[OE-core] [PATCH] dhcp-client: Ignore partial checksums

Rob Woolley rob.woolley at windriver.com
Fri Feb 6 02:56:39 UTC 2015


Hi Hongxu,

Have you had a chance to review this patch?  Do you have any questions about it?

Regards,
Rob


On Fri, Jan 30, 2015 at 04:55:09PM -0500, Rob Woolley wrote:
> dhclient will fail to get an IP address if run inside a guest when traffic is
> flowing over a virtual network interface.  The user will see the error
> message:
> 
>   5 bad udp checksums in 5 packets
>   No DHCPOFFERS received.
>   Unable to obtain a lease on first try.  Exiting.
>   Failed to bring up eth0.
> 
> This is because Linux only uses partial checksums for packets that go over
> virtual network interfaces and dhclient does not like this.
> 
>   See linux kernel commit 78ea85f17b15390e30d8b47488ec7b6cf0790663
>   ("net: skbuff: improve comment on checksumming")
> 
> An application can detect this behaviour by checking for the
> TP_STATUS_CSUMNOTREADY flag in the tp_status field.
> 
>   See linux kernel commit 8dc4194474159660d7f37c495e3fc3f10d0db8cc
>   ("Add optional checksum computation for recvmsg")
> 
> An extra parameter is added to decode_udp_ip_header() in dhclient to indicate
> whether or not dhclient should ignore partial checksums.  This is used
> when the TP_STATUS_CSUMNOTREADY bit is set by the guest kernel.
> 
> This fix has been included in Fedora and Ubuntu, however it has not yet been
> accepted by ISC upstream.  Likely because it is specific to behaviour in Linux
> and other UNIX variants do not seem to be affected.
> 
> The patch was imported from the dhcp source RPM in Fedora 21
>   (http://pkgs.fedoraproject.org/cgit/dhcp.git/tree/dhcp-xen-checksum.patch?h=f21)
> 
> Originally contributed to fedora-cvs-commit by David Cantrell on Jan 30 2007
>   (https://www.redhat.com/archives/fedora-cvs-commits/2007-January/msg01442.html)
> 
> Submitted to dhcp-bugs at isc.org - [ISC-Bugs #22806] - by Michael S. Tsirkin
>   (http://comments.gmane.org/gmane.comp.emulators.kvm.devel/65236)
>   (https://lists.isc.org/pipermail/dhcp-hackers/2010-April/001835.html)
> 
> Upstream-Status: Submitted [dhcp-bugs at isc.org]
> Signed-off-by: Rob Woolley <rob.woolley at windriver.com>
> ---
>  .../dhcp/dhcp/dhcp-xen-checksum.patch              | 282 +++++++++++++++++++++
>  meta/recipes-connectivity/dhcp/dhcp_4.3.1.bb       |   1 +
>  2 files changed, 283 insertions(+)
>  create mode 100644 meta/recipes-connectivity/dhcp/dhcp/dhcp-xen-checksum.patch
> 
> Index: b/meta/recipes-connectivity/dhcp/dhcp/dhcp-xen-checksum.patch
> ===================================================================
> --- /dev/null
> +++ b/meta/recipes-connectivity/dhcp/dhcp/dhcp-xen-checksum.patch
> @@ -0,0 +1,307 @@
> +dhcp-client: Ignore partial checksums
> +
> +dhclient will fail to get an IP address if run inside a guest when traffic is
> +flowing over a virtual network interface.  The user will see the error
> +message:
> +
> +  5 bad udp checksums in 5 packets
> +  No DHCPOFFERS received.
> +  Unable to obtain a lease on first try.  Exiting.
> +  Failed to bring up eth0.
> +
> +This is because Linux only uses partial checksums for packets that go over
> +virtual network interfaces and dhclient does not like this.
> +
> +  See linux kernel commit 78ea85f17b15390e30d8b47488ec7b6cf0790663
> +  ("net: skbuff: improve comment on checksumming")
> +
> +An application can detect this behaviour by checking for the
> +TP_STATUS_CSUMNOTREADY flag in the tp_status field.
> +
> +  See linux kernel commit 8dc4194474159660d7f37c495e3fc3f10d0db8cc
> +  ("Add optional checksum computation for recvmsg")
> +
> +An extra parameter is added to decode_udp_ip_header() in dhclient to indicate
> +whether or not dhclient should ignore partial checksums.  This is used
> +when the TP_STATUS_CSUMNOTREADY bit is set by the guest kernel.
> +
> +This fix has been included in Fedora and Ubuntu, however it has not yet been
> +accepted by ISC upstream.  Likely because it is specific to behaviour in Linux
> +and other UNIX variants do not seem to be affected.
> +
> +The patch was imported from the dhcp source RPM in Fedora 21
> +  (http://pkgs.fedoraproject.org/cgit/dhcp.git/tree/dhcp-xen-checksum.patch?h=f21)
> +
> +Originally contributed to fedora-cvs-commit by David Cantrell on Jan 30 2007
> +  (https://www.redhat.com/archives/fedora-cvs-commits/2007-January/msg01442.html)
> +
> +Submitted to dhcp-bugs at isc.org - [ISC-Bugs #22806] - by Michael S. Tsirkin
> +  (http://comments.gmane.org/gmane.comp.emulators.kvm.devel/65236)
> +  (https://lists.isc.org/pipermail/dhcp-hackers/2010-April/001835.html)
> +
> +Upstream-Status: Submitted [dhcp-bugs at isc.org]
> +Signed-off-by: Rob Woolley <rob.woolley at windriver.com>
> +--
> + common/bpf.c     |    2 -
> + common/dlpi.c    |    2 -
> + common/lpf.c     |   83 +++++++++++++++++++++++++++++++++++++++++--------------
> + common/nit.c     |    2 -
> + common/packet.c  |    4 +-
> + common/upf.c     |    2 -
> + includes/dhcpd.h |    2 -
> + 7 files changed, 70 insertions(+), 27 deletions(-)
> +
> +diff --git a/common/bpf.c b/common/bpf.c
> +--- a/common/bpf.c
> ++++ b/common/bpf.c
> +@@ -481,7 +481,7 @@ ssize_t receive_packet (interface, buf,
> + 		/* Decode the IP and UDP headers... */
> + 		offset = decode_udp_ip_header(interface, interface->rbuf,
> + 					       interface->rbuf_offset,
> +-  					       from, hdr.bh_caplen, &paylen);
> ++  					       from, hdr.bh_caplen, &paylen, 0);
> + 
> + 		/* If the IP or UDP checksum was bad, skip the packet... */
> + 		if (offset < 0) {
> +diff --git a/common/dlpi.c b/common/dlpi.c
> +--- a/common/dlpi.c
> ++++ b/common/dlpi.c
> +@@ -691,7 +691,7 @@ ssize_t receive_packet (interface, buf,
> + 	length -= offset;
> + #endif
> + 	offset = decode_udp_ip_header (interface, dbuf, bufix,
> +-				       from, length, &paylen);
> ++				       from, length, &paylen, 0);
> + 
> + 	/*
> + 	 * If the IP or UDP checksum was bad, skip the packet...
> +diff --git a/common/lpf.c b/common/lpf.c
> +--- a/common/lpf.c
> ++++ b/common/lpf.c
> +@@ -29,14 +29,15 @@
> + 
> + #include "dhcpd.h"
> + #if defined (USE_LPF_SEND) || defined (USE_LPF_RECEIVE)
> ++#include <sys/socket.h>
> + #include <sys/uio.h>
> + #include <errno.h>
> + 
> + #include <asm/types.h>
> + #include <linux/filter.h>
> + #include <linux/if_ether.h>
> ++#include <linux/if_packet.h>
> + #include <netinet/in_systm.h>
> +-#include <net/if_packet.h>
> + #include "includes/netinet/ip.h"
> + #include "includes/netinet/udp.h"
> + #include "includes/netinet/if_ether.h"
> +@@ -51,6 +52,19 @@
> + /* Reinitializes the specified interface after an address change.   This
> +    is not required for packet-filter APIs. */
> + 
> ++#ifndef PACKET_AUXDATA
> ++#define PACKET_AUXDATA 8
> ++
> ++struct tpacket_auxdata
> ++{
> ++	__u32		tp_status;
> ++	__u32		tp_len;
> ++	__u32		tp_snaplen;
> ++	__u16		tp_mac;
> ++	__u16		tp_net;
> ++};
> ++#endif
> ++
> + #ifdef USE_LPF_SEND
> + void if_reinitialize_send (info)
> + 	struct interface_info *info;
> +@@ -73,10 +87,14 @@ int if_register_lpf (info)
> + 	struct interface_info *info;
> + {
> + 	int sock;
> +-	struct sockaddr sa;
> ++	union {
> ++		struct sockaddr_ll ll;
> ++		struct sockaddr common;
> ++	} sa;
> ++	struct ifreq ifr;
> + 
> + 	/* Make an LPF socket. */
> +-	if ((sock = socket(PF_PACKET, SOCK_PACKET,
> ++	if ((sock = socket(PF_PACKET, SOCK_RAW,
> + 			   htons((short)ETH_P_ALL))) < 0) {
> + 		if (errno == ENOPROTOOPT || errno == EPROTONOSUPPORT ||
> + 		    errno == ESOCKTNOSUPPORT || errno == EPFNOSUPPORT ||
> +@@ -91,11 +109,17 @@ int if_register_lpf (info)
> + 		log_fatal ("Open a socket for LPF: %m");
> + 	}
> + 
> ++	memset (&ifr, 0, sizeof ifr);
> ++	strncpy (ifr.ifr_name, (const char *)info -> ifp, sizeof ifr.ifr_name);
> ++	ifr.ifr_name[IFNAMSIZ-1] = '\0';
> ++	if (ioctl (sock, SIOCGIFINDEX, &ifr))
> ++		log_fatal ("Failed to get interface index: %m");
> ++
> + 	/* Bind to the interface name */
> + 	memset (&sa, 0, sizeof sa);
> +-	sa.sa_family = AF_PACKET;
> +-	strncpy (sa.sa_data, (const char *)info -> ifp, sizeof sa.sa_data);
> +-	if (bind (sock, &sa, sizeof sa)) {
> ++	sa.ll.sll_family = AF_PACKET;
> ++	sa.ll.sll_ifindex = ifr.ifr_ifindex;
> ++	if (bind (sock, &sa.common, sizeof sa)) {
> + 		if (errno == ENOPROTOOPT || errno == EPROTONOSUPPORT ||
> + 		    errno == ESOCKTNOSUPPORT || errno == EPFNOSUPPORT ||
> + 		    errno == EAFNOSUPPORT || errno == EINVAL) {
> +@@ -177,9 +201,18 @@ static void lpf_gen_filter_setup (struct
> + void if_register_receive (info)
> + 	struct interface_info *info;
> + {
> ++	int val;
> ++
> + 	/* Open a LPF device and hang it on this interface... */
> + 	info -> rfdesc = if_register_lpf (info);
> + 
> ++	val = 1;
> ++	if (setsockopt (info -> rfdesc, SOL_PACKET, PACKET_AUXDATA, &val,
> ++			sizeof val) < 0) {
> ++		if (errno != ENOPROTOOPT)
> ++			log_fatal ("Failed to set auxiliary packet data: %m");
> ++	}
> ++
> + #if defined (HAVE_TR_SUPPORT)
> + 	if (info -> hw_address.hbuf [0] == HTYPE_IEEE802)
> + 		lpf_tr_filter_setup (info);
> +@@ -301,7 +334,6 @@ ssize_t send_packet (interface, packet,
> + 	double hh [16];
> + 	double ih [1536 / sizeof (double)];
> + 	unsigned char *buf = (unsigned char *)ih;
> +-	struct sockaddr_pkt sa;
> + 	int result;
> + 	int fudge;
> + 
> +@@ -322,17 +354,7 @@ ssize_t send_packet (interface, packet,
> + 				(unsigned char *)raw, len);
> + 	memcpy (buf + ibufp, raw, len);
> + 
> +-	/* For some reason, SOCK_PACKET sockets can't be connected,
> +-	   so we have to do a sentdo every time. */
> +-	memset (&sa, 0, sizeof sa);
> +-	sa.spkt_family = AF_PACKET;
> +-	strncpy ((char *)sa.spkt_device,
> +-		 (const char *)interface -> ifp, sizeof sa.spkt_device);
> +-	sa.spkt_protocol = htons(ETH_P_IP);
> +-
> +-	result = sendto (interface -> wfdesc,
> +-			 buf + fudge, ibufp + len - fudge, 0, 
> +-			 (const struct sockaddr *)&sa, sizeof sa);
> ++	result = write (interface -> wfdesc, buf + fudge, ibufp + len - fudge);
> + 	if (result < 0)
> + 		log_error ("send_packet: %m");
> + 	return result;
> +@@ -349,14 +371,35 @@ ssize_t receive_packet (interface, buf,
> + {
> + 	int length = 0;
> + 	int offset = 0;
> ++	int nocsum = 0;
> + 	unsigned char ibuf [1536];
> + 	unsigned bufix = 0;
> + 	unsigned paylen;
> ++	unsigned char cmsgbuf[CMSG_LEN(sizeof(struct tpacket_auxdata))];
> ++	struct iovec iov = {
> ++		.iov_base = ibuf,
> ++		.iov_len = sizeof ibuf,
> ++	};
> ++	struct msghdr msg = {
> ++		.msg_iov = &iov,
> ++		.msg_iovlen = 1,
> ++		.msg_control = cmsgbuf,
> ++		.msg_controllen = sizeof(cmsgbuf),
> ++	};
> ++	struct cmsghdr *cmsg;
> + 
> +-	length = read (interface -> rfdesc, ibuf, sizeof ibuf);
> ++	length = recvmsg (interface -> rfdesc, &msg, 0);
> + 	if (length <= 0)
> + 		return length;
> + 
> ++	for (cmsg = CMSG_FIRSTHDR(&msg); cmsg; cmsg = CMSG_NXTHDR(&msg, cmsg)) {
> ++		if (cmsg->cmsg_level == SOL_PACKET &&
> ++		    cmsg->cmsg_type == PACKET_AUXDATA) {
> ++			struct tpacket_auxdata *aux = (void *)CMSG_DATA(cmsg);
> ++			nocsum = aux->tp_status & TP_STATUS_CSUMNOTREADY;
> ++		}
> ++	}
> ++
> + 	bufix = 0;
> + 	/* Decode the physical header... */
> + 	offset = decode_hw_header (interface, ibuf, bufix, hfrom);
> +@@ -373,7 +416,7 @@ ssize_t receive_packet (interface, buf,
> + 
> + 	/* Decode the IP and UDP headers... */
> + 	offset = decode_udp_ip_header (interface, ibuf, bufix, from,
> +-				       (unsigned)length, &paylen);
> ++				       (unsigned)length, &paylen, nocsum);
> + 
> + 	/* If the IP or UDP checksum was bad, skip the packet... */
> + 	if (offset < 0)
> +diff --git a/common/nit.c b/common/nit.c
> +--- a/common/nit.c
> ++++ b/common/nit.c
> +@@ -363,7 +363,7 @@ ssize_t receive_packet (interface, buf,
> + 
> + 	/* Decode the IP and UDP headers... */
> + 	offset = decode_udp_ip_header (interface, ibuf, bufix,
> +-				       from, length, &paylen);
> ++				       from, length, &paylen, 0);
> + 
> + 	/* If the IP or UDP checksum was bad, skip the packet... */
> + 	if (offset < 0)
> +diff --git a/common/packet.c b/common/packet.c
> +--- a/common/packet.c
> ++++ b/common/packet.c
> +@@ -226,7 +226,7 @@ ssize_t
> + decode_udp_ip_header(struct interface_info *interface,
> + 		     unsigned char *buf, unsigned bufix,
> + 		     struct sockaddr_in *from, unsigned buflen,
> +-		     unsigned *rbuflen)
> ++		     unsigned *rbuflen, int nocsum)
> + {
> +   unsigned char *data;
> +   struct ip ip;
> +@@ -337,7 +337,7 @@ decode_udp_ip_header(struct interface_in
> + 					   8, IPPROTO_UDP + ulen))));
> + 
> +   udp_packets_seen++;
> +-  if (usum && usum != sum) {
> ++  if (!nocsum && usum && usum != sum) {
> + 	  udp_packets_bad_checksum++;
> + 	  if (udp_packets_seen > 4 &&
> + 	      (udp_packets_seen / udp_packets_bad_checksum) < 2) {
> +diff --git a/common/upf.c b/common/upf.c
> +--- a/common/upf.c
> ++++ b/common/upf.c
> +@@ -314,7 +314,7 @@ ssize_t receive_packet (interface, buf,
> + 
> + 	/* Decode the IP and UDP headers... */
> + 	offset = decode_udp_ip_header (interface, ibuf, bufix,
> +-				       from, length, &paylen);
> ++				       from, length, &paylen, 0);
> + 
> + 	/* If the IP or UDP checksum was bad, skip the packet... */
> + 	if (offset < 0)
> +diff --git a/includes/dhcpd.h b/includes/dhcpd.h
> +--- a/includes/dhcpd.h
> ++++ b/includes/dhcpd.h
> +@@ -2857,7 +2857,7 @@ ssize_t decode_hw_header (struct interfa
> + 			  unsigned, struct hardware *);
> + ssize_t decode_udp_ip_header (struct interface_info *, unsigned char *,
> + 			      unsigned, struct sockaddr_in *,
> +-			      unsigned, unsigned *);
> ++			      unsigned, unsigned *, int);
> + 
> + /* ethernet.c */
> + void assemble_ethernet_header (struct interface_info *, unsigned char *,
> +-- 
> +1.8.1.2
> +
> Index: b/meta/recipes-connectivity/dhcp/dhcp_4.3.1.bb
> ===================================================================
> --- a/meta/recipes-connectivity/dhcp/dhcp_4.3.1.bb
> +++ b/meta/recipes-connectivity/dhcp/dhcp_4.3.1.bb
> @@ -6,6 +6,7 @@ SRC_URI += "file://dhcp-3.0.3-dhclient-d
>              file://fixsepbuild.patch \
>              file://dhclient-script-drop-resolv.conf.dhclient.patch \
>              file://replace-ifconfig-route.patch \
> +            file://dhcp-xen-checksum.patch \
>             "
>  
>  SRC_URI[md5sum] = "b3a42ece3c7f2cd2e74a3e12ca881d20"
> -- 
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core at lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
> 



More information about the Openembedded-core mailing list