Merge tag 'iversion-v6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/jlayton/linux

Pull i_version updates from Jeff Layton:
 "This overhauls how we handle i_version queries from nfsd.

  Instead of having special routines and grabbing the i_version field
  directly out of the inode in some cases, we've moved most of the
  handling into the various filesystems' getattr operations. As a bonus,
  this makes ceph's change attribute usable by knfsd as well.

  This should pave the way for future work to make this value queryable
  by userland, and to make it more resilient against rolling back on a
  crash"

* tag 'iversion-v6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/jlayton/linux:
  nfsd: remove fetch_iversion export operation
  nfsd: use the getattr operation to fetch i_version
  nfsd: move nfsd4_change_attribute to nfsfh.c
  ceph: report the inode version in getattr if requested
  nfs: report the inode version in getattr if requested
  vfs: plumb i_version handling into struct kstat
  fs: clarify when the i_version counter must be updated
  fs: uninline inode_query_iversion

3 files changed, 31 insertions, 39 deletions

diff --git a/include/linux/exportfs.h b/include/linux/exportfs.h
index fe848901fcc3..9f4d4bcbf251 100644
--- a/include/linux/exportfs.h
+++ b/include/linux/exportfs.h
@@ -213,7 +213,6 @@ struct export_operations {
 			  bool write, u32 *device_generation);
 	int (*commit_blocks)(struct inode *inode, struct iomap *iomaps,
 			     int nr_iomaps, struct iattr *iattr);
-	u64 (*fetch_iversion)(struct inode *);
 #define	EXPORT_OP_NOWCC			(0x1) /* don't collect v3 wcc data */
 #define	EXPORT_OP_NOSUBTREECHK		(0x2) /* no subtree checking */
 #define	EXPORT_OP_CLOSE_BEFORE_UNLINK	(0x4) /* close files before unlink */
diff --git a/include/linux/iversion.h b/include/linux/iversion.h
index e27bd4f55d84..f174ff1b59ee 100644
--- a/include/linux/iversion.h
+++ b/include/linux/iversion.h
@@ -9,8 +9,26 @@
  * ---------------------------
  * The change attribute (i_version) is mandated by NFSv4 and is mostly for
  * knfsd, but is also used for other purposes (e.g. IMA). The i_version must
- * appear different to observers if there was a change to the inode's data or
- * metadata since it was last queried.
+ * appear larger to observers if there was an explicit change to the inode's
+ * data or metadata since it was last queried.
+ *
+ * An explicit change is one that would ordinarily result in a change to the
+ * inode status change time (aka ctime). i_version must appear to change, even
+ * if the ctime does not (since the whole point is to avoid missing updates due
+ * to timestamp granularity). If POSIX or other relevant spec mandates that the
+ * ctime must change due to an operation, then the i_version counter must be
+ * incremented as well.
+ *
+ * Making the i_version update completely atomic with the operation itself would
+ * be prohibitively expensive. Traditionally the kernel has updated the times on
+ * directories after an operation that changes its contents. For regular files,
+ * the ctime is usually updated before the data is copied into the cache for a
+ * write. This means that there is a window of time when an observer can
+ * associate a new timestamp with old file contents. Since the purpose of the
+ * i_version is to allow for better cache coherency, the i_version must always
+ * be updated after the results of the operation are visible. Updating it before
+ * and after a change is also permitted. (Note that no filesystems currently do
+ * this. Fixing that is a work-in-progress).
  *
  * Observers see the i_version as a 64-bit number that never decreases. If it
  * remains the same since it was last checked, then nothing has changed in the
@@ -234,42 +252,6 @@ inode_peek_iversion(const struct inode *inode)
 	return inode_peek_iversion_raw(inode) >> I_VERSION_QUERIED_SHIFT;
 }
 
-/**
- * inode_query_iversion - read i_version for later use
- * @inode: inode from which i_version should be read
- *
- * Read the inode i_version counter. This should be used by callers that wish
- * to store the returned i_version for later comparison. This will guarantee
- * that a later query of the i_version will result in a different value if
- * anything has changed.
- *
- * In this implementation, we fetch the current value, set the QUERIED flag and
- * then try to swap it into place with a cmpxchg, if it wasn't already set. If
- * that fails, we try again with the newly fetched value from the cmpxchg.
- */
-static inline u64
-inode_query_iversion(struct inode *inode)
-{
-	u64 cur, new;
-
-	cur = inode_peek_iversion_raw(inode);
-	do {
-		/* If flag is already set, then no need to swap */
-		if (cur & I_VERSION_QUERIED) {
-			/*
-			 * This barrier (and the implicit barrier in the
-			 * cmpxchg below) pairs with the barrier in
-			 * inode_maybe_inc_iversion().
-			 */
-			smp_mb();
-			break;
-		}
-
-		new = cur | I_VERSION_QUERIED;
-	} while (!atomic64_try_cmpxchg(&inode->i_version, &cur, new));
-	return cur >> I_VERSION_QUERIED_SHIFT;
-}
-
 /*
  * For filesystems without any sort of change attribute, the best we can
  * do is fake one up from the ctime:
@@ -283,6 +265,8 @@ static inline u64 time_to_chattr(struct timespec64 *t)
 	return chattr;
 }
 
+u64 inode_query_iversion(struct inode *inode);
+
 /**
  * inode_eq_iversion_raw - check whether the raw i_version counter has changed
  * @inode: inode to check
diff --git a/include/linux/stat.h b/include/linux/stat.h
index ff277ced50e9..52150570d37a 100644
--- a/include/linux/stat.h
+++ b/include/linux/stat.h
@@ -52,6 +52,15 @@ struct kstat {
 	u64		mnt_id;
 	u32		dio_mem_align;
 	u32		dio_offset_align;
+	u64		change_cookie;
 };
 
+/* These definitions are internal to the kernel for now. Mainly used by nfsd. */
+
+/* mask values */
+#define STATX_CHANGE_COOKIE		0x40000000U	/* Want/got stx_change_attr */
+
+/* file attribute values */
+#define STATX_ATTR_CHANGE_MONOTONIC	0x8000000000000000ULL /* version monotonically increases */
+
 #endif