April 2003 - ixp1200 - lists.cs.princeton.edu

IPsec implementation on IXP1200
by Ning Weng 09 May '03

09 May '03

Hi, Have anyone impelmented the IPsec on the IXP1200 network processor? Thanks. -Ning

2 1

Re: RE: [ixp1200] l3fwdr
by vikram nayak 05 May '03

05 May '03

I got the mac addresses form enp> test macadd show....it displayed all 4 macs addresses.. in the ixsys*.config file i added these macadd to the ports 0, 1, 2, 3... ( i havent modified the route* file, like some had suggested to at the GW, show I? ) ... did ./ixstart ixsys*.config.... ifconfig .. for the l3 sample displays all 4 interfaces for the count sample displays none of th interfaces ? in both cases when i try pinging the port from another machine directly connected by a crossover, it says network unreachable... any ideas?? when i did a tcpdump on the external machine it hung and now refuses to reboot? can i connect the port to a LAN through a straight cable and be able to notice activity?? thanx, vikram ----- Original Message ----- From: vikram nayak <vnayak(a)usc.edu> Date: Thursday, April 17, 2003 8:52 pm Subject: Re: RE: [ixp1200] l3fwdr > > how do i prog the PHY device so that its 10MB mode is not default. How do i > knowif the device is 10 or 100? > > ----- Original Message ----- > From: "Botham, Mick" <mick.botham(a)intel.com> > Date: Sunday, April 6, 2003 3:42 am > Subject: RE: [ixp1200] l3fwdr > > > Kamesh, > > > > I'm not sure if this applies to the ENP-2305, but it did affect the > > IXP1200EB. > > > > Some PHY devices default 10MB mode is also loop-back mode, so on a 10MB > > connection, as soon as a packet is sent by the IXP1200 it gets looped back > > to the IXP1200, and so the process goes on, the result is a packet flood. > > The fix was either to connect to a 100MB device so the PHY would > > auto-negotiate to 100MB, or to program the PHY so that in 10MB mode it > > wasn't set to loopback. > > > > Rgds > > > > Mick > > > > > > Hi, > > Ping was not running. In fact the PC kind of hangs ie mouse > > pointer disappears etc when I connnect. I am able to see the output only > > after I disconnect. > > Configuration: > > Ethernet port 1 of ENP-2505 is connected back-to-back to the eth1 of > > the host PC ( PC in which the card is plugged in) through a crossover > > cable. > > The IP address of the IXP port is 192.168.1.150 and the IP address of > > the PC port is 192.168.1.3 . > > > > Also is there any option is the ixsys.config file through which I can > > specify the rate ? > > Thanks, > > Kamesh. > > > > > > Abhijeet Joglekar wrote: > > > > >Hi Kameswara, > > > > > >I guess you have already tried to connect the PC port to the IXP > > >port without going thro' the hub as Cyrus suggested. Just to rule > > >out any autonegotiation problems betweem 10/100 hub and IXP port. > > > > > >So is ping running when you captured that tcpdump output? > > >Also, what are the interface addresses of the IXP port > > >and the PC port in your setup? > > > > > >-- abhijeet > > > > > >On Thu, 3 Apr 2003, Kameswara Avasarala wrote: > > > > > > > > > > > >>Hi, > > >> This is what I observe if I do a tcpdump. > > >> > > >> > > >>Cyrus Hall wrote: > > >> > > >> > > >> > > >>>Hi there- > > >>> > > >>>You mention a great deal of link level activity. Have you done a TCP > > >>>dump? I had my ENP-2505 hooked up to a 10 Mbps hub which was causing > > >>>all sorts of issues, including a packet flood to be issued from the card > > >>>whenever I tried to start up a uACE. Make sure all interfaces are 100 > > >>>Mbps and see if that helps... > > >>> > > >>> > > >>> > > > > > > > > > > > > > > > > >

4 4

Drivers for mapping memory regions.
by Kameswara Avasarala 29 Apr '03

29 Apr '03

Hi, The IXP1200 hardware manual gives the memory map of the IXP1200. Can ioremap be used to map the physical addresses given into the linux kernel's address space ? Kamesh.

2 2

Re: ixp
by Vikram P. Nayak 26 Apr '03

26 Apr '03

Hi again, I finally got the L3 appl running. I managed to force the eth. card on the Linux box to run at 100Mbps, and also i noticed that it takes two config files ixsys.config-l3 and ixsys.config. And ixsys.config, is the one that brings up the IXP interfaces. I'll be looking to modify the Count App thou' ... like i said ealier the interfaces are not listed when i do an ifconfig, also when I ping the card from an external box , ping returns nothing. But the count value on that port keeps incrementing.,, is this how it supposed to work? thanx, Vikram On Fri, 25 Apr 2003, Abhijeet Joglekar wrote: > Vikram, > > Sorry, couldnt get back earlier. > > I havent looked at it in detail, but I am guessing that the L3 > MicroAce does something special (like maybe adding device entries) > which the Count or Bridge MicroAce does not do. However, the > interfaces are anyway active if you specify so in the ixsys.config > file. > > > So are you able to run the L3/L2 samples yet? > > -- abhijeet > > On Thu, 24 Apr 2003, vikram nayak wrote: > > > Abhijeet, > > one quick question with the Count application. once i run the startup script > > shouldn't the interfaces be UP... ie, when i do a ifconfig I shud be able to see > > the 4 interfaces,eth1-eth4 (besides eth0 and lo) ,,right? > > the interfaces are displayed when I run the script for 'l3', .. . but not for > > the Count? > > vikram > > > > > >

2 2

Fw: [ixp1200] ENP2505 Example Issues
by Yong Liu 23 Apr '03

23 Apr '03

----- Original Message ----- From: "Yong Liu" <yongliu(a)ee.tamu.edu> To: "vikram nayak" <vnayak(a)usc.edu> Sent: Wednesday, April 23, 2003 12:09 PM Subject: Re: [ixp1200] ENP2505 Example Issues > Hi, Vikram > > I have applied the 2.01a patch and connected 2 of the ethernet ports on the > ENP-2505 card to two different 100Mbps switches. > > The current problem is that the PC connected to ENP-2505 through the 100Mbps > switch receives a lot of identical garbage packets > when I run the Count example. > > Because the network is 100Mbps now, it is different from Cyrus' experience. > > Anybody know how to program the PHY device to disable the loopback mode? > > Thanks a lot! > > Yong > > ----- Original Message ----- > From: "vikram nayak" <vnayak(a)usc.edu> > To: <yongliu(a)ee.tamu.edu> > Sent: Tuesday, April 22, 2003 7:30 PM > Subject: Re: [ixp1200] ENP2505 Example Issues > > > > Hi Liu, > > I tried running the Count application again today, and i was able to see > some packets... > > I'm running linux on the core, and using vmware.. r u doingthe same?? > > Anyways, this is what i did differently today, > > I ran the patch that comes with CD1. compiled the microcode(cygwin), > tranfered them to the linux host /arme-be... > > compiled the source for the core . > > the only modification in the config file is the "mac address" - i guess > becasue this is merely a layer 2 appl, ip address do not matter. > > I plugged in one of our straight cables( on our local LAN) into the IXP > port. Now the ixp is listening to all packets on the our LAN since its in > promosicous mode. > > I telnet into the card, run ./ixstart ixconfifg* ... and like 10 secs > after the images are loaded onto the card, the count on that particular port > kept incrementing. > > > > letme know if it works > > vikram > > > > P.S. do u have the port macs with you?? what are the mac addresses like? > what did u do to get them? > > >

1 0

Re: [ixp1200] ENP2505 Example Issues
by Yong Liu 22 Apr '03

22 Apr '03

Hi, Vikram I just begin to run these example code. I am facing the same problem to bring up the 4 ports. Anybody can tell us what we should do except modifying the config files? Thanks. ----- Original Message ----- From: "vikram nayak" <vnayak(a)usc.edu> To: <yongliu(a)ee.tamu.edu> Sent: Tuesday, April 22, 2003 3:47 AM Subject: Re: [ixp1200] ENP2505 Example Issues > Hi Liu, > I am working on the count/l3fwd that came with the CD. I noticed u are too. Are u able to run the the code on the card. How r u testing it? what are u using for generating packets. > I am facing some difficulties bring up the the 4 ports on the ixp, did u face any problem? did u modify the configure files that are passed to ixstart? > what mac address did u specify in the config files? > > Please help me out here, > thanx, > Vikram

2 1

ENP2505 Example Issues
by Cyrus Hall 21 Apr '03

21 Apr '03

I recently started working with the ENP2505 board that Intel is shipping for their IXP1200 academic program. After hacking their install scripts, dealing with kernel symbol issues, and hacking their boot scripts, I finally got the thing to boot Linux and communicate with the host system. I then made the encouraging discovery that all the sample projects are (seemingly) for other boards. Although they'll load, the board will proceed to spit out malformed Ethernet packets (each packet identical) as fast as the interface will allow. This is with the simple packet counting example. Is there a bare bones MicroACE, such as the counting example, that actually works on the ENP2505? Also, the documentation and tools provided with the board seems to indicate that using the ACE development paradigm is mandatory. Is it possible just to run microcode with no communication back to the ARM on ENP2505, ie, non-ACE code? Thanks for any pointers, Cyrus Hall hallcp(a)cs.colorado.edu University of Colorado Software Engineering Research Laboratory

4 4

SRAM commands queuing
by Yan Luo 20 Apr '03

20 Apr '03

Hi, I am interested in the SRAM command issuing and queuing in the SRAM functional unit. The hardware ref manul talked about how the command bus arbiter select commans from the command FIFOs. However, the exact timing is not stated in docs. For example, why does it take 3 cycles from issuing a sram reference command to enqueuing the command to the functional unit (as seen in the the thread history window) ? Also, it seems that removing a command from the SRAM queue cannot be done every cycle. It takes at least two cycles to dequeue the next command. Is it because the core and the sram bus operate at different frequency? I use IXP1200 now, I wonder how these parameter are different in IXP1200, IXP2400 and IXP2800? Could anyone please explain it or refer me to some useful documents? Any input will be appreciated. Thanks, Yan -- Yan Luo Dept. of CSE, Univ. of California Riverside, CA 92521, USA http://www.cs.ucr.edu/~yluo/

1 0

1 gigabit port support on microengines
by Dinesh Yaji 19 Apr '03

19 Apr '03

Hi All, I am a newbie to this group and apologise if this has been asked before. 1) The IXP200 SDk tutorial ( for the packet count example) says that 8 threads are needed for one gigabit port. It does not specify if 8 threads are needed for receive and 8 threads are needed for send or 8 threads are needed to support one gigabit port (i.e receive and send). 2) I intend to write code to support one gigabit port ( and no 10/100 ports). Do I need to have 4 microengines dedicated for receive/transmit alone? ( 2 for sending + 2 for receiving)? thanks! -Dinesh _________________________________________________________________ Protect your PC - get McAfee.com VirusScan Online http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

1 0

IXP1200 uEngine hangup problem
by shyam＠tejasnetworks.com 19 Apr '03

19 Apr '03

Hi all, We are facing some microengine hangup problem in IXP1200 as detailed out below. Has anybody encountered these problems in their projects. Thanks. Regards -Shyam It looks as if there is some problem with the Context Arbiter of IXP1200 or in the signal event handling mechanism. For example, for the RX scheduling portion in Rx threads, we do For the below code, hangup occurs in the following scenarios --- (1) In macro port_rx_receive() at the instruction ctx_arb[start_receive].. On examining the registers CTX_<N>_SIG_EVENTS ---- <N> is the context number & CTX_<N>_WAKEUP_EVENTS, it is found that the start_receive FBI signal event is reflected in both the registers (this clears up any problem being present in our recv_req issue). If context arbiter signal event event handling mechanism was OK, the uEngine should proceed forward. The strange fact is that reading RCV_CNTL register from the StrongARM core in a burst, results in things going forward. This implies that RCV_CNTL FIFO clears on read from StrongARM, and that uEngines have not been forcibly paused and rules out any chance of there being a memory corruption. So clearly, the FBI signal event is not getting received by uEngine threads, even though their registers reflect it several times. (2) Sometime the uEngine thread gets stuck in a ctx_arb[voluntary] instruction. It is found that other 3 uEngine threads keep on executing with context switching happening between them, ruling out any chance of a uEngine thread hogging the resources. Also the particular thread is not set into paused state as seen from CTX_ENABLES. (3) In macro port_rxrdy_chk() at the instrcution csr[read, $rec_rdy, rcv_rdy_lo], ctx_swap Here again as in case (1) above, the registers CTX_<N>_SIG_EVENTS & CTX_<N>_WAKEUP_EVENTS ---- <N> is the context number, reflect a FBI event for the context, while implies that the Rx Ready flags have been read. All the other contexts in this uEngine get scheduled one after the other in this interval, ruling out any chance of another context hogging the uEngine. All contexts are enabled to run as seen from CTX_ENABLES. Clearly the above 3 scenarios point to some problem occuring in context arbitration, particulary FBI events & the voluntary token. What could be the reason for strange occurences? -------------------Relevant MACROS----------------------------------------- // critsect_init #macro critsect_init(abs_reg) immed[abs_reg, 0xFFFFFFFF] #endm // critsect_enter #macro critsect_enter(abs_reg) begin#: alu[--, --, B, abs_reg] br<0[end#], guess_branch ctx_arb[voluntary] br[begin#] end#: immed[abs_reg, 0] #endm // critsect_exit #macro critsect_exit(abs_reg) immed[abs_reg, 0xFFFFFFFF] #endm // port_rxrdy_chk // Description: Check if receive port is ready and wait till it becomes so. // Outputs: // io_abs_rdyFlagsSem : (@GPR) Semaphore to restrict ready flag // reads to a single thread within a uEngine // Inputs: // in_rdyFlagOfPortSvced : (GPR) Ready Flag of port being serviced // by current context #macro port_rxrdy_chk(io_abs_rdyFlagsSem, in_rdyFlagOfPortSvced) .local $rec_rdy check_port#: // Block other contexts from reading rec ready, otherwise // threads processing packets wont get to use the FBI critsect_enter[io_abs_rdyFlagsSem] csr[read, $rec_rdy, rcv_rdy_lo], ctx_swap // Get port ready // Allow other contexts to read rec ready critsect_exit[io_abs_rdyFlagsSem] // Check if port has data alu[--, in_rdyFlagOfPortSvced, AND, $rec_rdy] br!=0[port_ready#] // If port has data then go process it ctx_arb[voluntary] // If port doesn't have data, loop br[check_port#] port_ready#: .endlocal #endm // issue_rx_request // Description: Issues an assembled receive request. // Inputs: // in_recReq : (GPR) Assembled receive request #macro issue_rx_request(in_recReq) .local $recReqCsr alu[$recReqCsr, --, B, in_recReq] csr[write, $recReqCsr, rcv_req], ctx_swap .endlocal #endm // port_rx_receive // Description: Waits till a receive request serviced by RSM, reads and // returns the RCV_CNTL register contents // Outputs: // out_rcvCntl : (GPR) Assembled Receive Request #macro port_rx_receive(out_rcvCntl) .local $rec_csr ctx_arb[start_receive] csr[read, $rec_csr, rcv_cntl], ctx_swap alu[out_rcvCntl, --, B, $rec_csr] .endlocal #endm -------------------End Relevant MACROS-------------------------------------- ----MAIN THREAD------------------- -------------------------------- --------------------------------- while (1) ---------------------------------------- -----Some Processing-------------------- ---------------------------------------- // Wait till port assigned to this context has data, create and issue // a request to service it, and wait for request to be serviced by RSM port_rxrdy_chk(@rdready_inflight, g_svcedPortRdyFlag) critsect_enter(@req_inflight) issue_rx_request(rcvReq) //critsect_exit(@req_inflight) port_rx_receive(rcvCntl) critsect_exit(@req_inflight) ---------------------------------------- -----Some Processing-------------------- ---------------------------------------- end while ----END MAIN THREAD-------------------

1 0