2.6.22.5 SATA Failure

I have been using 2.6.21.1. It seems working well, that is, all my
disk partitions are mapped as "/dev/sda*" and the performance looks
good. After I upgrade to 2.6.22.5 with the exact same configuration,
all the disk device turn to "/dev/hda*" and the performance degrade
obviously.

While I boot with 2.6.21.1, the logs regarding disk are:

Aug 26 06:11:26 localhost kernel: ata_piix 0000:00:1f.2: MAP [ P0 P2 IDE IDE ]
Aug 26 06:11:26 localhost kernel: ACPI: PCI Interrupt 0000:00:1f.2[B]
-> GSI 17 (level, low) -> IRQ 17
Aug 26 06:11:26 localhost kernel: ata: 0x170 IDE port busy
Aug 26 06:11:26 localhost kernel: ata: conflict with ide1
Aug 26 06:11:26 localhost kernel: ata1: SATA max UDMA/133 cmd
0x000101f0 ctl 0x000103f6 bmdma 0x0001bfa0 irq 14
Aug 26 06:11:26 localhost kernel: ata2: DUMMY
Aug 26 06:11:26 localhost kernel: scsi0 : ata_piix
Aug 26 06:11:26 localhost kernel: ata1.00: ATA-7: ST9120821AS, 8.03,
max UDMA/133
Aug 26 06:11:26 localhost kernel: ata1.00: 234441648 sectors, multi 8:
LBA48 NCQ (depth 0/32)
Aug 26 06:11:26 localhost kernel: ata1.00: configured for UDMA/133
Aug 26 06:11:26 localhost kernel: scsi1 : ata_piix
Aug 26 06:11:26 localhost kernel: scsi 0:0:0:0: Direct-Access     ATA
    ST9120821AS      8.03 PQ: 0 ANSI: 5
Aug 26 06:11:26 localhost kernel: SCSI device sda: 234441648 512-byte
hdwr sectors (120034 MB)
Aug 26 06:11:26 localhost kernel: sda: Write Protect is off
Aug 26 06:11:26 localhost kernel: SCSI device sda: write cache:
enabled, read cache: enabled, doesn't support DPO or FUA
Aug 26 06:11:26 localhost kernel: SCSI device sda: 234441648 512-byte
hdwr sectors (120034 MB)
Aug 26 06:11:26 localhost kernel: sda: Write Protect is off
Aug 26 06:11:26 localhost kernel: SCSI device sda: write cache:
enabled, read cache: enabled, doesn't support DPO or FUA
Aug 26 06:11:26 localhost kernel:  sda: sda1 sda2 sda3 < sda5 sda6 sda7 >



While booting with 2.6.22.5, the logs are:

Aug 26 09:30:04 localhost kernel: ACPI: PCI Interrupt 0000:00:1f.2[B]
-> GSI 17 (level, low) -> IRQ 17
Aug 26 09:30:04 localhost kernel: ata_piix 0000:00:1f.2: 0x1F0 IDE port busy
Aug 26 09:30:04 localhost kernel: ata_piix 0000:00:1f.2: 0x170 IDE port busy
Aug 26 09:30:04 localhost kernel: ata_piix 0000:00:1f.2: no available
legacy port



It seems the dramatic change in libata.c incurs this failure.