Search

Technical Discussion Group Forum

This forum is provided for user discussion. While Beacon EmbeddedWorks support staff and engineers participate, Beacon EmbeddedWorks does not guarantee the accuracy of all information within in the Technical Discussion Group (TDG).

The "Articles" forums provide brief Articles written by Beacon EmbeddedWorks engineers that address the most frequently asked technical questions.

To receive email notifications when updates are posted for a Beacon EmbeddedWorks product download, please subscribe to the TDG Forum of interest.

TDG Forum

PrevPrev Go to previous topic
NextNext Go to next topic
Last Post 25 Apr 2006 09:11 AM by  kwong
5474LITE Evaluation Problems
 5 Replies
Sort:
You are not authorized to post a reply.
Author Messages
gary.carlson
New Member
New Member
Posts:


--
09 Jan 2006 04:06 PM
    I am a principal engineer that is presently tasked with developing a next generation instrument platform within my company. Since I am leaning towards using the Freescale 5474 running a Linux OS, I decided to buy the evaluator board. Needless to say, this has not been a very rewarding experience.

    Problem #1: It is very confusing who exactly is responsible for the Linux build image. Freescale has instructions posted on their website for this board and Logic has theirs. Needless to say Freescale's website doesn't have a valid LTIB Linux ISO image posted (instead it plays the song “If God Permits” by Mark Daniel). Awful support from Freescale. They admit the image has a problem, but I haven't heard a peep from them since as to when they think it will fixed. Next I tried the Logic Linux image and my luck with that has been slightly better. One of the biggest problems I have encountered is that the Logic documentation on how to make this image work on the evaluator is fractured across several documents instead of one cohesive document. This lead to time being wasted chasing down known problems like "Freescale's debugger monitor doesn't play with our version of Linux" etc.

    Problem #2: After solving everything else, I am still having a problem downloading compiled build images into the evaluator using TFTP in a reasonable time frame. A typical 4M image is taking 12 minutes to download over a 100-Base-T connection! After starting a packet sniffer it became obvious to me that LOSH's load command is bursting chunks of data followed by long periods of silence. What is causing this behavior? A fast connection should rip through this in a few seconds.

    Problem #3: Although slow, eventually LOSH finishes its load process. Next I execute the image. A few log statements comes across on the serial port and then another period of silence begins that lasts for about 4 minutes. Once the silence ends, then it blazes through to startup. (See captured log below for details and location of quiet period). What is causing this problem?

    Hopefully you can help me quickly finish this evaluation and instill some confidence in this processor (and the evaluator). So far it has been a big disappointment.

    Gary Carlson



    Capture log following "exec" command in LOSH:

    vma_start: 0xc0000 vma_end: 0x4f8574
    Copy image [0x800428:0xc3899c] to [0xc0000:0x4f8574]
    Start image at 0xc0000

    ***4 minutes of silence occurs here****

    Linux version 2.4.26 (gcarlson@nevada) (gcc version 3.4.3) #1 Mon Jan 9 12:11:49 PST 2006
    starting up linux rev 0.2: startmem 0xc023e000, size 61MB
    initrd: c033e000 - c05fa574
    Flex CS0: ff800000 007f0001 00102180
    Flex CS1: e0000000 07ff0001 00102180
    Flex CS4: 40000000 0fff0001 00108580
    Flex CS5: 50000000 0fff0001 00103980
    On node 0 totalpages: 8192
    zone(0): 1024 pages.
    zone(1): 7168 pages.
    zone(2): 0 pages.
    Kernel command line: root=/dev/ram rootfstype=romfs
    Have no hardware time clock, returning ToD as 00:00:00 on 1 Jan 1970
    Calibrating delay loop... 263.78 BogoMIPS
    Memory: 60008k/65536k available (1024k kernel code, 4448k data, 56k init)
    kmem_create: Forcing size word alignment - mm_struct
    kmem_create: Forcing size word alignment - filp
    Dentry cache hash table entries: 8192 (order: 3, 65536 bytes)
    Inode cache hash table entries: 4096 (order: 2, 32768 bytes)
    Mount cache hash table entries: 1024 (order: 0, 8192 bytes)
    kmem_create: Forcing size word alignment - bdev_cache
    kmem_create: Forcing size word alignment - cdev_cache
    kmem_create: Forcing size word alignment - kiobuf
    Buffer cache hash table entries: 2048 (order: 0, 8192 bytes)
    Page-cache hash table entries: 8192 (order: 2, 32768 bytes)
    POSIX conformance testing by UNIFIX
    Linux/m68k PCI BIOS32 revision 0.06
    ColdFire PCI Host Bridge (Rev. 0) detected
    PCI: Probing PCI hardware
    layout_bus: starting bus 0
    layout_bus: disable_dev for bus 0
    layout_bus: starting bus 0 devices
    layout_bus: bus 0 finished
    Initialize Multi-channel DMA API Alpha v0.3 (2004-04-26)
    Linux NET4.0 for Linux 2.4
    Based upon Swansea University Computer Society NET3.039
    kmem_create: Forcing size word alignment - sock
    Initializing RT netlink socket
    Starting kswapd
    kmem_create: Forcing size word alignment - file_lock_cache
    devfs: v1.12c (20020818) Richard Gooch (rgooch@atnf.csiro.au)
    devfs: boot_options: 0x1
    JFFS2 version 2.1. (C) 2001 Red Hat, Inc., designed by Axis Communications AB.
    i2c-core.o: i2c core module version 2.6.1 (20010830)
    i2c-algo-mcf.o: I2C ColdFire algorithm module is loaded.
    pty: 256 Unix98 ptys configured
    ColdFire internal UART serial driver version 1.00
    ttyS0 at 0xe0008600 (irq = 99) is a builtin ColdFire UART
    ttyS1 at 0xe0008700 (irq = 98) is a builtin ColdFire UART
    tracek: ring buffer size: 64K
    kmem_create: Forcing size word alignment - blkdev_requests
    FireEgine CompactFlash IDE driver
    Flex CS4: 40000000 0fff0001 00108580
    Flex CS5: 50000000 0fff0001 00103980
    fire_engine_cpld_init:98 hw revision 58
    fire_engine_cpld_init:110 cpld_programmed 0
    cfide_init:554 present 0
    RAMDISK driver initialized: 16 RAM disks of 8192K size 1024 blocksize
    loop: loaded (max 8 devices)
    IPsec over IPv4 tunneling driver
    PPP generic driver version 2.4.2
    MCF547x/8x DSPI device driver installed
    Coldfire DSPI driver is loaded
    Flex CS0: ff800000 003f0001 00102180
    cfi_cmdset_0001: Erase suspend on write enabled
    Using word write method
    Creating 2 MTD partitions on "FireEngine BootFlash":
    0x00000000-0x000c0000 : "LoLo"
    0x000c0000-0x00400000 : "flash-boot"
    Initializing Cryptographic API
    NET4: Linux TCP/IP 1.0 for NET4.0
    IP Protocols: ICMP, UDP, TCP, IGMP
    kmem_create: Forcing size word alignment - ip_dst_cache
    IP: routing cache hash table of 1024 buckets, 8Kbytes
    TCP: Hash tables configured (established 8192 bind 16384)
    ip_conntrack version 2.1 (512 buckets, 4096 max) - 148 bytes per conntrack
    ip_tables: (C) 2000-2002 Netfilter core team
    arp_tables: (C) 2002 David S. Miller
    NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
    RAMDISK: Compressed image found at block 0

    decompressed 0256KB
    decompressed 0512KB
    decompressed 0768KB
    decompressed 1024KB
    decompressed 1280KB
    decompressed 1536KB
    decompressed 1792KB
    decompressed 2048KB
    decompressed 2304KB
    decompressed 2560KB
    Freeing initrd memory: 351k freed
    VFS: Mounted root (romfs filesystem) readonly.
    Mounted devfs on /dev
    starting: hwclock
    hwclock: Could not access RTC: No such file or directory
    starting: proc
    starting: local-fs
    starting: flash-fs
    starting: BASE
    starting: syslogd
    sourcing: load-drivers
    sourcing: start-applications
    starting: network-settings
    Setting eth0 MAC addr to 00:08:ee:33:44:55
    ColdFire internal FEC driver version 0.11
    Getting eth0 IP configuration via DHCP
    udhcpc (v0.9.9-pre) started
    Sending discover...
    Sending select for 172.16.2.145...
    Lease of 172.16.2.145 obtained, lease time 691200
    Setting eth1 MAC addr to 33:44:56
    starting: irda
    starting: ip-settings
    starting: inetd
    starting: hotplug
    ** can't synthesize input events - /proc/bus/input/devices missing
    ** can't synthesize pci hotplug events
    starting: devfsd
    devfsd v1.3.25 started for /dev
    starting: boas
    done
    BusyBox v1.00 (2006.01.09-20:14+0000) multi-call binary
    Anonymous
    Posts:


    --
    11 Jan 2006 03:46 PM
    Hi Gary,

    Can you provide a part number for the Fire Engine that you're using?

    https://www.logicpd.com/auth/serial_no.php

    Thanks,
    Anonymous
    Posts:


    --
    11 Jan 2006 03:50 PM
    Hi Gary,

    Can you also tell me what version of LogicLoader that you're running? The current release is 2.0.5 and that's what is recommended for customers to use.

    Thanks,
    willoc
    New Member
    New Member
    Posts:


    --
    25 Jan 2006 10:03 AM
    I am having the same exact problem through tftp. Silly thing is that when I go through teraterm (serial) , the upload takes forever, but the code boots instantly.
    Anonymous
    Posts:


    --
    25 Jan 2006 10:49 AM
    Gary,

    Can you please try doing an 'ifconfig sm0 down' in LogicLoader prior to running the 'exec' command to launch Linux and let me know if it helps with the Linux kernel boot time?

    Thanks,
    kwong
    New Member
    New Member
    Posts:


    --
    25 Apr 2006 09:11 AM
    I'm using the same evaluation board...

    I found the ltib iso for the board on Freescale's site... using the iso, I was able to get a linux 2.6.10 system up and running within a day or so (with some minor tweaks).

    I have to agree that the documentation is somewhat fractured but it wasn't too bad if you've got the patience to hunt it all down. If you're still struggling with some issues maybe I can help you out.
    You are not authorized to post a reply.