Date: Sun, 22 Jun 2014 16:55:00 +0000 (UTC) From: Michael Tuexen <tuexen@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-stable@freebsd.org, svn-src-stable-10@freebsd.org Subject: svn commit: r267736 - stable/10/sys/netinet Message-ID: <201406221655.s5MGt014049782@svn.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: tuexen Date: Sun Jun 22 16:54:59 2014 New Revision: 267736 URL: http://svnweb.freebsd.org/changeset/base/267736 Log: MFC r265691: For some UDP packets (for example with 200 byte payload) and IP options, the IP header and the UDP header are not in the same mbuf. Add code to in_delayed_cksum() to deal with this case. MFC r265713: Use KASSERTs as suggested by glebius@ Modified: stable/10/sys/netinet/ip_output.c Directory Properties: stable/10/ (props changed) Modified: stable/10/sys/netinet/ip_output.c ============================================================================== --- stable/10/sys/netinet/ip_output.c Sun Jun 22 16:50:46 2014 (r267735) +++ stable/10/sys/netinet/ip_output.c Sun Jun 22 16:54:59 2014 (r267736) @@ -864,17 +864,13 @@ in_delayed_cksum(struct mbuf *m) csum = 0xffff; offset += m->m_pkthdr.csum_data; /* checksum offset */ - if (offset + sizeof(u_short) > m->m_len) { - printf("delayed m_pullup, m->len: %d off: %d p: %d\n", - m->m_len, offset, ip->ip_p); - /* - * XXX - * this shouldn't happen, but if it does, the - * correct behavior may be to insert the checksum - * in the appropriate next mbuf in the chain. - */ - return; + /* find the mbuf in the chain where the checksum starts*/ + while ((m != NULL) && (offset >= m->m_len)) { + offset -= m->m_len; + m = m->m_next; } + KASSERT(m != NULL, ("in_delayed_cksum: checksum outside mbuf chain.")); + KASSERT(offset + sizeof(u_short) <= m->m_len, ("in_delayed_cksum: checksum split between mbufs.")); *(u_short *)(m->m_data + offset) = csum; }
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201406221655.s5MGt014049782>