From nobody@FreeBSD.org  Fri Mar 29 05:28:26 2002
Return-Path: <nobody@FreeBSD.org>
Received: from freefall.freebsd.org (freefall.FreeBSD.org [216.136.204.21])
	by hub.freebsd.org (Postfix) with ESMTP id 540C037B420
	for <freebsd-gnats-submit@FreeBSD.org>; Fri, 29 Mar 2002 05:28:26 -0800 (PST)
Received: (from nobody@localhost)
	by freefall.freebsd.org (8.11.6/8.11.6) id g2TDSQv24885;
	Fri, 29 Mar 2002 05:28:26 -0800 (PST)
	(envelope-from nobody)
Message-Id: <200203291328.g2TDSQv24885@freefall.freebsd.org>
Date: Fri, 29 Mar 2002 05:28:26 -0800 (PST)
From: Brian Candler <B.Candler@pobox.com>
To: freebsd-gnats-submit@FreeBSD.org
Subject: Multiport starfire card (sf/ukphy) doesn't work when full duplex selected
X-Send-Pr-Version: www-1.0

>Number:         36482
>Category:       kern
>Synopsis:       Multiport starfire card (sf/ukphy) doesn't work when full duplex selected
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    freebsd-bugs
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Fri Mar 29 05:30:01 PST 2002
>Closed-Date:    Sun Oct 03 14:39:34 GMT 2004
>Last-Modified:  Sun Oct 03 14:39:34 GMT 2004
>Originator:     Brian Candler
>Release:        4.5-RELEASE
>Organization:
>Environment:
FreeBSD xxx.xxx.xxx 4.5-RELEASE FreeBSD 4.5-RELEASE #1: Thu Feb 14 09:33:27 GMT 2002     root@:/usr/src/sys/compile/xxx  i386

>Description:
Problem machine has two quad-port Adaptec starfire cards
and is running 4.5-RELEASE.

If you try to nail a port to 100baseTX/full-duplex then it stops
working. ifconfig reports "no carrier" and "(10baseT/UTP)"

The uplink is into a Cisco Catalyst 2924 whose port is configured
as "duplex full", "speed 100", and has "span uplinkfast" set too,
so should be no problem there.

>How-To-Repeat:
# ifconfig sf3 media 100baseTX mediaopt full-duplex
# ifconfig sf3
...
        media: Ethernet 100baseTX <full-duplex> (10baseT/UTP)
        status: no carrier

It remains like this indefinitely. Obviously "no carrier" is a problem,
and strange it should be displaying "(10baseT/UTP)" even though I
have selected 100baseTX!

The dmesg output on the problem machine is:

pcib2: <DEC 21154 PCI-PCI bridge> at device 4.0 on pci0
pci1: <PCI bus> on pcib2
sf0: <Adaptec ANA-62044 10/100BaseTX> port 0xec00-0xecff mem 0xfe580000-0xfe5fff
ff irq 11 at device 4.0 on pci1
sf0: Ethernet address: 00:00:d1:ef:00:5d
miibus0: <MII bus> on sf0
ukphy0: <Generic IEEE 802.3u media interface> on miibus0
ukphy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
sf1: <Adaptec ANA-62044 10/100BaseTX> port 0xe800-0xe8ff mem 0xfe500000-0xfe57ff
ff irq 10 at device 5.0 on pci1
sf1: Ethernet address: 00:00:d1:ef:00:5e
miibus1: <MII bus> on sf1
ukphy1: <Generic IEEE 802.3u media interface> on miibus1
ukphy1:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
sf2: <Adaptec ANA-62044 10/100BaseTX> port 0xe400-0xe4ff mem 0xfe480000-0xfe4fff
ff irq 5 at device 6.0 on pci1
sf2: Ethernet address: 00:00:d1:ef:00:5f
miibus2: <MII bus> on sf2
ukphy2: <Generic IEEE 802.3u media interface> on miibus2
ukphy2:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
sf3: <Adaptec ANA-62044 10/100BaseTX> port 0xe000-0xe0ff mem 0xfe400000-0xfe47ff
ff irq 11 at device 7.0 on pci1
sf3: Ethernet address: 00:00:d1:ef:00:60
miibus3: <MII bus> on sf3
ukphy3: <Generic IEEE 802.3u media interface> on miibus3
ukphy3:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
pcib5: <DEC 21154 PCI-PCI bridge> at device 8.0 on pci0
pci2: <PCI bus> on pcib5
sf4: <Adaptec ANA-62044 10/100BaseTX> port 0xdc00-0xdcff mem 0xfe280000-0xfe2fff
ff irq 10 at device 4.0 on pci2
sf4: Ethernet address: 00:00:d1:ef:00:a1
miibus4: <MII bus> on sf4
ukphy4: <Generic IEEE 802.3u media interface> on miibus4
ukphy4:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
sf5: <Adaptec ANA-62044 10/100BaseTX> port 0xd800-0xd8ff mem 0xfe200000-0xfe27ff
ff irq 5 at device 5.0 on pci2
sf5: Ethernet address: 00:00:d1:ef:00:a2
miibus5: <MII bus> on sf5
ukphy5: <Generic IEEE 802.3u media interface> on miibus5
ukphy5:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
sf6: <Adaptec ANA-62044 10/100BaseTX> port 0xd400-0xd4ff mem 0xfe180000-0xfe1fff
ff irq 11 at device 6.0 on pci2
sf6: Ethernet address: 00:00:d1:ef:00:a3
miibus6: <MII bus> on sf6
ukphy6: <Generic IEEE 802.3u media interface> on miibus6
ukphy6:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
sf7: <Adaptec ANA-62044 10/100BaseTX> port 0xd000-0xd0ff mem 0xfe100000-0xfe17ff
ff irq 10 at device 7.0 on pci2
sf7: Ethernet address: 00:00:d1:ef:00:a4
miibus7: <MII bus> on sf7
ukphy7: <Generic IEEE 802.3u media interface> on miibus7
ukphy7:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto

>Fix:
      
>Release-Note:
>Audit-Trail:

From: Brian Candler <B.Candler@pobox.com>
To: FreeBSD-gnats-submit@FreeBSD.org
Cc:  
Subject: Re: kern/36482: Multiport starfire card (sf/ukphy) doesn't work when full duplex selected
Date: Fri, 29 Mar 2002 13:38:37 +0000

 Just for clarity, I should add that the card _does_ work when ports are in
 auto-neg mode. The port only stops working when nailed to 100baseTX/FD.

From: Robert Dal Santo <sysrds@detir.qld.gov.au>
To: freebsd-gnats-submit@FreeBSD.org
Cc: B.Candler@pobox.com
Subject: Re: kern/36482: Multiport starfire card (sf/ukphy) doesn't work when full duplex selected
Date: Tue, 15 Oct 2002 17:01:50 +1000

 	This problem still exists under 4.7 STABLE. The card does work if you 
 set it to 100baseTX but it goes off air with No Carrier still if you then try 
 to set it to 100baseTX full-duplex.
 
 	If you run the cards in 10baseT/UTP mode you get lots of TX ring full, 
 resetting messages and even the occasionaly sfN watchdog timeout.  Throughput 
 is terrible in 10Mbps mode due to all the resets. 
 
 	I'm quite willing to roll back to an earlier version of FreeBSD to get 
 this working if necessary if the card worked better under a previous release.
 
 Thanks,
 
 Robert Dal Santo
 sysrds@detir.qld.gov.au
 

From: Robert Dal Santo <sysrds@detir.qld.gov.au>
To: freebsd-gnats-submit@FreeBSD.org, B.Candler@pobox.com
Cc:  
Subject: Re: kern/36482: Multiport starfire card (sf/ukphy) doesn't work when
 full duplex selected
Date: Tue, 21 Jan 2003 17:37:39 +1000

     I upgraded a box to FreeBSD 5.0 yesterday. I now get the following 
 on boot:
 
 Jan 21 09:46:11 modium kernel: sf0: Ethernet address: 00:00:d1:ef:00:f1
 Jan 21 09:46:11 modium kernel: miibus2: <MII bus> on sf0
 Jan 21 09:46:12 modium kernel: ukphy0: <Generic IEEE 802.3u media 
 interface> on miibus2
 Jan 21 09:46:12 modium kernel: ukphy0:  10baseT, 10baseT-FDX, 100baseTX, 
 100baseTX-FDX, auto
 Jan 21 09:46:12 modium kernel: /usr/src/sys/vm/uma_core.c:1330: could 
 sleep with "sf0" locked from /usr/src/sys/pci/if_sf.c:688
 
 
 And later on I get:
 
 Jan 21 09:46:12 modium kernel: lock order reversal
 Jan 21 09:46:12 modium kernel: 1st 0xc6223380 sf1 (network driver) @ 
 /usr/src/sys/pci/if_sf.c:688
 Jan 21 09:46:12 modium kernel: 2nd 0xc05313a0 allproc (allproc) @ 
 /usr/src/sys/kern/kern_fork.c:321
 
 Hope this sheds some extra light on the issues with these cards.
 
 Robert
 

From: Tilman Linneweh <arved@FreeBSD.org>
To: freebsd-gnats-submit@FreeBSD.org, B.Candler@pobox.com
Cc: Robert Dal Santo <sysrds@detir.qld.gov.au>
Subject: Re: kern/36482: Multiport starfire card (sf/ukphy) doesn't work when full duplex selected
Date: Thu, 26 Aug 2004 16:40:58 +0200

 Robert,
 
 I think the LOR you reported was fixed before FreeBSD 5.1. Did you retry with 
 a recent version of 5.x ?
 Does the problem Brian reported occur with 5.x?
 (The driver is unchanged in RELENG_4, but several commits in HEAD may have 
 fixed it accidently.)
 

From: Brian Candler <B.Candler@pobox.com>
To: Tilman Linneweh <arved@FreeBSD.org>
Cc: freebsd-gnats-submit@FreeBSD.org,
	Robert Dal Santo <sysrds@detir.qld.gov.au>
Subject: Re: kern/36482: Multiport starfire card (sf/ukphy) doesn't work when full duplex selected
Date: Thu, 26 Aug 2004 16:15:17 +0100

 On Thu, Aug 26, 2004 at 04:40:58PM +0200, Tilman Linneweh wrote:
 > Does the problem Brian reported occur with 5.x?
 
 Aside: I've long since left the company where we had the machine with this
 card, so I'm afraid I can't test it! :-)
State-Changed-From-To: open->feedback 
State-Changed-By: arved 
State-Changed-When: Thu Aug 26 20:11:19 GMT 2004 
State-Changed-Why:  
As Brian can't test anymore, set to feedback to wait for Robert's 
report if the LOR is gone. 


http://www.freebsd.org/cgi/query-pr.cgi?pr=36482 
State-Changed-From-To: feedback->closed 
State-Changed-By: arved 
State-Changed-When: Sun Oct 3 14:38:30 GMT 2004 
State-Changed-Why:  
close (feedback timeout 1 month) 

http://www.freebsd.org/cgi/query-pr.cgi?pr=36482 
>Unformatted:
