I recently tried to upgrade to 5.2 from 5.1, and had some problems with the boot and supplimentary disks hanging when they tried to enable the PCMCIA port on my laptop. Using the 5.1 boot and supp disks, I was able to load 5.2, only to discover that the PCMCIA problems in the boot disks were also present in the Kernel. Checking the messages file when I do load the module (but before the system hangs), reveals little except the following laboriously re-typed messages (cause no PCMCIA == no network)... Linux PCMCIA Card Services 3.0.5 kernel build: 2.0.36 unknown options: [pci] [cardbus] Intel PCIC probe: Cirrus PD672x at port 0x3e0 ofs 0x00, 2 sockets host opts [0]: [ring] [65/6/0] [1/15/0] host opts [1]: [ring] [65/6/0] [1/15/0] ISA irqs (default) = 3,4,7,9,10,11 status change on irq 11 ...and then the cardmgr starts, and then the system hangs.
If necessary, I can bring the laptop in for verification, since I'm in the RTP, NC area.
This happens with the 2.0.36-3 kernel and associated modules. I haven't tried the 2.2 kernel yet.
This also happens on a NEC READY 330T. 5.1 works perfectly. Don't think this is tied to any one machine
I just now tried the 6.0 disks, and got the same thing (along with another interesting wrinkle that seems unrelated). Is there going to be any forward movement on this bug? Like I said, I can bring in the laptop for investigation, no problem. What looks like is happening is that the IO probe part of PCMCIA startup is tickling an unrelated port on the TI and causing it to hang (at least, that's what the PCMCIA howto says). The question I have, is how do I restrict PCMCIA port probing in a boot disk situation?
Scratch that, it's a different problem under 6.0. It looks like from the boot log, that the card services do not match when cardmgr starts. Does that mean that the PCMCIA system was complied incorrectly? Should I open another bug in the system to make sure this new thing is looked at?