Phil,
Not sure how your mailer works - I am assuming these are the same
patches i got CCed on.
On 16-03-04 07:11 AM, Phil Sutter wrote:
Cc: Felix Fietkau <n...@openwrt.org>
Signed-off-by: Phil Sutter <p...@nwl.cc>
---
man/man8/tc-connmark.8 | 55 ++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 55 insertions(+)
create mode 100644 man/man8/tc-connmark.8
diff --git a/man/man8/tc-connmark.8 b/man/man8/tc-connmark.8
new file mode 100644
index 0000000000000..bb4cf7543dfdb
--- /dev/null
+++ b/man/man8/tc-connmark.8
@@ -0,0 +1,55 @@
+.TH "Connmark retriever action in tc" 8 "11 Jan 2016" "iproute2" "Linux"
+
+.SH NAME
+connmark - netfilter connmark retriever action
+.SH SYNOPSIS
+.in +8
+.ti -8
+.BR tc " ... " "action connmark " [ " zone"
+.IR u16_zone_index " ] [ " BRANCH " ] ["
+.BI index " u32_index "
+]
+
+.ti -8
+.IR BRANCH " := { " reclassify " | " pipe " | " drop " | " continue " | " ok "
}"
It would be of benefit to have a general man page describing tc
actions/filters (may be one for tc action and other for tc filter).
I probably started the mess of calling this construct a "branch"
which actually is misleading.
These are controls ("if/else" are branches; a loop is not a branch).
Refer to my netdev01 paper. If you can extract details from hat paper
in a main man page or at minimal reference it in the action/classifier
man pages then the action specific man pages would just specify what
the default is.
An example usage of this from the commit message:
-----
...lets tag outgoing icmp with mark 0x10..
iptables -tmangle -A PREROUTING -p icmp -j CONNMARK --set-mark 0x10
..add on ingress of $ETH an extractor for connmark...
tc filter add dev $ETH parent ffff: prio 4 protocol ip \
u32 match ip protocol 1 0xff \
flowid 1:1 \
action connmark continue
...if the connmark was 0x11, we police to a ridic rate of 10Kbps
tc filter add dev $ETH parent ffff: prio 5 protocol ip \
handle 0x11 fw flowid 1:1 \
action police rate 10kbit burst 10k
----
cheers,
jamal
