public inbox for linux-trace-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Jeff Layton <jlayton@kernel.org>
To: Chuck Lever <chuck.lever@oracle.com>,
	Alexander Viro	 <viro@zeniv.linux.org.uk>,
	Christian Brauner <brauner@kernel.org>, Jan Kara <jack@suse.cz>,
	Steven Rostedt <rostedt@goodmis.org>,
	Masami Hiramatsu	 <mhiramat@kernel.org>,
	Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
	 NeilBrown <neil@brown.name>,
	Olga Kornievskaia <okorniev@redhat.com>,
	Dai Ngo <Dai.Ngo@oracle.com>,  Tom Talpey <tom@talpey.com>
Cc: Trond Myklebust <trondmy@hammerspace.com>,
	Anna Schumaker <anna@kernel.org>,
		linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
		linux-trace-kernel@vger.kernel.org, linux-nfs@vger.kernel.org
Subject: Re: [PATCH v2 3/7] nfsd: use ATTR_CTIME_SET for delegated ctime updates
Date: Sat, 26 Jul 2025 15:03:53 -0400	[thread overview]
Message-ID: <8ec5b19dc1d0ce26f1cd86d7db2ba5a2d260c073.camel@kernel.org> (raw)
In-Reply-To: <5f877de4-347c-484c-814f-33c08f1a5189@oracle.com>

On Sat, 2025-07-26 at 14:48 -0400, Chuck Lever wrote:
> Hi Jeff -
> 
> Thanks again for your focus on getting this straightened out!
> 
> 
> On 7/26/25 10:31 AM, Jeff Layton wrote:
> > Ensure that notify_change() doesn't clobber a delegated ctime update
> > with current_time() by setting ATTR_CTIME_SET for those updates.
> > 
> > Also, set the tv_nsec field the nfsd4_decode_fattr4 to the correct
> > value.
> 
> I don't yet see the connection of the above tv_nsec fix to the other
> changes in this patch. Wouldn't this be an independent fix?
> 

I felt like they were related. Yes, the ia_ctime field is currently
being set wrong, but it's also being clobbered by notify_change(), so
it doesn't matter much. I can break this into a separate patch (with a
Fixes: tag) if you prefer though.

> > Don't bother setting the timestamps in cb_getattr_update_times() in the
> > non-delegated case. notify_change() will do that itself.
> > 
> > Signed-off-by: Jeff Layton <jlayton@kernel.org>
> 
> General comments:
> 
> I don't feel that any of the patches in this series need to be tagged
> for stable, since there is already a Kconfig setting that defaults to
> leaving timestamp delegation disabled. But I would like to see Fixes:
> tags, where that makes sense?
> 

I don't think any of these need to go to stable since this is still
under a non-default Kconfig option, and the main effect of the bug is
wonky timestamps. I should be able to add some Fixes: tags though.

> Is this set on top of the set you posted a day or two ago with the new
> trace point? Or does this set replace that one?
> 

This set should replace those.

> 
> > ---
> >  fs/nfsd/nfs4state.c | 6 +++---
> >  fs/nfsd/nfs4xdr.c   | 5 +++--
> >  2 files changed, 6 insertions(+), 5 deletions(-)
> > 
> > diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
> > index 88c347957da5b8f352be63f84f207d2225f81cb9..77eea2ad93cc07939f045fc4b983b1ac00d068b8 100644
> > --- a/fs/nfsd/nfs4state.c
> > +++ b/fs/nfsd/nfs4state.c
> > @@ -9167,7 +9167,6 @@ static bool set_cb_time(struct timespec64 *cb, const struct timespec64 *orig,
> >  static int cb_getattr_update_times(struct dentry *dentry, struct nfs4_delegation *dp)
> >  {
> >  	struct inode *inode = d_inode(dentry);
> > -	struct timespec64 now = current_time(inode);
> >  	struct nfs4_cb_fattr *ncf = &dp->dl_cb_fattr;
> >  	struct iattr attrs = { };
> >  	int ret;
> > @@ -9175,6 +9174,7 @@ static int cb_getattr_update_times(struct dentry *dentry, struct nfs4_delegation
> >  	if (deleg_attrs_deleg(dp->dl_type)) {
> >  		struct timespec64 atime = inode_get_atime(inode);
> >  		struct timespec64 mtime = inode_get_mtime(inode);
> > +		struct timespec64 now = current_time(inode);
> >  
> >  		attrs.ia_atime = ncf->ncf_cb_atime;
> >  		attrs.ia_mtime = ncf->ncf_cb_mtime;
> > @@ -9183,12 +9183,12 @@ static int cb_getattr_update_times(struct dentry *dentry, struct nfs4_delegation
> >  			attrs.ia_valid |= ATTR_ATIME | ATTR_ATIME_SET;
> >  
> >  		if (set_cb_time(&attrs.ia_mtime, &mtime, &now)) {
> > -			attrs.ia_valid |= ATTR_CTIME | ATTR_MTIME | ATTR_MTIME_SET;
> > +			attrs.ia_valid |= ATTR_CTIME | ATTR_CTIME_SET |
> > +					  ATTR_MTIME | ATTR_MTIME_SET;
> >  			attrs.ia_ctime = attrs.ia_mtime;
> >  		}
> >  	} else {
> >  		attrs.ia_valid |= ATTR_MTIME | ATTR_CTIME;
> > -		attrs.ia_mtime = attrs.ia_ctime = now;
> >  	}
> >  
> >  	if (!attrs.ia_valid)
> > diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
> > index 8b68f74a8cf08c6aa1305a2a3093467656085e4a..c0a3c6a7c8bb70d62940115c3101e9f897401456 100644
> > --- a/fs/nfsd/nfs4xdr.c
> > +++ b/fs/nfsd/nfs4xdr.c
> > @@ -538,8 +538,9 @@ nfsd4_decode_fattr4(struct nfsd4_compoundargs *argp, u32 *bmval, u32 bmlen,
> >  		iattr->ia_mtime.tv_sec = modify.seconds;
> >  		iattr->ia_mtime.tv_nsec = modify.nseconds;
> >  		iattr->ia_ctime.tv_sec = modify.seconds;
> > -		iattr->ia_ctime.tv_nsec = modify.seconds;
> > -		iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME | ATTR_MTIME_SET | ATTR_DELEG;
> > +		iattr->ia_ctime.tv_nsec = modify.nseconds;
> > +		iattr->ia_valid |= ATTR_CTIME | ATTR_CTIME_SET |
> > +				   ATTR_MTIME | ATTR_MTIME_SET | ATTR_DELEG;
> >  	}
> >  
> >  	/* request sanity: did attrlist4 contain the expected number of words? */
> > 
> 
> 

-- 
Jeff Layton <jlayton@kernel.org>

  reply	other threads:[~2025-07-26 19:03 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-26 14:31 [PATCH v2 0/7] nfsd/vfs: fix handling of delegated timestamp updates Jeff Layton
2025-07-26 14:31 ` [PATCH v2 1/7] vfs: add ATTR_CTIME_SET flag Jeff Layton
2025-07-26 14:31 ` [PATCH v2 2/7] nfsd: ignore ATTR_DELEG when checking ia_valid before notify_change() Jeff Layton
2025-07-26 14:31 ` [PATCH v2 3/7] nfsd: use ATTR_CTIME_SET for delegated ctime updates Jeff Layton
2025-07-26 18:48   ` Chuck Lever
2025-07-26 19:03     ` Jeff Layton [this message]
2025-07-26 19:57       ` Chuck Lever
2025-07-26 22:25         ` Jeff Layton
2025-07-26 14:31 ` [PATCH v2 4/7] nfsd: track original timestamps in nfs4_delegation Jeff Layton
2025-07-26 14:31 ` [PATCH v2 5/7] nfsd: fix SETATTR updates for delegated timestamps Jeff Layton
2025-07-26 14:32 ` [PATCH v2 6/7] nfsd: fix timestamp updates in CB_GETATTR Jeff Layton
2025-07-26 14:32 ` [PATCH v2 7/7] vfs: remove inode_set_ctime_deleg() Jeff Layton

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=8ec5b19dc1d0ce26f1cd86d7db2ba5a2d260c073.camel@kernel.org \
    --to=jlayton@kernel.org \
    --cc=Dai.Ngo@oracle.com \
    --cc=anna@kernel.org \
    --cc=brauner@kernel.org \
    --cc=chuck.lever@oracle.com \
    --cc=jack@suse.cz \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-nfs@vger.kernel.org \
    --cc=linux-trace-kernel@vger.kernel.org \
    --cc=mathieu.desnoyers@efficios.com \
    --cc=mhiramat@kernel.org \
    --cc=neil@brown.name \
    --cc=okorniev@redhat.com \
    --cc=rostedt@goodmis.org \
    --cc=tom@talpey.com \
    --cc=trondmy@hammerspace.com \
    --cc=viro@zeniv.linux.org.uk \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox