Microcode Studio Plus compile/assmebly errors

I'm a newbie to the forum. In my brief time lurking, i've seen some pretty amazingly intelligent responses to good and not-so-good questions.

I've searched for this issue in the FAQs and forums, but haven't found it, and would appreciate any guidance.

I'm using PBP245 with an 18F4620. MicroCode studio Plus is calling PBP and MPASM for me when I click the compile button.

Up to a point (program size-wise) it all works fine. I got my prototype board working, figured out the 18F4620.inc entries I needed, and made part A of my software compile, assemble, and work properly.

When I started to add features to the base code, at about 3900 bytes (according to the Results window at the bottom of MCS+) I started getting Errors like:
Error[102] c:\design\pbp245\pbppic18.lib 551 : Out of memory
Error[128] c:\design\pbp245\pbppic18.lib 551 : Missing Argument(s)
Error[124] c:\design\pbp245\pbppic18.lib 551 : Illegal argument
Error[102] c:\design\pbp245\pbppic18.lib 596 : Out of memory

The status below the errors says: "Success : 3904 bytes used"
And though it says "success", I get no new Hex file.

If I kill a few lines of the basic code, the errors go away, the program compiles and assembles correctly, and I get a new Hex file.

Maybe "success" is a left-over from the previous successful compile...?

Anyway, the HEX file i get when truly successful is about 10.1kB long (accordingto windows explorer), which sounds "about right" based on the size of the basic code - which I think is just shy of 4k.

I'm only using about 32bytes worth of variables. Seems like I should have TONS of headroom for expansion. The 18F4620 is a 64k (program space) device with something like 3k of variable space.

Haven't tried doing the same thing in MPLAB - really wanted to use MicroCode Studio Plus as my IDE - and haven't learned HOW to use MPLAB yet.

Anyone recognize the errors? Am I underestimating the amount of overhead required by PBP? (wouldn't that be part of the hEX file?) Or is this error message because I have something set up wrong in MCS+ or PBP or MPASM...?

Thanks much!

Mike

Hi Mike,
Not sure on the other 2 errors - but I think the "out of memory" errors are about Windoze, not your PIC.

Try using pbpw.exe instead - somewhere buried deep in the PBP files is a text file containing info about pbpw and why you might need to use it for larger programs.

Thanks for the reply.

So far I've only been able to find settings for what folder microcode studio plus looks in for the compiler. It doesn't seem to want me to choose the program. So I made backup copy of pbp.exe and renamed pbpw.exe to pbp.exe... and used that, but when I tried to compile, I got the same "out of memory" and "bad argument" etc. errors...

So, at least based on this, PBPW doesn't appear to solve the issue.

Next I'll try to manually compile and assemble, using pbpw from the command line, and see what happens.

Any other ideas?

Best,

Mike

Well - thanks to your suggestion I tried using PBPW manually (no MCS+), and it worked.

C:\design\PBP245>pbpw -ac:\design\mpasm\mpasmwin.exe -p18f4620 x1-x10test.pbp

Produced no errors on the same file that had crashed during compile/assembly inside MicroCode Studio Plus.

Guess I need to get some Mecanique tech support!

Thanks for your help - I'll report back here with how their support works out.

Mike

Microcode Studio Plus apparently wants to see MPASM.exe, not MPASMWIN.exe.

My (downloaded through the link on MELABS site) MPASM folder only had MPASMWIN.exe in it. So MCS+ searched around my HD and found an old MPLAB installation with MPASM.exe in the folder! I guess that version or the accompanying inc files didn't support the 18F4620 - so it generated all kinds of bogus errors.

I renamed the (desired) MPASMWIN.exe to mpasm.exe and MCS+ stopped substituting the older version of MPASM, and all is good with the world now.

Thanks for pointing me in the right direction with the PBP.exe vs PBPW.exe. That doesn't seem to have been the issue, but it got me looking at the compiler/assembler settings. Now to gobble up all that newly realized program space! Sure beats the old 16c54!

Best,

Mike

Good to hear you got it working!