Crashing ONLY on debug

Langton · Joined: 29 Nov 2009 Posts: 25

Hi Guys (and Gals),

PCH 4.137
Windows 7
MPLABX
18F45K22

Everything was sweet (except a EEPROM write bug which I suspected may have been to do with the compiler 4.132) so I did an update to 4.137 and now it can clean and build but as soon as I go to debug it crashes. I can program after a build so my Real Ice has no dramas. The ONLY thing that has changed is the 4.137. I re-did the project from scratch in case there was a make file drama or something like that, but no joy there either. Searched Google, came up with a couple of things but no success. Any suggestions please?

here is what MPLABX says...

Ttelmah · Joined: 11 Mar 2010 Posts: 19537

How much ROM is your code using (when you build it without DEBUG)?. Is it using more with the new compiler than the old?.
Have you got the debug patch for PIC18F46K22 loaded?. This applies to your chip as well (it ought to be called PIC18xxK22), and causes a problem programming an area of memory just below the debug executive.

Best Wishes

gpsmikey · Joined: 16 Nov 2010 Posts: 588 Location: Kirkland, WA

Does that debug patch apply to Mplabx? From reading the post at Microchip, it seemed to indicate that it only affected MPLAB 8.63 and was going to be fixed in the February release. http://www.microchip.com/forums/m552373.aspx (I'm busy working on my little project with the 18f25k22, so this is very interesting - normally I trip over these things).

mikey
_________________
mikey
-- you can't have too many gadgets or too much disk space !
old engineering saying: 1+1 = 3 for sufficiently large values of 1 or small values of 3

Ttelmah · Joined: 11 Mar 2010 Posts: 19537

It is one of those 'some people get it' problems.
It is _only_ an issue if your code is putting data at the top of memory, right adjacent to the debug executive. Hence my question about 'size'. Have had loads of people working on code using these chips, without ever hitting this issue, but a few people have hit it, and then the fix seems to work.
My 'thought', was that the new compiler, might be making the code a fraction larger (possibly the EEPROM fix taking more space...), and hence hitting this problem. The problem vanishes when debug is not running, which would also 'fit' with the behaviour.
The fix, simply reprograms the address that MPLAB thinks is the size of the chip. It is meant not to apply to latter releases of MPLAB (they claimed it was fixed six months or more ago), but I've still seem people getting intermittent problems here.

Best Wishes

Langton · Joined: 29 Nov 2009 Posts: 25

I agree that is seem exactly like a similar patch is required. I checked and it is 8.63 only, that problem must have made a few people lose some hair. Yesterday I sent all the data to support and I've just (30min ago) went back to 4.132 which is now working perfectly, better the devil I know for now. Thanks guys for your replies, hopefully this will help someone else down the track