| View previous topic :: View next topic |
| Author |
Message |
jlh
Tux's lil' helper
Joined: 06 May 2007
Posts: 145
Location: Switzerland::Zürich
|
 Posted: Sun Apr 13, 2008 12:11 pm Post subject: Weird IDE problem regarding harddisk and cd drive |
 |
|
My setup is like this: hda is a hard disk that I don't use anymore, so usually, after I boot the system, I spin it down using hdparm. hdb is my cdrom drive. And hdc is my main hard disk with gentoo on it.
A while back, my system suddenly froze and hda spun up all by itself. Num lock LED and mouse were dead. So I was a bit shocked, because so far I only managed to get linux crashing that way while experimenting with unstable kernel drivers. I rebooted and examined the syslog to find out that it did not manage to write anything to it.
Then a bit later, something similar happened again. hda spun up again and the system rebooted automatically. Again, nothing in the syslog.
And today, everything freezes again, but strangely, I can still move the mouse and move windows, but they keyboard is dead. I had xchat running in the background which was set to beep on some channel messages, and since it was beeping, I knew the system was still alive. After a while, it all unfroze and I could use the system again, and this time there's something in the syslog:
Apr 13 13:21:09 bender hdb: status error: status=0x51 { DriveReady SeekComplete Error }
Apr 13 13:21:09 bender hdb: status error: error=0x04 { AbortedCommand }
Apr 13 13:21:09 bender ide: failed opcode was: unknown
Apr 13 13:21:09 bender hdb: status error: status=0x00 { }
...repeat above two lines lots of times...
Not only that, it continually repeats the last two lines about six times per second all the time (still doing it as I write this).
Also, these are errors about hdb, which is my cd drive that I was *not* using when this all happened (neither the earlier times) and there wasn't even a cd in it. Actually, the entire IDE bus 0 should have been idle, since hda was spun down and not being used either.
In the stream of those repeated two lines I found the following. That's exactly 2 minutes after the freeze and probably was the exact time when things unfroze. hdc having a problem was probably the reason for the freeze in the first place.
Apr 13 13:23:09 bender ide: failed opcode was: unknown
Apr 13 13:23:09 bender hdb: status error: status=0x00 { }
Apr 13 13:23:09 bender hdc: DMA timeout error
Apr 13 13:23:09 bender hdc: dma timeout error: status=0x58 { DriveReady SeekComplete DataRequest }
Apr 13 13:23:09 bender ide: failed opcode was: unknown
Apr 13 13:23:09 bender hdc: DMA disabled
Apr 13 13:23:09 bender ide1: reset: success
Apr 13 13:23:09 bender hdb: status error: status=0x00 { }
Apr 13 13:23:09 bender ide: failed opcode was: unknown
Then further down I find this:
...repeat those famous 2 lines lots of times...
Apr 13 13:23:09 bender hdc: dma_timer_expiry: dma status == 0x21
...repeat those famous 2 lines lots of times...
Apr 13 13:23:09 bender hdc: DMA timeout error
Apr 13 13:23:09 bender hdc: dma timeout error: status=0x58 { DriveReady SeekComplete DataRequest }
Apr 13 13:23:09 bender ide: failed opcode was: unknown
Apr 13 13:23:09 bender hdc: DMA disabled
Apr 13 13:23:09 bender hdb: status error: status=0x00 { }
Apr 13 13:23:09 bender ide: failed opcode was: unknown
Apr 13 13:23:09 bender hdb: status error: status=0x00 { }
Apr 13 13:23:09 bender ide: failed opcode was: unknown
Apr 13 13:23:09 bender hdb: status error: status=0x00 { }
Apr 13 13:23:09 bender ide: failed opcode was: unknown
Apr 13 13:23:09 bender ide1: reset: success
...repeat those famous 2 lines lots of times...
I can't mount any cds, when I try, I get this:
Apr 13 13:31:51 bender hdb: ATAPI reset complete
Apr 13 13:31:51 bender hdb: status error: status=0x00 { }
Apr 13 13:31:51 bender ide: failed opcode was: unknown
Apr 13 13:31:51 bender hdb: status error: status=0x00 { }
Apr 13 13:31:51 bender ide: failed opcode was: unknown
Apr 13 13:31:51 bender hdb: status error: status=0x00 { }
Apr 13 13:31:51 bender ide: failed opcode was: unknown
Apr 13 13:31:51 bender hdb: status error: status=0x00 { }
Apr 13 13:31:51 bender ide: failed opcode was: unknown
Apr 13 13:31:51 bender hdb: ATAPI reset complete
Apr 13 13:31:51 bender hdb: status error: status=0x00 { }
Apr 13 13:31:51 bender ide: failed opcode was: unknown
Apr 13 13:31:51 bender end_request: I/O error, dev hdb, sector 64
Apr 13 13:31:51 bender hdb: status error: status=0x00 { }
Apr 13 13:31:51 bender ide: failed opcode was: unknown
Apr 13 13:31:51 bender hdb: status error: status=0x00 { }
Apr 13 13:31:51 bender ide: failed opcode was: unknown
Apr 13 13:31:51 bender isofs_fill_super: bread failed, dev=hdb, iso_blknum=16, block=32
Computer info:
Mainboard: Shuttle Spacewalker AK32
CPU: AMD Athlon XP 2600+
hda: ATA Disk, Maxtor 4D040H2, 40GB
hdb: IDE CD-ROM, FX54++W
hdc: ATA Disk, Western Digital, WDC WD1600JB-22REA0, 160GB
uname -a: Linux bender 2.6.24-gentoo-r4 #1 PREEMPT Tue Apr 1 18:44:48 CEST 2008 i686 Unknown CPU Type AuthenticAMD GNU/Linux
Please ask if more info is needed.
Any ideas? |
|
| Back to top |
|
 |
pappy_mcfae
Watchman
Joined: 27 Dec 2007
Posts: 5999
Location: Pomona, California.
|
| Posted: Wed Apr 16, 2008 6:22 am Post subject: |
|
|
If you're not using the drive, get it out of the system. You will have to rename your drives, but it's a waste to leave an unused drive in the system, IMHO.
As to your problem, I'd say you have at least one CD ROM drive that has gone belly-up. To make absolutely sure, stick in your install CD in said drive, and see if it can boot the machine. If so, then the drive I told you to remove might be in the throes of dying. If not, the CD ROM drive is dead, or the IDE adapter is.
Considering that cheap DVD ROM drives are plentiful, now might be the time to look at an upgrade. Buying a new DVD drive and getting rid of the old unused hard drive might also be a good way of troubleshooting whether it's a drive problem, or a mobo/IDE adapter problem.
Blessed be!
Pappy
_________________
This space left intentionally blank, except for these ASCII symbols. |
|
| Back to top |
|
|
kimmie
Guru
Joined: 08 Sep 2004
Posts: 531
Location: Australia
|
| Posted: Wed Apr 16, 2008 8:33 am Post subject: |
|
|
| ... and check your IDE cable as well. I had DMA timeouts once which were fixed by replacing the cable. Apparently the cable I was using didn't meet the spec, and was too long, although I still have no idea why it worked for months. |
|
| Back to top |
|
|
Akkara
Bodhisattva
Joined: 28 Mar 2006
Posts: 6702
Location: &akkara
|
| Posted: Wed Apr 16, 2008 9:05 am Post subject: |
|
|
Check also your power. Your report sounds similar to what I saw after I installed one of those hard-drive hot-swap removable trays.
Something could be causing your hda to spin up, and the resulting spin-up power-glitch might be interfering with the other drives.
They were not fatal in my case, possibly because SATA is built for re-initializing the interface as a result of hot-swapping. IDE, from what I heard, isn't meant to be re-initialized. (See topic 648242 to read more about this and my solution.) |
|
| Back to top |
|
|
jlh
Tux's lil' helper
Joined: 06 May 2007
Posts: 145
Location: Switzerland::Zürich
|
| Posted: Wed Apr 16, 2008 8:53 pm Post subject: |
|
|
Here are updates: I wanted to reboot to get rid of the repeating error messages that have filled dozens of megabytes in /var/log/messages. But after I went to single user mode to clean up the log file before rebooting, I noticed that it stopped emitting the messages (and not just because syslogd was shut down). Looking at the logs, it must have been caused either by logging out of KDE, or restarting/shutting down the X server. After going to the default runlevel and logging back in, all worked fine again.
Excepted that I had to re-enable DMA mode for my hard disk using hdparms (took me a while to notice) and that my cd drive still won't work. If I try to mount something, it emits a few of the error messages I had before.
| pappy_mcfae wrote: | | If you're not using the drive, get it out of the system. You will have to rename your drives, but it's a waste to leave an unused drive in the system, IMHO. |
That's true, but I'm kinda too lazy for that.  And it's not much of a waste since it's spun down and I never reboot, unless I have to. However, will I really have to rename my drives? If my main disk is still on IDE 1 master, it will stay /dev/hdc, right?
| pappy_mcfae wrote: | | As to your problem, I'd say you have at least one CD ROM drive that has gone belly-up. |
Yes, this could be a possibility. In fact I can't remember for sure whether I ever used the drive again since the first time it happened. And it doesn't necessarily mean that the drive won't work again after a reboot. I'll try when I'll reboot the next time. But does this mean it's possible for a defective but idle CD drive with no CD in it to just randomly cause such an error?
| Akkara wrote: | | Check also your power. Your report sounds similar to what I saw after I installed one of those hard-drive hot-swap removable trays. Something could be causing your hda to spin up, and the resulting spin-up power-glitch might be interfering with the other drives. |
Mmh, but that would mean it could happen any time /dev/hda is spun up. In fact, hda isn't totally unused, it still contains the boot sector and every time I run lilo, it spins up and I spin it down again afterwards, this has never been a problem. Also, I can't think of anything that would ever spin up hda all by itself.
In any case, thanks a lot for the replies! My preferred solution would be to buy a new computer because I'm hitting on this one's limits all the time... but my money tree hasn't flowered yet. |
|
| Back to top |
|
|
pappy_mcfae
Watchman
Joined: 27 Dec 2007
Posts: 5999
Location: Pomona, California.
|
| Posted: Wed Apr 16, 2008 10:01 pm Post subject: |
|
|
| jlh wrote: | | pappy_mcfae wrote: | | If you're not using the drive, get it out of the system. You will have to rename your drives, but it's a waste to leave an unused drive in the system, IMHO. |
That's true, but I'm kinda too lazy for that. And it's not much of a waste since it's spun down and I never reboot, unless I have to. However, will I really have to rename my drives? If my main disk is still on IDE 1 master, it will stay /dev/hdc, right? |
If the drive is going bad, it can seriously effect your system. IDE drives tend to be overly polite to their neighbors that are going bad. In other words, if one IDE drive in your system is going bad, it can throw continuous interrupts to the IDE controller...that can cause, at the least, serious slow-downs. At worst, it can cause the system to become unresponsive, or even lock up. I have even seen bad IDE drives completely shut down a system...as in it won't even post. I understand laziness, but there are times it can be detrimental to your situation.
| Quote: | | pappy_mcfae wrote: | | As to your problem, I'd say you have at least one CD ROM drive that has gone belly-up. |
Yes, this could be a possibility. In fact I can't remember for sure whether I ever used the drive again since the first time it happened. And it doesn't necessarily mean that the drive won't work again after a reboot. I'll try when I'll reboot the next time. But does this mean it's possible for a defective but idle CD drive with no CD in it to just randomly cause such an error? |
Absolutely! The only difference between a CD ROM drive with a disk is that the drive control circuitry is completely activated. When there is no disk in the drive, it is still connected to power, and to the computer, and it is communicating with the IDE controller. Once again, I have seen CD ROM drives stop a system from posting...even when there was nothing in the drive.
I doubt that it's a power situation. That would cause even more erratic operation than what you are already experiencing. I'm not saying it's impossible, I'm saying it's more likely to be one or more of the drives.
And believe me, I can relate to the lack-o-fundage thing, too.
Blessed be!
Pappy
_________________
This space left intentionally blank, except for these ASCII symbols. |
|
| Back to top |
|
|
jlh
Tux's lil' helper
Joined: 06 May 2007
Posts: 145
Location: Switzerland::Zürich
|
| Posted: Sat Apr 19, 2008 3:09 pm Post subject: |
|
|
One interesting thing that happened is that while I was using blender, whenever I started the physics simulation (pressing 'p'), it would emit some more of those messages about hdb. I definitely can't explain that one. (syslog messages get sent to my screen all the time, that's why I noticed).
Ok, I rebooted with my gentoo minimal CD and that went all fine, CD drive works perfectly well again. One thing worth mentioning: Minimal gentoo apparently automatically tried to mount hda and it was hanging at that point for about half a minute. So something might be wrong with my old HD. But that's not important anymore, because my master boot sector moved to hdc and hda has been completely unplugged and is now going to be idle forever. I also checked the IDE cables. Rebooting, all happy so far, CD drive still works.
Now let's see what happens next. Thanks again for all the help. |
|
| Back to top |
|
|
pappy_mcfae
Watchman
Joined: 27 Dec 2007
Posts: 5999
Location: Pomona, California.
|
| Posted: Sat Apr 19, 2008 6:52 pm Post subject: |
|
|
You're welcome.
Blessed be!
Pappy
_________________
This space left intentionally blank, except for these ASCII symbols. |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
Kernel & Hardware
