- ixp1200 - lists.cs.princeton.edu

RE: [ixp1200] File info from nfs?
by David Mentre 16 Feb '04

16 Feb '04

Hello, > From: Øyvind Hvamstad [mailto:oyvindh@ifi.uio.no] > I'm writing a program, where I need the sizes of files. In > linux I would > call the function stat() to fetch this information. What > function would > I call on the strongARM? Without detailed explanation from your side (which platform, which OS, ...), I can only guess. If you are using the software framework provided by Intel in IXP1200, you have the choice between Linux and VxWorks. If it is Linux, use your standard stat() syscall. Otherwise (VxWorks or bare-bone code on the processor), you are on your own. Yours, d. -- David Mentré <mentre(a)tcl.ite.mee.com> - Research engineer (Ph.D.) Mitsubishi Electric ITE-TCL / European Telecommunication Research Lab Phone: +33 2 23 45 58 29 / Fax: +33 2 23 45 58 59 http://www.mitsubishi-electric-itce.fr

2 1

RE: [ixp1200] Help needed
by David Mentre 13 Feb '04

13 Feb '04

Hello sharmila, > From: sharmila radhakrishnan [mailto:shar_mila2002in@yahoo.co.in] > How to pass exceptional packets from the microengine to the > StrongArm.Can I use SDRAM & SRAM as we would do for > communication between microengines? I think so. You can also use scratch rings in scratch memory to exchange messages. > Since the address space > for StrongArm & Microengines are different do I need to do > the address translation myself? Intel provides certain facility to do that. Look at halMe_GetVirAddr() and following fonctions in XScale(tm) Support Libraries Reference Manual (IXP2400_IXP2800SRM.pdf) > I have one more doubt.Is it necessary to follow the ACE > programming framework for developing applications to download > the code on to the IXP1200 kit and make the code run? No. You can write your own code to download microcode into micro-engines. But, much simpler than that, Intel provides an Hardware Abstraction Layer (HAL) that allows you to do that. Look at UcLo - Microcode Loader Library, in the same manual as above. Yours, d. -- David Mentré <mentre(a)tcl.ite.mee.com> - Research engineer (Ph.D.) Mitsubishi Electric ITE-TCL / European Telecommunication Research Lab Phone: +33 2 23 45 58 29 / Fax: +33 2 23 45 58 59 http://www.mitsubishi-electric-itce.fr

1 0

Help needed
by sharmila radhakrishnan 13 Feb '04

13 Feb '04

Hello All, How to pass exceptional packets from the microengine to the StrongArm.Can I use SDRAM & SRAM as we would do for communication between microengines?Since the address space for StrongArm & Microengines are different do I need to do the address translation myself?I would be very thankful if you'll could help me out with this. I have one more doubt.Is it necessary to follow the ACE programming framework for developing applications to download the code on to the IXP1200 kit and make the code run? Thanks in advance, Regards, Sharmila Yahoo! India Education Special: Study in the UK now.

1 0

ENP-2611 board PCI-to-PCI bridge testing problem!
by smiley 12 Feb '04

12 Feb '04

question ? ~~~! Since I still could not load the module, I changed hhnet_drawb.c. ---from---- #define DRAWB_CFG_SETUP_DS3 0xB8 #define DRAWB_CFG_CHIP_CONTROL_1 0xCE #define HHNET_DRAWB_MEM_SIZE (1024 * 1024) if (primary) { ... bar = 4; irq_handler = hhnet_drawb_intr_primary; } ---to---- #define DRAWB_CFG_SETUP_DS3 0xAC #define DRAWB_CFG_CHIP_CONTROL_1 0xCC #define HHNET_DRAWB_MEM_SIZE (4 * 1024) if (primary) { ... bar = 0; irq_handler = hhnet_drawb_intr_primary; } --- After the modification, I COULD load the module on both side, server, and ENP. But still I could not make them to communicate. The dmesg & PCI info from the server side----------------- <dmesg> hhnet_drawb: mem_paddr = fe6b0000 hhnet_drawb: mem virt = f89d900 <pci> 03:08.0 Bridge: Intel Corp.: Unknown device b555 (rev 02) Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- Latency: 64 Interrupt: pin A routed to IRQ 24 Region 0: [virtual] Memory at fe6b0000 (32-bit, non-prefetchable) [size=4K] Region 1: I/O ports at 2400 [size=256] Region 2: [virtual] Memory at fe500000 (32-bit, non-prefetchable) [size=1M] Region 3: Memory at e0000000 (32-bit, prefetchable) [size=256M] Capabilities: [dc] Power Management version 2 Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-) Status: D0 PME-Enable- DSel=0 DScale=0 PME- Capabilities: [e4] Vital Product Data Capabilities: [ec] #06 [0080] The dmesg & PCI info from the ENP side----------------- <dmesg> hhnet_drawb: mem_paddr = 00000000 hhnet_drawb: mem virt = cb687000 <pci> Bus 0, device 1, function 0: Bridge: PCI device 8086:b555 (Intel Corp.) (rev 2). IRQ 41. Non-prefetchable 32 bit memory at 0xe4000000 [0xe4000fff]. I/O at 0x100000 [0x1000ff] As you can see, the mem_paddr of ENP is set as 00000000. I think this is the reason of the problem. I think the correct value is "0xe4000000". Am I correct? If then, how do I fix it? Thank you in advance. Regards, EunYoung Joo

2 1

ENP-2611 board PCI-to-PCI bridge testing problem!
by smiley 12 Feb '04

12 Feb '04

ENP-2611 board PCI-to-PCI bridge testing problem! 1. I downloaded the ENP_SDK_3.1_release3 for the PCI-to-PCI bridge communication. 2. I compiled it as described in ch. 6 in "ENP Software Development Kit Programmer's Guide". 3. The document said to load the modules in this manner. - insmod hhnet_eth insmod hhnet insmod hhnet_drawb ifconfig pci0 192.168.111.1 - When I load the module, the following message is displayed. - host/hhnet_drawb.o: init_module: No such device Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters - The output of "dmesg", is shown like this. - hhnet_drawb: Downstream setup 3 configuration invalid. (b:d:f = 3:8:0) - The host pci info is like below and it seems everything is configured correctly. - Bus 3, device 8, function 0: Bridge: PCI device 8086:b555 (Intel Corp.) (rev 2). IRQ 24. Non-prefetchable 32 bit memory at 0xfe6b0000 [0xfe6b0fff]. I/O at 0x2400 [0x24ff]. Non-prefetchable 32 bit memory at 0xfe500000 [0xfe5fffff]. Prefetchable 32 bit memory at 0xe0000000 [0xefffffff]. - When I look at the source code, In the function "hhnet_drawb_open_dev", you compare the setup value of "pci_read_config_dword()" and HHNET_DRAWB_MEM_MASK, and it seems like the setup value of the "pci_read_config_dword()" is zero. I guess this is the main reason for this problem. I also saw the same problem on the "target" side. What do I have to do to fix this problem?

1 0

RE: [ixp1200] File info from nfs?
by David Mentre 12 Feb '04

12 Feb '04

Hello Øyvind, > From: Øyvind Hvamstad [mailto:oyvindh@ifi.uio.no] > Sent: Wednesday, February 11, 2004 7:13 PM > Hi, I wonder how I might get file info, such as size from the > nfs mount > on the ixp1200. Is there something like the syscall stat(file, buf) > available? I don't understand your question. NFS mounts are seen as regular file system on Linux, so you can use usual Unix tools to get file size, access permission, etc. Yours, david -- David Mentré <mentre(a)tcl.ite.mee.com> - Research engineer (Ph.D.) Mitsubishi Electric ITE-TCL / European Telecommunication Research Lab Phone: +33 2 23 45 58 29 / Fax: +33 2 23 45 58 59 http://www.mitsubishi-electric-itce.fr

2 1

RE: [ixp1200] [In ixp2400/2800 programming] chapter 5. nextblock?
by David Mentre 12 Feb '04

12 Feb '04

Hello, > From: paragon [mailto:commoon@wins21.com] > Present, not trasmitted from RX to TX in enp-2611(ixp2400). > > To see in chapter 5, > 373 immed32(dl_next_block, SPI4_RX_NEXT_BLOCK) > > What's Nextblock? dl_next_block is a context sensitive registers that stores the identifier of the next microblock to which send the packet. It is used within the dl_sink[] macro, in Rx microblock. > #define SPI4_RX_NEXT_BLOCK 0x24 > > what's difference in ixp2400? I don't understand your question. > In chapter 5, what's difference between ixp2400 and ixp2800? > How to modify in chapter.5 ? Basically, the SPI4 mode does not exist on IXP2400. You should rewrite configurations steps. Maybe other things should be also changed. I don't know more as I haven't look at the code provided with the book. If you want to program on IXP2400, you'd rather start from SDK reference designs. You will find several designs on SDK Application CD-ROM. If you don't have it, ask your Intel representative. Yours, d. -- David MENTRE <mentre(a)tcl.ite.mee.com> - Research engineer (Ph.D.) Mitsubishi Electric ITE-TCL / European Telecommunication Research Lab Phone: +33 2 23 45 58 29 / Fax: +33 2 23 45 58 59 http://www.mitsubishi-electric-itce.fr

1 0

[In ixp2400/2800 programming] chapter 5. nextblock?
by paragon 12 Feb '04

12 Feb '04

Hello. Dear ixp1200. Present, not trasmitted from RX to TX in enp-2611(ixp2400). To see in chapter 5, 373 immed32(dl_next_block, SPI4_RX_NEXT_BLOCK) What's Nextblock? #define SPI4_RX_NEXT_BLOCK 0x24 what's difference in ixp2400? In chapter 5, what's difference between ixp2400 and ixp2800? How to modify in chapter.5 ? thank you~~^^.

1 0

[ixp2400] RX flow question ?
by paragon 11 Feb '04

11 Feb '04

Hi~~^^ I don't understand to following step! chapter 5 ixp2400/2800 programming original source --------------------------------------- _spi3_rx_move_rbuf_to_dram(rsw.w1.parts.element, cur_mpacket_addr, rsw.w1.parts.byte_count, &rbuf_to_dram_sig); // Update the buffer length dlMeta.bufferSize += rsw.w1.parts.byte_count; // Wait for the mpacket to move into DRAM __wait_for_all(&rbuf_to_dram_sig); _spi3_rx_free_rbuf(rsw.w1.parts.element); // If this is the EOP mpacket then return if (rsw.w1.parts.eop) { break; } // Update the reassembly pointer cur_mpacket_addr += rsw.w1.parts.byte_count; --------------------------------------- I think so to follow. --------------------------------------- _spi3_rx_move_rbuf_to_dram(rsw.w1.parts.element, cur_mpacket_addr, rsw.w1.parts.byte_count, &rbuf_to_dram_sig); __wait_for_all(&rbuf_to_dram_sig); dlMeta.bufferSize += rsw.w1.parts.byte_count; cur_mpacket_addr += rsw.w1.parts.byte_count; if (rsw.w1.parts.eop) { break; } _spi3_rx_free_rbuf(rsw.w1.parts.element); --------------------------------------- what do you think about? Thank you so much in advance!

1 0

File info from nfs?
by Øyvind Hvamstad 11 Feb '04

11 Feb '04

Hi, I wonder how I might get file info, such as size from the nfs mount on the ixp1200. Is there something like the syscall stat(file, buf) available? -- Øyvind Hvamstad <oyvindh(a)ifi.uio.no>

1 0