Wow. What a lengthy solution. This is true of "TRUE" hard drive failures. Often you will see messages like these. In this case however, I would say it's probably related to a buggy chipset/controller/mobo. Notice that he says only at bootup. If you see errors similair to these throughout your kernel messages then you are looking at a hard drive failure. However, this seems to be the kernel detecting a weird or buggy setup and working around it. Make sure you have the kernel drivers compiled into initrd for your particualr chipset and as modules for the kernel. Especially since you said it happened AFTER an upgrade. Possibly they added support for your buggy hardware and only now is it properly handling it. Perhaps before there were problems but went unnoticed. Have you tried a vanilla linux kernel from www.kernel.org? Hope that helps. On Friday 29 August 2003 01:56 pm, zentara wrote:

On Fri, 29 Aug 2003 15:50:47 +0100

"Roger James" wrote:

...

After upgrading my 8.1 kernel with the latest online updates I now get the following error messages atr boot up. <4>hda: drive_cmd: status=0x51 { DriveReady SeekComplete Error } <4>hda: drive_cmd: error=0x04 { DriveStatusError } <4>hda:saw err in ide_end_drive_cmd, disabling write barriers <4>hda: drive_cmd: status=0x51 { DriveReady SeekComplete Error } <4>hda: drive_cmd: error=0x04 { DriveStatusError } <4>hda:saw err in ide_end_drive_cmd, disabling write barriers Are they serious? Can I ignore them? Can I fix them?

From googling I think they may be something to do with ordered write

...

mode on a ReiserFS (jounalled) filesystem. But I have no idea what that means. Help..

I just had a similar problem. Someone else may know your answer, but it is symptomatic of big trouble. I sure hope someone more knowledgable than myself, can tell us what this is.

The first thing to do is try to make a backup, if you can even boot. Next, boot from an emergency cd and run reiserfsck on the partitions. If that dosn't work, you may be looking at a hard drive failure. My system wouldn't finish reiserfsck and wouldn't even boot afterwards. Reiserfsck told me I had "unremovable bad blocks".

When it happened a few weeks ago, I thought the harddrive was failing. It first started with those warnings you had, but the system would boot. After a few more boots, I would get program failures, displaying those messages, especially when moving files around. Then when I did boot and was running, the hard drive light would get "stuck on" like the hard drive was seeking when there was no activity. (I thought the RIAA was scanning my disk for mp3's....:-) ) Then finally, it had boot failure, where my /boot partition couldn't be found.

I have backups, so I thought what the heck..try to fix it. The alternative was throw the drive out.

<rant> My reasoning was I was playing with a dual boot I had to windows, and after that, the problem hit. So I knew from past experience windows will try to write hidden code in blocks marked "bad", and these seemed to be causing the problems. I had "bad blocks" which couldn't be deleted. Too weird and I suspected some sort of virus caused by windows. </rant>

My particular harddrive, an IBM, has a utilities disk which lets you "zero out" a harddrive. This usually wipes out the boot sector, the first 100 or so megabytes and the last 300 megabytes. Why they only go for the front and end is something of a mystery to me. It just make me more suspect of these "bad blocks". (As in the harddrive manufactureres know, and the government knows, but everyone else is left guessing)

Anyways, I ran it. It wiped the drive. Just to be sure I used my own similar wipe method. boot from a emergency cd dd if=/dev/zero of=/dev/hda bs=16384 count=100000

that will wack out the front of the disk.

Then I partitioned the disk into 4 equal partitions, and used dd and /dev/zero to completely fill the entire drive with 0's. As far as I can tell, it's very important to write 0's over the entire disk, because the "bad blocks" are probably written at certain places on the disk, and whoever is behind it, can reactivate them anytime, unless they are zeroed out. Then I refisked, and reinstalled from backup, and everything is running fine.

So what is the moral here? This very same problem has been showing up with increasing frequency, maybe there is some sort of secret virus going around. I don't want to badmouth these automatic update systems, but I wouldn't run anything as root, which lets some network connection have root rights to my disk. Network connections can be hijacked, and who knows whose in control? My instinct tells me there is a bunch of pro-windows network engineers who may be trying to cause trouble for linux users.

It sure opened my eyes, to have reiserfsck tell me I have "unremovable bad blocks", BUT have the blocks restored after writing zeroes over them.

I would be happy to hear what happens in your case. Good Luck.

-- I'm not really a human, but I play one on earth.

-- ---------------------- Eric Bambach Eric (at) CISU (dot) net ----------------------