| Message ID | 20210623194353.2021745-1-eric.dumazet@gmail.com |
|---|---|
| State | New |
| Headers | show |
| Series | [net] ipv6: fix out-of-bound access in ip6_parse_tlv() | expand |
On Wed, 2021-06-23 at 12:43 -0700, Eric Dumazet wrote: > From: Eric Dumazet <edumazet@google.com> > > First problem is that optlen is fetched without checking > there is more than one byte to parse. > > Fix this by taking care of IPV6_TLV_PAD1 before > fetching optlen (under appropriate sanity checks against len) > > Second problem is that IPV6_TLV_PADN checks of zero > padding are performed before the check of remaining length. > > Fixes: c1412fce7ecc ("net/ipv6/exthdrs.c: Strict PadN option checking") Perhaps even: Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") for the first issue? > Signed-off-by: Eric Dumazet <edumazet@google.com> > Cc: Paolo Abeni <pabeni@redhat.com> > Cc: Tom Herbert <tom@herbertland.com> > --- > > Only compiled, I would appreciate a solid review of this patch before merging it, thanks ! > > net/ipv6/exthdrs.c | 26 +++++++++++++------------- > 1 file changed, 13 insertions(+), 13 deletions(-) > > diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c > index 6f7da8f3e2e5849f917853984c69bf02a0f1e27c..007959d4d3748f1e21f83946024a9967d08b25b6 100644 > --- a/net/ipv6/exthdrs.c > +++ b/net/ipv6/exthdrs.c > @@ -135,18 +135,24 @@ static bool ip6_parse_tlv(const struct tlvtype_proc *procs, > len -= 2; > > while (len > 0) { > - int optlen = nh[off + 1] + 2; > - int i; > + int optlen, i; > > - switch (nh[off]) { > - case IPV6_TLV_PAD1: > + if (nh[off] == IPV6_TLV_PAD1) { > optlen = 1; It looks like the above assignment is not needed anymore. Other than that LGTM, Thanks! Paolo
On Thu, Jun 24, 2021 at 11:49 AM Paolo Abeni <pabeni@redhat.com> wrote: > > On Wed, 2021-06-23 at 12:43 -0700, Eric Dumazet wrote: > > From: Eric Dumazet <edumazet@google.com> > > > > First problem is that optlen is fetched without checking > > there is more than one byte to parse. > > > > Fix this by taking care of IPV6_TLV_PAD1 before > > fetching optlen (under appropriate sanity checks against len) > > > > Second problem is that IPV6_TLV_PADN checks of zero > > padding are performed before the check of remaining length. > > > > Fixes: c1412fce7ecc ("net/ipv6/exthdrs.c: Strict PadN option checking") > > Perhaps even: > > Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") > > for the first issue? > > + if (nh[off] == IPV6_TLV_PAD1) { > > optlen = 1; > > It looks like the above assignment is not needed anymore. > > Other than that LGTM, > Thanks for the review, I am sending the v2.
diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c index 6f7da8f3e2e5849f917853984c69bf02a0f1e27c..007959d4d3748f1e21f83946024a9967d08b25b6 100644 --- a/net/ipv6/exthdrs.c +++ b/net/ipv6/exthdrs.c @@ -135,18 +135,24 @@ static bool ip6_parse_tlv(const struct tlvtype_proc *procs, len -= 2; while (len > 0) { - int optlen = nh[off + 1] + 2; - int i; + int optlen, i; - switch (nh[off]) { - case IPV6_TLV_PAD1: + if (nh[off] == IPV6_TLV_PAD1) { optlen = 1; padlen++; if (padlen > 7) goto bad; - break; + off++; + len--; + continue; + } + if (len < 2) + goto bad; + optlen = nh[off + 1] + 2; + if (optlen > len) + goto bad; - case IPV6_TLV_PADN: + if (nh[off] == IPV6_TLV_PADN) { /* RFC 2460 states that the purpose of PadN is * to align the containing header to multiples * of 8. 7 is therefore the highest valid value. @@ -163,12 +169,7 @@ static bool ip6_parse_tlv(const struct tlvtype_proc *procs, if (nh[off + i] != 0) goto bad; } - break; - - default: /* Other TLV code so scan list */ - if (optlen > len) - goto bad; - + } else { tlv_count++; if (tlv_count > max_count) goto bad; @@ -188,7 +189,6 @@ static bool ip6_parse_tlv(const struct tlvtype_proc *procs, return false; padlen = 0; - break; } off += optlen; len -= optlen;