Skip to contentSkip to conversation

IDE error on 2.4.17

DISCUSSION Simon Turvey 12 participants 19 messages Feb 26, 2002 – Feb 27, 2002

lore
Share & tools
lore
#0 lore raw

After a large file (4gigs) transfer using Samba attempts to access the file (also across Samba) resulted in lots of the following type of message.

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=250746, sector=250680
end_request: I/O error, dev 03:01 (hda), sector 250680

Can anyone point me in the direction of a reason/solution?

Thanks,
Simon

#1 lore raw
Quoted 2 lines · Simon Turvey

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=250746,

                              ^^^^^^^^^^^^^^^^^^

If I'm not misstaking that's hardware error.

Signature · 1 line
André Dahlqvist <[email protected]>
#2 lore raw
In reply to Simon Turvey
Quoted 4 lines

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=250746, sector=250680
end_request: I/O error, dev 03:01 (hda), sector 250680

Uncorrectable error is a message from your disk, along the lines of "Hey pal I wonder if the warranty has expired yet"

#3 lore raw

The drive's less than a year old :-(

Should I try disabling some of the UDMA stuff?

----- Original Message -----
From: "Alan Cox" <[email protected]>
To: "Simon Turvey" <[email protected]>
Cc: "Linux Kernel Mailing List" <[email protected]>
Sent: Tuesday, February 26, 2002 6:27 PM
Subject: Re: IDE error on 2.4.17

Quoted 9 lines
Quoted 4 lines

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=250746, sector=250680
end_request: I/O error, dev 03:01 (hda), sector 250680

Uncorrectable error is a message from your disk, along the lines of "Hey pal I wonder if the warranty has expired yet"

#4 lore raw
Quoted 2 lines

The drive's less than a year old :-(
Should I try disabling some of the UDMA stuff?

If it amuses you. You might as well dye your hair pink and dance naked around it for all the difference it will make

Alan

#5 lore raw

LOL, I get the message, thanks Alan.

----- Original Message -----
From: "Alan Cox" <[email protected]>
To: "Simon Turvey" <[email protected]>
Cc: "Alan Cox" <[email protected]>; "Linux Kernel Mailing List" <[email protected]>
Sent: Tuesday, February 26, 2002 7:10 PM
Subject: Re: IDE error on 2.4.17

Quoted 13 lines
Quoted 2 lines

The drive's less than a year old :-(
Should I try disabling some of the UDMA stuff?

If it amuses you. You might as well dye your hair pink and dance naked around it for all the difference it will make

Alan
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

#6 lore raw
In reply to Alan Cox

Am Dienstag, 26. Februar 2002 19:50 schrieb Simon Turvey:

Quoted 3 lines

The drive's less than a year old :-(

Should I try disabling some of the UDMA stuff?

you should return it to the place you've bought it

Johnny

#7 lore raw
In reply to Alan Cox
Quoted 3 lines · Simon Turvey

The drive's less than a year old :-(

Should I try disabling some of the UDMA stuff?

Age of the disk doesn't matter, they'll die at any age. They seem to die most frequently either within one year, or after 3 years.

Turning off UDMA probably won't help, it looks like it's time to restore from backups. If you try to recover data from the disk, make sure you mount it in read-only mode if you can get the drive that far up if you reboot.

-Dave

#8 lore raw
In reply to Alan Cox

It's a dev box anyway - just for tinkering. Even so, it's annoying.

Someone asked what the drive model was: IBM-DTLA-305030.

Will send the bugger back in the morning.

----- Original Message -----
From: "David Rees" <[email protected]>
To: "Linux Kernel Mailing List" <[email protected]>
Sent: Tuesday, February 26, 2002 7:01 PM
Subject: Re: IDE error on 2.4.17

Quoted 10 lines · Simon Turvey
Quoted 3 lines

The drive's less than a year old :-(

Should I try disabling some of the UDMA stuff?

Age of the disk doesn't matter, they'll die at any age. They seem to die most frequently either within one year, or after 3 years.

Turning off UDMA probably won't help, it looks like it's time to restore from backups. If you try to recover data from the disk, make sure you

mount

Quoted 9 lines

it in read-only mode if you can get the drive that far up if you reboot.

(quoted diff) +0 -2
-Dave
-

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

#9 lore raw
In reply to Alan Cox

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 19 lines · Simon Turvey

The drive's less than a year old :-(

Should I try disabling some of the UDMA stuff?

----- Original Message -----
From: "Alan Cox" <[email protected]>
To: "Simon Turvey" <[email protected]>
Cc: "Linux Kernel Mailing List" <[email protected]>
Sent: Tuesday, February 26, 2002 6:27 PM
Subject: Re: IDE error on 2.4.17

Quoted 7 lines
Quoted 4 lines

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=250746, sector=250680
end_request: I/O error, dev 03:01 (hda), sector 250680

Uncorrectable error is a message from your disk, along the lines of "Hey pal I wonder if the warranty has expired yet"

I've seen 2 drives with UncorrectableErrors that were correctable.

One happened after booting Linux with APM enabled, and the other after running Windows with vendor-specific DMA Block mode enabled.

In each case a low level format took care of it. No bad sectors on either drive in over a year.

- --
Skip ID: 0x7EDDDB0A
-----BEGIN PGP SIGNATURE-----

iEYEARECAAYFAjx76nUACgkQBMKxVH7d2wpW4wCg+MQQBIOZfHRqR0QKUhNbuFSB kkIAoK/ThQ/pIeoj3/Qu8ZAd6S26E8M/
=/bRM
-----END PGP SIGNATURE-----

#10 lore raw
In reply to Alan Cox
Quoted 4 lines · Simon Turvey

It's a dev box anyway - just for tinkering. Even so, it's annoying.

Someone asked what the drive model was: IBM-DTLA-305030.

That's your problem right there. I've had two of the DTLA drives die, a 15GB and a 45GB. They're notorious. They sent me back a new 120GXP-series 60GB one to replace the 45 gigger. Hopefully it'll last a bit longer...

Signature · 4 lines
Chris Friesen                    | MailStop: 043/33/F10  
Nortel Networks                  | work: (613) 765-0557
3500 Carling Avenue              | fax:  (613) 765-2986
Nepean, ON K2H 8E9 Canada        | email: [email protected]
#11 lore raw
In reply to Alan Cox
Quoted 3 lines · Simon Turvey

The drive's less than a year old :-(

Should I try disabling some of the UDMA stuff?

That won't help, this indeed is a media error. The drive is heading to hell. You have about another six months of life before it goes belly up completely.

Any chance it's one of those fast IBM 30 or 45 gig drives? They seem to be dying pretty fast ...

Quoted 25 lines

----- Original Message -----
From: "Alan Cox" <[email protected]>
To: "Simon Turvey" <[email protected]>
Cc: "Linux Kernel Mailing List" <[email protected]>
Sent: Tuesday, February 26, 2002 6:27 PM
Subject: Re: IDE error on 2.4.17

Quoted 9 lines
Quoted 4 lines

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=250746, sector=250680
end_request: I/O error, dev 03:01 (hda), sector 250680

Uncorrectable error is a message from your disk, along the lines of "Hey pal I wonder if the warranty has expired yet"

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Signature · 2 lines
Vojtech Pavlik
SuSE Labs
#12 lore raw
In reply to Alan Cox
Quoted 6 lines · Vojtech Pavlik

That won't help, this indeed is a media error. The drive is heading to hell. You have about another six months of life before it goes belly up completely.

Any chance it's one of those fast IBM 30 or 45 gig drives? They seem to be dying pretty fast ...

I expect a patch like this would help avoid these sort of questions...

drivers/ide/hd.c renamed from drivers/ide.orig/hd.c +9 -0
diff -urN linux/drivers/ide.orig/hd.c linux/drivers/ide/hd.c
@@ -201,6 +201,12 @@
 				printk(", sector=%ld", CURRENT->sector);
 		}
 		printk("\n");
+		/* Make sure people realize that very bad things are
+		 * happening, so they can do something about it before
+		 * it is too late... */
+		if (hd_error & ECC_ERR) {
+		    printk("hd%c: You should make a backup, this drive may fail soon!\n", devc);
+		}
 	}
 #else
 	printk("hd%c: %s: status=0x%02x.\n", devc, msg, stat & 0xff);
@@ -209,6 +215,9 @@
 	} else {
 		hd_error = inb(HD_ERROR);
 		printk("hd%c: %s: error=0x%02x.\n", devc, msg, hd_error & 0xff);
+		if (hd_error & ECC_ERR) {
+		    printk("hd%c: You should make a backup, this drive may fail soon!\n", devc);
+		}
 	}
 #endif	/* verbose errors */
 	restore_flags (flags);
drivers/ide/ide.c renamed from drivers/ide.orig/ide.c +7 -0
diff -urN linux/drivers/ide.orig/ide.c linux/drivers/ide/ide.c
@@ -832,6 +832,13 @@
 				if (HWGROUP(drive) && HWGROUP(drive)->rq)
 					printk(", sector=%ld", HWGROUP(drive)->rq->sector);
 			}
+			if (hd_error & ECC_ERR) {
+			    printk("\nhd%c: You should make a backup, this drive may fail soon!", devc);
+			}
+		}
+#else
+		if (drive->media == ide_disk && hd_error & ECC_ERR) {
+		    printk("\nhd%c: You should make a backup, this drive may fail soon!", devc);
 		}
 #endif	/* FANCY_STATUS_DUMPS */
 		printk("\n");

-Erik

--
Erik B. Andersen http://codepoet-consulting.com/
--This message was written using 73% post-consumer electrons--

#13 lore raw
Quoted 5 lines
Quoted 2 lines

Any chance it's one of those fast IBM 30 or 45 gig drives? They seem to be dying pretty fast ...

I expect a patch like this would help avoid these sort of questions...

This is the wrong approach. That information is available properly if and when the vendors install the smart utilities

#14 lore raw
In reply to Erik Andersen
Quoted 2 lines · Alan Cox

This is the wrong approach. That information is available properly if and when the vendors install the smart utilities

Doesn't necessarily help. I recently saw a hard drive which made funny noises and got really slow reading some parts of the drive (~30MB/s on some parts, ~300kB/s on others), but ide-smart didn't report failed tests. Two days later the drive was dead...

It was an IBM 60GB drive, but I don't remember the exact type. It called itself "IC35L060AVER07-0".

Jan

#15 lore raw

Did you enable smart?
Did you run the captive smart tests?

If this is on you / drive DO NOT run captive mode, the kernel go into starvation mode.

Quoted 12 lines · Jan Niehusmann
Quoted 2 lines · Alan Cox

This is the wrong approach. That information is available properly if and when the vendors install the smart utilities

Doesn't necessarily help. I recently saw a hard drive which made funny noises and got really slow reading some parts of the drive (~30MB/s on some parts, ~300kB/s on others), but ide-smart didn't report failed tests. Two days later the drive was dead...

It was an IBM 60GB drive, but I don't remember the exact type. It called itself "IC35L060AVER07-0".

Quoted 8 lines

Jan

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Andre Hedrick
Linux Disk Certification Project Linux ATA Development

#16 lore raw
In reply to Jan Niehusmann
Quoted 2 lines · Andre Hedrick

Did you enable smart?
Did you run the captive smart tests?

I only ran ide-smart /dev/hdx on the drive, and got a list of 'Passed' tests. Do I need to enable something else for ide-smart to work? What are the captive smart tests?

Jan

#17 lore raw
In reply to Erik Andersen
Quoted 9 lines · Jan Niehusmann
Quoted 2 lines · Alan Cox

This is the wrong approach. That information is available properly if and when the vendors install the smart utilities

Doesn't necessarily help. I recently saw a hard drive which made funny noises and got really slow reading some parts of the drive (~30MB/s on some parts, ~300kB/s on others), but ide-smart didn't report failed tests. Two days later the drive was dead...

I had two DTLA-series IBM drives. Both of them started having problems, nothing showed up on IBM's own bootable floppy based SMART tests, and both of them died with no SMART warnings.

I'm not at all confident that SMART is all that useful.

Signature · 4 lines
Chris Friesen                    | MailStop: 043/33/F10  
Nortel Networks                  | work: (613) 765-0557
3500 Carling Avenue              | fax:  (613) 765-2986
Nepean, ON K2H 8E9 Canada        | email: [email protected]
#18 lore raw
In reply to Simon Turvey
Quoted 9 lines · Simon Turvey

After a large file (4gigs) transfer using Samba attempts to access the file (also across Samba) resulted in lots of the following type of message.

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=250746, sector=250680
end_request: I/O error, dev 03:01 (hda), sector 250680

Can anyone point me in the direction of a reason/solution?

I got a similar error yesterday and this morning and asked about it on the ext3 list because I wasn't sure if it was fs related or hw.

http://marc.theaimsgroup.com/?l=ext3-users&m=101473048203232&w=2

--

=====================================================================
Mohammad A. Haque http://www.haque.net/
[email protected]

"Alcohol and calculus don't mix. Developer/Project Lead
Don't drink and derive." --Unknown http://www.themes.org/
[email protected]
=====================================================================