Breakpoint Issue on V41 R9I

[Ángel Martin]

I'd like to report an issue on V41 R9I related to the breakpoints functionality.

I noticed the changes in the "Console" screen , now showing the bank detail for each of the 16 pages. That's a nice addition, but unfortunately it now doesn't allow addresses from secondary banks in the breakpoints screen!

This means we've lost the ability to trace the MCODE execution for secondary banks, which was a godsend for troubleshooting purposes.

So can you look into this issue and repair it to the status of the previous revisions?

[Christoph Giesselink]

(03-24-2023 08:46 AM)Ángel Martin Wrote:  I'd like to report an issue on V41 R9I related to the breakpoints functionality.

I noticed the changes in the "Console" screen , now showing the bank detail for each of the 16 pages. That's a nice addition, but unfortunately it now doesn't allow addresses from secondary banks in the breakpoints screen!

This means we've lost the ability to trace the MCODE execution for secondary banks, which was a godsend for troubleshooting purposes.

So can you look into this issue and repair it to the status of the previous revisions?

Sorry I don't understand the question. What has the "Console" screen todo with the breakpoints screen?

Until R9F the origin behavior of the Breakpoint i.e. 5100 was to stop at address 5100h and the actual bank doesn't matter. Since R9G 5100 means still address 5100 but only in bank1. To get the breakpoint at address 5100h in bank2 according to the HELP.TXT file enter 2#5100 in the breakpoint dialog.

But lets have a look at a practical example. I attached the 64K W&W RAMBOX II. The RAMBOX OS is at page 8 and the function PG<> is at address 8DBC in both banks. So I set the breakpoint "8DBC" in the breakpoints dialog and let the breakpoint dialog open.

I assume that CAT2 show "-RAMBOX 64a" so that bank1 is active. So I do XEQ PG<> and the console dialog opens at address 8DBC and stops execution. Closing the Console dialog continues execution. A CAT2 shows "-RAMBOX 64b" now, that mean bank2 in page 8 is active now.

Lets do a XEQ PG<> again. The command switched back to bank1 and was executed without entering the breakpoint because bank2 of page 8 was active!

So lets _disable_ the breakpoint "8DBC: 180 ENBANK2" (uncheck the checkbox in that line) and lets set a breakpoint at address 8DBC in bank2 "2#8DBC". The new breakpoint "2#8DBC: 04E C=0 ALL" occur which is enabled by default.

Lets do another XEQ PG<> and page 8 is switched to bank2 without entering the breakpoint (remember we disabled the breakpoint in bank1).

Lets switch back to page 8 bank1 with XEQ PG<> and the Console dialog opens at "8DBC: 04E C=0 ALL" and stop execution. We entered the breakpoint in bank2.

The syntax for bank3 is 3# and bank4 is 4#. You can also enter "1#8DBC" as breakpoint which is shorten to "8DBC".

So the breakpoint handling changed in a minor detail to get full control over breakpoints in different banks. All known symbolic breakpoints like DATE are in bank1.

[Ángel Martin]

(03-24-2023 07:22 PM)Christoph Giesselink Wrote:  Sorry I don't understand the question. What has the "Console" screen todo with the breakpoints screen?

Nothing, I was just mentioning the change/addition in the console screen about the same functionality.

(03-24-2023 07:22 PM)Christoph Giesselink Wrote:  Until R9F the origin behavior of the Breakpoint i.e. 5100 was to stop at address 5100h and the actual bank doesn't matter. Since R9G 5100 means still address 5100 but only in bank1. To get the breakpoint at address 5100h in bank2 according to the HELP.TXT file enter 2#5100 in the breakpoint dialog.

ok, so there's been a syntax change to the U/I, fair enough - it's all clear now.

Thanks,
ÁM