Re: [PATCH 15/79] bfs: switch to new ctime accessors

From: Jan Kara
Date: Thu Jun 22 2023 - 08:30:58 EST


On Wed 21-06-23 12:57:19, Jeff Layton wrote:
> On Wed, 2023-06-21 at 18:48 +0200, Jan Kara wrote:
> > On Wed 21-06-23 10:45:28, Jeff Layton wrote:
> > > In later patches, we're going to change how the ctime.tv_nsec field is
> > > utilized. Switch to using accessor functions instead of raw accesses of
> > > inode->i_ctime.
> > >
> > > Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx>
> >
> > ...
> >
> > > diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c
> > > index 1926bec2c850..c964316be32b 100644
> > > --- a/fs/bfs/inode.c
> > > +++ b/fs/bfs/inode.c
> > > @@ -82,10 +82,10 @@ struct inode *bfs_iget(struct super_block *sb, unsigned long ino)
> > > inode->i_blocks = BFS_FILEBLOCKS(di);
> > > inode->i_atime.tv_sec = le32_to_cpu(di->i_atime);
> > > inode->i_mtime.tv_sec = le32_to_cpu(di->i_mtime);
> > > - inode->i_ctime.tv_sec = le32_to_cpu(di->i_ctime);
> > > + inode_ctime_set_sec(inode, le32_to_cpu(di->i_ctime));
> > > inode->i_atime.tv_nsec = 0;
> > > inode->i_mtime.tv_nsec = 0;
> > > - inode->i_ctime.tv_nsec = 0;
> > > + inode_ctime_set_nsec(inode, 0);
> >
> > So I'm somewhat wondering here - in other filesystem you construct
> > timespec64 and then use inode_ctime_set(). Here you use
> > inode_ctime_set_sec() + inode_ctime_set_nsec(). What's the benefit? It
> > seems these two functions are not used that much some maybe we could just
> > live with just inode_ctime_set() and constructing timespec64 when needed?
> >
> > Honza
>
> The main advantage is that by using that, I didn't need to do quite so
> much of this conversion by hand. My coccinelle skills are pretty
> primitive. I went with whatever conversion was going to give minimal
> changes, to the existing accesses for the most part.
>
> We could certainly do it the way you suggest, it just means having to
> re-touch a lot of this code by hand, or someone with better coccinelle
> chops suggesting a way to declare a temporary variables in place.

Well, maybe temporary variables aren't that convenient but we could provide
function setting ctime from sec & nsec value without having to declare
temporary timespec64? Attached is a semantic patch that should deal with
that - at least it seems to handle all the cases I've found.

Honza
--
Jan Kara <jack@xxxxxxxx>
SUSE Labs, CR
@@
struct inode *inode;
expression val, E1, E2;
@@
- inode->i_ctime.tv_sec = val
+ inode_set_ctime(inode, val, 0)
...
(
- inode->i_ctime.tv_nsec = 0;
|
- E1 = inode->i_ctime.tv_nsec = 0
+ E1 = 0
|
- inode->i_ctime.tv_nsec = E1 = 0
+ E1 = 0
|
- inode->i_ctime.tv_nsec = E1 = E2 = 0
+ E1 = E2 = 0
)

@@
struct inode *inode;
expression val, val2;
@@
- inode->i_ctime.tv_sec = val
+ inode_set_ctime(inode, val, val2)
...
- inode->i_ctime.tv_nsec = val2;