Thanks Yinghai.  This still applies to a different line number in
v2.6.27-rc4 than what the patch indicates.  I also get a warning
about unused variable "i" (I commented it out in my previous fix,
but not this time).

At any rate, I still produces a kernel that hangs at the typical
location.


Dave W.
--

[ message continues ]

dmesg and lspci -vvxxx with hpet=disable?

please boot with "debug initcall_debug hpet=disable"

YH
--

[ message continues ]

David,

please check v3

deviceid should be 0x4385 instead.

YH
[ message continues ]