mounting root filesystem stuck

Hi,
same thing here. I've got a Thinkpad T60 with Intel Centrino Duo:
uname -a = 2.6.15-25-686 #1 SMP PREEMPT Wed Jun 14 11:34:19 UTC 2006 i686 GNU/Linux

My / is on /dev/sda3 and ext3, my /home is on /dev/sda6 and xfs. Swap is on /dev/sda5.

The "hanging" does not happen if booted with i386 Kernel.

From dmesg:
[17179578.864000] ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 169
[17179666.764000] PCI: Setting latency timer of device 0000:00:1f.2 to 64
[17179666.764000] ahci 0000:00:1f.2: AHCI 0001.0100 32 slots 4 ports 1.5 Gbps 0x1 impl SATA mode
[17179666.764000] ahci 0000:00:1f.2: flags: 64bit ncq pm led clo pio slum part
[17179666.764000] ata1: SATA max UDMA/133 cmd 0xF888E500 ctl 0x0 bmdma 0x0 irq 66
[17179666.764000] ata2: SATA max UDMA/133 cmd 0xF888E580 ctl 0x0 bmdma 0x0 irq 66
[17179666.764000] ata3: SATA max UDMA/133 cmd 0xF888E600 ctl 0x0 bmdma 0x0 irq 66
[17179666.764000] ata4: SATA max UDMA/133 cmd 0xF888E680 ctl 0x0 bmdma 0x0 irq 66
[17179667.048000] ata1: dev 0 cfg 00:045a 49:0f00 82:746b 83:7f69 84:6063 85:f469 86:3c49 87:6063 88:043f 93:0000
[17179667.048000] ata1: dev 0 ATA-7, max UDMA/100, 195371568 sectors: LBA48
[17179667.048000] sata_get_dev_handle: SATA dev addr=0x1f0002, handle=0xdffdf7e0
[17179667.056000] ata1: dev 0 configured for UDMA/100
[17179667.056000] sata_get_dev_handle: SATA dev addr=0x1f0002, handle=0xdffdf7e0
[17179667.056000] scsi0 : ahci
[17179667.640000] ata2: no device found (phy stat 00000000)
[17179667.640000] scsi1 : ahci
[17179667.928000] ata3: no device found (phy stat 00000000)
[17179667.928000] scsi2 : ahci
[17179668.216000] ata4: no device found (phy stat 00000000)
[17179668.216000] scsi3 : ahci
[17179668.216000] Vendor: ATA Model: HTS541010G9SA00 Rev: MBZI
[17179668.216000] Type: Direct-Access ANSI SCSI revision: 05
[17179668.220000] Driver 'sd' needs updating - please use bus_type methods
[17179668.220000] SCSI device sda: 195371568 512-byte hdwr sectors (100030 MB)
[17179668.220000] SCSI device sda: drive cache: write back
[17179668.220000] SCSI device sda: 195371568 512-byte hdwr sectors (100030 MB)
[17179668.220000] SCSI device sda: drive cache: write back
[17179668.220000] sda: sda1 sda2 sda3 sda4 < sda5 sda6 >
[17179668.612000] sd 0:0:0:0: Attached scsi disk sda

Further down:
[17179673.220000] EXT3-fs: mounted filesystem with ordered data mode.

and then:
[17179692.348000] Adding 979924k swap on /dev/sda5. Priority:-1 extents:1 across:979924k
[17179692.404000] EXT3 FS on sda3, internal journal
[17179692.580000] md: md driver 0.90.3 MAX_MD_DEVS=256, MD_SB_DISKS=27
[17179692.580000] md: bitmap version 4.39
[17179693.156000] device-mapper: 4.4.0-ioctl (2005-01-12) initialised: <email address hidden>
[17179693.548000] cdrom: open failed.
[17179694.036000] SGI XFS with ACLs, security attributes, realtime, large block numbers, no debug enabled
[17179694.036000] SGI XFS Quota Management subsystem
[17179694.036000] XFS mounting filesystem sda6
[17179694.112000] Ending clean XFS mount for filesystem: sda6

I have the same problem here, also with a Thinkpad X60s. This looks like an old bug was reintroduced into the kernel. I've seen the same behaviour some time ago, it was related to the sata controller timing out, when trying to detect disks. See point 2 in http://lkml.org/lkml/2006/1/12/461, I don't when this was fixed but some kernel update did it.

For me at least, this still happens on 2.6.15-26.

For me also, the same problem still exists on 2.6.15-26.

So it's fine for all of you using the -386 kernel?

With the -386 kernel it hangs for about 5 or 10 seconds as opposed to 2 or 3 minutes on my X60s.

I didn't try -386 kernel, but I understand that -386 does not support dual core. Am I right?

Yupp, still got it in 2.6.15-26, too (on T60).
With -386 it has a delay of some seconds as osch0001 said, whereas the -686 takes minutes.

If I change "Serial (ATA) mode" from "AHCI" to "Compatibility" in the BIOS, I don't experience this problem.

This solves the problem for me too (How did you think about it?).

Does this change have any implications: does it reduce speed, memory, etc. ?

The problem is solved for me too.

Thanks osch0001, with that change in BIOS I don't experience the problem any longer, too.

Well, it isn't an option in my BIOS. But booting using the 386 kernel works. It is only the
686, versions higher than 2.6.15-23 that hang when trying to mount the root file system.
The error message is regarding udevplug. There are only a few diffs in the config files between
version 23 and 26, so it shouldn't be too hard to nail down.

Is this still an issue in Feisty ?

Any news on this ?

We are closing this bug report because it lacks the information we need to investigate the problem, as described in the previous comments. Please reopen it if you can give us the missing information, and don't hesitate to submit bug reports in the future. To reopen the bug report you can click on the current status, under the Status column, and change the Status back to "New". Thanks again!