Please sign in to comment.
tcp: tcp_fragment() should apply sane memory limits
Jonathan Looney reported that a malicious peer can force a sender to fragment its retransmit queue into tiny skbs, inflating memory usage and/or overflow 32bit counters. TCP allows an application to queue up to sk_sndbuf bytes, so we need to give some allowance for non malicious splitting of retransmit queue. A new SNMP counter is added to monitor how many times TCP did not allow to split an skb if the allowance was exceeded. Note that this counter might increase in the case applications use SO_SNDBUF socket option to lower sk_sndbuf. CVE-2019-11478 : tcp_fragment, prevent fragmenting a packet when the socket is already using more than half the allowed space Signed-off-by: Eric Dumazet <firstname.lastname@example.org> Reported-by: Jonathan Looney <email@example.com> Acked-by: Neal Cardwell <firstname.lastname@example.org> Acked-by: Yuchung Cheng <email@example.com> Reviewed-by: Tyler Hicks <firstname.lastname@example.org> Cc: Bruce Curtis <email@example.com> Cc: Jonathan Lemon <firstname.lastname@example.org> Signed-off-by: David S. Miller <email@example.com>
- Loading branch information...
Showing with 7 additions and 0 deletions.