dtvmon/dtvboot 0.8

tlr

Trainee

Posts: 151

Date of registration: Nov 27th 2005

Location: Stockholm/Schweden

1

Friday, September 29th 2006, 6:41pm

Changes:

configuration options added.
The default action upon reset can be modified by changing a byte.
Video config may be hard coded to free up the userport.
Loading DTVMENU or INTRO upon pressing R-Button may be configured.
holding '1' preserves the memory $000000-$020000 on $100000.
added 'Z' to zero fill $0800-$ffff.
additional dtvtrans support for joystick 1, joystick 2, and the userport.
much improved init that keeps a lot more of $0000-$0300 intact after a reset.

Download: dtvmon-0.8.zip

http://www.kahlin.net/daniel/dtv/dtvmon/dtvmon-0.8_1.png

http://www.kahlin.net/daniel/dtv/dtvmon/dtvmon-0.8_2.png

Go to the top of the page

peiselulli

Professional

Posts: 944

Date of registration: Sep 14th 2006

2

Thursday, October 5th 2006, 11:05am

RE: dtvmon/dtvboot 0.8

Is it possible for you to add features/change some INSIDE the monitor ?
I have some points that will be nice to change there.

Go to the top of the page

tlr

Trainee

Posts: 151

Date of registration: Nov 27th 2005

Location: Stockholm/Schweden

3

Friday, October 6th 2006, 3:43pm

RE: dtvmon/dtvboot 0.8

Quoted

Originally posted by peiselulli
Is it possible for you to add features/change some INSIDE the monitor ?
I have some points that will be nice to change there.

ofcourse!

I already modified it quite heavily to make it work with banking and the extra instructions, nothing stops me from changing the rest.

I have some things in the TODO list, like a "TL - transfer long" command that uses DMA to copy between 24-bit addresses.

What did you have in mind?

This post has been edited 1 times, last edit by "tlr" (Oct 6th 2006, 3:48pm)

Go to the top of the page

peiselulli

Professional

Posts: 944

Date of registration: Sep 14th 2006

4

Friday, October 6th 2006, 8:49pm

RE: dtvmon/dtvboot 0.8

three things:

first : I have used the "i" command to look in my new kernal (will be released this weekend btw.) to look for free space in the kernal that I have marked with NOPs. But I have pushed the return key on a line instead of aborting with <shift-return>.
Then the ASCII code was interpeted and pushed back into memory BUT WITH THE
WRONG CONTEND !

to check this, try this :

m c000
:c000 70 71 72 73 74 75 76 77 <put this values there>

list the memory with
i c000
and just press return on this line.
Then look onto the memory again:
m c000
:c000 d0 d1 d2 d4 d4 d5 d6 d7 <<< ARGH

So putting back memory listed with 'I' that is not a printable character is not a good idea.

Second :
In another monitor I used sometimes in the past (I forgot the name, sorry it was back in the
eigthies) you have a command

i c000 c100 c050 04 (the 'I' stands for insert)

this command transfers c050-c100 to c054-c104, puts 4*EA to the addresses c050-c053
and adjusted the branches/jumps etc. to the new location of the second area.
If a branch cannot be recalculated (because the distance gets to far) then
a warning was printed.
A command for deleting bytes in the same way was also possible.
d c000 c100 c050 04 reverts the command above

Third:
A command to list memory binary, maybe in character format (1 byte per line)
and sprite format (3 byte per line). Maybe not only single color, but also multicolor
to present 2 bits as one pixel.

Maybe hard things to add, but I will feel happy to get this commands back again.

This post has been edited 1 times, last edit by "peiselulli" (Oct 6th 2006, 8:50pm)

Go to the top of the page

tlr

Trainee

Posts: 151

Date of registration: Nov 27th 2005

Location: Stockholm/Schweden

5

Friday, October 6th 2006, 10:17pm

RE: dtvmon/dtvboot 0.8

Quoted

Originally posted by peiselulli
three things:

first : I have used the "i" command to look in my new kernal (will be released this weekend btw.) to look for free space in the kernal that I have marked with NOPs. But I have pushed the return key on a line instead of aborting with <shift-return>.
Then the ASCII code was interpeted and pushed back into memory BUT WITH THE
WRONG CONTEND !

to check this, try this :

m c000
:c000 70 71 72 73 74 75 76 77 <put this values there>

list the memory with
i c000
and just press return on this line.
Then look onto the memory again:
m c000
:c000 d0 d1 d2 d4 d4 d5 d6 d7 <<< ARGH

So putting back memory listed with 'I' that is not a printable character is not a good idea.

Hmm... Actually I never noticed this! I used the original ccs-mon for everything I did in the 80's.
Very suprising that I didn't destroy anything.

Maybe I knew about it, but forgot. I'll have a look.

Quoted

Second :
In another monitor I used sometimes in the past (I forgot the name, sorry it was back in the
eigthies) you have a command

i c000 c100 c050 04 (the 'I' stands for insert)

this command transfers c050-c100 to c054-c104, puts 4*EA to the addresses c050-c053
and adjusted the branches/jumps etc. to the new location of the second area.
If a branch cannot be recalculated (because the distance gets to far) then
a warning was printed.
A command for deleting bytes in the same way was also possible.
d c000 c100 c050 04 reverts the command above

Real men use 'n' new locate.

(that doesn't fix relative branches though)

Appart from 'i' and 'd' is occupied this is doable, but I consider it a luxury feature.
i.e: I probably won't implement it because I don't have time, but you never know...

If I do it, it will probably be mapped as a flag to new locate, say "ni" and "nd".

Quoted

Third:
A command to list memory binary, maybe in character format (1 byte per line)
and sprite format (3 byte per line). Maybe not only single color, but also multicolor
to present 2 bits as one pixel.

Real men program sprites in hex.

Apart from that, it should be doable. I remember that this feature was present in the monitor to Merlin's Eprom programmer. I never used it myself though.

All the suggestions are good ideas, but for the next release I will probably concentrate on things that are really missing.
e.g:

transfer long
save with different load address.
flash directory
selectable device number for dir and disk command.
the interpret bug.
possibly the binary dump, but what letters should it map to? (MB/MS?) Suggestions...

I'd love to have some different monitors to look at so I can use similar names for the commands...

Go to the top of the page

peiselulli

Professional

Posts: 944

Date of registration: Sep 14th 2006

6

Friday, October 6th 2006, 10:44pm

RE: dtvmon/dtvboot 0.8

Quoted

Appart from 'i' and 'd' is occupied this is doable, but I consider it a luxury feature.
i.e: I probably won't implement it because I don't have time, but you never know...

while I programmed in the kernal, i thought to often:
"damned, I forgot one single command".
And then the "n" command is not really a help, espacially,
if your code cannot be displayed on one screen ...

But it is real much work to implement this compared with the other points, I know that

Quoted

Real men program sprites in hex.

I do often so in the past, too often, I think
And real men don't need mnemonics ;->
But it can help to see if some data is a sprite or not.

"MS" and "MB" is a good choice, becaue I think in the old
monitor I used it was "S" and "C" and this are occupied now.

This post has been edited 2 times, last edit by "peiselulli" (Oct 6th 2006, 10:48pm)

Go to the top of the page

tlr

Trainee

Posts: 151

Date of registration: Nov 27th 2005

Location: Stockholm/Schweden

7

Saturday, October 7th 2006, 1:59pm

RE: dtvmon/dtvboot 0.8

Quoted

Originally posted by tlr

Quoted

Originally posted by peiselulli
three things:

first : I have used the "i" command to look in my new kernal (will be released this weekend btw.) to look for free space in the kernal that I have marked with NOPs. But I have pushed the return key on a line instead of aborting with <shift-return>.
Then the ASCII code was interpeted and pushed back into memory BUT WITH THE
WRONG CONTEND !

to check this, try this :

m c000
:c000 70 71 72 73 74 75 76 77 <put this values there>

list the memory with
i c000
and just press return on this line.
Then look onto the memory again:
m c000
:c000 d0 d1 d2 d4 d4 d5 d6 d7 <<< ARGH

So putting back memory listed with 'I' that is not a printable character is not a good idea.

Hmm... Actually I never noticed this! I used the original ccs-mon for everything I did in the 80's.
Very suprising that I didn't destroy anything.
Maybe I knew about it, but forgot. I'll have a look.

I think I found it! I broke the screen code to mem conversion accidentally when adding the new banking code.
It was relying on the fact that the carry-flag was unmodified during the call to the mem-fetch routine, which it was in the original code.

Go to the top of the page

peiselulli

Professional

Posts: 944

Date of registration: Sep 14th 2006

8

Saturday, October 7th 2006, 2:28pm

RE: dtvmon/dtvboot 0.8

Maybe the problem in the "w" command too ?

Go to the top of the page

tlr

Trainee

Posts: 151

Date of registration: Nov 27th 2005

Location: Stockholm/Schweden

9

Saturday, October 7th 2006, 2:45pm

RE: dtvmon/dtvboot 0.8

Quoted

Originally posted by peiselulli
Maybe the problem in the "w" command too ?

No, this bug is more complex. If I remember correctly it wasn't working fully even in ccs-mon. In hesmon it works though.
I never tried to use it much myself.

I might have a go if this is something you need.

I sent you a PM with the I command fix for you to try.

This post has been edited 1 times, last edit by "tlr" (Oct 7th 2006, 2:46pm)