2.6.5 forcedeth sending .. trash? / intel 82540EM dropping packets?

From: Mark Rutherford (mark_at_maunzelectronics.com)
Date: 04/05/04

  • Next message: Jeff Sipek: "Re: 2.6.5-aa1"
    To: Linux Kernel <linux-kernel@vger.kernel.org>
    Date:	Mon, 5 Apr 2004 02:55:56 +0000
    
    

    Hi all,
    before I get into this, since I am not on list, I would appreciate any
    response be CC'd. thanks!

    I started having problems with my onboard intel gigabit controller, and
    decided to fire up the nforce ethernet driver in an attempt to isolate the
    problem. I was trying to determine if its a kernel setup, or something on the
    machine.. who knows.... I am trying to find out.

    but, the nforce ethernet driver appears to be inop. ( or the interface is
    hosed...)
    I tried compiling it in at first, it finds the interface and I can configure
    it.
    however I cannot use the interface...
    according to ifconfig, its transmitting the packets, its just not recieving
    any.
    now, the real fun is when I happened to look at my switch and note that the
    activity light is rapidly flashing.
    so, I telnet into the switch to find out that its rather upset about this and
    has isolated that port...
    when I unload the module it continues to send .. random stuff out, and the
    only way to stop this is to shut it down, and start the machine back up.

    I try using a different port, and cable and I get the same results.
    since I dont have Window$ I cant see if its actually working or not. (at least
    in Window$)

    however, that leads me to another question. I have this other interface, the
    intel gigabit that I used for the last few months just fine until the other
    day I noticed something odd start to happen.
    basically, what happens is anything that sends data rapidly, tends to 'vanish'
    for example, if I were to fire up a game browser and search for a server it
    would 'lose' chunks of the queries to other servers.
    in one in-game browser it makes the game flat out crash, and so I started
    trying to figure out why this all is happening....
    what I see a ton of in the strace for the game, before it dies is the
    following:
    gettimeofday({1081103264, 263945}, NULL) = 0
    ioctl(6, FIONREAD, [0]) = 0
    read(0, 0xbfff754f, 1) = -1 EAGAIN (Resource temporarily
    unavailable)
    recvfrom(13, 0x93848c0, 32768, 0, 0xbfff75a0, 0xbfff759c) = -1 EAGAIN
    (Resource temporarily unavailable)

    now, I googled around about this EAGAIN (Resource temporarily unavailable)
    error and came to some conclusion it had to do with the networking.
    correct me if im wrong.

    so, what im asking, am I having a problem with the card, the kernel, or both?
    (with regards to the intel)
    whats odd about it, is that im just noticing this problem. but it only occurs
    when a lot of connections out are made.
    some more interesting things, are:
    I ran sniffit on the linux nat box, and these packets never made it there.
    and, I run ping -f and ping the hell out of another machine, and it always got
    a response. I let it run for 2 hours, and when I ctrl+c'd it, it only lost 1
    response, out of like 10 million.
    and, the last interesting thing...
    I can try rapidly connecting to the machine with the intel nic from another
    machine and it seems to lose it then, as well.
    an example of how I can reproduce this is as follows:
    traceroute google.com, hit ctrl+c, uparrow, and enter rapidly.
    but not so fast you cant see the first 2 or 3 hops...
    every second or third try it will drop out on you. and its almost ALWAYS the
    second or third try.
    ex:

    traceroute: Warning: google.com has multiple addresses; using 216.239.37.99
    traceroute to 216.239.37.99 (216.239.37.99), 30 hops max, 40 byte packets
     1 192.168.0.1 (192.168.0.1) 0.239 ms 0.167 ms 0.207 ms
    traceroute: Warning: google.com has multiple addresses; using 216.239.39.99
    traceroute to 216.239.39.99 (216.239.39.99), 30 hops max, 40 byte packets
     1 192.168.0.1 (192.168.0.1) 0.230 ms * 0.241 ms

    it lost the second response there....

    im lost.

    and as far as the forcedeth, I dont know what to do with that one.

    someone, please give me some advice here. throw the board out? :)

    -- 
    Regards,
    mark@freequest.net
    Server Admin, FreeQuest IRC Network
    http://www.freequest.net
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at  http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at  http://www.tux.org/lkml/
    

  • Next message: Jeff Sipek: "Re: 2.6.5-aa1"

    Relevant Pages

    • Re: Program to interface
      ... Your example is a gray area of relationship navigation. ... even if it is embedded in the "owning" class. ... getChild interface of the collection class. ... The problem would not exist if the superclass had named the response the ...
      (comp.object)
    • =?ISO-8859-1?Q?Re=3A_cFS_-_Fujitsu_S7010D_Centrino_1?= =?ISO-8859-1?Q?=2E7GHz_La
      ... Intel Centrino 1.7Ghz CPU ... Onboard 56kps modem and 10/100/1000 Broadcom Network ... I am struggling to see how you keep in business given your totally OTT and unprofessional response - you being a responsible and reliable trader and all that. ... If you want to see rude and OTT you should have seen the diatribe DCA sent ...
      (uk.adverts.computer)
    • Re: Faulty Hard Drive?
      ... Maxtor was acquired by Seagate, so the Seagate web site is where you'd ... there is a "feature" where the interface drops down ... to PIO mode, instead of using DMA mode, if ... the drive to the Intel Southbridge. ...
      (alt.comp.periphs.mainboard.asus)
    • Re: Message-based vs. method-based interaction [was: Re: LSP and subtype]
      ... >>> interface is always the method signature. ... >>> client has done to trigger the response), ... Such separation is impossible to do with pure 3GL language ...
      (comp.object)
    • Re: showing user input errors in MVC
      ... these issues within MVC, MVP, Smalltalk or the general industry. ... application interface but instead should be based on a meta-description ... Relationship or composite domain model rules. ... response by the widget to an error notification from its model. ...
      (comp.lang.smalltalk)