Re: [PATCH v3 1/6] fs: add infrastructure for multigrain inode i_m/ctime

[Dave Chinner]

On Wed, May 03, 2023 at 10:20:32AM -0400, Jeff Layton wrote:
> The VFS always uses coarse-grained timestamp updates for filling out 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. 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 (e.g backup applications).
> 
> Switching to always using fine-grained timestamps would improve the
> situation, but that becomes rather expensive, as the underlying
> filesystem will 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.
> 
> The kernel always stores normalized ctime values, so only the first 30
> bits of the tv_nsec field are ever used. Whenever the mtime changes, the
> ctime must also change.
> 
> Use the 31st bit of the tv_nsec field to indicate that something has
> queried the inode for the i_mtime or i_ctime. When this flag is set, on
> the next timestamp update, the kernel can fetch a fine-grained timestamp
> instead of the usual coarse-grained one.
> 
> This patch adds the infrastructure this scheme. Filesytems can opt
> into it by setting the FS_MULTIGRAIN_TS flag in the fstype.
> 
> Later patches will convert individual filesystems over to use it.
> 
> Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx>
> ---
>  fs/inode.c         | 52 ++++++++++++++++++++++++++++++++++++---
>  fs/stat.c          | 32 ++++++++++++++++++++++++
>  include/linux/fs.h | 61 +++++++++++++++++++++++++++++++++++++++++++++-
>  3 files changed, 141 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/inode.c b/fs/inode.c
> index 4558dc2f1355..7f6189961d6a 100644
> --- a/fs/inode.c
> +++ b/fs/inode.c
> @@ -2030,6 +2030,7 @@ EXPORT_SYMBOL(file_remove_privs);
>  static int inode_needs_update_time(struct inode *inode, struct timespec64 *now)
>  {
>  	int sync_it = 0;
> +	struct timespec64 ctime;
>  
>  	/* First try to exhaust all avenues to not sync */
>  	if (IS_NOCMTIME(inode))
> @@ -2038,7 +2039,8 @@ static int inode_needs_update_time(struct inode *inode, struct timespec64 *now)
>  	if (!timespec64_equal(&inode->i_mtime, now))
>  		sync_it = S_MTIME;
>  
> -	if (!timespec64_equal(&inode->i_ctime, now))
> +	ctime = ctime_peek(inode);
> +	if (!timespec64_equal(&ctime, now))
>  		sync_it |= S_CTIME;
>  
>  	if (IS_I_VERSION(inode) && inode_iversion_need_inc(inode))
> @@ -2062,6 +2064,50 @@ static int __file_update_time(struct file *file, struct timespec64 *now,
>  	return ret;
>  }
>  
> +/**
> + * current_ctime - 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.
> + *
> + * For a multigrain timestamp, if the ctime is flagged as having been
> + * QUERIED, get a fine-grained timestamp.
> + */
> +struct timespec64 current_ctime(struct inode *inode)
> +{
> +	bool multigrain = is_multigrain_ts(inode);
> +	struct timespec64 now;
> +	long nsec = 0;
> +
> +	if (multigrain) {
> +		atomic_long_t *pnsec = (atomic_long_t *)&inode->i_ctime.tv_nsec;
> +
> +		nsec = atomic_long_fetch_andnot(I_CTIME_QUERIED, pnsec);
> +	}
> +
> +	if (nsec & I_CTIME_QUERIED) {
> +		ktime_get_real_ts64(&now);
> +	} else {
> +		ktime_get_coarse_real_ts64(&now);
> +
> +		if (multigrain) {
> +			/*
> +			 * If we've recently fetched a fine-grained timestamp
> +			 * then the coarse-grained one may be earlier than the
> +			 * existing one. Just keep the existing ctime if so.
> +			 */
> +			struct timespec64 ctime = ctime_peek(inode);
> +
> +			if (timespec64_compare(&ctime, &now) > 0)
> +				now = ctime;
> +		}
> +	}
> +
> +	return timestamp_truncate(now, inode);
> +}
> +EXPORT_SYMBOL(current_ctime);

I can't help but think this is easier to read/follow when structured
to separate multigrain vs coarse logic completely like so:

struct timespec64 current_ctime(struct inode *inode)
{
	struct timespec64 now, ctime;
	long nsec;

	if (!is_multigrain_ts(inode)) {
		ktime_get_coarse_real_ts64(&now);
		goto out_truncate;
	}

	nsec = atomic_long_fetch_andnot(I_CTIME_QUERIED,
			(atomic_long_t *)&inode->i_ctime.tv_nsec);

	if (nsec & I_CTIME_QUERIED) {
		ktime_get_real_ts64(&now);
		goto out_truncate;
	}

	/*
	 * If we've recently fetched a fine-grained timestamp then
	 * the coarse-grained one may be earlier than the existing
	 * one. Just keep the existing ctime if so.
	 */
	ktime_get_coarse_real_ts64(&now);
	ctime = ctime_peek(inode);
	if (timespec64_compare(&ctime, &now) > 0)
		now = ctime;

out_truncate:
	return timestamp_truncate(now, inode);
}

> diff --git a/fs/stat.c b/fs/stat.c
> index 7c238da22ef0..11a7e277f53e 100644
> --- a/fs/stat.c
> +++ b/fs/stat.c
> @@ -26,6 +26,38 @@
>  #include "internal.h"
>  #include "mount.h"
>  
> +/**
> + * generic_fill_multigrain_cmtime - Fill in the mtime and ctime and flag ctime as QUERIED
> + * @request_mask: STATX_* values requested
> + * @inode: inode from which to grab the c/mtime
> + * @stat: where to store the resulting values
> + *
> + * Given @inode, grab the ctime and mtime out if it and store the result
> + * in @stat. When fetching the value, flag it as queried so the next write
> + * will use a fine-grained timestamp.
> + */
> +void generic_fill_multigrain_cmtime(u32 request_mask,struct inode *inode,
> +					struct kstat *stat)
> +{
> +	atomic_long_t *pnsec = (atomic_long_t *)&inode->i_ctime.tv_nsec;
> +
> +	/* If neither time was requested, then just don't report it */
> +	if (!(request_mask & (STATX_CTIME|STATX_MTIME))) {
> +		stat->result_mask &= ~(STATX_CTIME|STATX_MTIME);
> +		return;
> +	}
> +
> +	stat->mtime = inode->i_mtime;
> +	stat->ctime.tv_sec = inode->i_ctime.tv_sec;
> +	/*
> +	 * Atomically set the QUERIED flag and fetch the new value with
> +	 * the flag masked off.
> +	 */
> +	stat->ctime.tv_nsec = atomic_long_fetch_or(I_CTIME_QUERIED, pnsec) &
> +					~I_CTIME_QUERIED;
> +}
> +EXPORT_SYMBOL(generic_fill_multigrain_cmtime);

Hmmm - why not just have a generic_fill_cmtime() function that hides
multigrain behaviour from all the statx callers?

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx