Re: 3Com 3c905C-TX

2002-05-02 Thread Danny Braniss

 
 Out of curiosity, do only 3c509's exibit this behavior, or is this
 the core problem with 3c59x's as well?  My experiences have not
 been consistent with these cards, and I had assumed it was due
 to buggy code in the 3-Com chipset.  I've noticed flaky behavior from the
 Vortex [3c59x] card as well.  
 
no problem with: (the ethernet is OnBoard)
Dell GX-115: 3Com 3c905C-TX Fast Etherlink XL
Dell GX-150: 3Com 3c905C-TX Fast Etherlink XL

(PXE et.all)

danny



To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message



Re: 3Com 3c905C-TX

2002-05-02 Thread Guido Kollerie

On Wed, May 01, 2002 at 11:49:02PM +0200, Sten wrote:

 The only way to use them reliably is to force
 both the card and the switch. We came to the
 conclusion that fxp's are a nicer option.

I have exchanged the 3Com NIC for an Intel one. I'm using an
Intel NIC at work and haven't had any problems with it under
FreeBSD. What remains strange though is that the 3Com NIC used to
work just fine. As said before the strange behaviour
(full-duplex - half-duplex) occurred about a month ago. Around
the same time I performed a 'make world'. Was the xl driver
changed somehow in the period before that?

--
Guido



msg37951/pgp0.pgp
Description: PGP signature


Re: 3Com 3c905C-TX

2002-05-02 Thread Hiten Pandya

--- Guido Kollerie [EMAIL PROTECTED] wrote:
 I have exchanged the 3Com NIC for an Intel one. I'm using an
 Intel NIC at work and haven't had any problems with it under
 FreeBSD. What remains strange though is that the 3Com NIC used to
 work just fine. As said before the strange behaviour
 (full-duplex - half-duplex) occurred about a month ago. Around
 the same time I performed a 'make world'. Was the xl driver
 changed somehow in the period before that?

It doesnt look like, accept for the following two deltas:

  o rev. 1.103 by [EMAIL PROTECTED]
CVS Log: Remove __P.
Branch: MAIN CVS Tags: HEAD

  o rev. 1.104 by [EMAIL PROTECTED]
CVS Log:
 Change callers of mtx_init() to pass in an appropriate lock type name. 
 In most cases NULL is passed, but in some cases such as network driver 
 locks (which use the MTX_NETWORK_LOCK macro) and UMA zone locks, a name 
 is used.
Branch: MAIN CVS Tags: HEAD

Those are the only two revisions made to the if_xl.c driver in the timeframe
you have provided; and I don't think they can cause the issue you have
described, IMHO.

  -- Hiten Pandya

__
Do You Yahoo!?
Yahoo! Health - your guide to health and wellness
http://health.yahoo.com

To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message



Re: 3Com 3c905C-TX

2002-05-01 Thread Guido Kollerie

On Mon, Apr 29, 2002 at 08:08:02PM +0300, Danny Braniss wrote:

 at 100baseTX full-duplex is slower than 10Mgb :-(

Same problem here with the xl0 driver. My 3Com 3c905C-TX Fast
Etherlink XL connects to a Cisco Micro Switch (1548, unmanaged).
According to the lights on the back of the switch everything runs
at 100 Mbits full-duplex initially (that is after a reboot).
However after a while the switch indicates that it is running at
100 Mbits half-duplex. I don't know what causes it, but this is
happening for at least a month.

When this happens 'ifconfig -a' will still report that everything
is running at 100 Mbits full-duplex. Judging from the performance
and what the Cisco switch indicates this is not true, it is
running half-duplex!

Unfortunately the switch is unmanaged hence I am not able to
explicitely set the switch to 100 Mbits full-duplex. Using
ifconfig to set the nic to 10baseT/UTP and then back to 100baseTX
full-duplex doesn't help. Only a reboot will bring the NIC back
to 100 Mbits full duplex mode.

--
Guido



msg37926/pgp0.pgp
Description: PGP signature


Re: 3Com 3c905C-TX

2002-05-01 Thread Sten

On Wed, 1 May 2002, Guido Kollerie wrote:


snip

 Unfortunately the switch is unmanaged hence I am not able to
 explicitely set the switch to 100 Mbits full-duplex. Using
 ifconfig to set the nic to 10baseT/UTP and then back to 100baseTX
 full-duplex doesn't help. Only a reboot will bring the NIC back
 to 100 Mbits full duplex mode.

Please note that due to vagaries in the auto-negotiation
spec 3com and cisco dont work well together.
And 3coms ( on linux atleast ) have the added
bonus of sometimes deciding to change
speed/duplex just for the heck of it.

The only way to use them reliably is to force
both the card and the switch. We came to the
conclusion that fxp's are a nicer option.

IMHO just creating a reliable and clearly defined
auto-negotiation protocol will do more for ethernet
speed than gigabit ethernet :).


-- 
Sten Spans

  What does one do with ones money,
   when there is no more empty rackspace ?


To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message



Re: 3Com 3c905C-TX

2002-05-01 Thread Glendon Gross


Out of curiosity, do only 3c509's exibit this behavior, or is this
the core problem with 3c59x's as well?  My experiences have not
been consistent with these cards, and I had assumed it was due
to buggy code in the 3-Com chipset.  I've noticed flaky behavior from the
Vortex [3c59x] card as well.  

Just now I have been wrestling with an ISA 3c509 which has
a Lucent 40-01304 chip on it.  At first the card was detected, and
later not detected [on a different OS.]I vote for the fxp's as
well, I've had hardly any problems with them.

Is there a way to lock down the card by hacking the driver, so it won't 
try to auto-negotiate the connection?

On Wed, 1 May 2002, Sten wrote:

 On Wed, 1 May 2002, Guido Kollerie wrote:
 
 
 snip
 
  Unfortunately the switch is unmanaged hence I am not able to
  explicitely set the switch to 100 Mbits full-duplex. Using
  ifconfig to set the nic to 10baseT/UTP and then back to 100baseTX
  full-duplex doesn't help. Only a reboot will bring the NIC back
  to 100 Mbits full duplex mode.
 
 Please note that due to vagaries in the auto-negotiation
 spec 3com and cisco dont work well together.
 And 3coms ( on linux atleast ) have the added
 bonus of sometimes deciding to change
 speed/duplex just for the heck of it.
 
 The only way to use them reliably is to force
 both the card and the switch. We came to the
 conclusion that fxp's are a nicer option.
 
 IMHO just creating a reliable and clearly defined
 auto-negotiation protocol will do more for ethernet
 speed than gigabit ethernet :).
 
 
 -- 
 Sten Spans
 
   What does one do with ones money,
when there is no more empty rackspace ?
 
 
 To Unsubscribe: send mail to [EMAIL PROTECTED]
 with unsubscribe freebsd-current in the body of the message
 


To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message



Re: 3Com 3c905C-TX

2002-05-01 Thread Sten

On Wed, 1 May 2002, Glendon Gross wrote:


 Out of curiosity, do only 3c509's exibit this behavior, or is this
 the core problem with 3c59x's as well?  My experiences have not
 been consistent with these cards, and I had assumed it was due
 to buggy code in the 3-Com chipset.  I've noticed flaky behavior from the
 Vortex [3c59x] card as well.

I would assume is the chipset, because just out of the blue redoing
negotiation doesnt seem like something that a sane driver would do.
The most probable thing is that the card interprets normal traffic
erronously as negotiation signals.

 Just now I have been wrestling with an ISA 3c509 which has
 a Lucent 40-01304 chip on it.  At first the card was detected, and
 later not detected [on a different OS.]I vote for the fxp's as
 well, I've had hardly any problems with them.

 Is there a way to lock down the card by hacking the driver, so it won't
 try to auto-negotiate the connection?

Like I said forcing it ( with the dos config tool ) helps,
and solves the problems in most cases.
But it's pretty workable when you force both sides.

-- 
Sten Spans

  What does one do with ones money,
   when there is no more empty rackspace ?


To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message



3Com 3c905C-TX

2002-04-29 Thread Danny Braniss

sorry for the cross posting, but this is true for both, stable and current.

the onboard ethernet (the motherboard is a tyan tiger dual MPX)

xl0: 3Com 3c905C-TX Fast Etherlink XL port 0x3000-0x307f mem 
0xfa001000-0xfa00107f irq 5 at device 8.0 on pci2

at 100baseTX full-duplex is slower than 10Mgb :-(

it's not the cable/switch, because i can connect the same cable to another
host, and it works as expected - fast.

btw, the PXE is also extreamly slow - still looking for a flash upgrade.

thanks,

danny



To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message



Re: 3Com 3c905C-TX

2002-04-29 Thread Paul Murphy

On Mon, 29 Apr 2002 20:08:02 +0300
Danny Braniss [EMAIL PROTECTED] wrote:

 xl0: 3Com 3c905C-TX Fast Etherlink XL port 0x3000-0x307f mem 
 0xfa001000-0xfa00107f irq 5 at device 8.0 on pci2
 
 at 100baseTX full-duplex is slower than 10Mgb :-(
 

 I recently had a similar problem. My switch (Netgear FS105) needs the
adaptor to *announce* itself as full duplex. Try turning OFF full-duplex
in your ifconfig and see if there is a speed-up.

-- 
Cogeco ergo sum


msg37860/pgp0.pgp
Description: PGP signature