• State Not Answered
  • Locked Locked
  • Replies 3 replies
  • Subscribers 22 subscribers
  • Views 1075 views
  • Users 0 members are here
Options
Related
How was your experience today?
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Nichelite TCP Stack misbehaving in network

Saurabh P

Dear all,

I took two MCB2300 keil boards, loaded them with
nichelite's TCP stack (keil ported version Downloaded from NXP site), and after that,

set IPs to both them 199.199.50.66 and 199.199.50.67 respectively. (as 199.199.50.x is my NW pattern)

[Off course Plug them in to office network].

I observe that at a time any of the a board can
communicate in office LAN ( other get down) , via
PING ( I checked PING from other PCs).

And I see that any a board can communicate with N
number of PCs but , when I plug 2 boards in network
then any one only can communicate !!

Is there any setting(s) I am missing that i should
make in any of header/config file of stack code?

[I checked IP values not conflicting, and I am
setting IP by what user commands they given like
setip etc, in non DHCP mode]

Pls help me, if anybody have exp. with nichelite.

Parents
  • 0 in reply to ImPer Westermark

    Thanks buddy,

    I get this work by MAC change.

    Now what issue I am facing is for PING (Network) connectivity.

    My Office network have around 80 PCs in LAN,
    When I am configuring Nichelite loaded board
    with IP 199.199.50.67 (unique ip), and plug it in network.

    Here I am doing PING test with any of the networked
    PC. It works well initially.

    But, After some time (say 1/2/3 hours) pc is not
    getting reply of the PING.

    At that time I debugged at code. What I see is
    that eth_isr0(void) __irq (emac.c) is not getting
    invoke (which should- I think) anymore then.

    I referred the implementation manual published at NXP
    site for nichelite implementation note.

    From which I could tried to alter NUMBIGBUFS,NUMLILBUFS,BIGBUFSIZE,LILBUFSIZE defines
    by diff values , I could see that it affects how much
    time the board can live in network.

    Is there any rugged method/patch which i can apply
    to make sure my MCB2300 board lives in the network
    24x7 with (continuous PING TEST form other PC)

    I will appreciate any help,
    -Saurabh

Reply
  • 0 in reply to ImPer Westermark

    Thanks buddy,

    I get this work by MAC change.

    Now what issue I am facing is for PING (Network) connectivity.

    My Office network have around 80 PCs in LAN,
    When I am configuring Nichelite loaded board
    with IP 199.199.50.67 (unique ip), and plug it in network.

    Here I am doing PING test with any of the networked
    PC. It works well initially.

    But, After some time (say 1/2/3 hours) pc is not
    getting reply of the PING.

    At that time I debugged at code. What I see is
    that eth_isr0(void) __irq (emac.c) is not getting
    invoke (which should- I think) anymore then.

    I referred the implementation manual published at NXP
    site for nichelite implementation note.

    From which I could tried to alter NUMBIGBUFS,NUMLILBUFS,BIGBUFSIZE,LILBUFSIZE defines
    by diff values , I could see that it affects how much
    time the board can live in network.

    Is there any rugged method/patch which i can apply
    to make sure my MCB2300 board lives in the network
    24x7 with (continuous PING TEST form other PC)

    I will appreciate any help,
    -Saurabh

Children
  • 0 in reply to saurabh p

    Additional info:

    I observe it every time after it dropped the
    PING connection, remaining OS contine to
    work, I can enter debug commands.

    And on givinh the buffer command, it shows all
    allocated buffers , at there it shows many
    duplicate buffers already allocated (bigbufs).

    I see each time that 5-7 duplicate bigbufs
    allocated (may be they have not been de-allocated
    may cause of communication failure)