LSI 1030 mpt doesn't work if I build a new kernel

[Michael Sinatra]

Matthew Jacob wrote:
> No- thank *you*. I specifically want to thank you for helping me narrow 
> it down.
> 
> On 9/5/06, Rod Person <rodperson at adelphia.net> wrote:
>> On Tue, 2006-09-05 at 17:03 -0700, Matthew Jacob wrote:
>> > The attached patch was integrated into -current- this should fix this
>> > issues for U320 chipsets. Let me know if it doesn't.
>>
>> Just cvsup'd and rebuilt kernel. Everything is working!
>> Great work Matt, Thanks!!!!

Sorry, I think this is the same problem I was having with the Tyan S2895 
and its LSI 1030 controller (see earlier thread "7-CURRENT 8/15/06 
hangs: Tyan s2895 LSI 1030 mpt").  I spent a lot of time trying to get 
the issues narrowed down, and was able to get the thing booting either 
by disabling the onboard LSI 1030 and putting in an add-on 1030-R 
(RAID/Mirroring capable) card or by cvsing back to prior to July 12. 
Unfortunately, I got caught up with other stuff over the weekend and 
didn't send the email.  I wish I could have been more help.  Anyway, it 
looks like Matthew has committed the patch he posted earlier and I am 
now able to boot a new kernel just fine on the S2895 with the onboard 
controller.

The really awesome news is that this kernel appears to work just fine 
with an out-of-the-box S2895 with SCSI disks; it requires none of the 
BIOS gyrations that 6-STABLE still does.  Moreover the nfe driver lets 
me push almost line rate gigabit through the onboard interfaces.

Great work all around!  I wish I could have been more help.

michael