Getting "kernel panic" after recompiling kernel

sumix · 03-02-2010, 08:22 AM

Hello,

I'm in an urgent need to recompile kernel (since I have to get pNFS to work), so I downloaded and patched a 2.6.33-rc6 vanilla kernel, configured and compiled. When I booted it, i got a kernel panic saying kernel was unable to access hda1 partition. Well, I told myself I had probably forgotten to compile PATA (piix) controller driver or ext3 driver in the kernel, which proved to be true. So I reconfigured it and recompiled, however I've been facing the same message:

Quote:

...
scsi 0:0:0:0 Direct-Access ATA MAXTOR ...
scsi 1:0:0:0 CD-ROM SAMSUNG ...
VFS: Cannot open root device "hda1" or unknown-block(0,0)
Please append a correct "root=" boot option: here are the available partitions:
Kernel panic - not syncing: VFS: unable to mount root fs on unknown-block(0,0)

I've googled a lot to solve this problem, haven't found any solution that works for me though. GRUB can see the hda1 (hd0,0) device, but kernel can't

. In menu.lst, I've the same configuration for the new kernel (with only changed kernel image name and commented initrd line) as the one for the default distribution kernel (I'm using Debian 5.0.4).

Above the kernel panic message I can see it seems like both PATA drives (CD-ROM and HDD) act like scsi devices, which might be the problem? But I don't know why (probably due to the kernel .config configuration - but which options exactly?) and I don't know whether it is really the reason of my problem. Maybe I could fix it with an init ramdisk, but I think it's not necessary here, so I would prefer to fix it without it... My .config can be found at http://charon.styxx.cz/tmp/.config.

Many thanks in advance!

JimBrewster · 03-02-2010, 08:33 AM

I've been reading a bit about block device naming changes in the new kernels. You might try changing the "root=/dev/hda1" to "root=/dev/sda1" in your kernel line of menu.lst.

sumix · 03-02-2010, 08:45 AM

Thanks for the advice... unfortunately that doesn't work for me

. I think there's something wrong with my kernel configuration, but I can't figure out what exactly it is

(I suspect something related to the PATA controller)

JimBrewster · 03-02-2010, 09:05 AM

Did you compare the old config with the new?

Something else in this section should be selected, it seems to me:

Code:

# CONFIG_PATA_ACPI is not set
# CONFIG_PATA_ALI is not set
# CONFIG_PATA_AMD is not set
# CONFIG_PATA_ARTOP is not set
# CONFIG_PATA_ATP867X is not set
# CONFIG_PATA_ATIIXP is not set
# CONFIG_PATA_CMD640_PCI is not set
# CONFIG_PATA_CMD64X is not set
# CONFIG_PATA_CS5520 is not set
# CONFIG_PATA_CS5530 is not set
# CONFIG_PATA_CS5535 is not set
# CONFIG_PATA_CS5536 is not set
# CONFIG_PATA_CYPRESS is not set
# CONFIG_PATA_EFAR is not set
CONFIG_ATA_GENERIC=y
# CONFIG_PATA_HPT366 is not set
# CONFIG_PATA_HPT37X is not set
# CONFIG_PATA_HPT3X2N is not set
# CONFIG_PATA_HPT3X3 is not set
# CONFIG_PATA_ISAPNP is not set
# CONFIG_PATA_IT821X is not set
# CONFIG_PATA_IT8213 is not set
# CONFIG_PATA_JMICRON is not set
# CONFIG_PATA_LEGACY is not set
# CONFIG_PATA_TRIFLEX is not set
# CONFIG_PATA_MARVELL is not set
# CONFIG_PATA_MPIIX is not set
# CONFIG_PATA_OLDPIIX is not set
# CONFIG_PATA_NETCELL is not set
# CONFIG_PATA_NINJA32 is not set
# CONFIG_PATA_NS87410 is not set
# CONFIG_PATA_NS87415 is not set
# CONFIG_PATA_OPTI is not set
# CONFIG_PATA_OPTIDMA is not set
# CONFIG_PATA_PDC2027X is not set
# CONFIG_PATA_PDC_OLD is not set
# CONFIG_PATA_QDI is not set
# CONFIG_PATA_RADISYS is not set
# CONFIG_PATA_RDC is not set
# CONFIG_PATA_RZ1000 is not set
# CONFIG_PATA_SC1200 is not set
# CONFIG_PATA_SERVERWORKS is not set
# CONFIG_PATA_SIL680 is not set
# CONFIG_PATA_SIS is not set
# CONFIG_PATA_TOSHIBA is not set
# CONFIG_PATA_VIA is not set
# CONFIG_PATA_WINBOND is not set
# CONFIG_PATA_WINBOND_VLB is not set
# CONFIG_PATA_SCH is not set

sumix · 03-02-2010, 09:18 AM

In the original configuration, the directives you mentioned are set to be "M" (drivers compiled as modules), which doesn't provide any useful information

(since it's a general distribution kernel which has to provide support for different devices and uses an initial ramdisk).

JimBrewster · 03-02-2010, 09:26 AM

Quote:

Originally Posted by sumix

In the original configuration, the directives you mentioned are set to be "M" (drivers compiled as modules), which doesn't provide any useful information

(since it's a general distribution kernel which has to provide support for different devices and uses an initial ramdisk).

Ah, well, then you'll need to enable those needed for your controller either set to "Y" to be built in to the kernel or to "M" and then included in an initial ramdisk. Making an initrd is not hard, but the commands and procedures might vary by distribution.

sumix · 03-02-2010, 09:48 AM

Yes, I know that the controller driver is missing (or rather some driver related to the controller or filesystem), but I hoped someone experienced in this area would tell me exactly what to fix

. I didn't want to mess with making initrd (well, I know it wouldn't probably a big issue, however it's still some studying on top of my problem

). Nevertheless, right now, it seems that trying to compile all the drivers as modules and making initrd is the only possible solution. Now I've to figure out how to make the initrd (I know some time ago mkinitrd was used for this purpose but AFAIK it's deprecated now and use of mkinitramfs with different arguments and properties is recommended). Anyways, thanks for the advices.

JimBrewster · 03-02-2010, 10:01 AM

One thing that might make your kernel configuration quicker and make sure you don't miss any modules (though you'll get more than you need...):

After you patch your kernel source, copy /boot/config-<old kernel version> to /path/to/clean/patched/kernel/source/.config, then run make oldconfig. It will only prompt you for options that have changed from the old kernel to the new+patches.

Then you'll still have to make an initrd, but this could save you a little trouble in the config stage...

sumix · 03-04-2010, 06:21 AM

Well, I finally got it to work. Just posting one of the possible solutions in case someone encounters the same problem and finds this thread: I've accomplished it in a similar way as you described here, Jim.

I copied the original .config coming with the debian kernel to my kernel source directory (just to make sure something wouldn't be missing once again):

Code:

cp /boot/config-2.6.22-2-686 /usr/src/linux/.config

then cleaned all the stuff I had compiled before, made a slight modification in the configuration (added pNFS support), recompiled and installed the kernel+modules:

Code:

make clean && make menuconfig && make && make install && make modules_install

and finally made init ramdisk (my new kernel was 2.6.33-rc6 with pNFS patch, ie. 2.6.33-rc6-pnfs):

Code:

mkinitramfs -o /boot/initrd.img-2.6.33-rc6-pnfs 2.6.33-rc6-pnfs

and

added initramdisk file path to my GRUB config (menu.lst) so the first boot option looked like that:

Code:

title           Debian GNU/Linux, kernel 2.6.33-rc6-pnfs
root            (hd0,0)
kernel          /boot/vmlinuz-2.6.33-rc6-pnfs root=/dev/hda1 ro
initrd          /boot/initrd.img-2.6.33-rc6-pnfs

and successfully rebooted.

However, as I've been adviced (just only too late

), my problem was probably caused by missing kernel config options - there were two possible ways to go:

1. The way of "old" IDE drivers (probably not a problem, but I was scared by the word "deprecated"

Code:

<*> ATA/ATAPI/MFM/RLL support[*]     ATA disk support
<*>   generic/default IDE chipset support
<*>   Generic PCI IDE Chipset Support
<*>   Intel PIIX/ICH chipsets support

which should result in root device being /dev/hda

2. The way of new PATA/SCSI (not sure if it's a totally good definition):

Code:

-*- SCSI device support 
<*> SCSI disk support
<*> Serial ATA (prod) and Parallel ATA (experimental) drivers
    =>[*]   ATA SFF support
       <*>     Intel ESB, ICH, PIIX3, PIIX4 PATA/SATA support

which should result in root device being /dev/sda.

I haven't tried the two possible solutions mentioned above, but it looks like it should work without problems. Hope that will be helpful to someone facing similar problems in the future.