[Last Call] Learn how to a build a cloud-first strategyRegister Now

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 255
  • Last Modified:

UDMA Disk Error Messages

Hello all,

I am trying to do a new SuSE 8.0 install on a new computer that I just bought.  The computer has an integrated IDE/VIDEO/Sound chip.  The SIS5513.  I did an install of suse  
using EXT3 as the file system and am getting the following errors when I view dmesg.

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x84 { DriveStatusError BadCRC }
hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x84 { DriveStatusError BadCRC }
hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x84 { DriveStatusError BadCRC }
hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x84 { DriveStatusError BadCRC }
hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x84 { DriveStatusError BadCRC }
ide0: reset: success


The machine locks up some times and just doesn't seem to work right.  

I found this page that suggest that a patch is needed but I am not sure how to apply it or deal with it on install.

http://gyver.homeip.net/sis5513/

I am using a UDMA 66 7200 RPM 60 GB IBM Deskstar drive.  I have tried the checking the disk with IBM's fitness tools and the disk seems ok.  I have also replaced the IDE cable with a high qualith shielded cable.

I'm not sure where to go from here.

Jason
0
jaalex
Asked:
jaalex
  • 5
  • 5
1 Solution
 
GnsCommented:
Ah, I used to have that with some ols SiS chipset. I thought the drive was scrap (Seagate crap... The first one I bought really _was_ a paperweight, but the (working) replacement did the same stuff).
IIRC, I turned off all DMA/UDMA with hdparm, and after that things ... started working... slowly (PIO mode 4. Sigh).
The "lockup" are due to the (rather timeconsuming) reset.
Can you choose "UDMA-ness" and "level" in the BIOS?

Anyway, to apply the patch, you need to compile your own kernel. Just get a tarball from some mirror of www.kernel.org, unpack it, (following the instruction from your link) plopp in the new file, configure, compile (don't forget the modules), install, reinstall your bootmanager (LILO?)...
If you haven't built a kernel before, the margin for error is rather large, so it might be wise to keep the "old" kernel around (and with a separate entry in the bootmanager configuration).

Or else wait for a kernel update for your distro. This might be the safest bet, since many distributions have "custom patched" kernels... I don't know what the case is with SuSE 8.0 (haven't tried it).

-- Glenn
0
 
jaalexAuthor Commented:
Umm,  Seems strange.  I compiled the 2.4.19-rc1 kernel last night and it seem to do the same thing with the new kernel.  
I stuck a 20Gig disk that is also ATA66 and can't seem to produce the errors.  So I am not really sure what do to.
I am still testing to see.

0
 
GnsCommented:
Did you check what "features" are turned on (with hdparm)?

hdparm -i (and -I) will tell what available UDMA modes there seem to be, and hdparm -X<some number> can be used to fiddle with the different transfer modes available. Read the manpage carefully though, there are some pitfalls that can lead to massive filesystem corruption.

-- Glenn
0
VIDEO: THE CONCERTO CLOUD FOR HEALTHCARE

Modern healthcare requires a modern cloud. View this brief video to understand how the Concerto Cloud for Healthcare can help your organization.

 
jaalexAuthor Commented:
I am purchasing a new drive a lunch today to test with.
I should know more tonight or tomorrow.
0
 
GnsCommented:
I constantly forget that (relatively speaking) drives are *cheap* these days. Seems like a good strategy.

The problem with your particular error is that it indicates an error _somewhere_, all the way from magnetic media, through the complete communication chain, to the OS "driver".
Makes it rather hard to pinpoint:-).
Lowering the (potential) transfer (speed) mode can sometimes alleviate the problem, if the CRC corruption is due to mistiming. The problem can also be "invisible" with another drive, but (this is just conjecture, mind you) still present in the IDE controller. Or the drive might be the culprit, in which case you'll (also) be all right with a new drive...
Heck, it might even be the cable that is marginal in some way (wrong length, marginal contacting etc).

So you see, although a new drive might solve your problem, it really doesn't indicate that the Deskstar is scrap metal.
If you have another system, and the disk persistently "misbehaves" in that too, then we'd know with some certainty that the drive is bad.

-- Glenn
0
 
jaalexAuthor Commented:
Well this is what drove me to the new disk.  I have a 20 Gig maxtor drive that I tried in the system.  Installed a very limited system on the disk.  I then ran several hdparm commands and the disk was detected and operates ok.
No errors.  I wrote a shell script to read and write large amounts of data to the disk. I didn't see any errors.  I replaced the deskstar and did the same with an identical system install.  The result was several errors when I ran the shell script.   I have an older machine with the SIS5513 IDE controler and I repeated the test in that computer with the same results.  

I purchaed the new disk but have been unable to find time to test it.  With the upcoming holiday I have some time off and will attempt to test it then and award points.

Jason

0
 
GnsCommented:
Aha! So the error follow the disk!!!

Good, then we can be fairly certain that the error is in fact due to the Deskstar malfunctioning, and that the new disk should "Just Work(tm)". If you have any warranty time left on the Deskstar, I'd suggest you try and use it (the warranty, that is:).

Happy trails.

-- Glenn
0
 
jaalexAuthor Commented:
Yea the Disk is brand new.  I am working on a SUSE install of with the new disk now.  I plan to stick the disk into my workstation running an ASUS board to see what it does there.
It might just be a timing issue that the IBM and the SIS don't get along.  So far the Western Digital seems to be doing fine but I am only about half way in the install

0
 
jaalexAuthor Commented:
Changing the disk fixed it.  I put the disk into another computer with an VIA chipset and it seemed to work just fine there with no problems.  Must be a timing thing.
0
 
GnsCommented:
Jupp, probably is... or some foolishness in the controller... or ....

Good that you have them both working though.

-- Glenn
0

Featured Post

Get quick recovery of individual SharePoint items

Free tool – Veeam Explorer for Microsoft SharePoint, enables fast, easy restores of SharePoint sites, documents, libraries and lists — all with no agents to manage and no additional licenses to buy.

  • 5
  • 5
Tackle projects and never again get stuck behind a technical roadblock.
Join Now