bexec* autostart

[tanteju]

After loading track 1 to $4a00 I found $0e in $4ce5.

Then
$4ce6 $b1 $e1

$00e1 is $1e and $00e2 is $2e so the buffer begin address look to be the same ($2e1e)...

This mean that the string 'RUN BEXEC*' should start at $4e2c which is not the case.
More curious is that I did not find any trace of "BEX" ($42 $45 $58) string between $4a00 to $4ff...

The starting value points toward the end of the buffer. I assume you find a $0d at $2e2c.

The buffer routine looks different... I have
$4ce8 $4c $6a
$4cea $3a $ea

after that same as in previous post.

That's a JMP $3a6a. Some extension. Don't have a listing from 3.3 at hand. What does this routine there do?

[bxdanny]

Wawa,

Here is a copy of your disk image, with the bytes needed to automatically RUN BEXEC* set to what they should be. It looks like you had saved track 1 from its "live" location of 2A00, instead of from 4A00, because the 65D input buffer on disk contained "SA 01,1=2A00/8". You may have done likewise with track 0.

[wawa.voun]

Thanks bxdanny !

I will give it a try tomorrow morning first hour !

Can you explain a little bit more the possible error I have done ?

Regards.
Philippe

[wawa.voun]

Hello,

The modified disk image provided by bxdanny is working.

Bexec* start at boot.

Could you provide more details about the modification you made ?

Regards.
Philippe

[bxdanny]

The modifications i made were as follows:

1) Place the string RUN"BEXE starting at location $2E25. (The "C*" and final carriage return were already there, since they had not been overwritten.)

2) Set locations $2321 and $2322 to $10 and $00, respectively (memory input and null output).

3) Verified that the memory-input pointer at 9098/9099 ($238A/$238B) was pointing to the RUN"BEXEC* string at $2E25. It was (that is, $238A contained $25, and $238B contained $2E), so this was not actually a change.

There was one more byte I changed which, on reflection, was completely unnecessary and superfluous (but harmless). That was to set 65D's pointer into its input buffer (at $2CE5, the operand of a Load Y Immediate instruction at $2CE4) to a value of 7, which is the offset from the start of the buffer at $2E1E to the start of the RUN"BEXEC* string at $2E25. But it is not 65D's input routine that reads that string, it's BASIC's, so that change (from $E) was actually pointless.

While I say I made changes starting at $2E25, and at $2321, etc., I actually made them at $4E25, $4321, etc., because I loaded tracks 0 and 1 into $4200 and $4A00, respectively, and saved the modified copies from those locations. The $2xxx locations are where the bytes in question load to when the disk is actually booted. It is important not to make such changes to the locations where your operating system is currently running, because other locations within the OS may have changed from their initial state as loaded from disk. You had apparently made changes directly to the $2xxx locations and saved the results from there, resulting in overwriting part of the RUN"BEXEC* string with the 65D command to SAve to track 1, and the memory input and null output I/O device settings with the settings for serial input and output (which BEXEC* sets them to when it gets control).

[wawa.voun]

Thanks again for your help.

Philippe