2.6.29-rc8: boot/kernel init duration regression from 2.6.28

[CaT]

I have a Eee and just decided to try rc8 with the result that kernel
init now takes around 20s longer than it did under 2.6.28. From dmesg
it takes place here:

[    2.646852] usb 5-1: New USB device found, idVendor=0b05, idProduct=b700
[    2.646943] usb 5-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[    2.647031] usb 5-1: Product: BT-253
[    2.647107] usb 5-1: Manufacturer: Broadcom Corp
[    2.647187] usb 5-1: SerialNumber: 002243ABFC41
[    2.647513] usb 5-1: configuration #1 chosen from 1 choice
[    3.981247] input: ImPS/2 Logitech Wheel Mouse as /devices/platform/i8042/serio1/input/input7
[   27.087523] eeepc: Hotkey init flags 0x41
[   27.090800] eeepc: Get control methods supported: 0x101713
[   27.091187] input: Asus EeePC extra buttons as /devices/virtual/input/input8
[   27.178188] Advanced Linux Sound Architecture Driver Version 1.0.18a.

I have included the full dmesg from 2.6.28 and 2.6.29-rc8 (bzip2 compressed
so that it can fit under any limits).

I can compile new kernels on this freely (though slowly) so if there are any rcs to
hint at to try, or whatever, please shout. I've CCed the input and Eee maintainers
as it seems to be around there that the pause happens.

-- 
  "A search of his car uncovered pornography, a homemade sex aid, women's 
  stockings and a Jack Russell terrier."
    - http://www.news.com.au/story/0%2C27574%2C24675808-421%2C00.html

Attachment: dmesg.2.6.28.bz2
Description: Binary data

Attachment: dmesg.2.6.29-rc8.bz2
Description: Binary data