chiark
/
gitweb
/
~mdw
/
firewall
/ commitdiff
commit
grep
author
committer
pickaxe
?
search:
re
summary
|
shortlog
|
log
|
commit
| commitdiff |
tree
raw
|
patch
|
inline
| side by side (from parent 1:
f01cdc7
)
fender.m4: Fix silly typo in comment.
author
Mark Wooding
<mdw@distorted.org.uk>
Fri, 1 Jul 2016 20:29:43 +0000
(21:29 +0100)
committer
Mark Wooding
<mdw@distorted.org.uk>
Fri, 1 Jul 2016 20:29:43 +0000
(21:29 +0100)
fender.m4
patch
|
blob
|
blame
|
history
diff --git
a/fender.m4
b/fender.m4
index 5f9597d6510428ada8f43a6ae34e79f30a63153a..f1d84d80bf245cfb59a6030177a92af84fa9e0d4 100644
(file)
--- a/
fender.m4
+++ b/
fender.m4
@@
-84,7
+84,7
@@
run ebtables -A check-bcp38 -j RETURN -p ip6 --ip6-source fe80::/10
run ebtables -A check-bcp38 -j bcp38 -p ip6
run ebtables -A FORWARD -j check-bcp38 -o bond0
run ebtables -A check-bcp38 -j bcp38 -p ip6
run ebtables -A FORWARD -j check-bcp38 -o bond0
-## There's a hideous bug in Linux
'
3.2.51-1's ebtables: for some reason it
+## There's a hideous bug in Linux 3.2.51-1's ebtables: for some reason it
## misparses (at least) locally originated multicast packets, and tries to
## extract IP header fields relative to the start of the Ethernet frame. The
## result is obviously a hideous mess. Don't try to do BCP38 checking for
## misparses (at least) locally originated multicast packets, and tries to
## extract IP header fields relative to the start of the Ethernet frame. The
## result is obviously a hideous mess. Don't try to do BCP38 checking for