This was reported by Matej a while ago, but we forgot to fix it. Even if the hypervisor is necessarily trusted by passt, as it can in any case terminate the guest or disrupt guest connectivity, it's a good idea to be robust against possible issues. Instead of resetting the connection to the hypervisor, just skip the offending frame, as we had a few cases where QEMU would get the length descriptor wrong, in the past. Reported-by: Matej Hrica <mhrica(a)redhat.com> Suggested-by: Matej Hrica <mhrica(a)redhat.com> Signed-off-by: Stefano Brivio <sbrivio(a)redhat.com> --- tap.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/tap.c b/tap.c index 7f8c26d..bb993e0 100644 --- a/tap.c +++ b/tap.c @@ -1026,6 +1026,9 @@ redo: p += sizeof(uint32_t); n -= sizeof(uint32_t); + if (l2len > (ssize_t)TAP_BUF_BYTES - n) + goto next; + /* At most one packet might not fit in a single read, and this * needs to be blocking. */ -- 2.43.0