Re: [PATCH 1/2] fs: preserve ACL_DONT_CACHE state in forget_cached_acl()

On Wed, Jul 15, 2026 at 11:02 AM Christian Brauner <[email protected]> wrote:
>
> > The ACL_DONT_CACHE state is meant to be a constant state for the inode
> > for filesystems that want to opt out of posix acl caching.
> >
> > Commit facd61053cff1 ("fuse: fixes after adapting to new posix acl api")
> > used this facility to opt out of posix acl caching for fuse inodes with
> > fuse server that does not negotiate FUSE_POSIX_ACL (fc->posix_acl).
> >
> > The commit also takes care to gate the forget_all_cached_acls() call in
> > fuse_set_acl() on fc->posix_acl because there is no need for it, but
> > there are other placed in fuse code which call forget_all_cached_acls()
> > unconditional to fc->posix_acl and those cause the loss of the
> > ACL_DONT_CACHE state.
> >
> > This is not only a functional bug. Properly timed, a get_acl() from this
> > fuse filesystem can return a stale cached value, as was observed in tests,
> > because set_acl() does not invalidate the unintentional acl cache.
> >
> > We could fix this in fuse, but it actually makes no sense for the vfs
> > helper forget_cached_acl() to invalidate the ACL_DONT_CACHE state, so
> > let it not do that to fix fuse and future users of ACL_DONT_CACHE.
> >
> > Fixes: facd61053cff1 ("fuse: fixes after adapting to new posix acl api")
> > Cc: [email protected]
> > Signed-off-by: Amir Goldstein <[email protected]>
> >
> > diff --git a/fs/posix_acl.c b/fs/posix_acl.c
> > index b4bfe4ddf64e..3dc62c1c2708 100644
> > --- a/fs/posix_acl.c
> > +++ b/fs/posix_acl.c
> > @@ -93,6 +93,13 @@ static void __forget_cached_acl(struct posix_acl **p)
> >  {
> >       struct posix_acl *old;
> >
> > +     /*
> > +      * ACL_DONT_CACHE is expected to be a "const" value and xchg it with
> > +      * ACL_NOT_CACHED would enable acl caching for the inode -
> > +      * clearly not what the caller has intended.
> > +      */
> > +     if (READ_ONCE(*p) == ACL_DONT_CACHE)
> > +             return;
>
> Still on vacation this week but I took a glimpse:
>
> If this isn't what the caller intended, having ACL_DONT_CACHE end up
> should be treated like a bug. So shouldn't this then be a WARN_ON_ONCE()
> and return?

The caller has requested to *forget* any *cached* *acl*.
The value ACL_DONT_CACHE means that there are no cached acls to forget
so the call is a success.
Therefore, WARN_ON is not called for IMO.

Sorry to interrupt vacation.
Let's continue when you get back.

Thanks,
Amir.