Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 26 Oct 1998 10:44:05 -0500
From:      Dave Chapeskie <dchapes@borderware.com>
To:        Drew Baxter <netmonger@genesis.ispace.com>
Cc:        hackers@FreeBSD.ORG
Subject:   Re: XL0
Message-ID:  <98Oct26.110114est.115594@gateway.borderware.com>
In-Reply-To: <4.1.0.67.19981022124020.00ad2100@genesis.ispace.com>; from Drew Baxter on Thu, Oct 22, 1998 at 12:43:17PM -0400
References:  <4.1.0.67.19981022124020.00ad2100@genesis.ispace.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Oct 22, 1998 at 12:43:17PM -0400, Drew Baxter wrote:
> Got my dmesg output for the day.. (this is probably my 2nd day of using my
> XL card instead of a EP0'd 3c509), and i see this.. Anyone able to shed
> some light on it?  It runs fine otherwise, not sure if this is killing
> performance when we have this happen or not.
> 
> > xl0: transmission error: 82
[repeats]
> 
> ---
> Drew "Droobie" Baxter
> Network Admin/Professional Computer Nerd(TM)
> OneEX: The OneNetwork Exchange 207-942-0275
> http://www.droo.orland.me.us
> My Latest Kernel: FreeBSD 3.0-CURRENT (ONEEX) #14: Mon Oct 19 22:36:58 EDT 1998

I had the exact same thing and this is the response I got from the
author of the xl0 driver:

> > a) What does transmission error 82 mean?
>
> The 2 part in this case indicates the error which, according to the
> manual is:
>
> #define XL_TXSTATUS_RECLAIM     0x02 /* 3c905B only */
>
> The manual says that this error happens because "the packet experienced
> a collision after the front of the packet had been reclaimed to the
> FIFO free space." In other words, the chip had sent the packet and had
> already erased part of it from its internal FIFO RAM when it detected
> a collision. The driver detects this error and, if the packet is still
> in the driver's outbound queue, it will attempt to retransmit it. If
> it isn't in the queue, the driver just acknowledges the TX error interrupt
> and continues. I don't think there should be any traffic interruptions.
>
> The 80 part means the transmission is complete:
>
> #define XL_TXSTATUS_COMPLETE    0x80
>
> > b) Is it something I should be concerned about?
>
> Not really. If the error message annoys you, you can hide it under an
> #ifdef DIAGNOSTIC/#endif pair. The driver recovers from the condition
> correctly, so it's really just an informational message. (There was a
> bug in older versions of the driver where the error handler failed to
> check if there really was a packet in the outbound queue and could panic
> when it tried to dereference the NULL queue head pointer, but I fixed
> that.)
>
> > c) Is it a potential problem with my cabling or hub?
>
> Mmm... possibly. I would expect to see this on a busy shared ethernet
> with lots of collisions, but there might be other causes. In my test
> environment I use an ethernet switch and don't ever see these errors.
>
> -Bill
>
> -- 
> =============================================================================
> -Bill Paul            (212) 854-6020 | System Manager, Master of Unix-Fu
> Work:         wpaul@ctr.columbia.edu | Center for Telecommunications Research
> Home:  wpaul@skynet.ctr.columbia.edu | Columbia University, New York City
> =============================================================================
>  "It is not I who am crazy; it is I who am mad!" - Ren Hoek, "Space Madness"
> =============================================================================

Hope this helps.
-- 
Dave Chapeskie <dchapes@borderware.com>

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?98Oct26.110114est.115594>