Am 23.05.2013 01:17, schrieb Matthew O'Connor: > This is the backported patch I submitted previously. Hopefully this > time around it won't be too messed up, I'm using Thunderbird instead of > the web interface. I have applied it successfully and without warnings > against 3.4.46. It builds, but is otherwise untested beyond what I did > when I originally submitted back in Feb. This patch applies only to the > 3.4 series kernel, although with minor changes it will work for 3.0, > 3.2, and 3.7. If you're interested, I can submit the other patches > shortly. If this submission still does not conform to standards, please > let me know where I went wrong. For what it's worth I dropped the patch > contents directly into the email, but I can attach it instead if that > would work better. > Hm, i tried to apply it to 3.4.46 but it does not work: node02:/ocfs2/usr/src/linux-3.4.46# patch -p1 <../bridge-patch-3.4.46 patching file drivers/net/bonding/bond_alb.c Hunk #1 FAILED at 704. 1 out of 1 hunk FAILED -- saving rejects to file drivers/net/bonding/bond_alb.c.rej patching file drivers/net/bonding/bonding.h patching file include/linux/etherdevice.h Hunk #1 FAILED at 277. 1 out of 1 hunk FAILED -- saving rejects to file include/linux/etherdevice.h.rej node02:/ocfs2/usr/src/linux-3.4.46# cat drivers/net/bonding/bond_alb.c.rej --- drivers/net/bonding/bond_alb.c 2013-01-27 23:51:45.000000000 -0500 +++ drivers/net/bonding/bond_alb.c 2013-01-30 15:37:25.121708311 -0500 @@ -704,6 +704,12 @@ struct arp_pkt *arp = arp_pkt(skb); struct slave *tx_slave = NULL; + /* Don't modify or load balance ARPs that do not originate locally + * (e.g.,arrive via a bridge). + */ + if (!bond_slave_has_mac(bond, arp->mac_src)) + return NULL; + if (arp->op_code == htons(ARPOP_REPLY)) { /* the arp must be sent on the selected * rx channel node02:/ocfs2/usr/src/linux-3.4.46# cat include/linux/etherdevice.h.rej --- include/linux/etherdevice.h 2013-01-27 23:51:45.000000000 -0500 +++ include/linux/etherdevice.h 2013-01-30 15:37:25.121708311 -0500 @@ -277,4 +277,37 @@ #endif } +/** + * ether_addr_equal_64bits - Compare two Ethernet addresses + * @addr1: Pointer to an array of 8 bytes + * @addr2: Pointer to an other array of 8 bytes + * + * Compare two Ethernet addresses, returns true if equal, false otherwise. + * + * The function doesn't need any conditional branches and possibly uses + * word memory accesses on CPU allowing cheap unaligned memory reads. + * arrays = { byte1, byte2, byte3, byte4, byte5, byte6, pad1, pad2 } + * + * Please note that alignment of addr1 & addr2 are only guaranteed to be 16 bits. + */ + +static inline bool ether_addr_equal_64bits(const u8 addr1[6+2], + const u8 addr2[6+2]) +{ +#ifdef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS + unsigned long fold = ((*(unsigned long *)addr1) ^ + (*(unsigned long *)addr2)); + + if (sizeof(fold) == 8) + return zap_last_2bytes(fold) == 0; + + fold |= zap_last_2bytes((*(unsigned long *)(addr1 + 4)) ^ + (*(unsigned long *)(addr2 + 4))); + return fold == 0; +#else + return ether_addr_equal(addr1, addr2); +#endif +} + + #endif /* _LINUX_ETHERDEVICE_H */ > > [ Upstream commit 567b871e503316b0927e54a3d7c86d50b722d955 ] > > bonding: rlb mode of bond should not alter ARP originating via bridge > > Do not modify or load balance ARP packets passing through balance-alb > mode (wherein the ARP did not originate locally, and arrived via a bridge). > > Modifying pass-through ARP replies causes an incorrect MAC address > to be placed into the ARP packet, rendering peers unable to communicate > with the actual destination from which the ARP reply originated. > > Load balancing pass-through ARP requests causes an entry to be > created for the peer in the rlb table, and bond_alb_monitor will > occasionally issue ARP updates to all peers in the table instrucing them > as to which MAC address they should communicate with; this occurs when > some event sets rx_ntt. In the bridged case, however, the MAC address > used for the update would be the MAC of the slave, not the actual source > MAC of the originating destination. This would render peers unable to > communicate with the destinations beyond the bridge. > > Signed-off-by: Matthew O'Connor > CC: Zheng Li > Cc: Jay Vosburgh > Cc: Andy Gospodarek > Cc: "David S. Miller" > > > diff -uprN linux-3.4.28/drivers/net/bonding/bond_alb.c > linux-3.4.28-patched/drivers/net/bonding/bond_alb.c > --- linux-3.4.28/drivers/net/bonding/bond_alb.c 2013-01-27 > 23:51:45.000000000 -0500 > +++ linux-3.4.28-patched/drivers/net/bonding/bond_alb.c 2013-01-30 > 15:37:25.121708311 -0500 > @@ -704,6 +704,12 @@ static struct slave *rlb_arp_xmit(struct > struct arp_pkt *arp = arp_pkt(skb); > struct slave *tx_slave = NULL; > > + /* Don't modify or load balance ARPs that do not originate locally > + * (e.g.,arrive via a bridge). > + */ > + if (!bond_slave_has_mac(bond, arp->mac_src)) > + return NULL; > + > if (arp->op_code == htons(ARPOP_REPLY)) { > /* the arp must be sent on the selected > * rx channel > diff -uprN linux-3.4.28/drivers/net/bonding/bonding.h > linux-3.4.28-patched/drivers/net/bonding/bonding.h > --- linux-3.4.28/drivers/net/bonding/bonding.h 2013-01-27 > 23:51:45.000000000 -0500 > +++ linux-3.4.28-patched/drivers/net/bonding/bonding.h 2013-01-30 > 15:37:25.121708311 -0500 > @@ -18,6 +18,7 @@ > #include > #include > #include > +#include > #include > #include > #include > @@ -450,6 +451,18 @@ static inline void bond_destroy_proc_dir > } > #endif > > +static inline struct slave *bond_slave_has_mac(struct bonding *bond, > + const u8 *mac) > +{ > + int i = 0; > + struct slave *tmp; > + > + bond_for_each_slave(bond, tmp, i) > + if (ether_addr_equal_64bits(mac, tmp->dev->dev_addr)) > + return tmp; > + > + return NULL; > +} > > /* exported from bond_main.c */ > extern int bond_net_id; > diff -uprN linux-3.4.28/include/linux/etherdevice.h > linux-3.4.28-patched/include/linux/etherdevice.h > --- linux-3.4.28/include/linux/etherdevice.h 2013-01-27 > 23:51:45.000000000 -0500 > +++ linux-3.4.28-patched/include/linux/etherdevice.h 2013-01-30 > 15:37:25.121708311 -0500 > @@ -277,4 +277,37 @@ static inline unsigned long compare_ethe > #endif > } > > +/** > + * ether_addr_equal_64bits - Compare two Ethernet addresses > + * @addr1: Pointer to an array of 8 bytes > + * @addr2: Pointer to an other array of 8 bytes > + * > + * Compare two Ethernet addresses, returns true if equal, false otherwise. > + * > + * The function doesn't need any conditional branches and possibly uses > + * word memory accesses on CPU allowing cheap unaligned memory reads. > + * arrays = { byte1, byte2, byte3, byte4, byte5, byte6, pad1, pad2 } > + * > + * Please note that alignment of addr1 & addr2 are only guaranteed to > be 16 bits. > + */ > + > +static inline bool ether_addr_equal_64bits(const u8 addr1[6+2], > + const u8 addr2[6+2]) > +{ > +#ifdef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS > + unsigned long fold = ((*(unsigned long *)addr1) ^ > + (*(unsigned long *)addr2)); > + > + if (sizeof(fold) == 8) > + return zap_last_2bytes(fold) == 0; > + > + fold |= zap_last_2bytes((*(unsigned long *)(addr1 + 4)) ^ > + (*(unsigned long *)(addr2 + 4))); > + return fold == 0; > +#else > + return ether_addr_equal(addr1, addr2); > +#endif > +} > + > + > #endif /* _LINUX_ETHERDEVICE_H */ > > > > > On 05/22/2013 05:26 PM, Smart Weblications GmbH - Florian Wiessner wrote: >> Hi Greg, >> >> >> Am 22.05.2013 22:04, schrieb Greg KH: >> >> >>>> https://kernel.googlesource.com/pub/scm/linux/kernel/git/jkirsher/net-next/+/567b871e503316b0927e54a3d7c86d50b722d955%5E!/ >>> Ok, that's what we need. >>> >>> Now, please cc: the developers / maintainers of that patch and ask them >>> to have it included in the 3.4-stable kernel series. >>> >>> Then, if they agree, the network maintainer will pick it up and send it >>> to me for inclusion. >>> >> i set committer David S. Miller in cc already, but do not >> know the network maintainer... >> >> this seems to me that "Matthew O'Connor" sent this to >> netdev on 2013-02-01: >> >> http://lists.openwall.net/netdev/2013/02/01/86 >> >> but i couldn't find a trace of the patch in 3.4.36?! >> >> Instead, i read another try to get it backported fail: >> >> http://permalink.gmane.org/gmane.linux.network/264198 >> >> >> >> >> >> > -- Mit freundlichen Grüßen, Florian Wiessner Smart Weblications GmbH Martinsberger Str. 1 D-95119 Naila fon.: +49 9282 9638 200 fax.: +49 9282 9638 205 24/7: +49 900 144 000 00 - 0,99 EUR/Min* http://www.smart-weblications.de -- Sitz der Gesellschaft: Naila Geschäftsführer: Florian Wiessner HRB-Nr.: HRB 3840 Amtsgericht Hof *aus dem dt. Festnetz, ggf. abweichende Preise aus dem Mobilfunknetz