Skip to content

Latest commit

 

History

History
133 lines (95 loc) · 4.97 KB

README.ssdebug.md

File metadata and controls

133 lines (95 loc) · 4.97 KB
Raw

Source-to-Source Debugger

The source-to-source debugger (ssdebug, ssdb) operates by performing a high-level transformation of Mercury source code to provide a rudimentary debugger interface. As such, it has numerous limitations (see below), but can potentially work with all backends. It is mainly intended for when you cannot use the Mercury debugger mdb, such as with the Java or high-level C backends.

ssdebug is still an experimental feature.

Installation

To use the source-to-source debugger you can install the grades containing the .ssdebug grade component. One way to do this is to invoke configure with the option --enable-ssdebug-grades. This will add the grades hlc.gc.ssdebug, csharp.ssdebug and java.ssdebug to the set of library grades to install.

Compilation

Compile your program in a grade with the .ssdebug grade component (e.g. java.ssdebug or hlc.gc.ssdebug). Your entire program will be then compiled with --ssdb-trace level of deep.

An alternative way is to use mmc --make --link-ssdb-libs to compile your program in any grade. You then just need to set --ssdb-trace shallow or --ssdb-trace deep on the modules you wish to debug.

Tracing Levels

  1. --ssdb-trace none

    None of the procedures in the module will generate trace events.

  2. --ssdb-trace shallow

    All the procedures in the interface of the module will generate events of trace level shallow. Events of trace level shallow are only displayed if the parent procedure in the call stack is compiled in trace level deep.

  3. --ssdb-trace deep

    All procedures in the module will generate events of trace level deep.

Using the source-to-source debugger

You may run the program as usual. To bring up the debugger prompt, set the environment variable SSDB beforehand.

    $ SSDB=1 ./calculator
        1:      1  1    CALL calculator.main
    ssdb>

If you set SSDB=0 then you will need to explicitly enable the debugger later in your code by calling ssdb.enable_debugging/2.

As in mdb, the three numbers are the event number, call sequence number (CSN) and the stack depth. Type help to show a list of commands. All commands act like their mdb counterparts (with reduced functionality), except for list which prints the source code at the call site, not of the called procedure.

The debugger will execute commands from $HOME/.ssdbrc and .ssdbrc in the current working directory, in order, at startup. You can put your aliases and settings in those files. Lines starting with the # character will be ignored.

Programs built in .ssdebug grades use much more stack space (tail call optimisation is prevented by the transformation). You will likely need to increase the stack size, e.g.

    $ JAVA='java -Xss10m' SSDB=1 ./calculator

Limitations

  • There are no internal events. The only events are CALL, EXIT, REDO, FAIL, and, for Java and C# grades, EXCP.

  • Standard library procedures are treated specially. Events will only be generated at the boundaries at which a user procedure calls and returns from a standard library procedure. No events will be generated when a standard library procedure calls another standard library procedure.

  • The retry command works by executing forwards until reaching the end of the call to retry, then recursively calling that procedure. Any side effects of continuing execution will be visible. If it is not possible to reach the end of the procedure to retry, the program will simply keep executing. Press ^C to get back the debugger prompt.

  • When running on Mono 2.8 and earlier, ^C can cause the program to hang. This is fixed in later versions.

  • Exceptions are only handled in Java and C# grades. Only a single EXCP event is generated when an exception is thrown, instead of propagating EXCP events up the call stack to the nearest exception handler.

  • In grades which don't handle exceptions, the debugger's internal state will be inconsistent with the program's execution after an exception, so you had better quit the program and restart.

  • Breakpoints currently match procedures by module and name only. Predicates or functions with the same name but different arities will still match. The debugger will not warn you if you set a breakpoint on a non-existent procedure.

  • We provide the filename and line number of call sites, but not the location of the source code for the called procedure itself. Use mtags.

  • The print goal command does not distinguish predicates and functions.

  • Procedures with arguments which are neither in nor out will not be transformed, hence will not generate events when called.

  • Many commands available in mdb are not yet implemented for ssdebug.

  • There is no tab completion.

  • There is no I/O tabling.

  • Debugging of multi-threaded programs is not supported.