Hi Jim,
jim wrote:
I'm still assuming this is local to my installation
I removed IEEE 1284 support from my kernel and left ppdev and parport_pc
as loadable modules. Now the gdbproxy finds the unit straight away.
One issue removed.
The Linux machines I have used all run RedHat (various versions), and
have no attached printers. This will mean they do not have the normal
printer modules loaded. I have never tested whether these might
interfere with ppdev. It sounds like they might do.
I still get an error when I try to erase flash ("Could not
preserve/restore device memory (12)").
When writing to memory (load test1) I get several "could not read
device memory (6)" errors for reads of locations 0xffff, 0xfffe, 0xffe0.
These sound like a genuine failure to communicate with the target. I
have had no trouble that wasn't genuine hardware related trouble. If you
are seeing the initial message where the software identifies the
attached device, then obviously communication can occur. It sounds like
you just have flaky communication, and that sounds hardware related. I
don't mean faulty hardware (although that is a possibility). I mean
hardware where the timing goes wrong. The FAQ lists some things people
seem to have had trouble with in this area.
I did install this under win98.... it seems to work just fine there.
Hooray! Someone says it works OK under Win98. There were problems with
Win98 a few months back. I have had zero feedback as to how well things
work now I fixed the key problem.
It seems the IEEE 1284 support causes issues. I have not tried
compiling in the parport_pc and ppdev. these are loadable modules at
this point.
Loadable modules are generally a good thing in modern Linux kernels.
Does anyone have any ideas what would causing this problem?
No. Not really. What Linux are you using. As I said, all testing was
done with RedHat 7.2, 8.0 and 9.0 and I never had any problems of this
type. The FAQ lists some hardware related issues (and workarounds)
people have fed back to me. What Linux are you using? Is Win98 working
on the same machine?
I'm really interested in resolving problems like this, as I'd like to
ensure these tools work more smoothly across a broad range of the latest
hardware. It does, however, seem that people are having more trouble
with the latest fast machines with mspgcc, and other MSP430 toolchains.
Regards,
Steve
