On Fri, Jan 12, 2024 at 01:07:36PM +0100, Pablo Neira Ayuso wrote:
> On Fri, Jan 12, 2024 at 12:41:47PM +0100, Phil Sutter wrote:
> > On Thu, Jan 11, 2024 at 11:25:27PM +0100, Pablo Neira Ayuso wrote:
> > > Allow to set a maximum limit of sizeof(s->desc.field_len) which is 16
> > > bytes, otherwise, bail out. Ensure s->desc.field_count does not go over
> > > the array boundary.
> > >
> > > Fixes: 7cd41b5387ac ("set: Add support for NFTA_SET_DESC_CONCAT attributes")
> > > Signed-off-by: Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx>
> > > ---
> > > src/set.c | 8 +++++++-
> > > 1 file changed, 7 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/src/set.c b/src/set.c
> > > index 719e59616e97..b51ff9e0ba64 100644
> > > --- a/src/set.c
> > > +++ b/src/set.c
> > > @@ -194,8 +194,14 @@ int nftnl_set_set_data(struct nftnl_set *s, uint16_t attr, const void *data,
> > > memcpy(&s->desc.size, data, sizeof(s->desc.size));
> > > break;
> > > case NFTNL_SET_DESC_CONCAT:
> > > + if (data_len > sizeof(s->desc.field_len))
> > > + return -1;
> > > +
> > > memcpy(&s->desc.field_len, data, data_len);
> > > - while (s->desc.field_len[++s->desc.field_count]);
> > > + while (s->desc.field_len[++s->desc.field_count]) {
> > > + if (s->desc.field_count >= NFT_REG32_COUNT)
> > > + break;
> > > + }
> >
> > Isn't the second check redundant if you adjust the first one like so:
> >
> > | if (data_len >= sizeof(s->desc.field_len))
> >
> > Or more explicit:
> >
> > | if (data_len > sizeof(s->desc.field_len) -
> > | sizeof(s->desc.field_len[0]))
>
> I see, you suggest to ensure last item in the array is always zero.
This is what's required by the while loop in the original form, I don't
see a real reason for it, though.
Cheers, Phil
