Transcript INF5060: Multimedia data communication using network processors

INF5060:
Multimedia data communication using network processors
Running A First Example:
The Web Bumper
3/9 - 2004
Overview
 Packet header and encapsulation review
 How to use the card
 programming model
 booting
 starting and stopping
 Lab setup
 Web bumper
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Packet Headers and
Encapsulation
Ethernet
48 bit address configured to an
interface on the NIC on the receiver
0
1
2
3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Destination Address
|
+
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
|
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+
|
Source address
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Frame type
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
describes content of ethernet frame,
e.g., 0x0800 indicates an IP datagram
INF5060 – multimedia data communication using network processors
48 bit address configured to an
interface on the NIC on the sender
2004 Carsten Griwodz & Pål Halvorsen
Internet Protocol version 4 (IPv4)
indication of the abstract parameters of the
quality of service desired – somehow treat
indicates the format of the internet header, i.e., version 4
high precedence traffic as more important –
length of the internet header in 32 bit words, and thus
tradeoff between low-delay, high-reliability, and
points to the beginning of the data (minimum value of 5) high-throughput – NOT used, bits now reused
first zero, fragments allowed
and last fragment
1
identifying
value to aid
assembly of
fragments
disable a packet
to circulate
forever,decrease
value by at least
1 in each node –
discarded if 0
0
2
3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|Version| IHL |Type of Service|
Total Length
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Identification
|Flags|
Fragment Offset
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Time to Live |
Protocol
|
Header Checksum
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Source Address
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Destination Address
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Options
|
Padding
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
indicates used transport
layer protocol
32-bit address fields. May be configured
differently from small to large networks
datagram length
(octets) including
header and data allows the length
of 65,535 octets
indicate where
this fragment
belongs in
datagram
checksum on the
header only – TCP,
UDP over payload.
Since some header
fields change(TTL),
this is recomputed
and verified at
each point
options may extend the header – indicated by IHL. If the options do
not end on a 32-bit boundary, the remaining fields are padded in the
padding field (0’s)
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
UDP
port to identify the
sending application
port to identify
receiving application
0
1
2
3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Source Port
|
Destination Port
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Length
|
Checksum
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
specifies the total length of the UDP
datagram in octets
INF5060 – multimedia data communication using network processors
contains a 1’s complement checksum over
UDP packet and an IP pseudo header with
source and destination address
2004 Carsten Griwodz & Pål Halvorsen
TCP
code bits: urgent, ack, push, reset, syn, fin
sequence number
for data in payload
port to identify the
sending application
port to identify
receiving application
acknowledgement
for data received
0
1
2
3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Source Port
|
Destination Port
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Sequence Number
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ receiver’s buffer
header length in |
Acknowledgment Number
| size for
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
32 bit units
additional data
| Header|
|U|A|P|R|S|F|
|
| length| Reserved |R|C|S|S|Y|I|
Window
|
|
|
|G|K|H|T|N|N|
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Checksum
|
Urgent Pointer
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|
Options
|
Padding
|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
contains a 1’s complement checksum over
UDP packet and an IP pseudo header with
source and destination address
pointer to urgent data in segment
options may extend the header. If the options do not end on a 32-bit
boundary, the remaining fields are padded in the padding field (0’s)
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Encapsulation
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Using IXP1200
Programming Model
 Active computing elements (ACEs):
 basic programming abstraction defined by Intel’s software
developer kit (SDK), not part of hardware
 asynchronous unit of execution
 a typical system contains at least three in a pipeline:
input
ports


ingress ACE
process ACE
egress ACE
output
ports
unidirectional queues between ACEs allow processors to run
independently
late bindings are used to bind ACE output at run-time
(unbounded targets can be used to drop packets)
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Programming Model
 Packet flow illustration for IP forwarding:
Stack ACE
ingress ACE
(core)
StrongARM
microengine
input
ports
erroneous packet, not IP, …
ingress ACE
(microblock)
IP ACE
(core)
egress ACE
(core)
packet for this host
IP ACE
(microblock)
INF5060 – multimedia data communication using network processors
egress ACE
(microblock)
output
ports
2004 Carsten Griwodz & Pål Halvorsen
Programming Model
 Often many possible communication paths,
but only one used per packet
 Division of work between ACEs running both on
microengines and StrongARM crucial for high
performance:


microengines comprise fast data path – the common case
StrongARM handles all exceptions
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Programming Model
 Challenge: achieve optimal parallelism
 efficient resource utilization  each part of the pipeline
should spend the same amount of time to process a packet,
reduce idle time

which microengine runs which ACE, many possible
configurations, e.g.,



exactly one microblock ACE per engine
a pipeline of microblocks for ACEs on one or more engines
microblock groups:
 set of microblocks that run on a single engine
 specified in a configuration file
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Programming Model
 Microblock group:
input
ports
ingress ACE
process ACE
microengine 1
egress ACE
output
ports
microengine 2
 Replicated microblock groups
input
ports
ingress ACE
process ACE
microengine 1
egress ACE
input
ports
ingress ACE
process ACE
output
ports
microengine 3
microengine 2
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Programming Model
 Microblock dispatch loop
 control packet flow among ACEs in a microblock group
Stack ACE

runs in an infinite loop, e.g., IP forwarder:
ingress ACE
(core)
IP ACE
(core)
egress ACE
(core)
ingress ACE
(microblock)
IP ACE
(microblock)
egress ACE
(microblock)
initialize microblocks
while (1) {
get next packet from input port
invoke ingress microblock
if not IP packet
send to ingress core
else {
invoke IP microblock
}
}
if packet not is for this host
send to egress microblock
else
send to IP core
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Programming Model
 Exceptions (interrupts raised by microengines)
 refer to packets passed from microblock components
(microengine) to a core component (StrongARM)

the symbolic constant IX_EXCEPTION can be used

components are identified using a tag

an exception code informs the core component of the type

core component must have an exception handler
determining how to process the packet
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Programming Model
 Crosscalls
 enable an ACE to invoke a function in another ACE
 similar to remote procedure calls (RPC)

must be programmed into both caller and callee
early binding, programmer determines type (not run-time)

specified by declaring a function

three types




deferred – caller do not block ; asynchronous return notification
oneway – caller do not block ; no return value
twoway – caller blocks until callee returns value
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
File Access, Communication and Booting
 PCI Ethernet emulation


physically, the card plugs into the host’s PCI bus
host and board communicates sending Ethernet frames over the PCI
 The disk is mounted using NFS over PCI, e.g.,
/opt/ixasdk is available at both host and card
 The card itself has no persistent memory




a RAM disk for the root file system is
when booting the card, an image is loaded from the host
bootixp performs these instructions and boots the card
allow some time after booting
 Rebooting the card from the board (reboot) only clears/reinitialize the card
 must run bootixp from host afterwards
 Running bootixp on a running card will crash the card


boot (reinitialize) card through minicom, followed by a bootixp, or
host reboot
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Environment Requirements
 The
Makefiles assume that the following environment
variables are set:

IXROOT = /opt/ixasdk

CONFIG = arm_be

… this is already set for the root user on the lab machines
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Starting and Stopping
 Telnet to the card: telnet ixp
 To start the example: ixstart <config_file>
 our example config files use relative paths
 cd $IXROOT/bin/arm_be ; ./ixstart ixsys.config
 To stop the example: ixstop
 stops whatever ACE that is running
 cd $IXROOT/bin/arm_be ; ./ixstop
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Lab Setup
Lab Setup
IXP lab
switch
switch
Internet
…
Student lab
switch
“reboot x”
INF5060 – multimedia data communication using network processors
power switch
2004 Carsten Griwodz & Pål Halvorsen
Lab Setup – Power Switch
IXP lab
switch
switch
…
power switch
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Lab Setup - Addresses
IXP lab
switch
switch
…
power switch
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Lab Setup - Addresses
IXP lab
129.240.67.111
switch
192.168.67.111
switch
129.240.67.112
192.168.67.112
129.240.67.113
192.168.67.113
…
129.240.67.110
…
…
129.240.67.118
192.168.67.5
192.168.67.118
power switch
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Lab Setup – Data Path
IXP lab
PCI
ssh connection
(129.240.67.112)
IO
switch
hub
switch
hub interface
IXP1200
To 192.168.67.5
system
bus
…
To 192.168.67.5
memory
hub
local network
(192.168.67.112)
RAM CPU
interface
memory
power switch
web bumper
(counting web packets and forwarding
all packets from one interface to another)
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Web Bumper
Lab Setup – Data Path
IXP lab
IO
switch
hub
ssh connection
(129.240.67.112)
IXP1200
To 192.168.67.5
…
memory
hub
local network
(192.168.67.112)
switch
CPU
memory
power switch
web bumper
(counting web packets and forwarding
all packets from one interface to another)
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
The Web Bumper
the wwbump coreACE checks a
packet forwarded by the microACE:
 count web packet - add 1 to counter
 send back to wwbump microACE
 provides access to counter via a
crosscall server
web bumper
IXP1200
wwbump
(core)
StrongARM
microengines
input
port
ingress ACE
(microblock)
the wwbump microACE checks all packets if
it is a web packet:
 if not, forward packet to egress microACE
(and sent to the other port)
 if web packet:
 send to wwbump coreACE (StrongARM)
 forward packets from core to egress microACE
(and sent to the other port)
wwbump
(microblock)
output
port
egress ACE
(microblock)
web bumper
(counting web packets and forwarding
all packets
toprocessing
another) encompasses all
ingress processing encompasses
all from one interface
egress
operations performed as packets arrive
operations applied as packets depart
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
IXP1200
The Web Bumper
the crosscall client reads the
web packet counter and prints
the current count
web bumper
wwbump
(core)
crosscall
client
wwbump
(microblock)
egress ACE
(microblock)
StrongARM
microengines
input
port
ingress ACE
(microblock)
INF5060 – multimedia data communication using network processors
output
port
2004 Carsten Griwodz & Pål Halvorsen
Configuration File
 It is run at boot time and specifies
 ACEs to be loaded
 which side to run an ACE (ingress or egress)
 number and speed of ports

see ixsys.config
 The interfaces that are to be started at system boot time:
interface 0 10.1.0.1 10.1.0.255 255.255.255.0 00:01:02:03:04:05 1
interface 1 10.2.0.1 10.2.0.255 255.255.255.0 00:01:02:03:04:06 1
…
 Using workbench or not (we do not)
mode 0
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Configuration File
 The microcode and microaces that are to be started at
system boot time
file 0 ./SlowIngressWWBump.uof


ingress microcode (type 0)
associated slow ports (10/100 Mbps)
microace ifaceInput ./ingressAce none 0 1





ifaceInput is a microace
executable
no config file
running on ingress side
type ingress
microace wwbump ./wwbump none 0 0



runs on ingress side (first zero)
has unknown type (second zero)
…
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Configuration File
 Bind configuration
bind static ifaceInput/default wwbump

binding for ingress (ifaceInput) is wwbump
bind static wwbump/default ifaceOutput

binding for wwbump output is ifaceOutput
 Shell commands to be run, e.g., add ARP entries in the cache
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Identifying Web Packets
These are the header
fields you need for
the web bumper:
 Ethernet type 0x800
 IP type 6
 TCP port 80
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
WWBump Dispatch Loop
initialize dispatch loop macros
do forever {
if packet has arrived from StrongARM
send to egress
if packet has arrived from Ethernet port {
invoke WWBump macro to process packet
if it is a web packet (exception specified)
send to StrongARM
else
send to egress
}
}
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
WWBump Macro
allocate 6 SDRAM registers to hold header data
get IXP input port  set IXP output port to the other
read first 24 bytes into the 6 SDRAM registers (etype, IP hlen, IP type)
if not frame type is IP (0x800), forward packet
if not IP type is TCP (6), forward packet
calculate port number address using IP header length
read TCP destination port number
if not TCP destination port is 80, forward packet
set exception code
set wwbump tag
set IX_EXCEPTION
return
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Testing
1.
Log in on the assigned IXP machine (129.240.67.xxx) using ssh as root
(twice, the example is easier using two different windows)
2.
Download code and place at a convenient place under $IXROOT/src, e.g.:


3.
cd $IXROOT/src/microaces/aces/
scp –r [email protected]:/hom/paalh/INF5060code/wwbump bumptest
Compile the code and make sure the targets are moved to $IXROOT/bin/arm_be


microcode for the microengines:
cd $IXROOT/src/microaces/aces/bumptest/ucbuild; make ; cp *.uof $IXROOT/bin/arm_be/.
c-code for the StrongARM:
cd .. ; make
(wwbump and getcount are automatically put right directory)
4.
Copy configuration file: cp ixsys.config $IXROOT/bin/arm_be/ixsys.config-bumptest
5.
Telnet to the IXP card as root: telnet ixp
6.
Enter bin directory: cd $IXROOT/bin/arm_be/
7.
Stop any running aces: ./ixstop
8.
Start the bumper: ./ixstart ixsys.config-bumptest
INF5060 – multimedia data communication using network processors
(you may already be there!!)
2004 Carsten Griwodz & Pål Halvorsen
Testing
The bumper should now be running:
1.
In IXP window, check web packet count (should be 0): ./getcount
2.
In HOST window

ping web server machine: ping 192.168.67.5
(packets should be forwarded through the card, but not counted, i.e.: ./getcount  0)

telnet to web server machine using port 80: telnet 192.168.67.5 80
(packets should be forwarded through the card, and counted, i.e.: ./getcount  4 (!!??))
Experiment with the card:

… start and stop the aces

… reboot the card

… remotely restart the machine using the power switch
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen
Summary
 We’ll use remote access to the IXP lab machines
 Bumper shows the basic concepts
 Intel SDK and programming model
 microengines and StrongARM
 much code even for a small, trivial task
 Next: short demo (here) and testing (in the lab)
 Next week: extending the bumper example
INF5060 – multimedia data communication using network processors
2004 Carsten Griwodz & Pål Halvorsen