Messages in this thread |  | | Date | Tue, 30 Apr 1996 15:01:55 -0600 (CST) | From | Colten Edwards <> | Subject | Re: IDE Disk speed enhancements |
| |
On Mon, 29 Apr 1996, David A Willmore wrote:
> On Apr 27, 3:59pm, mlord wrote: > > David A Willmore wrote: > > > I have tried this recently, but with my qd6580 VLB IDE interface, this > seems to > > > make no difference. I have the append "ide0=qd6580" in my lilo config and > the > > > kernel parses it correctly. Does this interface serialize and I don't know > > > about it? > > > > See qd6560.c for this info. The kernel does NOT serialize this > > interface, but then the kernel also only knows how to tune ide0 > > for the qd6560 -- does it actually have a second set of IDE ports?? > > Right, I looked at that, but, from my 'observation' of how the qd6580 DOS > driver works, the qd6580.c driver may be doing more than it thinks. I'll know > more when I get a copy of DOSEMU which logs I/O accesses. > > Yes, my 'Vision' brand VLB super I/O card has two IDE ports on it. One seems > to be an EIDE port and the second seems to be a normal ISA IDE port. I haven't > fully verified this.
This is basically how I got the information for this interface. My Vision VLB only had the primary ide port on it. I used a combination of tools to get the info for programming this interface, but dosemu IO logging will not allow the QD driver to finish loading. So I had to goto some DOS tools in order to get the rest of the information. The information I send to the card to initialize it is what the DOS driver sends for the primary ide port. On my system the speed increase was noticable so I submitted the info to Mark Lord. It seems to work for most people.
Colten Edwards
|  |