
 
-	for_each_possible_cpu(i)
+	for_each_possible_cpu(i) {
 		hits += per_cpu(nfsd_file_cache_hits, i);
+		acquisitions += per_cpu(nfsd_file_acquisitions, i);
+	}
 
 	seq_printf(m, "total entries: %u\n", count);
 	seq_printf(m, "longest chain: %u\n", longest);
 	seq_printf(m, "lru entries:   %lu\n", lru);
 	seq_printf(m, "cache hits:    %lu\n", hits);
+	seq_printf(m, "acquisitions:  %lu\n", acquisitions);
 	return 0;
 }
 
Signature · 1 line
2.43.0
author maintainer
#348 lore raw

PATCH 348/690 NFSD: Report count of freed filecache items

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit d63293272abb51c02457f1017dfd61c3270d9ae3 ]

Surface the count of freed nfsd_file items.

---
 fs/nfsd/filecache.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
fs/nfsd/filecache.c +6 -1
index 5a09b76ae25a8..5d31622c23040 100644
@@ -43,6 +43,7 @@ struct nfsd_fcache_bucket {
 
 static DEFINE_PER_CPU(unsigned long, nfsd_file_cache_hits);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_acquisitions);
+static DEFINE_PER_CPU(unsigned long, nfsd_file_releases);
 
 struct nfsd_fcache_disposal {
 	struct work_struct work;
@@ -195,6 +196,8 @@ nfsd_file_free(struct nfsd_file *nf)
 {
 	bool flush = false;
 
+	this_cpu_inc(nfsd_file_releases);
+
 	trace_nfsd_file_put_final(nf);
 	if (nf->nf_mark)
 		nfsd_file_mark_put(nf->nf_mark);
@@ -1049,7 +1052,7 @@ nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
  */
 static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 {
-	unsigned long hits = 0, acquisitions = 0;
+	unsigned long hits = 0, acquisitions = 0, releases = 0;
 	unsigned int i, count = 0, longest = 0;
 	unsigned long lru = 0;
 
@@ -1071,6 +1074,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 	for_each_possible_cpu(i) {
 		hits += per_cpu(nfsd_file_cache_hits, i);
 		acquisitions += per_cpu(nfsd_file_acquisitions, i);
+		releases += per_cpu(nfsd_file_releases, i);
 	}
 
 	seq_printf(m, "total entries: %u\n", count);
@@ -1078,6 +1082,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 	seq_printf(m, "lru entries:   %lu\n", lru);
 	seq_printf(m, "cache hits:    %lu\n", hits);
 	seq_printf(m, "acquisitions:  %lu\n", acquisitions);
+	seq_printf(m, "releases:      %lu\n", releases);
 	return 0;
 }
 
Signature · 1 line
2.43.0
author maintainer
#349 lore raw

PATCH 349/690 NFSD: Report average age of filecache items

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 904940e94a887701db24401e3ed6928a1d4e329f ]

This is a measure of how long items stay in the filecache, to help assess how efficient the cache is.

---
 fs/nfsd/filecache.c | 11 ++++++++++-
 fs/nfsd/filecache.h |  1 +
 2 files changed, 11 insertions(+), 1 deletion(-)
fs/nfsd/filecache.c +10 -1
index 5d31622c23040..0cd72c20fc12d 100644
@@ -44,6 +44,7 @@ struct nfsd_fcache_bucket {
 static DEFINE_PER_CPU(unsigned long, nfsd_file_cache_hits);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_acquisitions);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_releases);
+static DEFINE_PER_CPU(unsigned long, nfsd_file_total_age);
 
 struct nfsd_fcache_disposal {
 	struct work_struct work;
@@ -177,6 +178,7 @@ nfsd_file_alloc(struct inode *inode, unsigned int may, unsigned int hashval,
 	if (nf) {
 		INIT_HLIST_NODE(&nf->nf_node);
 		INIT_LIST_HEAD(&nf->nf_lru);
+		nf->nf_birthtime = ktime_get();
 		nf->nf_file = NULL;
 		nf->nf_cred = get_current_cred();
 		nf->nf_net = net;
@@ -194,9 +196,11 @@ nfsd_file_alloc(struct inode *inode, unsigned int may, unsigned int hashval,
 static bool
 nfsd_file_free(struct nfsd_file *nf)
 {
+	s64 age = ktime_to_ms(ktime_sub(ktime_get(), nf->nf_birthtime));
 	bool flush = false;
 
 	this_cpu_inc(nfsd_file_releases);
+	this_cpu_add(nfsd_file_total_age, age);
 
 	trace_nfsd_file_put_final(nf);
 	if (nf->nf_mark)
@@ -1054,7 +1058,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 {
 	unsigned long hits = 0, acquisitions = 0, releases = 0;
 	unsigned int i, count = 0, longest = 0;
-	unsigned long lru = 0;
+	unsigned long lru = 0, total_age = 0;
 
 	/*
 	 * No need for spinlocks here since we're not terribly interested in
@@ -1075,6 +1079,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 		hits += per_cpu(nfsd_file_cache_hits, i);
 		acquisitions += per_cpu(nfsd_file_acquisitions, i);
 		releases += per_cpu(nfsd_file_releases, i);
+		total_age += per_cpu(nfsd_file_total_age, i);
 	}
 
 	seq_printf(m, "total entries: %u\n", count);
@@ -1083,6 +1088,10 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 	seq_printf(m, "cache hits:    %lu\n", hits);
 	seq_printf(m, "acquisitions:  %lu\n", acquisitions);
 	seq_printf(m, "releases:      %lu\n", releases);
+	if (releases)
+		seq_printf(m, "mean age (ms): %ld\n", total_age / releases);
+	else
+		seq_printf(m, "mean age (ms): -\n");
 	return 0;
 }
 
fs/nfsd/filecache.h +1 -0
index c9e3c6eb4776e..c6ad5fe47f12f 100644
@@ -44,6 +44,7 @@ struct nfsd_file {
 	refcount_t		nf_ref;
 	unsigned char		nf_may;
 	struct nfsd_file_mark	*nf_mark;
+	ktime_t			nf_birthtime;
 };
 
 int nfsd_file_cache_init(void);
Signature · 1 line
2.43.0
author maintainer
#350 lore raw

PATCH 350/690 NFSD: Add nfsd_file_lru_dispose_list() helper

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 0bac5a264d9a923f5b01f3521e1519a8d0358342 ]

Refactor the invariant part of nfsd_file_lru_walk_list() into a separate helper function.

---
 fs/nfsd/filecache.c | 29 ++++++++++++++++++++++-------
 1 file changed, 22 insertions(+), 7 deletions(-)
fs/nfsd/filecache.c +22 -7
index 0cd72c20fc12d..ffe46f3f33495 100644
@@ -450,11 +450,31 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru,
 	return LRU_SKIP;
 }
 
+/*
+ * Unhash items on @dispose immediately, then queue them on the
+ * disposal workqueue to finish releasing them in the background.
+ *
+ * cel: Note that between the time list_lru_shrink_walk runs and
+ * now, these items are in the hash table but marked unhashed.
+ * Why release these outside of lru_cb ? There's no lock ordering
+ * problem since lru_cb currently takes no lock.
+ */
+static void nfsd_file_gc_dispose_list(struct list_head *dispose)
+{
+	struct nfsd_file *nf;
+
+	list_for_each_entry(nf, dispose, nf_lru) {
+		spin_lock(&nfsd_file_hashtbl[nf->nf_hashval].nfb_lock);
+		nfsd_file_do_unhash(nf);
+		spin_unlock(&nfsd_file_hashtbl[nf->nf_hashval].nfb_lock);
+	}
+	nfsd_file_dispose_list_delayed(dispose);
+}
+
 static unsigned long
 nfsd_file_lru_walk_list(struct shrink_control *sc)
 {
 	LIST_HEAD(head);
-	struct nfsd_file *nf;
 	unsigned long ret;
 
 	if (sc)
@@ -464,12 +484,7 @@ nfsd_file_lru_walk_list(struct shrink_control *sc)
 		ret = list_lru_walk(&nfsd_file_lru,
 				nfsd_file_lru_cb,
 				&head, LONG_MAX);
-	list_for_each_entry(nf, &head, nf_lru) {
-		spin_lock(&nfsd_file_hashtbl[nf->nf_hashval].nfb_lock);
-		nfsd_file_do_unhash(nf);
-		spin_unlock(&nfsd_file_hashtbl[nf->nf_hashval].nfb_lock);
-	}
-	nfsd_file_dispose_list_delayed(&head);
+	nfsd_file_gc_dispose_list(&head);
 	return ret;
 }
 
Signature · 1 line
2.43.0
author maintainer
#351 lore raw

PATCH 351/690 NFSD: Refactor nfsd_file_gc()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 3bc6d3470fe412f818f9bff6b71d1be3a76af8f3 ]

Refactor nfsd_file_gc() to use the new list_lru helper.

---
 fs/nfsd/filecache.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
fs/nfsd/filecache.c +5 -1
index ffe46f3f33495..656c94c779417 100644
@@ -491,7 +491,11 @@ nfsd_file_lru_walk_list(struct shrink_control *sc)
 static void
 nfsd_file_gc(void)
 {
-	nfsd_file_lru_walk_list(NULL);
+	LIST_HEAD(dispose);
+
+	list_lru_walk(&nfsd_file_lru, nfsd_file_lru_cb,
+		      &dispose, LONG_MAX);
+	nfsd_file_gc_dispose_list(&dispose);
 }
 
 static void
Signature · 1 line
2.43.0
author maintainer
#352 lore raw

PATCH 352/690 NFSD: Refactor nfsd_file_lru_scan()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 39f1d1ff8148902c5692ffb0e1c4479416ab44a7 ]

---
 fs/nfsd/filecache.c | 25 +++++++------------------
 1 file changed, 7 insertions(+), 18 deletions(-)
fs/nfsd/filecache.c +7 -18
index 656c94c779417..1d94491e5ddad 100644
@@ -471,23 +471,6 @@ static void nfsd_file_gc_dispose_list(struct list_head *dispose)
 	nfsd_file_dispose_list_delayed(dispose);
 }
 
-static unsigned long
-nfsd_file_lru_walk_list(struct shrink_control *sc)
-{
-	LIST_HEAD(head);
-	unsigned long ret;
-
-	if (sc)
-		ret = list_lru_shrink_walk(&nfsd_file_lru, sc,
-				nfsd_file_lru_cb, &head);
-	else
-		ret = list_lru_walk(&nfsd_file_lru,
-				nfsd_file_lru_cb,
-				&head, LONG_MAX);
-	nfsd_file_gc_dispose_list(&head);
-	return ret;
-}
-
 static void
 nfsd_file_gc(void)
 {
@@ -514,7 +497,13 @@ nfsd_file_lru_count(struct shrinker *s, struct shrink_control *sc)
 static unsigned long
 nfsd_file_lru_scan(struct shrinker *s, struct shrink_control *sc)
 {
-	return nfsd_file_lru_walk_list(sc);
+	LIST_HEAD(dispose);
+	unsigned long ret;
+
+	ret = list_lru_shrink_walk(&nfsd_file_lru, sc,
+				   nfsd_file_lru_cb, &dispose);
+	nfsd_file_gc_dispose_list(&dispose);
+	return ret;
 }
 
 static struct shrinker	nfsd_file_shrinker = {
Signature · 1 line
2.43.0
author maintainer
#353 lore raw

PATCH 353/690 NFSD: Report the number of items evicted by the LRU walk

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 94660cc19c75083af046b0f8362e3d3bc2eba21d ]

---
 fs/nfsd/filecache.c | 13 ++++++++++---
 fs/nfsd/trace.h     | 29 +++++++++++++++++++++++++++++
 2 files changed, 39 insertions(+), 3 deletions(-)
fs/nfsd/filecache.c +10 -3
index 1d94491e5ddad..e5bd9f06492c8 100644
@@ -45,6 +45,7 @@ static DEFINE_PER_CPU(unsigned long, nfsd_file_cache_hits);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_acquisitions);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_releases);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_total_age);
+static DEFINE_PER_CPU(unsigned long, nfsd_file_evictions);
 
 struct nfsd_fcache_disposal {
 	struct work_struct work;
@@ -445,6 +446,7 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru,
 		goto out_skip;
 
 	list_lru_isolate_move(lru, &nf->nf_lru, head);
+	this_cpu_inc(nfsd_file_evictions);
 	return LRU_REMOVED;
 out_skip:
 	return LRU_SKIP;
@@ -475,9 +477,11 @@ static void
 nfsd_file_gc(void)
 {
 	LIST_HEAD(dispose);
+	unsigned long ret;
 
-	list_lru_walk(&nfsd_file_lru, nfsd_file_lru_cb,
-		      &dispose, LONG_MAX);
+	ret = list_lru_walk(&nfsd_file_lru, nfsd_file_lru_cb,
+			    &dispose, LONG_MAX);
+	trace_nfsd_file_gc_removed(ret, list_lru_count(&nfsd_file_lru));
 	nfsd_file_gc_dispose_list(&dispose);
 }
 
@@ -502,6 +506,7 @@ nfsd_file_lru_scan(struct shrinker *s, struct shrink_control *sc)
 
 	ret = list_lru_shrink_walk(&nfsd_file_lru, sc,
 				   nfsd_file_lru_cb, &dispose);
+	trace_nfsd_file_shrinker_removed(ret, list_lru_count(&nfsd_file_lru));
 	nfsd_file_gc_dispose_list(&dispose);
 	return ret;
 }
@@ -1064,7 +1069,7 @@ nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
  */
 static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 {
-	unsigned long hits = 0, acquisitions = 0, releases = 0;
+	unsigned long hits = 0, acquisitions = 0, releases = 0, evictions = 0;
 	unsigned int i, count = 0, longest = 0;
 	unsigned long lru = 0, total_age = 0;
 
@@ -1088,6 +1093,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 		acquisitions += per_cpu(nfsd_file_acquisitions, i);
 		releases += per_cpu(nfsd_file_releases, i);
 		total_age += per_cpu(nfsd_file_total_age, i);
+		evictions += per_cpu(nfsd_file_evictions, i);
 	}
 
 	seq_printf(m, "total entries: %u\n", count);
@@ -1096,6 +1102,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 	seq_printf(m, "cache hits:    %lu\n", hits);
 	seq_printf(m, "acquisitions:  %lu\n", acquisitions);
 	seq_printf(m, "releases:      %lu\n", releases);
+	seq_printf(m, "evictions:     %lu\n", evictions);
 	if (releases)
 		seq_printf(m, "mean age (ms): %ld\n", total_age / releases);
 	else
fs/nfsd/trace.h +29 -0
index 593218d8a54d0..71919f7a31dc8 100644
@@ -894,6 +894,35 @@ TRACE_EVENT(nfsd_file_fsnotify_handle_event,
 			__entry->nlink, __entry->mode, __entry->mask)
 );
 
+DECLARE_EVENT_CLASS(nfsd_file_lruwalk_class,
+	TP_PROTO(
+		unsigned long removed,
+		unsigned long remaining
+	),
+	TP_ARGS(removed, remaining),
+	TP_STRUCT__entry(
+		__field(unsigned long, removed)
+		__field(unsigned long, remaining)
+	),
+	TP_fast_assign(
+		__entry->removed = removed;
+		__entry->remaining = remaining;
+	),
+	TP_printk("%lu entries removed, %lu remaining",
+		__entry->removed, __entry->remaining)
+);
+
+#define DEFINE_NFSD_FILE_LRUWALK_EVENT(name)				\
+DEFINE_EVENT(nfsd_file_lruwalk_class, name,				\
+	TP_PROTO(							\
+		unsigned long removed,					\
+		unsigned long remaining					\
+	),								\
+	TP_ARGS(removed, remaining))
+
+DEFINE_NFSD_FILE_LRUWALK_EVENT(nfsd_file_gc_removed);
+DEFINE_NFSD_FILE_LRUWALK_EVENT(nfsd_file_shrinker_removed);
+
 #include "cache.h"
 
 TRACE_DEFINE_ENUM(RC_DROPIT);
Signature · 1 line
2.43.0
author maintainer
#354 lore raw

PATCH 354/690 NFSD: Record number of flush calls

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit df2aff524faceaf743b7c5ab0f4fb86cb511f782 ]

---
 fs/nfsd/filecache.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
fs/nfsd/filecache.c +11 -2
index e5bd9f06492c8..b9941d4ef20d6 100644
@@ -45,6 +45,7 @@ static DEFINE_PER_CPU(unsigned long, nfsd_file_cache_hits);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_acquisitions);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_releases);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_total_age);
+static DEFINE_PER_CPU(unsigned long, nfsd_file_pages_flushed);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_evictions);
 
 struct nfsd_fcache_disposal {
@@ -242,7 +243,12 @@ nfsd_file_check_write_error(struct nfsd_file *nf)
 static void
 nfsd_file_flush(struct nfsd_file *nf)
 {
-	if (nf->nf_file && vfs_fsync(nf->nf_file, 1) != 0)
+	struct file *file = nf->nf_file;
+
+	if (!file || !(file->f_mode & FMODE_WRITE))
+		return;
+	this_cpu_add(nfsd_file_pages_flushed, file->f_mapping->nrpages);
+	if (vfs_fsync(file, 1) != 0)
 		nfsd_reset_write_verifier(net_generic(nf->nf_net, nfsd_net_id));
 }
 
@@ -1069,7 +1075,8 @@ nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
  */
 static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 {
-	unsigned long hits = 0, acquisitions = 0, releases = 0, evictions = 0;
+	unsigned long releases = 0, pages_flushed = 0, evictions = 0;
+	unsigned long hits = 0, acquisitions = 0;
 	unsigned int i, count = 0, longest = 0;
 	unsigned long lru = 0, total_age = 0;
 
@@ -1094,6 +1101,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 		releases += per_cpu(nfsd_file_releases, i);
 		total_age += per_cpu(nfsd_file_total_age, i);
 		evictions += per_cpu(nfsd_file_evictions, i);
+		pages_flushed += per_cpu(nfsd_file_pages_flushed, i);
 	}
 
 	seq_printf(m, "total entries: %u\n", count);
@@ -1107,6 +1115,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 		seq_printf(m, "mean age (ms): %ld\n", total_age / releases);
 	else
 		seq_printf(m, "mean age (ms): -\n");
+	seq_printf(m, "pages flushed: %lu\n", pages_flushed);
 	return 0;
 }
 
Signature · 1 line
2.43.0
author maintainer
#355 lore raw

PATCH 355/690 NFSD: Zero counters when the filecache is re-initialized

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 8b330f78040cbe16cf8029df70391b2a491f17e2 ]

If nfsd_file_cache_init() is called after a shutdown, be sure the stat counters are reset.

---
 fs/nfsd/filecache.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
fs/nfsd/filecache.c +11 -0
index b9941d4ef20d6..60c51a4d8e0d7 100644
@@ -823,6 +823,8 @@ nfsd_file_cache_shutdown_net(struct net *net)
 void
 nfsd_file_cache_shutdown(void)
 {
+	int i;
+
 	set_bit(NFSD_FILE_SHUTDOWN, &nfsd_file_lru_flags);
 
 	lease_unregister_notifier(&nfsd_file_lease_notifier);
@@ -846,6 +848,15 @@ nfsd_file_cache_shutdown(void)
 	nfsd_file_hashtbl = NULL;
 	destroy_workqueue(nfsd_filecache_wq);
 	nfsd_filecache_wq = NULL;
+
+	for_each_possible_cpu(i) {
+		per_cpu(nfsd_file_cache_hits, i) = 0;
+		per_cpu(nfsd_file_acquisitions, i) = 0;
+		per_cpu(nfsd_file_releases, i) = 0;
+		per_cpu(nfsd_file_total_age, i) = 0;
+		per_cpu(nfsd_file_pages_flushed, i) = 0;
+		per_cpu(nfsd_file_evictions, i) = 0;
+	}
 }
 
 static bool
Signature · 1 line
2.43.0
author maintainer
#356 lore raw

PATCH 356/690 NFSD: Hook up the filecache stat file

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 2e6c6e4c4375bfd3defa5b1ff3604d9f33d1c936 ]

There has always been the capability of exporting filecache metrics via /proc, but it was never hooked up. Let's surface these metrics to enable better observability of the filecache.

---
 fs/nfsd/nfsctl.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
fs/nfsd/nfsctl.c +10 -0
index 66c352bf61b1d..7002edbf26870 100644
@@ -25,6 +25,7 @@
 #include "state.h"
 #include "netns.h"
 #include "pnfs.h"
+#include "filecache.h"
 
 /*
  *	We have a single directory with several nodes in it.
@@ -45,6 +46,7 @@ enum {
 	NFSD_Ports,
 	NFSD_MaxBlkSize,
 	NFSD_MaxConnections,
+	NFSD_Filecache,
 	NFSD_SupportedEnctypes,
 	/*
 	 * The below MUST come last.  Otherwise we leave a hole in nfsd_files[]
@@ -229,6 +231,13 @@ static const struct file_operations reply_cache_stats_operations = {
 	.release	= single_release,
 };
 
+static const struct file_operations filecache_ops = {
+	.open		= nfsd_file_cache_stats_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
 /*----------------------------------------------------------------------------*/
 /*
  * payload - write methods
@@ -1370,6 +1379,7 @@ static int nfsd_fill_super(struct super_block *sb, struct fs_context *fc)
 		[NFSD_Ports] = {"portlist", &transaction_ops, S_IWUSR|S_IRUGO},
 		[NFSD_MaxBlkSize] = {"max_block_size", &transaction_ops, S_IWUSR|S_IRUGO},
 		[NFSD_MaxConnections] = {"max_connections", &transaction_ops, S_IWUSR|S_IRUGO},
+		[NFSD_Filecache] = {"filecache", &filecache_ops, S_IRUGO},
 #if defined(CONFIG_SUNRPC_GSS) || defined(CONFIG_SUNRPC_GSS_MODULE)
 		[NFSD_SupportedEnctypes] = {"supported_krb5_enctypes", &supported_enctypes_ops, S_IRUGO},
 #endif /* CONFIG_SUNRPC_GSS or CONFIG_SUNRPC_GSS_MODULE */
Signature · 1 line
2.43.0
author maintainer
#357 lore raw

PATCH 357/690 NFSD: WARN when freeing an item still linked via nf_lru

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 668ed92e651d3c25f9b6e8cb7ceca54d00daa96d ]

Add a guardrail to prevent freeing memory that is still on a list. This includes either a dispose list or the LRU list.

This is the sign of a bug, but this class of bugs can be detected so that they don't endanger system stability, especially while debugging.

---
 fs/nfsd/filecache.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)
fs/nfsd/filecache.c +10 -2
index 60c51a4d8e0d7..d9b5f1e183976 100644
@@ -213,6 +213,14 @@ nfsd_file_free(struct nfsd_file *nf)
 		fput(nf->nf_file);
 		flush = true;
 	}
+
+	/*
+	 * If this item is still linked via nf_lru, that's a bug.
+	 * WARN and leak it to preserve system stability.
+	 */
+	if (WARN_ON_ONCE(!list_empty(&nf->nf_lru)))
+		return flush;
+
 	call_rcu(&nf->nf_rcu, nfsd_file_slab_free);
 	return flush;
 }
@@ -342,7 +350,7 @@ nfsd_file_dispose_list(struct list_head *dispose)
 
 	while(!list_empty(dispose)) {
 		nf = list_first_entry(dispose, struct nfsd_file, nf_lru);
-		list_del(&nf->nf_lru);
+		list_del_init(&nf->nf_lru);
 		nfsd_file_flush(nf);
 		nfsd_file_put_noref(nf);
 	}
@@ -356,7 +364,7 @@ nfsd_file_dispose_list_sync(struct list_head *dispose)
 
 	while(!list_empty(dispose)) {
 		nf = list_first_entry(dispose, struct nfsd_file, nf_lru);
-		list_del(&nf->nf_lru);
+		list_del_init(&nf->nf_lru);
 		nfsd_file_flush(nf);
 		if (!refcount_dec_and_test(&nf->nf_ref))
 			continue;
Signature · 1 line
2.43.0
author maintainer
#358 lore raw

PATCH 358/690 NFSD: Trace filecache LRU activity

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit c46203acddd9b9200dbc53d0603c97355fd3a03b ]

Observe the operation of garbage collection and the lifetime of filecache items.

---
 fs/nfsd/filecache.c | 44 +++++++++++++++++++++++++++++++-------------
 fs/nfsd/trace.h     | 39 +++++++++++++++++++++++++++++++++++++++
 2 files changed, 70 insertions(+), 13 deletions(-)
fs/nfsd/filecache.c +31 -13
index d9b5f1e183976..a995a744a7481 100644
@@ -260,6 +260,18 @@ nfsd_file_flush(struct nfsd_file *nf)
 		nfsd_reset_write_verifier(net_generic(nf->nf_net, nfsd_net_id));
 }
 
+static void nfsd_file_lru_add(struct nfsd_file *nf)
+{
+	if (list_lru_add(&nfsd_file_lru, &nf->nf_lru))
+		trace_nfsd_file_lru_add(nf);
+}
+
+static void nfsd_file_lru_remove(struct nfsd_file *nf)
+{
+	if (list_lru_del(&nfsd_file_lru, &nf->nf_lru))
+		trace_nfsd_file_lru_del(nf);
+}
+
 static void
 nfsd_file_do_unhash(struct nfsd_file *nf)
 {
@@ -279,8 +291,7 @@ nfsd_file_unhash(struct nfsd_file *nf)
 {
 	if (test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags)) {
 		nfsd_file_do_unhash(nf);
-		if (!list_empty(&nf->nf_lru))
-			list_lru_del(&nfsd_file_lru, &nf->nf_lru);
+		nfsd_file_lru_remove(nf);
 		return true;
 	}
 	return false;
@@ -443,27 +454,34 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru,
 	 * counter. Here we check the counter and then test and clear the flag.
 	 * That order is deliberate to ensure that we can do this locklessly.
 	 */
-	if (refcount_read(&nf->nf_ref) > 1)
-		goto out_skip;
+	if (refcount_read(&nf->nf_ref) > 1) {
+		trace_nfsd_file_gc_in_use(nf);
+		return LRU_SKIP;
+	}
 
 	/*
 	 * Don't throw out files that are still undergoing I/O or
 	 * that have uncleared errors pending.
 	 */
-	if (nfsd_file_check_writeback(nf))
-		goto out_skip;
+	if (nfsd_file_check_writeback(nf)) {
+		trace_nfsd_file_gc_writeback(nf);
+		return LRU_SKIP;
+	}
 
-	if (test_and_clear_bit(NFSD_FILE_REFERENCED, &nf->nf_flags))
-		goto out_skip;
+	if (test_and_clear_bit(NFSD_FILE_REFERENCED, &nf->nf_flags)) {
+		trace_nfsd_file_gc_referenced(nf);
+		return LRU_SKIP;
+	}
 
-	if (!test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags))
-		goto out_skip;
+	if (!test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags)) {
+		trace_nfsd_file_gc_hashed(nf);
+		return LRU_SKIP;
+	}
 
 	list_lru_isolate_move(lru, &nf->nf_lru, head);
 	this_cpu_inc(nfsd_file_evictions);
+	trace_nfsd_file_gc_disposed(nf);
 	return LRU_REMOVED;
-out_skip:
-	return LRU_SKIP;
 }
 
 /*
@@ -1016,7 +1034,7 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	refcount_inc(&nf->nf_ref);
 	__set_bit(NFSD_FILE_HASHED, &nf->nf_flags);
 	__set_bit(NFSD_FILE_PENDING, &nf->nf_flags);
-	list_lru_add(&nfsd_file_lru, &nf->nf_lru);
+	nfsd_file_lru_add(nf);
 	hlist_add_head_rcu(&nf->nf_node, &nfsd_file_hashtbl[hashval].nfb_head);
 	++nfsd_file_hashtbl[hashval].nfb_count;
 	nfsd_file_hashtbl[hashval].nfb_maxcount = max(nfsd_file_hashtbl[hashval].nfb_maxcount,
fs/nfsd/trace.h +39 -0
index 71919f7a31dc8..c47f46d433ddb 100644
@@ -894,6 +894,45 @@ TRACE_EVENT(nfsd_file_fsnotify_handle_event,
 			__entry->nlink, __entry->mode, __entry->mask)
 );
 
+DECLARE_EVENT_CLASS(nfsd_file_gc_class,
+	TP_PROTO(
+		const struct nfsd_file *nf
+	),
+	TP_ARGS(nf),
+	TP_STRUCT__entry(
+		__field(void *, nf_inode)
+		__field(void *, nf_file)
+		__field(int, nf_ref)
+		__field(unsigned long, nf_flags)
+	),
+	TP_fast_assign(
+		__entry->nf_inode = nf->nf_inode;
+		__entry->nf_file = nf->nf_file;
+		__entry->nf_ref = refcount_read(&nf->nf_ref);
+		__entry->nf_flags = nf->nf_flags;
+	),
+	TP_printk("inode=%p ref=%d nf_flags=%s nf_file=%p",
+		__entry->nf_inode, __entry->nf_ref,
+		show_nf_flags(__entry->nf_flags),
+		__entry->nf_file
+	)
+);
+
+#define DEFINE_NFSD_FILE_GC_EVENT(name)					\
+DEFINE_EVENT(nfsd_file_gc_class, name,					\
+	TP_PROTO(							\
+		const struct nfsd_file *nf				\
+	),								\
+	TP_ARGS(nf))
+
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_lru_add);
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_lru_del);
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_in_use);
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_writeback);
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_referenced);
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_hashed);
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_disposed);
+
 DECLARE_EVENT_CLASS(nfsd_file_lruwalk_class,
 	TP_PROTO(
 		unsigned long removed,
Signature · 1 line
2.43.0
author maintainer
#359 lore raw

PATCH 359/690 NFSD: Leave open files out of the filecache LRU

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 4a0e73e635e3f36b616ad5c943e3d23debe4632f ]

There have been reports of problems when running fstests generic/531 against Linux NFS servers with NFSv4. The NFS server that hosts the test's SCRATCH_DEV suffers from CPU soft lock-ups during the test. Analysis shows that:

fs/nfsd/filecache.c
482 ret = list_lru_walk(&nfsd_file_lru,
483 nfsd_file_lru_cb,
484 &head, LONG_MAX);

causes nfsd_file_gc() to walk the entire length of the filecache LRU list every time it is called (which is quite frequently). The walk holds a spinlock the entire time that prevents other nfsd threads from accessing the filecache.

What's more, for NFSv4 workloads, none of the items that are visited during this walk may be evicted, since they are all files that are held OPEN by NFS clients.

Address this by ensuring that open files are not kept on the LRU list.

---
 fs/nfsd/filecache.c | 24 +++++++++++++++++++-----
 fs/nfsd/trace.h     |  2 ++
 2 files changed, 21 insertions(+), 5 deletions(-)
fs/nfsd/filecache.c +19 -5
index a995a744a7481..5c9e3ff6397b0 100644
@@ -262,6 +262,7 @@ nfsd_file_flush(struct nfsd_file *nf)
 
 static void nfsd_file_lru_add(struct nfsd_file *nf)
 {
+	set_bit(NFSD_FILE_REFERENCED, &nf->nf_flags);
 	if (list_lru_add(&nfsd_file_lru, &nf->nf_lru))
 		trace_nfsd_file_lru_add(nf);
 }
@@ -291,7 +292,6 @@ nfsd_file_unhash(struct nfsd_file *nf)
 {
 	if (test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags)) {
 		nfsd_file_do_unhash(nf);
-		nfsd_file_lru_remove(nf);
 		return true;
 	}
 	return false;
@@ -312,6 +312,7 @@ nfsd_file_unhash_and_release_locked(struct nfsd_file *nf, struct list_head *disp
 	if (refcount_dec_not_one(&nf->nf_ref))
 		return true;
 
+	nfsd_file_lru_remove(nf);
 	list_add(&nf->nf_lru, dispose);
 	return true;
 }
@@ -323,6 +324,7 @@ nfsd_file_put_noref(struct nfsd_file *nf)
 
 	if (refcount_dec_and_test(&nf->nf_ref)) {
 		WARN_ON(test_bit(NFSD_FILE_HASHED, &nf->nf_flags));
+		nfsd_file_lru_remove(nf);
 		nfsd_file_free(nf);
 	}
 }
@@ -332,7 +334,7 @@ nfsd_file_put(struct nfsd_file *nf)
 {
 	might_sleep();
 
-	set_bit(NFSD_FILE_REFERENCED, &nf->nf_flags);
+	nfsd_file_lru_add(nf);
 	if (test_bit(NFSD_FILE_HASHED, &nf->nf_flags) == 0) {
 		nfsd_file_flush(nf);
 		nfsd_file_put_noref(nf);
@@ -432,8 +434,18 @@ nfsd_file_dispose_list_delayed(struct list_head *dispose)
 	}
 }
 
-/*
+/**
+ * nfsd_file_lru_cb - Examine an entry on the LRU list
+ * @item: LRU entry to examine
+ * @lru: controlling LRU
+ * @lock: LRU list lock (unused)
+ * @arg: dispose list
+ *
  * Note this can deadlock with nfsd_file_cache_purge.
+ *
+ * Return values:
+ *   %LRU_REMOVED: @item was removed from the LRU
+ *   %LRU_SKIP: @item cannot be evicted
  */
 static enum lru_status
 nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru,
@@ -455,8 +467,9 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru,
 	 * That order is deliberate to ensure that we can do this locklessly.
 	 */
 	if (refcount_read(&nf->nf_ref) > 1) {
+		list_lru_isolate(lru, &nf->nf_lru);
 		trace_nfsd_file_gc_in_use(nf);
-		return LRU_SKIP;
+		return LRU_REMOVED;
 	}
 
 	/*
@@ -1013,6 +1026,7 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		goto retry;
 	}
 
+	nfsd_file_lru_remove(nf);
 	this_cpu_inc(nfsd_file_cache_hits);
 
 	status = nfserrno(nfsd_open_break_lease(file_inode(nf->nf_file), may_flags));
@@ -1034,7 +1048,6 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	refcount_inc(&nf->nf_ref);
 	__set_bit(NFSD_FILE_HASHED, &nf->nf_flags);
 	__set_bit(NFSD_FILE_PENDING, &nf->nf_flags);
-	nfsd_file_lru_add(nf);
 	hlist_add_head_rcu(&nf->nf_node, &nfsd_file_hashtbl[hashval].nfb_head);
 	++nfsd_file_hashtbl[hashval].nfb_count;
 	nfsd_file_hashtbl[hashval].nfb_maxcount = max(nfsd_file_hashtbl[hashval].nfb_maxcount,
@@ -1059,6 +1072,7 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	 */
 	if (status != nfs_ok || inode->i_nlink == 0) {
 		bool do_free;
+		nfsd_file_lru_remove(nf);
 		spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
 		do_free = nfsd_file_unhash(nf);
 		spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
fs/nfsd/trace.h +2 -0
index c47f46d433ddb..cc55a2b32e8cd 100644
@@ -926,7 +926,9 @@ DEFINE_EVENT(nfsd_file_gc_class, name,					\
 	TP_ARGS(nf))
 
 DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_lru_add);
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_lru_add_disposed);
 DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_lru_del);
+DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_lru_del_disposed);
 DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_in_use);
 DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_writeback);
 DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_referenced);
Signature · 1 line
2.43.0
author maintainer
#360 lore raw

PATCH 360/690 NFSD: Fix the filecache LRU shrinker

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit edead3a55804739b2e4af0f35e9c7326264e7b22 ]

Without LRU item rotation, the shrinker visits only a few items on the end of the LRU list, and those would always be long-term OPEN files for NFSv4 workloads. That makes the filecache shrinker completely ineffective.

Adopt the same strategy as the inode LRU by using LRU_ROTATE.

---
 fs/nfsd/filecache.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
fs/nfsd/filecache.c +3 -2
index 5c9e3ff6397b0..849c010c6ef61 100644
@@ -445,6 +445,7 @@ nfsd_file_dispose_list_delayed(struct list_head *dispose)
  *
  * Return values:
  *   %LRU_REMOVED: @item was removed from the LRU
+ *   %LRU_ROTATE: @item is to be moved to the LRU tail
  *   %LRU_SKIP: @item cannot be evicted
  */
 static enum lru_status
@@ -483,7 +484,7 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru,
 
 	if (test_and_clear_bit(NFSD_FILE_REFERENCED, &nf->nf_flags)) {
 		trace_nfsd_file_gc_referenced(nf);
-		return LRU_SKIP;
+		return LRU_ROTATE;
 	}
 
 	if (!test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags)) {
@@ -525,7 +526,7 @@ nfsd_file_gc(void)
 	unsigned long ret;
 
 	ret = list_lru_walk(&nfsd_file_lru, nfsd_file_lru_cb,
-			    &dispose, LONG_MAX);
+			    &dispose, list_lru_count(&nfsd_file_lru));
 	trace_nfsd_file_gc_removed(ret, list_lru_count(&nfsd_file_lru));
 	nfsd_file_gc_dispose_list(&dispose);
 }
Signature · 1 line
2.43.0
author maintainer
#361 lore raw

PATCH 361/690 NFSD: Never call nfsd_file_gc() in foreground paths

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 6df19411367a5fb4ef61854cbd1af269c077f917 ]

The checks in nfsd_file_acquire() and nfsd_file_put() that directly invoke filecache garbage collection are intended to keep cache occupancy between a low- and high-watermark. The reason to limit the capacity of the filecache is to keep filecache lookups reasonably fast.

However, invoking garbage collection at those points has some undesirable negative impacts. Files that are held open by NFSv4 clients often push the occupancy of the filecache over these watermarks. At that point:

- Every call to nfsd_file_acquire() and nfsd_file_put() results in an LRU walk. This has the same effect on lookup latency as long chains in the hash table.
- Garbage collection will then run on every nfsd thread, causing a lot of unnecessary lock contention.
- Limiting cache capacity pushes out files used only by NFSv3 clients, which are the type of files the filecache is supposed to help.

To address those negative impacts, remove the direct calls to the garbage collector. Subsequent patches will address maintaining lookup efficiency as cache capacity increases.

---
 fs/nfsd/filecache.c | 10 +---------
 1 file changed, 1 insertion(+), 9 deletions(-)
fs/nfsd/filecache.c +1 -9
index 849c010c6ef61..7a02ff11b9ec1 100644
@@ -28,8 +28,6 @@
 #define NFSD_LAUNDRETTE_DELAY		     (2 * HZ)
 
 #define NFSD_FILE_SHUTDOWN		     (1)
-#define NFSD_FILE_LRU_THRESHOLD		     (4096UL)
-#define NFSD_FILE_LRU_LIMIT		     (NFSD_FILE_LRU_THRESHOLD << 2)
 
 /* We only care about NFSD_MAY_READ/WRITE for this cache */
 #define NFSD_FILE_MAY_MASK	(NFSD_MAY_READ|NFSD_MAY_WRITE)
@@ -65,8 +63,6 @@ static struct fsnotify_group		*nfsd_file_fsnotify_group;
 static atomic_long_t			nfsd_filecache_count;
 static struct delayed_work		nfsd_filecache_laundrette;
 
-static void nfsd_file_gc(void);
-
 static void
 nfsd_file_schedule_laundrette(void)
 {
@@ -343,9 +339,6 @@ nfsd_file_put(struct nfsd_file *nf)
 		nfsd_file_schedule_laundrette();
 	} else
 		nfsd_file_put_noref(nf);
-
-	if (atomic_long_read(&nfsd_filecache_count) >= NFSD_FILE_LRU_LIMIT)
-		nfsd_file_gc();
 }
 
 struct nfsd_file *
@@ -1054,8 +1047,7 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	nfsd_file_hashtbl[hashval].nfb_maxcount = max(nfsd_file_hashtbl[hashval].nfb_maxcount,
 			nfsd_file_hashtbl[hashval].nfb_count);
 	spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
-	if (atomic_long_inc_return(&nfsd_filecache_count) >= NFSD_FILE_LRU_THRESHOLD)
-		nfsd_file_gc();
+	atomic_long_inc(&nfsd_filecache_count);
 
 	nf->nf_mark = nfsd_file_mark_find_or_create(nf);
 	if (nf->nf_mark) {
Signature · 1 line
2.43.0
author maintainer
#362 lore raw

PATCH 362/690 NFSD: No longer record nf_hashval in the trace log

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 54f7df7094b329ca35d9f9808692bb16c48b13e9 ]

I'm about to replace nfsd_file_hashtbl with an rhashtable. The individual hash values will no longer be visible or relevant, so remove them from the tracepoints.

---
 fs/nfsd/filecache.c | 15 ++++++++-------
 fs/nfsd/trace.h     | 45 +++++++++++++++++++++------------------------
 2 files changed, 29 insertions(+), 31 deletions(-)
fs/nfsd/filecache.c +8 -7
index 7a02ff11b9ec1..2d013a88e3565 100644
@@ -588,7 +588,7 @@ nfsd_file_close_inode_sync(struct inode *inode)
 	LIST_HEAD(dispose);
 
 	__nfsd_file_close_inode(inode, hashval, &dispose);
-	trace_nfsd_file_close_inode_sync(inode, hashval, !list_empty(&dispose));
+	trace_nfsd_file_close_inode_sync(inode, !list_empty(&dispose));
 	nfsd_file_dispose_list_sync(&dispose);
 }
 
@@ -608,7 +608,7 @@ nfsd_file_close_inode(struct inode *inode)
 	LIST_HEAD(dispose);
 
 	__nfsd_file_close_inode(inode, hashval, &dispose);
-	trace_nfsd_file_close_inode(inode, hashval, !list_empty(&dispose));
+	trace_nfsd_file_close_inode(inode, !list_empty(&dispose));
 	nfsd_file_dispose_list_delayed(&dispose);
 }
 
@@ -962,7 +962,7 @@ nfsd_file_is_cached(struct inode *inode)
 		}
 	}
 	rcu_read_unlock();
-	trace_nfsd_file_is_cached(inode, hashval, (int)ret);
+	trace_nfsd_file_is_cached(inode, (int)ret);
 	return ret;
 }
 
@@ -994,9 +994,8 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 	new = nfsd_file_alloc(inode, may_flags, hashval, net);
 	if (!new) {
-		trace_nfsd_file_acquire(rqstp, hashval, inode, may_flags,
-					NULL, nfserr_jukebox);
-		return nfserr_jukebox;
+		status = nfserr_jukebox;
+		goto out_status;
 	}
 
 	spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
@@ -1034,8 +1033,10 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		nf = NULL;
 	}
 
-	trace_nfsd_file_acquire(rqstp, hashval, inode, may_flags, nf, status);
+out_status:
+	trace_nfsd_file_acquire(rqstp, inode, may_flags, nf, status);
 	return status;
+
 open_file:
 	nf = new;
 	/* Take reference for the hashtable */
fs/nfsd/trace.h +21 -24

121 lines not on this page · raw message

Signature · 1 line
2.43.0
author maintainer
#363 lore raw

PATCH 363/690 NFSD: Remove lockdep assertion from unhash_and_release_locked()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit f53cef15dddec7203df702cdc62e554190385450 ]

IIUC, holding the hash bucket lock is needed only in nfsd_file_unhash, and there is already a lockdep assertion there.

---
 fs/nfsd/filecache.c | 2 --
 1 file changed, 2 deletions(-)
fs/nfsd/filecache.c +0 -2
index 2d013a88e3565..6a01de8677959 100644
@@ -299,8 +299,6 @@ nfsd_file_unhash(struct nfsd_file *nf)
 static bool
 nfsd_file_unhash_and_release_locked(struct nfsd_file *nf, struct list_head *dispose)
 {
-	lockdep_assert_held(&nfsd_file_hashtbl[nf->nf_hashval].nfb_lock);
-
 	trace_nfsd_file_unhash_and_release_locked(nf);
 	if (!nfsd_file_unhash(nf))
 		return false;
Signature · 1 line
2.43.0
author maintainer
#364 lore raw

PATCH 364/690 NFSD: nfsd_file_unhash can compute hashval from nf->nf_inode

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 8755326399f471ec3b31e2ab8c5074c0d28a0fb5 ]

Remove an unnecessary usage of nf_hashval.

---
 fs/nfsd/filecache.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
fs/nfsd/filecache.c +6 -2
index 6a01de8677959..d7c74b51eabf3 100644
@@ -272,13 +272,17 @@ static void nfsd_file_lru_remove(struct nfsd_file *nf)
 static void
 nfsd_file_do_unhash(struct nfsd_file *nf)
 {
-	lockdep_assert_held(&nfsd_file_hashtbl[nf->nf_hashval].nfb_lock);
+	struct inode *inode = nf->nf_inode;
+	unsigned int hashval = (unsigned int)hash_long(inode->i_ino,
+				NFSD_FILE_HASH_BITS);
+
+	lockdep_assert_held(&nfsd_file_hashtbl[hashval].nfb_lock);
 
 	trace_nfsd_file_unhash(nf);
 
 	if (nfsd_file_check_write_error(nf))
 		nfsd_reset_write_verifier(net_generic(nf->nf_net, nfsd_net_id));
-	--nfsd_file_hashtbl[nf->nf_hashval].nfb_count;
+	--nfsd_file_hashtbl[hashval].nfb_count;
 	hlist_del_rcu(&nf->nf_node);
 	atomic_long_dec(&nfsd_filecache_count);
 }
Signature · 1 line
2.43.0
author maintainer
#365 lore raw

PATCH 365/690 NFSD: Refactor __nfsd_file_close_inode()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit a845511007a63467fee575353c706806c21218b1 ]

The code that computes the hashval is the same in both callers.

To prevent them from going stale, reframe the documenting comments to remove descriptions of the underlying hash table structure, which is about to be replaced.

---
 fs/nfsd/filecache.c | 40 +++++++++++++++++++++-------------------
 fs/nfsd/trace.h     | 44 +++++++++++++++++++++++++++++++++-----------
 2 files changed, 54 insertions(+), 30 deletions(-)
fs/nfsd/filecache.c +21 -19
index d7c74b51eabf3..3925df9124c39 100644
@@ -558,39 +558,44 @@ static struct shrinker	nfsd_file_shrinker = {
 	.seeks = 1,
 };
 
-static void
-__nfsd_file_close_inode(struct inode *inode, unsigned int hashval,
-			struct list_head *dispose)
+/*
+ * Find all cache items across all net namespaces that match @inode and
+ * move them to @dispose. The lookup is atomic wrt nfsd_file_acquire().
+ */
+static unsigned int
+__nfsd_file_close_inode(struct inode *inode, struct list_head *dispose)
 {
+	unsigned int		hashval = (unsigned int)hash_long(inode->i_ino,
+						NFSD_FILE_HASH_BITS);
+	unsigned int		count = 0;
 	struct nfsd_file	*nf;
 	struct hlist_node	*tmp;
 
 	spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
 	hlist_for_each_entry_safe(nf, tmp, &nfsd_file_hashtbl[hashval].nfb_head, nf_node) {
-		if (inode == nf->nf_inode)
+		if (inode == nf->nf_inode) {
 			nfsd_file_unhash_and_release_locked(nf, dispose);
+			count++;
+		}
 	}
 	spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
+	return count;
 }
 
 /**
  * nfsd_file_close_inode_sync - attempt to forcibly close a nfsd_file
  * @inode: inode of the file to attempt to remove
  *
- * Walk the whole hash bucket, looking for any files that correspond to "inode".
- * If any do, then unhash them and put the hashtable reference to them and
- * destroy any that had their last reference put. Also ensure that any of the
- * fputs also have their final __fput done as well.
+ * Unhash and put, then flush and fput all cache items associated with @inode.
  */
 void
 nfsd_file_close_inode_sync(struct inode *inode)
 {
-	unsigned int		hashval = (unsigned int)hash_long(inode->i_ino,
-						NFSD_FILE_HASH_BITS);
 	LIST_HEAD(dispose);
+	unsigned int count;
 
-	__nfsd_file_close_inode(inode, hashval, &dispose);
-	trace_nfsd_file_close_inode_sync(inode, !list_empty(&dispose));
+	count = __nfsd_file_close_inode(inode, &dispose);
+	trace_nfsd_file_close_inode_sync(inode, count);
 	nfsd_file_dispose_list_sync(&dispose);
 }
 
@@ -598,19 +603,16 @@ nfsd_file_close_inode_sync(struct inode *inode)
  * nfsd_file_close_inode - attempt a delayed close of a nfsd_file
  * @inode: inode of the file to attempt to remove
  *
- * Walk the whole hash bucket, looking for any files that correspond to "inode".
- * If any do, then unhash them and put the hashtable reference to them and
- * destroy any that had their last reference put.
+ * Unhash and put all cache item associated with @inode.
  */
 static void
 nfsd_file_close_inode(struct inode *inode)
 {
-	unsigned int		hashval = (unsigned int)hash_long(inode->i_ino,
-						NFSD_FILE_HASH_BITS);
 	LIST_HEAD(dispose);
+	unsigned int count;
 
-	__nfsd_file_close_inode(inode, hashval, &dispose);
-	trace_nfsd_file_close_inode(inode, !list_empty(&dispose));
+	count = __nfsd_file_close_inode(inode, &dispose);
+	trace_nfsd_file_close_inode(inode, count);
 	nfsd_file_dispose_list_delayed(&dispose);
 }
 
fs/nfsd/trace.h +33 -11

63 lines not on this page · raw message

Signature · 1 line
2.43.0
author maintainer
#366 lore raw

PATCH 366/690 NFSD: nfsd_file_hash_remove can compute hashval

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit cb7ec76e73ff6640241c8f1f2f35c81d4005a2d6 ]

Remove an unnecessary use of nf_hashval.

---
 fs/nfsd/filecache.c | 19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)
fs/nfsd/filecache.c +14 -5
index 3925df9124c39..dd59deec8b011 100644
@@ -287,6 +287,18 @@ nfsd_file_do_unhash(struct nfsd_file *nf)
 	atomic_long_dec(&nfsd_filecache_count);
 }
 
+static void
+nfsd_file_hash_remove(struct nfsd_file *nf)
+{
+	struct inode *inode = nf->nf_inode;
+	unsigned int hashval = (unsigned int)hash_long(inode->i_ino,
+				NFSD_FILE_HASH_BITS);
+
+	spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
+	nfsd_file_do_unhash(nf);
+	spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
+}
+
 static bool
 nfsd_file_unhash(struct nfsd_file *nf)
 {
@@ -506,11 +518,8 @@ static void nfsd_file_gc_dispose_list(struct list_head *dispose)
 {
 	struct nfsd_file *nf;
 
-	list_for_each_entry(nf, dispose, nf_lru) {
-		spin_lock(&nfsd_file_hashtbl[nf->nf_hashval].nfb_lock);
-		nfsd_file_do_unhash(nf);
-		spin_unlock(&nfsd_file_hashtbl[nf->nf_hashval].nfb_lock);
-	}
+	list_for_each_entry(nf, dispose, nf_lru)
+		nfsd_file_hash_remove(nf);
 	nfsd_file_dispose_list_delayed(dispose);
 }
 
Signature · 1 line
2.43.0
author maintainer
#367 lore raw

PATCH 367/690 NFSD: Remove nfsd_file::nf_hashval

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit f0743c2b25c65debd4f599a7c861428cd9de5906 ]

The value in this field can always be computed from nf_inode, thus it is no longer used.

---
 fs/nfsd/filecache.c | 6 ++----
 fs/nfsd/filecache.h | 1 -
 2 files changed, 2 insertions(+), 5 deletions(-)
fs/nfsd/filecache.c +2 -4
index dd59deec8b011..29b1f57692a60 100644
@@ -167,8 +167,7 @@ nfsd_file_mark_find_or_create(struct nfsd_file *nf)
 }
 
 static struct nfsd_file *
-nfsd_file_alloc(struct inode *inode, unsigned int may, unsigned int hashval,
-		struct net *net)
+nfsd_file_alloc(struct inode *inode, unsigned int may, struct net *net)
 {
 	struct nfsd_file *nf;
 
@@ -182,7 +181,6 @@ nfsd_file_alloc(struct inode *inode, unsigned int may, unsigned int hashval,
 		nf->nf_net = net;
 		nf->nf_flags = 0;
 		nf->nf_inode = inode;
-		nf->nf_hashval = hashval;
 		refcount_set(&nf->nf_ref, 1);
 		nf->nf_may = may & NFSD_FILE_MAY_MASK;
 		nf->nf_mark = NULL;
@@ -1005,7 +1003,7 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (nf)
 		goto wait_for_construction;
 
-	new = nfsd_file_alloc(inode, may_flags, hashval, net);
+	new = nfsd_file_alloc(inode, may_flags, net);
 	if (!new) {
 		status = nfserr_jukebox;
 		goto out_status;
fs/nfsd/filecache.h +0 -1
index c6ad5fe47f12f..82051e1b8420d 100644
@@ -40,7 +40,6 @@ struct nfsd_file {
 #define NFSD_FILE_REFERENCED	(2)
 	unsigned long		nf_flags;
 	struct inode		*nf_inode;
-	unsigned int		nf_hashval;
 	refcount_t		nf_ref;
 	unsigned char		nf_may;
 	struct nfsd_file_mark	*nf_mark;
Signature · 1 line
2.43.0
author maintainer
#368 lore raw

PATCH 368/690 NFSD: Replace the "init once" mechanism

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit c7b824c3d06c85e054caf86e227255112c5e3c38 ]

In a moment, the nfsd_file_hashtbl global will be replaced with an rhashtable. Replace the one or two spots that need to check if the hash table is available. We can easily reuse the SHUTDOWN flag for this purpose.

Document that this mechanism relies on callers to hold the nfsd_mutex to prevent init, shutdown, and purging to run concurrently.

---
 fs/nfsd/filecache.c | 42 ++++++++++++++++++++++++++----------------
 1 file changed, 26 insertions(+), 16 deletions(-)
fs/nfsd/filecache.c +26 -16
index 29b1f57692a60..33bb4d31b4972 100644
@@ -27,7 +27,7 @@
 #define NFSD_FILE_HASH_SIZE                  (1 << NFSD_FILE_HASH_BITS)
 #define NFSD_LAUNDRETTE_DELAY		     (2 * HZ)
 
-#define NFSD_FILE_SHUTDOWN		     (1)
+#define NFSD_FILE_CACHE_UP		     (0)
 
 /* We only care about NFSD_MAY_READ/WRITE for this cache */
 #define NFSD_FILE_MAY_MASK	(NFSD_MAY_READ|NFSD_MAY_WRITE)
@@ -58,7 +58,7 @@ static struct kmem_cache		*nfsd_file_slab;
 static struct kmem_cache		*nfsd_file_mark_slab;
 static struct nfsd_fcache_bucket	*nfsd_file_hashtbl;
 static struct list_lru			nfsd_file_lru;
-static long				nfsd_file_lru_flags;
+static unsigned long			nfsd_file_flags;
 static struct fsnotify_group		*nfsd_file_fsnotify_group;
 static atomic_long_t			nfsd_filecache_count;
 static struct delayed_work		nfsd_filecache_laundrette;
@@ -66,9 +66,8 @@ static struct delayed_work		nfsd_filecache_laundrette;
 static void
 nfsd_file_schedule_laundrette(void)
 {
-	long count = atomic_long_read(&nfsd_filecache_count);
-
-	if (count == 0 || test_bit(NFSD_FILE_SHUTDOWN, &nfsd_file_lru_flags))
+	if ((atomic_long_read(&nfsd_filecache_count) == 0) ||
+	    test_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 0)
 		return;
 
 	queue_delayed_work(system_wq, &nfsd_filecache_laundrette,
@@ -697,9 +696,8 @@ nfsd_file_cache_init(void)
 	int		ret = -ENOMEM;
 	unsigned int	i;
 
-	clear_bit(NFSD_FILE_SHUTDOWN, &nfsd_file_lru_flags);
-
-	if (nfsd_file_hashtbl)
+	lockdep_assert_held(&nfsd_mutex);
+	if (test_and_set_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 1)
 		return 0;
 
 	nfsd_filecache_wq = alloc_workqueue("nfsd_filecache", 0, 0);
@@ -785,8 +783,8 @@ nfsd_file_cache_init(void)
 /*
  * Note this can deadlock with nfsd_file_lru_cb.
  */
-void
-nfsd_file_cache_purge(struct net *net)
+static void
+__nfsd_file_cache_purge(struct net *net)
 {
 	unsigned int		i;
 	struct nfsd_file	*nf;
@@ -794,9 +792,6 @@ nfsd_file_cache_purge(struct net *net)
 	LIST_HEAD(dispose);
 	bool del;
 
-	if (!nfsd_file_hashtbl)
-		return;
-
 	for (i = 0; i < NFSD_FILE_HASH_SIZE; i++) {
 		struct nfsd_fcache_bucket *nfb = &nfsd_file_hashtbl[i];
 
@@ -857,6 +852,19 @@ nfsd_file_cache_start_net(struct net *net)
 	return nn->fcache_disposal ? 0 : -ENOMEM;
 }
 
+/**
+ * nfsd_file_cache_purge - Remove all cache items associated with @net
+ * @net: target net namespace
+ *
+ */
+void
+nfsd_file_cache_purge(struct net *net)
+{
+	lockdep_assert_held(&nfsd_mutex);
+	if (test_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 1)
+		__nfsd_file_cache_purge(net);
+}
+
 void
 nfsd_file_cache_shutdown_net(struct net *net)
 {
@@ -869,7 +877,9 @@ nfsd_file_cache_shutdown(void)
 {
 	int i;
 
-	set_bit(NFSD_FILE_SHUTDOWN, &nfsd_file_lru_flags);
+	lockdep_assert_held(&nfsd_mutex);
+	if (test_and_clear_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 0)
+		return;
 
 	lease_unregister_notifier(&nfsd_file_lease_notifier);
 	unregister_shrinker(&nfsd_file_shrinker);
@@ -878,7 +888,7 @@ nfsd_file_cache_shutdown(void)
 	 * calling nfsd_file_cache_purge
 	 */
 	cancel_delayed_work_sync(&nfsd_filecache_laundrette);
-	nfsd_file_cache_purge(NULL);
+	__nfsd_file_cache_purge(NULL);
 	list_lru_destroy(&nfsd_file_lru);
 	rcu_barrier();
 	fsnotify_put_group(nfsd_file_fsnotify_group);
@@ -1142,7 +1152,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 	 * don't end up racing with server shutdown
 	 */
 	mutex_lock(&nfsd_mutex);
-	if (nfsd_file_hashtbl) {
+	if (test_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 1) {
 		for (i = 0; i < NFSD_FILE_HASH_SIZE; i++) {
 			count += nfsd_file_hashtbl[i].nfb_count;
 			longest = max(longest, nfsd_file_hashtbl[i].nfb_count);
Signature · 1 line
2.43.0
author maintainer
#369 lore raw

PATCH 369/690 NFSD: Set up an rhashtable for the filecache

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit fc22945ecc2a0a028f3683115f98a922d506c284 ]

Add code to initialize and tear down an rhashtable. The rhashtable is not used yet.

---
 fs/nfsd/filecache.c | 160 ++++++++++++++++++++++++++++++++++++++------
 fs/nfsd/filecache.h |   1 +
 2 files changed, 140 insertions(+), 21 deletions(-)
fs/nfsd/filecache.c +139 -21
index 33bb4d31b4972..95e7e15b567e2 100644
@@ -12,6 +12,7 @@
 #include <linux/fsnotify_backend.h>
 #include <linux/fsnotify.h>
 #include <linux/seq_file.h>
+#include <linux/rhashtable.h>
 
 #include "vfs.h"
 #include "nfsd.h"
@@ -62,6 +63,136 @@ static unsigned long			nfsd_file_flags;
 static struct fsnotify_group		*nfsd_file_fsnotify_group;
 static atomic_long_t			nfsd_filecache_count;
 static struct delayed_work		nfsd_filecache_laundrette;
+static struct rhashtable		nfsd_file_rhash_tbl
+						____cacheline_aligned_in_smp;
+
+enum nfsd_file_lookup_type {
+	NFSD_FILE_KEY_INODE,
+	NFSD_FILE_KEY_FULL,
+};
+
+struct nfsd_file_lookup_key {
+	struct inode			*inode;
+	struct net			*net;
+	const struct cred		*cred;
+	unsigned char			need;
+	enum nfsd_file_lookup_type	type;
+};
+
+/*
+ * The returned hash value is based solely on the address of an in-code
+ * inode, a pointer to a slab-allocated object. The entropy in such a
+ * pointer is concentrated in its middle bits.
+ */
+static u32 nfsd_file_inode_hash(const struct inode *inode, u32 seed)
+{
+	unsigned long ptr = (unsigned long)inode;
+	u32 k;
+
+	k = ptr >> L1_CACHE_SHIFT;
+	k &= 0x00ffffff;
+	return jhash2(&k, 1, seed);
+}
+
+/**
+ * nfsd_file_key_hashfn - Compute the hash value of a lookup key
+ * @data: key on which to compute the hash value
+ * @len: rhash table's key_len parameter (unused)
+ * @seed: rhash table's random seed of the day
+ *
+ * Return value:
+ *   Computed 32-bit hash value
+ */
+static u32 nfsd_file_key_hashfn(const void *data, u32 len, u32 seed)
+{
+	const struct nfsd_file_lookup_key *key = data;
+
+	return nfsd_file_inode_hash(key->inode, seed);
+}
+
+/**
+ * nfsd_file_obj_hashfn - Compute the hash value of an nfsd_file
+ * @data: object on which to compute the hash value
+ * @len: rhash table's key_len parameter (unused)
+ * @seed: rhash table's random seed of the day
+ *
+ * Return value:
+ *   Computed 32-bit hash value
+ */
+static u32 nfsd_file_obj_hashfn(const void *data, u32 len, u32 seed)
+{
+	const struct nfsd_file *nf = data;
+
+	return nfsd_file_inode_hash(nf->nf_inode, seed);
+}
+
+static bool
+nfsd_match_cred(const struct cred *c1, const struct cred *c2)
+{
+	int i;
+
+	if (!uid_eq(c1->fsuid, c2->fsuid))
+		return false;
+	if (!gid_eq(c1->fsgid, c2->fsgid))
+		return false;
+	if (c1->group_info == NULL || c2->group_info == NULL)
+		return c1->group_info == c2->group_info;
+	if (c1->group_info->ngroups != c2->group_info->ngroups)
+		return false;
+	for (i = 0; i < c1->group_info->ngroups; i++) {
+		if (!gid_eq(c1->group_info->gid[i], c2->group_info->gid[i]))
+			return false;
+	}
+	return true;
+}
+
+/**
+ * nfsd_file_obj_cmpfn - Match a cache item against search criteria
+ * @arg: search criteria
+ * @ptr: cache item to check
+ *
+ * Return values:
+ *   %0 - Item matches search criteria
+ *   %1 - Item does not match search criteria
+ */
+static int nfsd_file_obj_cmpfn(struct rhashtable_compare_arg *arg,
+			       const void *ptr)
+{
+	const struct nfsd_file_lookup_key *key = arg->key;
+	const struct nfsd_file *nf = ptr;
+
+	switch (key->type) {
+	case NFSD_FILE_KEY_INODE:
+		if (nf->nf_inode != key->inode)
+			return 1;
+		break;
+	case NFSD_FILE_KEY_FULL:
+		if (nf->nf_inode != key->inode)
+			return 1;
+		if (nf->nf_may != key->need)
+			return 1;
+		if (nf->nf_net != key->net)
+			return 1;
+		if (!nfsd_match_cred(nf->nf_cred, key->cred))
+			return 1;
+		if (test_bit(NFSD_FILE_HASHED, &nf->nf_flags) == 0)
+			return 1;
+		break;
+	}
+	return 0;
+}
+
+static const struct rhashtable_params nfsd_file_rhash_params = {
+	.key_len		= sizeof_field(struct nfsd_file, nf_inode),
+	.key_offset		= offsetof(struct nfsd_file, nf_inode),
+	.head_offset		= offsetof(struct nfsd_file, nf_rhash),
+	.hashfn			= nfsd_file_key_hashfn,
+	.obj_hashfn		= nfsd_file_obj_hashfn,
+	.obj_cmpfn		= nfsd_file_obj_cmpfn,
+	/* Reduce resizing churn on light workloads */
+	.min_size		= 512,		/* buckets */
+	.automatic_shrinking	= true,
+};
 
 static void
 nfsd_file_schedule_laundrette(void)
@@ -693,13 +824,18 @@ static const struct fsnotify_ops nfsd_file_fsnotify_ops = {
 int
 nfsd_file_cache_init(void)
 {
-	int		ret = -ENOMEM;
+	int		ret;
 	unsigned int	i;
 
 	lockdep_assert_held(&nfsd_mutex);
 	if (test_and_set_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 1)
 		return 0;
 
+	ret = rhashtable_init(&nfsd_file_rhash_tbl, &nfsd_file_rhash_params);
+	if (ret)
+		return ret;
+
+	ret = -ENOMEM;
 	nfsd_filecache_wq = alloc_workqueue("nfsd_filecache", 0, 0);
 	if (!nfsd_filecache_wq)
 		goto out;
@@ -777,6 +913,7 @@ nfsd_file_cache_init(void)
 	nfsd_file_hashtbl = NULL;
 	destroy_workqueue(nfsd_filecache_wq);
 	nfsd_filecache_wq = NULL;
+	rhashtable_destroy(&nfsd_file_rhash_tbl);
 	goto out;
 }
 
@@ -902,6 +1039,7 @@ nfsd_file_cache_shutdown(void)
 	nfsd_file_hashtbl = NULL;
 	destroy_workqueue(nfsd_filecache_wq);
 	nfsd_filecache_wq = NULL;
+	rhashtable_destroy(&nfsd_file_rhash_tbl);
 
 	for_each_possible_cpu(i) {
 		per_cpu(nfsd_file_cache_hits, i) = 0;
@@ -913,26 +1051,6 @@ nfsd_file_cache_shutdown(void)
 	}
 }
 
-static bool
-nfsd_match_cred(const struct cred *c1, const struct cred *c2)
-{
-	int i;
-
-	if (!uid_eq(c1->fsuid, c2->fsuid))
-		return false;
-	if (!gid_eq(c1->fsgid, c2->fsgid))
-		return false;
-	if (c1->group_info == NULL || c2->group_info == NULL)
-		return c1->group_info == c2->group_info;
-	if (c1->group_info->ngroups != c2->group_info->ngroups)
-		return false;
-	for (i = 0; i < c1->group_info->ngroups; i++) {
-		if (!gid_eq(c1->group_info->gid[i], c2->group_info->gid[i]))
-			return false;
-	}
-	return true;
-}
-
 static struct nfsd_file *
 nfsd_file_find_locked(struct inode *inode, unsigned int may_flags,
 			unsigned int hashval, struct net *net)
fs/nfsd/filecache.h +1 -0
index 82051e1b8420d..5cbfc61a7d7d9 100644
@@ -29,6 +29,7 @@ struct nfsd_file_mark {
  * never be dereferenced, only used for comparison.
  */
 struct nfsd_file {
+	struct rhash_head	nf_rhash;
 	struct hlist_node	nf_node;
 	struct list_head	nf_lru;
 	struct rcu_head		nf_rcu;
Signature · 1 line
2.43.0
author maintainer
#370 lore raw

PATCH 370/690 NFSD: Convert the filecache to use rhashtable

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit ce502f81ba884c1fe45dc0ebddbcaaa4ec0fc5fb ]

Enable the filecache hash table to start small, then grow with the workload. Smaller server deployments benefit because there should be lower memory utilization. Larger server deployments should see improved scaling with the number of open files.

---
 fs/nfsd/filecache.c | 265 +++++++++++++++++++-------------------------
 fs/nfsd/trace.h     |  63 ++++++++++-
 2 files changed, 179 insertions(+), 149 deletions(-)
fs/nfsd/filecache.c +117 -148
index 95e7e15b567e2..45dd4f3fa0905 100644
@@ -61,7 +61,6 @@ static struct nfsd_fcache_bucket	*nfsd_file_hashtbl;
 static struct list_lru			nfsd_file_lru;
 static unsigned long			nfsd_file_flags;
 static struct fsnotify_group		*nfsd_file_fsnotify_group;
-static atomic_long_t			nfsd_filecache_count;
 static struct delayed_work		nfsd_filecache_laundrette;
 static struct rhashtable		nfsd_file_rhash_tbl
 						____cacheline_aligned_in_smp;
@@ -197,7 +196,7 @@ static const struct rhashtable_params nfsd_file_rhash_params = {
 static void
 nfsd_file_schedule_laundrette(void)
 {
-	if ((atomic_long_read(&nfsd_filecache_count) == 0) ||
+	if ((atomic_read(&nfsd_file_rhash_tbl.nelems) == 0) ||
 	    test_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 0)
 		return;
 
@@ -297,7 +296,7 @@ nfsd_file_mark_find_or_create(struct nfsd_file *nf)
 }
 
 static struct nfsd_file *
-nfsd_file_alloc(struct inode *inode, unsigned int may, struct net *net)
+nfsd_file_alloc(struct nfsd_file_lookup_key *key, unsigned int may)
 {
 	struct nfsd_file *nf;
 
@@ -308,11 +307,14 @@ nfsd_file_alloc(struct inode *inode, unsigned int may, struct net *net)
 		nf->nf_birthtime = ktime_get();
 		nf->nf_file = NULL;
 		nf->nf_cred = get_current_cred();
-		nf->nf_net = net;
+		nf->nf_net = key->net;
 		nf->nf_flags = 0;
-		nf->nf_inode = inode;
-		refcount_set(&nf->nf_ref, 1);
-		nf->nf_may = may & NFSD_FILE_MAY_MASK;
+		__set_bit(NFSD_FILE_HASHED, &nf->nf_flags);
+		__set_bit(NFSD_FILE_PENDING, &nf->nf_flags);
+		nf->nf_inode = key->inode;
+		/* nf_ref is pre-incremented for hash table */
+		refcount_set(&nf->nf_ref, 2);
+		nf->nf_may = key->need;
 		nf->nf_mark = NULL;
 		trace_nfsd_file_alloc(nf);
 	}
@@ -398,40 +400,21 @@ static void nfsd_file_lru_remove(struct nfsd_file *nf)
 }
 
 static void
-nfsd_file_do_unhash(struct nfsd_file *nf)
+nfsd_file_hash_remove(struct nfsd_file *nf)
 {
-	struct inode *inode = nf->nf_inode;
-	unsigned int hashval = (unsigned int)hash_long(inode->i_ino,
-				NFSD_FILE_HASH_BITS);
-
-	lockdep_assert_held(&nfsd_file_hashtbl[hashval].nfb_lock);
-
 	trace_nfsd_file_unhash(nf);
 
 	if (nfsd_file_check_write_error(nf))
 		nfsd_reset_write_verifier(net_generic(nf->nf_net, nfsd_net_id));
-	--nfsd_file_hashtbl[hashval].nfb_count;
-	hlist_del_rcu(&nf->nf_node);
-	atomic_long_dec(&nfsd_filecache_count);
-}
-
-static void
-nfsd_file_hash_remove(struct nfsd_file *nf)
-{
-	struct inode *inode = nf->nf_inode;
-	unsigned int hashval = (unsigned int)hash_long(inode->i_ino,
-				NFSD_FILE_HASH_BITS);
-
-	spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
-	nfsd_file_do_unhash(nf);
-	spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
+	rhashtable_remove_fast(&nfsd_file_rhash_tbl, &nf->nf_rhash,
+			       nfsd_file_rhash_params);
 }
 
 static bool
 nfsd_file_unhash(struct nfsd_file *nf)
 {
 	if (test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags)) {
-		nfsd_file_do_unhash(nf);
+		nfsd_file_hash_remove(nf);
 		return true;
 	}
 	return false;
@@ -441,9 +424,9 @@ nfsd_file_unhash(struct nfsd_file *nf)
  * Return true if the file was unhashed.
  */
 static bool
-nfsd_file_unhash_and_release_locked(struct nfsd_file *nf, struct list_head *dispose)
+nfsd_file_unhash_and_dispose(struct nfsd_file *nf, struct list_head *dispose)
 {
-	trace_nfsd_file_unhash_and_release_locked(nf);
+	trace_nfsd_file_unhash_and_dispose(nf);
 	if (!nfsd_file_unhash(nf))
 		return false;
 	/* keep final reference for nfsd_file_lru_dispose */
@@ -702,20 +685,23 @@ static struct shrinker	nfsd_file_shrinker = {
 static unsigned int
 __nfsd_file_close_inode(struct inode *inode, struct list_head *dispose)
 {
-	unsigned int		hashval = (unsigned int)hash_long(inode->i_ino,
-						NFSD_FILE_HASH_BITS);
-	unsigned int		count = 0;
-	struct nfsd_file	*nf;
-	struct hlist_node	*tmp;
-
-	spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
-	hlist_for_each_entry_safe(nf, tmp, &nfsd_file_hashtbl[hashval].nfb_head, nf_node) {
-		if (inode == nf->nf_inode) {
-			nfsd_file_unhash_and_release_locked(nf, dispose);
-			count++;
-		}
-	}
-	spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
+	struct nfsd_file_lookup_key key = {
+		.type	= NFSD_FILE_KEY_INODE,
+		.inode	= inode,
+	};
+	unsigned int count = 0;
+	struct nfsd_file *nf;
+
+	rcu_read_lock();
+	do {
+		nf = rhashtable_lookup(&nfsd_file_rhash_tbl, &key,
+				       nfsd_file_rhash_params);
+		if (!nf)
+			break;
+		nfsd_file_unhash_and_dispose(nf, dispose);
+		count++;
+	} while (1);
+	rcu_read_unlock();
 	return count;
 }
 
@@ -923,30 +909,35 @@ nfsd_file_cache_init(void)
 static void
 __nfsd_file_cache_purge(struct net *net)
 {
-	unsigned int		i;
-	struct nfsd_file	*nf;
-	struct hlist_node	*next;
+	struct rhashtable_iter iter;
+	struct nfsd_file *nf;
 	LIST_HEAD(dispose);
 	bool del;
 
-	for (i = 0; i < NFSD_FILE_HASH_SIZE; i++) {
-		struct nfsd_fcache_bucket *nfb = &nfsd_file_hashtbl[i];
+	rhashtable_walk_enter(&nfsd_file_rhash_tbl, &iter);
+	do {
+		rhashtable_walk_start(&iter);
 
-		spin_lock(&nfb->nfb_lock);
-		hlist_for_each_entry_safe(nf, next, &nfb->nfb_head, nf_node) {
+		nf = rhashtable_walk_next(&iter);
+		while (!IS_ERR_OR_NULL(nf)) {
 			if (net && nf->nf_net != net)
 				continue;
-			del = nfsd_file_unhash_and_release_locked(nf, &dispose);
+			del = nfsd_file_unhash_and_dispose(nf, &dispose);
 
 			/*
 			 * Deadlock detected! Something marked this entry as
 			 * unhased, but hasn't removed it from the hash list.
 			 */
 			WARN_ON_ONCE(!del);
+
+			nf = rhashtable_walk_next(&iter);
 		}
-		spin_unlock(&nfb->nfb_lock);
-		nfsd_file_dispose_list(&dispose);
-	}
+
+		rhashtable_walk_stop(&iter);
+	} while (nf == ERR_PTR(-EAGAIN));
+	rhashtable_walk_exit(&iter);
+
+	nfsd_file_dispose_list(&dispose);
 }
 
 static struct nfsd_fcache_disposal *
@@ -1051,56 +1042,29 @@ nfsd_file_cache_shutdown(void)
 	}
 }
 
-static struct nfsd_file *
-nfsd_file_find_locked(struct inode *inode, unsigned int may_flags,
-			unsigned int hashval, struct net *net)
-{
-	struct nfsd_file *nf;
-	unsigned char need = may_flags & NFSD_FILE_MAY_MASK;
-
-	hlist_for_each_entry_rcu(nf, &nfsd_file_hashtbl[hashval].nfb_head,
-				 nf_node, lockdep_is_held(&nfsd_file_hashtbl[hashval].nfb_lock)) {
-		if (nf->nf_may != need)
-			continue;
-		if (nf->nf_inode != inode)
-			continue;
-		if (nf->nf_net != net)
-			continue;
-		if (!nfsd_match_cred(nf->nf_cred, current_cred()))
-			continue;
-		if (!test_bit(NFSD_FILE_HASHED, &nf->nf_flags))
-			continue;
-		if (nfsd_file_get(nf) != NULL)
-			return nf;
-	}
-	return NULL;
-}
-
 /**
- * nfsd_file_is_cached - are there any cached open files for this fh?
- * @inode: inode of the file to check
+ * nfsd_file_is_cached - are there any cached open files for this inode?
+ * @inode: inode to check
+ *
+ * The lookup matches inodes in all net namespaces and is atomic wrt
+ * nfsd_file_acquire().
  *
- * Scan the hashtable for open files that match this fh. Returns true if there
- * are any, and false if not.
+ * Return values:
+ *   %true: filecache contains at least one file matching this inode
+ *   %false: filecache contains no files matching this inode
  */
 bool
 nfsd_file_is_cached(struct inode *inode)
 {
-	bool			ret = false;
-	struct nfsd_file	*nf;
-	unsigned int		hashval;
-
-        hashval = (unsigned int)hash_long(inode->i_ino, NFSD_FILE_HASH_BITS);
-
-	rcu_read_lock();
-	hlist_for_each_entry_rcu(nf, &nfsd_file_hashtbl[hashval].nfb_head,
-				 nf_node) {
-		if (inode == nf->nf_inode) {
-			ret = true;
-			break;
-		}
-	}
-	rcu_read_unlock();
+	struct nfsd_file_lookup_key key = {
+		.type	= NFSD_FILE_KEY_INODE,
+		.inode	= inode,
+	};
+	bool ret = false;
+
+	if (rhashtable_lookup_fast(&nfsd_file_rhash_tbl, &key,
+				   nfsd_file_rhash_params) != NULL)
+		ret = true;
 	trace_nfsd_file_is_cached(inode, (int)ret);
 	return ret;
 }
@@ -1109,39 +1073,51 @@ static __be32
 nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		     unsigned int may_flags, struct nfsd_file **pnf, bool open)
 {
-	__be32	status;
-	struct net *net = SVC_NET(rqstp);
+	struct nfsd_file_lookup_key key = {
+		.type	= NFSD_FILE_KEY_FULL,
+		.need	= may_flags & NFSD_FILE_MAY_MASK,
+		.net	= SVC_NET(rqstp),
+	};
 	struct nfsd_file *nf, *new;
-	struct inode *inode;
-	unsigned int hashval;
 	bool retry = true;
+	__be32 status;
 
-	/* FIXME: skip this if fh_dentry is already set? */
 	status = fh_verify(rqstp, fhp, S_IFREG,
 				may_flags|NFSD_MAY_OWNER_OVERRIDE);
 	if (status != nfs_ok)
 		return status;
+	key.inode = d_inode(fhp->fh_dentry);
+	key.cred = get_current_cred();
 
-	inode = d_inode(fhp->fh_dentry);
-	hashval = (unsigned int)hash_long(inode->i_ino, NFSD_FILE_HASH_BITS);
 retry:
-	rcu_read_lock();
-	nf = nfsd_file_find_locked(inode, may_flags, hashval, net);
-	rcu_read_unlock();
+	/* Avoid allocation if the item is already in cache */
+	nf = rhashtable_lookup_fast(&nfsd_file_rhash_tbl, &key,
+				    nfsd_file_rhash_params);
+	if (nf)
+		nf = nfsd_file_get(nf);
 	if (nf)
 		goto wait_for_construction;
 
-	new = nfsd_file_alloc(inode, may_flags, net);
+	new = nfsd_file_alloc(&key, may_flags);
 	if (!new) {
 		status = nfserr_jukebox;
 		goto out_status;
 	}
 
-	spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
-	nf = nfsd_file_find_locked(inode, may_flags, hashval, net);
-	if (nf == NULL)
+	nf = rhashtable_lookup_get_insert_key(&nfsd_file_rhash_tbl,
+					      &key, &new->nf_rhash,
+					      nfsd_file_rhash_params);
+	if (!nf) {
+		nf = new;
+		goto open_file;
+	}
+	if (IS_ERR(nf))
+		goto insert_err;
+	nf = nfsd_file_get(nf);
+	if (nf == NULL) {
+		nf = new;
 		goto open_file;
-	spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
+	}
 	nfsd_file_slab_free(&new->nf_rcu);
 
 wait_for_construction:
@@ -1149,6 +1125,7 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 	/* Did construction of this file fail? */
 	if (!test_bit(NFSD_FILE_HASHED, &nf->nf_flags)) {
+		trace_nfsd_file_cons_err(rqstp, key.inode, may_flags, nf);
 		if (!retry) {
 			status = nfserr_jukebox;
 			goto out;
@@ -1173,22 +1150,11 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	}
 
 out_status:
-	trace_nfsd_file_acquire(rqstp, inode, may_flags, nf, status);
+	put_cred(key.cred);
+	trace_nfsd_file_acquire(rqstp, key.inode, may_flags, nf, status);
 	return status;
 
 open_file:
-	nf = new;
-	/* Take reference for the hashtable */
-	refcount_inc(&nf->nf_ref);
-	__set_bit(NFSD_FILE_HASHED, &nf->nf_flags);
-	__set_bit(NFSD_FILE_PENDING, &nf->nf_flags);
-	hlist_add_head_rcu(&nf->nf_node, &nfsd_file_hashtbl[hashval].nfb_head);
-	++nfsd_file_hashtbl[hashval].nfb_count;
-	nfsd_file_hashtbl[hashval].nfb_maxcount = max(nfsd_file_hashtbl[hashval].nfb_maxcount,
-			nfsd_file_hashtbl[hashval].nfb_count);
-	spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
-	atomic_long_inc(&nfsd_filecache_count);
-
 	nf->nf_mark = nfsd_file_mark_find_or_create(nf);
 	if (nf->nf_mark) {
 		if (open) {
@@ -1203,19 +1169,20 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	 * If construction failed, or we raced with a call to unlink()
 	 * then unhash.
 	 */
-	if (status != nfs_ok || inode->i_nlink == 0) {
-		bool do_free;
-		nfsd_file_lru_remove(nf);
-		spin_lock(&nfsd_file_hashtbl[hashval].nfb_lock);
-		do_free = nfsd_file_unhash(nf);
-		spin_unlock(&nfsd_file_hashtbl[hashval].nfb_lock);
-		if (do_free)
+	if (status != nfs_ok || key.inode->i_nlink == 0)
+		if (nfsd_file_unhash(nf))
 			nfsd_file_put_noref(nf);
-	}
 	clear_bit_unlock(NFSD_FILE_PENDING, &nf->nf_flags);
 	smp_mb__after_atomic();
 	wake_up_bit(&nf->nf_flags, NFSD_FILE_PENDING);
 	goto out;
+
+insert_err:
+	nfsd_file_slab_free(&new->nf_rcu);
+	trace_nfsd_file_insert_err(rqstp, key.inode, may_flags, PTR_ERR(nf));
+	nf = NULL;
+	status = nfserr_jukebox;
+	goto out_status;
 }
 
 /**
@@ -1261,21 +1228,23 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 {
 	unsigned long releases = 0, pages_flushed = 0, evictions = 0;
 	unsigned long hits = 0, acquisitions = 0;
-	unsigned int i, count = 0, longest = 0;
+	unsigned int i, count = 0, buckets = 0;
 	unsigned long lru = 0, total_age = 0;
 
-	/*
-	 * No need for spinlocks here since we're not terribly interested in
-	 * accuracy. We do take the nfsd_mutex simply to ensure that we
-	 * don't end up racing with server shutdown
-	 */
+	/* Serialize with server shutdown */
 	mutex_lock(&nfsd_mutex);
 	if (test_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 1) {
-		for (i = 0; i < NFSD_FILE_HASH_SIZE; i++) {
-			count += nfsd_file_hashtbl[i].nfb_count;
-			longest = max(longest, nfsd_file_hashtbl[i].nfb_count);
-		}
+		struct bucket_table *tbl;
+		struct rhashtable *ht;
+
 		lru = list_lru_count(&nfsd_file_lru);
+
+		rcu_read_lock();
+		ht = &nfsd_file_rhash_tbl;
+		count = atomic_read(&ht->nelems);
+		tbl = rht_dereference_rcu(ht->tbl, ht);
+		buckets = tbl->size;
+		rcu_read_unlock();
 	}
 	mutex_unlock(&nfsd_mutex);
 
@@ -1289,7 +1258,7 @@ static int nfsd_file_cache_stats_show(struct seq_file *m, void *v)
 	}
 
 	seq_printf(m, "total entries: %u\n", count);
-	seq_printf(m, "longest chain: %u\n", longest);
+	seq_printf(m, "hash buckets:  %u\n", buckets);
 	seq_printf(m, "lru entries:   %lu\n", lru);
 	seq_printf(m, "cache hits:    %lu\n", hits);
 	seq_printf(m, "acquisitions:  %lu\n", acquisitions);
fs/nfsd/trace.h +62 -1

75 lines not on this page · raw message

Signature · 1 line
2.43.0
author maintainer
#371 lore raw

PATCH 371/690 NFSD: Clean up unused code after rhashtable conversion

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 0ec8e9d1539a7b8109a554028bbce441052f847e ]

---
 fs/nfsd/filecache.c | 33 +--------------------------------
 fs/nfsd/filecache.h |  1 -
 2 files changed, 1 insertion(+), 33 deletions(-)
fs/nfsd/filecache.c +1 -32
index 45dd4f3fa0905..c6dc55c0f758b 100644
@@ -21,11 +21,6 @@
 #include "filecache.h"
 #include "trace.h"
 
-#define NFSDDBG_FACILITY	NFSDDBG_FH
-
-/* FIXME: dynamically size this for the machine somehow? */
-#define NFSD_FILE_HASH_BITS                   12
-#define NFSD_FILE_HASH_SIZE                  (1 << NFSD_FILE_HASH_BITS)
 #define NFSD_LAUNDRETTE_DELAY		     (2 * HZ)
 
 #define NFSD_FILE_CACHE_UP		     (0)
@@ -33,13 +28,6 @@
 /* We only care about NFSD_MAY_READ/WRITE for this cache */
 #define NFSD_FILE_MAY_MASK	(NFSD_MAY_READ|NFSD_MAY_WRITE)
 
-struct nfsd_fcache_bucket {
-	struct hlist_head	nfb_head;
-	spinlock_t		nfb_lock;
-	unsigned int		nfb_count;
-	unsigned int		nfb_maxcount;
-};
-
 static DEFINE_PER_CPU(unsigned long, nfsd_file_cache_hits);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_acquisitions);
 static DEFINE_PER_CPU(unsigned long, nfsd_file_releases);
@@ -57,7 +45,6 @@ static struct workqueue_struct *nfsd_filecache_wq __read_mostly;
 
 static struct kmem_cache		*nfsd_file_slab;
 static struct kmem_cache		*nfsd_file_mark_slab;
-static struct nfsd_fcache_bucket	*nfsd_file_hashtbl;
 static struct list_lru			nfsd_file_lru;
 static unsigned long			nfsd_file_flags;
 static struct fsnotify_group		*nfsd_file_fsnotify_group;
@@ -302,7 +289,6 @@ nfsd_file_alloc(struct nfsd_file_lookup_key *key, unsigned int may)
 
 	nf = kmem_cache_alloc(nfsd_file_slab, GFP_KERNEL);
 	if (nf) {
-		INIT_HLIST_NODE(&nf->nf_node);
 		INIT_LIST_HEAD(&nf->nf_lru);
 		nf->nf_birthtime = ktime_get();
 		nf->nf_file = NULL;
@@ -810,8 +796,7 @@ static const struct fsnotify_ops nfsd_file_fsnotify_ops = {
 int
 nfsd_file_cache_init(void)
 {
-	int		ret;
-	unsigned int	i;
+	int ret;
 
 	lockdep_assert_held(&nfsd_mutex);
 	if (test_and_set_bit(NFSD_FILE_CACHE_UP, &nfsd_file_flags) == 1)
@@ -826,13 +811,6 @@ nfsd_file_cache_init(void)
 	if (!nfsd_filecache_wq)
 		goto out;
 
-	nfsd_file_hashtbl = kvcalloc(NFSD_FILE_HASH_SIZE,
-				sizeof(*nfsd_file_hashtbl), GFP_KERNEL);
-	if (!nfsd_file_hashtbl) {
-		pr_err("nfsd: unable to allocate nfsd_file_hashtbl\n");
-		goto out_err;
-	}
-
 	nfsd_file_slab = kmem_cache_create("nfsd_file",
 				sizeof(struct nfsd_file), 0, 0, NULL);
 	if (!nfsd_file_slab) {
@@ -876,11 +854,6 @@ nfsd_file_cache_init(void)
 		goto out_notifier;
 	}
 
-	for (i = 0; i < NFSD_FILE_HASH_SIZE; i++) {
-		INIT_HLIST_HEAD(&nfsd_file_hashtbl[i].nfb_head);
-		spin_lock_init(&nfsd_file_hashtbl[i].nfb_lock);
-	}
-
 	INIT_DELAYED_WORK(&nfsd_filecache_laundrette, nfsd_file_gc_worker);
 out:
 	return ret;
@@ -895,8 +868,6 @@ nfsd_file_cache_init(void)
 	nfsd_file_slab = NULL;
 	kmem_cache_destroy(nfsd_file_mark_slab);
 	nfsd_file_mark_slab = NULL;
-	kvfree(nfsd_file_hashtbl);
-	nfsd_file_hashtbl = NULL;
 	destroy_workqueue(nfsd_filecache_wq);
 	nfsd_filecache_wq = NULL;
 	rhashtable_destroy(&nfsd_file_rhash_tbl);
@@ -1026,8 +997,6 @@ nfsd_file_cache_shutdown(void)
 	fsnotify_wait_marks_destroyed();
 	kmem_cache_destroy(nfsd_file_mark_slab);
 	nfsd_file_mark_slab = NULL;
-	kvfree(nfsd_file_hashtbl);
-	nfsd_file_hashtbl = NULL;
 	destroy_workqueue(nfsd_filecache_wq);
 	nfsd_filecache_wq = NULL;
 	rhashtable_destroy(&nfsd_file_rhash_tbl);
fs/nfsd/filecache.h +0 -1
index 5cbfc61a7d7d9..ee9ed99d8b8fa 100644
@@ -30,7 +30,6 @@ struct nfsd_file_mark {
  */
 struct nfsd_file {
 	struct rhash_head	nf_rhash;
-	struct hlist_node	nf_node;
 	struct list_head	nf_lru;
 	struct rcu_head		nf_rcu;
 	struct file		*nf_file;
Signature · 1 line
2.43.0
author maintainer
#372 lore raw

PATCH 372/690 NFSD: Separate tracepoints for acquire and create

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit be0230069fcbf7d332d010b57c1d0cfd623a84d6 ]

These tracepoints collect different information: the create case does not open a file, so there's no nf_file available.

---
 fs/nfsd/filecache.c |  9 ++++----
 fs/nfsd/nfs4state.c |  1 +
 fs/nfsd/trace.h     | 54 ++++++++++++++++++++++++++++++++++++++-------
 3 files changed, 52 insertions(+), 12 deletions(-)
fs/nfsd/filecache.c +5 -4
index c6dc55c0f758b..85813affb8abf 100644
@@ -1039,7 +1039,7 @@ nfsd_file_is_cached(struct inode *inode)
 }
 
 static __be32
-nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
+nfsd_file_do_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		     unsigned int may_flags, struct nfsd_file **pnf, bool open)
 {
 	struct nfsd_file_lookup_key key = {
@@ -1120,7 +1120,8 @@ nfsd_do_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 out_status:
 	put_cred(key.cred);
-	trace_nfsd_file_acquire(rqstp, key.inode, may_flags, nf, status);
+	if (open)
+		trace_nfsd_file_acquire(rqstp, key.inode, may_flags, nf, status);
 	return status;
 
 open_file:
@@ -1168,7 +1169,7 @@ __be32
 nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		  unsigned int may_flags, struct nfsd_file **pnf)
 {
-	return nfsd_do_file_acquire(rqstp, fhp, may_flags, pnf, true);
+	return nfsd_file_do_acquire(rqstp, fhp, may_flags, pnf, true);
 }
 
 /**
@@ -1185,7 +1186,7 @@ __be32
 nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		 unsigned int may_flags, struct nfsd_file **pnf)
 {
-	return nfsd_do_file_acquire(rqstp, fhp, may_flags, pnf, false);
+	return nfsd_file_do_acquire(rqstp, fhp, may_flags, pnf, false);
 }
 
 /*
fs/nfsd/nfs4state.c +1 -0
index 08700b6acba31..d349abf0821d6 100644
@@ -5121,6 +5121,7 @@ static __be32 nfs4_get_vfs_file(struct svc_rqst *rqstp, struct nfs4_file *fp,
 				goto out_put_access;
 			nf->nf_file = open->op_filp;
 			open->op_filp = NULL;
+			trace_nfsd_file_create(rqstp, access, nf);
 		}
 
 		spin_lock(&fp->fi_lock);
fs/nfsd/trace.h +46 -8

80 lines not on this page · raw message

Signature · 1 line
2.43.0
author maintainer
#373 lore raw

PATCH 373/690 NFSD: Move nfsd_file_trace_alloc() tracepoint

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit b40a2839470cd62ed68c4a32d72a18ee8975b1ac ]

Avoid recording the allocation of an nfsd_file item that is immediately released because a matching item was already inserted in the hash.

---
 fs/nfsd/filecache.c |  2 +-
 fs/nfsd/trace.h     | 25 ++++++++++++++++++++++++-
 2 files changed, 25 insertions(+), 2 deletions(-)
fs/nfsd/filecache.c +1 -1
index 85813affb8abf..26cfae138b906 100644
@@ -302,7 +302,6 @@ nfsd_file_alloc(struct nfsd_file_lookup_key *key, unsigned int may)
 		refcount_set(&nf->nf_ref, 2);
 		nf->nf_may = key->need;
 		nf->nf_mark = NULL;
-		trace_nfsd_file_alloc(nf);
 	}
 	return nf;
 }
@@ -1125,6 +1124,7 @@ nfsd_file_do_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	return status;
 
 open_file:
+	trace_nfsd_file_alloc(nf);
 	nf->nf_mark = nfsd_file_mark_find_or_create(nf);
 	if (nf->nf_mark) {
 		if (open) {
fs/nfsd/trace.h +24 -1
index 12dbc190e6595..c824ab30a758e 100644
@@ -773,12 +773,35 @@ DEFINE_EVENT(nfsd_file_class, name, \
 	TP_PROTO(struct nfsd_file *nf), \
 	TP_ARGS(nf))
 
-DEFINE_NFSD_FILE_EVENT(nfsd_file_alloc);
 DEFINE_NFSD_FILE_EVENT(nfsd_file_put_final);
 DEFINE_NFSD_FILE_EVENT(nfsd_file_unhash);
 DEFINE_NFSD_FILE_EVENT(nfsd_file_put);
 DEFINE_NFSD_FILE_EVENT(nfsd_file_unhash_and_dispose);
 
+TRACE_EVENT(nfsd_file_alloc,
+	TP_PROTO(
+		const struct nfsd_file *nf
+	),
+	TP_ARGS(nf),
+	TP_STRUCT__entry(
+		__field(const void *, nf_inode)
+		__field(unsigned long, nf_flags)
+		__field(unsigned long, nf_may)
+		__field(unsigned int, nf_ref)
+	),
+	TP_fast_assign(
+		__entry->nf_inode = nf->nf_inode;
+		__entry->nf_flags = nf->nf_flags;
+		__entry->nf_ref = refcount_read(&nf->nf_ref);
+		__entry->nf_may = nf->nf_may;
+	),
+	TP_printk("inode=%p ref=%u flags=%s may=%s",
+		__entry->nf_inode, __entry->nf_ref,
+		show_nf_flags(__entry->nf_flags),
+		show_nfsd_may_flags(__entry->nf_may)
+	)
+);
+
 TRACE_EVENT(nfsd_file_acquire,
 	TP_PROTO(
 		const struct svc_rqst *rqstp,
Signature · 1 line
2.43.0
author maintainer
#374 lore raw

PATCH 374/690 NFSD: NFSv4 CLOSE should release an nfsd_file immediately

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 5e138c4a750dc140d881dab4a8804b094bbc08d2 ]

The last close of a file should enable other accessors to open and use that file immediately. Leaving the file open in the filecache prevents other users from accessing that file until the filecache garbage-collects the file -- sometimes that takes several seconds.

---
 fs/nfsd/filecache.c | 18 ++++++++++++++++++
 fs/nfsd/filecache.h |  1 +
 fs/nfsd/nfs4state.c |  4 ++--
 3 files changed, 21 insertions(+), 2 deletions(-)
fs/nfsd/filecache.c +18 -0
index 26cfae138b906..7ad27655db699 100644
@@ -451,6 +451,24 @@ nfsd_file_put(struct nfsd_file *nf)
 		nfsd_file_put_noref(nf);
 }
 
+/**
+ * nfsd_file_close - Close an nfsd_file
+ * @nf: nfsd_file to close
+ *
+ * If this is the final reference for @nf, free it immediately.
+ * This reflects an on-the-wire CLOSE or DELEGRETURN into the
+ * VFS and exported filesystem.
+ */
+void nfsd_file_close(struct nfsd_file *nf)
+{
+	nfsd_file_put(nf);
+	if (refcount_dec_if_one(&nf->nf_ref)) {
+		nfsd_file_unhash(nf);
+		nfsd_file_lru_remove(nf);
+		nfsd_file_free(nf);
+	}
+}
+
 struct nfsd_file *
 nfsd_file_get(struct nfsd_file *nf)
 {
fs/nfsd/filecache.h +1 -0
index ee9ed99d8b8fa..28145f1628923 100644
@@ -52,6 +52,7 @@ void nfsd_file_cache_shutdown(void);
 int nfsd_file_cache_start_net(struct net *net);
 void nfsd_file_cache_shutdown_net(struct net *net);
 void nfsd_file_put(struct nfsd_file *nf);
+void nfsd_file_close(struct nfsd_file *nf);
 struct nfsd_file *nfsd_file_get(struct nfsd_file *nf);
 void nfsd_file_close_inode_sync(struct inode *inode);
 bool nfsd_file_is_cached(struct inode *inode);
fs/nfsd/nfs4state.c +2 -2
index d349abf0821d6..923eec2716d75 100644
@@ -831,9 +831,9 @@ static void __nfs4_file_put_access(struct nfs4_file *fp, int oflag)
 			swap(f2, fp->fi_fds[O_RDWR]);
 		spin_unlock(&fp->fi_lock);
 		if (f1)
-			nfsd_file_put(f1);
+			nfsd_file_close(f1);
 		if (f2)
-			nfsd_file_put(f2);
+			nfsd_file_close(f2);
 	}
 }
 
Signature · 1 line
2.43.0
author maintainer
#375 lore raw

PATCH 375/690 NFSD: Ensure nf_inode is never dereferenced

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 427f5f83a3191cbf024c5aea6e5b601cdf88d895 ]

The documenting comment for struct nf_file states:

/*
* A representation of a file that has been opened by knfsd. These are hashed
* in the hashtable by inode pointer value. Note that this object doesn't
* hold a reference to the inode by itself, so the nf_inode pointer should
* never be dereferenced, only used for comparison.
*/

Replace the two existing dereferences to make the comment always true.

---
 fs/nfsd/filecache.c | 5 ++---
 fs/nfsd/filecache.h | 2 +-
 fs/nfsd/nfs4state.c | 2 +-
 3 files changed, 4 insertions(+), 5 deletions(-)
fs/nfsd/filecache.c +2 -3
index 7ad27655db699..55478d411e5a0 100644
@@ -227,12 +227,11 @@ nfsd_file_mark_put(struct nfsd_file_mark *nfm)
 }
 
 static struct nfsd_file_mark *
-nfsd_file_mark_find_or_create(struct nfsd_file *nf)
+nfsd_file_mark_find_or_create(struct nfsd_file *nf, struct inode *inode)
 {
 	int			err;
 	struct fsnotify_mark	*mark;
 	struct nfsd_file_mark	*nfm = NULL, *new;
-	struct inode *inode = nf->nf_inode;
 
 	do {
 		fsnotify_group_lock(nfsd_file_fsnotify_group);
@@ -1143,7 +1142,7 @@ nfsd_file_do_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 open_file:
 	trace_nfsd_file_alloc(nf);
-	nf->nf_mark = nfsd_file_mark_find_or_create(nf);
+	nf->nf_mark = nfsd_file_mark_find_or_create(nf, key.inode);
 	if (nf->nf_mark) {
 		if (open) {
 			status = nfsd_open_verified(rqstp, fhp, may_flags,
fs/nfsd/filecache.h +1 -1
index 28145f1628923..8e8c0c47d67df 100644
@@ -39,7 +39,7 @@ struct nfsd_file {
 #define NFSD_FILE_PENDING	(1)
 #define NFSD_FILE_REFERENCED	(2)
 	unsigned long		nf_flags;
-	struct inode		*nf_inode;
+	struct inode		*nf_inode;	/* don't deref */
 	refcount_t		nf_ref;
 	unsigned char		nf_may;
 	struct nfsd_file_mark	*nf_mark;
fs/nfsd/nfs4state.c +1 -1
index 923eec2716d75..9d344164f814f 100644
@@ -2577,7 +2577,7 @@ static void nfs4_show_fname(struct seq_file *s, struct nfsd_file *f)
 
 static void nfs4_show_superblock(struct seq_file *s, struct nfsd_file *f)
 {
-	struct inode *inode = f->nf_inode;
+	struct inode *inode = file_inode(f->nf_file);
 
 	seq_printf(s, "superblock: \"%02x:%02x:%ld\"",
 					MAJOR(inode->i_sb->s_dev),
Signature · 1 line
2.43.0
author maintainer
#376 lore raw

PATCH 376/690 NFSD: refactoring v4 specific code to a helper in nfs4state.c

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Dai Ngo <[email protected]>

[ Upstream commit 6867137ebcf4155fe25f2ecf7c29b9fb90a76d1d ]

This patch moves the v4 specific code from nfsd_init_net() to nfsd4_init_leases_net() helper in nfs4state.c

---
 fs/nfsd/nfs4state.c | 12 ++++++++++++
 fs/nfsd/nfsctl.c    |  9 +--------
 fs/nfsd/nfsd.h      |  4 ++++
 3 files changed, 17 insertions(+), 8 deletions(-)
fs/nfsd/nfs4state.c +12 -0
index 9d344164f814f..a75f3f7c94d50 100644
@@ -4347,6 +4347,18 @@ nfsd4_init_slabs(void)
 	return -ENOMEM;
 }
 
+void nfsd4_init_leases_net(struct nfsd_net *nn)
+{
+	nn->nfsd4_lease = 90;	/* default lease time */
+	nn->nfsd4_grace = 90;
+	nn->somebody_reclaimed = false;
+	nn->track_reclaim_completes = false;
+	nn->clverifier_counter = prandom_u32();
+	nn->clientid_base = prandom_u32();
+	nn->clientid_counter = nn->clientid_base + 1;
+	nn->s2s_cp_cl_id = nn->clientid_counter++;
+}
+
 static void init_nfs4_replay(struct nfs4_replay *rp)
 {
 	rp->rp_status = nfserr_serverfault;
fs/nfsd/nfsctl.c +1 -8
index 7002edbf26870..164c822ae3ae9 100644
@@ -1484,14 +1484,7 @@ static __net_init int nfsd_init_net(struct net *net)
 	retval = nfsd_reply_cache_init(nn);
 	if (retval)
 		goto out_drc_error;
-	nn->nfsd4_lease = 90;	/* default lease time */
-	nn->nfsd4_grace = 90;
-	nn->somebody_reclaimed = false;
-	nn->track_reclaim_completes = false;
-	nn->clverifier_counter = prandom_u32();
-	nn->clientid_base = prandom_u32();
-	nn->clientid_counter = nn->clientid_base + 1;
-	nn->s2s_cp_cl_id = nn->clientid_counter++;
+	nfsd4_init_leases_net(nn);
 
 	get_random_bytes(&nn->siphash_key, sizeof(nn->siphash_key));
 	seqlock_init(&nn->writeverf_lock);
fs/nfsd/nfsd.h +4 -0
index 9a8b09afc1733..ef8087691138a 100644
@@ -496,12 +496,16 @@ extern void unregister_cld_notifier(void);
 extern void nfsd4_ssc_init_umount_work(struct nfsd_net *nn);
 #endif
 
+extern void nfsd4_init_leases_net(struct nfsd_net *nn);
+
 #else /* CONFIG_NFSD_V4 */
 static inline int nfsd4_is_junction(struct dentry *dentry)
 {
 	return 0;
 }
 
+static inline void nfsd4_init_leases_net(struct nfsd_net *nn) {};
+
 #define register_cld_notifier() 0
 #define unregister_cld_notifier() do { } while(0)
 
Signature · 1 line
2.43.0
author maintainer
#377 lore raw

PATCH 377/690 NFSD: keep track of the number of v4 clients in the system

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Dai Ngo <[email protected]>

[ Upstream commit 0926c39515aa065a296e97dfc8790026f1e53f86 ]

Add counter nfs4_client_count to keep track of the total number of v4 clients, including courtesy clients, in the system.

---
 fs/nfsd/netns.h     |  2 ++
 fs/nfsd/nfs4state.c | 10 ++++++++--
 2 files changed, 10 insertions(+), 2 deletions(-)
fs/nfsd/netns.h +2 -0
index 1b1a962a18041..ce864f001a3ee 100644
@@ -189,6 +189,8 @@ struct nfsd_net {
 	struct nfsd_fcache_disposal *fcache_disposal;
 
 	siphash_key_t		siphash_key;
+
+	atomic_t		nfs4_client_count;
 };
 
 /* Simple check to find out if a given net was properly initialized */
fs/nfsd/nfs4state.c +8 -2
index a75f3f7c94d50..3d5ef021632b9 100644
@@ -2066,7 +2066,8 @@ STALE_CLIENTID(clientid_t *clid, struct nfsd_net *nn)
  * This type of memory management is somewhat inefficient, but we use it
  * anyway since SETCLIENTID is not a common operation.
  */
-static struct nfs4_client *alloc_client(struct xdr_netobj name)
+static struct nfs4_client *alloc_client(struct xdr_netobj name,
+				struct nfsd_net *nn)
 {
 	struct nfs4_client *clp;
 	int i;
@@ -2089,6 +2090,7 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name)
 	atomic_set(&clp->cl_rpc_users, 0);
 	clp->cl_cb_state = NFSD4_CB_UNKNOWN;
 	clp->cl_state = NFSD4_ACTIVE;
+	atomic_inc(&nn->nfs4_client_count);
 	atomic_set(&clp->cl_delegs_in_recall, 0);
 	INIT_LIST_HEAD(&clp->cl_idhash);
 	INIT_LIST_HEAD(&clp->cl_openowners);
@@ -2196,6 +2198,7 @@ static __be32 mark_client_expired_locked(struct nfs4_client *clp)
 static void
 __destroy_client(struct nfs4_client *clp)
 {
+	struct nfsd_net *nn = net_generic(clp->net, nfsd_net_id);
 	int i;
 	struct nfs4_openowner *oo;
 	struct nfs4_delegation *dp;
@@ -2239,6 +2242,7 @@ __destroy_client(struct nfs4_client *clp)
 	nfsd4_shutdown_callback(clp);
 	if (clp->cl_cb_conn.cb_xprt)
 		svc_xprt_put(clp->cl_cb_conn.cb_xprt);
+	atomic_add_unless(&nn->nfs4_client_count, -1, 0);
 	free_client(clp);
 	wake_up_all(&expiry_wq);
 }
@@ -2865,7 +2869,7 @@ static struct nfs4_client *create_client(struct xdr_netobj name,
 	struct nfsd_net *nn = net_generic(net, nfsd_net_id);
 	struct dentry *dentries[ARRAY_SIZE(client_files)];
 
-	clp = alloc_client(name);
+	clp = alloc_client(name, nn);
 	if (clp == NULL)
 		return NULL;
 
@@ -4357,6 +4361,8 @@ void nfsd4_init_leases_net(struct nfsd_net *nn)
 	nn->clientid_base = prandom_u32();
 	nn->clientid_counter = nn->clientid_base + 1;
 	nn->s2s_cp_cl_id = nn->clientid_counter++;
+
+	atomic_set(&nn->nfs4_client_count, 0);
 }
 
 static void init_nfs4_replay(struct nfs4_replay *rp)
Signature · 1 line
2.43.0
author maintainer
#378 lore raw

PATCH 378/690 NFSD: limit the number of v4 clients to 1024 per 1GB of system memory

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Dai Ngo <[email protected]>

[ Upstream commit 4271c2c0887562318a0afef97d32d8a71cbe0743 ]

Currently there is no limit on how many v4 clients are supported by the system. This can be a problem in systems with small memory configuration to function properly when a very large number of clients exist that creates memory shortage conditions.

This patch enforces a limit of 1024 NFSv4 clients, including courtesy clients, per 1GB of system memory. When the number of the clients reaches the limit, requests that create new clients are returned with NFS4ERR_DELAY and the laundromat is kicked start to trim old clients. Due to the overhead of the upcall to remove the client record, the maximun number of clients the laundromat removes on each run is limited to 128. This is done to ensure the laundromat can still process the other tasks in a timely manner.

Since there is now a limit of the number of clients, the 24-hr idle time limit of courtesy client is no longer needed and was removed.

---
 fs/nfsd/netns.h     |  1 +
 fs/nfsd/nfs4state.c | 27 +++++++++++++++++++++------
 fs/nfsd/nfsd.h      |  2 ++
 3 files changed, 24 insertions(+), 6 deletions(-)
fs/nfsd/netns.h +1 -0
index ce864f001a3ee..ffe17743cc74b 100644
@@ -191,6 +191,7 @@ struct nfsd_net {
 	siphash_key_t		siphash_key;
 
 	atomic_t		nfs4_client_count;
+	int			nfs4_max_clients;
 };
 
 /* Simple check to find out if a given net was properly initialized */
fs/nfsd/nfs4state.c +21 -6
index 3d5ef021632b9..340d533dcafd3 100644
@@ -2072,6 +2072,10 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name,
 	struct nfs4_client *clp;
 	int i;
 
+	if (atomic_read(&nn->nfs4_client_count) >= nn->nfs4_max_clients) {
+		mod_delayed_work(laundry_wq, &nn->laundromat_work, 0);
+		return NULL;
+	}
 	clp = kmem_cache_zalloc(client_slab, GFP_KERNEL);
 	if (clp == NULL)
 		return NULL;
@@ -4353,6 +4357,9 @@ nfsd4_init_slabs(void)
 
 void nfsd4_init_leases_net(struct nfsd_net *nn)
 {
+	struct sysinfo si;
+	u64 max_clients;
+
 	nn->nfsd4_lease = 90;	/* default lease time */
 	nn->nfsd4_grace = 90;
 	nn->somebody_reclaimed = false;
@@ -4363,6 +4370,10 @@ void nfsd4_init_leases_net(struct nfsd_net *nn)
 	nn->s2s_cp_cl_id = nn->clientid_counter++;
 
 	atomic_set(&nn->nfs4_client_count, 0);
+	si_meminfo(&si);
+	max_clients = (u64)si.totalram * si.mem_unit / (1024 * 1024 * 1024);
+	max_clients *= NFS4_CLIENTS_PER_GB;
+	nn->nfs4_max_clients = max_t(int, max_clients, NFS4_CLIENTS_PER_GB);
 }
 
 static void init_nfs4_replay(struct nfs4_replay *rp)
@@ -5828,9 +5839,12 @@ static void
 nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist,
 				struct laundry_time *lt)
 {
+	unsigned int maxreap, reapcnt = 0;
 	struct list_head *pos, *next;
 	struct nfs4_client *clp;
 
+	maxreap = (atomic_read(&nn->nfs4_client_count) >= nn->nfs4_max_clients) ?
+			NFSD_CLIENT_MAX_TRIM_PER_RUN : 0;
 	INIT_LIST_HEAD(reaplist);
 	spin_lock(&nn->client_lock);
 	list_for_each_safe(pos, next, &nn->client_lru) {
@@ -5841,14 +5855,15 @@ nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist,
 			break;
 		if (!atomic_read(&clp->cl_rpc_users))
 			clp->cl_state = NFSD4_COURTESY;
-		if (!client_has_state(clp) ||
-				ktime_get_boottime_seconds() >=
-				(clp->cl_time + NFSD_COURTESY_CLIENT_TIMEOUT))
+		if (!client_has_state(clp))
 			goto exp_client;
-		if (nfs4_anylock_blockers(clp)) {
+		if (!nfs4_anylock_blockers(clp))
+			if (reapcnt >= maxreap)
+				continue;
 exp_client:
-			if (!mark_client_expired_locked(clp))
-				list_add(&clp->cl_lru, reaplist);
+		if (!mark_client_expired_locked(clp)) {
+			list_add(&clp->cl_lru, reaplist);
+			reapcnt++;
 		}
 	}
 	spin_unlock(&nn->client_lock);
fs/nfsd/nfsd.h +2 -0
index ef8087691138a..57a468ed85c35 100644
@@ -341,6 +341,8 @@ void		nfsd_lockd_shutdown(void);
 
 #define NFSD_LAUNDROMAT_MINTIMEOUT      1   /* seconds */
 #define	NFSD_COURTESY_CLIENT_TIMEOUT	(24 * 60 * 60)	/* seconds */
+#define	NFSD_CLIENT_MAX_TRIM_PER_RUN	128
+#define	NFS4_CLIENTS_PER_GB		1024
 
 /*
  * The following attributes are currently not supported by the NFSv4 server:
Signature · 1 line
2.43.0
author maintainer
#379 lore raw

PATCH 379/690 nfsd: silence extraneous printk on nfsd.ko insertion

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jeff Layton <[email protected]>

[ Upstream commit 3a5940bfa17fb9964bf9688b4356ca643a8f5e2d ]

This printk pops every time nfsd.ko gets plugged in. Most kmods don't do that and this one is not very informative. Olaf's email address seems to be defunct at this point anyway. Just drop it.

---
 fs/nfsd/nfsctl.c | 1 -
 1 file changed, 1 deletion(-)
fs/nfsd/nfsctl.c +0 -1
index 164c822ae3ae9..917fa1892fd2d 100644
@@ -1519,7 +1519,6 @@ static struct pernet_operations nfsd_net_ops = {
 static int __init init_nfsd(void)
 {
 	int retval;
-	printk(KERN_INFO "Installing knfsd (copyright (C) 1996 [email protected]).\n");
 
 	retval = nfsd4_init_slabs();
 	if (retval)
Signature · 1 line
2.43.0
author maintainer
#380 lore raw

PATCH 380/690 NFSD: Optimize nfsd4_encode_operation()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 095a764b7afb06c9499b798c04eaa3cbf70ebe2d ]

write_bytes_to_xdr_buf() is a generic way to place a variable-length data item in an already-reserved spot in the encoding buffer. However, it is costly, and here, it is unnecessary because the data item is fixed in size, the buffer destination address is always word-aligned, and the destination location is already in @p.

---
 fs/nfsd/nfs4xdr.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/nfsd/nfs4xdr.c +1 -2
index 07f891d7fa0ae..9eba9afe13794 100644
@@ -5378,8 +5378,7 @@ nfsd4_encode_operation(struct nfsd4_compoundres *resp, struct nfsd4_op *op)
 						so->so_replay.rp_buf, len);
 	}
 status:
-	/* Note that op->status is already in network byte order: */
-	write_bytes_to_xdr_buf(xdr->buf, post_err_offset - 4, &op->status, 4);
+	*p = op->status;
 }
 
 /* 
Signature · 1 line
2.43.0
author maintainer
#381 lore raw

PATCH 381/690 NFSD: Optimize nfsd4_encode_fattr()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit ab04de60ae1cc64ae16b77feae795311b97720c7 ]

write_bytes_to_xdr_buf() is a generic way to place a variable-length data item in an already-reserved spot in the encoding buffer.

However, it is costly. In nfsd4_encode_fattr(), it is unnecessary because the data item is fixed in size and the buffer destination address is always word-aligned.

---
 fs/nfsd/nfs4xdr.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)
fs/nfsd/nfs4xdr.c +4 -7
index 9eba9afe13794..d5a4aa0da32be 100644
@@ -2825,10 +2825,9 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
 	struct kstat stat;
 	struct svc_fh *tempfh = NULL;
 	struct kstatfs statfs;
-	__be32 *p;
+	__be32 *p, *attrlen_p;
 	int starting_len = xdr->buf->len;
 	int attrlen_offset;
-	__be32 attrlen;
 	u32 dummy;
 	u64 dummy64;
 	u32 rdattr_err = 0;
@@ -2916,10 +2915,9 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
 		goto out;
 
 	attrlen_offset = xdr->buf->len;
-	p = xdr_reserve_space(xdr, 4);
-	if (!p)
+	attrlen_p = xdr_reserve_space(xdr, XDR_UNIT);
+	if (!attrlen_p)
 		goto out_resource;
-	p++;                /* to be backfilled later */
 
 	if (bmval0 & FATTR4_WORD0_SUPPORTED_ATTRS) {
 		u32 supp[3];
@@ -3341,8 +3339,7 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
 		*p++ = cpu_to_be32(err == 0);
 	}
 
-	attrlen = htonl(xdr->buf->len - attrlen_offset - 4);
-	write_bytes_to_xdr_buf(xdr->buf, attrlen_offset, &attrlen, 4);
+	*attrlen_p = cpu_to_be32(xdr->buf->len - attrlen_offset - XDR_UNIT);
 	status = nfs_ok;
 
 out:
Signature · 1 line
2.43.0
author maintainer
#382 lore raw

PATCH 382/690 NFSD: Clean up SPLICE_OK in nfsd4_encode_read()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit c738b218a2e5a753a336b4b7fee6720b902c7ace ]

Do the test_bit() once -- this reduces the number of locked-bus operations and makes the function a little easier to read.

---
 fs/nfsd/nfs4xdr.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)
fs/nfsd/nfs4xdr.c +4 -5
index d5a4aa0da32be..afc8a51cf60f1 100644
@@ -3988,6 +3988,7 @@ static __be32
 nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr,
 		  struct nfsd4_read *read)
 {
+	bool splice_ok = test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags);
 	unsigned long maxcount;
 	struct xdr_stream *xdr = resp->xdr;
 	struct file *file;
@@ -4000,11 +4001,10 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr,
 
 	p = xdr_reserve_space(xdr, 8); /* eof flag and byte count */
 	if (!p) {
-		WARN_ON_ONCE(test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags));
+		WARN_ON_ONCE(splice_ok);
 		return nfserr_resource;
 	}
-	if (resp->xdr->buf->page_len &&
-	    test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags)) {
+	if (resp->xdr->buf->page_len && splice_ok) {
 		WARN_ON_ONCE(1);
 		return nfserr_serverfault;
 	}
@@ -4013,8 +4013,7 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr,
 	maxcount = min_t(unsigned long, read->rd_length,
 			 (xdr->buf->buflen - xdr->buf->len));
 
-	if (file->f_op->splice_read &&
-	    test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags))
+	if (file->f_op->splice_read && splice_ok)
 		nfserr = nfsd4_encode_splice_read(resp, read, file, maxcount);
 	else
 		nfserr = nfsd4_encode_readv(resp, read, file, maxcount);
Signature · 1 line
2.43.0
author maintainer
#383 lore raw

PATCH 383/690 NFSD: Add an nfsd4_read::rd_eof field

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 24c7fb85498eda1d4c6b42cc4886328429814990 ]

Refactor: Make the EOF result available in the entire NFSv4 READ path.

---
 fs/nfsd/nfs4xdr.c | 11 +++++------
 fs/nfsd/xdr4.h    |  5 +++--
 2 files changed, 8 insertions(+), 8 deletions(-)
fs/nfsd/nfs4xdr.c +5 -6
index afc8a51cf60f1..b31103221fee9 100644
@@ -3887,7 +3887,6 @@ static __be32 nfsd4_encode_splice_read(
 	struct xdr_stream *xdr = resp->xdr;
 	struct xdr_buf *buf = xdr->buf;
 	int status, space_left;
-	u32 eof;
 	__be32 nfserr;
 	__be32 *p = xdr->p - 2;
 
@@ -3896,7 +3895,8 @@ static __be32 nfsd4_encode_splice_read(
 		return nfserr_resource;
 
 	nfserr = nfsd_splice_read(read->rd_rqstp, read->rd_fhp,
-				  file, read->rd_offset, &maxcount, &eof);
+				  file, read->rd_offset, &maxcount,
+				  &read->rd_eof);
 	read->rd_length = maxcount;
 	if (nfserr)
 		goto out_err;
@@ -3907,7 +3907,7 @@ static __be32 nfsd4_encode_splice_read(
 		goto out_err;
 	}
 
-	*(p++) = htonl(eof);
+	*(p++) = htonl(read->rd_eof);
 	*(p++) = htonl(maxcount);
 
 	buf->page_len = maxcount;
@@ -3951,7 +3951,6 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp,
 				 struct file *file, unsigned long maxcount)
 {
 	struct xdr_stream *xdr = resp->xdr;
-	u32 eof;
 	int starting_len = xdr->buf->len - 8;
 	__be32 nfserr;
 	__be32 tmp;
@@ -3963,7 +3962,7 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp,
 
 	nfserr = nfsd_readv(resp->rqstp, read->rd_fhp, file, read->rd_offset,
 			    resp->rqstp->rq_vec, read->rd_vlen, &maxcount,
-			    &eof);
+			    &read->rd_eof);
 	read->rd_length = maxcount;
 	if (nfserr)
 		return nfserr;
@@ -3971,7 +3970,7 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp,
 		return nfserr_io;
 	xdr_truncate_encode(xdr, starting_len + 8 + xdr_align_size(maxcount));
 
-	tmp = htonl(eof);
+	tmp = htonl(read->rd_eof);
 	write_bytes_to_xdr_buf(xdr->buf, starting_len    , &tmp, 4);
 	tmp = htonl(maxcount);
 	write_bytes_to_xdr_buf(xdr->buf, starting_len + 4, &tmp, 4);
fs/nfsd/xdr4.h +3 -2
index 77286e8c9ab02..0737f81c1004e 100644
@@ -302,9 +302,10 @@ struct nfsd4_read {
 	u32			rd_length;          /* request */
 	int			rd_vlen;
 	struct nfsd_file	*rd_nf;
-	
+
 	struct svc_rqst		*rd_rqstp;          /* response */
-	struct svc_fh		*rd_fhp;             /* response */
+	struct svc_fh		*rd_fhp;            /* response */
+	u32			rd_eof;             /* response */
 };
 
 struct nfsd4_readdir {
Signature · 1 line
2.43.0
author maintainer
#384 lore raw

PATCH 384/690 NFSD: Optimize nfsd4_encode_readv()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 28d5bc468efe74b790e052f758ce083a5015c665 ]

write_bytes_to_xdr_buf() is pretty expensive to use for inserting an XDR data item that is always 1 XDR_UNIT at an address that is always XDR word-aligned.

Since both the readv and splice read paths encode EOF and maxcount values, move both to a common code path.

---
 fs/nfsd/nfs4xdr.c | 18 ++++++------------
 1 file changed, 6 insertions(+), 12 deletions(-)
fs/nfsd/nfs4xdr.c +6 -12
index b31103221fee9..b7a3c770d436b 100644
@@ -3888,7 +3888,6 @@ static __be32 nfsd4_encode_splice_read(
 	struct xdr_buf *buf = xdr->buf;
 	int status, space_left;
 	__be32 nfserr;
-	__be32 *p = xdr->p - 2;
 
 	/* Make sure there will be room for padding if needed */
 	if (xdr->end - xdr->p < 1)
@@ -3907,9 +3906,6 @@ static __be32 nfsd4_encode_splice_read(
 		goto out_err;
 	}
 
-	*(p++) = htonl(read->rd_eof);
-	*(p++) = htonl(maxcount);
-
 	buf->page_len = maxcount;
 	buf->len += maxcount;
 	xdr->page_ptr += (buf->page_base + maxcount + PAGE_SIZE - 1)
@@ -3970,11 +3966,6 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp,
 		return nfserr_io;
 	xdr_truncate_encode(xdr, starting_len + 8 + xdr_align_size(maxcount));
 
-	tmp = htonl(read->rd_eof);
-	write_bytes_to_xdr_buf(xdr->buf, starting_len    , &tmp, 4);
-	tmp = htonl(maxcount);
-	write_bytes_to_xdr_buf(xdr->buf, starting_len + 4, &tmp, 4);
-
 	tmp = xdr_zero;
 	pad = (maxcount&3) ? 4 - (maxcount&3) : 0;
 	write_bytes_to_xdr_buf(xdr->buf, starting_len + 8 + maxcount,
@@ -4016,11 +4007,14 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr,
 		nfserr = nfsd4_encode_splice_read(resp, read, file, maxcount);
 	else
 		nfserr = nfsd4_encode_readv(resp, read, file, maxcount);
-
-	if (nfserr)
+	if (nfserr) {
 		xdr_truncate_encode(xdr, starting_len);
+		return nfserr;
+	}
 
-	return nfserr;
+	p = xdr_encode_bool(p, read->rd_eof);
+	*p = cpu_to_be32(read->rd_length);
+	return nfs_ok;
 }
 
 static __be32
Signature · 1 line
2.43.0
author maintainer
#385 lore raw

PATCH 385/690 NFSD: Simplify starting_len

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 071ae99feadfc55979f89287d6ad2c6a315cb46d ]

Clean-up: Now that nfsd4_encode_readv() does not have to encode the EOF or rd_length values, it no longer needs to subtract 8 from @starting_len.

---
 fs/nfsd/nfs4xdr.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)
fs/nfsd/nfs4xdr.c +4 -5
index b7a3c770d436b..310321b9b94cd 100644
@@ -3947,7 +3947,7 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp,
 				 struct file *file, unsigned long maxcount)
 {
 	struct xdr_stream *xdr = resp->xdr;
-	int starting_len = xdr->buf->len - 8;
+	unsigned int starting_len = xdr->buf->len;
 	__be32 nfserr;
 	__be32 tmp;
 	int pad;
@@ -3962,14 +3962,13 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp,
 	read->rd_length = maxcount;
 	if (nfserr)
 		return nfserr;
-	if (svc_encode_result_payload(resp->rqstp, starting_len + 8, maxcount))
+	if (svc_encode_result_payload(resp->rqstp, starting_len, maxcount))
 		return nfserr_io;
-	xdr_truncate_encode(xdr, starting_len + 8 + xdr_align_size(maxcount));
+	xdr_truncate_encode(xdr, starting_len + xdr_align_size(maxcount));
 
 	tmp = xdr_zero;
 	pad = (maxcount&3) ? 4 - (maxcount&3) : 0;
-	write_bytes_to_xdr_buf(xdr->buf, starting_len + 8 + maxcount,
-								&tmp, pad);
+	write_bytes_to_xdr_buf(xdr->buf, starting_len + maxcount, &tmp, pad);
 	return 0;
 
 }
Signature · 1 line
2.43.0
author maintainer
#386 lore raw

PATCH 386/690 NFSD: Use xdr_pad_size()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 5e64d85c7d0c59cfcd61d899720b8ccfe895d743 ]

Clean up: Use a helper instead of open-coding the calculation of the XDR pad size.

---
 fs/nfsd/nfs4xdr.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)
fs/nfsd/nfs4xdr.c +4 -7
index 310321b9b94cd..88e8192f9a75d 100644
@@ -3948,9 +3948,8 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp,
 {
 	struct xdr_stream *xdr = resp->xdr;
 	unsigned int starting_len = xdr->buf->len;
+	__be32 zero = xdr_zero;
 	__be32 nfserr;
-	__be32 tmp;
-	int pad;
 
 	read->rd_vlen = xdr_reserve_space_vec(xdr, resp->rqstp->rq_vec, maxcount);
 	if (read->rd_vlen < 0)
@@ -3966,11 +3965,9 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp,
 		return nfserr_io;
 	xdr_truncate_encode(xdr, starting_len + xdr_align_size(maxcount));
 
-	tmp = xdr_zero;
-	pad = (maxcount&3) ? 4 - (maxcount&3) : 0;
-	write_bytes_to_xdr_buf(xdr->buf, starting_len + maxcount, &tmp, pad);
-	return 0;
-
+	write_bytes_to_xdr_buf(xdr->buf, starting_len + maxcount, &zero,
+			       xdr_pad_size(maxcount));
+	return nfs_ok;
 }
 
 static __be32
Signature · 1 line
2.43.0
author maintainer
#387 lore raw

PATCH 387/690 NFSD: Clean up nfsd4_encode_readlink()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 99b002a1fa00d90e66357315757e7277447ce973 ]

Similar changes to nfsd4_encode_readv(), all bundled into a single patch.

---
 fs/nfsd/nfs4xdr.c | 24 +++++++++---------------
 1 file changed, 9 insertions(+), 15 deletions(-)
fs/nfsd/nfs4xdr.c +9 -15
index 88e8192f9a75d..a5ab6ea475423 100644
@@ -4016,16 +4016,13 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr,
 static __be32
 nfsd4_encode_readlink(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_readlink *readlink)
 {
-	int maxcount;
-	__be32 wire_count;
-	int zero = 0;
+	__be32 *p, *maxcount_p, zero = xdr_zero;
 	struct xdr_stream *xdr = resp->xdr;
 	int length_offset = xdr->buf->len;
-	int status;
-	__be32 *p;
+	int maxcount, status;
 
-	p = xdr_reserve_space(xdr, 4);
-	if (!p)
+	maxcount_p = xdr_reserve_space(xdr, XDR_UNIT);
+	if (!maxcount_p)
 		return nfserr_resource;
 	maxcount = PAGE_SIZE;
 
@@ -4050,14 +4047,11 @@ nfsd4_encode_readlink(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd
 		nfserr = nfserrno(status);
 		goto out_err;
 	}
-
-	wire_count = htonl(maxcount);
-	write_bytes_to_xdr_buf(xdr->buf, length_offset, &wire_count, 4);
-	xdr_truncate_encode(xdr, length_offset + 4 + ALIGN(maxcount, 4));
-	if (maxcount & 3)
-		write_bytes_to_xdr_buf(xdr->buf, length_offset + 4 + maxcount,
-						&zero, 4 - (maxcount&3));
-	return 0;
+	*maxcount_p = cpu_to_be32(maxcount);
+	xdr_truncate_encode(xdr, length_offset + 4 + xdr_align_size(maxcount));
+	write_bytes_to_xdr_buf(xdr->buf, length_offset + 4 + maxcount, &zero,
+			       xdr_pad_size(maxcount));
+	return nfs_ok;
 
 out_err:
 	xdr_truncate_encode(xdr, length_offset);
Signature · 1 line
2.43.0
author maintainer
#388 lore raw

PATCH 388/690 NFSD: Fix strncpy() fortify warning

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 5304877936c0a67e1a01464d113bae4c81eacdb6 ]

In function ‘strncpy’,
inlined from ‘nfsd4_ssc_setup_dul’ at /home/cel/src/linux/manet/fs/nfsd/nfs4proc.c:1392:3,
inlined from ‘nfsd4_interssc_connect’ at /home/cel/src/linux/manet/fs/nfsd/nfs4proc.c:1489:11:
/home/cel/src/linux/manet/include/linux/fortify-string.h:52:33: warning: ‘__builtin_strncpy’ specified bound 63 equals destination size [-Wstringop-truncation]

   52 | #define __underlying_strncpy    __builtin_strncpy
      |                                 ^
/home/cel/src/linux/manet/include/linux/fortify-string.h:89:16: note: in expansion of macro ‘__underlying_strncpy’
   89 |         return __underlying_strncpy(p, q, size);
      |                ^~~~~~~~~~~~~~~~~~~~
---
 fs/nfsd/nfs4proc.c      | 2 +-
 include/linux/nfs_ssc.h | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
fs/nfsd/nfs4proc.c +1 -1
index 15991eb9b8d8c..5dce18fe99085 100644
@@ -1392,7 +1392,7 @@ static __be32 nfsd4_ssc_setup_dul(struct nfsd_net *nn, char *ipaddr,
 		return 0;
 	}
 	if (work) {
-		strncpy(work->nsui_ipaddr, ipaddr, sizeof(work->nsui_ipaddr));
+		strlcpy(work->nsui_ipaddr, ipaddr, sizeof(work->nsui_ipaddr) - 1);
 		refcount_set(&work->nsui_refcnt, 2);
 		work->nsui_busy = true;
 		list_add_tail(&work->nsui_list, &nn->nfsd_ssc_mount_list);
include/linux/nfs_ssc.h +1 -1
index 222ae8883e854..75843c00f326a 100644
@@ -64,7 +64,7 @@ struct nfsd4_ssc_umount_item {
 	refcount_t nsui_refcnt;
 	unsigned long nsui_expire;
 	struct vfsmount *nsui_vfsmount;
-	char nsui_ipaddr[RPC_MAX_ADDRBUFLEN];
+	char nsui_ipaddr[RPC_MAX_ADDRBUFLEN + 1];
 };
 #endif
 
Signature · 1 line
2.43.0
author maintainer
#389 lore raw

PATCH 389/690 NFSD: nfserrno(-ENOMEM) is nfserr_jukebox

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit bb4d842722b84a2731257054b6405f2d866fc5f3 ]

---
 fs/nfsd/nfs4xdr.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
fs/nfsd/nfs4xdr.c +2 -2
index a5ab6ea475423..fb891249694c3 100644
@@ -1810,7 +1810,7 @@ nfsd4_decode_test_stateid(struct nfsd4_compoundargs *argp, struct nfsd4_test_sta
 	for (i = 0; i < test_stateid->ts_num_ids; i++) {
 		stateid = svcxdr_tmpalloc(argp, sizeof(*stateid));
 		if (!stateid)
-			return nfserrno(-ENOMEM);	/* XXX: not jukebox? */
+			return nfserr_jukebox;
 		INIT_LIST_HEAD(&stateid->ts_id_list);
 		list_add_tail(&stateid->ts_id_list, &test_stateid->ts_stateid_list);
 		status = nfsd4_decode_stateid4(argp, &stateid->ts_id_stateid);
@@ -1933,7 +1933,7 @@ nfsd4_decode_copy(struct nfsd4_compoundargs *argp, struct nfsd4_copy *copy)
 
 	ns_dummy = kmalloc(sizeof(struct nl4_server), GFP_KERNEL);
 	if (ns_dummy == NULL)
-		return nfserrno(-ENOMEM);	/* XXX: jukebox? */
+		return nfserr_jukebox;
 	for (i = 0; i < count - 1; i++) {
 		status = nfsd4_decode_nl4_server(argp, ns_dummy);
 		if (status) {
Signature · 1 line
2.43.0
author maintainer
#390 lore raw

PATCH 390/690 NFSD: Shrink size of struct nfsd4_copy_notify

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 09426ef2a64ee189ca1e3298f1e874842dbf35ea ]

struct nfsd4_copy_notify is part of struct nfsd4_op, which resides in an 8-element array.

sizeof(struct nfsd4_op):

Before: /* size: 2208, cachelines: 35, members: 5 */
After:  /* size: 1696, cachelines: 27, members: 5 */
---
 fs/nfsd/nfs4proc.c |  4 ++--
 fs/nfsd/nfs4xdr.c  | 12 ++++++++++--
 fs/nfsd/xdr4.h     |  4 ++--
 3 files changed, 14 insertions(+), 6 deletions(-)
fs/nfsd/nfs4proc.c +2 -2
index 5dce18fe99085..80ec51a89d5b5 100644
@@ -1947,9 +1947,9 @@ nfsd4_copy_notify(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	/* For now, only return one server address in cpn_src, the
 	 * address used by the client to connect to this server.
 	 */
-	cn->cpn_src.nl4_type = NL4_NETADDR;
+	cn->cpn_src->nl4_type = NL4_NETADDR;
 	status = nfsd4_set_netaddr((struct sockaddr *)&rqstp->rq_daddr,
-				 &cn->cpn_src.u.nl4_addr);
+				 &cn->cpn_src->u.nl4_addr);
 	WARN_ON_ONCE(status);
 	if (status) {
 		nfs4_put_cpntf_state(nn, cps);
fs/nfsd/nfs4xdr.c +10 -2
index fb891249694c3..515edc1b662e1 100644
@@ -1952,10 +1952,17 @@ nfsd4_decode_copy_notify(struct nfsd4_compoundargs *argp,
 {
 	__be32 status;
 
+	cn->cpn_src = svcxdr_tmpalloc(argp, sizeof(*cn->cpn_src));
+	if (cn->cpn_src == NULL)
+		return nfserr_jukebox;
+	cn->cpn_dst = svcxdr_tmpalloc(argp, sizeof(*cn->cpn_dst));
+	if (cn->cpn_dst == NULL)
+		return nfserr_jukebox;
+
 	status = nfsd4_decode_stateid4(argp, &cn->cpn_src_stateid);
 	if (status)
 		return status;
-	return nfsd4_decode_nl4_server(argp, &cn->cpn_dst);
+	return nfsd4_decode_nl4_server(argp, cn->cpn_dst);
 }
 
 static __be32
@@ -4903,7 +4910,8 @@ nfsd4_encode_copy_notify(struct nfsd4_compoundres *resp, __be32 nfserr,
 
 	*p++ = cpu_to_be32(1);
 
-	return nfsd42_encode_nl4_server(resp, &cn->cpn_src);
+	nfserr = nfsd42_encode_nl4_server(resp, cn->cpn_src);
+	return nfserr;
 }
 
 static __be32
fs/nfsd/xdr4.h +2 -2
index 0737f81c1004e..6673c2980c77e 100644
@@ -595,13 +595,13 @@ struct nfsd4_offload_status {
 struct nfsd4_copy_notify {
 	/* request */
 	stateid_t		cpn_src_stateid;
-	struct nl4_server	cpn_dst;
+	struct nl4_server	*cpn_dst;
 
 	/* response */
 	stateid_t		cpn_cnr_stateid;
 	u64			cpn_sec;
 	u32			cpn_nsec;
-	struct nl4_server	cpn_src;
+	struct nl4_server	*cpn_src;
 };
 
 struct nfsd4_op {
Signature · 1 line
2.43.0
author maintainer
#391 lore raw

PATCH 391/690 NFSD: Shrink size of struct nfsd4_copy

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 87689df694916c40e8e6c179ab1c8710f65cb6c6 ]

struct nfsd4_copy is part of struct nfsd4_op, which resides in an 8-element array.

sizeof(struct nfsd4_op):

Before: /* size: 1696, cachelines: 27, members: 5 */
After:  /* size: 672, cachelines: 11, members: 5 */
---
 fs/nfsd/nfs4proc.c | 8 ++++++--
 fs/nfsd/nfs4xdr.c  | 5 ++++-
 fs/nfsd/xdr4.h     | 2 +-
 3 files changed, 11 insertions(+), 4 deletions(-)
fs/nfsd/nfs4proc.c +6 -2
index 80ec51a89d5b5..f0722d4ed0810 100644
@@ -1291,6 +1291,7 @@ void nfs4_put_copy(struct nfsd4_copy *copy)
 {
 	if (!refcount_dec_and_test(&copy->refcount))
 		return;
+	kfree(copy->cp_src);
 	kfree(copy);
 }
 
@@ -1545,7 +1546,7 @@ nfsd4_setup_inter_ssc(struct svc_rqst *rqstp,
 	if (status)
 		goto out;
 
-	status = nfsd4_interssc_connect(&copy->cp_src, rqstp, mount);
+	status = nfsd4_interssc_connect(copy->cp_src, rqstp, mount);
 	if (status)
 		goto out;
 
@@ -1753,7 +1754,7 @@ static void dup_copy_fields(struct nfsd4_copy *src, struct nfsd4_copy *dst)
 		dst->nf_src = nfsd_file_get(src->nf_src);
 
 	memcpy(&dst->cp_stateid, &src->cp_stateid, sizeof(src->cp_stateid));
-	memcpy(&dst->cp_src, &src->cp_src, sizeof(struct nl4_server));
+	memcpy(dst->cp_src, src->cp_src, sizeof(struct nl4_server));
 	memcpy(&dst->stateid, &src->stateid, sizeof(src->stateid));
 	memcpy(&dst->c_fh, &src->c_fh, sizeof(src->c_fh));
 	dst->ss_mnt = src->ss_mnt;
@@ -1847,6 +1848,9 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		async_copy = kzalloc(sizeof(struct nfsd4_copy), GFP_KERNEL);
 		if (!async_copy)
 			goto out_err;
+		async_copy->cp_src = kmalloc(sizeof(*async_copy->cp_src), GFP_KERNEL);
+		if (!async_copy->cp_src)
+			goto out_err;
 		if (!nfs4_init_copy_state(nn, copy))
 			goto out_err;
 		refcount_set(&async_copy->refcount, 1);
fs/nfsd/nfs4xdr.c +4 -1
index 515edc1b662e1..dad744d44b0a2 100644
@@ -1920,6 +1920,9 @@ nfsd4_decode_copy(struct nfsd4_compoundargs *argp, struct nfsd4_copy *copy)
 
 	if (xdr_stream_decode_u32(argp->xdr, &count) < 0)
 		return nfserr_bad_xdr;
+	copy->cp_src = svcxdr_tmpalloc(argp, sizeof(*copy->cp_src));
+	if (copy->cp_src == NULL)
+		return nfserr_jukebox;
 	copy->cp_intra = false;
 	if (count == 0) { /* intra-server copy */
 		copy->cp_intra = true;
@@ -1927,7 +1930,7 @@ nfsd4_decode_copy(struct nfsd4_compoundargs *argp, struct nfsd4_copy *copy)
 	}
 
 	/* decode all the supplied server addresses but use only the first */
-	status = nfsd4_decode_nl4_server(argp, &copy->cp_src);
+	status = nfsd4_decode_nl4_server(argp, copy->cp_src);
 	if (status)
 		return status;
 
fs/nfsd/xdr4.h +1 -1
index 6673c2980c77e..ed90843a55293 100644
@@ -540,7 +540,7 @@ struct nfsd4_copy {
 	u64			cp_src_pos;
 	u64			cp_dst_pos;
 	u64			cp_count;
-	struct nl4_server	cp_src;
+	struct nl4_server	*cp_src;
 	bool			cp_intra;
 
 	/* both */
Signature · 1 line
2.43.0
author maintainer
#392 lore raw

PATCH 392/690 NFSD: Reorder the fields in struct nfsd4_op

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit d314309425ad5dc1b6facdb2d456580fb5fa5e3a ]

Pack the fields to reduce the size of struct nfsd4_op, which is used an array in struct nfsd4_compoundargs.

sizeof(struct nfsd4_op):

Before: /* size: 672, cachelines: 11, members: 5 */
After:  /* size: 640, cachelines: 10, members: 5 */
---
 fs/nfsd/xdr4.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
fs/nfsd/xdr4.h +2 -2
index ed90843a55293..87b4270af59ef 100644
@@ -606,8 +606,9 @@ struct nfsd4_copy_notify {
 
 struct nfsd4_op {
 	u32					opnum;
-	const struct nfsd4_operation *		opdesc;
 	__be32					status;
+	const struct nfsd4_operation		*opdesc;
+	struct nfs4_replay			*replay;
 	union nfsd4_op_u {
 		struct nfsd4_access		access;
 		struct nfsd4_close		close;
@@ -671,7 +672,6 @@ struct nfsd4_op {
 		struct nfsd4_listxattrs		listxattrs;
 		struct nfsd4_removexattr	removexattr;
 	} u;
-	struct nfs4_replay *			replay;
 };
 
 bool nfsd4_cache_this_op(struct nfsd4_op *);
Signature · 1 line
2.43.0
author maintainer
#393 lore raw

PATCH 393/690 NFSD: Make nfs4_put_copy() static

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 8ea6e2c90bb0eb74a595a12e23a1dff9abbc760a ]

Clean up: All call sites are in fs/nfsd/nfs4proc.c.

---
 fs/nfsd/nfs4proc.c | 2 +-
 fs/nfsd/state.h    | 1 -
 2 files changed, 1 insertion(+), 2 deletions(-)
fs/nfsd/nfs4proc.c +1 -1
index f0722d4ed0810..3e4b0fb44c7b7 100644
@@ -1287,7 +1287,7 @@ nfsd4_clone(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	return status;
 }
 
-void nfs4_put_copy(struct nfsd4_copy *copy)
+static void nfs4_put_copy(struct nfsd4_copy *copy)
 {
 	if (!refcount_dec_and_test(&copy->refcount))
 		return;
fs/nfsd/state.h +0 -1
index f3d6313914ed0..ae596dbf86675 100644
@@ -703,7 +703,6 @@ extern struct nfs4_client_reclaim *nfs4_client_to_reclaim(struct xdr_netobj name
 extern bool nfs4_has_reclaimed_state(struct xdr_netobj name, struct nfsd_net *nn);
 
 void put_nfs4_file(struct nfs4_file *fi);
-extern void nfs4_put_copy(struct nfsd4_copy *copy);
 extern struct nfsd4_copy *
 find_async_copy(struct nfs4_client *clp, stateid_t *staetid);
 extern void nfs4_put_cpntf_state(struct nfsd_net *nn,
Signature · 1 line
2.43.0
author maintainer
#394 lore raw

PATCH 394/690 NFSD: Replace boolean fields in struct nfsd4_copy

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 1913cdf56cb5bfbc8170873728d13598cbecda23 ]

Clean up: saves 8 bytes, and we can replace check_and_set_stop_copy() with an atomic bitop.

[ cel: adjusted to apply to v5.15.y ]
---
 fs/nfsd/nfs4proc.c | 49 +++++++++++++++++-----------------------------
 fs/nfsd/nfs4xdr.c  | 12 ++++++------
 fs/nfsd/xdr4.h     | 33 ++++++++++++++++++++++++++-----
 3 files changed, 52 insertions(+), 42 deletions(-)
fs/nfsd/nfs4proc.c +18 -31
index 3e4b0fb44c7b7..4fd6611d29ce4 100644
@@ -1295,23 +1295,9 @@ static void nfs4_put_copy(struct nfsd4_copy *copy)
 	kfree(copy);
 }
 
-static bool
-check_and_set_stop_copy(struct nfsd4_copy *copy)
-{
-	bool value;
-
-	spin_lock(&copy->cp_clp->async_lock);
-	value = copy->stopped;
-	if (!copy->stopped)
-		copy->stopped = true;
-	spin_unlock(&copy->cp_clp->async_lock);
-	return value;
-}
-
 static void nfsd4_stop_copy(struct nfsd4_copy *copy)
 {
-	/* only 1 thread should stop the copy */
-	if (!check_and_set_stop_copy(copy))
+	if (!test_and_set_bit(NFSD4_COPY_F_STOPPED, &copy->cp_flags))
 		kthread_stop(copy->copy_task);
 	nfs4_put_copy(copy);
 }
@@ -1669,8 +1655,9 @@ static const struct nfsd4_callback_ops nfsd4_cb_offload_ops = {
 static void nfsd4_init_copy_res(struct nfsd4_copy *copy, bool sync)
 {
 	copy->cp_res.wr_stable_how =
-		copy->committed ? NFS_FILE_SYNC : NFS_UNSTABLE;
-	copy->cp_synchronous = sync;
+		test_bit(NFSD4_COPY_F_COMMITTED, &copy->cp_flags) ?
+			NFS_FILE_SYNC : NFS_UNSTABLE;
+	nfsd4_copy_set_sync(copy, sync);
 	gen_boot_verifier(&copy->cp_res.wr_verifier, copy->cp_clp->net);
 }
 
@@ -1700,16 +1687,16 @@ static ssize_t _nfsd_copy_file_range(struct nfsd4_copy *copy)
 		copy->cp_res.wr_bytes_written += bytes_copied;
 		src_pos += bytes_copied;
 		dst_pos += bytes_copied;
-	} while (bytes_total > 0 && !copy->cp_synchronous);
+	} while (bytes_total > 0 && nfsd4_copy_is_async(copy));
 	/* for a non-zero asynchronous copy do a commit of data */
-	if (!copy->cp_synchronous && copy->cp_res.wr_bytes_written > 0) {
+	if (nfsd4_copy_is_async(copy) && copy->cp_res.wr_bytes_written > 0) {
 		since = READ_ONCE(dst->f_wb_err);
 		end = copy->cp_dst_pos + copy->cp_res.wr_bytes_written - 1;
 		status = vfs_fsync_range(dst, copy->cp_dst_pos, end, 0);
 		if (!status)
 			status = filemap_check_wb_err(dst->f_mapping, since);
 		if (!status)
-			copy->committed = true;
+			set_bit(NFSD4_COPY_F_COMMITTED, &copy->cp_flags);
 	}
 	return bytes_copied;
 }
@@ -1730,7 +1717,7 @@ static __be32 nfsd4_do_copy(struct nfsd4_copy *copy, bool sync)
 		status = nfs_ok;
 	}
 
-	if (!copy->cp_intra) /* Inter server SSC */
+	if (nfsd4_ssc_is_inter(copy))
 		nfsd4_cleanup_inter_ssc(copy->ss_mnt, copy->nf_src,
 					copy->nf_dst);
 	else
@@ -1744,13 +1731,13 @@ static void dup_copy_fields(struct nfsd4_copy *src, struct nfsd4_copy *dst)
 	dst->cp_src_pos = src->cp_src_pos;
 	dst->cp_dst_pos = src->cp_dst_pos;
 	dst->cp_count = src->cp_count;
-	dst->cp_synchronous = src->cp_synchronous;
+	dst->cp_flags = src->cp_flags;
 	memcpy(&dst->cp_res, &src->cp_res, sizeof(src->cp_res));
 	memcpy(&dst->fh, &src->fh, sizeof(src->fh));
 	dst->cp_clp = src->cp_clp;
 	dst->nf_dst = nfsd_file_get(src->nf_dst);
-	dst->cp_intra = src->cp_intra;
-	if (src->cp_intra) /* for inter, file_src doesn't exist yet */
+	/* for inter, nf_src doesn't exist yet */
+	if (!nfsd4_ssc_is_inter(src))
 		dst->nf_src = nfsd_file_get(src->nf_src);
 
 	memcpy(&dst->cp_stateid, &src->cp_stateid, sizeof(src->cp_stateid));
@@ -1764,7 +1751,7 @@ static void cleanup_async_copy(struct nfsd4_copy *copy)
 {
 	nfs4_free_copy_state(copy);
 	nfsd_file_put(copy->nf_dst);
-	if (copy->cp_intra)
+	if (!nfsd4_ssc_is_inter(copy))
 		nfsd_file_put(copy->nf_src);
 	spin_lock(&copy->cp_clp->async_lock);
 	list_del(&copy->copies);
@@ -1777,7 +1764,7 @@ static int nfsd4_do_async_copy(void *data)
 	struct nfsd4_copy *copy = (struct nfsd4_copy *)data;
 	struct nfsd4_copy *cb_copy;
 
-	if (!copy->cp_intra) { /* Inter server SSC */
+	if (nfsd4_ssc_is_inter(copy)) {
 		copy->nf_src = kzalloc(sizeof(struct nfsd_file), GFP_KERNEL);
 		if (!copy->nf_src) {
 			copy->nfserr = nfserr_serverfault;
@@ -1809,7 +1796,7 @@ static int nfsd4_do_async_copy(void *data)
 			      &copy->fh, copy->cp_count, copy->nfserr);
 	nfsd4_run_cb(&cb_copy->cp_cb);
 out:
-	if (!copy->cp_intra)
+	if (nfsd4_ssc_is_inter(copy))
 		kfree(copy->nf_src);
 	cleanup_async_copy(copy);
 	return 0;
@@ -1823,8 +1810,8 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	__be32 status;
 	struct nfsd4_copy *async_copy = NULL;
 
-	if (!copy->cp_intra) { /* Inter server SSC */
-		if (!inter_copy_offload_enable || copy->cp_synchronous) {
+	if (nfsd4_ssc_is_inter(copy)) {
+		if (!inter_copy_offload_enable || nfsd4_copy_is_sync(copy)) {
 			status = nfserr_notsupp;
 			goto out;
 		}
@@ -1841,7 +1828,7 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	copy->cp_clp = cstate->clp;
 	memcpy(&copy->fh, &cstate->current_fh.fh_handle,
 		sizeof(struct knfsd_fh));
-	if (!copy->cp_synchronous) {
+	if (nfsd4_copy_is_async(copy)) {
 		struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id);
 
 		status = nfserrno(-ENOMEM);
@@ -2607,7 +2594,7 @@ check_if_stalefh_allowed(struct nfsd4_compoundargs *args)
 				return;
 			}
 			putfh = (struct nfsd4_putfh *)&saved_op->u;
-			if (!copy->cp_intra)
+			if (nfsd4_ssc_is_inter(copy))
 				putfh->no_verify = true;
 		}
 	}
fs/nfsd/nfs4xdr.c +6 -6
index dad744d44b0a2..5aafbd0f7ae30 100644
@@ -1896,8 +1896,8 @@ static __be32 nfsd4_decode_nl4_server(struct nfsd4_compoundargs *argp,
 static __be32
 nfsd4_decode_copy(struct nfsd4_compoundargs *argp, struct nfsd4_copy *copy)
 {
+	u32 consecutive, i, count, sync;
 	struct nl4_server *ns_dummy;
-	u32 consecutive, i, count;
 	__be32 status;
 
 	status = nfsd4_decode_stateid4(argp, &copy->cp_src_stateid);
@@ -1915,17 +1915,17 @@ nfsd4_decode_copy(struct nfsd4_compoundargs *argp, struct nfsd4_copy *copy)
 	/* ca_consecutive: we always do consecutive copies */
 	if (xdr_stream_decode_u32(argp->xdr, &consecutive) < 0)
 		return nfserr_bad_xdr;
-	if (xdr_stream_decode_u32(argp->xdr, &copy->cp_synchronous) < 0)
+	if (xdr_stream_decode_bool(argp->xdr, &sync) < 0)
 		return nfserr_bad_xdr;
+	nfsd4_copy_set_sync(copy, sync);
 
 	if (xdr_stream_decode_u32(argp->xdr, &count) < 0)
 		return nfserr_bad_xdr;
 	copy->cp_src = svcxdr_tmpalloc(argp, sizeof(*copy->cp_src));
 	if (copy->cp_src == NULL)
 		return nfserr_jukebox;
-	copy->cp_intra = false;
 	if (count == 0) { /* intra-server copy */
-		copy->cp_intra = true;
+		__set_bit(NFSD4_COPY_F_INTRA, &copy->cp_flags);
 		return nfs_ok;
 	}
 
@@ -4709,13 +4709,13 @@ nfsd4_encode_copy(struct nfsd4_compoundres *resp, __be32 nfserr,
 	__be32 *p;
 
 	nfserr = nfsd42_encode_write_res(resp, &copy->cp_res,
-					 !!copy->cp_synchronous);
+					 nfsd4_copy_is_sync(copy));
 	if (nfserr)
 		return nfserr;
 
 	p = xdr_reserve_space(resp->xdr, 4 + 4);
 	*p++ = xdr_one; /* cr_consecutive */
-	*p++ = cpu_to_be32(copy->cp_synchronous);
+	*p = nfsd4_copy_is_sync(copy) ? xdr_one : xdr_zero;
 	return 0;
 }
 
fs/nfsd/xdr4.h +28 -5
index 87b4270af59ef..37d1b6d0486b3 100644
@@ -541,10 +541,12 @@ struct nfsd4_copy {
 	u64			cp_dst_pos;
 	u64			cp_count;
 	struct nl4_server	*cp_src;
-	bool			cp_intra;
 
-	/* both */
-	u32			cp_synchronous;
+	unsigned long		cp_flags;
+#define NFSD4_COPY_F_STOPPED		(0)
+#define NFSD4_COPY_F_INTRA		(1)
+#define NFSD4_COPY_F_SYNCHRONOUS	(2)
+#define NFSD4_COPY_F_COMMITTED		(3)
 
 	/* response */
 	struct nfsd42_write_res	cp_res;
@@ -564,14 +566,35 @@ struct nfsd4_copy {
 	struct list_head	copies;
 	struct task_struct	*copy_task;
 	refcount_t		refcount;
-	bool			stopped;
 
 	struct vfsmount		*ss_mnt;
 	struct nfs_fh		c_fh;
 	nfs4_stateid		stateid;
-	bool			committed;
 };
 
+static inline void nfsd4_copy_set_sync(struct nfsd4_copy *copy, bool sync)
+{
+	if (sync)
+		set_bit(NFSD4_COPY_F_SYNCHRONOUS, &copy->cp_flags);
+	else
+		clear_bit(NFSD4_COPY_F_SYNCHRONOUS, &copy->cp_flags);
+}
+
+static inline bool nfsd4_copy_is_sync(const struct nfsd4_copy *copy)
+{
+	return test_bit(NFSD4_COPY_F_SYNCHRONOUS, &copy->cp_flags);
+}
+
+static inline bool nfsd4_copy_is_async(const struct nfsd4_copy *copy)
+{
+	return !test_bit(NFSD4_COPY_F_SYNCHRONOUS, &copy->cp_flags);
+}
+
+static inline bool nfsd4_ssc_is_inter(const struct nfsd4_copy *copy)
+{
+	return !test_bit(NFSD4_COPY_F_INTRA, &copy->cp_flags);
+}
+
 struct nfsd4_seek {
 	/* request */
 	stateid_t	seek_stateid;
Signature · 1 line
2.43.0
author maintainer
#395 lore raw

PATCH 395/690 NFSD: Refactor nfsd4_cleanup_inter_ssc() (1/2)

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 24d796ea383b8a4c8234e06d1b14bbcd371192ea ]

The @src parameter is sometimes a pointer to a struct nfsd_file and sometimes a pointer to struct file hiding in a phony struct nfsd_file. Refactor nfsd4_cleanup_inter_ssc() so the @src parameter is always an explicit struct file.

---
 fs/nfsd/nfs4proc.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
fs/nfsd/nfs4proc.c +5 -5
index 4fd6611d29ce4..238df435b395d 100644
@@ -1550,7 +1550,7 @@ nfsd4_setup_inter_ssc(struct svc_rqst *rqstp,
 }
 
 static void
-nfsd4_cleanup_inter_ssc(struct vfsmount *ss_mnt, struct nfsd_file *src,
+nfsd4_cleanup_inter_ssc(struct vfsmount *ss_mnt, struct file *filp,
 			struct nfsd_file *dst)
 {
 	bool found = false;
@@ -1559,9 +1559,9 @@ nfsd4_cleanup_inter_ssc(struct vfsmount *ss_mnt, struct nfsd_file *src,
 	struct nfsd4_ssc_umount_item *ni = NULL;
 	struct nfsd_net *nn = net_generic(dst->nf_net, nfsd_net_id);
 
-	nfs42_ssc_close(src->nf_file);
+	nfs42_ssc_close(filp);
 	nfsd_file_put(dst);
-	fput(src->nf_file);
+	fput(filp);
 
 	if (!nn) {
 		mntput(ss_mnt);
@@ -1604,7 +1604,7 @@ nfsd4_setup_inter_ssc(struct svc_rqst *rqstp,
 }
 
 static void
-nfsd4_cleanup_inter_ssc(struct vfsmount *ss_mnt, struct nfsd_file *src,
+nfsd4_cleanup_inter_ssc(struct vfsmount *ss_mnt, struct file *filp,
 			struct nfsd_file *dst)
 {
 }
@@ -1718,7 +1718,7 @@ static __be32 nfsd4_do_copy(struct nfsd4_copy *copy, bool sync)
 	}
 
 	if (nfsd4_ssc_is_inter(copy))
-		nfsd4_cleanup_inter_ssc(copy->ss_mnt, copy->nf_src,
+		nfsd4_cleanup_inter_ssc(copy->ss_mnt, copy->nf_src->nf_file,
 					copy->nf_dst);
 	else
 		nfsd4_cleanup_intra_ssc(copy->nf_src, copy->nf_dst);
Signature · 1 line
2.43.0
author maintainer
#396 lore raw

PATCH 396/690 NFSD: Refactor nfsd4_cleanup_inter_ssc() (2/2)

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 478ed7b10d875da2743d1a22822b9f8a82df8f12 ]

Move the nfsd4_cleanup_*() call sites out of nfsd4_do_copy(). A subsequent patch will modify one of the new call sites to avoid the need to manufacture the phony struct nfsd_file.

---
 fs/nfsd/nfs4proc.c | 15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)
fs/nfsd/nfs4proc.c +7 -8
index 238df435b395d..10130f0b088ef 100644
@@ -1716,13 +1716,6 @@ static __be32 nfsd4_do_copy(struct nfsd4_copy *copy, bool sync)
 		nfsd4_init_copy_res(copy, sync);
 		status = nfs_ok;
 	}
-
-	if (nfsd4_ssc_is_inter(copy))
-		nfsd4_cleanup_inter_ssc(copy->ss_mnt, copy->nf_src->nf_file,
-					copy->nf_dst);
-	else
-		nfsd4_cleanup_intra_ssc(copy->nf_src, copy->nf_dst);
-
 	return status;
 }
 
@@ -1778,9 +1771,14 @@ static int nfsd4_do_async_copy(void *data)
 			/* ss_mnt will be unmounted by the laundromat */
 			goto do_callback;
 		}
+		copy->nfserr = nfsd4_do_copy(copy, 0);
+		nfsd4_cleanup_inter_ssc(copy->ss_mnt, copy->nf_src->nf_file,
+					copy->nf_dst);
+	} else {
+		copy->nfserr = nfsd4_do_copy(copy, 0);
+		nfsd4_cleanup_intra_ssc(copy->nf_src, copy->nf_dst);
 	}
 
-	copy->nfserr = nfsd4_do_copy(copy, 0);
 do_callback:
 	cb_copy = kzalloc(sizeof(struct nfsd4_copy), GFP_KERNEL);
 	if (!cb_copy)
@@ -1856,6 +1854,7 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		status = nfs_ok;
 	} else {
 		status = nfsd4_do_copy(copy, 1);
+		nfsd4_cleanup_intra_ssc(copy->nf_src, copy->nf_dst);
 	}
 out:
 	return status;
Signature · 1 line
2.43.0
author maintainer
#397 lore raw

PATCH 397/690 NFSD: Refactor nfsd4_do_copy()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 3b7bf5933cada732783554edf0dc61283551c6cf ]

Refactor: Now that nfsd4_do_copy() no longer calls the cleanup helpers, plumb the use of struct file pointers all the way down to _nfsd_copy_file_range().

---
 fs/nfsd/nfs4proc.c | 22 ++++++++++++++--------
 1 file changed, 14 insertions(+), 8 deletions(-)
fs/nfsd/nfs4proc.c +14 -8
index 10130f0b088ef..24c7d5e6c8c33 100644
@@ -1661,10 +1661,10 @@ static void nfsd4_init_copy_res(struct nfsd4_copy *copy, bool sync)
 	gen_boot_verifier(&copy->cp_res.wr_verifier, copy->cp_clp->net);
 }
 
-static ssize_t _nfsd_copy_file_range(struct nfsd4_copy *copy)
+static ssize_t _nfsd_copy_file_range(struct nfsd4_copy *copy,
+				     struct file *dst,
+				     struct file *src)
 {
-	struct file *dst = copy->nf_dst->nf_file;
-	struct file *src = copy->nf_src->nf_file;
 	errseq_t since;
 	ssize_t bytes_copied = 0;
 	u64 bytes_total = copy->cp_count;
@@ -1701,12 +1701,15 @@ static ssize_t _nfsd_copy_file_range(struct nfsd4_copy *copy)
 	return bytes_copied;
 }
 
-static __be32 nfsd4_do_copy(struct nfsd4_copy *copy, bool sync)
+static __be32 nfsd4_do_copy(struct nfsd4_copy *copy,
+			    struct file *src, struct file *dst,
+			    bool sync)
 {
 	__be32 status;
 	ssize_t bytes;
 
-	bytes = _nfsd_copy_file_range(copy);
+	bytes = _nfsd_copy_file_range(copy, dst, src);
+
 	/* for async copy, we ignore the error, client can always retry
 	 * to get the error
 	 */
@@ -1771,11 +1774,13 @@ static int nfsd4_do_async_copy(void *data)
 			/* ss_mnt will be unmounted by the laundromat */
 			goto do_callback;
 		}
-		copy->nfserr = nfsd4_do_copy(copy, 0);
+		copy->nfserr = nfsd4_do_copy(copy, copy->nf_src->nf_file,
+					     copy->nf_dst->nf_file, false);
 		nfsd4_cleanup_inter_ssc(copy->ss_mnt, copy->nf_src->nf_file,
 					copy->nf_dst);
 	} else {
-		copy->nfserr = nfsd4_do_copy(copy, 0);
+		copy->nfserr = nfsd4_do_copy(copy, copy->nf_src->nf_file,
+					     copy->nf_dst->nf_file, false);
 		nfsd4_cleanup_intra_ssc(copy->nf_src, copy->nf_dst);
 	}
 
@@ -1853,7 +1858,8 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		wake_up_process(async_copy->copy_task);
 		status = nfs_ok;
 	} else {
-		status = nfsd4_do_copy(copy, 1);
+		status = nfsd4_do_copy(copy, copy->nf_src->nf_file,
+				       copy->nf_dst->nf_file, true);
 		nfsd4_cleanup_intra_ssc(copy->nf_src, copy->nf_dst);
 	}
 out:
Signature · 1 line
2.43.0
author maintainer
#398 lore raw

PATCH 398/690 NFSD: Remove kmalloc from nfsd4_do_async_copy()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit ad1e46c9b07b13659635ee5405f83ad0df143116 ]

Instead of manufacturing a phony struct nfsd_file, pass the struct file returned by nfs42_ssc_open() directly to nfsd4_do_copy().

[ cel: adjusted to apply to v5.15.y ]
---
 fs/nfsd/nfs4proc.c | 28 ++++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)
fs/nfsd/nfs4proc.c +14 -14
index 24c7d5e6c8c33..f63c3c4c10ca7 100644
@@ -1755,29 +1755,31 @@ static void cleanup_async_copy(struct nfsd4_copy *copy)
 	nfs4_put_copy(copy);
 }
 
+/**
+ * nfsd4_do_async_copy - kthread function for background server-side COPY
+ * @data: arguments for COPY operation
+ *
+ * Return values:
+ *   %0: Copy operation is done.
+ */
 static int nfsd4_do_async_copy(void *data)
 {
 	struct nfsd4_copy *copy = (struct nfsd4_copy *)data;
 	struct nfsd4_copy *cb_copy;
 
 	if (nfsd4_ssc_is_inter(copy)) {
-		copy->nf_src = kzalloc(sizeof(struct nfsd_file), GFP_KERNEL);
-		if (!copy->nf_src) {
-			copy->nfserr = nfserr_serverfault;
-			/* ss_mnt will be unmounted by the laundromat */
-			goto do_callback;
-		}
-		copy->nf_src->nf_file = nfs42_ssc_open(copy->ss_mnt, &copy->c_fh,
-					      &copy->stateid);
-		if (IS_ERR(copy->nf_src->nf_file)) {
+		struct file *filp;
+
+		filp = nfs42_ssc_open(copy->ss_mnt, &copy->c_fh,
+				      &copy->stateid);
+		if (IS_ERR(filp)) {
 			copy->nfserr = nfserr_offload_denied;
 			/* ss_mnt will be unmounted by the laundromat */
 			goto do_callback;
 		}
-		copy->nfserr = nfsd4_do_copy(copy, copy->nf_src->nf_file,
+		copy->nfserr = nfsd4_do_copy(copy, filp,
 					     copy->nf_dst->nf_file, false);
-		nfsd4_cleanup_inter_ssc(copy->ss_mnt, copy->nf_src->nf_file,
-					copy->nf_dst);
+		nfsd4_cleanup_inter_ssc(copy->ss_mnt, filp, copy->nf_dst);
 	} else {
 		copy->nfserr = nfsd4_do_copy(copy, copy->nf_src->nf_file,
 					     copy->nf_dst->nf_file, false);
@@ -1799,8 +1801,6 @@ static int nfsd4_do_async_copy(void *data)
 			      &copy->fh, copy->cp_count, copy->nfserr);
 	nfsd4_run_cb(&cb_copy->cp_cb);
 out:
-	if (nfsd4_ssc_is_inter(copy))
-		kfree(copy->nf_src);
 	cleanup_async_copy(copy);
 	return 0;
 }
Signature · 1 line
2.43.0
author maintainer
#399 lore raw

PATCH 399/690 NFSD: Add nfsd4_send_cb_offload()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit e72f9bc006c08841c46d27747a4debc747a8fe13 ]

Refactor for legibility.

---
 fs/nfsd/nfs4proc.c | 37 ++++++++++++++++++++++---------------
 1 file changed, 22 insertions(+), 15 deletions(-)
fs/nfsd/nfs4proc.c +22 -15
index f63c3c4c10ca7..be51338deda46 100644
@@ -1755,6 +1755,27 @@ static void cleanup_async_copy(struct nfsd4_copy *copy)
 	nfs4_put_copy(copy);
 }
 
+static void nfsd4_send_cb_offload(struct nfsd4_copy *copy)
+{
+	struct nfsd4_copy *cb_copy;
+
+	cb_copy = kzalloc(sizeof(struct nfsd4_copy), GFP_KERNEL);
+	if (!cb_copy)
+		return;
+
+	refcount_set(&cb_copy->refcount, 1);
+	memcpy(&cb_copy->cp_res, &copy->cp_res, sizeof(copy->cp_res));
+	cb_copy->cp_clp = copy->cp_clp;
+	cb_copy->nfserr = copy->nfserr;
+	memcpy(&cb_copy->fh, &copy->fh, sizeof(copy->fh));
+
+	nfsd4_init_cb(&cb_copy->cp_cb, cb_copy->cp_clp,
+			&nfsd4_cb_offload_ops, NFSPROC4_CLNT_CB_OFFLOAD);
+	trace_nfsd_cb_offload(copy->cp_clp, &copy->cp_res.cb_stateid,
+			      &copy->fh, copy->cp_count, copy->nfserr);
+	nfsd4_run_cb(&cb_copy->cp_cb);
+}
+
 /**
  * nfsd4_do_async_copy - kthread function for background server-side COPY
  * @data: arguments for COPY operation
@@ -1765,7 +1786,6 @@ static void cleanup_async_copy(struct nfsd4_copy *copy)
 static int nfsd4_do_async_copy(void *data)
 {
 	struct nfsd4_copy *copy = (struct nfsd4_copy *)data;
-	struct nfsd4_copy *cb_copy;
 
 	if (nfsd4_ssc_is_inter(copy)) {
 		struct file *filp;
@@ -1787,20 +1807,7 @@ static int nfsd4_do_async_copy(void *data)
 	}
 
 do_callback:
-	cb_copy = kzalloc(sizeof(struct nfsd4_copy), GFP_KERNEL);
-	if (!cb_copy)
-		goto out;
-	refcount_set(&cb_copy->refcount, 1);
-	memcpy(&cb_copy->cp_res, &copy->cp_res, sizeof(copy->cp_res));
-	cb_copy->cp_clp = copy->cp_clp;
-	cb_copy->nfserr = copy->nfserr;
-	memcpy(&cb_copy->fh, &copy->fh, sizeof(copy->fh));
-	nfsd4_init_cb(&cb_copy->cp_cb, cb_copy->cp_clp,
-			&nfsd4_cb_offload_ops, NFSPROC4_CLNT_CB_OFFLOAD);
-	trace_nfsd_cb_offload(copy->cp_clp, &copy->cp_res.cb_stateid,
-			      &copy->fh, copy->cp_count, copy->nfserr);
-	nfsd4_run_cb(&cb_copy->cp_cb);
-out:
+	nfsd4_send_cb_offload(copy);
 	cleanup_async_copy(copy);
 	return 0;
 }
Signature · 1 line
2.43.0
author maintainer
#400 lore raw

PATCH 400/690 NFSD: Move copy offload callback arguments into a separate structure

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit a11ada99ce93a79393dc6683d22f7915748c8f6b ]

Refactor so that CB_OFFLOAD arguments can be passed without allocating a whole struct nfsd4_copy object. On my system (x86_64) this removes another 96 bytes from struct nfsd4_copy.

[ cel: adjusted to apply to v5.15.y ]
---
 fs/nfsd/nfs4callback.c | 37 +++++++++++++++++------------------
 fs/nfsd/nfs4proc.c     | 44 +++++++++++++++++++++---------------------
 fs/nfsd/xdr4.h         | 11 +++++++----
 3 files changed, 47 insertions(+), 45 deletions(-)
fs/nfsd/nfs4callback.c +18 -19
index e1272a7f45220..face8908a40b1 100644
@@ -679,7 +679,7 @@ static int nfs4_xdr_dec_cb_notify_lock(struct rpc_rqst *rqstp,
  *	case NFS4_OK:
  *		write_response4	coa_resok4;
  *	default:
- *	length4		coa_bytes_copied;
+ *		length4		coa_bytes_copied;
  * };
  * struct CB_OFFLOAD4args {
  *	nfs_fh4		coa_fh;
@@ -688,21 +688,22 @@ static int nfs4_xdr_dec_cb_notify_lock(struct rpc_rqst *rqstp,
  * };
  */
 static void encode_offload_info4(struct xdr_stream *xdr,
-				 __be32 nfserr,
-				 const struct nfsd4_copy *cp)
+				 const struct nfsd4_cb_offload *cbo)
 {
 	__be32 *p;
 
 	p = xdr_reserve_space(xdr, 4);
-	*p++ = nfserr;
-	if (!nfserr) {
+	*p = cbo->co_nfserr;
+	switch (cbo->co_nfserr) {
+	case nfs_ok:
 		p = xdr_reserve_space(xdr, 4 + 8 + 4 + NFS4_VERIFIER_SIZE);
 		p = xdr_encode_empty_array(p);
-		p = xdr_encode_hyper(p, cp->cp_res.wr_bytes_written);
-		*p++ = cpu_to_be32(cp->cp_res.wr_stable_how);
-		p = xdr_encode_opaque_fixed(p, cp->cp_res.wr_verifier.data,
+		p = xdr_encode_hyper(p, cbo->co_res.wr_bytes_written);
+		*p++ = cpu_to_be32(cbo->co_res.wr_stable_how);
+		p = xdr_encode_opaque_fixed(p, cbo->co_res.wr_verifier.data,
 					    NFS4_VERIFIER_SIZE);
-	} else {
+		break;
+	default:
 		p = xdr_reserve_space(xdr, 8);
 		/* We always return success if bytes were written */
 		p = xdr_encode_hyper(p, 0);
@@ -710,18 +711,16 @@ static void encode_offload_info4(struct xdr_stream *xdr,
 }
 
 static void encode_cb_offload4args(struct xdr_stream *xdr,
-				   __be32 nfserr,
-				   const struct knfsd_fh *fh,
-				   const struct nfsd4_copy *cp,
+				   const struct nfsd4_cb_offload *cbo,
 				   struct nfs4_cb_compound_hdr *hdr)
 {
 	__be32 *p;
 
 	p = xdr_reserve_space(xdr, 4);
-	*p++ = cpu_to_be32(OP_CB_OFFLOAD);
-	encode_nfs_fh4(xdr, fh);
-	encode_stateid4(xdr, &cp->cp_res.cb_stateid);
-	encode_offload_info4(xdr, nfserr, cp);
+	*p = cpu_to_be32(OP_CB_OFFLOAD);
+	encode_nfs_fh4(xdr, &cbo->co_fh);
+	encode_stateid4(xdr, &cbo->co_res.cb_stateid);
+	encode_offload_info4(xdr, cbo);
 
 	hdr->nops++;
 }
@@ -731,8 +730,8 @@ static void nfs4_xdr_enc_cb_offload(struct rpc_rqst *req,
 				    const void *data)
 {
 	const struct nfsd4_callback *cb = data;
-	const struct nfsd4_copy *cp =
-		container_of(cb, struct nfsd4_copy, cp_cb);
+	const struct nfsd4_cb_offload *cbo =
+		container_of(cb, struct nfsd4_cb_offload, co_cb);
 	struct nfs4_cb_compound_hdr hdr = {
 		.ident = 0,
 		.minorversion = cb->cb_clp->cl_minorversion,
@@ -740,7 +739,7 @@ static void nfs4_xdr_enc_cb_offload(struct rpc_rqst *req,
 
 	encode_cb_compound4args(xdr, &hdr);
 	encode_cb_sequence4args(xdr, cb, &hdr);
-	encode_cb_offload4args(xdr, cp->nfserr, &cp->fh, cp, &hdr);
+	encode_cb_offload4args(xdr, cbo, &hdr);
 	encode_cb_nops(&hdr);
 }
 
fs/nfsd/nfs4proc.c +22 -22

83 lines not on this page · raw message

fs/nfsd/xdr4.h +7 -4
index 37d1b6d0486b3..adb9aef26d7f1 100644
@@ -533,6 +533,13 @@ struct nfsd42_write_res {
 	stateid_t		cb_stateid;
 };
 
+struct nfsd4_cb_offload {
+	struct nfsd4_callback	co_cb;
+	struct nfsd42_write_res	co_res;
+	__be32			co_nfserr;
+	struct knfsd_fh		co_fh;
+};
+
 struct nfsd4_copy {
 	/* request */
 	stateid_t		cp_src_stateid;
@@ -550,10 +557,6 @@ struct nfsd4_copy {
 
 	/* response */
 	struct nfsd42_write_res	cp_res;
-
-	/* for cb_offload */
-	struct nfsd4_callback	cp_cb;
-	__be32			nfserr;
 	struct knfsd_fh		fh;
 
 	struct nfs4_client      *cp_clp;
Signature · 1 line
2.43.0
author maintainer
#401 lore raw

PATCH 401/690 NFSD: drop fh argument from alloc_init_deleg

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jeff Layton <[email protected]>

[ Upstream commit bbf936edd543e7220f60f9cbd6933b916550396d ]

Currently, we pass the fh of the opened file down through several functions so that alloc_init_deleg can pass it to delegation_blocked. The filehandle of the open file is available in the nfs4_file however, so there's no need to pass it in a separate argument.

Drop the argument from alloc_init_deleg, nfs4_open_delegation and nfs4_set_delegation.

---
 fs/nfsd/nfs4state.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)
fs/nfsd/nfs4state.c +6 -8
index 340d533dcafd3..2b333f9259a03 100644
@@ -1143,7 +1143,6 @@ static void block_delegations(struct knfsd_fh *fh)
 
 static struct nfs4_delegation *
 alloc_init_deleg(struct nfs4_client *clp, struct nfs4_file *fp,
-		 struct svc_fh *current_fh,
 		 struct nfs4_clnt_odstate *odstate)
 {
 	struct nfs4_delegation *dp;
@@ -1153,7 +1152,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_file *fp,
 	n = atomic_long_inc_return(&num_delegations);
 	if (n < 0 || n > max_delegations)
 		goto out_dec;
-	if (delegation_blocked(&current_fh->fh_handle))
+	if (delegation_blocked(&fp->fi_fhandle))
 		goto out_dec;
 	dp = delegstateid(nfs4_alloc_stid(clp, deleg_slab, nfs4_free_deleg));
 	if (dp == NULL)
@@ -5307,7 +5306,7 @@ static int nfsd4_check_conflicting_opens(struct nfs4_client *clp,
 }
 
 static struct nfs4_delegation *
-nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh,
+nfs4_set_delegation(struct nfs4_client *clp,
 		    struct nfs4_file *fp, struct nfs4_clnt_odstate *odstate)
 {
 	int status = 0;
@@ -5352,7 +5351,7 @@ nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh,
 		return ERR_PTR(status);
 
 	status = -ENOMEM;
-	dp = alloc_init_deleg(clp, fp, fh, odstate);
+	dp = alloc_init_deleg(clp, fp, odstate);
 	if (!dp)
 		goto out_delegees;
 
@@ -5420,8 +5419,7 @@ static void nfsd4_open_deleg_none_ext(struct nfsd4_open *open, int status)
  * proper support for them.
  */
 static void
-nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open,
-			struct nfs4_ol_stateid *stp)
+nfs4_open_delegation(struct nfsd4_open *open, struct nfs4_ol_stateid *stp)
 {
 	struct nfs4_delegation *dp;
 	struct nfs4_openowner *oo = openowner(stp->st_stateowner);
@@ -5453,7 +5451,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open,
 		default:
 			goto out_no_deleg;
 	}
-	dp = nfs4_set_delegation(clp, fh, stp->st_stid.sc_file, stp->st_clnt_odstate);
+	dp = nfs4_set_delegation(clp, stp->st_stid.sc_file, stp->st_clnt_odstate);
 	if (IS_ERR(dp))
 		goto out_no_deleg;
 
@@ -5585,7 +5583,7 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf
 	* Attempt to hand out a delegation. No error return, because the
 	* OPEN succeeds even if we fail.
 	*/
-	nfs4_open_delegation(current_fh, open, stp);
+	nfs4_open_delegation(open, stp);
 nodeleg:
 	status = nfs_ok;
 	trace_nfsd_open(&stp->st_stid.sc_stateid);
Signature · 1 line
2.43.0
author maintainer
#402 lore raw

PATCH 402/690 NFSD: verify the opened dentry after setting a delegation

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jeff Layton <[email protected]>

[ Upstream commit 876c553cb41026cb6ad3cef970a35e5f69c42a25 ]

Between opening a file and setting a delegation on it, someone could rename or unlink the dentry. If this happens, we do not want to grant a delegation on the open.

On a CLAIM_NULL open, we're opening by filename, and we may (in the non-create case) or may not (in the create case) be holding i_rwsem when attempting to set a delegation. The latter case allows a race.

After getting a lease, redo the lookup of the file being opened and validate that the resulting dentry matches the one in the open file description.

To properly redo the lookup we need an rqst pointer to pass to nfsd_lookup_dentry(), so make sure that is available.

---
 fs/nfsd/nfs4proc.c  |  1 +
 fs/nfsd/nfs4state.c | 54 ++++++++++++++++++++++++++++++++++++++++-----
 fs/nfsd/xdr4.h      |  1 +
 3 files changed, 51 insertions(+), 5 deletions(-)
fs/nfsd/nfs4proc.c +1 -0
index 46ec66f4ec9e7..ae0948271da9c 100644
@@ -547,6 +547,7 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		open->op_openowner);
 
 	open->op_filp = NULL;
+	open->op_rqstp = rqstp;
 
 	/* This check required by spec. */
 	if (open->op_create && open->op_claim_type != NFS4_OPEN_CLAIM_NULL)
fs/nfsd/nfs4state.c +49 -5

97 lines not on this page · raw message

fs/nfsd/xdr4.h +1 -0
index adb9aef26d7f1..466e2786fc976 100644
@@ -279,6 +279,7 @@ struct nfsd4_open {
 	struct nfs4_clnt_odstate *op_odstate; /* used during processing */
 	struct nfs4_acl *op_acl;
 	struct xdr_netobj op_label;
+	struct svc_rqst *op_rqstp;
 };
 
 struct nfsd4_open_confirm {
Signature · 1 line
2.43.0
author maintainer
#403 lore raw

PATCH 403/690 NFSD: introduce struct nfsd_attrs

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit 7fe2a71dda349a1afa75781f0cc7975be9784d15 ]

The attributes that nfsd might want to set on a file include 'struct iattr' as well as an ACL and security label.
The latter two are passed around quite separately from the first, in part because they are only needed for NFSv4. This leads to some clumsiness in the code, such as the attributes NOT being set in nfsd_create_setattr().

We need to keep the directory locked until all attributes are set to ensure the file is never visibile without all its attributes. This need combined with the inconsistent handling of attributes leads to more clumsiness.

As a first step towards tidying this up, introduce 'struct nfsd_attrs'. This is passed (by reference) to vfs.c functions that work with attributes, and is assembled by the various nfs*proc functions which call them. As yet only iattr is included, but future patches will expand this.

---
 fs/nfsd/nfs3proc.c  | 20 ++++++++++++++++----
 fs/nfsd/nfs4proc.c  | 23 ++++++++++++++++-------
 fs/nfsd/nfs4state.c |  5 ++++-
 fs/nfsd/nfsproc.c   | 17 +++++++++++++----
 fs/nfsd/vfs.c       | 24 ++++++++++++++----------
 fs/nfsd/vfs.h       | 12 ++++++++----
 6 files changed, 71 insertions(+), 30 deletions(-)
fs/nfsd/nfs3proc.c +16 -4
index 57854ca022d18..113567b3a98a5 100644
@@ -67,12 +67,15 @@ nfsd3_proc_setattr(struct svc_rqst *rqstp)
 {
 	struct nfsd3_sattrargs *argp = rqstp->rq_argp;
 	struct nfsd3_attrstat *resp = rqstp->rq_resp;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &argp->attrs,
+	};
 
 	dprintk("nfsd: SETATTR(3)  %s\n",
 				SVCFH_fmt(&argp->fh));
 
 	fh_copy(&resp->fh, &argp->fh);
-	resp->status = nfsd_setattr(rqstp, &resp->fh, &argp->attrs,
+	resp->status = nfsd_setattr(rqstp, &resp->fh, &attrs,
 				    argp->check_guard, argp->guardtime);
 	return rpc_success;
 }
@@ -233,6 +236,9 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 {
 	struct iattr *iap = &argp->attrs;
 	struct dentry *parent, *child;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= iap,
+	};
 	__u32 v_mtime, v_atime;
 	struct inode *inode;
 	__be32 status;
@@ -331,7 +337,7 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	}
 
 set_attr:
-	status = nfsd_create_setattr(rqstp, fhp, resfhp, iap);
+	status = nfsd_create_setattr(rqstp, fhp, resfhp, &attrs);
 
 out:
 	fh_unlock(fhp);
@@ -368,6 +374,9 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 {
 	struct nfsd3_createargs *argp = rqstp->rq_argp;
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &argp->attrs,
+	};
 
 	dprintk("nfsd: MKDIR(3)    %s %.*s\n",
 				SVCFH_fmt(&argp->fh),
@@ -378,7 +387,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 	fh_copy(&resp->dirfh, &argp->fh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
 	resp->status = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len,
-				   &argp->attrs, S_IFDIR, 0, &resp->fh);
+				   &attrs, S_IFDIR, 0, &resp->fh);
 	fh_unlock(&resp->dirfh);
 	return rpc_success;
 }
@@ -428,6 +437,9 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 {
 	struct nfsd3_mknodargs *argp = rqstp->rq_argp;
 	struct nfsd3_diropres  *resp = rqstp->rq_resp;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &argp->attrs,
+	};
 	int type;
 	dev_t	rdev = 0;
 
@@ -453,7 +465,7 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 
 	type = nfs3_ftypes[argp->ftype];
 	resp->status = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len,
-				   &argp->attrs, type, rdev, &resp->fh);
+				   &attrs, type, rdev, &resp->fh);
 	fh_unlock(&resp->dirfh);
 out:
 	return rpc_success;
fs/nfsd/nfs4proc.c +16 -7
index ae0948271da9c..9b04611a318d7 100644
@@ -286,6 +286,9 @@ 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 nfsd_attrs attrs = {
+		.na_iattr	= iap,
+	};
 	struct dentry *parent, *child;
 	__u32 v_mtime, v_atime;
 	struct inode *inode;
@@ -404,7 +407,7 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	}
 
 set_attr:
-	status = nfsd_create_setattr(rqstp, fhp, resfhp, iap);
+	status = nfsd_create_setattr(rqstp, fhp, resfhp, &attrs);
 
 out:
 	fh_unlock(fhp);
@@ -787,6 +790,9 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	     union nfsd4_op_u *u)
 {
 	struct nfsd4_create *create = &u->create;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &create->cr_iattr,
+	};
 	struct svc_fh resfh;
 	__be32 status;
 	dev_t rdev;
@@ -818,7 +824,7 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 			goto out_umask;
 		status = nfsd_create(rqstp, &cstate->current_fh,
 				     create->cr_name, create->cr_namelen,
-				     &create->cr_iattr, S_IFBLK, rdev, &resfh);
+				     &attrs, S_IFBLK, rdev, &resfh);
 		break;
 
 	case NF4CHR:
@@ -829,26 +835,26 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 			goto out_umask;
 		status = nfsd_create(rqstp, &cstate->current_fh,
 				     create->cr_name, create->cr_namelen,
-				     &create->cr_iattr, S_IFCHR, rdev, &resfh);
+				     &attrs, S_IFCHR, rdev, &resfh);
 		break;
 
 	case NF4SOCK:
 		status = nfsd_create(rqstp, &cstate->current_fh,
 				     create->cr_name, create->cr_namelen,
-				     &create->cr_iattr, S_IFSOCK, 0, &resfh);
+				     &attrs, S_IFSOCK, 0, &resfh);
 		break;
 
 	case NF4FIFO:
 		status = nfsd_create(rqstp, &cstate->current_fh,
 				     create->cr_name, create->cr_namelen,
-				     &create->cr_iattr, S_IFIFO, 0, &resfh);
+				     &attrs, S_IFIFO, 0, &resfh);
 		break;
 
 	case NF4DIR:
 		create->cr_iattr.ia_valid &= ~ATTR_SIZE;
 		status = nfsd_create(rqstp, &cstate->current_fh,
 				     create->cr_name, create->cr_namelen,
-				     &create->cr_iattr, S_IFDIR, 0, &resfh);
+				     &attrs, S_IFDIR, 0, &resfh);
 		break;
 
 	default:
@@ -1142,6 +1148,9 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	      union nfsd4_op_u *u)
 {
 	struct nfsd4_setattr *setattr = &u->setattr;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &setattr->sa_iattr,
+	};
 	__be32 status = nfs_ok;
 	int err;
 
@@ -1174,7 +1183,7 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 				&setattr->sa_label);
 	if (status)
 		goto out;
-	status = nfsd_setattr(rqstp, &cstate->current_fh, &setattr->sa_iattr,
+	status = nfsd_setattr(rqstp, &cstate->current_fh, &attrs,
 				0, (time64_t)0);
 out:
 	fh_drop_write(&cstate->current_fh);
fs/nfsd/nfs4state.c +4 -1
index 7122ebc50a035..a299aeaa0de07 100644
@@ -5077,11 +5077,14 @@ nfsd4_truncate(struct svc_rqst *rqstp, struct svc_fh *fh,
 		.ia_valid = ATTR_SIZE,
 		.ia_size = 0,
 	};
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &iattr,
+	};
 	if (!open->op_truncate)
 		return 0;
 	if (!(open->op_share_access & NFS4_SHARE_ACCESS_WRITE))
 		return nfserr_inval;
-	return nfsd_setattr(rqstp, fh, &iattr, 0, (time64_t)0);
+	return nfsd_setattr(rqstp, fh, &attrs, 0, (time64_t)0);
 }
 
 static __be32 nfs4_get_vfs_file(struct svc_rqst *rqstp, struct nfs4_file *fp,
fs/nfsd/nfsproc.c +13 -4
index f65eba938a57d..c75d83bc3f21b 100644
@@ -51,6 +51,9 @@ nfsd_proc_setattr(struct svc_rqst *rqstp)
 	struct nfsd_sattrargs *argp = rqstp->rq_argp;
 	struct nfsd_attrstat *resp = rqstp->rq_resp;
 	struct iattr *iap = &argp->attrs;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= iap,
+	};
 	struct svc_fh *fhp;
 
 	dprintk("nfsd: SETATTR  %s, valid=%x, size=%ld\n",
@@ -100,7 +103,7 @@ nfsd_proc_setattr(struct svc_rqst *rqstp)
 		}
 	}
 
-	resp->status = nfsd_setattr(rqstp, fhp, iap, 0, (time64_t)0);
+	resp->status = nfsd_setattr(rqstp, fhp, &attrs, 0, (time64_t)0);
 	if (resp->status != nfs_ok)
 		goto out;
 
@@ -261,6 +264,9 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	svc_fh		*dirfhp = &argp->fh;
 	svc_fh		*newfhp = &resp->fh;
 	struct iattr	*attr = &argp->attrs;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= attr,
+	};
 	struct inode	*inode;
 	struct dentry	*dchild;
 	int		type, mode;
@@ -386,7 +392,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	if (!inode) {
 		/* File doesn't exist. Create it and set attrs */
 		resp->status = nfsd_create_locked(rqstp, dirfhp, argp->name,
-						  argp->len, attr, type, rdev,
+						  argp->len, &attrs, type, rdev,
 						  newfhp);
 	} else if (type == S_IFREG) {
 		dprintk("nfsd:   existing %s, valid=%x, size=%ld\n",
@@ -397,7 +403,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 		 */
 		attr->ia_valid &= ATTR_SIZE;
 		if (attr->ia_valid)
-			resp->status = nfsd_setattr(rqstp, newfhp, attr, 0,
+			resp->status = nfsd_setattr(rqstp, newfhp, &attrs, 0,
 						    (time64_t)0);
 	}
 
@@ -512,6 +518,9 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 {
 	struct nfsd_createargs *argp = rqstp->rq_argp;
 	struct nfsd_diropres *resp = rqstp->rq_resp;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &argp->attrs,
+	};
 
 	dprintk("nfsd: MKDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
 
@@ -523,7 +532,7 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 	argp->attrs.ia_valid &= ~ATTR_SIZE;
 	fh_init(&resp->fh, NFS_FHSIZE);
 	resp->status = nfsd_create(rqstp, &argp->fh, argp->name, argp->len,
-				   &argp->attrs, S_IFDIR, 0, &resp->fh);
+				   &attrs, S_IFDIR, 0, &resp->fh);
 	fh_put(&argp->fh);
 	if (resp->status != nfs_ok)
 		goto out;
fs/nfsd/vfs.c +14 -10
index 6689ad5bb790d..489225de05a2a 100644
@@ -350,11 +350,13 @@ nfsd_get_write_access(struct svc_rqst *rqstp, struct svc_fh *fhp,
  * Set various file attributes.  After this call fhp needs an fh_put.
  */
 __be32
-nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap,
+nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
+	     struct nfsd_attrs *attr,
 	     int check_guard, time64_t guardtime)
 {
 	struct dentry	*dentry;
 	struct inode	*inode;
+	struct iattr	*iap = attr->na_iattr;
 	int		accmode = NFSD_MAY_SATTR;
 	umode_t		ftype = 0;
 	__be32		err;
@@ -1203,14 +1205,15 @@ nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp, u64 offset,
  * @rqstp: RPC transaction being executed
  * @fhp: NFS filehandle of parent directory
  * @resfhp: NFS filehandle of new object
- * @iap: requested attributes of new object
+ * @attrs: requested attributes of new object
  *
  * Returns nfs_ok on success, or an nfsstat in network byte order.
  */
 __be32
 nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		    struct svc_fh *resfhp, struct iattr *iap)
+		    struct svc_fh *resfhp, struct nfsd_attrs *attrs)
 {
+	struct iattr *iap = attrs->na_iattr;
 	__be32 status;
 
 	/*
@@ -1231,7 +1234,7 @@ nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	 * if the attributes have not changed.
 	 */
 	if (iap->ia_valid)
-		status = nfsd_setattr(rqstp, resfhp, iap, 0, (time64_t)0);
+		status = nfsd_setattr(rqstp, resfhp, attrs, 0, (time64_t)0);
 	else
 		status = nfserrno(commit_metadata(resfhp));
 
@@ -1270,11 +1273,12 @@ nfsd_check_ignore_resizing(struct iattr *iap)
 /* The parent directory should already be locked: */
 __be32
 nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		char *fname, int flen, struct iattr *iap,
-		int type, dev_t rdev, struct svc_fh *resfhp)
+		   char *fname, int flen, struct nfsd_attrs *attrs,
+		   int type, dev_t rdev, struct svc_fh *resfhp)
 {
 	struct dentry	*dentry, *dchild;
 	struct inode	*dirp;
+	struct iattr	*iap = attrs->na_iattr;
 	__be32		err;
 	int		host_err;
 
@@ -1348,7 +1352,7 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (host_err < 0)
 		goto out_nfserr;
 
-	err = nfsd_create_setattr(rqstp, fhp, resfhp, iap);
+	err = nfsd_create_setattr(rqstp, fhp, resfhp, attrs);
 
 out:
 	dput(dchild);
@@ -1367,8 +1371,8 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
  */
 __be32
 nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		char *fname, int flen, struct iattr *iap,
-		int type, dev_t rdev, struct svc_fh *resfhp)
+	    char *fname, int flen, struct nfsd_attrs *attrs,
+	    int type, dev_t rdev, struct svc_fh *resfhp)
 {
 	struct dentry	*dentry, *dchild = NULL;
 	__be32		err;
@@ -1400,7 +1404,7 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	dput(dchild);
 	if (err)
 		return err;
-	return nfsd_create_locked(rqstp, fhp, fname, flen, iap, type,
+	return nfsd_create_locked(rqstp, fhp, fname, flen, attrs, type,
 					rdev, resfhp);
 }
 
fs/nfsd/vfs.h +8 -4
index 26347d76f44a0..d8b1a36fca956 100644
@@ -42,6 +42,10 @@ struct nfsd_file;
 typedef int (*nfsd_filldir_t)(void *, const char *, int, loff_t, u64, unsigned);
 
 /* nfsd/vfs.c */
+struct nfsd_attrs {
+	struct iattr		*na_iattr;	/* input */
+};
+
 int		nfsd_cross_mnt(struct svc_rqst *rqstp, struct dentry **dpp,
 		                struct svc_export **expp);
 __be32		nfsd_lookup(struct svc_rqst *, struct svc_fh *,
@@ -50,7 +54,7 @@ __be32		 nfsd_lookup_dentry(struct svc_rqst *, struct svc_fh *,
 				const char *, unsigned int,
 				struct svc_export **, struct dentry **);
 __be32		nfsd_setattr(struct svc_rqst *, struct svc_fh *,
-				struct iattr *, int, time64_t);
+				struct nfsd_attrs *, int, time64_t);
 int nfsd_mountpoint(struct dentry *, struct svc_export *);
 #ifdef CONFIG_NFSD_V4
 __be32          nfsd4_set_nfs4_label(struct svc_rqst *, struct svc_fh *,
@@ -63,14 +67,14 @@ __be32		nfsd4_clone_file_range(struct svc_rqst *rqstp,
 				       u64 count, bool sync);
 #endif /* CONFIG_NFSD_V4 */
 __be32		nfsd_create_locked(struct svc_rqst *, struct svc_fh *,
-				char *name, int len, struct iattr *attrs,
+				char *name, int len, struct nfsd_attrs *attrs,
 				int type, dev_t rdev, struct svc_fh *res);
 __be32		nfsd_create(struct svc_rqst *, struct svc_fh *,
-				char *name, int len, struct iattr *attrs,
+				char *name, int len, struct nfsd_attrs *attrs,
 				int type, dev_t rdev, struct svc_fh *res);
 __be32		nfsd_access(struct svc_rqst *, struct svc_fh *, u32 *, u32 *);
 __be32		nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
-				struct svc_fh *resfhp, struct iattr *iap);
+				struct svc_fh *resfhp, struct nfsd_attrs *iap);
 __be32		nfsd_commit(struct svc_rqst *rqst, struct svc_fh *fhp,
 				u64 offset, u32 count, __be32 *verf);
 #ifdef CONFIG_NFSD_V4
Signature · 1 line
2.43.0
author maintainer
#404 lore raw

PATCH 404/690 NFSD: set attributes when creating symlinks

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit 93adc1e391a761441d783828b93979b38093d011 ]

The NFS protocol includes attributes when creating symlinks. Linux does store attributes for symlinks and allows them to be set, though they are not used for permission checking.

NFSD currently doesn't set standard (struct iattr) attributes when creating symlinks, but for NFSv4 it does set ACLs and security labels. This is inconsistent.

To improve consistency, pass the provided attributes into nfsd_symlink() and call nfsd_create_setattr() to set them.

NOTE: this results in a behaviour change for all NFS versions when the client sends non-default attributes with a SYMLINK request. With the Linux client, the only attributes are:

        attr.ia_mode = S_IFLNK | S_IRWXUGO;
        attr.ia_valid = ATTR_MODE;

so the final outcome will be unchanged. Other clients might sent different attributes, and if they did they probably expect them to be honoured.

We ignore any error from nfsd_create_setattr(). It isn't really clear what should be done if a file is successfully created, but the attributes cannot be set. NFS doesn't allow partial success to be reported. Reporting failure is probably more misleading than reporting success, so the status is ignored.

---
 fs/nfsd/nfs3proc.c |  5 ++++-
 fs/nfsd/nfs4proc.c |  2 +-
 fs/nfsd/nfsproc.c  |  5 ++++-
 fs/nfsd/vfs.c      | 25 ++++++++++++++++++-------
 fs/nfsd/vfs.h      |  5 +++--
 5 files changed, 30 insertions(+), 12 deletions(-)
fs/nfsd/nfs3proc.c +4 -1
index 113567b3a98a5..cb91088bce2e8 100644
@@ -397,6 +397,9 @@ nfsd3_proc_symlink(struct svc_rqst *rqstp)
 {
 	struct nfsd3_symlinkargs *argp = rqstp->rq_argp;
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &argp->attrs,
+	};
 
 	if (argp->tlen == 0) {
 		resp->status = nfserr_inval;
@@ -423,7 +426,7 @@ nfsd3_proc_symlink(struct svc_rqst *rqstp)
 	fh_copy(&resp->dirfh, &argp->ffh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
 	resp->status = nfsd_symlink(rqstp, &resp->dirfh, argp->fname,
-				    argp->flen, argp->tname, &resp->fh);
+				    argp->flen, argp->tname, &attrs, &resp->fh);
 	kfree(argp->tname);
 out:
 	return rpc_success;
fs/nfsd/nfs4proc.c +1 -1
index 9b04611a318d7..96f6fe4f86fd8 100644
@@ -813,7 +813,7 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	case NF4LNK:
 		status = nfsd_symlink(rqstp, &cstate->current_fh,
 				      create->cr_name, create->cr_namelen,
-				      create->cr_data, &resfh);
+				      create->cr_data, &attrs, &resfh);
 		break;
 
 	case NF4BLK:
fs/nfsd/nfsproc.c +4 -1
index c75d83bc3f21b..09afd188099be 100644
@@ -479,6 +479,9 @@ nfsd_proc_symlink(struct svc_rqst *rqstp)
 {
 	struct nfsd_symlinkargs *argp = rqstp->rq_argp;
 	struct nfsd_stat *resp = rqstp->rq_resp;
+	struct nfsd_attrs attrs = {
+		.na_iattr	= &argp->attrs,
+	};
 	struct svc_fh	newfh;
 
 	if (argp->tlen > NFS_MAXPATHLEN) {
@@ -500,7 +503,7 @@ nfsd_proc_symlink(struct svc_rqst *rqstp)
 
 	fh_init(&newfh, NFS_FHSIZE);
 	resp->status = nfsd_symlink(rqstp, &argp->ffh, argp->fname, argp->flen,
-				    argp->tname, &newfh);
+				    argp->tname, &attrs, &newfh);
 
 	kfree(argp->tname);
 	fh_put(&argp->ffh);
fs/nfsd/vfs.c +18 -7
index 489225de05a2a..bfdb42aa23a01 100644
@@ -1446,15 +1446,25 @@ nfsd_readlink(struct svc_rqst *rqstp, struct svc_fh *fhp, char *buf, int *lenp)
 	return 0;
 }
 
-/*
- * Create a symlink and look up its inode
+/**
+ * nfsd_symlink - Create a symlink and look up its inode
+ * @rqstp: RPC transaction being executed
+ * @fhp: NFS filehandle of parent directory
+ * @fname: filename of the new symlink
+ * @flen: length of @fname
+ * @path: content of the new symlink (NUL-terminated)
+ * @attrs: requested attributes of new object
+ * @resfhp: NFS filehandle of new object
+ *
  * N.B. After this call _both_ fhp and resfhp need an fh_put
+ *
+ * Returns nfs_ok on success, or an nfsstat in network byte order.
  */
 __be32
 nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
-				char *fname, int flen,
-				char *path,
-				struct svc_fh *resfhp)
+	     char *fname, int flen,
+	     char *path, struct nfsd_attrs *attrs,
+	     struct svc_fh *resfhp)
 {
 	struct dentry	*dentry, *dnew;
 	__be32		err, cerr;
@@ -1484,13 +1494,14 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
 
 	host_err = vfs_symlink(&init_user_ns, d_inode(dentry), dnew, path);
 	err = nfserrno(host_err);
+	cerr = fh_compose(resfhp, fhp->fh_export, dnew, fhp);
+	if (!err)
+		nfsd_create_setattr(rqstp, fhp, resfhp, attrs);
 	fh_unlock(fhp);
 	if (!err)
 		err = nfserrno(commit_metadata(fhp));
-
 	fh_drop_write(fhp);
 
-	cerr = fh_compose(resfhp, fhp->fh_export, dnew, fhp);
 	dput(dnew);
 	if (err==0) err = cerr;
 out:
fs/nfsd/vfs.h +3 -2
index d8b1a36fca956..5047cec4c423c 100644
@@ -114,8 +114,9 @@ __be32		nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp,
 __be32		nfsd_readlink(struct svc_rqst *, struct svc_fh *,
 				char *, int *);
 __be32		nfsd_symlink(struct svc_rqst *, struct svc_fh *,
-				char *name, int len, char *path,
-				struct svc_fh *res);
+			     char *name, int len, char *path,
+			     struct nfsd_attrs *attrs,
+			     struct svc_fh *res);
 __be32		nfsd_link(struct svc_rqst *, struct svc_fh *,
 				char *, int, struct svc_fh *);
 ssize_t		nfsd_copy_file_range(struct file *, u64,
Signature · 1 line
2.43.0
author maintainer
#405 lore raw

PATCH 405/690 NFSD: add security label to struct nfsd_attrs

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit d6a97d3f589a3a46a16183e03f3774daee251317 ]

nfsd_setattr() now sets a security label if provided, and nfsv4 provides it in the 'open' and 'create' paths and the 'setattr' path. If setting the label failed (including because the kernel doesn't support labels), an error field in 'struct nfsd_attrs' is set, and the caller can respond. The open/create callers clear FATTR4_WORD2_SECURITY_LABEL in the returned attr set in this case. The setattr caller returns the error.

---
 fs/nfsd/nfs4proc.c | 49 +++++++++-------------------------------------
 fs/nfsd/vfs.c      | 29 +++------------------------
 fs/nfsd/vfs.h      |  5 +++--
 3 files changed, 15 insertions(+), 68 deletions(-)
fs/nfsd/nfs4proc.c +9 -40
index 96f6fe4f86fd8..7ebf807f33d98 100644
@@ -64,36 +64,6 @@ MODULE_PARM_DESC(nfsd4_ssc_umount_timeout,
 		"idle msecs before unmount export from source server");
 #endif
 
-#ifdef CONFIG_NFSD_V4_SECURITY_LABEL
-#include <linux/security.h>
-
-static inline void
-nfsd4_security_inode_setsecctx(struct svc_fh *resfh, struct xdr_netobj *label, u32 *bmval)
-{
-	struct inode *inode = d_inode(resfh->fh_dentry);
-	int status;
-
-	inode_lock(inode);
-	status = security_inode_setsecctx(resfh->fh_dentry,
-		label->data, label->len);
-	inode_unlock(inode);
-
-	if (status)
-		/*
-		 * XXX: We should really fail the whole open, but we may
-		 * already have created a new file, so it may be too
-		 * late.  For now this seems the least of evils:
-		 */
-		bmval[2] &= ~FATTR4_WORD2_SECURITY_LABEL;
-
-	return;
-}
-#else
-static inline void
-nfsd4_security_inode_setsecctx(struct svc_fh *resfh, struct xdr_netobj *label, u32 *bmval)
-{ }
-#endif
-
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
 static u32 nfsd_attrmask[] = {
@@ -288,6 +258,7 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	struct iattr *iap = &open->op_iattr;
 	struct nfsd_attrs attrs = {
 		.na_iattr	= iap,
+		.na_seclabel	= &open->op_label,
 	};
 	struct dentry *parent, *child;
 	__u32 v_mtime, v_atime;
@@ -409,6 +380,8 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 set_attr:
 	status = nfsd_create_setattr(rqstp, fhp, resfhp, &attrs);
 
+	if (attrs.na_labelerr)
+		open->op_bmval[2] &= ~FATTR4_WORD2_SECURITY_LABEL;
 out:
 	fh_unlock(fhp);
 	if (child && !IS_ERR(child))
@@ -450,9 +423,6 @@ do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stru
 		status = nfsd4_create_file(rqstp, current_fh, *resfh, open);
 		current->fs->umask = 0;
 
-		if (!status && open->op_label.len)
-			nfsd4_security_inode_setsecctx(*resfh, &open->op_label, open->op_bmval);
-
 		/*
 		 * Following rfc 3530 14.2.16, and rfc 5661 18.16.4
 		 * use the returned bitmask to indicate which attributes
@@ -792,6 +762,7 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	struct nfsd4_create *create = &u->create;
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &create->cr_iattr,
+		.na_seclabel	= &create->cr_label,
 	};
 	struct svc_fh resfh;
 	__be32 status;
@@ -864,8 +835,8 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	if (status)
 		goto out;
 
-	if (create->cr_label.len)
-		nfsd4_security_inode_setsecctx(&resfh, &create->cr_label, create->cr_bmval);
+	if (attrs.na_labelerr)
+		create->cr_bmval[2] &= ~FATTR4_WORD2_SECURITY_LABEL;
 
 	if (create->cr_acl != NULL)
 		do_set_nfs4_acl(rqstp, &resfh, create->cr_acl,
@@ -1150,6 +1121,7 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	struct nfsd4_setattr *setattr = &u->setattr;
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &setattr->sa_iattr,
+		.na_seclabel	= &setattr->sa_label,
 	};
 	__be32 status = nfs_ok;
 	int err;
@@ -1178,13 +1150,10 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 					    setattr->sa_acl);
 	if (status)
 		goto out;
-	if (setattr->sa_label.len)
-		status = nfsd4_set_nfs4_label(rqstp, &cstate->current_fh,
-				&setattr->sa_label);
-	if (status)
-		goto out;
 	status = nfsd_setattr(rqstp, &cstate->current_fh, &attrs,
 				0, (time64_t)0);
+	if (!status)
+		status = nfserrno(attrs.na_labelerr);
 out:
 	fh_drop_write(&cstate->current_fh);
 	return status;
fs/nfsd/vfs.c +3 -26
index bfdb42aa23a01..f9f62282d91f8 100644
@@ -459,6 +459,9 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	host_err = notify_change(&init_user_ns, dentry, iap, NULL);
 
 out_unlock:
+	if (attr->na_seclabel && attr->na_seclabel->len)
+		attr->na_labelerr = security_inode_setsecctx(dentry,
+			attr->na_seclabel->data, attr->na_seclabel->len);
 	fh_unlock(fhp);
 	if (size_change)
 		put_write_access(inode);
@@ -497,32 +500,6 @@ int nfsd4_is_junction(struct dentry *dentry)
 		return 0;
 	return 1;
 }
-#ifdef CONFIG_NFSD_V4_SECURITY_LABEL
-__be32 nfsd4_set_nfs4_label(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		struct xdr_netobj *label)
-{
-	__be32 error;
-	int host_error;
-	struct dentry *dentry;
-
-	error = fh_verify(rqstp, fhp, 0 /* S_IFREG */, NFSD_MAY_SATTR);
-	if (error)
-		return error;
-
-	dentry = fhp->fh_dentry;
-
-	inode_lock(d_inode(dentry));
-	host_error = security_inode_setsecctx(dentry, label->data, label->len);
-	inode_unlock(d_inode(dentry));
-	return nfserrno(host_error);
-}
-#else
-__be32 nfsd4_set_nfs4_label(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		struct xdr_netobj *label)
-{
-	return nfserr_notsupp;
-}
-#endif
 
 static struct nfsd4_compound_state *nfsd4_get_cstate(struct svc_rqst *rqstp)
 {
fs/nfsd/vfs.h +3 -2
index 5047cec4c423c..d5d4cfe37c933 100644
@@ -44,6 +44,9 @@ typedef int (*nfsd_filldir_t)(void *, const char *, int, loff_t, u64, unsigned);
 /* nfsd/vfs.c */
 struct nfsd_attrs {
 	struct iattr		*na_iattr;	/* input */
+	struct xdr_netobj	*na_seclabel;	/* input */
+
+	int			na_labelerr;	/* output */
 };
 
 int		nfsd_cross_mnt(struct svc_rqst *rqstp, struct dentry **dpp,
@@ -57,8 +60,6 @@ __be32		nfsd_setattr(struct svc_rqst *, struct svc_fh *,
 				struct nfsd_attrs *, int, time64_t);
 int nfsd_mountpoint(struct dentry *, struct svc_export *);
 #ifdef CONFIG_NFSD_V4
-__be32          nfsd4_set_nfs4_label(struct svc_rqst *, struct svc_fh *,
-		    struct xdr_netobj *);
 __be32		nfsd4_vfs_fallocate(struct svc_rqst *, struct svc_fh *,
 				    struct file *, loff_t, loff_t, int);
 __be32		nfsd4_clone_file_range(struct svc_rqst *rqstp,
Signature · 1 line
2.43.0
author maintainer
#406 lore raw

PATCH 406/690 NFSD: add posix ACLs to struct nfsd_attrs

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit c0cbe70742f4a70893cd6e5f6b10b6e89b6db95b ]

pacl and dpacl pointers are added to struct nfsd_attrs, which requires that we have an nfsd_attrs_free() function to free them. Those nfsv4 functions that can set ACLs now set up these pointers based on the passed in NFSv4 ACL.

nfsd_setattr() sets the acls as appropriate.

Errors are handled as with security labels.

---
 fs/nfsd/acl.h      |  6 ++++--
 fs/nfsd/nfs4acl.c  | 46 +++++++---------------------------------------
 fs/nfsd/nfs4proc.c | 46 ++++++++++++++++------------------------------
 fs/nfsd/vfs.c      |  9 +++++++++
 fs/nfsd/vfs.h      | 11 +++++++++++
 5 files changed, 47 insertions(+), 71 deletions(-)
fs/nfsd/acl.h +4 -2
index ba14d2f4b64f4..4b7324458a94e 100644
@@ -38,6 +38,8 @@
 struct nfs4_acl;
 struct svc_fh;
 struct svc_rqst;
+struct nfsd_attrs;
+enum nfs_ftype4;
 
 int nfs4_acl_bytes(int entries);
 int nfs4_acl_get_whotype(char *, u32);
@@ -45,7 +47,7 @@ __be32 nfs4_acl_write_who(struct xdr_stream *xdr, int who);
 
 int nfsd4_get_nfs4_acl(struct svc_rqst *rqstp, struct dentry *dentry,
 		struct nfs4_acl **acl);
-__be32 nfsd4_set_nfs4_acl(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		struct nfs4_acl *acl);
+__be32 nfsd4_acl_to_attr(enum nfs_ftype4 type, struct nfs4_acl *acl,
+			 struct nfsd_attrs *attr);
 
 #endif /* LINUX_NFS4_ACL_H */
fs/nfsd/nfs4acl.c +7 -39

65 lines not on this page · raw message

fs/nfsd/nfs4proc.c +16 -30

108 lines not on this page · raw message

fs/nfsd/vfs.c +9 -0
index f9f62282d91f8..e91ac3bc68764 100644
@@ -462,6 +462,15 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (attr->na_seclabel && attr->na_seclabel->len)
 		attr->na_labelerr = security_inode_setsecctx(dentry,
 			attr->na_seclabel->data, attr->na_seclabel->len);
+	if (IS_ENABLED(CONFIG_FS_POSIX_ACL) && attr->na_pacl)
+		attr->na_aclerr = set_posix_acl(&init_user_ns,
+						inode, ACL_TYPE_ACCESS,
+						attr->na_pacl);
+	if (IS_ENABLED(CONFIG_FS_POSIX_ACL) &&
+	    !attr->na_aclerr && attr->na_dpacl && S_ISDIR(inode->i_mode))
+		attr->na_aclerr = set_posix_acl(&init_user_ns,
+						inode, ACL_TYPE_DEFAULT,
+						attr->na_dpacl);
 	fh_unlock(fhp);
 	if (size_change)
 		put_write_access(inode);
fs/nfsd/vfs.h +11 -0
index d5d4cfe37c933..c95cd414b4bb0 100644
@@ -6,6 +6,8 @@
 #ifndef LINUX_NFSD_VFS_H
 #define LINUX_NFSD_VFS_H
 
+#include <linux/fs.h>
+#include <linux/posix_acl.h>
 #include "nfsfh.h"
 #include "nfsd.h"
 
@@ -45,10 +47,19 @@ typedef int (*nfsd_filldir_t)(void *, const char *, int, loff_t, u64, unsigned);
 struct nfsd_attrs {
 	struct iattr		*na_iattr;	/* input */
 	struct xdr_netobj	*na_seclabel;	/* input */
+	struct posix_acl	*na_pacl;	/* input */
+	struct posix_acl	*na_dpacl;	/* input */
 
 	int			na_labelerr;	/* output */
+	int			na_aclerr;	/* output */
 };
 
+static inline void nfsd_attrs_free(struct nfsd_attrs *attrs)
+{
+	posix_acl_release(attrs->na_pacl);
+	posix_acl_release(attrs->na_dpacl);
+}
+
 int		nfsd_cross_mnt(struct svc_rqst *rqstp, struct dentry **dpp,
 		                struct svc_export **expp);
 __be32		nfsd_lookup(struct svc_rqst *, struct svc_fh *,
Signature · 1 line
2.43.0
author maintainer
#407 lore raw

PATCH 407/690 NFSD: change nfsd_create()/nfsd_symlink() to unlock directory before returning.

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit 927bfc5600cd6333c9ef9f090f19e66b7d4c8ee1 ]

nfsd_create() usually returns with the directory still locked. nfsd_symlink() usually returns with it unlocked. This is clumsy.

Until recently nfsd_create() needed to keep the directory locked until ACLs and security label had been set. These are now set inside nfsd_create() (in nfsd_setattr()) so this need is gone.

So change nfsd_create() and nfsd_symlink() to always unlock, and remove any fh_unlock() calls that follow calls to these functions.

---
 fs/nfsd/nfs3proc.c |  2 --
 fs/nfsd/nfs4proc.c |  2 --
 fs/nfsd/vfs.c      | 38 +++++++++++++++++++++-----------------
 3 files changed, 21 insertions(+), 21 deletions(-)
fs/nfsd/nfs3proc.c +0 -2
index cb91088bce2e8..c7c2c7db30f54 100644
@@ -388,7 +388,6 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 	fh_init(&resp->fh, NFS3_FHSIZE);
 	resp->status = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len,
 				   &attrs, S_IFDIR, 0, &resp->fh);
-	fh_unlock(&resp->dirfh);
 	return rpc_success;
 }
 
@@ -469,7 +468,6 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 	type = nfs3_ftypes[argp->ftype];
 	resp->status = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len,
 				   &attrs, type, rdev, &resp->fh);
-	fh_unlock(&resp->dirfh);
 out:
 	return rpc_success;
 }
fs/nfsd/nfs4proc.c +0 -2
index ffa2806fd5d3b..fb2487ceac46e 100644
@@ -823,8 +823,6 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		create->cr_bmval[2] &= ~FATTR4_WORD2_SECURITY_LABEL;
 	if (attrs.na_aclerr)
 		create->cr_bmval[0] &= ~FATTR4_WORD0_ACL;
-
-	fh_unlock(&cstate->current_fh);
 	set_change_info(&create->cr_cinfo, &cstate->current_fh);
 	fh_dup2(&cstate->current_fh, &resfh);
 out:
fs/nfsd/vfs.c +21 -17
index e91ac3bc68764..90bd6968fbf68 100644
@@ -1380,8 +1380,10 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	fh_lock_nested(fhp, I_MUTEX_PARENT);
 	dchild = lookup_one_len(fname, dentry, flen);
 	host_err = PTR_ERR(dchild);
-	if (IS_ERR(dchild))
-		return nfserrno(host_err);
+	if (IS_ERR(dchild)) {
+		err = nfserrno(host_err);
+		goto out_unlock;
+	}
 	err = fh_compose(resfhp, fhp->fh_export, dchild, fhp);
 	/*
 	 * We unconditionally drop our ref to dchild as fh_compose will have
@@ -1389,9 +1391,12 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	 */
 	dput(dchild);
 	if (err)
-		return err;
-	return nfsd_create_locked(rqstp, fhp, fname, flen, attrs, type,
-					rdev, resfhp);
+		goto out_unlock;
+	err = nfsd_create_locked(rqstp, fhp, fname, flen, attrs, type,
+				 rdev, resfhp);
+out_unlock:
+	fh_unlock(fhp);
+	return err;
 }
 
 /*
@@ -1468,16 +1473,19 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		goto out;
 
 	host_err = fh_want_write(fhp);
-	if (host_err)
-		goto out_nfserr;
+	if (host_err) {
+		err = nfserrno(host_err);
+		goto out;
+	}
 
 	fh_lock(fhp);
 	dentry = fhp->fh_dentry;
 	dnew = lookup_one_len(fname, dentry, flen);
-	host_err = PTR_ERR(dnew);
-	if (IS_ERR(dnew))
-		goto out_nfserr;
-
+	if (IS_ERR(dnew)) {
+		err = nfserrno(PTR_ERR(dnew));
+		fh_unlock(fhp);
+		goto out_drop_write;
+	}
 	host_err = vfs_symlink(&init_user_ns, d_inode(dentry), dnew, path);
 	err = nfserrno(host_err);
 	cerr = fh_compose(resfhp, fhp->fh_export, dnew, fhp);
@@ -1486,16 +1494,12 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	fh_unlock(fhp);
 	if (!err)
 		err = nfserrno(commit_metadata(fhp));
-	fh_drop_write(fhp);
-
 	dput(dnew);
 	if (err==0) err = cerr;
+out_drop_write:
+	fh_drop_write(fhp);
 out:
 	return err;
-
-out_nfserr:
-	err = nfserrno(host_err);
-	goto out;
 }
 
 /*
Signature · 1 line
2.43.0
author maintainer
#408 lore raw

PATCH 408/690 NFSD: always drop directory lock in nfsd_unlink()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit b677c0c63a135a916493c064906582e9f3ed4802 ]

Some error paths in nfsd_unlink() allow it to exit without unlocking the directory. This is not a problem in practice as the directory will be locked with an fh_put(), but it is untidy and potentially confusing.

This allows us to remove all the fh_unlock() calls that are immediately after nfsd_unlink() calls.

---
 fs/nfsd/nfs3proc.c | 2 --
 fs/nfsd/nfs4proc.c | 4 +---
 fs/nfsd/vfs.c      | 7 +++++--
 3 files changed, 6 insertions(+), 7 deletions(-)
fs/nfsd/nfs3proc.c +0 -2
index c7c2c7db30f54..fbdc109fbd067 100644
@@ -490,7 +490,6 @@ nfsd3_proc_remove(struct svc_rqst *rqstp)
 	fh_copy(&resp->fh, &argp->fh);
 	resp->status = nfsd_unlink(rqstp, &resp->fh, -S_IFDIR,
 				   argp->name, argp->len);
-	fh_unlock(&resp->fh);
 	return rpc_success;
 }
 
@@ -511,7 +510,6 @@ nfsd3_proc_rmdir(struct svc_rqst *rqstp)
 	fh_copy(&resp->fh, &argp->fh);
 	resp->status = nfsd_unlink(rqstp, &resp->fh, S_IFDIR,
 				   argp->name, argp->len);
-	fh_unlock(&resp->fh);
 	return rpc_success;
 }
 
fs/nfsd/nfs4proc.c +1 -3
index fb2487ceac46e..26cd2479e30cf 100644
@@ -1002,10 +1002,8 @@ nfsd4_remove(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		return nfserr_grace;
 	status = nfsd_unlink(rqstp, &cstate->current_fh, 0,
 			     remove->rm_name, remove->rm_namelen);
-	if (!status) {
-		fh_unlock(&cstate->current_fh);
+	if (!status)
 		set_change_info(&remove->rm_cinfo, &cstate->current_fh);
-	}
 	return status;
 }
 
fs/nfsd/vfs.c +5 -2
index 90bd6968fbf68..4b1304fe718fd 100644
@@ -1754,12 +1754,12 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	rdentry = lookup_one_len(fname, dentry, flen);
 	host_err = PTR_ERR(rdentry);
 	if (IS_ERR(rdentry))
-		goto out_drop_write;
+		goto out_unlock;
 
 	if (d_really_is_negative(rdentry)) {
 		dput(rdentry);
 		host_err = -ENOENT;
-		goto out_drop_write;
+		goto out_unlock;
 	}
 	rinode = d_inode(rdentry);
 	ihold(rinode);
@@ -1797,6 +1797,9 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	}
 out:
 	return err;
+out_unlock:
+	fh_unlock(fhp);
+	goto out_drop_write;
 }
 
 /*
Signature · 1 line
2.43.0
author maintainer
#409 lore raw

PATCH 409/690 NFSD: only call fh_unlock() once in nfsd_link()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit e18bcb33bc5b69bccc2b532075aa00bb49cc01c5 ]

On non-error paths, nfsd_link() calls fh_unlock() twice. This is safe because fh_unlock() records that the unlock has been done and doesn't repeat it.
However it makes the code a little confusing and interferes with changes that are planned for directory locking.

So rearrange the code to ensure fh_unlock() is called exactly once if fh_lock() was called.

---
 fs/nfsd/vfs.c | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)
fs/nfsd/vfs.c +10 -9
index 4b1304fe718fd..ac716ced1fd5f 100644
@@ -1542,9 +1542,10 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 	dirp = d_inode(ddir);
 
 	dnew = lookup_one_len(name, ddir, len);
-	host_err = PTR_ERR(dnew);
-	if (IS_ERR(dnew))
-		goto out_nfserr;
+	if (IS_ERR(dnew)) {
+		err = nfserrno(PTR_ERR(dnew));
+		goto out_unlock;
+	}
 
 	dold = tfhp->fh_dentry;
 
@@ -1563,17 +1564,17 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 		else
 			err = nfserrno(host_err);
 	}
-out_dput:
 	dput(dnew);
-out_unlock:
-	fh_unlock(ffhp);
+out_drop_write:
 	fh_drop_write(tfhp);
 out:
 	return err;
 
-out_nfserr:
-	err = nfserrno(host_err);
-	goto out_unlock;
+out_dput:
+	dput(dnew);
+out_unlock:
+	fh_unlock(ffhp);
+	goto out_drop_write;
 }
 
 static void
Signature · 1 line
2.43.0
author maintainer
#410 lore raw

PATCH 410/690 NFSD: reduce locking in nfsd_lookup()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit 19d008b46941b8c668402170522e0f7a9258409c ]

nfsd_lookup() takes an exclusive lock on the parent inode, but no callers want the lock and it may not be needed at all if the result is in the dcache.

Change nfsd_lookup_dentry() to not take the lock, and call lookup_one_len_locked() which takes lock only if needed.

nfsd4_open() currently expects the lock to still be held, but that isn't necessary as nfsd_validate_delegated_dentry() provides required guarantees without the lock.

NOTE: NFSv4 requires directory changeinfo for OPEN even when a create
wasn't requested and no change happened. Now that nfsd_lookup() doesn't use fh_lock(), we need to explicitly fill the attributes when no create happens. A new fh_fill_both_attrs() is provided for that task.

---
 fs/nfsd/nfs4proc.c  | 20 ++++++++++++--------
 fs/nfsd/nfs4state.c |  3 ---
 fs/nfsd/nfsfh.c     | 19 +++++++++++++++++++
 fs/nfsd/nfsfh.h     |  2 +-
 fs/nfsd/vfs.c       | 34 ++++++++++++++--------------------
 5 files changed, 46 insertions(+), 32 deletions(-)
fs/nfsd/nfs4proc.c +12 -8
index 26cd2479e30cf..b6df56fb6755d 100644
@@ -302,6 +302,11 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (d_really_is_positive(child)) {
 		status = nfs_ok;
 
+		/* NFSv4 protocol requires change attributes even though
+		 * no change happened.
+		 */
+		fh_fill_both_attrs(fhp);
+
 		switch (open->op_createmode) {
 		case NFS4_CREATE_UNCHECKED:
 			if (!d_is_reg(child))
@@ -417,15 +422,15 @@ do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stru
 		if (nfsd4_create_is_exclusive(open->op_createmode) && status == 0)
 			open->op_bmval[1] |= (FATTR4_WORD1_TIME_ACCESS |
 						FATTR4_WORD1_TIME_MODIFY);
-	} else
-		/*
-		 * Note this may exit with the parent still locked.
-		 * We will hold the lock until nfsd4_open's final
-		 * lookup, to prevent renames or unlinks until we've had
-		 * a chance to an acquire a delegation if appropriate.
-		 */
+	} else {
 		status = nfsd_lookup(rqstp, current_fh,
 				     open->op_fname, open->op_fnamelen, *resfh);
+		if (!status)
+			/* NFSv4 protocol requires change attributes even though
+			 * no change happened.
+			 */
+			fh_fill_both_attrs(current_fh);
+	}
 	if (status)
 		goto out;
 	status = nfsd_check_obj_isreg(*resfh);
@@ -1043,7 +1048,6 @@ nfsd4_secinfo(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 				    &exp, &dentry);
 	if (err)
 		return err;
-	fh_unlock(&cstate->current_fh);
 	if (d_really_is_negative(dentry)) {
 		exp_put(exp);
 		err = nfserr_noent;
fs/nfsd/nfs4state.c +0 -3
index a299aeaa0de07..f66fb39714893 100644
@@ -5321,9 +5321,6 @@ nfsd4_verify_deleg_dentry(struct nfsd4_open *open, struct nfs4_file *fp,
 	struct dentry *child;
 	__be32 err;
 
-	/* parent may already be locked, and it may get unlocked by
-	 * this call, but that is safe.
-	 */
 	err = nfsd_lookup_dentry(open->op_rqstp, parent,
 				 open->op_fname, open->op_fnamelen,
 				 &exp, &child);
fs/nfsd/nfsfh.c +19 -0
index 5e2ed4b2a925c..cd2946a88d727 100644
@@ -672,6 +672,25 @@ void fh_fill_post_attrs(struct svc_fh *fhp)
 			nfsd4_change_attribute(&fhp->fh_post_attr, inode);
 }
 
+/**
+ * fh_fill_both_attrs - Fill pre-op and post-op attributes
+ * @fhp: file handle to be updated
+ *
+ * This is used when the directory wasn't changed, but wcc attributes
+ * are needed anyway.
+ */
+void fh_fill_both_attrs(struct svc_fh *fhp)
+{
+	fh_fill_post_attrs(fhp);
+	if (!fhp->fh_post_saved)
+		return;
+	fhp->fh_pre_change = fhp->fh_post_change;
+	fhp->fh_pre_mtime = fhp->fh_post_attr.mtime;
+	fhp->fh_pre_ctime = fhp->fh_post_attr.ctime;
+	fhp->fh_pre_size = fhp->fh_post_attr.size;
+	fhp->fh_pre_saved = true;
+}
+
 /*
  * Release a file handle.
  */
fs/nfsd/nfsfh.h +1 -1
index fb9d358a267e5..28a4f9a94e2c8 100644
@@ -322,7 +322,7 @@ static inline u64 nfsd4_change_attribute(struct kstat *stat,
 
 extern void fh_fill_pre_attrs(struct svc_fh *fhp);
 extern void fh_fill_post_attrs(struct svc_fh *fhp);
-
+extern void fh_fill_both_attrs(struct svc_fh *fhp);
 
 /*
  * Lock a file handle/inode
fs/nfsd/vfs.c +14 -20
index ac716ced1fd5f..c07fe50d6bdfb 100644
@@ -198,27 +198,13 @@ nfsd_lookup_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp,
 				goto out_nfserr;
 		}
 	} else {
-		/*
-		 * In the nfsd4_open() case, this may be held across
-		 * subsequent open and delegation acquisition which may
-		 * need to take the child's i_mutex:
-		 */
-		fh_lock_nested(fhp, I_MUTEX_PARENT);
-		dentry = lookup_one_len(name, dparent, len);
+		dentry = lookup_one_len_unlocked(name, dparent, len);
 		host_err = PTR_ERR(dentry);
 		if (IS_ERR(dentry))
 			goto out_nfserr;
 		if (nfsd_mountpoint(dentry, exp)) {
-			/*
-			 * We don't need the i_mutex after all.  It's
-			 * still possible we could open this (regular
-			 * files can be mountpoints too), but the
-			 * i_mutex is just there to prevent renames of
-			 * something that we might be about to delegate,
-			 * and a mountpoint won't be renamed:
-			 */
-			fh_unlock(fhp);
-			if ((host_err = nfsd_cross_mnt(rqstp, &dentry, &exp))) {
+			host_err = nfsd_cross_mnt(rqstp, &dentry, &exp);
+			if (host_err) {
 				dput(dentry);
 				goto out_nfserr;
 			}
@@ -233,7 +219,15 @@ nfsd_lookup_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	return nfserrno(host_err);
 }
 
-/*
+/**
+ * nfsd_lookup - look up a single path component for nfsd
+ *
+ * @rqstp:   the request context
+ * @fhp:     the file handle of the directory
+ * @name:    the component name, or %NULL to look up parent
+ * @len:     length of name to examine
+ * @resfh:   pointer to pre-initialised filehandle to hold result.
+ *
  * Look up one component of a pathname.
  * N.B. After this call _both_ fhp and resfh need an fh_put
  *
@@ -243,11 +237,11 @@ nfsd_lookup_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp,
  * returned. Otherwise the covered directory is returned.
  * NOTE: this mountpoint crossing is not supported properly by all
  *   clients and is explicitly disallowed for NFSv3
- *      NeilBrown <[email protected]>
+ *
  */
 __be32
 nfsd_lookup(struct svc_rqst *rqstp, struct svc_fh *fhp, const char *name,
-				unsigned int len, struct svc_fh *resfh)
+	    unsigned int len, struct svc_fh *resfh)
 {
 	struct svc_export	*exp;
 	struct dentry		*dentry;
Signature · 1 line
2.43.0
author maintainer
#411 lore raw

PATCH 411/690 NFSD: use explicit lock/unlock for directory ops

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit debf16f0c671cb8db154a9ebcd6014cfff683b80 ]

When creating or unlinking a name in a directory use explicit inode_lock_nested() instead of fh_lock(), and explicit calls to fh_fill_pre_attrs() and fh_fill_post_attrs(). This is already done for renames, with lock_rename() as the explicit locking.

Also move the 'fill' calls closer to the operation that might change the attributes. This way they are avoided on some error paths.

For the v2-only code in nfsproc.c, the fill calls are not replaced as they aren't needed.

Making the locking explicit will simplify proposed future changes to locking for directories. It also makes it easily visible exactly where pre/post attributes are used - not all callers of fh_lock() actually need the pre/post attributes.

---
 fs/nfsd/nfs3proc.c |  6 ++++--
 fs/nfsd/nfs4proc.c |  6 ++++--
 fs/nfsd/nfsproc.c  |  5 ++---
 fs/nfsd/vfs.c      | 30 +++++++++++++++++++-----------
 4 files changed, 29 insertions(+), 18 deletions(-)
fs/nfsd/nfs3proc.c +4 -2
index fbdc109fbd067..5b1e771238b35 100644
@@ -260,7 +260,7 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (host_err)
 		return nfserrno(host_err);
 
-	fh_lock_nested(fhp, I_MUTEX_PARENT);
+	inode_lock_nested(inode, I_MUTEX_PARENT);
 
 	child = lookup_one_len(argp->name, parent, argp->len);
 	if (IS_ERR(child)) {
@@ -318,11 +318,13 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (!IS_POSIXACL(inode))
 		iap->ia_mode &= ~current_umask();
 
+	fh_fill_pre_attrs(fhp);
 	host_err = vfs_create(&init_user_ns, inode, child, iap->ia_mode, true);
 	if (host_err < 0) {
 		status = nfserrno(host_err);
 		goto out;
 	}
+	fh_fill_post_attrs(fhp);
 
 	/* A newly created file already has a file size of zero. */
 	if ((iap->ia_valid & ATTR_SIZE) && (iap->ia_size == 0))
@@ -340,7 +342,7 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	status = nfsd_create_setattr(rqstp, fhp, resfhp, &attrs);
 
 out:
-	fh_unlock(fhp);
+	inode_unlock(inode);
 	if (child && !IS_ERR(child))
 		dput(child);
 	fh_drop_write(fhp);
fs/nfsd/nfs4proc.c +4 -2
index b6df56fb6755d..5e4b7858b2e50 100644
@@ -264,7 +264,7 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (is_create_with_attrs(open))
 		nfsd4_acl_to_attr(NF4REG, open->op_acl, &attrs);
 
-	fh_lock_nested(fhp, I_MUTEX_PARENT);
+	inode_lock_nested(inode, I_MUTEX_PARENT);
 
 	child = lookup_one_len(open->op_fname, parent, open->op_fnamelen);
 	if (IS_ERR(child)) {
@@ -348,10 +348,12 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (!IS_POSIXACL(inode))
 		iap->ia_mode &= ~current_umask();
 
+	fh_fill_pre_attrs(fhp);
 	status = nfsd4_vfs_create(fhp, child, open);
 	if (status != nfs_ok)
 		goto out;
 	open->op_created = true;
+	fh_fill_post_attrs(fhp);
 
 	/* A newly created file already has a file size of zero. */
 	if ((iap->ia_valid & ATTR_SIZE) && (iap->ia_size == 0))
@@ -373,7 +375,7 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (attrs.na_aclerr)
 		open->op_bmval[0] &= ~FATTR4_WORD0_ACL;
 out:
-	fh_unlock(fhp);
+	inode_unlock(inode);
 	nfsd_attrs_free(&attrs);
 	if (child && !IS_ERR(child))
 		dput(child);
fs/nfsd/nfsproc.c +2 -3
index 09afd188099be..4b19cc727ea50 100644
@@ -292,7 +292,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 		goto done;
 	}
 
-	fh_lock_nested(dirfhp, I_MUTEX_PARENT);
+	inode_lock_nested(dirfhp->fh_dentry->d_inode, I_MUTEX_PARENT);
 	dchild = lookup_one_len(argp->name, dirfhp->fh_dentry, argp->len);
 	if (IS_ERR(dchild)) {
 		resp->status = nfserrno(PTR_ERR(dchild));
@@ -408,8 +408,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	}
 
 out_unlock:
-	/* We don't really need to unlock, as fh_put does it. */
-	fh_unlock(dirfhp);
+	inode_unlock(dirfhp->fh_dentry->d_inode);
 	fh_drop_write(dirfhp);
 done:
 	fh_put(dirfhp);
fs/nfsd/vfs.c +19 -11
index c07fe50d6bdfb..7de76b37a9bc2 100644
@@ -1371,7 +1371,7 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (host_err)
 		return nfserrno(host_err);
 
-	fh_lock_nested(fhp, I_MUTEX_PARENT);
+	inode_lock_nested(dentry->d_inode, I_MUTEX_PARENT);
 	dchild = lookup_one_len(fname, dentry, flen);
 	host_err = PTR_ERR(dchild);
 	if (IS_ERR(dchild)) {
@@ -1386,10 +1386,12 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	dput(dchild);
 	if (err)
 		goto out_unlock;
+	fh_fill_pre_attrs(fhp);
 	err = nfsd_create_locked(rqstp, fhp, fname, flen, attrs, type,
 				 rdev, resfhp);
+	fh_fill_post_attrs(fhp);
 out_unlock:
-	fh_unlock(fhp);
+	inode_unlock(dentry->d_inode);
 	return err;
 }
 
@@ -1472,20 +1474,22 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		goto out;
 	}
 
-	fh_lock(fhp);
 	dentry = fhp->fh_dentry;
+	inode_lock_nested(dentry->d_inode, I_MUTEX_PARENT);
 	dnew = lookup_one_len(fname, dentry, flen);
 	if (IS_ERR(dnew)) {
 		err = nfserrno(PTR_ERR(dnew));
-		fh_unlock(fhp);
+		inode_unlock(dentry->d_inode);
 		goto out_drop_write;
 	}
+	fh_fill_pre_attrs(fhp);
 	host_err = vfs_symlink(&init_user_ns, d_inode(dentry), dnew, path);
 	err = nfserrno(host_err);
 	cerr = fh_compose(resfhp, fhp->fh_export, dnew, fhp);
 	if (!err)
 		nfsd_create_setattr(rqstp, fhp, resfhp, attrs);
-	fh_unlock(fhp);
+	fh_fill_post_attrs(fhp);
+	inode_unlock(dentry->d_inode);
 	if (!err)
 		err = nfserrno(commit_metadata(fhp));
 	dput(dnew);
@@ -1531,9 +1535,9 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 		goto out;
 	}
 
-	fh_lock_nested(ffhp, I_MUTEX_PARENT);
 	ddir = ffhp->fh_dentry;
 	dirp = d_inode(ddir);
+	inode_lock_nested(dirp, I_MUTEX_PARENT);
 
 	dnew = lookup_one_len(name, ddir, len);
 	if (IS_ERR(dnew)) {
@@ -1546,8 +1550,10 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 	err = nfserr_noent;
 	if (d_really_is_negative(dold))
 		goto out_dput;
+	fh_fill_pre_attrs(ffhp);
 	host_err = vfs_link(dold, &init_user_ns, dirp, dnew, NULL);
-	fh_unlock(ffhp);
+	fh_fill_post_attrs(ffhp);
+	inode_unlock(dirp);
 	if (!host_err) {
 		err = nfserrno(commit_metadata(ffhp));
 		if (!err)
@@ -1567,7 +1573,7 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 out_dput:
 	dput(dnew);
 out_unlock:
-	fh_unlock(ffhp);
+	inode_unlock(dirp);
 	goto out_drop_write;
 }
 
@@ -1742,9 +1748,9 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	if (host_err)
 		goto out_nfserr;
 
-	fh_lock_nested(fhp, I_MUTEX_PARENT);
 	dentry = fhp->fh_dentry;
 	dirp = d_inode(dentry);
+	inode_lock_nested(dirp, I_MUTEX_PARENT);
 
 	rdentry = lookup_one_len(fname, dentry, flen);
 	host_err = PTR_ERR(rdentry);
@@ -1762,6 +1768,7 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	if (!type)
 		type = d_inode(rdentry)->i_mode & S_IFMT;
 
+	fh_fill_pre_attrs(fhp);
 	if (type != S_IFDIR) {
 		if (rdentry->d_sb->s_export_op->flags & EXPORT_OP_CLOSE_BEFORE_UNLINK)
 			nfsd_close_cached_files(rdentry);
@@ -1769,8 +1776,9 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	} else {
 		host_err = vfs_rmdir(&init_user_ns, dirp, rdentry);
 	}
+	fh_fill_post_attrs(fhp);
 
-	fh_unlock(fhp);
+	inode_unlock(dirp);
 	if (!host_err)
 		host_err = commit_metadata(fhp);
 	dput(rdentry);
@@ -1793,7 +1801,7 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 out:
 	return err;
 out_unlock:
-	fh_unlock(fhp);
+	inode_unlock(dirp);
 	goto out_drop_write;
 }
 
Signature · 1 line
2.43.0
author maintainer
#412 lore raw

PATCH 412/690 NFSD: use (un)lock_inode instead of fh_(un)lock for file operations

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit bb4d53d66e4b8c8b8e5634802262e53851a2d2db ]

When locking a file to access ACLs and xattrs etc, use explicit locking with inode_lock() instead of fh_lock(). This means that the calls to fh_fill_pre/post_attr() are also explicit which improves readability and allows us to place them only where they are needed. Only the xattr calls need pre/post information.

When locking a file we don't need I_MUTEX_PARENT as the file is not a parent of anything, so we can use inode_lock() directly rather than the inode_lock_nested() call that fh_lock() uses.

---
 fs/nfsd/nfs2acl.c   |  6 +++---
 fs/nfsd/nfs3acl.c   |  4 ++--
 fs/nfsd/nfs4state.c |  9 +++++----
 fs/nfsd/vfs.c       | 34 ++++++++++++++++++++--------------
 4 files changed, 30 insertions(+), 23 deletions(-)
fs/nfsd/nfs2acl.c +3 -3
index efcd429b0f28e..87f224cd30a85 100644
@@ -111,7 +111,7 @@ static __be32 nfsacld_proc_setacl(struct svc_rqst *rqstp)
 	if (error)
 		goto out_errno;
 
-	fh_lock(fh);
+	inode_lock(inode);
 
 	error = set_posix_acl(&init_user_ns, inode, ACL_TYPE_ACCESS,
 			      argp->acl_access);
@@ -122,7 +122,7 @@ static __be32 nfsacld_proc_setacl(struct svc_rqst *rqstp)
 	if (error)
 		goto out_drop_lock;
 
-	fh_unlock(fh);
+	inode_unlock(inode);
 
 	fh_drop_write(fh);
 
@@ -136,7 +136,7 @@ static __be32 nfsacld_proc_setacl(struct svc_rqst *rqstp)
 	return rpc_success;
 
 out_drop_lock:
-	fh_unlock(fh);
+	inode_unlock(inode);
 	fh_drop_write(fh);
 out_errno:
 	resp->status = nfserrno(error);
fs/nfsd/nfs3acl.c +2 -2
index 35b2ebda14dac..9446c67436649 100644
@@ -101,7 +101,7 @@ static __be32 nfsd3_proc_setacl(struct svc_rqst *rqstp)
 	if (error)
 		goto out_errno;
 
-	fh_lock(fh);
+	inode_lock(inode);
 
 	error = set_posix_acl(&init_user_ns, inode, ACL_TYPE_ACCESS,
 			      argp->acl_access);
@@ -111,7 +111,7 @@ static __be32 nfsd3_proc_setacl(struct svc_rqst *rqstp)
 			      argp->acl_default);
 
 out_drop_lock:
-	fh_unlock(fh);
+	inode_unlock(inode);
 	fh_drop_write(fh);
 out_errno:
 	resp->status = nfserrno(error);
fs/nfsd/nfs4state.c +5 -4
index f66fb39714893..66cf8217ebe57 100644
@@ -7415,21 +7415,22 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 static __be32 nfsd_test_lock(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file_lock *lock)
 {
 	struct nfsd_file *nf;
+	struct inode *inode;
 	__be32 err;
 
 	err = nfsd_file_acquire(rqstp, fhp, NFSD_MAY_READ, &nf);
 	if (err)
 		return err;
-	fh_lock(fhp); /* to block new leases till after test_lock: */
-	err = nfserrno(nfsd_open_break_lease(fhp->fh_dentry->d_inode,
-							NFSD_MAY_READ));
+	inode = fhp->fh_dentry->d_inode;
+	inode_lock(inode); /* to block new leases till after test_lock: */
+	err = nfserrno(nfsd_open_break_lease(inode, NFSD_MAY_READ));
 	if (err)
 		goto out;
 	lock->fl_file = nf->nf_file;
 	err = nfserrno(vfs_test_lock(nf->nf_file, lock));
 	lock->fl_file = NULL;
 out:
-	fh_unlock(fhp);
+	inode_unlock(inode);
 	nfsd_file_put(nf);
 	return err;
 }
fs/nfsd/vfs.c +20 -14
index 7de76b37a9bc2..73a153be6a5ad 100644
@@ -417,7 +417,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 			return err;
 	}
 
-	fh_lock(fhp);
+	inode_lock(inode);
 	if (size_change) {
 		/*
 		 * RFC5661, Section 18.30.4:
@@ -465,7 +465,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		attr->na_aclerr = set_posix_acl(&init_user_ns,
 						inode, ACL_TYPE_DEFAULT,
 						attr->na_dpacl);
-	fh_unlock(fhp);
+	inode_unlock(inode);
 	if (size_change)
 		put_write_access(inode);
 out:
@@ -2156,13 +2156,16 @@ nfsd_listxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char **bufp,
 	return err;
 }
 
-/*
- * Removexattr and setxattr need to call fh_lock to both lock the inode
- * and set the change attribute. Since the top-level vfs_removexattr
- * and vfs_setxattr calls already do their own inode_lock calls, call
- * the _locked variant. Pass in a NULL pointer for delegated_inode,
- * and let the client deal with NFS4ERR_DELAY (same as with e.g.
- * setattr and remove).
+/**
+ * nfsd_removexattr - Remove an extended attribute
+ * @rqstp: RPC transaction being executed
+ * @fhp: NFS filehandle of object with xattr to remove
+ * @name: name of xattr to remove (NUL-terminate)
+ *
+ * Pass in a NULL pointer for delegated_inode, and let the client deal
+ * with NFS4ERR_DELAY (same as with e.g. setattr and remove).
+ *
+ * Returns nfs_ok on success, or an nfsstat in network byte order.
  */
 __be32
 nfsd_removexattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name)
@@ -2178,12 +2181,14 @@ nfsd_removexattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name)
 	if (ret)
 		return nfserrno(ret);
 
-	fh_lock(fhp);
+	inode_lock(fhp->fh_dentry->d_inode);
+	fh_fill_pre_attrs(fhp);
 
 	ret = __vfs_removexattr_locked(&init_user_ns, fhp->fh_dentry,
 				       name, NULL);
 
-	fh_unlock(fhp);
+	fh_fill_post_attrs(fhp);
+	inode_unlock(fhp->fh_dentry->d_inode);
 	fh_drop_write(fhp);
 
 	return nfsd_xattr_errno(ret);
@@ -2203,12 +2208,13 @@ nfsd_setxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name,
 	ret = fh_want_write(fhp);
 	if (ret)
 		return nfserrno(ret);
-	fh_lock(fhp);
+	inode_lock(fhp->fh_dentry->d_inode);
+	fh_fill_pre_attrs(fhp);
 
 	ret = __vfs_setxattr_locked(&init_user_ns, fhp->fh_dentry, name, buf,
 				    len, flags, NULL);
-
-	fh_unlock(fhp);
+	fh_fill_post_attrs(fhp);
+	inode_unlock(fhp->fh_dentry->d_inode);
 	fh_drop_write(fhp);
 
 	return nfsd_xattr_errno(ret);
Signature · 1 line
2.43.0
author maintainer
#413 lore raw

PATCH 413/690 NFSD: discard fh_locked flag and fh_lock/fh_unlock

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit dd8dd403d7b223cc77ee89d8d09caf045e90e648 ]

As all inode locking is now fully balanced, fh_put() does not need to call fh_unlock().
fh_lock() and fh_unlock() are no longer used, so discard them. These are the only real users of ->fh_locked, so discard that too.

---
 fs/nfsd/nfsfh.c |  3 +--
 fs/nfsd/nfsfh.h | 56 ++++---------------------------------------------
 fs/nfsd/vfs.c   | 17 +--------------
 3 files changed, 6 insertions(+), 70 deletions(-)
fs/nfsd/nfsfh.c +1 -2
index cd2946a88d727..a5b71526cee0f 100644
@@ -549,7 +549,7 @@ fh_compose(struct svc_fh *fhp, struct svc_export *exp, struct dentry *dentry,
 	if (ref_fh == fhp)
 		fh_put(ref_fh);
 
-	if (fhp->fh_locked || fhp->fh_dentry) {
+	if (fhp->fh_dentry) {
 		printk(KERN_ERR "fh_compose: fh %pd2 not initialized!\n",
 		       dentry);
 	}
@@ -700,7 +700,6 @@ fh_put(struct svc_fh *fhp)
 	struct dentry * dentry = fhp->fh_dentry;
 	struct svc_export * exp = fhp->fh_export;
 	if (dentry) {
-		fh_unlock(fhp);
 		fhp->fh_dentry = NULL;
 		dput(dentry);
 		fh_clear_pre_post_attrs(fhp);
fs/nfsd/nfsfh.h +4 -52

84 lines not on this page · raw message

fs/nfsd/vfs.c +1 -16
index 73a153be6a5ad..66d4a126f20ab 100644
@@ -1266,13 +1266,6 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	dirp = d_inode(dentry);
 
 	dchild = dget(resfhp->fh_dentry);
-	if (!fhp->fh_locked) {
-		WARN_ONCE(1, "nfsd_create: parent %pd2 not locked!\n",
-				dentry);
-		err = nfserr_io;
-		goto out;
-	}
-
 	err = nfsd_permission(rqstp, fhp->fh_export, dentry, NFSD_MAY_CREATE);
 	if (err)
 		goto out;
@@ -1641,10 +1634,7 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
 		goto out;
 	}
 
-	/* cannot use fh_lock as we need deadlock protective ordering
-	 * so do it by hand */
 	trap = lock_rename(tdentry, fdentry);
-	ffhp->fh_locked = tfhp->fh_locked = true;
 	fh_fill_pre_attrs(ffhp);
 	fh_fill_pre_attrs(tfhp);
 
@@ -1694,17 +1684,12 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
 	dput(odentry);
  out_nfserr:
 	err = nfserrno(host_err);
-	/*
-	 * We cannot rely on fh_unlock on the two filehandles,
-	 * as that would do the wrong thing if the two directories
-	 * were the same, so again we do it by hand.
-	 */
+
 	if (!close_cached) {
 		fh_fill_post_attrs(ffhp);
 		fh_fill_post_attrs(tfhp);
 	}
 	unlock_rename(tdentry, fdentry);
-	ffhp->fh_locked = tfhp->fh_locked = false;
 	fh_drop_write(ffhp);
 
 	/*
Signature · 1 line
2.43.0
author maintainer
#414 lore raw

PATCH 414/690 NFSD: fix regression with setting ACLs.

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit 00801cd92d91e94aa04d687f9bb9a9104e7c3d46 ]

A recent patch moved ACL setting into nfsd_setattr(). Unfortunately it didn't work as nfsd_setattr() aborts early if iap->ia_valid is 0.

Remove this test, and instead avoid calling notify_change() when ia_valid is 0.

This means that nfsd_setattr() will now *always* lock the inode. Previously it didn't if only a ATTR_MODE change was requested on a symlink (see Commit 15b7a1b86d66 ("[PATCH] knfsd: fix setattr-on-symlink error return")). I don't think this change really matters.

Fixes c0cbe70742f4 ("NFSD: add posix ACLs to struct nfsd_attrs")
---
 fs/nfsd/vfs.c | 19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)
fs/nfsd/vfs.c +9 -10
index 66d4a126f20ab..ad689215b1f37 100644
@@ -299,6 +299,10 @@ commit_metadata(struct svc_fh *fhp)
 static void
 nfsd_sanitize_attrs(struct inode *inode, struct iattr *iap)
 {
+	/* Ignore mode updates on symlinks */
+	if (S_ISLNK(inode->i_mode))
+		iap->ia_valid &= ~ATTR_MODE;
+
 	/* sanitize the mode change */
 	if (iap->ia_valid & ATTR_MODE) {
 		iap->ia_mode &= S_IALLUGO;
@@ -354,7 +358,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	int		accmode = NFSD_MAY_SATTR;
 	umode_t		ftype = 0;
 	__be32		err;
-	int		host_err;
+	int		host_err = 0;
 	bool		get_write_count;
 	bool		size_change = (iap->ia_valid & ATTR_SIZE);
 
@@ -392,13 +396,6 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	dentry = fhp->fh_dentry;
 	inode = d_inode(dentry);
 
-	/* Ignore any mode updates on symlinks */
-	if (S_ISLNK(inode->i_mode))
-		iap->ia_valid &= ~ATTR_MODE;
-
-	if (!iap->ia_valid)
-		return 0;
-
 	nfsd_sanitize_attrs(inode, iap);
 
 	if (check_guard && guardtime != inode->i_ctime.tv_sec)
@@ -449,8 +446,10 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 			goto out_unlock;
 	}
 
-	iap->ia_valid |= ATTR_CTIME;
-	host_err = notify_change(&init_user_ns, dentry, iap, NULL);
+	if (iap->ia_valid) {
+		iap->ia_valid |= ATTR_CTIME;
+		host_err = notify_change(&init_user_ns, dentry, iap, NULL);
+	}
 
 out_unlock:
 	if (attr->na_seclabel && attr->na_seclabel->len)
Signature · 1 line
2.43.0
author maintainer
#415 lore raw

PATCH 415/690 nfsd_splice_actor(): handle compound pages

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Al Viro <[email protected]>

[ Upstream commit bfbfb6182ad1d7d184b16f25165faad879147f79 ]

pipe_buffer might refer to a compound page (and contain more than a PAGE_SIZE worth of data). Theoretically it had been possible since way back, but nfsd_splice_actor() hadn't run into that until copy_page_to_iter() change. Fortunately, the only thing that changes for compound pages is that we need to stuff each relevant subpage in and convert the offset into offset in the first subpage.

Fixes f0f6b614f83d "copy_page_to_iter(): don't split high-order page in case of ITER_PIPE"
---
 fs/nfsd/vfs.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)
fs/nfsd/vfs.c +9 -4
index ad689215b1f37..343af6341e5e1 100644
@@ -846,10 +846,15 @@ nfsd_splice_actor(struct pipe_inode_info *pipe, struct pipe_buffer *buf,
 		  struct splice_desc *sd)
 {
 	struct svc_rqst *rqstp = sd->u.data;
-
-	svc_rqst_replace_page(rqstp, buf->page);
-	if (rqstp->rq_res.page_len == 0)
-		rqstp->rq_res.page_base = buf->offset;
+	struct page *page = buf->page;	// may be a compound one
+	unsigned offset = buf->offset;
+	int i;
+
+	page += offset / PAGE_SIZE;
+	for (i = sd->len; i > 0; i -= PAGE_SIZE)
+		svc_rqst_replace_page(rqstp, page++);
+	if (rqstp->rq_res.page_len == 0)	// first call
+		rqstp->rq_res.page_base = offset % PAGE_SIZE;
 	rqstp->rq_res.page_len += sd->len;
 	return sd->len;
 }
Signature · 1 line
2.43.0
author maintainer
#416 lore raw

PATCH 416/690 NFSD: move from strlcpy with unused retval to strscpy

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Wolfram Sang <[email protected]>

[ Upstream commit 72f78ae00a8e5d7abe13abac8305a300f6afd74b ]

Follow the advice of the below link and prefer 'strscpy' in this subsystem. Conversion is 1:1 because the return value is not used. Generated by a coccinelle script.

---
 fs/nfsd/nfs4idmap.c | 8 ++++----
 fs/nfsd/nfs4proc.c  | 2 +-
 fs/nfsd/nfssvc.c    | 2 +-
 3 files changed, 6 insertions(+), 6 deletions(-)
fs/nfsd/nfs4idmap.c +4 -4
index f92161ce1f97d..e70a1a2999b7b 100644
@@ -82,8 +82,8 @@ ent_init(struct cache_head *cnew, struct cache_head *citm)
 	new->id = itm->id;
 	new->type = itm->type;
 
-	strlcpy(new->name, itm->name, sizeof(new->name));
-	strlcpy(new->authname, itm->authname, sizeof(new->authname));
+	strscpy(new->name, itm->name, sizeof(new->name));
+	strscpy(new->authname, itm->authname, sizeof(new->authname));
 }
 
 static void
@@ -548,7 +548,7 @@ idmap_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen
 		return nfserr_badowner;
 	memcpy(key.name, name, namelen);
 	key.name[namelen] = '\0';
-	strlcpy(key.authname, rqst_authname(rqstp), sizeof(key.authname));
+	strscpy(key.authname, rqst_authname(rqstp), sizeof(key.authname));
 	ret = idmap_lookup(rqstp, nametoid_lookup, &key, nn->nametoid_cache, &item);
 	if (ret == -ENOENT)
 		return nfserr_badowner;
@@ -584,7 +584,7 @@ static __be32 idmap_id_to_name(struct xdr_stream *xdr,
 	int ret;
 	struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id);
 
-	strlcpy(key.authname, rqst_authname(rqstp), sizeof(key.authname));
+	strscpy(key.authname, rqst_authname(rqstp), sizeof(key.authname));
 	ret = idmap_lookup(rqstp, idtoname_lookup, &key, nn->idtoname_cache, &item);
 	if (ret == -ENOENT)
 		return encode_ascii_id(xdr, id);
fs/nfsd/nfs4proc.c +1 -1
index 5e4b7858b2e50..b2e6fa962f7d9 100644
@@ -1346,7 +1346,7 @@ static __be32 nfsd4_ssc_setup_dul(struct nfsd_net *nn, char *ipaddr,
 		return 0;
 	}
 	if (work) {
-		strlcpy(work->nsui_ipaddr, ipaddr, sizeof(work->nsui_ipaddr) - 1);
+		strscpy(work->nsui_ipaddr, ipaddr, sizeof(work->nsui_ipaddr) - 1);
 		refcount_set(&work->nsui_refcnt, 2);
 		work->nsui_busy = true;
 		list_add_tail(&work->nsui_list, &nn->nfsd_ssc_mount_list);
fs/nfsd/nfssvc.c +1 -1
index 011c556caa1e7..8b1afde192118 100644
@@ -799,7 +799,7 @@ nfsd_svc(int nrservs, struct net *net, const struct cred *cred)
 	if (nrservs == 0 && nn->nfsd_serv == NULL)
 		goto out;
 
-	strlcpy(nn->nfsd_name, utsname()->nodename,
+	strscpy(nn->nfsd_name, utsname()->nodename,
 		sizeof(nn->nfsd_name));
 
 	error = nfsd_create_serv(net);
Signature · 1 line
2.43.0
author maintainer
#417 lore raw

PATCH 417/690 lockd: move from strlcpy with unused retval to strscpy

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Wolfram Sang <[email protected]>

[ Upstream commit 97f8e62572555f8ad578d7b1739ba64d5d2cac0f ]

Follow the advice of the below link and prefer 'strscpy' in this subsystem. Conversion is 1:1 because the return value is not used. Generated by a coccinelle script.

---
 fs/lockd/host.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/lockd/host.c +1 -1
index f802223e71abe..cdc8e12cdac44 100644
@@ -164,7 +164,7 @@ static struct nlm_host *nlm_alloc_host(struct nlm_lookup_host_info *ni,
 	host->h_addrbuf    = nsm->sm_addrbuf;
 	host->net	   = ni->net;
 	host->h_cred	   = get_cred(ni->cred);
-	strlcpy(host->nodename, utsname()->nodename, sizeof(host->nodename));
+	strscpy(host->nodename, utsname()->nodename, sizeof(host->nodename));
 
 out:
 	return host;
Signature · 1 line
2.43.0
author maintainer
#418 lore raw

PATCH 418/690 NFSD enforce filehandle check for source file in COPY

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Olga Kornievskaia <[email protected]>

[ Upstream commit 754035ff79a14886e68c0c9f6fa80adb21f12b53 ]

If the passed in filehandle for the source file in the COPY operation is not a regular file, the server MUST return NFS4ERR_WRONG_TYPE.

[ cel: adjusted to apply to v5.15.y ]
---
 fs/nfsd/nfs4proc.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
fs/nfsd/nfs4proc.c +7 -1
index b2e6fa962f7d9..b2bfe540c1cb0 100644
@@ -1760,7 +1760,13 @@ static int nfsd4_do_async_copy(void *data)
 		filp = nfs42_ssc_open(copy->ss_mnt, &copy->c_fh,
 				      &copy->stateid);
 		if (IS_ERR(filp)) {
-			nfserr = nfserr_offload_denied;
+			switch (PTR_ERR(filp)) {
+			case -EBADF:
+				nfserr = nfserr_wrong_type;
+				break;
+			default:
+				nfserr = nfserr_offload_denied;
+			}
 			/* ss_mnt will be unmounted by the laundromat */
 			goto do_callback;
 		}
Signature · 1 line
2.43.0
author maintainer
#419 lore raw

PATCH 419/690 NFSD: remove redundant variable status

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jinpeng Cui <[email protected]>

[ Upstream commit 4ab3442ca384a02abf8b1f2b3449a6c547851873 ]

Return value directly from fh_verify() do_open_permission() exp_pseudoroot() instead of getting value from redundant variable status.

---
 fs/nfsd/nfs4proc.c | 16 ++++------------
 1 file changed, 4 insertions(+), 12 deletions(-)
fs/nfsd/nfs4proc.c +4 -12
index b2bfe540c1cb0..69d3013fb1b26 100644
@@ -141,7 +141,6 @@ fh_dup2(struct svc_fh *dst, struct svc_fh *src)
 static __be32
 do_open_permission(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_open *open, int accmode)
 {
-	__be32 status;
 
 	if (open->op_truncate &&
 		!(open->op_share_access & NFS4_SHARE_ACCESS_WRITE))
@@ -156,9 +155,7 @@ do_open_permission(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfs
 	if (open->op_share_deny & NFS4_SHARE_DENY_READ)
 		accmode |= NFSD_MAY_WRITE;
 
-	status = fh_verify(rqstp, current_fh, S_IFREG, accmode);
-
-	return status;
+	return fh_verify(rqstp, current_fh, S_IFREG, accmode);
 }
 
 static __be32 nfsd_check_obj_isreg(struct svc_fh *fh)
@@ -454,7 +451,6 @@ static __be32
 do_open_fhandle(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_open *open)
 {
 	struct svc_fh *current_fh = &cstate->current_fh;
-	__be32 status;
 	int accmode = 0;
 
 	/* We don't know the target directory, and therefore can not
@@ -479,9 +475,7 @@ do_open_fhandle(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, str
 	if (open->op_claim_type == NFS4_OPEN_CLAIM_DELEG_CUR_FH)
 		accmode = NFSD_MAY_OWNER_OVERRIDE;
 
-	status = do_open_permission(rqstp, current_fh, open, accmode);
-
-	return status;
+	return do_open_permission(rqstp, current_fh, open, accmode);
 }
 
 static void
@@ -668,11 +662,9 @@ static __be32
 nfsd4_putrootfh(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		union nfsd4_op_u *u)
 {
-	__be32 status;
-
 	fh_put(&cstate->current_fh);
-	status = exp_pseudoroot(rqstp, &cstate->current_fh);
-	return status;
+
+	return exp_pseudoroot(rqstp, &cstate->current_fh);
 }
 
 static __be32
Signature · 1 line
2.43.0
author maintainer
#420 lore raw

PATCH 420/690 nfsd: Avoid some useless tests

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Christophe JAILLET <[email protected]>

[ Upstream commit d44899b8bb0b919f923186c616a84f0e70e04772 ]

memdup_user() can't return NULL, so there is no point for checking for it.

Simplify some tests accordingly.

---
 fs/nfsd/nfs4recover.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
fs/nfsd/nfs4recover.c +3 -3
index 8f24485e0f04f..4edfc95806412 100644
@@ -807,7 +807,7 @@ __cld_pipe_inprogress_downcall(const struct cld_msg_v2 __user *cmsg,
 			if (get_user(namelen, &ci->cc_name.cn_len))
 				return -EFAULT;
 			name.data = memdup_user(&ci->cc_name.cn_id, namelen);
-			if (IS_ERR_OR_NULL(name.data))
+			if (IS_ERR(name.data))
 				return -EFAULT;
 			name.len = namelen;
 			get_user(princhashlen, &ci->cc_princhash.cp_len);
@@ -815,7 +815,7 @@ __cld_pipe_inprogress_downcall(const struct cld_msg_v2 __user *cmsg,
 				princhash.data = memdup_user(
 						&ci->cc_princhash.cp_data,
 						princhashlen);
-				if (IS_ERR_OR_NULL(princhash.data)) {
+				if (IS_ERR(princhash.data)) {
 					kfree(name.data);
 					return -EFAULT;
 				}
@@ -829,7 +829,7 @@ __cld_pipe_inprogress_downcall(const struct cld_msg_v2 __user *cmsg,
 			if (get_user(namelen, &cnm->cn_len))
 				return -EFAULT;
 			name.data = memdup_user(&cnm->cn_id, namelen);
-			if (IS_ERR_OR_NULL(name.data))
+			if (IS_ERR(name.data))
 				return -EFAULT;
 			name.len = namelen;
 		}
Signature · 1 line
2.43.0
author maintainer
#421 lore raw

PATCH 421/690 nfsd: Propagate some error code returned by memdup_user()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Christophe JAILLET <[email protected]>

[ Upstream commit 30a30fcc3fc1ad4c5d017c9fcb75dc8f59e7bdad ]

Propagate the error code returned by memdup_user() instead of a hard coded -EFAULT.

---
 fs/nfsd/nfs4recover.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
fs/nfsd/nfs4recover.c +3 -3
index 4edfc95806412..5d680045fa2c7 100644
@@ -808,7 +808,7 @@ __cld_pipe_inprogress_downcall(const struct cld_msg_v2 __user *cmsg,
 				return -EFAULT;
 			name.data = memdup_user(&ci->cc_name.cn_id, namelen);
 			if (IS_ERR(name.data))
-				return -EFAULT;
+				return PTR_ERR(name.data);
 			name.len = namelen;
 			get_user(princhashlen, &ci->cc_princhash.cp_len);
 			if (princhashlen > 0) {
@@ -817,7 +817,7 @@ __cld_pipe_inprogress_downcall(const struct cld_msg_v2 __user *cmsg,
 						princhashlen);
 				if (IS_ERR(princhash.data)) {
 					kfree(name.data);
-					return -EFAULT;
+					return PTR_ERR(princhash.data);
 				}
 				princhash.len = princhashlen;
 			} else
@@ -830,7 +830,7 @@ __cld_pipe_inprogress_downcall(const struct cld_msg_v2 __user *cmsg,
 				return -EFAULT;
 			name.data = memdup_user(&cnm->cn_id, namelen);
 			if (IS_ERR(name.data))
-				return -EFAULT;
+				return PTR_ERR(name.data);
 			name.len = namelen;
 		}
 		if (name.len > 5 && memcmp(name.data, "hash:", 5) == 0) {
Signature · 1 line
2.43.0
author maintainer
#422 lore raw

PATCH 422/690 NFSD: Increase NFSD_MAX_OPS_PER_COMPOUND

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 80e591ce636f3ae6855a0ca26963da1fdd6d4508 ]

When attempting an NFSv4 mount, a Solaris NFSv4 client builds a single large COMPOUND that chains a series of LOOKUPs to get to the pseudo filesystem root directory that is to be mounted. The Linux NFS server's current maximum of 16 operations per NFSv4 COMPOUND is not large enough to ensure that this works for paths that are more than a few components deep.

Since NFSD_MAX_OPS_PER_COMPOUND is mostly a sanity check, and most NFSv4 COMPOUNDS are between 3 and 6 operations (thus they do not trigger any re-allocation of the operation array on the server), increasing this maximum should result in little to no impact.

The ops array can get large now, so allocate it via vmalloc() to help ensure memory fragmentation won't cause an allocation failure.

---
 fs/nfsd/nfs4xdr.c | 7 ++++---
 fs/nfsd/state.h   | 2 +-
 2 files changed, 5 insertions(+), 4 deletions(-)
fs/nfsd/nfs4xdr.c +4 -3
index 5aafbd0f7ae30..0f30d93577e7b 100644
@@ -42,6 +42,8 @@
 #include <linux/sunrpc/svcauth_gss.h>
 #include <linux/sunrpc/addr.h>
 #include <linux/xattr.h>
+#include <linux/vmalloc.h>
+
 #include <uapi/linux/xattr.h>
 
 #include "idmap.h"
@@ -2364,10 +2366,9 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp)
 			    NFSD_MAX_OPS_PER_COMPOUND);
 
 	if (argp->opcnt > ARRAY_SIZE(argp->iops)) {
-		argp->ops = kzalloc(argp->opcnt * sizeof(*argp->ops), GFP_KERNEL);
+		argp->ops = vcalloc(argp->opcnt, sizeof(*argp->ops));
 		if (!argp->ops) {
 			argp->ops = argp->iops;
-			dprintk("nfsd: couldn't allocate room for COMPOUND\n");
 			return false;
 		}
 	}
@@ -5399,7 +5400,7 @@ void nfsd4_release_compoundargs(struct svc_rqst *rqstp)
 	struct nfsd4_compoundargs *args = rqstp->rq_argp;
 
 	if (args->ops != args->iops) {
-		kfree(args->ops);
+		vfree(args->ops);
 		args->ops = args->iops;
 	}
 	while (args->to_free) {
fs/nfsd/state.h +1 -1
index ae596dbf86675..5d28beb290fef 100644
@@ -175,7 +175,7 @@ static inline struct nfs4_delegation *delegstateid(struct nfs4_stid *s)
 /* Maximum number of slots per session. 160 is useful for long haul TCP */
 #define NFSD_MAX_SLOTS_PER_SESSION     160
 /* Maximum number of operations per session compound */
-#define NFSD_MAX_OPS_PER_COMPOUND	16
+#define NFSD_MAX_OPS_PER_COMPOUND	50
 /* Maximum  session per slot cache size */
 #define NFSD_SLOT_CACHE_SIZE		2048
 /* Maximum number of NFSD_SLOT_CACHE_SIZE slots per session */
Signature · 1 line
2.43.0
author maintainer
#423 lore raw

PATCH 423/690 NFSD: drop fname and flen args from nfsd_create_locked()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

[ Upstream commit 9558f9304ca1903090fa5d995a3269a8e82804b4 ]

nfsd_create_locked() does not use the "fname" and "flen" arguments, so drop them from declaration and all callers.

---
 fs/nfsd/nfsproc.c | 5 ++---
 fs/nfsd/vfs.c     | 5 ++---
 fs/nfsd/vfs.h     | 4 ++--
 3 files changed, 6 insertions(+), 8 deletions(-)
fs/nfsd/nfsproc.c +2 -3
index 4b19cc727ea50..ee02ede74bf53 100644
@@ -391,9 +391,8 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	resp->status = nfs_ok;
 	if (!inode) {
 		/* File doesn't exist. Create it and set attrs */
-		resp->status = nfsd_create_locked(rqstp, dirfhp, argp->name,
-						  argp->len, &attrs, type, rdev,
-						  newfhp);
+		resp->status = nfsd_create_locked(rqstp, dirfhp, &attrs, type,
+						  rdev, newfhp);
 	} else if (type == S_IFREG) {
 		dprintk("nfsd:   existing %s, valid=%x, size=%ld\n",
 			argp->name, attr->ia_valid, (long) attr->ia_size);
fs/nfsd/vfs.c +2 -3
index 343af6341e5e1..77f8ab3826d75 100644
@@ -1257,7 +1257,7 @@ nfsd_check_ignore_resizing(struct iattr *iap)
 /* The parent directory should already be locked: */
 __be32
 nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
-		   char *fname, int flen, struct nfsd_attrs *attrs,
+		   struct nfsd_attrs *attrs,
 		   int type, dev_t rdev, struct svc_fh *resfhp)
 {
 	struct dentry	*dentry, *dchild;
@@ -1384,8 +1384,7 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (err)
 		goto out_unlock;
 	fh_fill_pre_attrs(fhp);
-	err = nfsd_create_locked(rqstp, fhp, fname, flen, attrs, type,
-				 rdev, resfhp);
+	err = nfsd_create_locked(rqstp, fhp, attrs, type, rdev, resfhp);
 	fh_fill_post_attrs(fhp);
 out_unlock:
 	inode_unlock(dentry->d_inode);
fs/nfsd/vfs.h +2 -2
index c95cd414b4bb0..120521bc7b247 100644
@@ -79,8 +79,8 @@ __be32		nfsd4_clone_file_range(struct svc_rqst *rqstp,
 				       u64 count, bool sync);
 #endif /* CONFIG_NFSD_V4 */
 __be32		nfsd_create_locked(struct svc_rqst *, struct svc_fh *,
-				char *name, int len, struct nfsd_attrs *attrs,
-				int type, dev_t rdev, struct svc_fh *res);
+				struct nfsd_attrs *attrs, int type, dev_t rdev,
+				struct svc_fh *res);
 __be32		nfsd_create(struct svc_rqst *, struct svc_fh *,
 				char *name, int len, struct nfsd_attrs *attrs,
 				int type, dev_t rdev, struct svc_fh *res);
Signature · 1 line
2.43.0
author maintainer
#424 lore raw

PATCH 424/690 nfsd: clean up mounted_on_fileid handling

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jeff Layton <[email protected]>

[ Upstream commit 6106d9119b6599fa23dc556b429d887b4c2d9f62 ]

We only need the inode number for this, not a full rack of attributes. Rename this function make it take a pointer to a u64 instead of struct kstat, and change it to just request STATX_INO.

[ cel: renamed get_mounted_on_ino() ]
---
 fs/nfsd/nfs4xdr.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
fs/nfsd/nfs4xdr.c +9 -7
index 0f30d93577e7b..3ad9b41c51730 100644
@@ -2772,9 +2772,10 @@ static __be32 fattr_handle_absent_fs(u32 *bmval0, u32 *bmval1, u32 *bmval2, u32
 }
 
 
-static int get_parent_attributes(struct svc_export *exp, struct kstat *stat)
+static int nfsd4_get_mounted_on_ino(struct svc_export *exp, u64 *pino)
 {
 	struct path path = exp->ex_path;
+	struct kstat stat;
 	int err;
 
 	path_get(&path);
@@ -2782,8 +2783,10 @@ static int get_parent_attributes(struct svc_export *exp, struct kstat *stat)
 		if (path.dentry != path.mnt->mnt_root)
 			break;
 	}
-	err = vfs_getattr(&path, stat, STATX_BASIC_STATS, AT_STATX_SYNC_AS_STAT);
+	err = vfs_getattr(&path, &stat, STATX_INO, AT_STATX_SYNC_AS_STAT);
 	path_put(&path);
+	if (!err)
+		*pino = stat.ino;
 	return err;
 }
 
@@ -3280,22 +3283,21 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
 		*p++ = cpu_to_be32(stat.btime.tv_nsec);
 	}
 	if (bmval1 & FATTR4_WORD1_MOUNTED_ON_FILEID) {
-		struct kstat parent_stat;
 		u64 ino = stat.ino;
 
 		p = xdr_reserve_space(xdr, 8);
 		if (!p)
                 	goto out_resource;
 		/*
-		 * Get parent's attributes if not ignoring crossmount
-		 * and this is the root of a cross-mounted filesystem.
+		 * Get ino of mountpoint in parent filesystem, if not ignoring
+		 * crossmount and this is the root of a cross-mounted
+		 * filesystem.
 		 */
 		if (ignore_crossmnt == 0 &&
 		    dentry == exp->ex_path.mnt->mnt_root) {
-			err = get_parent_attributes(exp, &parent_stat);
+			err = nfsd4_get_mounted_on_ino(exp, &ino);
 			if (err)
 				goto out_nfserr;
-			ino = parent_stat.ino;
 		}
 		p = xdr_encode_hyper(p, ino);
 	}
Signature · 1 line
2.43.0
author maintainer
#425 lore raw

PATCH 425/690 nfsd: remove nfsd4_prepare_cb_recall() declaration

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Gaosheng Cui <[email protected]>

[ Upstream commit 18224dc58d960c65446971930d0487fc72d00598 ]

nfsd4_prepare_cb_recall() has been removed since commit 0162ac2b978e ("nfsd: introduce nfsd4_callback_ops"), so remove it.

---
 fs/nfsd/state.h | 1 -
 1 file changed, 1 deletion(-)
fs/nfsd/state.h +0 -1
index 5d28beb290fef..4155be65d8069 100644
@@ -697,7 +697,6 @@ extern int nfsd4_create_callback_queue(void);
 extern void nfsd4_destroy_callback_queue(void);
 extern void nfsd4_shutdown_callback(struct nfs4_client *);
 extern void nfsd4_shutdown_copy(struct nfs4_client *clp);
-extern void nfsd4_prepare_cb_recall(struct nfs4_delegation *dp);
 extern struct nfs4_client_reclaim *nfs4_client_to_reclaim(struct xdr_netobj name,
 				struct xdr_netobj princhash, struct nfsd_net *nn);
 extern bool nfs4_has_reclaimed_state(struct xdr_netobj name, struct nfsd_net *nn);
Signature · 1 line
2.43.0
author maintainer
#426 lore raw

PATCH 426/690 NFSD: Replace dprintk() call site in fh_verify()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 948755efc951de75c87d4fa916d9d36b58299295 ]

Record permission errors in the trace log. Note that the new trace event is conditional, so it will only record non-zero return values from nfsd_permission().

---
 fs/nfsd/nfsfh.c |  8 +-------
 fs/nfsd/trace.h | 48 +++++++++++++++++++++++++++++++++++++++++++++---
 2 files changed, 46 insertions(+), 10 deletions(-)
fs/nfsd/nfsfh.c +1 -7
index a5b71526cee0f..d73434200df98 100644
@@ -392,13 +392,7 @@ fh_verify(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, int access)
 skip_pseudoflavor_check:
 	/* Finally, check access permissions. */
 	error = nfsd_permission(rqstp, exp, dentry, access);
-
-	if (error) {
-		dprintk("fh_verify: %pd2 permission failure, "
-			"acc=%x, error=%d\n",
-			dentry,
-			access, ntohl(error));
-	}
+	trace_nfsd_fh_verify_err(rqstp, fhp, type, access, error);
 out:
 	if (error == nfserr_stale)
 		nfsd_stats_fh_stale_inc(exp);
fs/nfsd/trace.h +45 -3

65 lines not on this page · raw message

Signature · 1 line
2.43.0
author maintainer
#427 lore raw

PATCH 427/690 NFSD: Trace NFSv4 COMPOUND tags

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit de29cf7e6cbbe236c3a51999c188fcd467762899 ]

The Linux NFSv4 client implementation does not use COMPOUND tags, but the Solaris and MacOS implementations do, and so does pynfs. Record these eye-catchers in the server's trace buffer to annotate client requests while troubleshooting.

---
 fs/nfsd/nfs4proc.c |  2 +-
 fs/nfsd/trace.h    | 21 ++++++++++++++-------
 2 files changed, 15 insertions(+), 8 deletions(-)
fs/nfsd/nfs4proc.c +1 -1
index 69d3013fb1b26..e4c0dc577fe35 100644
@@ -2637,7 +2637,7 @@ nfsd4_proc_compound(struct svc_rqst *rqstp)
 
 	rqstp->rq_lease_breaker = (void **)&cstate->clp;
 
-	trace_nfsd_compound(rqstp, args->client_opcnt);
+	trace_nfsd_compound(rqstp, args->tag, args->taglen, args->client_opcnt);
 	while (!status && resp->opcnt < args->opcnt) {
 		op = &args->ops[resp->opcnt++];
 
fs/nfsd/trace.h +14 -7
index 297bf9ddc5090..c5d4a258680c3 100644
@@ -84,19 +84,26 @@ DEFINE_NFSD_XDR_ERR_EVENT(cant_encode);
 		{ NFSD_MAY_64BIT_COOKIE,	"64BIT_COOKIE" })
 
 TRACE_EVENT(nfsd_compound,
-	TP_PROTO(const struct svc_rqst *rqst,
-		 u32 args_opcnt),
-	TP_ARGS(rqst, args_opcnt),
+	TP_PROTO(
+		const struct svc_rqst *rqst,
+		const char *tag,
+		u32 taglen,
+		u32 opcnt
+	),
+	TP_ARGS(rqst, tag, taglen, opcnt),
 	TP_STRUCT__entry(
 		__field(u32, xid)
-		__field(u32, args_opcnt)
+		__field(u32, opcnt)
+		__string_len(tag, tag, taglen)
 	),
 	TP_fast_assign(
 		__entry->xid = be32_to_cpu(rqst->rq_xid);
-		__entry->args_opcnt = args_opcnt;
+		__entry->opcnt = opcnt;
+		__assign_str_len(tag, tag, taglen);
 	),
-	TP_printk("xid=0x%08x opcnt=%u",
-		__entry->xid, __entry->args_opcnt)
+	TP_printk("xid=0x%08x opcnt=%u tag=%s",
+		__entry->xid, __entry->opcnt, __get_str(tag)
+	)
 )
 
 TRACE_EVENT(nfsd_compound_status,
Signature · 1 line
2.43.0
author maintainer
#428 lore raw

PATCH 428/690 NFSD: Add tracepoints to report NFSv4 callback completions

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 1035d65446a018ca2dd179e29a2fcd6d29057781 ]

Wireshark has always been lousy about dissecting NFSv4 callbacks, especially NFSv4.0 backchannel requests. Add tracepoints so we can surgically capture these events in the trace log.

Tracepoints are time-stamped and ordered so that we can now observe the timing relationship between a CB_RECALL Reply and the client's DELEGRETURN Call. Example:

            nfsd-1153  [002]   211.986391: nfsd_cb_recall:       addr=192.168.1.67:45767 client 62ea82e4:fee7492a stateid 00000003:00000001

            nfsd-1153  [002]   212.095634: nfsd_compound:        xid=0x0000002c opcnt=2
            nfsd-1153  [002]   212.095647: nfsd_compound_status: op=1/2 OP_PUTFH status=0
            nfsd-1153  [002]   212.095658: nfsd_file_put:        hash=0xf72 inode=0xffff9291148c7410 ref=3 flags=HASHED|REFERENCED may=READ file=0xffff929103b3ea00
            nfsd-1153  [002]   212.095661: nfsd_compound_status: op=2/2 OP_DELEGRETURN status=0

kworker/u25:8-148 [002] 212.096713: nfsd_cb_recall_done: client 62ea82e4:fee7492a stateid 00000003:00000001 status=0

---
 fs/nfsd/nfs4layouts.c |  2 +-
 fs/nfsd/nfs4proc.c    |  4 ++++
 fs/nfsd/nfs4state.c   |  4 ++++
 fs/nfsd/trace.h       | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 48 insertions(+), 1 deletion(-)
fs/nfsd/nfs4layouts.c +1 -1
index 7018d209b784a..e4e23b2a3e655 100644
@@ -657,7 +657,7 @@ nfsd4_cb_layout_done(struct nfsd4_callback *cb, struct rpc_task *task)
 	ktime_t now, cutoff;
 	const struct nfsd4_layout_ops *ops;
 
-
+	trace_nfsd_cb_layout_done(&ls->ls_stid.sc_stateid, task);
 	switch (task->tk_status) {
 	case 0:
 	case -NFS4ERR_DELAY:
fs/nfsd/nfs4proc.c +4 -0
index e4c0dc577fe35..ce8062c959315 100644
@@ -1604,6 +1604,10 @@ static void nfsd4_cb_offload_release(struct nfsd4_callback *cb)
 static int nfsd4_cb_offload_done(struct nfsd4_callback *cb,
 				 struct rpc_task *task)
 {
+	struct nfsd4_cb_offload *cbo =
+		container_of(cb, struct nfsd4_cb_offload, co_cb);
+
+	trace_nfsd_cb_offload_done(&cbo->co_res.cb_stateid, task);
 	return 1;
 }
 
fs/nfsd/nfs4state.c +4 -0
index 66cf8217ebe57..a0c5658599ead 100644
@@ -357,6 +357,8 @@ nfsd4_cb_notify_lock_prepare(struct nfsd4_callback *cb)
 static int
 nfsd4_cb_notify_lock_done(struct nfsd4_callback *cb, struct rpc_task *task)
 {
+	trace_nfsd_cb_notify_lock_done(&zero_stateid, task);
+
 	/*
 	 * Since this is just an optimization, we don't try very hard if it
 	 * turns out not to succeed. We'll requeue it on NFS4ERR_DELAY, and
@@ -4760,6 +4762,8 @@ static int nfsd4_cb_recall_done(struct nfsd4_callback *cb,
 {
 	struct nfs4_delegation *dp = cb_to_delegation(cb);
 
+	trace_nfsd_cb_recall_done(&dp->dl_stid.sc_stateid, task);
+
 	if (dp->dl_stid.sc_type == NFS4_CLOSED_DELEG_STID ||
 	    dp->dl_stid.sc_type == NFS4_REVOKED_DELEG_STID)
 	        return 1;
fs/nfsd/trace.h +39 -0
index c5d4a258680c3..d832429e575e4 100644
@@ -1449,6 +1449,45 @@ TRACE_EVENT(nfsd_cb_offload,
 		__entry->fh_hash, __entry->count, __entry->status)
 );
 
+DECLARE_EVENT_CLASS(nfsd_cb_done_class,
+	TP_PROTO(
+		const stateid_t *stp,
+		const struct rpc_task *task
+	),
+	TP_ARGS(stp, task),
+	TP_STRUCT__entry(
+		__field(u32, cl_boot)
+		__field(u32, cl_id)
+		__field(u32, si_id)
+		__field(u32, si_generation)
+		__field(int, status)
+	),
+	TP_fast_assign(
+		__entry->cl_boot = stp->si_opaque.so_clid.cl_boot;
+		__entry->cl_id = stp->si_opaque.so_clid.cl_id;
+		__entry->si_id = stp->si_opaque.so_id;
+		__entry->si_generation = stp->si_generation;
+		__entry->status = task->tk_status;
+	),
+	TP_printk("client %08x:%08x stateid %08x:%08x status=%d",
+		__entry->cl_boot, __entry->cl_id, __entry->si_id,
+		__entry->si_generation, __entry->status
+	)
+);
+
+#define DEFINE_NFSD_CB_DONE_EVENT(name)			\
+DEFINE_EVENT(nfsd_cb_done_class, name,			\
+	TP_PROTO(					\
+		const stateid_t *stp,			\
+		const struct rpc_task *task		\
+	),						\
+	TP_ARGS(stp, task))
+
+DEFINE_NFSD_CB_DONE_EVENT(nfsd_cb_recall_done);
+DEFINE_NFSD_CB_DONE_EVENT(nfsd_cb_notify_lock_done);
+DEFINE_NFSD_CB_DONE_EVENT(nfsd_cb_layout_done);
+DEFINE_NFSD_CB_DONE_EVENT(nfsd_cb_offload_done);
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
2.43.0
author maintainer
#429 lore raw

PATCH 429/690 NFSD: Add a mechanism to wait for a DELEGRETURN

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit c035362eb935fe9381d9d1cc453bc2a37460e24c ]

Subsequent patches will use this mechanism to wake up an operation that is waiting for a client to return a delegation.

The new tracepoint records whether the wait timed out or was properly awoken by the expected DELEGRETURN:

            nfsd-1155  [002] 83799.493199: nfsd_delegret_wakeup: xid=0x14b7d6ef fh_hash=0xf6826792 (timed out)
---
 fs/nfsd/nfs4state.c | 30 ++++++++++++++++++++++++++++++
 fs/nfsd/nfsd.h      |  7 +++++++
 fs/nfsd/trace.h     | 23 +++++++++++++++++++++++
 3 files changed, 60 insertions(+)
fs/nfsd/nfs4state.c +30 -0
index a0c5658599ead..6cb654e308787 100644
@@ -4734,6 +4734,35 @@ nfs4_share_conflict(struct svc_fh *current_fh, unsigned int deny_type)
 	return ret;
 }
 
+static bool nfsd4_deleg_present(const struct inode *inode)
+{
+	struct file_lock_context *ctx = smp_load_acquire(&inode->i_flctx);
+
+	return ctx && !list_empty_careful(&ctx->flc_lease);
+}
+
+/**
+ * nfsd_wait_for_delegreturn - wait for delegations to be returned
+ * @rqstp: the RPC transaction being executed
+ * @inode: in-core inode of the file being waited for
+ *
+ * The timeout prevents deadlock if all nfsd threads happen to be
+ * tied up waiting for returning delegations.
+ *
+ * Return values:
+ *   %true: delegation was returned
+ *   %false: timed out waiting for delegreturn
+ */
+bool nfsd_wait_for_delegreturn(struct svc_rqst *rqstp, struct inode *inode)
+{
+	long __maybe_unused timeo;
+
+	timeo = wait_var_event_timeout(inode, !nfsd4_deleg_present(inode),
+				       NFSD_DELEGRETURN_TIMEOUT);
+	trace_nfsd_delegret_wakeup(rqstp, inode, timeo);
+	return timeo > 0;
+}
+
 static void nfsd4_cb_recall_prepare(struct nfsd4_callback *cb)
 {
 	struct nfs4_delegation *dp = cb_to_delegation(cb);
@@ -6797,6 +6826,7 @@ nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	if (status)
 		goto put_stateid;
 
+	wake_up_var(d_inode(cstate->current_fh.fh_dentry));
 	destroy_delegation(dp);
 put_stateid:
 	nfs4_put_stid(&dp->dl_stid);
fs/nfsd/nfsd.h +7 -0
index 57a468ed85c35..6ab4ad41ae84e 100644
@@ -164,6 +164,7 @@ char * nfs4_recoverydir(void);
 bool nfsd4_spo_must_allow(struct svc_rqst *rqstp);
 int nfsd4_create_laundry_wq(void);
 void nfsd4_destroy_laundry_wq(void);
+bool nfsd_wait_for_delegreturn(struct svc_rqst *rqstp, struct inode *inode);
 #else
 static inline int nfsd4_init_slabs(void) { return 0; }
 static inline void nfsd4_free_slabs(void) { }
@@ -179,6 +180,11 @@ static inline bool nfsd4_spo_must_allow(struct svc_rqst *rqstp)
 }
 static inline int nfsd4_create_laundry_wq(void) { return 0; };
 static inline void nfsd4_destroy_laundry_wq(void) {};
+static inline bool nfsd_wait_for_delegreturn(struct svc_rqst *rqstp,
+					      struct inode *inode)
+{
+	return false;
+}
 #endif
 
 /*
@@ -343,6 +349,7 @@ void		nfsd_lockd_shutdown(void);
 #define	NFSD_COURTESY_CLIENT_TIMEOUT	(24 * 60 * 60)	/* seconds */
 #define	NFSD_CLIENT_MAX_TRIM_PER_RUN	128
 #define	NFS4_CLIENTS_PER_GB		1024
+#define NFSD_DELEGRETURN_TIMEOUT	(HZ / 34)	/* 30ms */
 
 /*
  * The following attributes are currently not supported by the NFSv4 server:
fs/nfsd/trace.h +23 -0
index d832429e575e4..1229502b6e9e0 100644
@@ -538,6 +538,29 @@ DEFINE_NFSD_COPY_ERR_EVENT(clone_file_range_err);
 #include "filecache.h"
 #include "vfs.h"
 
+TRACE_EVENT(nfsd_delegret_wakeup,
+	TP_PROTO(
+		const struct svc_rqst *rqstp,
+		const struct inode *inode,
+		long timeo
+	),
+	TP_ARGS(rqstp, inode, timeo),
+	TP_STRUCT__entry(
+		__field(u32, xid)
+		__field(const void *, inode)
+		__field(long, timeo)
+	),
+	TP_fast_assign(
+		__entry->xid = be32_to_cpu(rqstp->rq_xid);
+		__entry->inode = inode;
+		__entry->timeo = timeo;
+	),
+	TP_printk("xid=0x%08x inode=%p%s",
+		  __entry->xid, __entry->inode,
+		  __entry->timeo == 0 ? " (timed out)" : ""
+	)
+);
+
 DECLARE_EVENT_CLASS(nfsd_stateid_class,
 	TP_PROTO(stateid_t *stp),
 	TP_ARGS(stp),
Signature · 1 line
2.43.0
author maintainer
#430 lore raw

PATCH 430/690 NFSD: Refactor nfsd_setattr()

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit c0aa1913db57219e91a0a8832363cbafb3a9cf8f ]

Move code that will be retried (in a subsequent patch) into a helper function.

---
 fs/nfsd/vfs.c | 97 ++++++++++++++++++++++++++++++---------------------
 1 file changed, 57 insertions(+), 40 deletions(-)
fs/nfsd/vfs.c +57 -40
index 77f8ab3826d75..392df2353556e 100644
@@ -344,8 +344,61 @@ nfsd_get_write_access(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	return nfserrno(get_write_access(inode));
 }
 
-/*
- * Set various file attributes.  After this call fhp needs an fh_put.
+static int __nfsd_setattr(struct dentry *dentry, struct iattr *iap)
+{
+	int host_err;
+
+	if (iap->ia_valid & ATTR_SIZE) {
+		/*
+		 * RFC5661, Section 18.30.4:
+		 *   Changing the size of a file with SETATTR indirectly
+		 *   changes the time_modify and change attributes.
+		 *
+		 * (and similar for the older RFCs)
+		 */
+		struct iattr size_attr = {
+			.ia_valid	= ATTR_SIZE | ATTR_CTIME | ATTR_MTIME,
+			.ia_size	= iap->ia_size,
+		};
+
+		if (iap->ia_size < 0)
+			return -EFBIG;
+
+		host_err = notify_change(&init_user_ns, dentry, &size_attr, NULL);
+		if (host_err)
+			return host_err;
+		iap->ia_valid &= ~ATTR_SIZE;
+
+		/*
+		 * Avoid the additional setattr call below if the only other
+		 * attribute that the client sends is the mtime, as we update
+		 * it as part of the size change above.
+		 */
+		if ((iap->ia_valid & ~ATTR_MTIME) == 0)
+			return 0;
+	}
+
+	if (!iap->ia_valid)
+		return 0;
+
+	iap->ia_valid |= ATTR_CTIME;
+	return notify_change(&init_user_ns, dentry, iap, NULL);
+}
+
+/**
+ * nfsd_setattr - Set various file attributes.
+ * @rqstp: controlling RPC transaction
+ * @fhp: filehandle of target
+ * @attr: attributes to set
+ * @check_guard: set to 1 if guardtime is a valid timestamp
+ * @guardtime: do not act if ctime.tv_sec does not match this timestamp
+ *
+ * This call may adjust the contents of @attr (in particular, this
+ * call may change the bits in the na_iattr.ia_valid field).
+ *
+ * Returns nfs_ok on success, otherwise an NFS status code is
+ * returned. Caller must release @fhp by calling fh_put in either
+ * case.
  */
 __be32
 nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
@@ -358,7 +411,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	int		accmode = NFSD_MAY_SATTR;
 	umode_t		ftype = 0;
 	__be32		err;
-	int		host_err = 0;
+	int		host_err;
 	bool		get_write_count;
 	bool		size_change = (iap->ia_valid & ATTR_SIZE);
 
@@ -415,43 +468,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	}
 
 	inode_lock(inode);
-	if (size_change) {
-		/*
-		 * RFC5661, Section 18.30.4:
-		 *   Changing the size of a file with SETATTR indirectly
-		 *   changes the time_modify and change attributes.
-		 *
-		 * (and similar for the older RFCs)
-		 */
-		struct iattr size_attr = {
-			.ia_valid	= ATTR_SIZE | ATTR_CTIME | ATTR_MTIME,
-			.ia_size	= iap->ia_size,
-		};
-
-		host_err = -EFBIG;
-		if (iap->ia_size < 0)
-			goto out_unlock;
-
-		host_err = notify_change(&init_user_ns, dentry, &size_attr, NULL);
-		if (host_err)
-			goto out_unlock;
-		iap->ia_valid &= ~ATTR_SIZE;
-
-		/*
-		 * Avoid the additional setattr call below if the only other
-		 * attribute that the client sends is the mtime, as we update
-		 * it as part of the size change above.
-		 */
-		if ((iap->ia_valid & ~ATTR_MTIME) == 0)
-			goto out_unlock;
-	}
-
-	if (iap->ia_valid) {
-		iap->ia_valid |= ATTR_CTIME;
-		host_err = notify_change(&init_user_ns, dentry, iap, NULL);
-	}
-
-out_unlock:
+	host_err = __nfsd_setattr(dentry, iap);
 	if (attr->na_seclabel && attr->na_seclabel->len)
 		attr->na_labelerr = security_inode_setsecctx(dentry,
 			attr->na_seclabel->data, attr->na_seclabel->len);
Signature · 1 line
2.43.0
author maintainer
#431 lore raw

PATCH 431/690 NFSD: Make nfsd4_setattr() wait before returning NFS4ERR_DELAY

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 34b91dda7124fc3259e4b2ae53e0c933dedfec01 ]

nfsd_setattr() can kick off a CB_RECALL (via notify_change() -> break_lease()) if a delegation is present. Before returning NFS4ERR_DELAY, give the client holding that delegation a chance to return it and then retry the nfsd_setattr() again, once.

---
 fs/nfsd/vfs.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
fs/nfsd/vfs.c +8 -1
index 392df2353556e..e8329051dde01 100644
@@ -414,6 +414,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	int		host_err;
 	bool		get_write_count;
 	bool		size_change = (iap->ia_valid & ATTR_SIZE);
+	int		retries;
 
 	if (iap->ia_valid & ATTR_SIZE) {
 		accmode |= NFSD_MAY_WRITE|NFSD_MAY_OWNER_OVERRIDE;
@@ -468,7 +469,13 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	}
 
 	inode_lock(inode);
-	host_err = __nfsd_setattr(dentry, iap);
+	for (retries = 1;;) {
+		host_err = __nfsd_setattr(dentry, iap);
+		if (host_err != -EAGAIN || !retries--)
+			break;
+		if (!nfsd_wait_for_delegreturn(rqstp, inode))
+			break;
+	}
 	if (attr->na_seclabel && attr->na_seclabel->len)
 		attr->na_labelerr = security_inode_setsecctx(dentry,
 			attr->na_seclabel->data, attr->na_seclabel->len);
Signature · 1 line
2.43.0
author maintainer
#432 lore raw

PATCH 432/690 NFSD: Make nfsd4_rename() wait before returning NFS4ERR_DELAY

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 68c522afd0b1936b48a03a4c8b81261e7597c62d ]

nfsd_rename() can kick off a CB_RECALL (via vfs_rename() -> leases_conflict()) if a delegation is present. Before returning NFS4ERR_DELAY, give the client holding that delegation a chance to return it and then retry the nfsd_rename() again, once.

This version of the patch handles renaming an existing file, but does not deal with renaming onto an existing file. That case will still always trigger an NFS4ERR_DELAY.

---
 fs/nfsd/vfs.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)
fs/nfsd/vfs.c +9 -1
index e8329051dde01..4c5cc142562b2 100644
@@ -1698,7 +1698,15 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
 			.new_dir	= tdir,
 			.new_dentry	= ndentry,
 		};
-		host_err = vfs_rename(&rd);
+		int retries;
+
+		for (retries = 1;;) {
+			host_err = vfs_rename(&rd);
+			if (host_err != -EAGAIN || !retries--)
+				break;
+			if (!nfsd_wait_for_delegreturn(rqstp, d_inode(odentry)))
+				break;
+		}
 		if (!host_err) {
 			host_err = commit_metadata(tfhp);
 			if (!host_err)
Signature · 1 line
2.43.0
author maintainer
#433 lore raw

PATCH 433/690 NFSD: Make nfsd4_remove() wait before returning NFS4ERR_DELAY

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

[ Upstream commit 5f5f8b6d655fd947e899b1771c2f7cb581a06764 ]

nfsd_unlink() can kick off a CB_RECALL (via vfs_unlink() -> leases_conflict()) if a delegation is present. Before returning NFS4ERR_DELAY, give the client holding that delegation a chance to return it and then retry the nfsd_unlink() again, once.

---
 fs/nfsd/vfs.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)
fs/nfsd/vfs.c +10 -1
index 4c5cc142562b2..d17377148b669 100644
@@ -1790,9 +1790,18 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 
 	fh_fill_pre_attrs(fhp);
 	if (type != S_IFDIR) {
+		int retries;
+
 		if (rdentry->d_sb->s_export_op->flags & EXPORT_OP_CLOSE_BEFORE_UNLINK)
 			nfsd_close_cached_files(rdentry);
-		host_err = vfs_unlink(&init_user_ns, dirp, rdentry, NULL);
+
+		for (retries = 1;;) {
+			host_err = vfs_unlink(&init_user_ns, dirp, rdentry, NULL);
+			if (host_err != -EAGAIN || !retries--)
+				break;
+			if (!nfsd_wait_for_delegreturn(rqstp, rinode))
+				break;
+		}
 	} else {
 		host_err = vfs_rmdir(&init_user_ns, dirp, rdentry);
 	}
Signature · 1 line
2.43.0
author maintainer
#434 lore raw

PATCH 434/690 NFSD: keep track of the number of courtesy clients in the system

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Dai Ngo <[email protected]>

[ Upstream commit 3a4ea23d86a317c4b68b9a69d51f7e84e1e04357 ]

Add counter nfs4_courtesy_client_count to nfsd_net to keep track of the number of courtesy clients in the system.

---
 fs/nfsd/netns.h     |  2 ++
 fs/nfsd/nfs4state.c | 17 ++++++++++++++++-
 2 files changed, 18 insertions(+), 1 deletion(-)
fs/nfsd/netns.h +2 -0
index ffe17743cc74b..55c7006d6109a 100644
@@ -192,6 +192,8 @@ struct nfsd_net {
 
 	atomic_t		nfs4_client_count;
 	int			nfs4_max_clients;
+
+	atomic_t		nfsd_courtesy_clients;
 };
 
 /* Simple check to find out if a given net was properly initialized */
fs/nfsd/nfs4state.c +16 -1
index 6cb654e308787..6a7a99511111d 100644
@@ -160,6 +160,13 @@ static bool is_client_expired(struct nfs4_client *clp)
 	return clp->cl_time == 0;
 }
 
+static void nfsd4_dec_courtesy_client_count(struct nfsd_net *nn,
+					struct nfs4_client *clp)
+{
+	if (clp->cl_state != NFSD4_ACTIVE)
+		atomic_add_unless(&nn->nfsd_courtesy_clients, -1, 0);
+}
+
 static __be32 get_client_locked(struct nfs4_client *clp)
 {
 	struct nfsd_net *nn = net_generic(clp->net, nfsd_net_id);
@@ -169,6 +176,7 @@ static __be32 get_client_locked(struct nfs4_client *clp)
 	if (is_client_expired(clp))
 		return nfserr_expired;
 	atomic_inc(&clp->cl_rpc_users);
+	nfsd4_dec_courtesy_client_count(nn, clp);
 	clp->cl_state = NFSD4_ACTIVE;
 	return nfs_ok;
 }
@@ -190,6 +198,7 @@ renew_client_locked(struct nfs4_client *clp)
 
 	list_move_tail(&clp->cl_lru, &nn->client_lru);
 	clp->cl_time = ktime_get_boottime_seconds();
+	nfsd4_dec_courtesy_client_count(nn, clp);
 	clp->cl_state = NFSD4_ACTIVE;
 }
 
@@ -2248,6 +2257,7 @@ __destroy_client(struct nfs4_client *clp)
 	if (clp->cl_cb_conn.cb_xprt)
 		svc_xprt_put(clp->cl_cb_conn.cb_xprt);
 	atomic_add_unless(&nn->nfs4_client_count, -1, 0);
+	nfsd4_dec_courtesy_client_count(nn, clp);
 	free_client(clp);
 	wake_up_all(&expiry_wq);
 }
@@ -4375,6 +4385,8 @@ void nfsd4_init_leases_net(struct nfsd_net *nn)
 	max_clients = (u64)si.totalram * si.mem_unit / (1024 * 1024 * 1024);
 	max_clients *= NFS4_CLIENTS_PER_GB;
 	nn->nfs4_max_clients = max_t(int, max_clients, NFS4_CLIENTS_PER_GB);
+
+	atomic_set(&nn->nfsd_courtesy_clients, 0);
 }
 
 static void init_nfs4_replay(struct nfs4_replay *rp)
@@ -5928,8 +5940,11 @@ nfs4_get_client_reaplist(struct nfsd_net *nn, struct list_head *reaplist,
 			goto exp_client;
 		if (!state_expired(lt, clp->cl_time))
 			break;
-		if (!atomic_read(&clp->cl_rpc_users))
+		if (!atomic_read(&clp->cl_rpc_users)) {
+			if (clp->cl_state == NFSD4_ACTIVE)
+				atomic_inc(&nn->nfsd_courtesy_clients);
 			clp->cl_state = NFSD4_COURTESY;
+		}
 		if (!client_has_state(clp))
 			goto exp_client;
 		if (!nfs4_anylock_blockers(clp))
Signature · 1 line
2.43.0
author maintainer
#435 lore raw

PATCH 435/690 NFSD: add shrinker to reap courtesy clients on low memory condition

In reply to Greg Kroah-Hartman

5.15-stable review patch. If anyone has any objections, please let me know.

------------------

From: Dai Ngo <[email protected]>

[ Upstream commit 7746b32f467b3813fb61faaab3258de35806a7ac ]

Add courtesy_client_reaper to react to low memory condition triggered by the system memory shrinker.

The delayed_work for the courtesy_client_reaper is scheduled on the shrinker's count callback using the laundry_wq.

The shrinker's scan callback is not used for expiring the courtesy clients due to potential deadlocks.

---
 fs/nfsd/netns.h     |  2 +
 fs/nfsd/nfs4state.c | 94 +++++++++++++++++++++++++++++++++++++++++----
 fs/nfsd/nfsctl.c    |  6 ++-
 fs/nfsd/nfsd.h      |  6 ++-
 4 files changed, 96 insertions(+), 12 deletions(-)
fs/nfsd/netns.h +2 -0
index 55c7006d6109a..8c854ba3285bb 100644
@@ -194,6 +194,8 @@ struct nfsd_net {
 	int			nfs4_max_clients;
 
 	atomic_t		nfsd_courtesy_clients;
+	struct shrinker		nfsd_client_shrinker;
+	struct delayed_work	nfsd_shrinker_work;
 };
 
 /* Simple check to find out if a given net was properly initialized */
fs/nfsd/nfs4state.c +86 -8

133 lines not on this page · raw message

fs/nfsd/nfsctl.c +4 -2
index 917fa1892fd2d..597a26ad4183f 100644
@@ -1481,11 +1481,12 @@ static __net_init int nfsd_init_net(struct net *net)
 		goto out_idmap_error;
 	nn->nfsd_versions = NULL;
 	nn->nfsd4_minorversions = NULL;
+	retval = nfsd4_init_leases_net(nn);
+	if (retval)
+		goto out_drc_error;
 	retval = nfsd_reply_cache_init(nn);
 	if (retval)
 		goto out_drc_error;
-	nfsd4_init_leases_net(nn);
-
 	get_random_bytes(&nn->siphash_key, sizeof(nn->siphash_key));
 	seqlock_init(&nn->writeverf_lock);
 
@@ -1507,6 +1508,7 @@ static __net_exit void nfsd_exit_net(struct net *net)
 	nfsd_idmap_shutdown(net);
 	nfsd_export_shutdown(net);
 	nfsd_netns_free_versions(net_generic(net, nfsd_net_id));
+	nfsd4_leases_net_shutdown(nn);
 }
 
 static struct pernet_operations nfsd_net_ops = {
fs/nfsd/nfsd.h +4 -2
index 6ab4ad41ae84e..09726c5b9a317 100644
@@ -505,7 +505,8 @@ extern void unregister_cld_notifier(void);
 extern void nfsd4_ssc_init_umount_work(struct nfsd_net *nn);
 #endif
 
-extern void nfsd4_init_leases_net(struct nfsd_net *nn);
+extern int nfsd4_init_leases_net(struct nfsd_net *nn);
+extern void nfsd4_leases_net_shutdown(struct nfsd_net *nn);
 
 #else /* CONFIG_NFSD_V4 */
 static inline int nfsd4_is_junction(struct dentry *dentry)
@@ -513,7 +514,8 @@ static inline int nfsd4_is_junction(struct dentry *dentry)
 	return 0;
 }
 
-static inline void nfsd4_init_leases_net(struct nfsd_net *nn) {};
+static inline int nfsd4_init_leases_net(struct nfsd_net *nn) { return 0; };
+static inline void nfsd4_leases_net_shutdown(struct nfsd_net *nn) {};
 
 #define register_cld_notifier() 0
 #define unregister_cld_notifier() do { } while(0)
Signature · 1 line
2.43.0