In accordance with Murphy's law, 4.0-RELEASE seems to have broken pccard
support for me. I have a WaveLAN wireless LAN card and use the optional
ISA-to-pccard bridge. On a two week old -current, it was working just
fine, displaying the following:
pcic: polling mode
pcic: polling mode
pcic0:
: pcic0: Vadem 469 at port 0x3e0 iomem 0xd on isa0
This is a slight lie.
You might want to try this at 0x3e2. It is worth a shot.
Bingo, it works! Now I'm wondering, GENERIC says:
device pcic0 at isa? irq 10 port 0x3e0 iomem 0xd
device pcic1 at isa? irq 11 port
In message [EMAIL PROTECTED] Blaz Zupan
writes:
: : pcic0: Vadem 469 at port 0x3e0 iomem 0xd on isa0
: This is a slight lie.
: You might want to try this at 0x3e2. It is worth a shot.
:
: Bingo, it works! Now I'm wondering, GENERIC says:
:
: device pcic0 at isa? irq 10 port
The old code wouldn't even look at the second pcic entry. It will
just blindly probe at both 3e0 and 3e2 and claim that it was really at
3e0. It would try to share interupts between these two entries and
would generally not work at all on systems that had multiple pcic
cards in them (I
In message [EMAIL PROTECTED] Blaz Zupan writes:
: Notice that I don't specify an irq for pcic0, because I am really short of
: them. So pcic is run in polling mode. I also tried specifying an irq with
: exactly the same result.
irqs shouldn't matter one way or the other.
: With a 4.0-RELEASE
In message [EMAIL PROTECTED] Blaz Zupan
writes:
: I'm wondering, why use the management irq at all? Is polling in this
: situation really so bad? It really shouldn't matter much if the system
: detects the card removal right away or half a second later, does it?
For the install disk, it likely