<html><head></head>
<body>Just a theory, no idea if it's related. Since about 2.3.40-pre4 I
have had problems with my cardbus controller. It's seems that the changes
to yenta_probe_irq would 1. report incorrect IRQ's and 2. render my pcmcia
network card unusable. I have an IBM thinpad 600E, which I believe has
a TI PCI1251 cardbus controller. Normally, this controller is on IRQ11,
however, the new probe routine reports about six different IRQ's. In addition,
the following line causes a "lockup" of the controller:<br> cb_writel(socket, CB_SOCKET_FORCE, CB_FCARDSTS);<br>If
I comment out this line and return a mask of cb_irq, the problem still persists
until I physically eject the card in the socket, reinsert it and reboot.
The card is the notorious ibm etherjet aka xircom cbe network adapter.
Additionally, on boot, the following is reported when the problem is encountered:<br> Socket status: 30000020<br>and the following is reported when the above fix is made and all works well:<br> Socket status: 30000920<br><br>I
don't know if any of this is related. I don't experience keyboard lockups.
However, from what I can gather, the lockups occured at approximately the
same kernel revision and are related to the yenta driver.<br><br>Please cc: me with any replies.<br><br>Tim Raymond<br></body>
</html>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/