]> git.proxmox.com Git - mirror_ubuntu-bionic-kernel.git/blame - Documentation/scsi/dtc3x80.txt
Merge branch 'for-linus' into for-next
[mirror_ubuntu-bionic-kernel.git] / Documentation / scsi / dtc3x80.txt
CommitLineData
1da177e4
LT
1README file for the Linux DTC3180/3280 scsi driver.
2by Ray Van Tassle (rayvt@comm.mot.com) March 1996
3Based on the generic & core NCR5380 code by Drew Eckhard
4
5SCSI device driver for the DTC 3180/3280.
6Data Technology Corp---a division of Qume.
7
8The 3280 has a standard floppy interface.
9
10The 3180 does not. Otherwise, they are identical.
11
12The DTC3x80 does not support DMA but it does have Pseudo-DMA which is
13supported by the driver.
14
a33f3224 15Its DTC406 scsi chip is supposedly compatible with the NCR 53C400.
1da177e4
LT
16It is memory mapped, uses an IRQ, but no dma or io-port. There is
17internal DMA, between SCSI bus and an on-chip 128-byte buffer. Double
18buffering is done automagically by the chip. Data is transferred
19between the on-chip buffer and CPU/RAM via memory moves.
20
21The driver detects the possible memory addresses (jumper selectable):
22 CC00, DC00, C800, and D800
23The possible IRQ's (jumper selectable) are:
24 IRQ 10, 11, 12, 15
25Parity is supported by the chip, but not by this driver.
26Information can be obtained from /proc/scsi/dtc3c80/N.
27
28Note on interrupts:
29
30The documentation says that it can be set to interrupt whenever the
31on-chip buffer needs CPU attention. I couldn't get this to work. So
32the driver polls for data-ready in the pseudo-DMA transfer routine.
33The interrupt support routines in the NCR3280.c core modules handle
34scsi disconnect/reconnect, and this (mostly) works. However..... I
35have tested it with 4 totally different hard drives (both SCSI-1 and
36SCSI-2), and one CDROM drive. Interrupts works great for all but one
37specific hard drive. For this one, the driver will eventually hang in
38the transfer state. I have tested with: "dd bs=4k count=2k
39of=/dev/null if=/dev/sdb". It reads ok for a while, then hangs.
40After beating my head against this for a couple of weeks, getting
41nowhere, I give up. So.....This driver does NOT use interrupts, even
42if you have the card jumpered to an IRQ. Probably nobody will ever
43care.