System freezes after loading ide driver
 

I am trying to recompile kernel 2.6.5 in sarge however I seem to get stuck at bootup.

...
...
...
/dev/ide/host0/bus0/target0/lun0: p1 p2 < p5 >
hdb: max request size: 128KiB
hdb: 120103200 sectors (61492 MB) w/2048KiB Cache, CHS=65535/16/63, UDMA(133)
/dev/ide/host0/bus0/target1/lun0: p1 p2 p3 p4
hdd: ATAPI 40X DVD-ROM DVD-R CD-R/RW drive, 2048kB Cache
Uniform CD-ROM driver Revision: 3.20

and thats it. it freezes.

I decided to unpplug the power to the CDROM drive so it wont detect it. well that didnt work. it got stuck at hdb now.


My system is currently down. Can someone please help out?

Processor:
AMD ATHLON XP 2500

Well in future, it's always a good idea to keep your old kernel around (and bootable) for just these kinds of occassions. Mind you that doesn't help you now.

Is your primary hard disc IDE or SCSI? If IDE, did you make sure to compile IDE support monolithically into the kernel? It looks like you did but I'm making sure. Did you also compile your primary filesystem(s) monolithically into the kernel?

yes and yes

actually, the default bootloader debian installed for me was grub.

i installed lilo manually and i swear i did setup an emergency kernel however even thats not booting but its error is completely different.

What's the previous kernel's error (and version)? It could be hardware. If the two kernels are sufficiently different then the error they report may be quite different.

i am able to boot from a similar debian image kernel, 2.6.5-1-k7.

but because that one is full of everything i wanted to recompile my own.

i have baffled by whats going on. I have tried many distros.

Heck i nuked gentoo in favor of debian.

Have you tried to install a different distro since this error occured? If More distros report such errors then it's probably hardware. It's funny how the hardware can shit itself just when you change around the software.

I decided to download kernel 2.6.6 last night ... heres the error 2.6.6 gives me now.

hdb: task_no_data_intr: status=0x51 { DriveReady SeekComplete Error }
hdb: task_no_data_intr: error=0x04 { DriveStatusError }
hdb: Write Cache FAILED flushing!

hdd: ATAPI 40X DVD-ROM drive......
Uniform CD-ROM Driver revision 3.20

then it just sits there.

There is no error like that with kernel 2.6.5, it just lists the drive specs.

I am able to boot into the system with the debian 2.6.5-1-k7 which is pre-compiled by debian but now i think i screwed up my kde cause that just sits there saying "intializing system" crap so i did a gnome terminal and did startkde and noticed it was giving tons of warnings about not being able to load ICONS from the /usr/share dir.

you know what also on a side note I always knew my HDs were crammed on each other and they might be causing heat So i removed the floppy drive completely and put the first hd where the floppy disk used to be. a bit more ream for cooling.

by the way, this is on a new MSI KT600 mb which has SATA however SATA is disabled.

http://www.msicomputer.com/product/p...V-LSR&class=mb


---------------
UPDATE
---------------
well, i started PQmagic and was given an error (shown in image).

Click here

I hate to tell you but it sounds lke your hard disc is borked. You'll need a new one.

:mad: :mad:
oh no no.
you cant be saying that
its working
ya
its great.
it works fine in winblows.
also, the image kernel by debian boots just fine.

there must be something else thats wrong. ya thats it. something else.:scratch:

If you have even the possibility that your HDD is beginning to fail (the fact that it works in Windows is no proof of continued stability), and you have one Linux boot that still works, I would seriously consider backing up important data while you have the chance before it stops working.

Partition Magic will claim to fix errors like the one you show (this has happened to me), but it won't/can't (it claims to work with ext3 partitions, but this is not true; the only Linux filesystem PM works reliably with is ext2).

The drive may be OK as a whole, but that partition is likely borked. I would suggest backing anything necessary off of it ASAP, because there's a fairly good chance that the error will migrate to your working boot, and you'll be forced to reformat the partition in any case. I think you ought to do so anyway, while you still have a modicum of control over the situation.

What flavor of Debian is this, anyway?

I'd go a little further even. I'd back up everything valuable and DEFINETLY format the partition. It looks like that partition (at the least) is cactus. You can try a shutdown -Fr now which will reboot and run fsck on startup. That may fix you up or give you some idea of what's going on.

well i made a new discovery in my bios that for some reason i did not see in the past AFTER i had deleted all partitions on that drive (hdb).

In the bios under Standard settings --> HD (A) there was a 32 bit transfer or disk access mode. That for some reason was set to off by the manufacturer of the board (MSI).

I turned that on and so far I've noticed improvement in performance. I also reinstalled Debian (Sarge).

I have yet to recompile my kernel again to duplicate this issue but i will soon.

I also did research new SATA hard drives and I am not sure I am ready to afford a new HD.

You may have solved the whole problem already....hope that turns out to be the case.

And even if you did need a new HDD, you wouldn't necessarily have to buy a SATA drive, since your current drive is IDE, right? Or are you out of IDE controllers and only have SATA ones left?

my new kt600 has sata so why not but its like a parallel ata drive 120gig is 89. sata 80 is $139 ...

and besides, ata 133 vs sata 150 ? i dont really see a huge margin. i'll stick w/ regular.

aren't those errors (task_no_data_intr: status=0x51 and so on) related to a problem with 2.6.6 kernel-final and ide-drive.c? switching it back solved my problem (my system booted even if throwing those errors)
http://marc.theaimsgroup.com/?l=linu...8108213445&w=2

from http://www.varlock.com/ is possible to download a patch to revert the file back to the previous safe version