Bug 102948

Summary: PCMCIA socket doesn't work
Product: [openSUSE] SUSE LINUX 10.0 Reporter: Andreas Schneider <asn>
Component: KernelAssignee: Christian Zoz <zoz>
Status: RESOLVED INVALID QA Contact: E-mail List <qa-bugs>
Severity: Normal    
Priority: P5 - None    
Version: Beta 2   
Target Milestone: ---   
Hardware: Other   
OS: All   
Whiteboard:
Found By: Other Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---
Bug Depends on: 99630    
Bug Blocks:    
Attachments: logging howto
script
udev boot log
hardware files output

Description Andreas Schneider 2005-08-09 08:10:05 UTC
The yenta_socket module doesn't get loaded

0000:02:04.0 CardBus bridge: Texas Instruments PCI1620 PC Card Controller (rev 01)
        Subsystem: Hewlett-Packard Company: Unknown device 006d
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR+ FastB2B-
        Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
        Latency: 64, cache line size 08
        Interrupt: pin A routed to IRQ 177
        Region 0: Memory at e0104000 (32-bit, non-prefetchable) [size=4K]
        Bus: primary=02, secondary=03, subordinate=06, sec-latency=176
        Memory window 0: 20000000-21fff000 (prefetchable)
        I/O window 0: 00003000-00003fff
        I/O window 1: 00004000-00004fff
        BridgeCtl: Parity- SERR- ISA- VGA- MAbort- >Reset+ 16bInt- PostWrite-
        16-bit legacy interface ports at 0001

0000:02:04.1 CardBus bridge: Texas Instruments PCI1620 PC Card Controller (rev 01)
        Subsystem: Hewlett-Packard Company: Unknown device 006d
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR+ FastB2B-
        Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
        Latency: 64, cache line size 08
        Interrupt: pin B routed to IRQ 185
        Region 0: Memory at e0105000 (32-bit, non-prefetchable) [size=4K]
        Bus: primary=02, secondary=07, subordinate=0a, sec-latency=176
        Memory window 0: 22000000-23fff000 (prefetchable)
        Memory window 1: e0c00000-e0fff000 (prefetchable)
        I/O window 0: 00005000-00005fff
        I/O window 1: 00006000-00006fff
        BridgeCtl: Parity- SERR- ISA- VGA- MAbort- >Reset+ 16bInt- PostWrite-
        16-bit legacy interface ports at 0001
Comment 1 Hubert Mantel 2005-08-09 15:00:22 UTC
Thanks for the extremely detailed bug report.
Comment 2 Andreas Schneider 2005-08-09 15:19:37 UTC
Do you need more information? The cardbus bridge is detected by the kernel, but
the yenta_socket module isn't loaded. If I add it manually all works just fine.
Comment 3 Christian Zoz 2005-08-10 07:51:35 UTC
No wait, currently there is no proper way of debugging device initialisation via
udev. See bug 99630
Comment 4 Kay Sievers 2005-08-10 11:22:44 UTC
Don't wait! Start! Wrap the tool in a shell, if this is really an issue...
Comment 5 Christian Zoz 2005-08-19 08:07:44 UTC
Is yenta_socket still not loaded in beta2?
Comment 6 Andreas Schneider 2005-08-19 12:25:18 UTC
The yenta_socket module still doesn't get loaded. I don't know how to provide
more information, haven't found something in the whole bunch of logs...
Comment 7 Christian Zoz 2005-08-19 12:42:27 UTC
I'll attach a hotplug logging howto and a script. With that check the pci events
at boottime. 
Comment 8 Christian Zoz 2005-08-19 12:43:07 UTC
Created attachment 46684 [details]
logging howto
Comment 9 Christian Zoz 2005-08-19 12:45:17 UTC
Created attachment 46685 [details]
script

When you boot as described in the howto with a fresh messages file, then
'show_event_log 1 200 pci' should contain averything of interest. Please attach
that output.
Comment 10 Andreas Schneider 2005-08-19 13:49:16 UTC
Created attachment 46692 [details]
udev boot log

Here is the log file
Comment 11 Christian Zoz 2005-08-24 16:33:12 UTC
The interesting events are 32 and 33, but unfortunately i don't see anything
that hwup is doing. Maybe you have some old config files in
/etc/sysconfig/hardware. Please attach output from 'cd /etc/sysconfig/hardware;
grep -H . hwcfg*'.

Also try
modprobe -v `cat /sys/bus/pci/devices/0000:02:04.0/modalias`
This should load yenta_socket.
Comment 12 Andreas Schneider 2005-08-24 16:53:26 UTC
Created attachment 47440 [details]
hardware files output

cd /etc/sysconfig/hardware; grep -H . hwcfg*
Comment 13 Christian Zoz 2005-08-24 19:49:27 UTC
That was an update, not a fresh install, yes? Please add that information next
time in a bug report. This makes it much easier.

Just remove all the hwcfg-vpid, if you didn't created them by intention. Should
be fixed then for you. Please confirm, if so.

The remaining question is, why you had configuration files for your Cardbus
bridges, that have an empty MODULE= line. On 9.3.or older the modules have been
loaded nanyway from the pcmcia startscript. I will check if that happened
somewhere else, too.
Comment 14 Andreas Schneider 2005-08-25 11:53:34 UTC
It works now. Thanks.
Comment 15 Christian Zoz 2005-08-25 11:56:40 UTC
And i checked several other 9.3 installations and found nowhere a hecfg file for
the CardBus bridge with empty variable MODULE. I guess this was an individual
problem.