[PATCH v8 00/12] fs: multigrain timestamp redux

[PATCH v8 00/12] fs: multigrain timestamp redux

[Jeff Layton]

This is a replacement for the v6 series sitting in Christian's
vfs.mgtime branch.

The kernel test robot reported a performance regression in v6 due to the
changes to current_time(). This patchset addresses that by moving the
ctime floor handling into the timekeeper code, which allows us to avoid
multiple seqcount loops when fetching and converting times. The basic
approach is still the same. The only difference is in where the
timestamp floor is handled, and in how we get new timestamps.

This reduces the changes to fs/inode.c and avoids a lot of the messiness
of handling both timespec64's and ktime_t values.

The pipe1_threads test shows these averages on my test rig:

    v6.11-rc7				103561295 (baseline)
    v6.11-rc7 + v6 series		95995565  (~7% slower)
    v6.11-rc7 + v7 series		101357203 (~2% slower)

...so the performance difference here is significant.

The main difference between v6 and v7 is in the first two patches, so
I've dropped the R-b's from those. The rest I left intact.

Note that there is one additional patch in this series (#4) that adds
support for handling delegated timestamps. The patches that make use of
that are in Chuck's nfsd-next branch. Taking that in here should make
that merge easier.

R-b's would be welcome (particularly from the timekeeper folks).

Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
Changes in v8:
- split patch that adds percpu counters into fs and timekeeping patches
- convert percpu counters to unsigned longs
- directly access the offs_real value in timekeeper instead of going
  through offsets array
- drop WARN_ON's in timekeeping patches
- better changelogs and more comments for the timekeeping bits
- better document how backward realtime clock jumps affect things
- Link to v7: https://lore.kernel.org/r/20240913-mgtime-v7-0-92d4020e3b00@kernel.org

Changes in v7:
- move the floor value handling into timekeeper for better performance
- Link to v6: https://lore.kernel.org/r/20240715-mgtime-v6-0-48e5d34bd2ba@kernel.org

Changes in v6:
- Normalize timespec64 in inode_set_ctime_to_ts
- use DEFINE_PER_CPU counters for better vfs consistency
- skip ctime cmpxchg if the result means nothing will change
- add trace_ctime_xchg_skip to track skipped ctime updates
- use __print_flags in ctime_ns_xchg tracepoint
- Link to v5: https://lore.kernel.org/r/20240711-mgtime-v5-0-37bb5b465feb@kernel.org

Changes in v5:
- refetch coarse time in coarse_ctime if not returning floor
- timestamp_truncate before swapping new ctime value into place
- track floor value as atomic64_t
- cleanups to Documentation file
- Link to v4: https://lore.kernel.org/r/20240708-mgtime-v4-0-a0f3c6fb57f3@kernel.org

Changes in v4:
- reordered tracepoint fields for better packing
- rework percpu counters again to also count fine grained timestamps
- switch to try_cmpxchg for better efficiency
- Link to v3: https://lore.kernel.org/r/20240705-mgtime-v3-0-85b2daa9b335@kernel.org

Changes in v3:
- Drop the conversion of i_ctime fields to ktime_t, and use an unused bit
  of the i_ctime_nsec field as QUERIED flag.
- Better tracepoints for tracking floor and ctime updates
- Reworked percpu counters to be more useful
- Track floor as monotonic value, which eliminates clock-jump problem

Changes in v2:
- Added Documentation file
- Link to v1: https://lore.kernel.org/r/20240626-mgtime-v1-0-a189352d0f8f@kernel.org

---
Jeff Layton (12):
      timekeeping: add interfaces for handling timestamps with a floor value
      fs: add infrastructure for multigrain timestamps
      fs: have setattr_copy handle multigrain timestamps appropriately
      fs: handle delegated timestamps in setattr_copy_mgtime
      fs: tracepoints around multigrain timestamp events
      fs: add percpu counters for significant multigrain timestamp events
      timekeeping: add percpu counter for tracking floor swap events
      Documentation: add a new file documenting multigrain timestamps
      xfs: switch to multigrain timestamps
      ext4: switch to multigrain timestamps
      btrfs: convert to multigrain timestamps
      tmpfs: add support for multigrain timestamps

 Documentation/filesystems/index.rst         |   1 +
 Documentation/filesystems/multigrain-ts.rst | 125 +++++++++++++
 fs/attr.c                                   |  60 +++++-
 fs/btrfs/file.c                             |  25 +--
 fs/btrfs/super.c                            |   3 +-
 fs/ext4/super.c                             |   2 +-
 fs/inode.c                                  | 278 +++++++++++++++++++++++++---
 fs/stat.c                                   |  42 ++++-
 fs/xfs/libxfs/xfs_trans_inode.c             |   6 +-
 fs/xfs/xfs_iops.c                           |  10 +-
 fs/xfs/xfs_super.c                          |   2 +-
 include/linux/fs.h                          |  36 +++-
 include/linux/timekeeping.h                 |   5 +
 include/trace/events/timestamp.h            | 124 +++++++++++++
 kernel/time/timekeeping.c                   |  97 ++++++++++
 kernel/time/timekeeping_debug.c             |  13 ++
 kernel/time/timekeeping_internal.h          |   9 +
 mm/shmem.c                                  |   2 +-
 18 files changed, 767 insertions(+), 73 deletions(-)
---
base-commit: 9852d85ec9d492ebef56dc5f229416c925758edc
change-id: 20240913-mgtime-20c98bcda88e

Best regards,
-- 
Jeff Layton <jlayton@kernel.org>

[PATCH v8 01/12] timekeeping: add interfaces for handling timestamps with a floor value

[Jeff Layton]

Multigrain timestamps allow the kernel to use fine-grained timestamps
when an inode's attributes is being actively observed via ->getattr().
With this support, it's possible for a file to get a fine-grained
timestamp, and another modified after it to get a coarse-grained stamp
that is earlier than the fine-grained time.  If this happens then the
files can appear to have been modified in reverse order, which breaks
VFS ordering guarantees.

To prevent this, maintain a floor value for multigrain timestamps.
Whenever a fine-grained timestamp is handed out, record it, and when
coarse-grained stamps are handed out, ensure they are not earlier than
that value. If the coarse-grained timestamp is earlier than the
fine-grained floor, return the floor value instead.

Add a static singleton atomic64_t into timekeeper.c that we can use to
keep track of the latest fine-grained time ever handed out. This is
tracked as a monotonic ktime_t value to ensure that it isn't affected by
clock jumps. Because it is updated at different times than the rest of
the timekeeper object, the floor value is managed independently of the
timekeeper via a cmpxchg() operation, and sits on its own cacheline.

This patch also adds two new public interfaces:

- ktime_get_coarse_real_ts64_mg() fills a timespec64 with the later of the
  coarse-grained clock and the floor time

- ktime_get_real_ts64_mg() gets the fine-grained clock value, and tries
  to swap it into the floor. A timespec64 is filled with the result.

Since the floor is global, take care to avoid updating it unless it's
absolutely necessary. If we do the cmpxchg and find that the value has
been updated since we fetched it, then we discard the fine-grained time
that was fetched in favor of the recent update.

Note that the VFS ordering guarantees assume that the realtime clock
does not experience a backward jump. POSIX requires that we stamp files
using realtime clock values, so if a backward clock jump occurs, then
files can appear to have been modified in reverse order.

Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 include/linux/timekeeping.h |  4 ++
 kernel/time/timekeeping.c   | 96 +++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 100 insertions(+)

diff --git a/include/linux/timekeeping.h b/include/linux/timekeeping.h
index fc12a9ba2c88..7aa85246c183 100644
--- a/include/linux/timekeeping.h
+++ b/include/linux/timekeeping.h
@@ -45,6 +45,10 @@ extern void ktime_get_real_ts64(struct timespec64 *tv);
 extern void ktime_get_coarse_ts64(struct timespec64 *ts);
 extern void ktime_get_coarse_real_ts64(struct timespec64 *ts);
 
+/* Multigrain timestamp interfaces */
+extern void ktime_get_coarse_real_ts64_mg(struct timespec64 *ts);
+extern void ktime_get_real_ts64_mg(struct timespec64 *ts);
+
 void getboottime64(struct timespec64 *ts);
 
 /*
diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
index 7e6f409bf311..37004a4758cf 100644
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -114,6 +114,22 @@ static struct tk_fast tk_fast_raw  ____cacheline_aligned = {
 	.base[1] = FAST_TK_INIT,
 };
 
+/*
+ * Multigrain timestamps require that we keep track of the latest fine-grained
+ * timestamp that has been issued, and never return a coarse-grained timestamp
+ * that is earlier than that value.
+ *
+ * mg_floor represents the latest fine-grained time that we have handed out as
+ * a timestamp on the system. Tracked as a monotonic ktime_t, and converted to
+ * the realtime clock on an as-needed basis.
+ *
+ * This ensures that we never issue a timestamp earlier than one that has
+ * already been issued, as long as the realtime clock never experiences a
+ * backward clock jump. If the realtime clock is set to an earlier time, then
+ * realtime timestamps can appear to go backward.
+ */
+static __cacheline_aligned_in_smp atomic64_t mg_floor;
+
 static inline void tk_normalize_xtime(struct timekeeper *tk)
 {
 	while (tk->tkr_mono.xtime_nsec >= ((u64)NSEC_PER_SEC << tk->tkr_mono.shift)) {
@@ -2394,6 +2410,86 @@ void ktime_get_coarse_real_ts64(struct timespec64 *ts)
 }
 EXPORT_SYMBOL(ktime_get_coarse_real_ts64);
 
+/**
+ * ktime_get_coarse_real_ts64_mg - return latter of coarse grained time or floor
+ * @ts: timespec64 to be filled
+ *
+ * Fetch the global mg_floor value, convert it to realtime and
+ * compare it to the current coarse-grained time. Fill @ts with
+ * whichever is latest. Note that this is a filesystem-specific
+ * interface and should be avoided outside of that context.
+ */
+void ktime_get_coarse_real_ts64_mg(struct timespec64 *ts)
+{
+	struct timekeeper *tk = &tk_core.timekeeper;
+	u64 floor = atomic64_read(&mg_floor);
+	ktime_t f_real, offset, coarse;
+	unsigned int seq;
+
+	do {
+		seq = read_seqcount_begin(&tk_core.seq);
+		*ts = tk_xtime(tk);
+		offset = tk_core.timekeeper.offs_real;
+	} while (read_seqcount_retry(&tk_core.seq, seq));
+
+	coarse = timespec64_to_ktime(*ts);
+	f_real = ktime_add(floor, offset);
+	if (ktime_after(f_real, coarse))
+		*ts = ktime_to_timespec64(f_real);
+}
+EXPORT_SYMBOL_GPL(ktime_get_coarse_real_ts64_mg);
+
+/**
+ * ktime_get_real_ts64_mg - attempt to update floor value and return result
+ * @ts:		pointer to the timespec to be set
+ *
+ * Get a monotonic fine-grained time value and attempt to swap it into the
+ * floor. If it succeeds then accept the new floor value. If it fails
+ * then another task raced in during the interim time and updated the floor.
+ * That value is just as valid, so accept that value in this case.
+ *
+ * @ts will be filled with the resulting floor value, regardless of
+ * the outcome of the swap. Note that this is a filesystem specific interface
+ * and should be avoided outside of that context.
+ */
+void ktime_get_real_ts64_mg(struct timespec64 *ts)
+{
+	struct timekeeper *tk = &tk_core.timekeeper;
+	ktime_t old = atomic64_read(&mg_floor);
+	ktime_t offset, mono;
+	unsigned int seq;
+	u64 nsecs;
+
+	do {
+		seq = read_seqcount_begin(&tk_core.seq);
+
+		ts->tv_sec = tk->xtime_sec;
+		mono = tk->tkr_mono.base;
+		nsecs = timekeeping_get_ns(&tk->tkr_mono);
+		offset = tk_core.timekeeper.offs_real;
+	} while (read_seqcount_retry(&tk_core.seq, seq));
+
+	mono = ktime_add_ns(mono, nsecs);
+
+	/*
+	 * Attempt to update the floor with the new time value. Accept the
+	 * resulting floor value regardless of the outcome of the swap.
+	 */
+	if (atomic64_try_cmpxchg(&mg_floor, &old, mono)) {
+		ts->tv_nsec = 0;
+		timespec64_add_ns(ts, nsecs);
+	} else {
+		/*
+		 * Something has changed mg_floor since "old" was
+		 * fetched. "old" has now been updated with the
+		 * current value of mg_floor, so use that to return
+		 * the current coarse floor value.
+		 */
+		*ts = ktime_to_timespec64(ktime_add(old, offset));
+	}
+}
+EXPORT_SYMBOL_GPL(ktime_get_real_ts64_mg);
+
 void ktime_get_coarse_ts64(struct timespec64 *ts)
 {
 	struct timekeeper *tk = &tk_core.timekeeper;

-- 
2.46.2

[PATCH v8 02/12] fs: add infrastructure for multigrain timestamps

[Jeff Layton]

The VFS has always used coarse-grained timestamps when updating the
ctime and mtime after a change. This has the benefit of allowing
filesystems to optimize away a lot metadata updates, down to around 1
per jiffy, even when a file is under heavy writes.

Unfortunately, this has always been an issue when we're exporting via
NFSv3, which relies on timestamps to validate caches. A lot of changes
can happen in a jiffy, so timestamps aren't sufficient to help the
client decide when to invalidate the cache. Even with NFSv4, a lot of
exported filesystems don't properly support a change attribute and are
subject to the same problems with timestamp granularity. Other
applications have similar issues with timestamps (e.g backup
applications).

If we were to always use fine-grained timestamps, that would improve the
situation, but that becomes rather expensive, as the underlying
filesystem would have to log a lot more metadata updates.

What we need is a way to only use fine-grained timestamps when they are
being actively queried. Use the (unused) top bit in inode->i_ctime_nsec
as a flag that indicates whether the current timestamps have been
queried via stat() or the like. When it's set, we allow the kernel to
use a fine-grained timestamp iff it's necessary to make the ctime show
a different value.

This solves the problem of being able to distinguish the timestamp
between updates, but introduces a new problem: it's now possible for a
file being changed to get a fine-grained timestamp. A file that is
altered just a bit later can then get a coarse-grained one that appears
older than the earlier fine-grained time. This violates timestamp
ordering guarantees.

To remedy this, keep a global monotonic atomic64_t value that acts as a
timestamp floor.  When we go to stamp a file, we first get the latter of
the current floor value and the current coarse-grained time. If the
inode ctime hasn't been queried then we just attempt to stamp it with
that value.

If it has been queried, then first see whether the current coarse time
is later than the existing ctime. If it is, then we accept that value.
If it isn't, then we get a fine-grained timestamp.

Filesystems can opt into this by setting the FS_MGTIME fstype flag.
Others should be unaffected (other than being subject to the same floor
value as multigrain filesystems).

Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/inode.c         | 137 +++++++++++++++++++++++++++++++++++++++++++----------
 fs/stat.c          |  39 ++++++++++++++-
 include/linux/fs.h |  34 +++++++++----
 3 files changed, 175 insertions(+), 35 deletions(-)

diff --git a/fs/inode.c b/fs/inode.c
index 471ae4a31549..4ec1e71e9a9d 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -2209,19 +2209,58 @@ int file_remove_privs(struct file *file)
 }
 EXPORT_SYMBOL(file_remove_privs);
 
+/**
+ * current_time - Return FS time (possibly fine-grained)
+ * @inode: inode.
+ *
+ * Return the current time truncated to the time granularity supported by
+ * the fs, as suitable for a ctime/mtime change. If the ctime is flagged
+ * as having been QUERIED, get a fine-grained timestamp, but don't update
+ * the floor.
+ *
+ * For a multigrain inode, this is effectively an estimate of the timestamp
+ * that a file would receive. An actual update must go through
+ * inode_set_ctime_current().
+ */
+struct timespec64 current_time(struct inode *inode)
+{
+	struct timespec64 now;
+	u32 cns;
+
+	ktime_get_coarse_real_ts64_mg(&now);
+
+	if (!is_mgtime(inode))
+		goto out;
+
+	/* If nothing has queried it, then coarse time is fine */
+	cns = smp_load_acquire(&inode->i_ctime_nsec);
+	if (cns & I_CTIME_QUERIED) {
+		/*
+		 * If there is no apparent change, then get a fine-grained
+		 * timestamp.
+		 */
+		if (now.tv_nsec == (cns & ~I_CTIME_QUERIED))
+			ktime_get_real_ts64(&now);
+	}
+out:
+	return timestamp_truncate(now, inode);
+}
+EXPORT_SYMBOL(current_time);
+
 static int inode_needs_update_time(struct inode *inode)
 {
+	struct timespec64 now, ts;
 	int sync_it = 0;
-	struct timespec64 now = current_time(inode);
-	struct timespec64 ts;
 
 	/* First try to exhaust all avenues to not sync */
 	if (IS_NOCMTIME(inode))
 		return 0;
 
+	now = current_time(inode);
+
 	ts = inode_get_mtime(inode);
 	if (!timespec64_equal(&ts, &now))
-		sync_it = S_MTIME;
+		sync_it |= S_MTIME;
 
 	ts = inode_get_ctime(inode);
 	if (!timespec64_equal(&ts, &now))
@@ -2598,6 +2637,15 @@ void inode_nohighmem(struct inode *inode)
 }
 EXPORT_SYMBOL(inode_nohighmem);
 
+struct timespec64 inode_set_ctime_to_ts(struct inode *inode, struct timespec64 ts)
+{
+	set_normalized_timespec64(&ts, ts.tv_sec, ts.tv_nsec);
+	inode->i_ctime_sec = ts.tv_sec;
+	inode->i_ctime_nsec = ts.tv_nsec;
+	return ts;
+}
+EXPORT_SYMBOL(inode_set_ctime_to_ts);
+
 /**
  * timestamp_truncate - Truncate timespec to a granularity
  * @t: Timespec
@@ -2630,36 +2678,75 @@ struct timespec64 timestamp_truncate(struct timespec64 t, struct inode *inode)
 EXPORT_SYMBOL(timestamp_truncate);
 
 /**
- * current_time - Return FS time
- * @inode: inode.
+ * inode_set_ctime_current - set the ctime to current_time
+ * @inode: inode
  *
- * Return the current time truncated to the time granularity supported by
- * the fs.
+ * Set the inode's ctime to the current value for the inode. Returns the
+ * current value that was assigned. If this is not a multigrain inode, then we
+ * set it to the later of the coarse time and floor value.
  *
- * Note that inode and inode->sb cannot be NULL.
- * Otherwise, the function warns and returns time without truncation.
+ * If it is multigrain, then we first see if the coarse-grained timestamp is
+ * distinct from what we have. If so, then we'll just use that. If we have to
+ * get a fine-grained timestamp, then do so, and try to swap it into the floor.
+ * We accept the new floor value regardless of the outcome of the cmpxchg.
+ * After that, we try to swap the new value into i_ctime_nsec. Again, we take
+ * the resulting ctime, regardless of the outcome of the swap.
  */
-struct timespec64 current_time(struct inode *inode)
+struct timespec64 inode_set_ctime_current(struct inode *inode)
 {
 	struct timespec64 now;
+	u32 cns, cur;
 
-	ktime_get_coarse_real_ts64(&now);
-	return timestamp_truncate(now, inode);
-}
-EXPORT_SYMBOL(current_time);
+	ktime_get_coarse_real_ts64_mg(&now);
+	now = timestamp_truncate(now, inode);
 
-/**
- * inode_set_ctime_current - set the ctime to current_time
- * @inode: inode
- *
- * Set the inode->i_ctime to the current value for the inode. Returns
- * the current value that was assigned to i_ctime.
- */
-struct timespec64 inode_set_ctime_current(struct inode *inode)
-{
-	struct timespec64 now = current_time(inode);
+	/* Just return that if this is not a multigrain fs */
+	if (!is_mgtime(inode)) {
+		inode_set_ctime_to_ts(inode, now);
+		goto out;
+	}
 
-	inode_set_ctime_to_ts(inode, now);
+	/*
+	 * We only need a fine-grained time if someone has queried it,
+	 * and the current coarse grained time isn't later than what's
+	 * already there.
+	 */
+	cns = smp_load_acquire(&inode->i_ctime_nsec);
+	if (cns & I_CTIME_QUERIED) {
+		struct timespec64 ctime = { .tv_sec = inode->i_ctime_sec,
+					    .tv_nsec = cns & ~I_CTIME_QUERIED };
+
+		if (timespec64_compare(&now, &ctime) <= 0) {
+			ktime_get_real_ts64_mg(&now);
+			now = timestamp_truncate(now, inode);
+		}
+	}
+
+	/* No need to cmpxchg if it's exactly the same */
+	if (cns == now.tv_nsec && inode->i_ctime_sec == now.tv_sec)
+		goto out;
+	cur = cns;
+retry:
+	/* Try to swap the nsec value into place. */
+	if (try_cmpxchg(&inode->i_ctime_nsec, &cur, now.tv_nsec)) {
+		/* If swap occurred, then we're (mostly) done */
+		inode->i_ctime_sec = now.tv_sec;
+	} else {
+		/*
+		 * Was the change due to someone marking the old ctime QUERIED?
+		 * If so then retry the swap. This can only happen once since
+		 * the only way to clear I_CTIME_QUERIED is to stamp the inode
+		 * with a new ctime.
+		 */
+		if (!(cns & I_CTIME_QUERIED) && (cns | I_CTIME_QUERIED) == cur) {
+			cns = cur;
+			goto retry;
+		}
+		/* Otherwise, keep the existing ctime */
+		now.tv_sec = inode->i_ctime_sec;
+		now.tv_nsec = cur & ~I_CTIME_QUERIED;
+	}
+out:
 	return now;
 }
 EXPORT_SYMBOL(inode_set_ctime_current);
diff --git a/fs/stat.c b/fs/stat.c
index 41e598376d7e..381926fb405f 100644
--- a/fs/stat.c
+++ b/fs/stat.c
@@ -26,6 +26,35 @@
 #include "internal.h"
 #include "mount.h"
 
+/**
+ * fill_mg_cmtime - Fill in the mtime and ctime and flag ctime as QUERIED
+ * @stat: where to store the resulting values
+ * @request_mask: STATX_* values requested
+ * @inode: inode from which to grab the c/mtime
+ *
+ * Given @inode, grab the ctime and mtime out if it and store the result
+ * in @stat. When fetching the value, flag it as QUERIED (if not already)
+ * so the next write will record a distinct timestamp.
+ */
+void fill_mg_cmtime(struct kstat *stat, u32 request_mask, struct inode *inode)
+{
+	atomic_t *pcn = (atomic_t *)&inode->i_ctime_nsec;
+
+	/* If neither time was requested, then don't report them */
+	if (!(request_mask & (STATX_CTIME|STATX_MTIME))) {
+		stat->result_mask &= ~(STATX_CTIME|STATX_MTIME);
+		return;
+	}
+
+	stat->mtime = inode_get_mtime(inode);
+	stat->ctime.tv_sec = inode->i_ctime_sec;
+	stat->ctime.tv_nsec = (u32)atomic_read(pcn);
+	if (!(stat->ctime.tv_nsec & I_CTIME_QUERIED))
+		stat->ctime.tv_nsec = ((u32)atomic_fetch_or(I_CTIME_QUERIED, pcn));
+	stat->ctime.tv_nsec &= ~I_CTIME_QUERIED;
+}
+EXPORT_SYMBOL(fill_mg_cmtime);
+
 /**
  * generic_fillattr - Fill in the basic attributes from the inode struct
  * @idmap:		idmap of the mount the inode was found from
@@ -58,8 +87,14 @@ void generic_fillattr(struct mnt_idmap *idmap, u32 request_mask,
 	stat->rdev = inode->i_rdev;
 	stat->size = i_size_read(inode);
 	stat->atime = inode_get_atime(inode);
-	stat->mtime = inode_get_mtime(inode);
-	stat->ctime = inode_get_ctime(inode);
+
+	if (is_mgtime(inode)) {
+		fill_mg_cmtime(stat, request_mask, inode);
+	} else {
+		stat->ctime = inode_get_ctime(inode);
+		stat->mtime = inode_get_mtime(inode);
+	}
+
 	stat->blksize = i_blocksize(inode);
 	stat->blocks = inode->i_blocks;
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index e3c603d01337..23908bad166c 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1653,6 +1653,17 @@ static inline struct timespec64 inode_set_mtime(struct inode *inode,
 	return inode_set_mtime_to_ts(inode, ts);
 }
 
+/*
+ * Multigrain timestamps
+ *
+ * Conditionally use fine-grained ctime and mtime timestamps when there
+ * are users actively observing them via getattr. The primary use-case
+ * for this is NFS clients that use the ctime to distinguish between
+ * different states of the file, and that are often fooled by multiple
+ * operations that occur in the same coarse-grained timer tick.
+ */
+#define I_CTIME_QUERIED		((u32)BIT(31))
+
 static inline time64_t inode_get_ctime_sec(const struct inode *inode)
 {
 	return inode->i_ctime_sec;
@@ -1660,7 +1671,7 @@ static inline time64_t inode_get_ctime_sec(const struct inode *inode)
 
 static inline long inode_get_ctime_nsec(const struct inode *inode)
 {
-	return inode->i_ctime_nsec;
+	return inode->i_ctime_nsec & ~I_CTIME_QUERIED;
 }
 
 static inline struct timespec64 inode_get_ctime(const struct inode *inode)
@@ -1671,13 +1682,7 @@ static inline struct timespec64 inode_get_ctime(const struct inode *inode)
 	return ts;
 }
 
-static inline struct timespec64 inode_set_ctime_to_ts(struct inode *inode,
-						      struct timespec64 ts)
-{
-	inode->i_ctime_sec = ts.tv_sec;
-	inode->i_ctime_nsec = ts.tv_nsec;
-	return ts;
-}
+struct timespec64 inode_set_ctime_to_ts(struct inode *inode, struct timespec64 ts);
 
 /**
  * inode_set_ctime - set the ctime in the inode
@@ -2542,6 +2547,7 @@ struct file_system_type {
 #define FS_USERNS_MOUNT		8	/* Can be mounted by userns root */
 #define FS_DISALLOW_NOTIFY_PERM	16	/* Disable fanotify permission events */
 #define FS_ALLOW_IDMAP         32      /* FS has been updated to handle vfs idmappings. */
+#define FS_MGTIME		64	/* FS uses multigrain timestamps */
 #define FS_RENAME_DOES_D_MOVE	32768	/* FS will handle d_move() during rename() internally. */
 	int (*init_fs_context)(struct fs_context *);
 	const struct fs_parameter_spec *parameters;
@@ -2565,6 +2571,17 @@ struct file_system_type {
 
 #define MODULE_ALIAS_FS(NAME) MODULE_ALIAS("fs-" NAME)
 
+/**
+ * is_mgtime: is this inode using multigrain timestamps
+ * @inode: inode to test for multigrain timestamps
+ *
+ * Return true if the inode uses multigrain timestamps, false otherwise.
+ */
+static inline bool is_mgtime(const struct inode *inode)
+{
+	return inode->i_sb->s_type->fs_flags & FS_MGTIME;
+}
+
 extern struct dentry *mount_bdev(struct file_system_type *fs_type,
 	int flags, const char *dev_name, void *data,
 	int (*fill_super)(struct super_block *, void *, int));
@@ -3321,6 +3338,7 @@ extern void page_put_link(void *);
 extern int page_symlink(struct inode *inode, const char *symname, int len);
 extern const struct inode_operations page_symlink_inode_operations;
 extern void kfree_link(void *);
+void fill_mg_cmtime(struct kstat *stat, u32 request_mask, struct inode *inode);
 void generic_fillattr(struct mnt_idmap *, u32, struct inode *, struct kstat *);
 void generic_fill_statx_attr(struct inode *inode, struct kstat *stat);
 void generic_fill_statx_atomic_writes(struct kstat *stat,

-- 
2.46.2

[PATCH v8 03/12] fs: have setattr_copy handle multigrain timestamps appropriately

[Jeff Layton]

The setattr codepath is still using coarse-grained timestamps, even on
multigrain filesystems. To fix this, we need to fetch the timestamp for
ctime updates later, at the point where the assignment occurs in
setattr_copy.

On a multigrain inode, ignore the ia_ctime in the attrs, and always
update the ctime to the current clock value. Update the atime and mtime
with the same value (if needed) unless they are being set to other
specific values, a'la utimes().

Note that we don't want to do this universally however, as some
filesystems (e.g. most networked fs) want to do an explicit update
elsewhere before updating the local inode.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/attr.c | 52 ++++++++++++++++++++++++++++++++++++++++++++++------
 1 file changed, 46 insertions(+), 6 deletions(-)

diff --git a/fs/attr.c b/fs/attr.c
index c04d19b58f12..3bcbc45708a3 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -271,6 +271,42 @@ int inode_newsize_ok(const struct inode *inode, loff_t offset)
 }
 EXPORT_SYMBOL(inode_newsize_ok);
 
+/**
+ * setattr_copy_mgtime - update timestamps for mgtime inodes
+ * @inode: inode timestamps to be updated
+ * @attr: attrs for the update
+ *
+ * With multigrain timestamps, we need to take more care to prevent races
+ * when updating the ctime. Always update the ctime to the very latest
+ * using the standard mechanism, and use that to populate the atime and
+ * mtime appropriately (unless we're setting those to specific values).
+ */
+static void setattr_copy_mgtime(struct inode *inode, const struct iattr *attr)
+{
+	unsigned int ia_valid = attr->ia_valid;
+	struct timespec64 now;
+
+	/*
+	 * If the ctime isn't being updated then nothing else should be
+	 * either.
+	 */
+	if (!(ia_valid & ATTR_CTIME)) {
+		WARN_ON_ONCE(ia_valid & (ATTR_ATIME|ATTR_MTIME));
+		return;
+	}
+
+	now = inode_set_ctime_current(inode);
+	if (ia_valid & ATTR_ATIME_SET)
+		inode_set_atime_to_ts(inode, attr->ia_atime);
+	else if (ia_valid & ATTR_ATIME)
+		inode_set_atime_to_ts(inode, now);
+
+	if (ia_valid & ATTR_MTIME_SET)
+		inode_set_mtime_to_ts(inode, attr->ia_mtime);
+	else if (ia_valid & ATTR_MTIME)
+		inode_set_mtime_to_ts(inode, now);
+}
+
 /**
  * setattr_copy - copy simple metadata updates into the generic inode
  * @idmap:	idmap of the mount the inode was found from
@@ -303,12 +339,6 @@ void setattr_copy(struct mnt_idmap *idmap, struct inode *inode,
 
 	i_uid_update(idmap, attr, inode);
 	i_gid_update(idmap, attr, inode);
-	if (ia_valid & ATTR_ATIME)
-		inode_set_atime_to_ts(inode, attr->ia_atime);
-	if (ia_valid & ATTR_MTIME)
-		inode_set_mtime_to_ts(inode, attr->ia_mtime);
-	if (ia_valid & ATTR_CTIME)
-		inode_set_ctime_to_ts(inode, attr->ia_ctime);
 	if (ia_valid & ATTR_MODE) {
 		umode_t mode = attr->ia_mode;
 		if (!in_group_or_capable(idmap, inode,
@@ -316,6 +346,16 @@ void setattr_copy(struct mnt_idmap *idmap, struct inode *inode,
 			mode &= ~S_ISGID;
 		inode->i_mode = mode;
 	}
+
+	if (is_mgtime(inode))
+		return setattr_copy_mgtime(inode, attr);
+
+	if (ia_valid & ATTR_ATIME)
+		inode_set_atime_to_ts(inode, attr->ia_atime);
+	if (ia_valid & ATTR_MTIME)
+		inode_set_mtime_to_ts(inode, attr->ia_mtime);
+	if (ia_valid & ATTR_CTIME)
+		inode_set_ctime_to_ts(inode, attr->ia_ctime);
 }
 EXPORT_SYMBOL(setattr_copy);
 

-- 
2.46.2

[PATCH v8 04/12] fs: handle delegated timestamps in setattr_copy_mgtime

[Jeff Layton]

When updating the ctime on an inode for a SETATTR with a multigrain
filesystem, we usually want to take the latest time we can get for the
ctime. The exception to this rule is when there is a nfsd write
delegation and the server is proxying timestamps from the client.

When nfsd gets a CB_GETATTR response, we want to update the timestamp
value in the inode to the values that the client is tracking. The client
doesn't send a ctime value (since that's always determined by the
exported filesystem), but it can send a mtime value. In the case where
it does, then we may need to update the ctime to a value commensurate
with that instead of the current time.

If ATTR_DELEG is set, then use ia_ctime value instead of setting the
timestamp to the current time.

With the addition of delegated timestamps we can also receive a request
to update only the atime, but we may not need to set the ctime. Trust
the ATTR_CTIME flag in the update and only update the ctime when it's
set.

Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/attr.c          | 28 +++++++++++++--------
 fs/inode.c         | 72 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/fs.h |  2 ++
 3 files changed, 92 insertions(+), 10 deletions(-)

diff --git a/fs/attr.c b/fs/attr.c
index 3bcbc45708a3..392eb62aa609 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -286,16 +286,20 @@ static void setattr_copy_mgtime(struct inode *inode, const struct iattr *attr)
 	unsigned int ia_valid = attr->ia_valid;
 	struct timespec64 now;
 
-	/*
-	 * If the ctime isn't being updated then nothing else should be
-	 * either.
-	 */
-	if (!(ia_valid & ATTR_CTIME)) {
-		WARN_ON_ONCE(ia_valid & (ATTR_ATIME|ATTR_MTIME));
-		return;
+	if (ia_valid & ATTR_CTIME) {
+		/*
+		 * In the case of an update for a write delegation, we must respect
+		 * the value in ia_ctime and not use the current time.
+		 */
+		if (ia_valid & ATTR_DELEG)
+			now = inode_set_ctime_deleg(inode, attr->ia_ctime);
+		else
+			now = inode_set_ctime_current(inode);
+	} else {
+		/* If ATTR_CTIME isn't set, then ATTR_MTIME shouldn't be either. */
+		WARN_ON_ONCE(ia_valid & ATTR_MTIME);
 	}
 
-	now = inode_set_ctime_current(inode);
 	if (ia_valid & ATTR_ATIME_SET)
 		inode_set_atime_to_ts(inode, attr->ia_atime);
 	else if (ia_valid & ATTR_ATIME)
@@ -354,8 +358,12 @@ void setattr_copy(struct mnt_idmap *idmap, struct inode *inode,
 		inode_set_atime_to_ts(inode, attr->ia_atime);
 	if (ia_valid & ATTR_MTIME)
 		inode_set_mtime_to_ts(inode, attr->ia_mtime);
-	if (ia_valid & ATTR_CTIME)
-		inode_set_ctime_to_ts(inode, attr->ia_ctime);
+	if (ia_valid & ATTR_CTIME) {
+		if (ia_valid & ATTR_DELEG)
+			inode_set_ctime_deleg(inode, attr->ia_ctime);
+		else
+			inode_set_ctime_to_ts(inode, attr->ia_ctime);
+	}
 }
 EXPORT_SYMBOL(setattr_copy);
 
diff --git a/fs/inode.c b/fs/inode.c
index 4ec1e71e9a9d..7a324d999816 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -2751,6 +2751,78 @@ struct timespec64 inode_set_ctime_current(struct inode *inode)
 }
 EXPORT_SYMBOL(inode_set_ctime_current);
 
+/**
+ * inode_set_ctime_deleg - try to update the ctime on a delegated inode
+ * @inode: inode to update
+ * @update: timespec64 to set the ctime
+ *
+ * Attempt to atomically update the ctime on behalf of a delegation holder.
+ *
+ * The nfs server can call back the holder of a delegation to get updated
+ * inode attributes, including the mtime. When updating the mtime we may
+ * need to update the ctime to a value at least equal to that.
+ *
+ * This can race with concurrent updates to the inode, in which
+ * case we just don't do the update.
+ *
+ * Note that this works even when multigrain timestamps are not enabled,
+ * so use it in either case.
+ */
+struct timespec64 inode_set_ctime_deleg(struct inode *inode, struct timespec64 update)
+{
+	struct timespec64 now, cur_ts;
+	u32 cur, old;
+
+	/* pairs with try_cmpxchg below */
+	cur = smp_load_acquire(&inode->i_ctime_nsec);
+	cur_ts.tv_nsec = cur & ~I_CTIME_QUERIED;
+	cur_ts.tv_sec = inode->i_ctime_sec;
+
+	/* If the update is older than the existing value, skip it. */
+	if (timespec64_compare(&update, &cur_ts) <= 0)
+		return cur_ts;
+
+	ktime_get_coarse_real_ts64_mg(&now);
+
+	/* Clamp the update to "now" if it's in the future */
+	if (timespec64_compare(&update, &now) > 0)
+		update = now;
+
+	update = timestamp_truncate(update, inode);
+
+	/* No need to update if the values are already the same */
+	if (timespec64_equal(&update, &cur_ts))
+		return cur_ts;
+
+	/*
+	 * Try to swap the nsec value into place. If it fails, that means
+	 * we raced with an update due to a write or similar activity. That
+	 * stamp takes precedence, so just skip the update.
+	 */
+retry:
+	old = cur;
+	if (try_cmpxchg(&inode->i_ctime_nsec, &cur, update.tv_nsec)) {
+		inode->i_ctime_sec = update.tv_sec;
+		mgtime_counter_inc(mg_ctime_swaps);
+		return update;
+	}
+
+	/*
+	 * Was the change due to someone marking the old ctime QUERIED?
+	 * If so then retry the swap. This can only happen once since
+	 * the only way to clear I_CTIME_QUERIED is to stamp the inode
+	 * with a new ctime.
+	 */
+	if (!(old & I_CTIME_QUERIED) && (cur == (old | I_CTIME_QUERIED)))
+		goto retry;
+
+	/* Otherwise, it was a new timestamp. */
+	cur_ts.tv_sec = inode->i_ctime_sec;
+	cur_ts.tv_nsec = cur & ~I_CTIME_QUERIED;
+	return cur_ts;
+}
+EXPORT_SYMBOL(inode_set_ctime_deleg);
+
 /**
  * in_group_or_capable - check whether caller is CAP_FSETID privileged
  * @idmap:	idmap of the mount @inode was found from
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 23908bad166c..b1a3bd07711b 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1584,6 +1584,8 @@ static inline bool fsuidgid_has_mapping(struct super_block *sb,
 
 struct timespec64 current_time(struct inode *inode);
 struct timespec64 inode_set_ctime_current(struct inode *inode);
+struct timespec64 inode_set_ctime_deleg(struct inode *inode,
+					struct timespec64 update);
 
 static inline time64_t inode_get_atime_sec(const struct inode *inode)
 {

-- 
2.46.2

[PATCH v8 05/12] fs: tracepoints around multigrain timestamp events

[Jeff Layton]

Add some tracepoints around various multigrain timestamp events.

Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/inode.c                       |   9 ++-
 fs/stat.c                        |   3 +
 include/trace/events/timestamp.h | 124 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 135 insertions(+), 1 deletion(-)

diff --git a/fs/inode.c b/fs/inode.c
index 7a324d999816..1a7eff2a40e2 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -22,6 +22,9 @@
 #include <linux/iversion.h>
 #include <linux/rw_hint.h>
 #include <trace/events/writeback.h>
+#define CREATE_TRACE_POINTS
+#include <trace/events/timestamp.h>
+
 #include "internal.h"
 
 /*
@@ -2639,6 +2642,7 @@ EXPORT_SYMBOL(inode_nohighmem);
 
 struct timespec64 inode_set_ctime_to_ts(struct inode *inode, struct timespec64 ts)
 {
+	trace_inode_set_ctime_to_ts(inode, &ts);
 	set_normalized_timespec64(&ts, ts.tv_sec, ts.tv_nsec);
 	inode->i_ctime_sec = ts.tv_sec;
 	inode->i_ctime_nsec = ts.tv_nsec;
@@ -2723,14 +2727,17 @@ struct timespec64 inode_set_ctime_current(struct inode *inode)
 	}
 
 	/* No need to cmpxchg if it's exactly the same */
-	if (cns == now.tv_nsec && inode->i_ctime_sec == now.tv_sec)
+	if (cns == now.tv_nsec && inode->i_ctime_sec == now.tv_sec) {
+		trace_ctime_xchg_skip(inode, &now);
 		goto out;
+	}
 	cur = cns;
 retry:
 	/* Try to swap the nsec value into place. */
 	if (try_cmpxchg(&inode->i_ctime_nsec, &cur, now.tv_nsec)) {
 		/* If swap occurred, then we're (mostly) done */
 		inode->i_ctime_sec = now.tv_sec;
+		trace_ctime_ns_xchg(inode, cns, now.tv_nsec, cur);
 	} else {
 		/*
 		 * Was the change due to someone marking the old ctime QUERIED?
diff --git a/fs/stat.c b/fs/stat.c
index 381926fb405f..72914b6624a5 100644
--- a/fs/stat.c
+++ b/fs/stat.c
@@ -23,6 +23,8 @@
 #include <linux/uaccess.h>
 #include <asm/unistd.h>
 
+#include <trace/events/timestamp.h>
+
 #include "internal.h"
 #include "mount.h"
 
@@ -52,6 +54,7 @@ void fill_mg_cmtime(struct kstat *stat, u32 request_mask, struct inode *inode)
 	if (!(stat->ctime.tv_nsec & I_CTIME_QUERIED))
 		stat->ctime.tv_nsec = ((u32)atomic_fetch_or(I_CTIME_QUERIED, pcn));
 	stat->ctime.tv_nsec &= ~I_CTIME_QUERIED;
+	trace_fill_mg_cmtime(inode, &stat->ctime, &stat->mtime);
 }
 EXPORT_SYMBOL(fill_mg_cmtime);
 
diff --git a/include/trace/events/timestamp.h b/include/trace/events/timestamp.h
new file mode 100644
index 000000000000..c9e5ec930054
--- /dev/null
+++ b/include/trace/events/timestamp.h
@@ -0,0 +1,124 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM timestamp
+
+#if !defined(_TRACE_TIMESTAMP_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_TIMESTAMP_H
+
+#include <linux/tracepoint.h>
+#include <linux/fs.h>
+
+#define CTIME_QUERIED_FLAGS \
+	{ I_CTIME_QUERIED, "Q" }
+
+DECLARE_EVENT_CLASS(ctime,
+	TP_PROTO(struct inode *inode,
+		 struct timespec64 *ctime),
+
+	TP_ARGS(inode, ctime),
+
+	TP_STRUCT__entry(
+		__field(dev_t,		dev)
+		__field(ino_t,		ino)
+		__field(time64_t,	ctime_s)
+		__field(u32,		ctime_ns)
+		__field(u32,		gen)
+	),
+
+	TP_fast_assign(
+		__entry->dev		= inode->i_sb->s_dev;
+		__entry->ino		= inode->i_ino;
+		__entry->gen		= inode->i_generation;
+		__entry->ctime_s	= ctime->tv_sec;
+		__entry->ctime_ns	= ctime->tv_nsec;
+	),
+
+	TP_printk("ino=%d:%d:%ld:%u ctime=%lld.%u",
+		MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __entry->gen,
+		__entry->ctime_s, __entry->ctime_ns
+	)
+);
+
+DEFINE_EVENT(ctime, inode_set_ctime_to_ts,
+		TP_PROTO(struct inode *inode,
+			 struct timespec64 *ctime),
+		TP_ARGS(inode, ctime));
+
+DEFINE_EVENT(ctime, ctime_xchg_skip,
+		TP_PROTO(struct inode *inode,
+			 struct timespec64 *ctime),
+		TP_ARGS(inode, ctime));
+
+TRACE_EVENT(ctime_ns_xchg,
+	TP_PROTO(struct inode *inode,
+		 u32 old,
+		 u32 new,
+		 u32 cur),
+
+	TP_ARGS(inode, old, new, cur),
+
+	TP_STRUCT__entry(
+		__field(dev_t,		dev)
+		__field(ino_t,		ino)
+		__field(u32,		gen)
+		__field(u32,		old)
+		__field(u32,		new)
+		__field(u32,		cur)
+	),
+
+	TP_fast_assign(
+		__entry->dev		= inode->i_sb->s_dev;
+		__entry->ino		= inode->i_ino;
+		__entry->gen		= inode->i_generation;
+		__entry->old		= old;
+		__entry->new		= new;
+		__entry->cur		= cur;
+	),
+
+	TP_printk("ino=%d:%d:%ld:%u old=%u:%s new=%u cur=%u:%s",
+		MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __entry->gen,
+		__entry->old & ~I_CTIME_QUERIED,
+		__print_flags(__entry->old & I_CTIME_QUERIED, "|", CTIME_QUERIED_FLAGS),
+		__entry->new,
+		__entry->cur & ~I_CTIME_QUERIED,
+		__print_flags(__entry->cur & I_CTIME_QUERIED, "|", CTIME_QUERIED_FLAGS)
+	)
+);
+
+TRACE_EVENT(fill_mg_cmtime,
+	TP_PROTO(struct inode *inode,
+		 struct timespec64 *ctime,
+		 struct timespec64 *mtime),
+
+	TP_ARGS(inode, ctime, mtime),
+
+	TP_STRUCT__entry(
+		__field(dev_t,		dev)
+		__field(ino_t,		ino)
+		__field(time64_t,	ctime_s)
+		__field(time64_t,	mtime_s)
+		__field(u32,		ctime_ns)
+		__field(u32,		mtime_ns)
+		__field(u32,		gen)
+	),
+
+	TP_fast_assign(
+		__entry->dev		= inode->i_sb->s_dev;
+		__entry->ino		= inode->i_ino;
+		__entry->gen		= inode->i_generation;
+		__entry->ctime_s	= ctime->tv_sec;
+		__entry->mtime_s	= mtime->tv_sec;
+		__entry->ctime_ns	= ctime->tv_nsec;
+		__entry->mtime_ns	= mtime->tv_nsec;
+	),
+
+	TP_printk("ino=%d:%d:%ld:%u ctime=%lld.%u mtime=%lld.%u",
+		MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __entry->gen,
+		__entry->ctime_s, __entry->ctime_ns,
+		__entry->mtime_s, __entry->mtime_ns
+	)
+);
+#endif /* _TRACE_TIMESTAMP_H */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>

-- 
2.46.2

[PATCH v8 06/12] fs: add percpu counters for significant multigrain timestamp events

[Jeff Layton]

New percpu counters for counting various stats around multigrain
timestamp events, and a new debugfs file for displaying them when
CONFIG_DEBUG_FS is enabled:

- number of attempted ctime updates
- number of successful i_ctime_nsec swaps
- number of fine-grained timestamp fetches

Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/inode.c | 75 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 71 insertions(+), 4 deletions(-)

diff --git a/fs/inode.c b/fs/inode.c
index 1a7eff2a40e2..e46f7170851b 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -21,6 +21,8 @@
 #include <linux/list_lru.h>
 #include <linux/iversion.h>
 #include <linux/rw_hint.h>
+#include <linux/seq_file.h>
+#include <linux/debugfs.h>
 #include <trace/events/writeback.h>
 #define CREATE_TRACE_POINTS
 #include <trace/events/timestamp.h>
@@ -101,6 +103,69 @@ long get_nr_dirty_inodes(void)
 	return nr_dirty > 0 ? nr_dirty : 0;
 }
 
+#ifdef CONFIG_DEBUG_FS
+static DEFINE_PER_CPU(long, mg_ctime_updates);
+static DEFINE_PER_CPU(long, mg_fine_stamps);
+static DEFINE_PER_CPU(long, mg_ctime_swaps);
+
+static unsigned long get_mg_ctime_updates(void)
+{
+	int i;
+	unsigned long sum = 0;
+
+	for_each_possible_cpu(i)
+		sum += per_cpu(mg_ctime_updates, i);
+	return sum;
+}
+
+static unsigned long get_mg_fine_stamps(void)
+{
+	int i;
+	unsigned long sum = 0;
+
+	for_each_possible_cpu(i)
+		sum += per_cpu(mg_fine_stamps, i);
+	return sum;
+}
+
+static unsigned long get_mg_ctime_swaps(void)
+{
+	int i;
+	unsigned long sum = 0;
+
+	for_each_possible_cpu(i)
+		sum += per_cpu(mg_ctime_swaps, i);
+	return sum;
+}
+
+#define mgtime_counter_inc(__var)	this_cpu_inc(__var)
+
+static int mgts_show(struct seq_file *s, void *p)
+{
+	unsigned long ctime_updates = get_mg_ctime_updates();
+	unsigned long ctime_swaps = get_mg_ctime_swaps();
+	unsigned long fine_stamps = get_mg_fine_stamps();
+
+	seq_printf(s, "%lu %lu %lu\n",
+		   ctime_updates, ctime_swaps, fine_stamps);
+	return 0;
+}
+
+DEFINE_SHOW_ATTRIBUTE(mgts);
+
+static int __init mg_debugfs_init(void)
+{
+	debugfs_create_file("multigrain_timestamps", S_IFREG | S_IRUGO, NULL, NULL, &mgts_fops);
+	return 0;
+}
+late_initcall(mg_debugfs_init);
+
+#else /* ! CONFIG_DEBUG_FS */
+
+#define mgtime_counter_inc(__var)	do { } while (0)
+
+#endif /* CONFIG_DEBUG_FS */
+
 /*
  * Handle nr_inode sysctl
  */
@@ -2691,10 +2756,9 @@ EXPORT_SYMBOL(timestamp_truncate);
  *
  * If it is multigrain, then we first see if the coarse-grained timestamp is
  * distinct from what we have. If so, then we'll just use that. If we have to
- * get a fine-grained timestamp, then do so, and try to swap it into the floor.
- * We accept the new floor value regardless of the outcome of the cmpxchg.
- * After that, we try to swap the new value into i_ctime_nsec. Again, we take
- * the resulting ctime, regardless of the outcome of the swap.
+ * get a fine-grained timestamp, then do so. After that, we try to swap the new
+ * value into i_ctime_nsec. We take the resulting ctime, regardless of the
+ * outcome of the swap.
  */
 struct timespec64 inode_set_ctime_current(struct inode *inode)
 {
@@ -2723,8 +2787,10 @@ struct timespec64 inode_set_ctime_current(struct inode *inode)
 		if (timespec64_compare(&now, &ctime) <= 0) {
 			ktime_get_real_ts64_mg(&now);
 			now = timestamp_truncate(now, inode);
+			mgtime_counter_inc(mg_fine_stamps);
 		}
 	}
+	mgtime_counter_inc(mg_ctime_updates);
 
 	/* No need to cmpxchg if it's exactly the same */
 	if (cns == now.tv_nsec && inode->i_ctime_sec == now.tv_sec) {
@@ -2738,6 +2804,7 @@ struct timespec64 inode_set_ctime_current(struct inode *inode)
 		/* If swap occurred, then we're (mostly) done */
 		inode->i_ctime_sec = now.tv_sec;
 		trace_ctime_ns_xchg(inode, cns, now.tv_nsec, cur);
+		mgtime_counter_inc(mg_ctime_swaps);
 	} else {
 		/*
 		 * Was the change due to someone marking the old ctime QUERIED?

-- 
2.46.2

[PATCH v8 07/12] timekeeping: add percpu counter for tracking floor swap events

[Jeff Layton]

The mgtime_floor value is a global variable for tracking the latest
fine-grained timestamp handed out. Because it's a global, track the
number of times that a new floor value is assigned.

Add a new percpu counter to the timekeeping code to track the number of
floor swap events that have occurred. Display that as a fourth integer
in /sys/kernel/debug/multigrain_timestamps.

Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/inode.c                         |  5 +++--
 include/linux/timekeeping.h        |  1 +
 kernel/time/timekeeping.c          |  1 +
 kernel/time/timekeeping_debug.c    | 13 +++++++++++++
 kernel/time/timekeeping_internal.h |  9 +++++++++
 5 files changed, 27 insertions(+), 2 deletions(-)

diff --git a/fs/inode.c b/fs/inode.c
index e46f7170851b..93a9365cc22d 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -145,9 +145,10 @@ static int mgts_show(struct seq_file *s, void *p)
 	unsigned long ctime_updates = get_mg_ctime_updates();
 	unsigned long ctime_swaps = get_mg_ctime_swaps();
 	unsigned long fine_stamps = get_mg_fine_stamps();
+	unsigned long floor_swaps = timekeeping_get_mg_floor_swaps();
 
-	seq_printf(s, "%lu %lu %lu\n",
-		   ctime_updates, ctime_swaps, fine_stamps);
+	seq_printf(s, "%lu %lu %lu %lu\n",
+		   ctime_updates, ctime_swaps, fine_stamps, floor_swaps);
 	return 0;
 }
 
diff --git a/include/linux/timekeeping.h b/include/linux/timekeeping.h
index 7aa85246c183..84a035e86ac8 100644
--- a/include/linux/timekeeping.h
+++ b/include/linux/timekeeping.h
@@ -48,6 +48,7 @@ extern void ktime_get_coarse_real_ts64(struct timespec64 *ts);
 /* Multigrain timestamp interfaces */
 extern void ktime_get_coarse_real_ts64_mg(struct timespec64 *ts);
 extern void ktime_get_real_ts64_mg(struct timespec64 *ts);
+extern unsigned long timekeeping_get_mg_floor_swaps(void);
 
 void getboottime64(struct timespec64 *ts);
 
diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
index 37004a4758cf..c8a934c1c8aa 100644
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -2478,6 +2478,7 @@ void ktime_get_real_ts64_mg(struct timespec64 *ts)
 	if (atomic64_try_cmpxchg(&mg_floor, &old, mono)) {
 		ts->tv_nsec = 0;
 		timespec64_add_ns(ts, nsecs);
+		timekeeping_inc_mg_floor_swaps();
 	} else {
 		/*
 		 * Something has changed mg_floor since "old" was
diff --git a/kernel/time/timekeeping_debug.c b/kernel/time/timekeeping_debug.c
index b73e8850e58d..b731621ad811 100644
--- a/kernel/time/timekeeping_debug.c
+++ b/kernel/time/timekeeping_debug.c
@@ -17,6 +17,9 @@
 
 #define NUM_BINS 32
 
+/* incremented every time mg_floor is updated */
+DEFINE_PER_CPU(unsigned long, timekeeping_mg_floor_swaps);
+
 static unsigned int sleep_time_bin[NUM_BINS] = {0};
 
 static int tk_debug_sleep_time_show(struct seq_file *s, void *data)
@@ -53,3 +56,13 @@ void tk_debug_account_sleep_time(const struct timespec64 *t)
 			   (s64)t->tv_sec, t->tv_nsec / NSEC_PER_MSEC);
 }
 
+unsigned long timekeeping_get_mg_floor_swaps(void)
+{
+	int i;
+	unsigned long sum = 0;
+
+	for_each_possible_cpu(i)
+		sum += per_cpu(timekeeping_mg_floor_swaps, i);
+	return sum < 0 ? 0 : sum;
+}
+
diff --git a/kernel/time/timekeeping_internal.h b/kernel/time/timekeeping_internal.h
index 4ca2787d1642..f53e76d5ee7c 100644
--- a/kernel/time/timekeeping_internal.h
+++ b/kernel/time/timekeeping_internal.h
@@ -10,9 +10,18 @@
  * timekeeping debug functions
  */
 #ifdef CONFIG_DEBUG_FS
+DECLARE_PER_CPU(unsigned long, timekeeping_mg_floor_swaps);
+static inline void timekeeping_inc_mg_floor_swaps(void)
+{
+	this_cpu_inc(timekeeping_mg_floor_swaps);
+}
+
 extern void tk_debug_account_sleep_time(const struct timespec64 *t);
 #else
 #define tk_debug_account_sleep_time(x)
+static inline void timekeeping_inc_mg_floor_swaps(void)
+{
+}
 #endif
 
 #ifdef CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE

-- 
2.46.2

[PATCH v8 08/12] Documentation: add a new file documenting multigrain timestamps

[Jeff Layton]

Add a high-level document that describes how multigrain timestamps work,
rationale for them, and some info about implementation and tradeoffs.

Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 Documentation/filesystems/index.rst         |   1 +
 Documentation/filesystems/multigrain-ts.rst | 125 ++++++++++++++++++++++++++++
 2 files changed, 126 insertions(+)

diff --git a/Documentation/filesystems/index.rst b/Documentation/filesystems/index.rst
index e8e496d23e1d..44e9e77ffe0d 100644
--- a/Documentation/filesystems/index.rst
+++ b/Documentation/filesystems/index.rst
@@ -29,6 +29,7 @@ algorithms work.
    fiemap
    files
    locks
+   multigrain-ts
    mount_api
    quota
    seq_file
diff --git a/Documentation/filesystems/multigrain-ts.rst b/Documentation/filesystems/multigrain-ts.rst
new file mode 100644
index 000000000000..a150c93471de
--- /dev/null
+++ b/Documentation/filesystems/multigrain-ts.rst
@@ -0,0 +1,125 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+=====================
+Multigrain Timestamps
+=====================
+
+Introduction
+============
+Historically, the kernel has always used coarse time values to stamp inodes.
+This value is updated every jiffy, so any change that happens within that jiffy
+will end up with the same timestamp.
+
+When the kernel goes to stamp an inode (due to a read or write), it first gets
+the current time and then compares it to the existing timestamp(s) to see
+whether anything will change. If nothing changed, then it can avoid updating
+the inode's metadata.
+
+Coarse timestamps are therefore good from a performance standpoint, since they
+reduce the need for metadata updates, but bad from the standpoint of
+determining whether anything has changed, since a lot of things can happen in a
+jiffy.
+
+They are particularly troublesome with NFSv3, where unchanging timestamps can
+make it difficult to tell whether to invalidate caches. NFSv4 provides a
+dedicated change attribute that should always show a visible change, but not
+all filesystems implement this properly, causing the NFS server to substitute
+the ctime in many cases.
+
+Multigrain timestamps aim to remedy this by selectively using fine-grained
+timestamps when a file has had its timestamps queried recently, and the current
+coarse-grained time does not cause a change.
+
+Inode Timestamps
+================
+There are currently 3 timestamps in the inode that are updated to the current
+wallclock time on different activity:
+
+ctime:
+  The inode change time. This is stamped with the current time whenever
+  the inode's metadata is changed. Note that this value is not settable
+  from userland.
+
+mtime:
+  The inode modification time. This is stamped with the current time
+  any time a file's contents change.
+
+atime:
+  The inode access time. This is stamped whenever an inode's contents are
+  read. Widely considered to be a terrible mistake. Usually avoided with
+  options like noatime or relatime.
+
+Updating the mtime always implies a change to the ctime, but updating the
+atime due to a read request does not.
+
+Multigrain timestamps are only tracked for the ctime and the mtime. atimes are
+not affected and always use the coarse-grained value (subject to the floor).
+
+Inode Timestamp Ordering
+========================
+
+In addition to just providing info about changes to individual files, file
+timestamps also serve an important purpose in applications like "make". These
+programs measure timestamps in order to determine whether source files might be
+newer than cached objects.
+
+Userland applications like make can only determine ordering based on
+operational boundaries. For a syscall those are the syscall entry and exit
+points. For io_uring or nfsd operations, that's the request submission and
+response. In the case of concurrent operations, userland can make no
+determination about the order in which things will occur.
+
+For instance, if a single thread modifies one file, and then another file in
+sequence, the second file must show an equal or later mtime than the first. The
+same is true if two threads are issuing similar operations that do not overlap
+in time.
+
+If however, two threads have racing syscalls that overlap in time, then there
+is no such guarantee, and the second file may appear to have been modified
+before, after or at the same time as the first, regardless of which one was
+submitted first.
+
+Note that the above assumes that the system doesn't experience a backward jump
+of the realtime clock. If that occurs at an inopportune time, then timestamps
+can appear to go backward, even on a properly functioning system.
+
+Multigrain Timestamp Implementation
+===================================
+Multigrain timestamps are aimed at ensuring that changes to a single file are
+always recognizable, without violating the ordering guarantees when multiple
+different files are modified. This affects the mtime and the ctime, but the
+atime will always use coarse-grained timestamps.
+
+It uses an unused bit in the i_ctime_nsec field to indicate whether the mtime
+or ctime has been queried. If either or both have, then the kernel takes
+special care to ensure the next timestamp update will display a visible change.
+This ensures tight cache coherency for use-cases like NFS, without sacrificing
+the benefits of reduced metadata updates when files aren't being watched.
+
+The Ctime Floor Value
+=====================
+It's not sufficient to simply use fine or coarse-grained timestamps based on
+whether the mtime or ctime has been queried. A file could get a fine grained
+timestamp, and then a second file modified later could get a coarse-grained one
+that appears earlier than the first, which would break the kernel's timestamp
+ordering guarantees.
+
+To mitigate this problem, we maintain a global floor value that ensures that
+this can't happen. The two files in the above example may appear to have been
+modified at the same time in such a case, but they will never show the reverse
+order. To avoid problems with realtime clock jumps, the floor is managed as a
+monotonic ktime_t, and the values are converted to realtime clock values as
+needed.
+
+Implementation Notes
+====================
+Multigrain timestamps are intended for use by local filesystems that get
+ctime values from the local clock. This is in contrast to network filesystems
+and the like that just mirror timestamp values from a server.
+
+For most filesystems, it's sufficient to just set the FS_MGTIME flag in the
+fstype->fs_flags in order to opt-in, providing the ctime is only ever set via
+inode_set_ctime_current(). If the filesystem has a ->getattr routine that
+doesn't call generic_fillattr, then you should have it call fill_mg_cmtime to
+fill those values. For setattr, it should use setattr_copy() to update the
+timestamps, or otherwise mimic its behavior.

-- 
2.46.2

[PATCH v8 09/12] xfs: switch to multigrain timestamps

[Jeff Layton]

Enable multigrain timestamps, which should ensure that there is an
apparent change to the timestamp whenever it has been written after
being actively observed via getattr.

Also, anytime the mtime changes, the ctime must also change, and those
are now the only two options for xfs_trans_ichgtime. Have that function
unconditionally bump the ctime, and ASSERT that XFS_ICHGTIME_CHG is
always set.

Finally, stop setting STATX_CHANGE_COOKIE in getattr, since the ctime
should give us better semantics now.

Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/xfs/libxfs/xfs_trans_inode.c |  6 +++---
 fs/xfs/xfs_iops.c               | 10 +++-------
 fs/xfs/xfs_super.c              |  2 +-
 3 files changed, 7 insertions(+), 11 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_trans_inode.c b/fs/xfs/libxfs/xfs_trans_inode.c
index 3c40f37e82c7..c962ad64b0c1 100644
--- a/fs/xfs/libxfs/xfs_trans_inode.c
+++ b/fs/xfs/libxfs/xfs_trans_inode.c
@@ -62,12 +62,12 @@ xfs_trans_ichgtime(
 	ASSERT(tp);
 	xfs_assert_ilocked(ip, XFS_ILOCK_EXCL);
 
-	tv = current_time(inode);
+	/* If the mtime changes, then ctime must also change */
+	ASSERT(flags & XFS_ICHGTIME_CHG);
 
+	tv = inode_set_ctime_current(inode);
 	if (flags & XFS_ICHGTIME_MOD)
 		inode_set_mtime_to_ts(inode, tv);
-	if (flags & XFS_ICHGTIME_CHG)
-		inode_set_ctime_to_ts(inode, tv);
 	if (flags & XFS_ICHGTIME_ACCESS)
 		inode_set_atime_to_ts(inode, tv);
 	if (flags & XFS_ICHGTIME_CREATE)
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index ee79cf161312..b5d0c5c157e7 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -597,8 +597,9 @@ xfs_vn_getattr(
 	stat->gid = vfsgid_into_kgid(vfsgid);
 	stat->ino = ip->i_ino;
 	stat->atime = inode_get_atime(inode);
-	stat->mtime = inode_get_mtime(inode);
-	stat->ctime = inode_get_ctime(inode);
+
+	fill_mg_cmtime(stat, request_mask, inode);
+
 	stat->blocks = XFS_FSB_TO_BB(mp, ip->i_nblocks + ip->i_delayed_blks);
 
 	if (xfs_has_v3inodes(mp)) {
@@ -608,11 +609,6 @@ xfs_vn_getattr(
 		}
 	}
 
-	if ((request_mask & STATX_CHANGE_COOKIE) && IS_I_VERSION(inode)) {
-		stat->change_cookie = inode_query_iversion(inode);
-		stat->result_mask |= STATX_CHANGE_COOKIE;
-	}
-
 	/*
 	 * Note: If you add another clause to set an attribute flag, please
 	 * update attributes_mask below.
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index fbb3a1594c0d..fda75db739b1 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -2063,7 +2063,7 @@ static struct file_system_type xfs_fs_type = {
 	.init_fs_context	= xfs_init_fs_context,
 	.parameters		= xfs_fs_parameters,
 	.kill_sb		= xfs_kill_sb,
-	.fs_flags		= FS_REQUIRES_DEV | FS_ALLOW_IDMAP,
+	.fs_flags		= FS_REQUIRES_DEV | FS_ALLOW_IDMAP | FS_MGTIME,
 };
 MODULE_ALIAS_FS("xfs");
 

-- 
2.46.2

[PATCH v8 10/12] ext4: switch to multigrain timestamps

[Jeff Layton]

Enable multigrain timestamps, which should ensure that there is an
apparent change to the timestamp whenever it has been written after
being actively observed via getattr.

For ext4, we only need to enable the FS_MGTIME flag.

Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/ext4/super.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 16a4ce704460..b77acba4a719 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -7329,7 +7329,7 @@ static struct file_system_type ext4_fs_type = {
 	.init_fs_context	= ext4_init_fs_context,
 	.parameters		= ext4_param_specs,
 	.kill_sb		= ext4_kill_sb,
-	.fs_flags		= FS_REQUIRES_DEV | FS_ALLOW_IDMAP,
+	.fs_flags		= FS_REQUIRES_DEV | FS_ALLOW_IDMAP | FS_MGTIME,
 };
 MODULE_ALIAS_FS("ext4");
 

-- 
2.46.2

[PATCH v8 11/12] btrfs: convert to multigrain timestamps

[Jeff Layton]

Enable multigrain timestamps, which should ensure that there is an
apparent change to the timestamp whenever it has been written after
being actively observed via getattr.

Beyond enabling the FS_MGTIME flag, this patch eliminates
update_time_for_write, which goes to great pains to avoid in-memory
stores. Just have it overwrite the timestamps unconditionally.

Note that this also drops the IS_I_VERSION check and unconditionally
bumps the change attribute, since SB_I_VERSION is always set on btrfs.

Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/btrfs/file.c  | 25 ++++---------------------
 fs/btrfs/super.c |  3 ++-
 2 files changed, 6 insertions(+), 22 deletions(-)

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 4fb521d91b06..e5384ceb8acf 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1120,26 +1120,6 @@ void btrfs_check_nocow_unlock(struct btrfs_inode *inode)
 	btrfs_drew_write_unlock(&inode->root->snapshot_lock);
 }
 
-static void update_time_for_write(struct inode *inode)
-{
-	struct timespec64 now, ts;
-
-	if (IS_NOCMTIME(inode))
-		return;
-
-	now = current_time(inode);
-	ts = inode_get_mtime(inode);
-	if (!timespec64_equal(&ts, &now))
-		inode_set_mtime_to_ts(inode, now);
-
-	ts = inode_get_ctime(inode);
-	if (!timespec64_equal(&ts, &now))
-		inode_set_ctime_to_ts(inode, now);
-
-	if (IS_I_VERSION(inode))
-		inode_inc_iversion(inode);
-}
-
 int btrfs_write_check(struct kiocb *iocb, struct iov_iter *from, size_t count)
 {
 	struct file *file = iocb->ki_filp;
@@ -1170,7 +1150,10 @@ int btrfs_write_check(struct kiocb *iocb, struct iov_iter *from, size_t count)
 	 * need to start yet another transaction to update the inode as we will
 	 * update the inode when we finish writing whatever data we write.
 	 */
-	update_time_for_write(inode);
+	if (!IS_NOCMTIME(inode)) {
+		inode_set_mtime_to_ts(inode, inode_set_ctime_current(inode));
+		inode_inc_iversion(inode);
+	}
 
 	start_pos = round_down(pos, fs_info->sectorsize);
 	oldsize = i_size_read(inode);
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 98fa0f382480..d423acfe11d0 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -2198,7 +2198,8 @@ static struct file_system_type btrfs_fs_type = {
 	.init_fs_context	= btrfs_init_fs_context,
 	.parameters		= btrfs_fs_parameters,
 	.kill_sb		= btrfs_kill_super,
-	.fs_flags		= FS_REQUIRES_DEV | FS_BINARY_MOUNTDATA | FS_ALLOW_IDMAP,
+	.fs_flags		= FS_REQUIRES_DEV | FS_BINARY_MOUNTDATA |
+				  FS_ALLOW_IDMAP | FS_MGTIME,
  };
 
 MODULE_ALIAS_FS("btrfs");

-- 
2.46.2

[PATCH v8 12/12] tmpfs: add support for multigrain timestamps

[Jeff Layton]

Enable multigrain timestamps, which should ensure that there is an
apparent change to the timestamp whenever it has been written after
being actively observed via getattr.

tmpfs only requires the FS_MGTIME flag.

Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 mm/shmem.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index 4f11b5506363..3444efc275b1 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -4951,7 +4951,7 @@ static struct file_system_type shmem_fs_type = {
 	.parameters	= shmem_fs_parameters,
 #endif
 	.kill_sb	= kill_litter_super,
-	.fs_flags	= FS_USERNS_MOUNT | FS_ALLOW_IDMAP,
+	.fs_flags	= FS_USERNS_MOUNT | FS_ALLOW_IDMAP | FS_MGTIME,
 };
 
 void __init shmem_init(void)

-- 
2.46.2

Re: [PATCH v8 02/12] fs: add infrastructure for multigrain timestamps

[Jan Kara]

On Tue 01-10-24 06:58:56, Jeff Layton wrote:
> The VFS has always used coarse-grained timestamps when updating the
> ctime and mtime after a change. This has the benefit of allowing
> filesystems to optimize away a lot metadata updates, down to around 1
> per jiffy, even when a file is under heavy writes.
> 
> Unfortunately, this has always been an issue when we're exporting via
> NFSv3, which relies on timestamps to validate caches. A lot of changes
> can happen in a jiffy, so timestamps aren't sufficient to help the
> client decide when to invalidate the cache. Even with NFSv4, a lot of
> exported filesystems don't properly support a change attribute and are
> subject to the same problems with timestamp granularity. Other
> applications have similar issues with timestamps (e.g backup
> applications).
> 
> If we were to always use fine-grained timestamps, that would improve the
> situation, but that becomes rather expensive, as the underlying
> filesystem would have to log a lot more metadata updates.
> 
> What we need is a way to only use fine-grained timestamps when they are
> being actively queried. Use the (unused) top bit in inode->i_ctime_nsec
> as a flag that indicates whether the current timestamps have been
> queried via stat() or the like. When it's set, we allow the kernel to
> use a fine-grained timestamp iff it's necessary to make the ctime show
> a different value.
> 
> This solves the problem of being able to distinguish the timestamp
> between updates, but introduces a new problem: it's now possible for a
> file being changed to get a fine-grained timestamp. A file that is
> altered just a bit later can then get a coarse-grained one that appears
> older than the earlier fine-grained time. This violates timestamp
> ordering guarantees.
> 
> To remedy this, keep a global monotonic atomic64_t value that acts as a
> timestamp floor.  When we go to stamp a file, we first get the latter of
> the current floor value and the current coarse-grained time. If the
> inode ctime hasn't been queried then we just attempt to stamp it with
> that value.
> 
> If it has been queried, then first see whether the current coarse time
> is later than the existing ctime. If it is, then we accept that value.
> If it isn't, then we get a fine-grained timestamp.
> 
> Filesystems can opt into this by setting the FS_MGTIME fstype flag.
> Others should be unaffected (other than being subject to the same floor
> value as multigrain filesystems).
> 
> Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
> Signed-off-by: Jeff Layton <jlayton@kernel.org>

Mostly looks good. Some smaller comments below.

> +/**
> + * current_time - Return FS time (possibly fine-grained)
> + * @inode: inode.
> + *
> + * Return the current time truncated to the time granularity supported by
> + * the fs, as suitable for a ctime/mtime change. If the ctime is flagged
> + * as having been QUERIED, get a fine-grained timestamp, but don't update
> + * the floor.
> + *
> + * For a multigrain inode, this is effectively an estimate of the timestamp
> + * that a file would receive. An actual update must go through
> + * inode_set_ctime_current().
> + */
> +struct timespec64 current_time(struct inode *inode)
> +{
> +	struct timespec64 now;
> +	u32 cns;
> +
> +	ktime_get_coarse_real_ts64_mg(&now);
> +
> +	if (!is_mgtime(inode))
> +		goto out;
> +
> +	/* If nothing has queried it, then coarse time is fine */
> +	cns = smp_load_acquire(&inode->i_ctime_nsec);
> +	if (cns & I_CTIME_QUERIED) {
> +		/*
> +		 * If there is no apparent change, then get a fine-grained
> +		 * timestamp.
> +		 */
> +		if (now.tv_nsec == (cns & ~I_CTIME_QUERIED))
> +			ktime_get_real_ts64(&now);
> +	}
> +out:
> +	return timestamp_truncate(now, inode);
> +}
> +EXPORT_SYMBOL(current_time);
> +
>  static int inode_needs_update_time(struct inode *inode)
>  {
> +	struct timespec64 now, ts;
>  	int sync_it = 0;
> -	struct timespec64 now = current_time(inode);
> -	struct timespec64 ts;
>  
>  	/* First try to exhaust all avenues to not sync */
>  	if (IS_NOCMTIME(inode))
>  		return 0;
>  
> +	now = current_time(inode);
> +
>  	ts = inode_get_mtime(inode);
>  	if (!timespec64_equal(&ts, &now))
> -		sync_it = S_MTIME;
> +		sync_it |= S_MTIME;
>  
>  	ts = inode_get_ctime(inode);
>  	if (!timespec64_equal(&ts, &now))
> @@ -2598,6 +2637,15 @@ void inode_nohighmem(struct inode *inode)
>  }
>  EXPORT_SYMBOL(inode_nohighmem);
>  
> +struct timespec64 inode_set_ctime_to_ts(struct inode *inode, struct timespec64 ts)
> +{
> +	set_normalized_timespec64(&ts, ts.tv_sec, ts.tv_nsec);
> +	inode->i_ctime_sec = ts.tv_sec;
> +	inode->i_ctime_nsec = ts.tv_nsec;
> +	return ts;
> +}
> +EXPORT_SYMBOL(inode_set_ctime_to_ts);
> +
>  /**
>   * timestamp_truncate - Truncate timespec to a granularity
>   * @t: Timespec
> @@ -2630,36 +2678,75 @@ struct timespec64 timestamp_truncate(struct timespec64 t, struct inode *inode)
>  EXPORT_SYMBOL(timestamp_truncate);
>  
>  /**
> - * current_time - Return FS time
> - * @inode: inode.
> + * inode_set_ctime_current - set the ctime to current_time
> + * @inode: inode
>   *
> - * Return the current time truncated to the time granularity supported by
> - * the fs.
> + * Set the inode's ctime to the current value for the inode. Returns the
> + * current value that was assigned. If this is not a multigrain inode, then we
> + * set it to the later of the coarse time and floor value.
>   *
> - * Note that inode and inode->sb cannot be NULL.
> - * Otherwise, the function warns and returns time without truncation.
> + * If it is multigrain, then we first see if the coarse-grained timestamp is
> + * distinct from what we have. If so, then we'll just use that. If we have to
> + * get a fine-grained timestamp, then do so, and try to swap it into the floor.
> + * We accept the new floor value regardless of the outcome of the cmpxchg.
> + * After that, we try to swap the new value into i_ctime_nsec. Again, we take
> + * the resulting ctime, regardless of the outcome of the swap.

This comment seems outdated now. No floor in this function anymore...

> -struct timespec64 current_time(struct inode *inode)
> +struct timespec64 inode_set_ctime_current(struct inode *inode)
>  {
>  	struct timespec64 now;
> +	u32 cns, cur;
...

> diff --git a/fs/stat.c b/fs/stat.c
> index 41e598376d7e..381926fb405f 100644
> --- a/fs/stat.c
> +++ b/fs/stat.c
> @@ -26,6 +26,35 @@
>  #include "internal.h"
>  #include "mount.h"
>  
> +/**
> + * fill_mg_cmtime - Fill in the mtime and ctime and flag ctime as QUERIED
> + * @stat: where to store the resulting values
> + * @request_mask: STATX_* values requested
> + * @inode: inode from which to grab the c/mtime
> + *
> + * Given @inode, grab the ctime and mtime out if it and store the result
						 ^^ of

> + * in @stat. When fetching the value, flag it as QUERIED (if not already)
> + * so the next write will record a distinct timestamp.
> + */
> +void fill_mg_cmtime(struct kstat *stat, u32 request_mask, struct inode *inode)
> +{

Given how things worked out in the end, it seems this function doesn't need
to handle mtime at all and we can move mtime handling back to shared generic
code?

> +	atomic_t *pcn = (atomic_t *)&inode->i_ctime_nsec;
> +
> +	/* If neither time was requested, then don't report them */
> +	if (!(request_mask & (STATX_CTIME|STATX_MTIME))) {
> +		stat->result_mask &= ~(STATX_CTIME|STATX_MTIME);
> +		return;
> +	}
> +
> +	stat->mtime = inode_get_mtime(inode);
> +	stat->ctime.tv_sec = inode->i_ctime_sec;
> +	stat->ctime.tv_nsec = (u32)atomic_read(pcn);
> +	if (!(stat->ctime.tv_nsec & I_CTIME_QUERIED))
> +		stat->ctime.tv_nsec = ((u32)atomic_fetch_or(I_CTIME_QUERIED, pcn));
> +	stat->ctime.tv_nsec &= ~I_CTIME_QUERIED;
> +}
> +EXPORT_SYMBOL(fill_mg_cmtime);
> +
>  /**
>   * generic_fillattr - Fill in the basic attributes from the inode struct
>   * @idmap:		idmap of the mount the inode was found from
> @@ -58,8 +87,14 @@ void generic_fillattr(struct mnt_idmap *idmap, u32 request_mask,
>  	stat->rdev = inode->i_rdev;
>  	stat->size = i_size_read(inode);
>  	stat->atime = inode_get_atime(inode);
> -	stat->mtime = inode_get_mtime(inode);
> -	stat->ctime = inode_get_ctime(inode);
> +
> +	if (is_mgtime(inode)) {
> +		fill_mg_cmtime(stat, request_mask, inode);
> +	} else {
> +		stat->ctime = inode_get_ctime(inode);
> +		stat->mtime = inode_get_mtime(inode);
> +	}
> +
>  	stat->blksize = i_blocksize(inode);
>  	stat->blocks = inode->i_blocks;
>  
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index e3c603d01337..23908bad166c 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -1653,6 +1653,17 @@ static inline struct timespec64 inode_set_mtime(struct inode *inode,
>  	return inode_set_mtime_to_ts(inode, ts);
>  }
>  
> +/*
> + * Multigrain timestamps
> + *
> + * Conditionally use fine-grained ctime and mtime timestamps when there
> + * are users actively observing them via getattr. The primary use-case
> + * for this is NFS clients that use the ctime to distinguish between
> + * different states of the file, and that are often fooled by multiple
> + * operations that occur in the same coarse-grained timer tick.

Again, mtime seems unaffected by mgtime changes now.

> + */
> +#define I_CTIME_QUERIED		((u32)BIT(31))
> +

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH v8 04/12] fs: handle delegated timestamps in setattr_copy_mgtime

[Jan Kara]

On Tue 01-10-24 06:58:58, Jeff Layton wrote:
> When updating the ctime on an inode for a SETATTR with a multigrain
> filesystem, we usually want to take the latest time we can get for the
> ctime. The exception to this rule is when there is a nfsd write
> delegation and the server is proxying timestamps from the client.
> 
> When nfsd gets a CB_GETATTR response, we want to update the timestamp
> value in the inode to the values that the client is tracking. The client
> doesn't send a ctime value (since that's always determined by the
> exported filesystem), but it can send a mtime value. In the case where
> it does, then we may need to update the ctime to a value commensurate
> with that instead of the current time.
> 
> If ATTR_DELEG is set, then use ia_ctime value instead of setting the
> timestamp to the current time.
> 
> With the addition of delegated timestamps we can also receive a request
> to update only the atime, but we may not need to set the ctime. Trust
> the ATTR_CTIME flag in the update and only update the ctime when it's
> set.
> 
> Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
> Signed-off-by: Jeff Layton <jlayton@kernel.org>

Looks good. Feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

								Honza

> ---
>  fs/attr.c          | 28 +++++++++++++--------
>  fs/inode.c         | 72 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
>  include/linux/fs.h |  2 ++
>  3 files changed, 92 insertions(+), 10 deletions(-)
> 
> diff --git a/fs/attr.c b/fs/attr.c
> index 3bcbc45708a3..392eb62aa609 100644
> --- a/fs/attr.c
> +++ b/fs/attr.c
> @@ -286,16 +286,20 @@ static void setattr_copy_mgtime(struct inode *inode, const struct iattr *attr)
>  	unsigned int ia_valid = attr->ia_valid;
>  	struct timespec64 now;
>  
> -	/*
> -	 * If the ctime isn't being updated then nothing else should be
> -	 * either.
> -	 */
> -	if (!(ia_valid & ATTR_CTIME)) {
> -		WARN_ON_ONCE(ia_valid & (ATTR_ATIME|ATTR_MTIME));
> -		return;
> +	if (ia_valid & ATTR_CTIME) {
> +		/*
> +		 * In the case of an update for a write delegation, we must respect
> +		 * the value in ia_ctime and not use the current time.
> +		 */
> +		if (ia_valid & ATTR_DELEG)
> +			now = inode_set_ctime_deleg(inode, attr->ia_ctime);
> +		else
> +			now = inode_set_ctime_current(inode);
> +	} else {
> +		/* If ATTR_CTIME isn't set, then ATTR_MTIME shouldn't be either. */
> +		WARN_ON_ONCE(ia_valid & ATTR_MTIME);
>  	}
>  
> -	now = inode_set_ctime_current(inode);
>  	if (ia_valid & ATTR_ATIME_SET)
>  		inode_set_atime_to_ts(inode, attr->ia_atime);
>  	else if (ia_valid & ATTR_ATIME)
> @@ -354,8 +358,12 @@ void setattr_copy(struct mnt_idmap *idmap, struct inode *inode,
>  		inode_set_atime_to_ts(inode, attr->ia_atime);
>  	if (ia_valid & ATTR_MTIME)
>  		inode_set_mtime_to_ts(inode, attr->ia_mtime);
> -	if (ia_valid & ATTR_CTIME)
> -		inode_set_ctime_to_ts(inode, attr->ia_ctime);
> +	if (ia_valid & ATTR_CTIME) {
> +		if (ia_valid & ATTR_DELEG)
> +			inode_set_ctime_deleg(inode, attr->ia_ctime);
> +		else
> +			inode_set_ctime_to_ts(inode, attr->ia_ctime);
> +	}
>  }
>  EXPORT_SYMBOL(setattr_copy);
>  
> diff --git a/fs/inode.c b/fs/inode.c
> index 4ec1e71e9a9d..7a324d999816 100644
> --- a/fs/inode.c
> +++ b/fs/inode.c
> @@ -2751,6 +2751,78 @@ struct timespec64 inode_set_ctime_current(struct inode *inode)
>  }
>  EXPORT_SYMBOL(inode_set_ctime_current);
>  
> +/**
> + * inode_set_ctime_deleg - try to update the ctime on a delegated inode
> + * @inode: inode to update
> + * @update: timespec64 to set the ctime
> + *
> + * Attempt to atomically update the ctime on behalf of a delegation holder.
> + *
> + * The nfs server can call back the holder of a delegation to get updated
> + * inode attributes, including the mtime. When updating the mtime we may
> + * need to update the ctime to a value at least equal to that.
> + *
> + * This can race with concurrent updates to the inode, in which
> + * case we just don't do the update.
> + *
> + * Note that this works even when multigrain timestamps are not enabled,
> + * so use it in either case.
> + */
> +struct timespec64 inode_set_ctime_deleg(struct inode *inode, struct timespec64 update)
> +{
> +	struct timespec64 now, cur_ts;
> +	u32 cur, old;
> +
> +	/* pairs with try_cmpxchg below */
> +	cur = smp_load_acquire(&inode->i_ctime_nsec);
> +	cur_ts.tv_nsec = cur & ~I_CTIME_QUERIED;
> +	cur_ts.tv_sec = inode->i_ctime_sec;
> +
> +	/* If the update is older than the existing value, skip it. */
> +	if (timespec64_compare(&update, &cur_ts) <= 0)
> +		return cur_ts;
> +
> +	ktime_get_coarse_real_ts64_mg(&now);
> +
> +	/* Clamp the update to "now" if it's in the future */
> +	if (timespec64_compare(&update, &now) > 0)
> +		update = now;
> +
> +	update = timestamp_truncate(update, inode);
> +
> +	/* No need to update if the values are already the same */
> +	if (timespec64_equal(&update, &cur_ts))
> +		return cur_ts;
> +
> +	/*
> +	 * Try to swap the nsec value into place. If it fails, that means
> +	 * we raced with an update due to a write or similar activity. That
> +	 * stamp takes precedence, so just skip the update.
> +	 */
> +retry:
> +	old = cur;
> +	if (try_cmpxchg(&inode->i_ctime_nsec, &cur, update.tv_nsec)) {
> +		inode->i_ctime_sec = update.tv_sec;
> +		mgtime_counter_inc(mg_ctime_swaps);
> +		return update;
> +	}
> +
> +	/*
> +	 * Was the change due to someone marking the old ctime QUERIED?
> +	 * If so then retry the swap. This can only happen once since
> +	 * the only way to clear I_CTIME_QUERIED is to stamp the inode
> +	 * with a new ctime.
> +	 */
> +	if (!(old & I_CTIME_QUERIED) && (cur == (old | I_CTIME_QUERIED)))
> +		goto retry;
> +
> +	/* Otherwise, it was a new timestamp. */
> +	cur_ts.tv_sec = inode->i_ctime_sec;
> +	cur_ts.tv_nsec = cur & ~I_CTIME_QUERIED;
> +	return cur_ts;
> +}
> +EXPORT_SYMBOL(inode_set_ctime_deleg);
> +
>  /**
>   * in_group_or_capable - check whether caller is CAP_FSETID privileged
>   * @idmap:	idmap of the mount @inode was found from
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index 23908bad166c..b1a3bd07711b 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -1584,6 +1584,8 @@ static inline bool fsuidgid_has_mapping(struct super_block *sb,
>  
>  struct timespec64 current_time(struct inode *inode);
>  struct timespec64 inode_set_ctime_current(struct inode *inode);
> +struct timespec64 inode_set_ctime_deleg(struct inode *inode,
> +					struct timespec64 update);
>  
>  static inline time64_t inode_get_atime_sec(const struct inode *inode)
>  {
> 
> -- 
> 2.46.2
> 
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH v8 02/12] fs: add infrastructure for multigrain timestamps

[Jeff Layton]

On Tue, 2024-10-01 at 15:20 +0200, Jan Kara wrote:
> On Tue 01-10-24 06:58:56, Jeff Layton wrote:
> > The VFS has always used coarse-grained timestamps when updating the
> > ctime and mtime after a change. This has the benefit of allowing
> > filesystems to optimize away a lot metadata updates, down to around 1
> > per jiffy, even when a file is under heavy writes.
> > 
> > Unfortunately, this has always been an issue when we're exporting via
> > NFSv3, which relies on timestamps to validate caches. A lot of changes
> > can happen in a jiffy, so timestamps aren't sufficient to help the
> > client decide when to invalidate the cache. Even with NFSv4, a lot of
> > exported filesystems don't properly support a change attribute and are
> > subject to the same problems with timestamp granularity. Other
> > applications have similar issues with timestamps (e.g backup
> > applications).
> > 
> > If we were to always use fine-grained timestamps, that would improve the
> > situation, but that becomes rather expensive, as the underlying
> > filesystem would have to log a lot more metadata updates.
> > 
> > What we need is a way to only use fine-grained timestamps when they are
> > being actively queried. Use the (unused) top bit in inode->i_ctime_nsec
> > as a flag that indicates whether the current timestamps have been
> > queried via stat() or the like. When it's set, we allow the kernel to
> > use a fine-grained timestamp iff it's necessary to make the ctime show
> > a different value.
> > 
> > This solves the problem of being able to distinguish the timestamp
> > between updates, but introduces a new problem: it's now possible for a
> > file being changed to get a fine-grained timestamp. A file that is
> > altered just a bit later can then get a coarse-grained one that appears
> > older than the earlier fine-grained time. This violates timestamp
> > ordering guarantees.
> > 
> > To remedy this, keep a global monotonic atomic64_t value that acts as a
> > timestamp floor.  When we go to stamp a file, we first get the latter of
> > the current floor value and the current coarse-grained time. If the
> > inode ctime hasn't been queried then we just attempt to stamp it with
> > that value.
> > 
> > If it has been queried, then first see whether the current coarse time
> > is later than the existing ctime. If it is, then we accept that value.
> > If it isn't, then we get a fine-grained timestamp.
> > 
> > Filesystems can opt into this by setting the FS_MGTIME fstype flag.
> > Others should be unaffected (other than being subject to the same floor
> > value as multigrain filesystems).
> > 
> > Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
> > Signed-off-by: Jeff Layton <jlayton@kernel.org>
> 
> Mostly looks good. Some smaller comments below.
> 
> > +/**
> > + * current_time - Return FS time (possibly fine-grained)
> > + * @inode: inode.
> > + *
> > + * Return the current time truncated to the time granularity supported by
> > + * the fs, as suitable for a ctime/mtime change. If the ctime is flagged
> > + * as having been QUERIED, get a fine-grained timestamp, but don't update
> > + * the floor.
> > + *
> > + * For a multigrain inode, this is effectively an estimate of the timestamp
> > + * that a file would receive. An actual update must go through
> > + * inode_set_ctime_current().
> > + */
> > +struct timespec64 current_time(struct inode *inode)
> > +{
> > +	struct timespec64 now;
> > +	u32 cns;
> > +
> > +	ktime_get_coarse_real_ts64_mg(&now);
> > +
> > +	if (!is_mgtime(inode))
> > +		goto out;
> > +
> > +	/* If nothing has queried it, then coarse time is fine */
> > +	cns = smp_load_acquire(&inode->i_ctime_nsec);
> > +	if (cns & I_CTIME_QUERIED) {
> > +		/*
> > +		 * If there is no apparent change, then get a fine-grained
> > +		 * timestamp.
> > +		 */
> > +		if (now.tv_nsec == (cns & ~I_CTIME_QUERIED))
> > +			ktime_get_real_ts64(&now);
> > +	}
> > +out:
> > +	return timestamp_truncate(now, inode);
> > +}
> > +EXPORT_SYMBOL(current_time);
> > +
> >  static int inode_needs_update_time(struct inode *inode)
> >  {
> > +	struct timespec64 now, ts;
> >  	int sync_it = 0;
> > -	struct timespec64 now = current_time(inode);
> > -	struct timespec64 ts;
> >  
> >  	/* First try to exhaust all avenues to not sync */
> >  	if (IS_NOCMTIME(inode))
> >  		return 0;
> >  
> > +	now = current_time(inode);
> > +
> >  	ts = inode_get_mtime(inode);
> >  	if (!timespec64_equal(&ts, &now))
> > -		sync_it = S_MTIME;
> > +		sync_it |= S_MTIME;
> >  
> >  	ts = inode_get_ctime(inode);
> >  	if (!timespec64_equal(&ts, &now))
> > @@ -2598,6 +2637,15 @@ void inode_nohighmem(struct inode *inode)
> >  }
> >  EXPORT_SYMBOL(inode_nohighmem);
> >  
> > +struct timespec64 inode_set_ctime_to_ts(struct inode *inode, struct timespec64 ts)
> > +{
> > +	set_normalized_timespec64(&ts, ts.tv_sec, ts.tv_nsec);
> > +	inode->i_ctime_sec = ts.tv_sec;
> > +	inode->i_ctime_nsec = ts.tv_nsec;
> > +	return ts;
> > +}
> > +EXPORT_SYMBOL(inode_set_ctime_to_ts);
> > +
> >  /**
> >   * timestamp_truncate - Truncate timespec to a granularity
> >   * @t: Timespec
> > @@ -2630,36 +2678,75 @@ struct timespec64 timestamp_truncate(struct timespec64 t, struct inode *inode)
> >  EXPORT_SYMBOL(timestamp_truncate);
> >  
> >  /**
> > - * current_time - Return FS time
> > - * @inode: inode.
> > + * inode_set_ctime_current - set the ctime to current_time
> > + * @inode: inode
> >   *
> > - * Return the current time truncated to the time granularity supported by
> > - * the fs.
> > + * Set the inode's ctime to the current value for the inode. Returns the
> > + * current value that was assigned. If this is not a multigrain inode, then we
> > + * set it to the later of the coarse time and floor value.
> >   *
> > - * Note that inode and inode->sb cannot be NULL.
> > - * Otherwise, the function warns and returns time without truncation.
> > + * If it is multigrain, then we first see if the coarse-grained timestamp is
> > + * distinct from what we have. If so, then we'll just use that. If we have to
> > + * get a fine-grained timestamp, then do so, and try to swap it into the floor.
> > + * We accept the new floor value regardless of the outcome of the cmpxchg.
> > + * After that, we try to swap the new value into i_ctime_nsec. Again, we take
> > + * the resulting ctime, regardless of the outcome of the swap.
> 
> This comment seems outdated now. No floor in this function anymore...
> 

True. Will fix.

> > -struct timespec64 current_time(struct inode *inode)
> > +struct timespec64 inode_set_ctime_current(struct inode *inode)
> >  {
> >  	struct timespec64 now;
> > +	u32 cns, cur;
> ...
> 
> > diff --git a/fs/stat.c b/fs/stat.c
> > index 41e598376d7e..381926fb405f 100644
> > --- a/fs/stat.c
> > +++ b/fs/stat.c
> > @@ -26,6 +26,35 @@
> >  #include "internal.h"
> >  #include "mount.h"
> >  
> > +/**
> > + * fill_mg_cmtime - Fill in the mtime and ctime and flag ctime as QUERIED
> > + * @stat: where to store the resulting values
> > + * @request_mask: STATX_* values requested
> > + * @inode: inode from which to grab the c/mtime
> > + *
> > + * Given @inode, grab the ctime and mtime out if it and store the result
> 						 ^^ of
> 
> > + * in @stat. When fetching the value, flag it as QUERIED (if not already)
> > + * so the next write will record a distinct timestamp.
> > + */
> > +void fill_mg_cmtime(struct kstat *stat, u32 request_mask, struct inode *inode)
> > +{
> 
> Given how things worked out in the end, it seems this function doesn't need
> to handle mtime at all and we can move mtime handling back to shared generic
> code?
> 

I don't think we can. The mtime is effectively derived from the ctime.

If I query only the mtime, I think it's reasonable to expect that it
will change if there is another write, even if I don't query the ctime.
We won't get that unless we can also set the flag in the ctime when
only the mtime is requested.

> > +	atomic_t *pcn = (atomic_t *)&inode->i_ctime_nsec;
> > +
> > +	/* If neither time was requested, then don't report them */
> > +	if (!(request_mask & (STATX_CTIME|STATX_MTIME))) {
> > +		stat->result_mask &= ~(STATX_CTIME|STATX_MTIME);
> > +		return;
> > +	}
> > +
> > +	stat->mtime = inode_get_mtime(inode);
> > +	stat->ctime.tv_sec = inode->i_ctime_sec;
> > +	stat->ctime.tv_nsec = (u32)atomic_read(pcn);
> > +	if (!(stat->ctime.tv_nsec & I_CTIME_QUERIED))
> > +		stat->ctime.tv_nsec = ((u32)atomic_fetch_or(I_CTIME_QUERIED, pcn));
> > +	stat->ctime.tv_nsec &= ~I_CTIME_QUERIED;
> > +}
> > +EXPORT_SYMBOL(fill_mg_cmtime);
> > +
> >  /**
> >   * generic_fillattr - Fill in the basic attributes from the inode struct
> >   * @idmap:		idmap of the mount the inode was found from
> > @@ -58,8 +87,14 @@ void generic_fillattr(struct mnt_idmap *idmap, u32 request_mask,
> >  	stat->rdev = inode->i_rdev;
> >  	stat->size = i_size_read(inode);
> >  	stat->atime = inode_get_atime(inode);
> > -	stat->mtime = inode_get_mtime(inode);
> > -	stat->ctime = inode_get_ctime(inode);
> > +
> > +	if (is_mgtime(inode)) {
> > +		fill_mg_cmtime(stat, request_mask, inode);
> > +	} else {
> > +		stat->ctime = inode_get_ctime(inode);
> > +		stat->mtime = inode_get_mtime(inode);
> > +	}
> > +
> >  	stat->blksize = i_blocksize(inode);
> >  	stat->blocks = inode->i_blocks;
> >  
> > diff --git a/include/linux/fs.h b/include/linux/fs.h
> > index e3c603d01337..23908bad166c 100644
> > --- a/include/linux/fs.h
> > +++ b/include/linux/fs.h
> > @@ -1653,6 +1653,17 @@ static inline struct timespec64 inode_set_mtime(struct inode *inode,
> >  	return inode_set_mtime_to_ts(inode, ts);
> >  }
> >  
> > +/*
> > + * Multigrain timestamps
> > + *
> > + * Conditionally use fine-grained ctime and mtime timestamps when there
> > + * are users actively observing them via getattr. The primary use-case
> > + * for this is NFS clients that use the ctime to distinguish between
> > + * different states of the file, and that are often fooled by multiple
> > + * operations that occur in the same coarse-grained timer tick.
> 
> Again, mtime seems unaffected by mgtime changes now.
> 

I still think we need this.
 
> > + */
> > +#define I_CTIME_QUERIED		((u32)BIT(31))
> > +
> 
> 								Honza

-- 
Jeff Layton <jlayton@kernel.org>

Re: [PATCH v8 01/12] timekeeping: add interfaces for handling timestamps with a floor value

[John Stultz]

On Tue, Oct 1, 2024 at 3:59 AM Jeff Layton <jlayton@kernel.org> wrote:
>
> Multigrain timestamps allow the kernel to use fine-grained timestamps
> when an inode's attributes is being actively observed via ->getattr().
> With this support, it's possible for a file to get a fine-grained
> timestamp, and another modified after it to get a coarse-grained stamp
> that is earlier than the fine-grained time.  If this happens then the
> files can appear to have been modified in reverse order, which breaks
> VFS ordering guarantees.
>
> To prevent this, maintain a floor value for multigrain timestamps.
> Whenever a fine-grained timestamp is handed out, record it, and when
> coarse-grained stamps are handed out, ensure they are not earlier than
> that value. If the coarse-grained timestamp is earlier than the
> fine-grained floor, return the floor value instead.
>
> Add a static singleton atomic64_t into timekeeper.c that we can use to
> keep track of the latest fine-grained time ever handed out. This is
> tracked as a monotonic ktime_t value to ensure that it isn't affected by
> clock jumps. Because it is updated at different times than the rest of
> the timekeeper object, the floor value is managed independently of the
> timekeeper via a cmpxchg() operation, and sits on its own cacheline.
>
> This patch also adds two new public interfaces:
>
> - ktime_get_coarse_real_ts64_mg() fills a timespec64 with the later of the
>   coarse-grained clock and the floor time
>
> - ktime_get_real_ts64_mg() gets the fine-grained clock value, and tries
>   to swap it into the floor. A timespec64 is filled with the result.
>
> Since the floor is global, take care to avoid updating it unless it's
> absolutely necessary. If we do the cmpxchg and find that the value has
> been updated since we fetched it, then we discard the fine-grained time
> that was fetched in favor of the recent update.
>
> Note that the VFS ordering guarantees assume that the realtime clock
> does not experience a backward jump. POSIX requires that we stamp files
> using realtime clock values, so if a backward clock jump occurs, then
> files can appear to have been modified in reverse order.
>
> Tested-by: Randy Dunlap <rdunlap@infradead.org> # documentation bits
> Signed-off-by: Jeff Layton <jlayton@kernel.org>

Acked-by: John Stultz <jstultz@google.com>

Re: [PATCH v8 02/12] fs: add infrastructure for multigrain timestamps

[Jan Kara]

On Tue 01-10-24 09:34:18, Jeff Layton wrote:
> On Tue, 2024-10-01 at 15:20 +0200, Jan Kara wrote:
> > > diff --git a/fs/stat.c b/fs/stat.c
> > > index 41e598376d7e..381926fb405f 100644
> > > --- a/fs/stat.c
> > > +++ b/fs/stat.c
> > > @@ -26,6 +26,35 @@
> > >  #include "internal.h"
> > >  #include "mount.h"
> > >  
> > > +/**
> > > + * fill_mg_cmtime - Fill in the mtime and ctime and flag ctime as QUERIED
> > > + * @stat: where to store the resulting values
> > > + * @request_mask: STATX_* values requested
> > > + * @inode: inode from which to grab the c/mtime
> > > + *
> > > + * Given @inode, grab the ctime and mtime out if it and store the result
> > 						 ^^ of
> > 
> > > + * in @stat. When fetching the value, flag it as QUERIED (if not already)
> > > + * so the next write will record a distinct timestamp.
> > > + */
> > > +void fill_mg_cmtime(struct kstat *stat, u32 request_mask, struct inode *inode)
> > > +{
> > 
> > Given how things worked out in the end, it seems this function doesn't need
> > to handle mtime at all and we can move mtime handling back to shared generic
> > code?
> > 
> 
> I don't think we can. The mtime is effectively derived from the ctime.
> 
> If I query only the mtime, I think it's reasonable to expect that it
> will change if there is another write, even if I don't query the ctime.
> We won't get that unless we can also set the flag in the ctime when
> only the mtime is requested.

Aha, right. I already forgot about this :). Can you please add to the
comment the above explanation so that we remember next time somebody wants
to "clean this up" like me ;)? Thanks!

Also feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH v8 02/12] fs: add infrastructure for multigrain timestamps

[Jeff Layton]

On Wed, 2024-10-02 at 11:14 +0200, Jan Kara wrote:
> On Tue 01-10-24 09:34:18, Jeff Layton wrote:
> > On Tue, 2024-10-01 at 15:20 +0200, Jan Kara wrote:
> > > > diff --git a/fs/stat.c b/fs/stat.c
> > > > index 41e598376d7e..381926fb405f 100644
> > > > --- a/fs/stat.c
> > > > +++ b/fs/stat.c
> > > > @@ -26,6 +26,35 @@
> > > >  #include "internal.h"
> > > >  #include "mount.h"
> > > >  
> > > > +/**
> > > > + * fill_mg_cmtime - Fill in the mtime and ctime and flag ctime as QUERIED
> > > > + * @stat: where to store the resulting values
> > > > + * @request_mask: STATX_* values requested
> > > > + * @inode: inode from which to grab the c/mtime
> > > > + *
> > > > + * Given @inode, grab the ctime and mtime out if it and store the result
> > > 						 ^^ of
> > > 
> > > > + * in @stat. When fetching the value, flag it as QUERIED (if not already)
> > > > + * so the next write will record a distinct timestamp.
> > > > + */
> > > > +void fill_mg_cmtime(struct kstat *stat, u32 request_mask, struct inode *inode)
> > > > +{
> > > 
> > > Given how things worked out in the end, it seems this function doesn't need
> > > to handle mtime at all and we can move mtime handling back to shared generic
> > > code?
> > > 
> > 
> > I don't think we can. The mtime is effectively derived from the ctime.
> > 
> > If I query only the mtime, I think it's reasonable to expect that it
> > will change if there is another write, even if I don't query the ctime.
> > We won't get that unless we can also set the flag in the ctime when
> > only the mtime is requested.
> 
> Aha, right. I already forgot about this :). Can you please add to the
> comment the above explanation so that we remember next time somebody wants
> to "clean this up" like me ;)? Thanks!
> 
> Also feel free to add:
> 
> Reviewed-by: Jan Kara <jack@suse.cz>
> 
> 								Honza

Done, and the revised version is in my tree.

Thanks for the review!
-- 
Jeff Layton <jlayton@kernel.org>

Re: [PATCH v8 01/12] timekeeping: add interfaces for handling timestamps with a floor value

[Thomas Gleixner]

On Tue, Oct 01 2024 at 06:58, Jeff Layton wrote:

V8 ? I remember that I reviewed v* already :)

Also the sentence after the susbsystem prefix starts with an uppercase
letter.

> Multigrain timestamps allow the kernel to use fine-grained timestamps
> when an inode's attributes is being actively observed via ->getattr().
> With this support, it's possible for a file to get a fine-grained
> timestamp, and another modified after it to get a coarse-grained stamp
> that is earlier than the fine-grained time.  If this happens then the
> files can appear to have been modified in reverse order, which breaks
> VFS ordering guarantees.
>
> To prevent this, maintain a floor value for multigrain timestamps.
> Whenever a fine-grained timestamp is handed out, record it, and when
> coarse-grained stamps are handed out, ensure they are not earlier than
> that value. If the coarse-grained timestamp is earlier than the
> fine-grained floor, return the floor value instead.
>
> Add a static singleton atomic64_t into timekeeper.c that we can use to

s/we can use/is used/

> keep track of the latest fine-grained time ever handed out. This is
> tracked as a monotonic ktime_t value to ensure that it isn't affected by
> clock jumps. Because it is updated at different times than the rest of
> the timekeeper object, the floor value is managed independently of the
> timekeeper via a cmpxchg() operation, and sits on its own cacheline.
>
> This patch also adds two new public interfaces:

git grep 'This patch' Docuementation/process/

> - ktime_get_coarse_real_ts64_mg() fills a timespec64 with the later of the
>   coarse-grained clock and the floor time
>
> - ktime_get_real_ts64_mg() gets the fine-grained clock value, and tries
>   to swap it into the floor. A timespec64 is filled with the result.
>
> Since the floor is global, take care to avoid updating it unless it's
> absolutely necessary. If we do the cmpxchg and find that the value has

We do nothing. Please describe your changes in passive voice and do not
impersonate code.

> been updated since we fetched it, then we discard the fine-grained time
> that was fetched in favor of the recent update.

This still is confusing. Something like this:

  The floor value is global and updated via a single try_cmpxchg(). If
  that fails then the operation raced with a concurrent update. It does
  not matter whether the new value is later than the value, which the
  failed cmpxchg() tried to write, or not. Add sensible technical
  explanation.

> +/*
> + * Multigrain timestamps require that we keep track of the latest fine-grained
> + * timestamp that has been issued, and never return a coarse-grained timestamp
> + * that is earlier than that value.
> + *
> + * mg_floor represents the latest fine-grained time that we have handed out as
> + * a timestamp on the system. Tracked as a monotonic ktime_t, and converted to
> + * the realtime clock on an as-needed basis.
> + *
> + * This ensures that we never issue a timestamp earlier than one that has
> + * already been issued, as long as the realtime clock never experiences a
> + * backward clock jump. If the realtime clock is set to an earlier time, then
> + * realtime timestamps can appear to go backward.
> + */
> +static __cacheline_aligned_in_smp atomic64_t mg_floor;
> +
>  static inline void tk_normalize_xtime(struct timekeeper *tk)
>  {
>  	while (tk->tkr_mono.xtime_nsec >= ((u64)NSEC_PER_SEC << tk->tkr_mono.shift)) {
> @@ -2394,6 +2410,86 @@ void ktime_get_coarse_real_ts64(struct timespec64 *ts)
>  }
>  EXPORT_SYMBOL(ktime_get_coarse_real_ts64);
>  
> +/**
> + * ktime_get_coarse_real_ts64_mg - return latter of coarse grained time or floor
> + * @ts: timespec64 to be filled
> + *
> + * Fetch the global mg_floor value, convert it to realtime and
> + * compare it to the current coarse-grained time. Fill @ts with
> + * whichever is latest. Note that this is a filesystem-specific
> + * interface and should be avoided outside of that context.
> + */
> +void ktime_get_coarse_real_ts64_mg(struct timespec64 *ts)
> +{
> +	struct timekeeper *tk = &tk_core.timekeeper;
> +	u64 floor = atomic64_read(&mg_floor);
> +	ktime_t f_real, offset, coarse;
> +	unsigned int seq;
> +
> +	do {
> +		seq = read_seqcount_begin(&tk_core.seq);
> +		*ts = tk_xtime(tk);
> +		offset = tk_core.timekeeper.offs_real;
> +	} while (read_seqcount_retry(&tk_core.seq, seq));
> +
> +	coarse = timespec64_to_ktime(*ts);
> +	f_real = ktime_add(floor, offset);
> +	if (ktime_after(f_real, coarse))
> +		*ts = ktime_to_timespec64(f_real);
> +}
> +EXPORT_SYMBOL_GPL(ktime_get_coarse_real_ts64_mg);
> +
> +/**
> + * ktime_get_real_ts64_mg - attempt to update floor value and return result
> + * @ts:		pointer to the timespec to be set
> + *
> + * Get a monotonic fine-grained time value and attempt to swap it into the
> + * floor. If it succeeds then accept the new floor value. If it fails
> + * then another task raced in during the interim time and updated the floor.
> + * That value is just as valid, so accept that value in this case.

Why? 'just as valid' does not tell me anything.

> + * @ts will be filled with the resulting floor value, regardless of
> + * the outcome of the swap. Note that this is a filesystem specific interface
> + * and should be avoided outside of that context.
> + */
> +void ktime_get_real_ts64_mg(struct timespec64 *ts)
> +{
> +	struct timekeeper *tk = &tk_core.timekeeper;
> +	ktime_t old = atomic64_read(&mg_floor);
> +	ktime_t offset, mono;
> +	unsigned int seq;
> +	u64 nsecs;
> +
> +	do {
> +		seq = read_seqcount_begin(&tk_core.seq);
> +
> +		ts->tv_sec = tk->xtime_sec;
> +		mono = tk->tkr_mono.base;
> +		nsecs = timekeeping_get_ns(&tk->tkr_mono);
> +		offset = tk_core.timekeeper.offs_real;
> +	} while (read_seqcount_retry(&tk_core.seq, seq));
> +
> +	mono = ktime_add_ns(mono, nsecs);
> +
> +	/*
> +	 * Attempt to update the floor with the new time value. Accept the
> +	 * resulting floor value regardless of the outcome of the swap.
> +	 */
> +	if (atomic64_try_cmpxchg(&mg_floor, &old, mono)) {
> +		ts->tv_nsec = 0;
> +		timespec64_add_ns(ts, nsecs);
> +	} else {
> +		/*
> +		 * Something has changed mg_floor since "old" was

I complained about 'something has changed ...' before. Can we please
have proper technical explanations?

> +		 * fetched. "old" has now been updated with the
> +		 * current value of mg_floor, so use that to return
> +		 * the current coarse floor value.

This still does not explain why the new floor value is valid even when
it is before the value in @mono.

Thanks,

        tglx