[3com905b freeze Alpha SMP 2.4.2] FullDuplex issue ?

[3com905b freeze Alpha SMP 2.4.2] FullDuplex issue ?

[Cabaniols, Sebastien]

[-- Attachment #1: Type: text/plain, Size: 1009 bytes --]

Hello,

********************************************************************
my hardware configuration is:

2 Alphaserver ES40 running kernel 2.4.2smp
   with 3com905b FastEthernet PCI 


The configuration is switched, 100 Full Duplex autonegotiation.
********************************************************************


I insert the 3c59x module with debug=7.


I start an ftp transfer from machine A to B:

	machine A			machine B	

	ftp B				ftpd answering

	get bigfile		

bigfile is 500 Megabytes, and transfer fine at 11 MegaBytes/s perfectly


Now I want to stress/test fullduplex:

	machine A			machine B
	
	ftp B				ftp A
	ftpd answers to B		ftpd answer to A
	get bigFile			get bigFile2

The first of the above machines launching the get freezes.	

I include a file with my logs, the output of the crashed machine after
reboot.
I also include the output of uname -a, lspci -vx, vortex --aaee.

Thanks for any help, If some information is missing there do not hesitate to
ask.


Sebastien Cabaniols


[-- Attachment #2: log3Com.txt --]
[-- Type: text/plain, Size: 5351 bytes --]



uname-a

Linux es40-06 2.4.2smp #1 SMP Wed May 2 13:58:01 EDT 2001 alpha unknown
 



kernel log with debug=7 passed to modprobe




May  2 16:45:52 es40-06 kernel: 3c59x.c:LK1.1.12 06 Jan 2000  Donald Becker and others. http://www.scyld.com/network/vortex.html $Revision: 1.102.2.46 $
May  2 16:45:52 es40-06 kernel: See Documentation/networking/vortex.txt
May  2 16:45:52 es40-06 kernel: eth0: 3Com PCI 3c905B Cyclone 100baseTx at 0x8400,  00:01:02:d9:94:f0, IRQ 32
May  2 16:45:52 es40-06 kernel:   product code 'CG' rev 00.12 date 09-22-00
May  2 16:45:52 es40-06 kernel:   8K byte-wide RAM 5:3 Rx:Tx split, autoselect/Autonegotiate interface.
May  2 16:45:52 es40-06 kernel:   MII transceiver found at address 24, status 786d.
May  2 16:45:52 es40-06 kernel: 3c59x: Wake-on-LAN functions disabled
May  2 16:45:52 es40-06 kernel:   Enabling bus-master transmits and whole-frame receives.
May  2 16:46:07 es40-06 sysctl: error: 'net.ipv4.ip_always_defrag' is an unknown key
May  2 16:46:07 es40-06 sysctl: net.ipv4.ip_forward = 0
May  2 16:46:07 es40-06 sysctl: net.ipv4.conf.all.rp_filter = 1
May  2 16:46:07 es40-06 sysctl: kernel.sysrq = 0
May  2 16:46:07 es40-06 network: Setting network parameters:  succeeded
May  2 16:46:07 es40-06 ifup: SIOCADDRT: Network is unreachable
May  2 16:46:07 es40-06 network: Bringing up interface lo:  succeeded
May  2 16:46:07 es40-06 kernel: eth0: using NWAY autonegotiation
May  2 16:46:07 es40-06 kernel: eth0: MII #24 status 786d, link partner capability 41e1, setting full-duplex.
May  2 16:46:07 es40-06 network: Bringing up interface eth0:  succeeded
May  2 16:47:02 es40-06 ftpd[1225]: FTP LOGIN FROM es40-05.idris.domain [10.1.1.5], toto
May  2 16:47:29 es40-06 kernel: <74 ticks.
May  2 16:47:29 es40-06 kernel: <7tart_xmit()
May  2 16:47:29 es40-06 kernel: < e401.
May  2 16:47:29 es40-06 kernel: <77>eth0: exiting interrupt, status e000.
May  2 16:47:29 es40-06 kernel: t size 66 status 60008042.
May  2 16:47:29 es40-06 kernel: <nterrupt. status=0xe401
May  2 16:47:29 es40-06 kernel: <t()
May  2 16:47:29 es40-06 kernel: <<7>boomerang_rx(): status e001
May  2 16:47:29 es40-06 kernel:  interrupt loop, status e401.
May  2 16:47:29 es40-06 kernel: <7errupt, status e201, latency 3 ticks.
May  2 16:47:29 es40-06 kernel: <_xmit()
May  2 16:47:29 es40-06 kernel: <status e401.
May  2 16:47:29 es40-06 kernel: <rt_xmit()


next line is syslog restart after the freeze of the machine

lspci -vx

00:03.0 Ethernet controller: 3Com Corporation 3c905B 100BaseTX [Cyclone] (rev 30)
        Subsystem: 3Com Corporation 3C905B Fast Etherlink XL 10/100
        Flags: bus master, medium devsel, latency 32, IRQ 32
        I/O ports at 8400 [size=128]
        Memory at 000000000a0c3000 (32-bit, non-prefetchable) [size=128]
        Expansion ROM at 000000000a0a0000 [disabled] [size=128K]
        Capabilities: [dc] Power Management version 1
00: b7 10 55 90 07 00 10 02 30 00 00 02 10 20 00 00
10: 01 84 00 00 00 30 0c 0a 00 00 00 00 00 00 00 00
20: 00 00 00 00 00 00 00 00 00 00 00 00 b7 10 55 90
30: 00 00 0a 0a dc 00 00 00 00 00 00 00 20 01 0a 0a
vortex --aaee

 ./vortex -aaee
vortex-diag.c:v2.04 1/8/2001 Donald Becker (becker@scyld.com)
 http://www.scyld.com/diag/index.html
Index #1: Found a 3c905B Cyclone 100baseTx adapter at 0x8400.
The Vortex chip may be active, so FIFO registers will not be read.
To see all register values use the '-f' flag.
Initial window 7, registers values by window:
  Window 0: 0000 0000 0000 0000 f5f5 00bf 0000 0000.
  Window 1: FIFO FIFO 0000 0000 0000 0000 0000 2000.
  Window 2: 0100 d902 1a93 0000 0000 0000 000a 4000.
  Window 3: 0000 0180 05ea 0020 000a 0800 0800 6000.
  Window 4: 0000 0000 0000 0cd8 0003 8880 0000 8000.
  Window 5: 1ffc 0000 0000 0600 0807 06ce 06c6 a000.
  Window 6: 0000 0000 0000 0900 1000 0252 0a87 c000.
  Window 7: 0000 0000 0000 0000 0000 0000 0000 e000.
Vortex chip registers at 0x8400
  0x8410: **FIFO** 00000000 00000011 *STATUS*
  0x8420: 00000020 00000000 00080000 00000004
  0x8430: 00000000 234ddcb3 c1d88090 00080004
 Indication enable is 06c6, interrupt enable is 06ce.
 No interrupt sources are pending.
 Transceiver/media interfaces available:  100baseTx 10baseT.
Transceiver type in use:  Autonegotiate.
 MAC settings: full-duplex.
 Station address set to 00:01:02:d9:93:1a.
 Configuration options 000a.
EEPROM contents (64 words, offset 0):
 0x000: 0001 02d9 931a 9055 0135 0048 4743 6d50
 0x008: 2971 0000 0001 02d9 931a 0010 0000 0022
 0x010: 32a2 0000 0000 0180 0000 0000 0000 10b7
 0x018: 9055 000a 0000 0000 0000 0000 0000 0000
 0x020: 0093 0000 0000 0000 0000 0000 0000 0000
 0x028: 0000 0000 0000 0000 0000 0000 0000 0000
 0x030: 0000 0000 0000 0000 0000 0000 0000 0000
 0x038: 0000 0000 0000 0000 0000 0000 0000 0000
 The word-wide EEPROM checksum is 0x71bb.
Parsing the EEPROM of a 3Com Vortex/Boomerang:
 3Com Node Address 00:01:02:D9:93:1A (used as a unique ID only).
 OEM Station address 00:01:02:D9:93:1A (used as the ethernet address).
 Manufacture date (MM/DD/YYYY) 9/21/2000, division H, product CG.
Options: none.
  Vortex format checksum is incorrect (00fb vs. 10b7).
  Cyclone format checksum is correct (0x93 vs. 0x93).
  Hurricane format checksum is correct (0x93 vs. 0x93).

Re: [3com905b freeze Alpha SMP 2.4.2] FullDuplex issue ?

[Bogdan Costescu]

On Wed, 2 May 2001, Cabaniols, Sebastien wrote:

> I insert the 3c59x module with debug=7.

Why ? debug=7 is the highest debug level and produces _lots_ of debug data
for high network activity. Do you have problems when insmod-ing without
any option and use a higher debug level just to see what's going on?

> The first of the above machines launching the get freezes.

Why do you believe that the card/driver is responsible for the freeze ?
The outputs that you provided show no problems to me.

A duplex mismatch would not freeze a computer. You would get crappy
transfer rates, usually some error messages from the driver, but
everything should otherwise work. To verify the media settings, you might
want to use mii-diag (from ftp.scyld.com).

Sincerely,

Bogdan Costescu

IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
E-mail: Bogdan.Costescu@IWR.Uni-Heidelberg.De

Re: [3com905b freeze Alpha SMP 2.4.2] FullDuplex issue ?

[Andrew Morton]

"Cabaniols, Sebastien" wrote:
> 
> [ SMP Alpha dies running TCP Rx+Tx ]
>

Sebastien, I'd be suspecting the 2.4.2/Alpha kernel
more than the ethernet driver.  Are you able to reproduce
this with more recent kernels?

-

RE: [3com905b freeze Alpha SMP 2.4.2] FullDuplex issue ?

[Cabaniols, Sebastien]

> -----Original Message-----
> From: Bogdan Costescu [mailto:bogdan.costescu@iwr.uni-heidelberg.de]
> Sent: mercredi 2 mai 2001 17:14
> To: Cabaniols, Sebastien
> Cc: 'andrewm@uow.edu.au'; 'netdev@oss.sgi.com';
> 'linux-kernel@vger.kernel.org'
> Subject: Re: [3com905b freeze Alpha SMP 2.4.2] FullDuplex issue ?
> 
> 
> On Wed, 2 May 2001, Cabaniols, Sebastien wrote:
> 
> > I insert the 3c59x module with debug=7.
> 
> Why ? debug=7 is the highest debug level and produces _lots_ 
> of debug data
> for high network activity. Do you have problems when 
> insmod-ing without
> any option and use a higher debug level just to see what's going on?

It is only to produce verbose logs. (See attached file in original email)


> 
> > The first of the above machines launching the get freezes.
> 
> Why do you believe that the card/driver is responsible for 
> the freeze ?

Excellent question: I don't know, should I suspect the wu-ftpd daemon ?
Since the machine is crashed I guess I can suspect some code doing
system stuff, drivers/kernel or daemons ?


> The outputs that you provided show no problems to me.
> 
> A duplex mismatch would not freeze a computer. You would get crappy
> transfer rates, usually some error messages from the driver, but
> everything should otherwise work. To verify the media 
> settings, you might
> want to use mii-diag (from ftp.scyld.com).

Ok, let me see how we compile this and send you the output.


> 
> Sincerely,
> 
> Bogdan Costescu
> 
> IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
> Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
> Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
> E-mail: Bogdan.Costescu@IWR.Uni-Heidelberg.De
> 
>