Bug 165657 – [2.6.19 regression] sata_sil/libata status handling problems

• Product: Gentoo Linux
• Status: RESOLVED NEEDINFO

[Description Jimmy.Jazz 2007-02-06 19:07:08 UTC]

Hello,

my box boots from a sata disk composed with 2 "dos" partitions sda1 and sda3. sda3 was configured with evms. It is a lvm2 volume group.
The boot partition sda1 has a ext2 filesystem and is accessible directly from grub.
The kernel configuration uses the pivot_root=/dev/evms/root.
All the tests were made with the same disk, the same .config file and without reinstalling the system.

Currently i'm writing the report with that disk installed and booting and the kernel 2.6.18 gentoo-r4

Since 2.6.18 gentoo r5 release, i'm unable to boot from the same sata disk. The micro kernel installed in memory has recognized the partitions but has been unable to access them, even for sda1 used to boot the kernel.

dmesg gives the following messages under ash shell when the kernel asks me to specify a valid root partition.

::shell
dmesg
sda, sector 160086400 
sd 0:0:0:0: scsi error: return code = 0x00040000
end_request : I/O error, dev sda, sector 160086512
end_request : I/O error, dev sda, sector 0
end_request : I/O error, dev sda, sector 8 
end_request : I/O error, dev sda, sector 160086400

lvm tool returns,

lvm> lvs
/dev/sda read failed after 0 of 4096 at 0: Input/output error
/dev/sda1 read failed after 0 of 2048 at 0: Input/output error
/dev/sda3 read failed after 0 of 2048 at 0: Input/output error

The motherboard is a Asus A7N Deluxe version 2.

That motherboard has a nforce2 pata controller and a Sil 3112A sata controller. 

The Sil 3112 driver cannot access the device. The problem appends also with  2.6.19 and 2.6.20 gentoo kernels as well. The last git source gives the same result.

lspci
00:00.0 Host bridge: nVidia Corporation nForce2 AGP (different version?) (rev c1)
00:00.1 RAM memory: nVidia Corporation nForce2 Memory Controller 1 (rev c1)
00:00.2 RAM memory: nVidia Corporation nForce2 Memory Controller 4 (rev c1)
00:00.3 RAM memory: nVidia Corporation nForce2 Memory Controller 3 (rev c1)
00:00.4 RAM memory: nVidia Corporation nForce2 Memory Controller 2 (rev c1)
00:00.5 RAM memory: nVidia Corporation nForce2 Memory Controller 5 (rev c1)
00:01.0 ISA bridge: nVidia Corporation nForce2 ISA Bridge (rev a4)
00:01.1 SMBus: nVidia Corporation nForce2 SMBus (MCP) (rev a2)
00:02.0 USB Controller: nVidia Corporation nForce2 USB Controller (rev a4)
00:02.1 USB Controller: nVidia Corporation nForce2 USB Controller (rev a4)
00:02.2 USB Controller: nVidia Corporation nForce2 USB Controller (rev a4)
00:04.0 Ethernet controller: nVidia Corporation nForce2 Ethernet Controller (rev a1)
00:05.0 Multimedia audio controller: nVidia Corporation nForce Audio Processing Unit (rev a2)
00:06.0 Multimedia audio controller: nVidia Corporation nForce2 AC97 Audio Controler (MCP) (rev a1)
00:08.0 PCI bridge: nVidia Corporation nForce2 External PCI Bridge (rev a3)
00:09.0 IDE interface: nVidia Corporation nForce2 IDE (rev a2)
00:0c.0 PCI bridge: nVidia Corporation nForce2 PCI Bridge (rev a3)
00:0d.0 FireWire (IEEE 1394): nVidia Corporation nForce2 FireWire (IEEE 1394) Controller (rev a3)
00:1e.0 PCI bridge: nVidia Corporation nForce2 AGP (rev c1)
01:0b.0 RAID bus controller: Silicon Image, Inc. SiI 3112 [SATALink/SATARaid] Serial ATA Controller (rev 02)
02:01.0 Ethernet controller: 3Com Corporation 3C920B-EMB Integrated Fast Ethernet Controller [Tornado] (rev 40)
03:00.0 VGA compatible controller: ATI Technologies Inc Radeon R300 ND [Radeon 9700 Pro]
03:00.1 Display controller: ATI Technologies Inc Radeon R300 [Radeon 9700 Pro] (Secondary)





Reproducible: Always

Steps to Reproduce:
1. boot kernel 2.6.18-gentoo-r5 and above
2. the "micro" kernel is unable to open the sda sata device but /dev/md-x and /dev/sdx exist.
3. booting from kernel 2.6.18-gentoo-r4 works without changing the .config file.

Actual Results:  
Cannot upgrade the kernel above  2.6.18 gentoo-r4 for instance


That certainly affects other nforce2 motherboards using lvm, md or evms root partition.

[Comment 1 Daniel Drake (RETIRED) 2007-03-06 18:28:21 UTC]

Am I right in saying that 2.6.20 does not work either?

[Comment 2 Jimmy.Jazz 2007-03-06 19:17:33 UTC]

(In reply to comment #1)
> Am I right in saying that 2.6.20 does not work either?
> 

You are right. The problem stays the same with 2.6.19 and 2.6.20 kernels.

[Comment 3 Daniel Drake (RETIRED) 2007-03-06 19:39:58 UTC]

Please apply this patch on top of 2.6.18-gentoo-r4 and confirm that it BREAKS:
http://dev.gentoo.org/~dsd/genpatches/trunk/2.6.18/2100_libata-status.patch

Of course, keep a working kernel around as well :)

[Comment 4 Jimmy.Jazz 2007-03-06 22:33:54 UTC]

(In reply to comment #3)
> Please apply this patch on top of 2.6.18-gentoo-r4 and confirm that it BREAKS:
> http://dev.gentoo.org/~dsd/genpatches/trunk/2.6.18/2100_libata-status.patch
> 
> Of course, keep a working kernel around as well :)
> 

I confirm.
Hopefully, i was clever enough to change the EXTRAVERSION name in the Makefile and read twice your message before proceeding.
Indeed, the patch should have been called "how to break my gentoo in 1 lesson" ;)

I'm sure for now there is a evil status register that does the bad job :)

FYI The message returned by evms is as before:

Engine: Error Code 5 (Input/output error) when reading the {primary,secondary} copy of feature header an object sda.


Anyway, i would thank you to take your time to find out that bug. I was kind of desperate not to be able to upgrade the kernel ;)

Jj

[Comment 5 Jimmy.Jazz 2007-03-20 10:09:16 UTC]

(In reply to comment #3)

Hello,

i have switched back to the last lvm2 and raid modules/packages instead of evms and i didn't have to modify libata-core.c again. I was using an initramfs like before, but not with genkernel. Also, i needed to modify slightly /sbin/rc and some other start addons like udev lvm and raid to get it work. But that's not related.

Anyway, it seems to be some incompatibilities between evms_activate itself and how it handles/calls the kernel driver.

I believe, you should forward the bug report to the evms gentoo team instead.

Thank you,

Jj

[Comment 6 Maarten Bressers (RETIRED) 2007-09-21 16:50:09 UTC]

Any news here? Do you still have this problem? Are you able to reproduce this with the latest development kernel, 2.6.23-rc7 as of this writing? If so, please attach  your .config and dmesg output. Thanks.

[Comment 7 Mike Pagano 2007-10-05 13:46:46 UTC]

If this is still an issue, please feel free to reopen.