Re: Problem with IDE hd after instlalling potato with kernel 2.2.15

[markm@student.unsw.edu.au]

Hi Wojtek,

> However now, after the upgrade sometimes I receive the series of errors:

I don't think it is the ugrade / scratch install difference. I think
this may be due to a possible kernel bugs.
 
> ************ QUOTING FROM /var/log/syslog
> Jun 21 19:42:36 wzab kernel: hdb: read_intr: status=0x59 { DriveReady
> SeekComplete DataRequest Error }
[snipage]

> and the machine slows down (in fact almost freezes).
> After the hard reset, I get the "first hdd controller error" from BIOS.
> Only switching the power off solves the problem.

I haven't seen this my self, but I have seen some of your symtoms.

> My IDE configuration is as follows: (quoting from "dmesg" output)
[]
> hda: ST33240A, ATA DISK drive
> hdb: SAMSUNG SV1824D, ATA DISK drive
> hdc: CD-540E, ATAPI CDROM drive
> hdd: IOMEGA ZIP 100 ATAPI, ATAPI FLOPPY drive
[]
> hda: ST33240A, 3077MB w/0kB Cache, CHS=781/128/63
> hdb: SAMSUNG SV1824D, 17386MB w/472kB Cache, CHS=35324/16/63, (U)DMA
> hdc: ATAPI 40X CD-ROM drive, 128kB Cache
> Uniform CDROM driver Revision: 2.56
> hdd: 98304kB, 32/64/96 CHS, 4096 kBps, 512 sector size, 2941 rpm

I have:
hda: ST33221A, 3077MB w/128kB Cache, CHS=781/128/63, UDMA
hdc: WDC AC1210F, 202MB w/64kB Cache, CHS=989/12/35

[]
> 0% [Working]hdc: timeout waiting for DMA
> hdc: irq timeout: status=0x58 { DriveReady SeekComplete DataRequest }
> 0% [Working]hdc: timeout waiting for DMA
> hdc: irq timeout: status=0x58 { DriveReady SeekComplete DataRequest
> hdc: timeout waiting for DMA
> hdc: DMA disabled

I have seen this too - under heavy load while accessing both drives at
once. eg.

drive2:/usr/src$ bzip2 -cd /drive1/kernel2.2.tar.bz2 | tar -x 

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

However, both drives work fine independently. Also, I have the one
drive on each ide controller whereas you have filled up both master
& slave on both, which would make the problem worse.

I have a via MVP4 board and hence completely different ide controllers
(unless via reverse engineered intel). That is why I think it may be a
kernel bug.

Do your drives work under kernel 2.0.x? Have you looked to see if Andre
Hedrick (the linux ide guy) has any patches under 
your_kernel_mirror/kernel/people/hedrick (that is what i will do
when i have time).

Regards,
Mark.