MFp4 bz_ipv6_fast:

  Add support for delayed checksum calculations in the IPv6
  output path.  We currently cannot offload to the card if we
  add extension headers (which incl. fragmentation).

  Fix two SCTP offload support copy&paste bugs: calculate
  checksums if fragmenting and no need to flag IPv4 header
  checksums in the IPv6 forwarding path.

  Sponsored by:	The FreeBSD Foundation
  Sponsored by:	iXsystems

Reviewed by:	gnn (as part of the whole)
MFC After:	3 days

2 files changed, 63 insertions, 12 deletions

diff --git a/sys/netinet6/ip6_forward.c b/sys/netinet6/ip6_forward.c
index bae418a..dfecdc1 100644
--- a/sys/netinet6/ip6_forward.c
+++ b/sys/netinet6/ip6_forward.c
@@ -586,8 +586,6 @@ skip_routing:
 				    CSUM_DATA_VALID | CSUM_PSEUDO_HDR;
 				m->m_pkthdr.csum_data = 0xffff;
 			}
-			m->m_pkthdr.csum_flags |=
-			    CSUM_IP_CHECKED | CSUM_IP_VALID;
 #ifdef SCTP
 			if (m->m_pkthdr.csum_flags & CSUM_SCTP)
 				m->m_pkthdr.csum_flags |= CSUM_SCTP_VALID;
diff --git a/sys/netinet6/ip6_output.c b/sys/netinet6/ip6_output.c
index d15d7c2..f2dd552 100644
--- a/sys/netinet6/ip6_output.c
+++ b/sys/netinet6/ip6_output.c
@@ -83,6 +83,8 @@ __FBSDID("$FreeBSD$");
 #include <sys/syslog.h>
 #include <sys/ucred.h>
 
+#include <machine/in_cksum.h>
+
 #include <net/if.h>
 #include <net/netisr.h>
 #include <net/route.h>
@@ -182,6 +184,29 @@ static int copypktopts(struct ip6_pktopts *, struct ip6_pktopts *, int);
 	}\
     } while (/*CONSTCOND*/ 0)
 
+static void
+in6_delayed_cksum(struct mbuf *m, uint32_t plen, u_short offset)
+{
+	u_short csum;
+
+	csum = in_cksum_skip(m, ntohl(plen), offset);
+	if (m->m_pkthdr.csum_flags & CSUM_UDP && csum == 0)
+		csum = 0xffff;
+	offset += m->m_pkthdr.csum_data;	/* checksum offset */
+
+	if (offset + sizeof(u_short) > m->m_len) {
+		printf("%s: delayed m_pullup, m->len: %d  off: %d\n",
+		    __func__, m->m_len, offset);
+		/*
+		 * XXX this should not happen, but if it does, the correct
+		 * behavior may be to insert the checksum in the appropriate
+		 * next mbuf in the chain.
+		 */
+		return;
+	}
+	*(u_short *)(m->m_data + offset) = csum;
+}
+
 /*
  * IP6 output. The packet in mbuf chain m contains a skeletal IP6
  * header (with pri, len, nxt, hlim, src, dst).
@@ -221,9 +246,7 @@ ip6_output(struct mbuf *m0, struct ip6_pktopts *opt,
 	int flevalid = 0;
 	int hdrsplit = 0;
 	int needipsec = 0;
-#ifdef SCTP
-	int sw_csum;
-#endif
+	int sw_csum, tso;
 #ifdef IPSEC
 	struct ipsec_output_state state;
 	struct ip6_rthdr *rh = NULL;
@@ -867,8 +890,6 @@ again:
 				    CSUM_DATA_VALID | CSUM_PSEUDO_HDR;
 				m->m_pkthdr.csum_data = 0xffff;
 			}
-			m->m_pkthdr.csum_flags |=
-			    CSUM_IP_CHECKED | CSUM_IP_VALID;
 #ifdef SCTP
 			if (m->m_pkthdr.csum_flags & CSUM_SCTP)
 				m->m_pkthdr.csum_flags |= CSUM_SCTP_VALID;
@@ -891,7 +912,7 @@ again:
 		}
 #ifdef SCTP
 		if (m->m_pkthdr.csum_flags & CSUM_SCTP)
-		m->m_pkthdr.csum_flags |= CSUM_SCTP_VALID;
+			m->m_pkthdr.csum_flags |= CSUM_SCTP_VALID;
 #endif   
 		error = netisr_queue(NETISR_IPV6, m);
 		goto done;
@@ -927,16 +948,32 @@ passout:
 	 * 4: if dontfrag == 1 && alwaysfrag == 1
 	 *	error, as we cannot handle this conflicting request
 	 */
+	sw_csum = m->m_pkthdr.csum_flags;
+	if (!hdrsplit) {
+		tso = ((sw_csum & ifp->if_hwassist & CSUM_TSO) != 0) ? 1 : 0;
+		sw_csum &= ~ifp->if_hwassist;
+	} else
+		tso = 0;
+	/*
+	 * If we added extension headers, we will not do TSO and calculate the
+	 * checksums ourselves for now.
+	 * XXX-BZ  Need a framework to know when the NIC can handle it, even
+	 * with ext. hdrs.
+	 */
+	if (sw_csum & CSUM_DELAY_DATA) {
+		sw_csum &= ~CSUM_DELAY_DATA;
+		in6_delayed_cksum(m, ip6->ip6_plen, sizeof(struct ip6_hdr));
+	}
 #ifdef SCTP
-	sw_csum = m->m_pkthdr.csum_flags & ~ifp->if_hwassist;
 	if (sw_csum & CSUM_SCTP) {
-		sctp_delayed_cksum(m, sizeof(struct ip6_hdr));
 		sw_csum &= ~CSUM_SCTP;
+		sctp_delayed_cksum(m, sizeof(struct ip6_hdr));
 	}
 #endif
+	m->m_pkthdr.csum_flags &= ifp->if_hwassist;
 	tlen = m->m_pkthdr.len;
 
-	if (opt && (opt->ip6po_flags & IP6PO_DONTFRAG))
+	if ((opt && (opt->ip6po_flags & IP6PO_DONTFRAG)) || tso)
 		dontfrag = 1;
 	else
 		dontfrag = 0;
@@ -945,7 +982,7 @@ passout:
 		error = EMSGSIZE;
 		goto bad;
 	}
-	if (dontfrag && tlen > IN6_LINKMTU(ifp)) {	/* case 2-b */
+	if (dontfrag && tlen > IN6_LINKMTU(ifp) && !tso) {	/* case 2-b */
 		/*
 		 * Even if the DONTFRAG option is specified, we cannot send the
 		 * packet when the data length is larger than the MTU of the
@@ -1033,6 +1070,22 @@ passout:
 			goto bad;
 		}
 
+
+		/*
+		 * If the interface will not calculate checksums on
+		 * fragmented packets, then do it here.
+		 * XXX-BZ handle the hw offloading case.  Need flags.
+		 */
+		if (m->m_pkthdr.csum_flags & CSUM_DELAY_DATA) {
+			in6_delayed_cksum(m, ip6->ip6_plen, sizeof(*ip6));
+			m->m_pkthdr.csum_flags &= ~CSUM_DELAY_DATA;
+		}
+#ifdef SCTP
+		if (m->m_pkthdr.csum_flags & CSUM_SCTP) {
+			sctp_delayed_cksum(m, hlen);
+			m->m_pkthdr.csum_flags &= ~CSUM_SCTP;
+		}
+#endif
 		mnext = &m->m_nextpkt;
 
 		/*