On Fri 23-11-18 10:21:35, Dan Carpenter wrote:
> We had intended to only print dentry->d_name.len characters but there is
> a width vs precision typo so if the name isn't NUL terminated it will
> read past the end of the buffer.
OK, it took me quite some time to grasp what you mean here. The code
works as expected because d_name.len and dname.name are in sync so there
no spacing going to happen. Anyway what you propose is formally more
correct I guess.
> Fixes: 408ddbc22be3 ("mm: print more information about mapping in __dump_page")
This sha is an unstable sha for mmotm patch.
> Signed-off-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
Acked-by: Michal Hocko <mhocko@xxxxxxxx>
Thanks!
> ---
> mm/debug.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/debug.c b/mm/debug.c
> index d18c5cea3320..faf856b652b6 100644
> --- a/mm/debug.c
> +++ b/mm/debug.c
> @@ -80,7 +80,7 @@ void __dump_page(struct page *page, const char *reason)
> if (mapping->host->i_dentry.first) {
> struct dentry *dentry;
> dentry = container_of(mapping->host->i_dentry.first, struct dentry, d_u.d_alias);
> - pr_warn("name:\"%*s\" ", dentry->d_name.len, dentry->d_name.name);
> + pr_warn("name:\"%.*s\" ", dentry->d_name.len, dentry->d_name.name);
> }
> }
> BUILD_BUG_ON(ARRAY_SIZE(pageflag_names) != __NR_PAGEFLAGS + 1);
> --
> 2.11.0
>
--
Michal Hocko
SUSE Labs
