Re: best way to debug a 'changed emulated program counter' ?

2002-07-25 Thread Martin Elzen
Hi all. Well, just now I finally found out where things were going south with that 'changed emulated program counter' error I was getting with Poser. And, now that I've found it, I realised if I'd told you guys a bit more about what was going wrong, you'd probably have been able to help me on

Re: best way to debug a 'changed emulated program counter' ?

2002-07-16 Thread Martin Elzen
Joe Malone [EMAIL PROTECTED] wrote: Small clarification: *local* array. The changed emulated program counter error can be caused by writing beyond the end of a local array (which is stored on the stack) thereby messing up the return address (because the address to return to gets pushed on

Re: best way to debug a 'changed emulated program counter' ?

2002-07-12 Thread Martin Elzen
Hi all. I'd like to thank everyone who answered my question. The project I was working on *is* a multi-segment app, but compiling it under either smart or the 32-bit-pointer models doesn't help get rid of the problem. And Joe, that post you found by Ben Combee was exactly the one I'd found in

Re: best way to debug a 'changed emulated program counter' ?

2002-07-12 Thread Joe Malone
--- Martin Elzen [EMAIL PROTECTED] wrote: ... but from what it looks like to me now, my problem isn't due to an incorrect function pointer, so it seems likely that there's a write past the end of a global array. Small clarification: *local* array. The changed emulated program counter

Re: best way to debug a 'changed emulated program counter' ?

2002-07-11 Thread Daniel Morais
On Wednesday 10 July 2002 19:55, Joe Malone wrote: Actually, the usual problem here is one of 1) using an invalid function pointer 2) wrote too much data to a local array so that you overwrote your return addess. You won't get a line number, since when the crash happens, you are outside

RE: best way to debug a 'changed emulated program counter' ?

2002-07-10 Thread Paul Gibson
I would try to step through to find which line precisely results in the error. In the event of corrupt memory problems (as this seems to be) that is the only method that I have ever found to work. Of course, I have not encountered this particular problem, so this advice may not be worth what

RE: best way to debug a 'changed emulated program counter' ?

2002-07-10 Thread Kevin OKeefe
Is your application multi-segmented? All of the times that I've seen this error it was because I was calling a function in a segment other than the first segment in a condition where those segments are not loaded (e.g., startup, non-global launch codes, etc.). Kevin -Original

RE: best way to debug a 'changed emulated program counter' ?

2002-07-10 Thread Joe Malone
-Original Message- From: Martin Elzen ... MedEurope (7.24) just changed the emulated program counter to 0xFEFE. This address is invalid because it's in neither ROM nor RAM. Looking it up in the archives gives me the info that somewhere the program counter is being