On Mon, Nov 8, 2021 at 10:38 PM Nicolas Iooss <nicolas.iooss@xxxxxxx> wrote:
>
> On Tue, Oct 19, 2021 at 5:13 PM Christian Göttsche
> <cgzones@xxxxxxxxxxxxxx> wrote:
> >
> > For the first iteration `mod->perm_map[sclassi]` is NULL, thus do not
> > use it as source of a memcpy(3), even with a size of 0. memcpy(3) might
> > be annotated with the function attribute nonnull and UBSan then
> > complains:
> >
> > link.c:193:3: runtime error: null pointer passed as argument 2, which is declared to never be null
> >
> > Signed-off-by: Christian Göttsche <cgzones@xxxxxxxxxxxxxx>
>
> For these 3 patches:
>
> Acked-by: Nicolas Iooss <nicolas.iooss@xxxxxxx>
>
> Sorry for the delay, I have been busy with other topics.
>
> Thanks!
> Nicolas
I applied these 3 patches.
Thanks!
Nicolas
> > ---
> > v2:
> > drop realloc rewrite, just check for 0 size
> > ---
> > libsepol/src/link.c | 5 +++--
> > 1 file changed, 3 insertions(+), 2 deletions(-)
> >
> > diff --git a/libsepol/src/link.c b/libsepol/src/link.c
> > index 7512a4d9..b14240d5 100644
> > --- a/libsepol/src/link.c
> > +++ b/libsepol/src/link.c
> > @@ -190,8 +190,9 @@ static int permission_copy_callback(hashtab_key_t key, hashtab_datum_t datum,
> > ERR(state->handle, "Out of memory!");
> > return -1;
> > }
> > - memcpy(newmap, mod->perm_map[sclassi],
> > - mod->perm_map_len[sclassi] * sizeof(*newmap));
> > + if (mod->perm_map_len[sclassi] > 0) {
> > + memcpy(newmap, mod->perm_map[sclassi], mod->perm_map_len[sclassi] * sizeof(*newmap));
> > + }
> > free(mod->perm_map[sclassi]);
> > mod->perm_map[sclassi] = newmap;
> > mod->perm_map_len[sclassi] = perm->s.value;
> > --
> > 2.33.0
> >
