Re: ACPI call for testing (take two)

[Piotr Meyer <aniou%smutek.pl@localhost>]

On Tue, Jan 11, 2011 at 05:10:55AM +0200, Jukka Ruohonen wrote:
 
> So it indeed seems to cause a storm on pin 9, which is most likely the SCI.
> Can you try booting with options -x -v (see boothowto(8)) and posting the
> dmesg, so that we can see whether it is doing something stupid like trying
> to route the SCI as a PCI interrupt.

I tested Your kernel on my old Dell2950 and got same behaviour:

# vmstat  -i                                                                    
                                         
interrupt                                     total     rate
cpu0 timer                                    20060      100
global TLB IPI                                 1452        7
ioapic0 pin 9                               5915264    29724
ioapic1 pin 15                                 1948        9
ioapic1 pin 0                                  1179        5
ioapic0 pin 14                                    6        0
Total                                       5939909    29848

# uptime
 9:19AM  up 3 mins, 1 user, load averages: 0.01, 0.02, 0.00


My previous kernel, after three days reported:

# cat irq_old.txt                                                               
                                         
interrupt                                     total     rate
cpu0 timer                                 22219752       99
global TLB IPI                              3440778       15
ioapic1 pin 15                              1763169        7
ioapic1 pin 0                               4237799       19
ioapic0 pin 14                                    6        0
Total                                      31661504      142


Current dmesg (with -x and -v) is located at:
http://smutek.pl/netbsd/dmesg-dell2950-acpi-patch.txt

-- 
Piotr 'aniou' Meyer