| 1 |
1626 |
jcastillo |
README file for the Linux DTC3180/3280 scsi driver.
|
| 2 |
|
|
by Ray Van Tassle (rayvt@comm.mot.com) March 1996
|
| 3 |
|
|
Based on the generic & core NCR5380 code by Drew Eckhard
|
| 4 |
|
|
|
| 5 |
|
|
SCSI device driver for the DTC 3180/3280.
|
| 6 |
|
|
Data Technology Corp---a division of Qume.
|
| 7 |
|
|
|
| 8 |
|
|
The 3280 has a standard floppy interface.
|
| 9 |
|
|
The 3180 does not. Otherwise, they are identical.
|
| 10 |
|
|
The DTC3x80 does not support DMA but it does have Pseudo-DMA which is
|
| 11 |
|
|
supported by the driver.
|
| 12 |
|
|
It's DTC406 scsi chip is supposedly compatible with the NCR 53C400.
|
| 13 |
|
|
It is memory mapped, uses an IRQ, but no dma or io-port. There is
|
| 14 |
|
|
internal DMA, between SCSI bus and an on-chip 128-byte buffer. Double
|
| 15 |
|
|
buffering is done automagically by the chip.
|
| 16 |
|
|
Data is transferred between the on-chip buffer and CPU/RAM via
|
| 17 |
|
|
memory moves.
|
| 18 |
|
|
|
| 19 |
|
|
The driver detects the possible memory addresses (jumper selectable):
|
| 20 |
|
|
CC00, DC00, C800, and D800
|
| 21 |
|
|
The possible IRQ's (jumper selectable) are:
|
| 22 |
|
|
IRQ 10, 11, 12, 15
|
| 23 |
|
|
Parity is supported by the chip, but not by this driver.
|
| 24 |
|
|
Information can be obtained from /proc/scsi/dtc3c80/N.
|
| 25 |
|
|
|
| 26 |
|
|
Note on interrupts:
|
| 27 |
|
|
The documentation says that it can be set to interrupt whenever the
|
| 28 |
|
|
on-chip buffer needs CPU attention. I couldn't get this to work.
|
| 29 |
|
|
So the driver polls for data-ready in the pseudo-DMA transfer routine.
|
| 30 |
|
|
The interrupt support routines in the NCR3280.c core modules handle
|
| 31 |
|
|
scsi disconnect/reconnect, and this (mostly) works.
|
| 32 |
|
|
However.....
|
| 33 |
|
|
I have tested it with 4 totally different hard drives (both SCSI-1
|
| 34 |
|
|
and SCSI-2), and one CDROM drive.
|
| 35 |
|
|
Interrupts works great for all but one specific hard drive. For this one,
|
| 36 |
|
|
the driver will eventually hang in the transfer state.
|
| 37 |
|
|
I have tested with: "dd bs=4k count=2k of=/dev/null if=/dev/sdb". It
|
| 38 |
|
|
reads ok for a while, then hangs. After beating my head against this for a
|
| 39 |
|
|
couple of weeks, getting nowhere, I give up.
|
| 40 |
|
|
So.....This driver does NOT use interrupts, even if you have the card
|
| 41 |
|
|
jumpered to an IRQ. Probably nobody will ever care.
|
| 42 |
|
|
Nor will I when the $380 2.5GB IDE drives hit the market in early 1996!
|
| 43 |
|
|
|
| 44 |
|
|
|
