Loading ...
Sorry, an error occurred while loading the content.
 

Re: Debugger Lauterbach

Expand Messages
  • 42Bastian Schick
    ... I did not mean, the use of the frame pointer does affect the calling convention, but anyway as we discussed some time ago, it is not completly possible to
    Message 1 of 10 , Aug 1 3:46 AM
      >> Not true:
      >> 1) gcc usese a frame pointer, IAR,MW and Cosmic not
      >
      > I disagree. The use of a frame pointer does not affect your calling
      > convention. You can still use -fomit-frame-pointer in a function and
      > call/use it from another one with a frame.

      I did not mean, the use of the frame pointer does affect the calling
      convention, but anyway as we discussed some time ago, it is not
      completly possible to tell the gcc not to use it.


      --
      42Bastian
    • 42Bastian Schick
      ... Hmm, at least I can see my source and can set breakpoints. With HiWave it is not possible to single-step through gcc compiled programs. And you need to get
      Message 2 of 10 , Aug 3 11:29 PM
        > According to your expérience on the simulator, Bastian, do
        > you think that it is possible to use the Lauterbach tools with all
        > these limitations ?

        Hmm, at least I can see my source and can set breakpoints.
        With HiWave it is not possible to single-step through gcc compiled
        programs.
        And you need to get the address from a map file in order to set
        breakpoints.

        I did some more tests and it looks like, Lauterbach has problems to find
        local variables.

        Globals (even local to functions) can be seen.

        (My first try was wrong: I did v.v (sciopta_t *)sciopta, but sciopta was
        no pointer, so I should have used v.v (sciopta_t)sciopta.
        BTW: sciopta is a variable declared in assembly, that's why I have to tell
        Trace32 the type.)

        --
        42Bastian
      • 42Bastian Schick
        ... So far, they have been very quick with answers and bug-fixes (mainly in some simulators). I will tell them. Let s see ... -- 42Bastian
        Message 3 of 10 , Aug 3 11:30 PM
          > It says on their web page, "TRACE32, one system fits all", so howabout
          > tell them to live up to their claim. They would have to either
          > understand other formats when they try to access code from other
          > compilers, or they would not be able to "fit all".

          So far, they have been very quick with answers and bug-fixes (mainly in
          some simulators).
          I will tell them. Let's see ...

          --
          42Bastian
        Your message has been successfully submitted and would be delivered to recipients shortly.