Bug 112907

Summary: SLP Installation fails with IBM iSeries ThinkPad
Product: [openSUSE] SUSE LINUX 10.0 Reporter: Joachim Werner <joe>
Component: InstallationAssignee: Christian Zoz <zoz>
Status: VERIFIED FIXED QA Contact: Klaus Kämpf <kkaempf>
Severity: Major    
Priority: P5 - None CC: b.jonsson, snwint, yast2-maintainers
Version: Beta 3   
Target Milestone: ---   
Hardware: i586   
OS: All   
Whiteboard:
Found By: Development Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---

Description Joachim Werner 2005-08-25 09:34:10 UTC 
I've got an iSeries ThinkPad that does not install via SLP: 
 
Hardware: 
- "Type 2651" ThinkPad, P III 500 MHz, 256 MB 
- "level one" 10 MBit PCMCIA Ethernet adapter 
 
How to reproduce: 
- start mini ISO 
- select "Installation" 
- SLP installation starts with loading kernel from mini ISO 
- PCMCIA device is activated (connect and traffic LEDs indicate correct 
  function) 
- Installation fails: "Could not find the SUSE Linux Installation Source. 
  Activating manual setup program." 
 
Console output (Ctrl-Alt-F3): 
 
insmod /modules/piix.ko 
Ok, that seems to have worked. :-) 
Looking for a network server... 
Ok, that didn't work; see if we can activate another network device... 
No further network cards found; giving up. 
Automatic setup not possible.
Comment 1 Arvin Schnell 2005-08-25 09:36:37 UTC 
Linuxrc problem.  Maybe your network is broken.
Comment 2 Steffen Winterfeldt 2005-08-25 09:54:06 UTC 
Try manual mode and load the pcmcia module yourself. Does it work then? 
What does 'hwinfo --pcmcia' say?
Comment 3 Joachim Werner 2005-08-25 10:02:44 UTC 
I've checked with Christian Zoz. Loading the module seems to work (no errors),  
but network does not work afterwards. We tried with a tulip card. That one  
works.  
  
So the issue is with the network card, not the laptop as a whole.  
  
I agree that it is no blocker as only some PCMCIA cards seem to be affected, 
but it's at least major because it is a regression during installation, with no 
known workaround.
Comment 4 Steffen Winterfeldt 2005-08-25 10:12:15 UTC 
Christian seems to have problems with his own test machine as well. 
For me it works.
Comment 5 Christian Zoz 2005-09-02 08:17:28 UTC 
*** Bug 114029 has been marked as a duplicate of this bug. ***
Comment 6 Christian Zoz 2005-09-02 08:21:07 UTC 
PCMCIA bridges behave differently. With some of them you have to initialize the
socket explicitely (add io and mem resources) before the card can be initialized.
But since there is another a way to detect the presence of a card, we now
initialize sockets, if there is a card. We don't want to initialize them always,
because this might also crash some machines, if static io and mem ranges in
config.opts do fit that machine.

Joeachim, please confirm, that it works with beta4. Your PC-card is at my table.
Comment 7 Christian Zoz 2005-09-02 08:22:59 UTC 
Last sentence in first paragrah: " ... do _not_ fit ..."