Comment 4 for bug 550382

I checked again. Here are my findings:

Out of 5 attempts lucid booted 2 times successfully with 2.6.32-17 and froze 3 times with that version.
Lucid booted 8 times successfully with 2.6.32-16 (2.6.32-16-generic #25-Ubuntu SMP Tue Mar 9 16:33:52 UTC 2010 i686 GNU/Linux) and no freezes could be observed

I'm pretty sure that the regression was introduced with -17.

I'll attach the dmesg for -16 and -17 and the diff (after timing output has been removed).

The one difference I could find is that -16 prints something like
 scsi0 : ahci
 scsi1 : ahci
 scsi2 : ahci
 ata1: SATA max UDMA/133 abar m2048@0xec009000 port 0xec009100 irq 31
 ata2: SATA max UDMA/133 abar m2048@0xec009000 port 0xec009180 irq 31
 ata3: SATA max UDMA/133 abar m2048@0xec009000 port 0xec009200 irq 31

and -17 prints
  scsi0 : ata_piix
 scsi1 : ata_piix
 ata1: PATA max UDMA/100 cmd 0x1f0 ctl 0x3f6 bmdma 0x70a0 irq 14
 ata2: PATA max UDMA/100 cmd 0x170 ctl 0x376 bmdma 0x70a8 irq 15

Since the ata_piix message has as far as I remember always been the last line I suspect that the issue could have to do with that output.

What can I check next?