Re: [SLUG] Burning CD with Redhat 7.1 Kernal 2.4.3-12

From: Bryan-TheBS-Smith (b.j.smith@ieee.org)
Date: Fri Sep 21 2001 - 13:40:07 EDT


Mike Manchester wrote:
> track 01: 12 of 31 MB written (fifo 100%) hdb: timeout waiting for data
> ide_dmaproc: chipset supported ide_data_timeout func only: 14
> hdb: status error: status=0x58 {driveready seekcomplete datareq user}
> hdb: drive not ready for command
> <... cut ...>
> hda: Maxtor 90576D4, ATA DISK drive
> hdb: Maxtor 53073H4, ATA DISK drive
> hdc: MATSHITA CR-587, ATAPI CD/DVD-ROM drive
> hdd: HP CD-Writer+ 7200, ATAPI CD/DVD-ROM drive
> ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
> ide1 at 0x170-0x177,0x376 on irq 15
> hda: 11255328 sectors (5763 MB) w/256KiB Cache, CHS=700/255/63, UDMA(33)
> hdb: 60030432 sectors (30736 MB) w/2048KiB Cache, CHS=3736/255/63, UDMA(33)

Hmmm, your 90576D4 and 53073H4 might be fighting over DMA control.
The "D" is for Ultra33, and the "H" is for Ultra100 ("U" would be
for Ultra66). Several UltraDMA drives from different manufacturer's
take issue when the drive is either paired with a drive or
controller of a different UltraDMA speed.

Since your on-board controller, PIIX4 (i440LX/BX/GX southbridge),
obviously only supports Ultra33, I'd get the Maxtor "update66" or
"update100" utility to force it to only use Ultra33 (UltraDMA mode
2). I'm not sure if this is on the MaxBlast disk, but if you goto
Maxtor's support FTP site, it is somewhere on there.

The other issue is that you are using multiple drives on a single
channel. If a process gets a good burst of data, especially from
the master drive, the slave drive can timeout and/or miss near-real
time transfer requirements, such as when burning.

Consider using a large FIFO buffer like 20MB using the "fs=" option
to cdrecord.

> at this point the system is locked up tight as a drum and the green
> light on the cd rom is flashing. The write lite is off.

Of course, you've had a serious DMA issue. Heck, there are tons of
IDE DMA issues under kernel 2.4.x, especially early releases. This
is compounded by the fact you're using more than one device on a
channel, which is a performance issue in many cases.

To review, try (from easiest to hardest):
- Large "fs=" option to cdrecord
- Maxtor update utility to force 53073H4 to only use "Ultra33"
(UltraDMA mode 2)
- Only one device per channel, at least in the case of the hard
drives

-- TheBS

-- 
Bryan "TheBS" Smith   mailto:b.j.smith@ieee.org    chat:thebs413
Engineer  AbsoluteValue Systems, Inc.  http://www.linux-wlan.org
President    SmithConcepts, Inc.    http://www.SmithConcepts.com



This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 19:56:02 EDT