Driver does not work...
Nelson Sampaio Araujo Junior
nsjunior@nutecnet.com.br
Fri Oct 16 09:12:50 1998
This is a cryptographically signed message in MIME format.
--------------msECD4026917008D2AE3AF1A91
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Hello ALL,
I'm have serious trouble with Lite-On boards on a single system. They
are 10/100Mb boards and come with drivers only for Nt/95/SCO (as usual)
:(. I've put Mr. Becker's tulip driver and detected the boards. I can
sometimes send/receive by the boards, but they are instable (works only
when they want).
I'm sending you the system log and debug information to show the
problems are ocurring.
With the current Donald Becker's tulip driver (0.89K), the boards are
detected at boot time. But the CSR5 and CSR6 registers changes their
value with certain pattern and the results are: work a little, stop a
little.
I've tried to fix the full_duplex setting the board. That corrected the
server that have 2 boards. But the server that has 3 of them, share the
same interrupt, and only one board work.
verde (means green in portuguese) is the server with 2 boards. preto
(means black) is the server with 3 boards...
Here follow the logs:
--- messages ---
Oct 15 13:17:35 verde kernel: tulip.c:v0.89Ka 8/8/98
becker@cesdis.gsfc.nasa.gov
Oct 15 13:17:35 verde kernel: eth0: Lite-On 82c168 PNIC at 0xd000, 00 a0
ff ff 19 f1, IRQ 9.
Oct 15 13:17:35 verde kernel: eth0: Checking for MII transceivers...
Oct 15 13:17:35 verde kernel: eth0: MII transceiver found at MDIO
address 1, config 3100 status 7829.
Oct 15 13:17:35 verde kernel: eth1: Lite-On 82c168 PNIC at 0xb800, 00 a0
cc 20 04 2c, IRQ 12.
Oct 15 13:17:35 verde kernel: eth1: Checking for MII transceivers...
Oct 15 13:17:35 verde kernel: eth1: MII transceiver found at MDIO
address 1, config 3100 status 7829.
Oct 15 13:18:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0000.
Oct 15 13:19:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
Oct 15 13:20:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0000.
Oct 15 13:22:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
Oct 15 13:23:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0000.
Oct 15 13:32:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
Oct 15 13:33:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0000.
Oct 15 13:34:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
Oct 15 13:35:39 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0000.
--- debug ---
Oct 15 12:42:47 verde kernel: eth0: Advertising 01e1 on PHY 1,
previously advertising 3100.
Oct 15 12:52:15 verde kernel: eth0: Advertising 01e1 on PHY 1,
previously advertising 0000.
Oct 15 12:52:15 verde kernel: eth1: Advertising 01e1 on PHY 1,
previously advertising 3100.
Oct 15 13:17:35 verde kernel: eth0: Advertising 01e1 on PHY 1,
previously advertising 01e1.
Oct 15 13:17:35 verde kernel: eth1: Advertising 01e1 on PHY 1,
previously advertising 01e1.
When the CSR6 is ...0200 the board works, otherwise not.
--- messages ---
Oct 15 13:40:14 verde kernel: tulip.c:v0.89I 7/26/98
becker@cesdis.gsfc.nasa.gov
Oct 15 13:40:14 verde kernel: eth0: Lite-On 82c168 PNIC at 0xd000, 00 a0
ff ff 19 f1, IRQ 9.
Oct 15 13:40:14 verde kernel: eth0: MII transceiver found at MDIO
address 1, config 0000 status 7829.
Oct 15 13:40:14 verde kernel: eth1: Lite-On 82c168 PNIC at 0xb800, 00 a0
cc 20 04 2c, IRQ 12.
Oct 15 13:40:14 verde kernel: eth1: MII transceiver found at MDIO
address 1, config 0000 status 7829.
...
Oct 15 13:40:17 verde kernel: eth0: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
Oct 15 13:40:17 verde kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
--- debug ---
Oct 15 13:40:14 verde kernel: eth0: Advertising 01e1 on PHY 1,
previously advertising ffff.
Oct 15 13:40:14 verde kernel: eth1: Advertising 01e1 on PHY 1,
previously advertising ffff.
The above had worked when I compiled with the #define TUPIP_FULL_DUPLEX
(-DTUPIP_FULL_DUPLEX).
That made the verde server to work well. The 2 board started working
correctly. So, why am I bothering you with these ? That's because I've
made the same thing on the black server, that the hardware is the same,
except it has one more board attached. The logs of black server:
============ PRETO ===============
--- messages ---
Oct 14 15:06:12 gw kernel: tulip.c:v0.89I 7/26/98
becker@cesdis.gsfc.nasa.gov
Oct 14 15:06:12 gw kernel: eth1: Lite-On 82c168 PNIC at 0xd000, 00 a0 cc
20 1c 47, IRQ 15.
Oct 14 15:06:12 gw kernel: eth1: MII transceiver found at MDIO address
1, config 3100 status 782d.
Oct 14 15:06:12 gw kernel: eth2: Lite-On 82c168 PNIC at 0xb800, 00 a0 cc
20 48 56, IRQ 12.
Oct 14 15:06:12 gw kernel: eth2: MII transceiver found at MDIO address
1, config 820f status 8219.
Oct 14 15:06:12 gw kernel: eth2: MII transceiver found at MDIO address
2, config 0000 status 841f.
Oct 14 15:06:12 gw kernel: eth2: MII transceiver found at MDIO address
4, config 0000 status 881f.
Oct 14 15:06:12 gw kernel: eth2: MII transceiver found at MDIO address
5, config 0000 status 8a1f.
Oct 14 15:06:12 gw kernel: eth3: Lite-On 82c168 PNIC at 0xb000, 00 a0 cc
20 48 a1, IRQ 15.
Oct 14 15:06:12 gw kernel: eth3: MII transceiver found at MDIO address
1, config 0000 status 782d.
...
Oct 14 15:06:23 gw kernel: eth1: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
Oct 14 15:06:23 gw kernel: eth2: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
Oct 14 15:06:23 gw kernel: eth3: Changing PNIC configuration to
full-duplex, CSR6 816e0200.
--- debug ---
Oct 14 14:08:05 gw kernel: eth1: Advertising 01e1 on PHY 1, previously
advertising ffff.
Oct 14 14:08:05 gw kernel: eth3: Advertising 01e1 on PHY 1, previously
advertising 0000.
Oct 14 14:11:14 gw kernel: eth1: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Oct 14 14:12:06 gw kernel: eth1: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Oct 14 14:12:06 gw kernel: eth2: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Oct 14 14:12:06 gw kernel: eth3: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Oct 14 14:12:55 gw kernel: eth1: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Oct 14 14:12:55 gw kernel: eth2: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Oct 14 14:12:55 gw kernel: eth3: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Oct 14 14:15:48 gw kernel: eth1: Advertising 01e1 on PHY 1, previously
advertising 0000.
Oct 14 14:15:48 gw kernel: eth3: Advertising 01e1 on PHY 1, previously
advertising 3100.
Oct 14 15:04:24 gw kernel: eth2: Advertising 01e1 on PHY 1, previously
advertising 0000.
Oct 14 15:04:24 gw kernel: eth4: Advertising 01e1 on PHY 1, previously
advertising 3100.
Oct 14 15:06:12 gw kernel: eth1: Advertising 01e1 on PHY 1, previously
advertising 0000.
Oct 14 15:06:12 gw kernel: eth2: Advertising 0201 on PHY 1, previously
advertising 0000.
Oct 14 15:06:12 gw kernel: eth2: Advertising 0201 on PHY 2, previously
advertising 844f.
Oct 14 15:06:12 gw kernel: eth2: Advertising 0221 on PHY 4, previously
advertising 884f.
Oct 14 15:06:12 gw kernel: eth2: Advertising 0221 on PHY 5, previously
advertising 8a4f.
Oct 14 15:06:12 gw kernel: eth3: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Oct 14 15:11:59 gw kernel: eth1: Advertising 01e1 on PHY 1, previously
advertising 3100.
Oct 14 15:11:59 gw kernel: eth2: Advertising 01e1 on PHY 1, previously
advertising ffff.
Oct 14 15:11:59 gw kernel: eth3: Advertising 01e1 on PHY 1, previously
advertising 01e1.
Only the eth1 board works. They are all conected on a 3Com SuperStack
Hub and a 3Com. The Hub is a 10Mb one. The switch has 2 100Mb ports and
8 10Mb ports. Is there any problem with this ?
Thanks for any help,
--
Nelson Sampaio Araujo Junior
===================================================================
Modulo Informatica - Analista de Desenvolvimento
Aluno de Mestrado - COPPE Sistemas cos.ufrj.br
-------------------------------------------------------------------
Unix System Administrator nce.ufrj.br
NetWare System Administrator lci.ufrj.br
Windows NT System Administrator cetem.gov.br
contact: Pager: Teletrim: +55 (021) 546-1636 [Code: 7005543]
===================================================================
Nelson Sampaio Araujo Junior wrote:
>
> Hello,
>
> The driver you sent me enclosed on your previous e-mail does not work.
>
> This is the tulip driver, written by Donald Becker of CESDIS. I've
> already contacted him on the past, and he gave me some guidelines to
> solve my problem.
>
> Your boards are detected by the driver. That's ok. But I have 3 boards
> on each server, and this configuration does not work. On the primary
> server (linux slakware 3.5 [kernel 2.0.34]), the first board is assigned
> to the same interrupt of the third. The board configuration, and the
> debug messages I'll send you when I arrive at work.
>
> On the second system, the boards worked a while, and then stopped
> forever. In this case, each board have a private IRQ line for them, and
> different IO address too.
>
> I'll be e-mailing you again by afternoon, with more diagnostics
> messages.
>
> Question: Does the board has a MII transceiver ???
>
> Thanks for the attention,
--------------msECD4026917008D2AE3AF1A91
Content-Type: application/x-pkcs7-signature; name="smime.p7s"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="smime.p7s"
Content-Description: S/MIME Cryptographic Signature
MIIFeQYJKoZIhvcNAQcCoIIFajCCBWYCAQExCzAJBgUrDgMCGgUAMAsGCSqGSIb3DQEHAaCC
A2kwggNlMIICzqADAgECAggZmAAAAjEAATANBgkqhkiG9w0BAQQFADByMQswCQYDVQQGEwJC
UjEuMCwGA1UEChMlQ2VydGlzaWduIENlcnRpZmljYWRvcmEgRGlnaXRhbCBMdGRhLjEzMDEG
A1UECxMqQ2VydGlzaWduIC0gQXV0b3JpZGFkZSBDZXJ0aWZpY2Fkb3JhIC0gQUMxMB4XDTk4
MTAwMTAwMDAwMVoXDTk5MTAwMTIzNTk1OVowgawxCzAJBgNVBAYTAkJSMS0wKwYDVQQKEyRD
ZXJ0aXNpZ24gQ2VydGlmaWNhZG9yYSBEaWdpdGFsIEx0ZGExHjAcBgNVBAsTFUNlcnRpZmlj
YWRvIGRlIEUtTWFpbDElMCMGA1UEAxMcTkVMU09OIFNBTVBBSU8gQVJBVUpPIEpVTklPUjEn
MCUGCSqGSIb3DQEJARYYbnNqdW5pb3JAbnV0ZWNuZXQuY29tLmJyMIGfMA0GCSqGSIb3DQEB
AQUAA4GNADCBiQKBgQCsCOXHBXGOsJMekcuwUZfqP+2pclllXThAYQYTU0m6Uf6niWH6w6uE
KuSnKVhtW/Yf/0fSRZecm6XwPUvZCjOpWuZhxPADKgsq4ekem5gIOxFd7Z6FoBdpY407ReCI
vHcfkGZTL4RoAzQMyWTU47M48HkSHfrn0Tm0qxVE1f9grwIDAQABo4HIMIHFMCUGA1UdEQQe
MByhGhMYbnNqdW5pb3JAbnV0ZWNuZXQuY29tLmJyMAkGA1UdEwQCMAAwEQYJYIZIAYb4QgEB
BAQDAgCAMCwGCWCGSAGG+EIBAgQfFh1odHRwczovL3d3dy5jZXJ0aXNpZ24uY29tLmJyLzAk
BglghkgBhvhCAQMEFxYVY2hlY2stdXNlcmNlcnQucGh0bWw/MCoGCWCGSAGG+EIBCAQdFhtp
bmZvL0NlcnRpU2lnbkNBUG9saWN5Lmh0bWwwDQYJKoZIhvcNAQEEBQADgYEALfTx6bgQVFrw
FbgE5q1uKhk/Sj0ifoTxK2cLSOi/Ajkjc8qSl2KsJtXQD8JuUoljl9nwGLXq8Lpgw8mRg/yY
aT9JCAnqNBvsx35J5zceCNSSrB04xRldXAHI3taiHgSS8hgxx1X7zrjAy1RRfOJaFzq7Ow1O
jwmQIxLa8M/9L7gxggHYMIIB1AIBATB+MHIxCzAJBgNVBAYTAkJSMS4wLAYDVQQKEyVDZXJ0
aXNpZ24gQ2VydGlmaWNhZG9yYSBEaWdpdGFsIEx0ZGEuMTMwMQYDVQQLEypDZXJ0aXNpZ24g
LSBBdXRvcmlkYWRlIENlcnRpZmljYWRvcmEgLSBBQzECCBmYAAACMQABMAkGBSsOAwIaBQCg
gbEwGAYJKoZIhvcNAQkDMQsGCSqGSIb3DQEHATAcBgkqhkiG9w0BCQUxDxcNOTgxMDE2MTQx
MjEzWjAjBgkqhkiG9w0BCQQxFgQU3xuOJgj242+49HMdimWcjuRRz0cwUgYJKoZIhvcNAQkP
MUUwQzAKBggqhkiG9w0DBzAOBggqhkiG9w0DAgICAIAwBwYFKw4DAgcwDQYIKoZIhvcNAwIC
AUAwDQYIKoZIhvcNAwICASgwDQYJKoZIhvcNAQEBBQAEgYCNy7yTtlikEa68s3xXyZMskeb1
QCeHsqkHC812NIUqOSyxcsdIjgvCoulrtK3TZWNlCJsA82dJvjUJy8tbSzbcrQkj/WSgArJf
a4axpY5atzL0gMTc1vtDHxvnCe3WgV5N15nZeVyuBpJt5L5B2aNG/BUdPTTgCdy3L2qFpKuJ
KQ==
--------------msECD4026917008D2AE3AF1A91--