Eclipse Community Forums: C / C++ IDE (CDT) » MSP430 MSPGCC Debugging in Eclipse: Variable miss?

Home » Language IDEs » C / C++ IDE (CDT) » MSP430 MSPGCC Debugging in Eclipse: Variable miss?(Eclipse miscalculates variable addresses)

Show: Today's Messages :: Show Polls :: Message Navigator

MSP430 MSPGCC Debugging in Eclipse: Variable miss? [message #708662]

Wed, 03 August 2011 03:50

iamthemik3

Messages: 2
Registered: August 2011

Junior Member

Hello all,
I have been slowly chipping away at getting MSPGCC and MSPDebug working with Eclipse and i'm 99% there except for one odd hangup. When debugging, Eclipse doesn't seem to be able to calculate the correct address for local variables (when displayed in the variables window). Global variables seem to be fine since they always have an absolute address. Local ones, which depend on the value of the stack pointer, are off by a few. (If I force a local variable into a CPU register, it works fine)
For example: lets say I declare a variable "x" in a function. Viewing how the code addresses the variable in disassembly, i deduce that the address of '"x" is 0x027A. However when i type "&x" in the expressions pane to get what Eclipse "thinks" the address is, i get 0x0286.
Otherwise, programs run completely fine. It seems like the issue is somewhere in Eclipse and that it doesn't seem to be handling the stack pointer properly when debugging.
Any ideas? I'm completely stumped

Report message to a moderator

Re: MSP430 MSPGCC Debugging in Eclipse: Variable miss? [message #709048 is a reply to message #708662]

Wed, 03 August 2011 12:50

Marc Khouzam

Messages: 357
Registered: July 2009

Senior Member

Assuming you are using DSF-GDB, it is not Eclipse that calculates the variables or their addresses, it is GDB. You can look at the requests sent to GDB to see if there is a reason GDB gives you an unexpected answer:
http://wiki.eclipse.org/CDT/User/FAQ#I.27ve_been_asked_for_.27gdb_traces.27.2C_where_can_I_find_them.3F

Marc

Report message to a moderator