large code. Strange behavior

[PCM programmer]

Read this thread (2 pages). Someone else had a similar problem.
http://www.ccsinfo.com/forum/viewtopic.php?t=23991

If that doesn't help, post if you're doing anything unusual, such as:
1. Using the XINST fuse.
2. Using an #INT_GLOBAL routine.
3. Enabling Global interrupts inside an interrupt service routine.
4. Enabling interrupts for which you don't have an interrupt service
routine.
5. Running a non-LF PIC at 3.3v, when the data sheet says to use
the "LF" version.
6. Using the #separate directive.
7. Using Goto statements.
8. Using the incorrect oscillator fuse for the crystal frequency.
9. Letting the program run off the closing brace in main(), where it
will hit a hidden SLEEP instruction (placed there by the compiler)
and go to sleep, thus appearing to lock-up (or failing to print an
entire string).
10. Failing to place 100 nF ceramic capacitors between all Vdd pins and
ground, close to the chip.
11. Using a PIC which has errata that may cause your problem.
12. Enabling the Watchdog Timer, but not clearing it often enough, or
in the correct place(s) in the program.
13. Having a poorly regulated Vdd voltage, allowing it to droop down
to a low voltage level during the execution of the program, or it
has excessive noise on it.