Bit Format Watch Directives

pjv

Hello All;

Can someone shed some light on bit format watch directives for me.

I have a flag byte, and I wish to "watch" each of the bits with their own name, so I create:

flags········· ds···· 1

sync········· equ··· flags.0
error········· equ····flags.1
exit···········equ····flags.2

etc....

I make watch directives as:

watch sync,1,ubin
watch error,1,ubin
watch exit,1,ubin

When running, all the watch bits reflect the value of flag bit zero, instead of their respective bits.

So I change the watch directives to

watch sync.0,1,ubin
watch error.1,1,ubin
watch exit.2,1,ubin

And get the same result.

So I try watch directives as:

watch flags.0,1,ubin
watch flags.1,1,ubin
watch flags.2,1,ubin

Again with the same result.

Am I missing something, or does the watch directive not properly support bit formats ?
I'm positive this worked in earlier renditions of the IDE.

Comments please.

Peter


Post Edited (pjv) : 2/11/2005 10:26:14 PM GMT

PJMonty

Peter,

Here's the short answer:

It doesn't matter how you format it because there's a bug in the SXKey IDE that makes all bit watches use bit zero regardless of what you tell them to use.

Here's the longer answer:

I found the bug, and am currently testing the fix it to verify that it works correctly.

Here's the work around:

Use a UBIN watch where the number of bits = the bit you want to watch + 1. For example, if you want to watch flags.2, then the watch would look like this:

watch flags, 3, UBIN

or like this:

watch flags.2, 3, UBIN

This bug fix will be incorporated into the next point release of the SXKey IDE. Unfortunately, I don't know when that release will be, so until then, you'll have to use the workaround. It makes debugging more of a pain, but at least you can do it. Just remember that if you use the technique I suggest, the leftmost bit will be the one you want to monitor.

Thanks, PeterM

pjv

Thanks Peter, I'll try that.

If there is going to be a new IDE release in the (near ?) future, can we offer up some other buggy or annoyances that are looking for a fix?

I dont want whine too much because generally I think the tools are great, but the assembler address calculation error and the incorrect RTCC indication during interrupt are a nuisance.

As far as wish lists; keeping the "blue bar" near the center of the code screen while single stepping or "walking" would be a HUGE improvement in convenience.

Keep up the good work,

Peter (pjv)

PJMonty

Peter,

I'm not sure when the next release is. If you find bugs, please let me know about them. As for suggestions, they are welcome too. It's hard to say what bugs/suggestions will get fixed on any given release (a lot depends on the ratio of the amount of work to the number of people it will positively affect), but if you don't at least bring it up, it'll never change.

Thanks, PeterM

PS - Can you give me some detail on the RTCC during interrupt issue?

pjv

Hi Peter;

Thanks for being responsive.

I'll do some homework to localize and characterize the RTCC issue so you don't have to wade through a bunch of unrelated stuff. Might take a couple of days before I get time.

Peter