On 29 Sep, Tim Waugh wrote:
> On Fri, Sep 29, 2000 at 07:20:15AM -0400, rjh@world.std.com wrote:
>
>> Any hints about likely places to check or configuration items that might
>> disable the read drivers?
>
> Compiler bugs? But something that large would surely break
> everything..
>
It now looks like a possible mismatch in patches. I got an hour on the
Turbo system and found that the port structure was wrong. It contained:
port->base = 0x378
port->hi_base = 0x0
and ECR present bit was set. The actual hardware has an ECR, so hi_base
should have been 0x778. This is why the wierd problems. The ECR
related commands and macros, like ECONTROL(), were going to the wrong
I/O address.
Was there a patch or addition related to this that came out in the
2.2.14 time period?
I just got another disk drive so that I can examine a Turbo system in my
lab. The problem Turbo system is inconveniently located on the other
side of an ocean.
I was guessing it was configuration related at first. It is unlikely to
be compiler related. I looked at the relevant assembler for my part of
the driver. It looked correct. Also, when I replaced the ECONTROL
macro with one that just inserted 0x77a most of the problems went away.
I think that if I also replaced the FIFO macro the rest would go away.
But that is not a proper answer. The real answer is fixing the
initialization of the port structure.
R Horn
-- To unsubscribe, send mail to: linux-parport-request@torque.net --
-- with the single word "unsubscribe" in the body of the message. --
This archive was generated by hypermail 2b29 : Fri Sep 29 2000 - 15:17:46 EDT