: dma_intr: status=0x51 { DriveReady SeekComplete Error }

carboncopy · 06-30-2004, 12:11 AM

Hi!

After I move my /root to reiserfs, problem started and I get the following message during boot-up (edited):

Code:

hde: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hde: dma_intr: error=0x84 { DriveStatusError BadCRC }
hdf: DMA disabled
ide2: reset: success

I would think this has to do with reiserfs which was formatted under 2.4.26 during installation AND then I moved on to 2.6.7 after installation.

I did google the error, Linus was answering some question on in but not a fix.

I don't understand exactly what is the problem which causes this. And yes, I don't have DMA on either my /dev/hde or /dev/hdf. And it can switch in between and suddenly both can loose DMA. And I will get SUCKY read time:

Code:

root@mtop:~# hdparm -t /dev/hdf2

/dev/hdf2:
 Timing buffered disk reads:  158 MB in  3.01 seconds =  52.46 MB/sec
root@mtop:~# hdparm -t /dev/hde1

/dev/hde1:
 Timing buffered disk reads:    6 MB in  3.75 seconds =   1.60 MB/sec
root@mtop:~#

Ok, I want to know, how can I avoid this if I want to use reiserfs.

If I can't avoid it, how can I avoid it, if I just use ext3?

And also what causes the problem, so I can fix it and avoid it.

carboncopy · 06-30-2004, 03:31 AM

Update on actions taken:

1. Delete all partition on /dev/hde
2. Repartition /dev/hde
3. Install Slackware 10.0 on /dev/hde1
4. Delete partition /dev/hdf2 (reiserfs; used to be /).

I rebooted AND I still get the same error and same shity slow data transfer rate This time on /dev/hdf

ARGHHH!!! which drive is the one which is causing the problem? /dev/hde or /dev/hdf?

Just for your info /dev/hde has been used by my Mac Os X. (hfs+ journal).
When I fdisk it the first time I can't see any partition in it. I didn't know I can switch fdisk to bsd style partition view [b]. Frustrated.

carboncopy · 07-02-2004, 06:29 AM

Urr.. Anybody can suggest something? Now i have an 80G ATA 133 harddisk doing nothing in my machine.

keefaz · 07-02-2004, 06:37 AM

I would check if driver for IDE shipset controler on the motherboard is correct (if you have via IDE shipset, make sure kernel was configured with it for example). I would check the bios too but it should be fine as it runned fine under 2.4.26. Did you try to format the whole drive at last ressort ? (just delete partition may be unefficient)

carboncopy · 07-02-2004, 09:51 PM

The odd thing about all this is that:

When I boot into /dev/hde, then /dev/hdf DMA would be disabled. If I boot into /dev/hdf then /dev/hde DMA would be disabled.

Do you thing it is something to do with my PCI-IDE card? I have a Sil680 card. And yes, I compiled both my motherboard (VIA) and SIL680 driver into the kernel.

And I just spend around US$10 for a NEW IDE cable by CoolerMaster. Suppose to have shielding and what not.

By the way, I have repartition hde and format it to ext3. But I still have not done anything to hdf because All my data is in there and I am not willing to transfer the data at 1.6Mbit/second.

When I mount any of the partition from hdf (I am booting into hde) The machine feels like a grandma crawling.

pH* · 07-03-2004, 03:03 PM

carbon copy: i found something similar to your problem (i think) in the archives...

click here

the fix suggested was to go under

ATA/IDE/MFM/RLL Support
-- IDE,ATA,Atapi Block Devices
----Use Multi-Mode by Default <------ select THAT to yes in your kernal config

then recompile, obviously.

not sure if it worked for him, and it didn't work for me, but my error is a bit of the same, and a bit different... I'll include it here just for reference sake.

Code:

hdd: status error: status=0x58 { DriveReady SeekComplete DataRequest }
hdd: status error: error=0x00
hdd: drive not ready for command
hdd: status error: status=0x58 { DriveReady SeekComplete DataRequest }
hdd: status error: error=0x00
hdd: drive not ready for command
hdd: status error: status=0x58 { DriveReady SeekComplete DataRequest }
hdd: status error: error=0x00
hdd: drive not ready for command
hdd: ATAPI reset complete
hdd: status error: status=0x59 { DriveReady SeekComplete DataRequest Error }
hdd: status error: error=0x20LastFailedSense 0x02
hdd: drive not ready for command
hdd: status error: status=0x58 { DriveReady SeekComplete DataRequest }
hdd: status error: error=0x00
hdd: drive not ready for command
hdd: status error: status=0x58 { DriveReady SeekComplete DataRequest }
hdd: status error: error=0x00
hdd: drive not ready for command
hdd: status error: status=0x58 { DriveReady SeekComplete DataRequest }
hdd: status error: error=0x00
hdd: drive not ready for command
hdd: ATAPI reset complete

which just repeats for a while... can't even get the full output of dmesg this takes up so much room...

keefaz · 07-03-2004, 03:43 PM

what is your output for cat /usr/src/linux/.config | grep -i dma ?
[edit]
try to add append="ide=nodma" to your lilo config, run lilo, reboot and see what happen

carboncopy · 07-07-2004, 04:02 AM

I think I found the culprit!

It is my pci-ide card. Now, I am running both my harddisk off my on-board IDE with no slow-down, infact I am getting a higher reading with buffer-cache reads. Damn.