From: Claus Just Rasmussen (Claus_at_just-rasmussen.dk)
Date: 2006-04-02 21:54:42
Hi Nicolas,
Just retried with only 1 retry - it's a lot faster, and each sector ends
up with a "3" - don't know what it stands for, guess read-error?
The result was a 533.248 bytes large file, empty according to Vice.
Well... this might be an empty disk, but it shouldn't be...
I then retried set to 255 retries, it's still writing "3" and it's a lot
slower. Might indicate that my guess (read-error) is correct?!
It's really taking a long time now, so I decided to stop it, because I
saw your new program :-D
If I recall it correct, I get an "ibcmd err" if I wait long enough.
//Claus
Nicolas Welte wrote:
> Hi,
>
> I just had another idea. What error did you get exactly? What does the
> program say when it exits? It's possible that the 8250 takes a very
> long time to retry on a read error, and in this case, the GPIB driver
> will assume a timeout and the program will exit. I set the timeout
> value to 10 seconds which should be enough, but I can prepare a
> program version for testing with infinite timeout, or later with
> configurable timeout.
>
> For a first test, you can reduce the number of retries to 1 or so.
> That could also help, because the retries are done in the drive, not
> by the PC, IIRC.
>
> Nicolas
>
> Claus Just Rasmussen wrote:
>> Hi Nicolas,
>>
>> I have a problem reading one of the 8250 disks - one of the sectors
>> has a readerror.
>>
>> Question: Would it be enough to remove the "exit(1);" line in the
>> codesnip below to ignore the error?!
>>
>> chkin(unit,3);
>> if (ibrd (bd,buf,256L) & ERR) {
>> gpiberr("ibrd Error");
>> exit(1);
>> }
>> clrchn();
>>
>> Best regards
>> Claus
>>
>> Claus Just Rasmussen wrote:
>>> [ Note: This mail was converted from HTML to text by majordomo.
>>> Formatting might be poor. ]
>>>
>>> Hi Nicolas,
>>>
>>> It would be really nice if you'd help and modify the code. I haven't
>>> got any problem to look into it myself - except that I need to
>>> install a compiler, and I wouldn't know what to change in the
>>> 6502-code to make it work on 4040.
>>>
>>> Best regards,
>>> Claus
>>>
>>> Nicolas Welte wrote:
>>>
>>> --- Ursprüngliche Nachricht ---
>>> Von: Claus Just Rasmussen <Claus@just-rasmussen.dk>
>>> An: cbm-hackers@ling.gu.se Betreff: Re: 8250 and C64 disks
>>> Datum: Tue, 28 Mar 2006 21:35:13 +0200
>>> A combination - some 4040 disks and also imgrd, I'm
>>> afraid. The 4040 disks explains itself, the imgrd problem
>>> was/is that if I specify a filename in actual path, then it'll
>>> be overwritten by something <2KB.
>>> Strange. I never had this problem here. But that's
>>> for using a program on
>>> only a single machine. You never know what happens on the next
>>> one! But I'll
>>> have another look into the source if there is something obvious.
>>> I managed to get my hands on a 4040 so now
>>> I've got the HW in place to read the disks. Really nice.
>>> Next question: Do a 4040-version of imgrd exist? If not
>>> - will I need to change more than the #sectors,tracks?
>>> (bitclock?)
>>> Of course the map of sectors per track and number
>>> of total tracks has to be
>>> fixed. Then some minor changes for the filename extension (d64),
>>> and the
>>> autodetection of doublesided disks has to be skipped. I use some
>>> 6502 code
>>> that gets downloaded into the drive, I have to check if this is
>>> usable on
>>> the 4040.
>>> If you don't feel like hacking into the source code, I can
>>> have a look and
>>> add an option -40 to support the 4040 drive and disk layout.
>>> (I'm sure a
>>> 2031 drive can't be supported so easily, because my 6502 drive
>>> code had to
>>> be changed to the completely different memory layout of the drive).
>>> Nicolas
>>> Message was sent through the cbm-hackers mailing
>>> list
>>>
>>> -- ------------------------------------------------------------
>>> Claus Just Rasmussen, BSc.EE. www.just-rasmussen.dk
>>> ------------------------------------------------------------
>>>
>>>
>>> Message was sent through the cbm-hackers mailing list
>>>
>>>
>>>
>>
>
>
>
> Message was sent through the cbm-hackers mailing list
>
>
--
------------------------------------------------------------
Claus Just Rasmussen, BSc.EE.
www.just-rasmussen.dk
------------------------------------------------------------
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.