Re: [PATCH 1/2] fs: initialize inode->__i_ctime to the epoch

[Jeff Layton]

On Fri, 2023-09-08 at 12:42 +0200, Jan Kara wrote:
> On Thu 07-09-23 12:33:47, Jeff Layton wrote:
> > With the advent of multigrain timestamps, we use inode_set_ctime_current
> > to set the ctime, which can skip updating if the existing ctime appears
> > to be in the future. Because we don't initialize this field at
> > allocation time, that could prevent the ctime from being initialized
> > properly when the inode is instantiated.
> > 
> > Always initialize the ctime field to the epoch so that the filesystem
> > can set the timestamps properly later.
> > 
> > Reported-by: kernel test robot <oliver.sang@xxxxxxxxx>
> > Closes: https://lore.kernel.org/oe-lkp/202309071017.a64aca5e-oliver.sang@xxxxxxxxx
> > Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx>
> 
> Looks good but don't you need the same treatment to atime after your patch
> 2/2?
> 
> 

I don't think so. Most filesystems are doing something along the lines
of this when allocating a new inode:

    inode->i_atime = inode->i_mtime = inode_set_ctime_current(inode);

...and I think they pretty much all have to initialize i_atime properly,
since someone could stat the inode before an atime update occurs.

> > ---
> >  fs/inode.c | 2 ++
> >  1 file changed, 2 insertions(+)
> > 
> > diff --git a/fs/inode.c b/fs/inode.c
> > index 35fd688168c5..54237f4242ff 100644
> > --- a/fs/inode.c
> > +++ b/fs/inode.c
> > @@ -168,6 +168,8 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
> >  	inode->i_fop = &no_open_fops;
> >  	inode->i_ino = 0;
> >  	inode->__i_nlink = 1;
> > +	inode->__i_ctime.tv_sec = 0;
> > +	inode->__i_ctime.tv_nsec = 0;
> >  	inode->i_opflags = 0;
> >  	if (sb->s_xattr)
> >  		inode->i_opflags |= IOP_XATTR;
> > 
> > -- 
> > 2.41.0
> > 

-- 
Jeff Layton <jlayton@xxxxxxxxxx>