From nobody@FreeBSD.org  Thu Mar 16 00:06:52 2006
Return-Path: <nobody@FreeBSD.org>
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id 6A48E16A424
	for <freebsd-gnats-submit@FreeBSD.org>; Thu, 16 Mar 2006 00:06:52 +0000 (UTC)
	(envelope-from nobody@FreeBSD.org)
Received: from www.freebsd.org (www.freebsd.org [216.136.204.117])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 25C0A43D4C
	for <freebsd-gnats-submit@FreeBSD.org>; Thu, 16 Mar 2006 00:06:52 +0000 (GMT)
	(envelope-from nobody@FreeBSD.org)
Received: from www.freebsd.org (localhost [127.0.0.1])
	by www.freebsd.org (8.13.1/8.13.1) with ESMTP id k2G06pnc032947
	for <freebsd-gnats-submit@FreeBSD.org>; Thu, 16 Mar 2006 00:06:51 GMT
	(envelope-from nobody@www.freebsd.org)
Received: (from nobody@localhost)
	by www.freebsd.org (8.13.1/8.13.1/Submit) id k2G06p4t032946;
	Thu, 16 Mar 2006 00:06:51 GMT
	(envelope-from nobody)
Message-Id: <200603160006.k2G06p4t032946@www.freebsd.org>
Date: Thu, 16 Mar 2006 00:06:51 GMT
From: Davide <ifrit@anubics.org>
To: freebsd-gnats-submit@FreeBSD.org
Subject: src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25 21:57:03 bz NOT WORK
X-Send-Pr-Version: www-2.3

>Number:         94524
>Category:       kern
>Synopsis:       [nve] src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25 21:57:03 bz NOT WORK
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    bz
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Thu Mar 16 15:08:44 GMT 2006
>Closed-Date:    Fri Apr 28 08:07:14 GMT 2006
>Last-Modified:  Fri Apr 28 08:07:14 GMT 2006
>Originator:     Davide
>Release:        FreeBSD 6.1-PRERELEASE #0
>Organization:
>Environment:
FreeBSD eL33t 6.1-PRERELEASE FreeBSD 6.1-PRERELEASE #0: Mon Mar 13 21:43:50 CET 2006     root@eL33t:/usr/obj/usr/src/sys/OPOSSUM  i386

>Description:
$ dmesg
[...]
nve0: device timeout (1)
nve0: link state changed to DOWN
nve0: link state changed to UP
nve0: device timeout (1)
nve0: link state changed to DOWN
nve0: link state changed to UP
nve0: device timeout (1)
nve0: link state changed to DOWN
nve0: link state changed to UP
nve0: device timeout (2)
nve0: link state changed to DOWN
nve0: link state changed to UP
nve0: device timeout (1)
nve0: link state changed to DOWN
nve0: link state changed to UP


 src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25 21:57:03 bz NOT WORK


 cause system freeze
>How-To-Repeat:

>Fix:

>Release-Note:
>Audit-Trail:
Responsible-Changed-From-To: freebsd-bugs->bz 
Responsible-Changed-By: glebius 
Responsible-Changed-When: Thu Mar 16 15:17:32 UTC 2006 
Responsible-Changed-Why:  
Assign to author of the problematic commit. 

http://www.freebsd.org/cgi/query-pr.cgi?pr=94524 

From: "Bjoern A. Zeeb" <bz@FreeBSD.org>
To: bug-followup@FreeBSD.org, ifrit@anubics.org
Cc:  
Subject: Re: kern/94524 : src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25
 21:57:03 bz NOT WORK
Date: Thu, 16 Mar 2006 17:29:09 +0000 (UTC)

 Could you be a bit more specific about your problem:
 
 a) did it work before?
  	- if so what revision of if_nve.c was that?
 
 b) pciconf -lv ( output for the nve entry only!! )
 
 c) what board do you have?
 
 d) are you running i386 or amd64?
 
 e) from what I can see the interface seems to work before
     the system freezes. Is that correct?
 
 f) does the system freeze hard (only unpowering helps) or
     does only the interface freeze?
 
 g) does the interface just stall for some seconds and works again
     afterwards?
 
 h) or does ifconfig nve0 down && ifconfig nve0 up help?
 

From: Alexander <msx9@mail.ru>
To: bug-followup@FreeBSD.org, ifrit@anubics.org, bz@FreeBSD.org
Cc:  
Subject: Re: kern/94524: [nve] src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25 21:57:03 bz NOT WORK
Date: Mon, 20 Mar 2006 22:26:48 +0600

 This is reply to bug 94524:  (see my text after the quotes)
 
 > http://www.freebsd.org/cgi/query-pr.cgi?pr=3D94524
 >=20
 > From: "Bjoern A. Zeeb" <bz@FreeBSD.org>
 > To: bug-followup@FreeBSD.org, ifrit@anubics.org
 > Cc: =20
 > Subject: Re: kern/94524 : src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25
 >  21:57:03 bz NOT WORK
 > Date: Thu, 16 Mar 2006 17:29:09 +0000 (UTC)
 >=20
 >  Could you be a bit more specific about your problem:
 > =20
 >  a) did it work before?
 >         - if so what revision of if_nve.c was that?
 > =20
 >  b) pciconf -lv ( output for the nve entry only!! )
 > =20
 >  c) what board do you have?
 > =20
 >  d) are you running i386 or amd64?
 > =20
 >  e) from what I can see the interface seems to work before
 >      the system freezes. Is that correct?
 > =20
 >  f) does the system freeze hard (only unpowering helps) or
 >      does only the interface freeze?
 > =20
 >  g) does the interface just stall for some seconds and works again
 >      afterwards?
 > =20
 >  h) or does ifconfig nve0 down && ifconfig nve0 up help?
 
 
 I have similar system
 
 FreeBSD  6.1-PRERELEASE FreeBSD 6.1-PRERELEASE #0: Thu Mar 16 20:48:21 OMST=
  2006     root@:/usr/obj/usr/src/sys/GENERIC  i386
 
 ...and I have similar bug in nve0 NIC.
 
 My processor is Sempron 2800+ (1600-MHz 686-class CPU).
 My motherboard is ECS (EliteGroup) NFORCE3-A
 My NIC (MAC chip is NVidia NForce3 250, PHY chip is RealTek 8201):
 
  nve0: <NVIDIA nForce MCP7 Networking Adapter> port 0xb800-0xb807 mem 0xe50=
 00000-0xe5000fff irq 20 at device 5.0 on pci0
  nve0: Ethernet address 00:14:2a:7d:9e:a4
  miibus0: <MII bus> on nve0
  rlphy0: <RTL8201L 10/100 media interface> on miibus0
  rlphy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
  nve0: Ethernet address: 00:14:2a:7d:9e:a4
 
 Before cvsup (i.e. pure 6.0 RELEASE i386) NIC was work with many
 "device timeout" messages and interface DOWN/UP events. After update
 to RELENG_6 (if_nve 1.7.2.8) device timeout and down/up events occurs
 more rare.
 
 
 pciconf -lv:
 
 nve0@pci0:5:0:  class=3D0x068000 card=3D0x25011019 chip=3D0x00df10de rev=3D=
 0xa2 hdr=3D0x00
     vendor   =3D 'NVIDIA Corporation'
     device   =3D 'Network Bus Enumerator'
     class    =3D bridge
 
 
 system board: ECS (EliteGroup) NFORCE3-A, chipset is NForce3 250
 
 
 FreeBSD code: i386 (not amd64); CPU is AMD64 (Sempron)
 
 
 My interface (nve0) seems to work before the system freezes (really, I
 have no high system load and I have no long uptime -- then my system
 was not freeze)
 
 
 Right, my interface (nve0) just stall for 1 -- 1.5 (?) seconds and
 works again afterwards (if_nve 1.7.2.8).
 
 
 
 --=20
 =D1 =F3=E2=E0=E6=E5=ED=E8=E5=EC,
  Alexander                          mailto:msx9@mail.ru
 

From: "Bjoern A. Zeeb" <bz@FreeBSD.org>
To: Alexander <msx9@mail.ru>
Cc: bug-followup@FreeBSD.org, ifrit@anubics.org
Subject: Re: kern/94524: [nve] src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25
 21:57:03 bz NOT WORK
Date: Mon, 20 Mar 2006 17:48:12 +0000 (UTC)

 Just to clarify: this problem was there before my commit and is not
 related to it at all. Further the problem is well known.
 The difference is for most people seeing this before the interfaces
 hadn't worked at all and now does with the symptoms you are seeing.

From: Anish Mistry <mistry.7@osu.edu>
To: bug-followup@freebsd.org, ifrit@anubics.org
Cc:  
Subject: Re: kern/94524: [nve] src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25 21:57:03 bz NOT WORK
Date: Mon, 27 Mar 2006 02:58:35 -0500

 --nextPart23380577.derATJl388
 Content-Type: text/plain;
   charset="us-ascii"
 Content-Transfer-Encoding: quoted-printable
 Content-Disposition: inline
 
 Just adding another data point.
 =2D Branch: RELENG_6 - 6.1-PRERELEASE Fri Mar 24 22:57:43 UTC 2006
 =2D AMD64
 =2D __FBSDID("$FreeBSD: src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25=20
 21:57:03 bz Exp $");
 =2D No patches.
 =2D nve0@pci0:5:0:  class=3D0x068000 card=3D0x0c1110de chip=3D0x00df10de=20
 rev=3D0xa2 hdr=3D0x00
     vendor   =3D 'NVIDIA Corporation'
     device   =3D 'Network Bus Enumerator'
     class    =3D bridge
 =2D The interface works for a little bit of traffic (a few bytes) and=20
 then when you do something as "intensive" as visiting yahoo.com the=20
 device will display:
 nve0: device timeout (1)
 nve0: link state changed to DOWN
 nve0: link state changed to UP
 When this happens the computer freezes for a second or two.  This=20
 happens for every "intensive" burst of traffic.  So the interface=20
 works, but the freezing makes using the device problematic.
 =2D-=20
 Anish Mistry
 
 --nextPart23380577.derATJl388
 Content-Type: application/pgp-signature
 
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1.4.2.2 (FreeBSD)
 
 iD8DBQBEJ5s0xqA5ziudZT0RAjnEAKCxUIAR6NVEXG5Cc+uAGmtGowy+vQCgvYhQ
 gJqRlfGh5yvEwJ/Eys4iyrU=
 =WEyd
 -----END PGP SIGNATURE-----
 
 --nextPart23380577.derATJl388--

From: Dave Lawrence <tale@dd.org>
To: bug-followup@FreeBSD.org,ifrit@anubics.org,bz@freebsd.org
Cc:  
Subject: Re: kern/94524: [nve] src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25 21:57:03 bz NOT WORK
Date: Mon, 17 Apr 2006 12:08:10 -0400

 I was rather distressed when I migrated to my new server and put it
 into production to find it is regularly giving these errors:
 
 nve0: device timeout (1)
 nve0: link state changed to DOWN
 nve0: link state changed to UP
 nve0: device timeout (1)
 nve0: link state changed to DOWN
 nve0: link state changed to UP
 
 It has momentary freezes when this happens, though for the most part
 it is up and running and maintains its TCP connexions.  I'm still
 concerned though; sure would be nice to not have to worry on it.
 
 System is built on 6.1 beta4:
 
 FreeBSD gro.dd.org 6.1-BETA4 FreeBSD 6.1-BETA4 #0: Mon Mar 13 19:16:58 UTC 2006     root@x64.samsco.home:/usr/obj/usr/src/sys/SMP  amd64
 
 a) I didn't notice the problem when I first installed the system,
 before I put it into production.  When I first brought it up in
 production, having switched from a DHCP address to a static address,
 I had "media 100baseTX" on the interface line in rc.conf and the
 interface was constantly resetting and the machine was unusable.  I
 did a manual change to media autoselect and it got a little better,
 though some of the problem persists.
 
 b) pciconf -lv:
 
 
 nve0@pci0:10:0: class=0x068000 card=0x81411043 chip=0x005710de
 rev=0xa3 hdr=0x00
     vendor   = 'NVIDIA Corporation'
     device   = 'nForce4 Ethernet Controller'
     class    = bridge
 pcib2@pci0:11:0:        class=0x060400 card=0x00000040 chip=0x005d10de rev=0xa3 hdr=0x01
     vendor   = 'NVIDIA Corporation'
     device   = 'nForce4 PCI Express Bridge'
     class    = bridge
     subclass = PCI-PCI
 
 c) Mobo is an ASUS A8N5X.
 
 d) Running amd64 build.
 
 e) Yes, the interface is mostly working.  netstat also shows nothing
 remarkable; 0 for errs and collisions.
 
 f) Interface freeze.  Thankfully.  
 
 g) Yes, stalls then works again.
 
 h) I've not manually brought it down and back up, since it seems to
 reset itself.
 
 

From: Paul Blair <toadlife@toadlife.net>
To: bug-followup@FreeBSD.org, ifrit@anubics.org
Cc:  
Subject: Re: kern/94524: [nve] src/sys/dev/nve/if_nve.c,v 1.7.2.8 2005/12/25
 21:57:03 bz NOT WORK
Date: Sun, 23 Apr 2006 02:21:19 -0700

 I have this problem on my new workstation. The motherboard is an ASUS 
 A8N 32 SLI Deluxe, the processor an Athlon 64 X2 4400+, and I'm running 
 i386.
 
 FreeBSD  6.1-RC Sun Apr 23 00:42:00 PDT 2006
 
 nve0: <NVIDIA nForce MCP9 Networking Adapter> port 0xd880-0xd887 mem 
 0xfebfc000-0xfebfcfff irq 20 at device 19.0 on pci0
 
 The motherboard has dual gigabit lan ports. I have only the first one 
 enabled in the BIOS. The port is connected to a 8 port Netgear Gigabit 
 switch. The switch lights up green, which indicates that the media is 
 set to 1000mbps, but ifconfig says this:
 
 media: Ethernet autoselect (100baseTX <full-duplex>)
 
 I tried setting the NIC manually via ifconfig to 100baseTX, 1000baseTX, 
 and even 10baseT/UTP, full and half-duplex, but none of those settings 
 make any difference.
 
 What happens is similar to other reports. When I get the device time 
 message, the machine freezes up completely for about 1 or 2 seconds and 
 then resumes normal operation. The freeze never lasts for more than 2 
 seconds and this issue has never caused the machine to freeze 
 permanently or crash completely. When the timeout error occurs the light 
 on the switch turns off, as if the network cable is unplugged.
 
 I started with FreeBSD 6.1 BETA4 and have been tracking the release. 
 I've had this problem from the first time I installed BETA4.
 
 -Paul
State-Changed-From-To: open->closed 
State-Changed-By: bz 
State-Changed-When: Fri Apr 28 08:03:39 UTC 2006 
State-Changed-Why:  
Thanks for all the input. Trying to avoid multiple open 
PRs I am closing this one. You can find a patch in 
PR kern/96391. Please test and follow-up there. 

http://www.freebsd.org/cgi/query-pr.cgi?pr=94524 
>Unformatted:
