Bug 235990 - Broadcom network card fails on Fedora 7 test 3 (b44)
Summary: Broadcom network card fails on Fedora 7 test 3 (b44)
Keywords:
Status: CLOSED RAWHIDE
Alias: None
Product: Fedora
Classification: Fedora
Component: kernel
Version: rawhide
Hardware: i686
OS: Linux
medium
urgent
Target Milestone: ---
Assignee: John W. Linville
QA Contact: Brian Brock
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2007-04-11 10:25 UTC by Valent Turkovic
Modified: 2007-11-30 22:12 UTC (History)
1 user (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2007-04-25 18:53:13 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
dmesg output (27.73 KB, text/plain)
2007-04-17 07:35 UTC, Valent Turkovic
no flags Details
modprobe.conf (307 bytes, text/plain)
2007-04-17 07:38 UTC, Valent Turkovic
no flags Details

Description Valent Turkovic 2007-04-11 10:25:21 UTC
Description of problem:
I install Fedora 7 test 3 and run 'ifconfig' and no eth0 device exists!
I have a broadcom network card using b44 module.
I remove and reload b44 module - than eth0 device exists for a minute or two and
I can use it - browse the net and get updates.
After no more than 2 minutes eth0 device dissapears and I can't use it anymore.
I have tested this on default kernel and also the last one that is available
today. On both kernels I see the same behavior.

Version-Release number of selected component (if applicable):


How reproducible:
Every time.


Steps to Reproduce:
1. Install Fedora 7 test 3
2. Try to use the network card
3.
  
Actual results:


Expected results:
Working network.

Additional info:

Comment 1 John W. Linville 2007-04-13 19:53:48 UTC
Could you attach the output of running 'dmesg' and also 'lspci'?  Thanks!

Comment 2 Valent Turkovic 2007-04-17 07:32:37 UTC
Ok, I found out something else.

It seams that there is a bug somewhere but not the one I described in the first
post. Sorry for that.

The problem was that now my broadcom network card, lspci:
02:0e.0 Ethernet controller: Broadcom Corporation BCM4401-B0 100Base-TX (rev 02)
is eth1 and not etho anymore!

This probably happened after first kernel upgrade, because after the install it
was eth0.

I deleted all network scripts in:
/etc/sysconfig/network-scripts/
and created new ones.

But I still have problems, after the boot I don't have working network. I remove
and reload b44 kernel module and network work from then on.

Now I have 2.6.20-1.3059.fc7 kernel.



Comment 3 Valent Turkovic 2007-04-17 07:35:04 UTC
Created attachment 152775 [details]
dmesg output

dmesg output

Comment 4 Valent Turkovic 2007-04-17 07:38:16 UTC
Created attachment 152776 [details]
modprobe.conf

modprobe config file

Comment 5 Valent Turkovic 2007-04-17 07:40:14 UTC
any ideas?

Comment 6 John W. Linville 2007-04-17 15:01:19 UTC
Your modprobe.conf file has "alias eth1 b44" and has "#alias eth0 b44" (i.e. 
the eth0 version is commented out).  Is this a change you made intentionally?

Comment 7 Valent Turkovic 2007-04-17 17:32:00 UTC
Yes I added comment for eth0.

The default version was with both "alias eth1 b44" and also "alias eth0 b44" I
didn't look at the modprobe.conf before this issue so I don't know if before
kernel upgrade there were both eth0 and eth1 tied with b44.

I hoped that atleast eth1 would work fine if I comment eth0 but that didn't help.
I'll try not to delete again all networking scripts and leave only eth0 alias in
modprobe.conf

be back after the reboot.

Comment 8 Valent Turkovic 2007-04-18 08:31:13 UTC
I updated to another kernel - 2.6.20-1.3079.fc7

THis is not a kernel bug. Because everything works fine now. My card is eth0 and
works after the reboot with no problem.

Maybe it is a system-config-network bug?

This is a snow-stopper for new linux users so should be escalated to a high
level - and fixed swiftly.

Comment 9 Dave Jones 2007-04-23 16:02:30 UTC
the eth0/eth1 switching was caused by an anaconda bug which should now be fixed.

do a reinstall when test4 comes out, it should be fixed.
(or you could try a rawhide install before then).

Comment 10 Valent Turkovic 2007-04-27 13:51:50 UTC
Ok, I will. thanks.


Note You need to log in before you can comment on or make changes to this bug.