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

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.

Thanks,
Qu

>   }
>   
>   static struct btrfs_path *alloc_path_for_send(void)