Hello: Here are a couple knowledge base articles on the subject. I will have about 100 unused interrupt vectors. The first article creates a different ISR for each unused interrupt and calls a common fault handling function. This generates way too much code for me. The second article seems like what I want but I'm not sure because I cannot get it to assemble. I get a syntax error when it sees the macro definition. I did try something very similar. I have an assembly file that ORGs a JMPS instruction to the location of each unused vector, skipping the ones I use. This reserves a contiguous block of memory from my first JMPS to the last one including the ones I skip, so I get a memory overlap error for each vector I am using. http://www.keil.com/support/docs/216.htm http://www.keil.com/support/docs/941.htm Does anyone have a neat way you have figured out to do this that you can share with me. Any help would be appreciated, Walt
Just an idea: drop the automatically generated interrupt vector table and build your own one. Use the macro processor to generate the necessary assembly code using a list of ISRs kept in an include file. I never had to use the macro processor, but having seen some Keil's assembly code I got the impression that it's much more powerful than the C preprocessor. For example, this is how the interrupt vector table is generated in Monitor166 source code:
%SET(COUNT,4) ; SET UP INTERRUPT TABLE %WHILE(%COUNT LE 01FCH) (%IF (%COUNT EQ (%BRK_T_ADR*4)) THEN (JMP FAR BRKP ) ELSE (%IF (%COUNT EQ (2BH*4)) THEN (JMP FAR SER_ISR ) ELSE (JMPS INT_ADR_SEG,INT_ADR_OFF + %COUNT ) FI ) FI %SET(COUNT,%COUNT + 4) )