You are not logged in.
I have an WD IDE disk attached to one ITE8212 RAID card. It's the only disk in that card so it doesn't work in RAID mode. Since yesterday's update I noticed that when I try to write large amount of data (aprox. 4GB) in that disk after a while I get errors like
dmesg |tail
ide: failed opcode was: unknown
hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x84 { DriveStatusError BadCRC }
ide: failed opcode was: unknown
hda: recal_intr: status=0x51 { DriveReady SeekComplete Error }
hda: recal_intr: error=0x04 { DriveStatusError }
ide: failed opcode was: unknown
hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x84 { DriveStatusError BadCRC }
ide: failed opcode was: unknown
Dec 11 00:24:32 arch end_request: I/O error, dev hdc, sector 12375and after that the drive disappears. The /dev entries are still there but when I umount and mount that partition back again I can't see any files. When I reboot everything works again for a while and then the problem arises again.
Just for the sake of it I tried to transfer the drive to motherboard's IDE controller and I was able to copy 10GB of data with no problem so I don't think it's a bad drive or a bad cable.
Another thing I have noticed is that when the kernel loads I see messages like
Dec 11 00:17:18 arch hdc:hdc: recal_intr: status=0x51 { DriveReady SeekComplete Error }
Dec 11 00:17:18 arch hdc: recal_intr: error=0x04 { DriveStatusError }This was happening since the first day I installed that card but since everything worked fine I didn't bother to investigate further.
I am using the beyond kernel but the same thing happens with stock kernel also.
Any idea on what could be wrong?
Offline
You could try if the new pata drivers for your device work better than the ide drivers. Just a guess though.
Offline
After your post I tried the new pata drivers. The error messages are gone from the boot log but now drive's performance is very slow. It is like the drive is working in PIO mode.
[root@arch ~]# hdparm /dev/sdb
/dev/sdb:
IO_support = 0 (default 16-bit)
readonly = 0 (off)
readahead = 256 (on)
geometry = 24321/255/63, sectors = 390721968, start = 0
[root@arch ~]# hdparm -d 1 /dev/sdb
/dev/sdb:
setting using_dma to 1 (on)
HDIO_SET_DMA failed: Inappropriate ioctl for deviceHow can I fix this? Google is not helping much.
Offline
PATA drivers are experimental, that means that DMA support could still be missing for some of them.
It seems your IDE problem is also related to DMA though. Maybe this is fixes in 2.6.19.1 (in the repos later today).
Offline
I found kernel 2.6.18-ARCH in the pacman's cache and installed it and switched back to ide drivers. The original problem just came back. My next step is to try with some live cd. Let's see what will happen there.
Offline
Ok, I finally found the solution and I post it here just in case someone else needs it.
That card has 2 different BIOS versions, one acts as a RAID controller and th other acts as an ATAPI controller. I bought it with the RAID version and since I used it with only one drive attached, there was no RAID to build. I flashed with the ATAPI version and it works OK now. I don't see the ugly messages I posted earlier and I copied 16GB of data without problems.
However now I am a little puzzled by this
#cat /var/log/messages.log |grep status
Dec 15 18:29:56 arch ATA: abnormal status 0x7F on port 0x9E7
Dec 15 18:29:56 arch ATA: abnormal status 0x7F on port 0x967
Dec 15 18:29:56 arch ATA: abnormal status 0x7F on port 0xC407
Dec 15 18:29:56 arch ATA: abnormal status 0x7F on port 0xBC07
Dec 15 18:29:56 arch ATA: abnormal status 0x8 on port 0xA407but since everything seems to work fine I think I can leave it as it is.
Offline