Re: [PATCH v2] btrfs: send: fix is_current_inode_path() to avoid path resets for common prefixes

On Tue, Jun 23, 2026 at 11:38 PM Qu Wenruo <[email protected]> wrote:
>
>
>
> 在 2026/6/16 02:44, [email protected] 写道:
> > From: Filipe Manana <[email protected]>
> >
> > In case the current inode's path is a prefix of the given path, the helper
> > is_current_inode_path() will return true, which causes the single caller
> > to reset the current inode's path. While this is not a functional issue,
> > it makes the caller recompute the current inode's path later. It could
> > also become a problem in the future in case get new callers for
> > is_current_inode_path() in more sensitive contexts.
> >
> > Example: the current inode path is "/foo/bar" and the path we compare
> > against is "/foo/bar_xyz".
> >
> > Fix this by returning true only if we have exact matches.
> >
> > Signed-off-by: Filipe Manana <[email protected]>
> > ---
> >
> > V2: Simplify and avoid false positives in case of something like
> >      "/foo/bar" vs "foo/bar/xyz".
> >
> >   fs/btrfs/send.c | 9 ++++++++-
> >   1 file changed, 8 insertions(+), 1 deletion(-)
> >
> > diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
> > index a3c8f2d889bb..d664e470fd77 100644
> > --- a/fs/btrfs/send.c
> > +++ b/fs/btrfs/send.c
> > @@ -627,7 +627,14 @@ static inline bool is_current_inode_path(const struct send_ctx *sctx,
> >   {
> >       const struct fs_path *cur = &sctx->cur_inode_path;
> >
> > -     return (strncmp(path->start, cur->start, fs_path_len(cur)) == 0);
> > +     /*
> > +      * Path may be a prefix of the current inode's path, so return false if
> > +      * the lengths are different.
> > +      */
> > +     if (fs_path_len(path) != fs_path_len(cur))
> > +             return false;
> > +
> > +     return (strcmp(path->start, cur->start) == 0);
>
> Can we just use strcmp without string length check? It looks like
> fs_path is always NUL terminated.
>
> E.g. fs_path_prepare_for_add() always reset *p->end to 0.
>
> So we can just use strcmp() to compare two NUL terminated paths.

Right, it's safe here. Updated in for-next, thanks.

>
> Thanks,
> Qu
>
> >   }
> >
> >   static struct btrfs_path *alloc_path_for_send(void)
>