gm2
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gm2] Build problem on Redhat 9


From: Gaius Mulley
Subject: Re: [Gm2] Build problem on Redhat 9
Date: 23 Apr 2003 21:19:25 +0100
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2

Ben Stuyts <address@hidden> writes:

> 
> You mean pge, right? Do you have any hints on how to accoplish this? The  
> output from pge is HUGE... And I have no idea where to start.

Hi Ben,

yes pge, sorry..

I normally move the rule that is causing the problem

> ImplementationOrProgramModule :=                                           %  
> PushAutoOff %
>                                  ( ImplementationModule | ProgramModule )  %  
> PopAuto %
>                               =:

to the beginning of all the rules, ie above `FileUnit'. Then check
that the problem still exists. Then you could run pge by hand and pass
the `-c' option to pge which tells it not to generate any code. It
will generate comments (containing the rules) - if the `Bx' is still
in the comments then, great, we can bypass much of pge!

Now you could try running pge under gdb and setting a break point at
`Modula2Code' (note that there are two procedures called this in
gm2/gm2-compiler/ppg.mod - of course only one is compiled with pge -
the error checking variant) then at run pge, at the end of this
function you could `print WriteCodeHunkList(p)' which hopefully might
display the `Bx' if so rerun the program and single step this function
etc.. if a bug does exist in this function it is highly likely to exist in
the non error checking version of Modula2Code in the same source file.

hope this helps a little..

Gaius



reply via email to

[Prev in Thread] Current Thread [Next in Thread]