Generic packet editor action in tc(8) Linux Generic packet editor action in tc(8)

pedit - generic packet editor action


tc ... action pedit [ex] munge { RAW_OP | LAYERED_OP | EXTENDED_LAYERED_OP } [ CONTROL ]

RAW_OP := offset OFFSET { u8 | u16 | u32 } [ AT_SPEC ] CMD_SPEC

AT_SPEC := at AT offmask MASK shift SHIFT

LAYERED_OP := { ip IPHDR_FIELD | ip BEYOND_IPHDR_FIELD } CMD_SPEC

EXTENDED_LAYERED_OP := { eth ETHHDR_FIELD | ip IPHDR_FIELD | ip EX_IPHDR_FIELD | ip6 IP6HDR_FIELD | tcp TCPHDR_FIELD | udp UDPHDR_FIELD } CMD_SPEC

ETHHDR_FIELD := { src | dst | type }

IPHDR_FIELD := { src | dst | tos | dsfield | ihl | protocol | precedence | nofrag | firstfrag | ce | df }

BEYOND_IPHDR_FIELD := { dport | sport | icmp_type | icmp_code }

EX_IPHDR_FIELD := { ttl }

IP6HDR_FIELD := { src | dst | traffic_class | flow_lbl | payload_len | nexthdr | hoplimit }

TCPHDR_FIELD := { sport | dport | flags }

UDPHDR_FIELD := { sport | dport }

CMD_SPEC := { clear | invert | set VAL | add VAL | decrement | preserve } [ retain RVAL ]

CONTROL := { reclassify | pipe | drop | shot | continue | pass | goto chain CHAIN_INDEX }

The pedit action can be used to change arbitrary packet data. The location of data to change can either be specified by giving an offset and size as in RAW_OP, or for header values by naming the header and field to edit the size is then chosen automatically based on the header field size.

ex
Use extended pedit. EXTENDED_LAYERED_OP and the add/decrement CMD_SPEC are allowed only in this mode.
offset OFFSET { u32 | u16 | u8 }
Specify the offset at which to change data. OFFSET is a signed integer, it's base is automatically chosen (e.g. hex if prefixed by 0x or octal if prefixed by 0). The second argument specifies the length of data to change, that is four bytes (u32), two bytes (u16) or a single byte (u8).
at AT offmask MASK shift SHIFT
This is an optional part of RAW_OP which allows to have a variable OFFSET depending on packet data at offset AT, which is binary ANDed with MASK and right-shifted by SHIFT before adding it to OFFSET.
eth ETHHDR_FIELD
Change an ETH header field. The supported keywords for ETHHDR_FIELD are:
src
dst
Source or destination MAC address in the standard format: XX:XX:XX:XX:XX:XX
type
Ether-type in numeric value
ip IPHDR_FIELD
Change an IPv4 header field. The supported keywords for IPHDR_FIELD are:
src
dst
Source or destination IP address, a four-byte value.
tos
dsfield
precedence
Type Of Service field, an eight-bit value.
ihl
Change the IP Header Length field, a four-bit value.
protocol
Next-layer Protocol field, an eight-bit value.
nofrag
firstfrag
ce
df
mf
Change IP header flags. Note that the value to pass to the set command is not just a bit value, but the full byte including the flags field. Though only the relevant bits of that value are respected, the rest ignored.
ip BEYOND_IPHDR_FIELD
Supported only for non-extended layered op. It is passed to the kernel as offsets relative to the beginning of the IP header and assumes the IP header is of minimum size (20 bytes). The supported keywords for BEYOND_IPHDR_FIELD are:
dport
sport
Destination or source port numbers, a 16-bit value. Indeed, IPv4 headers don't contain this information. Instead, this will set an offset which suits at least TCP and UDP if the IP header is of minimum size (20 bytes). If not, this will do unexpected things.
icmp_type
icmp_code
Again, this allows to change data past the actual IP header itself. It assumes an ICMP header is present immediately following the (minimal sized) IP header. If it is not or the latter is bigger than the minimum of 20 bytes, this will do unexpected things. These fields are eight-bit values.
ip EX_IPHDR_FIELD
Supported only when ex is used. The supported keywords for EX_IPHDR_FIELD are:
ttl
ip6 IP6HDR_FIELD
The supported keywords for IP6HDR_FIELD are:
src
dst
traffic_class
flow_lbl
payload_len
nexthdr
hoplimit
tcp TCPHDR_FIELD
The supported keywords for TCPHDR_FIELD are:
sport
dport
Source or destination TCP port number, a 16-bit value.
flags
udp UDPHDR_FIELD
The supported keywords for UDPHDR_FIELD are:
sport
dport
Source or destination TCP port number, a 16-bit value.
clear
Clear the addressed data (i.e., set it to zero).
invert
Swap every bit in the addressed data.
set VAL
Set the addressed data to a specific value. The size of VAL is defined by either one of the u32, u16 or u8 keywords in RAW_OP, or the size of the addressed header field in LAYERED_OP.
add VAL
Add the addressed data by a specific value. The size of VAL is defined by the size of the addressed header field in EXTENDED_LAYERED_OP. This operation is supported only for extended layered op.
decrement
Decrease the addressed data by one. This operation is supported only for ip ttl and ip6 hoplimit.
preserve
Keep the addressed data as is.
retain RVAL
This optional extra part of CMD_SPEC allows to exclude bits from being changed. Supported only for 32 bits fields or smaller.
CONTROL
The following keywords allow to control how the tree of qdisc, classes, filters and actions is further traversed after this action.
reclassify
Restart with the first filter in the current list.
pipe
Continue with the next action attached to the same filter.
drop
shot
Drop the packet.
continue
Continue classification with the next filter in line.
pass
Finish classification process and return to calling qdisc for further packet processing. This is the default.

Being able to edit packet data, one could do all kinds of things, such as e.g. implementing port redirection. Certainly not the most useful application, but as an example it should do:

First, qdiscs need to be set up to attach filters to. For the receive path, a simple ingress qdisc will do, for transmit path a classful qdisc (HTB in this case) is necessary:

tc qdisc replace dev eth0 root handle 1: htb
tc qdisc add dev eth0 ingress handle ffff:

Finally, a filter with pedit action can be added for each direction. In this case, u32 is used matching on the port number to redirect from, while pedit then does the actual rewriting:

tc filter add dev eth0 parent 1: u32 \
	match ip dport 23 0xffff \
	action pedit pedit munge ip dport set 22
tc filter add dev eth0 parent ffff: u32 \
	match ip sport 22 0xffff \
	action pedit pedit munge ip sport set 23
tc filter add dev eth0 parent ffff: u32 \
	match ip sport 22 0xffff \
	action pedit ex munge ip dst set 192.168.1.199
tc filter add dev eth0 parent ffff: u32 \
	match ip sport 22 0xffff \
	action pedit ex munge ip6 dst set fe80::dacb:8aff:fec7:320e
tc filter add dev eth0 parent ffff: u32 \
	match ip sport 22 0xffff \
	action pedit ex munge eth dst set 11:22:33:44:55:66
tc filter add dev eth0 parent ffff: u32 \
	match ip dport 23 0xffff \
	action pedit ex munge tcp dport set 22

To rewrite just part of a field, use the retain directive. E.g. to overwrite the DSCP part of a dsfield with $DSCP, without touching ECN:

tc filter add dev eth0 ingress flower ... \
	action pedit ex munge ip dsfield set $((DSCP << 2)) retain 0xfc

And vice versa, to set ECN to e.g. 1 without impacting DSCP:

tc filter add dev eth0 ingress flower ... \
	action pedit ex munge ip dsfield set 1 retain 0x3

tc(8), tc-htb(8), tc-u32(8)
12 Jan 2015 iproute2