Using GDB, it is straightforward to debug and monitor a target program by setting a break-point at a specific instruction, since the instruction addresses are known.
However, we have two concepts, static instruction, and dynamic instruction. It is possible to have a program with ten static instructions, that execute several times, ten million dynamic instructions, for the same program.
Is it possible to set a break-point at a specific dynamic instruction? For example, in some computer architecture simulators, it is possible to rise a trap at a specific clock cycle, or certain tick, like in Gem5.
I'm certainly not a GDB expert, but in certain instances you could emulate this behaviour through Python.
For example, if you want to target a certain iteration of a loop, you could create a Python script that runs within GDB and does
If speed is not a factor and you really want to target the N'th executed dynamic instruction, you could write a Python script that steps through the program and counts the number of executed instructions. Once your target N is reached, you can do what you want to do.
As said, I'm not a GDB expert and even hardly use it, so people who are more acquainted with GDB might provide you with tips and tricks inherent to GDB.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators