Hello, I am running CCS v3.3 on F2812. I set up a break point at asm(" ESTOP0"); in interrupt void ILLEGAL_ISR(void). Is there a trace facility on CCS I can trace down the code caused ILLEGAL_ISR? If not, I've opened CPU register on CCS, the PC pointed to the ILLEGAL_ISR address, which CPU register on F2812 CCS hold the code address which caused ILLEGAL_ISR? Thank you. Jim
Illegal operation TRAP on F2812
Started by ●March 13, 2007
Reply by ●March 14, 20072007-03-14
On Mar 13, 2:14 pm, "JimHugh" <y...@bizmail.com.au> wrote:> Hello, > > I am running CCS v3.3 on F2812. I set up a break point at asm(" > ESTOP0"); in interrupt void ILLEGAL_ISR(void). Is there a trace facility > on CCS I can trace down the code caused ILLEGAL_ISR? > > If not, I've opened CPU register on CCS, the PC pointed to the ILLEGAL_ISR > address, which CPU register on F2812 CCS hold the code address which caused > ILLEGAL_ISR? > > Thank you. > > JimHi Jim Either it is on the stack (document SPRU430D, page 3-14 describing automatic context save), or it is in the RPC (return program counter). How is the CCS 3.3 working out for you? A friend of mind finds it slow compared to 3.1 and it has practically nothing new to offer for C2000 family. If they would fix the watch window, so that you could set font size I'd switch immediately (working on a 1600x1200 desktop). Regards, Mitja
Reply by ●March 14, 20072007-03-14
Korenje wrote:> On Mar 13, 2:14 pm, "JimHugh" <y...@bizmail.com.au> wrote: >> Hello, >> >> I am running CCS v3.3 on F2812. I set up a break point at asm(" >> ESTOP0"); in interrupt void ILLEGAL_ISR(void). Is there a trace facility >> on CCS I can trace down the code caused ILLEGAL_ISR? >> >> If not, I've opened CPU register on CCS, the PC pointed to the ILLEGAL_ISR >> address, which CPU register on F2812 CCS hold the code address which caused >> ILLEGAL_ISR? >> >> Thank you. >> >> Jim > > Hi Jim > > Either it is on the stack (document SPRU430D, page 3-14 describing > automatic context save), or it is in the RPC (return program counter). > > How is the CCS 3.3 working out for you? A friend of mind finds it slow > compared to 3.1 and it has practically nothing new to offer for C2000 > family. If they would fix the watch window, so that you could set font > size I'd switch immediately (working on a 1600x1200 desktop). > > Regards, Mitja > >Section 3.6 of spru430d describes the "illegal instruction trap" which operates whenever an illegal instruction is decoded. This is functionally equivalent to executing "TRAP #19" in software. The TRAP instruction is discussed in section 3.5.2 and in the tables/figures following it shows that the return address would be on top of the stack.