Re: Q: Sun DBX fixes for DDD

[Arnaud Desitter]

----- Original Message ----- 
From: "Andrew Gaylard" <address@hidden>
Newsgroups: gmane.comp.debugging.ddd.general
Cc: <address@hidden>
Sent: Wednesday, September 01, 2004 7:32 PM
Subject: Re: Q: Sun DBX fixes for DDD


> Ingo Dahm wrote:
> > Hi Andrew,
> >
> > attd. the diffs and - for completness - the complete files. Please let
me know
> > if i can test some other stuff, if you need it.
> >
> > The stuff compiles correctly if HAVE_SUNDBX set (most likly through
config.h,
> > but i don't know how to do that best). Threads will be enabled and the
nasty
> > jump to the top of the file after setting a breakpoint is ''disabled''
:-)
> >
> > Cheers
> >
> > -ingo
> >
> > PS: Please use the email address above in the future.
>
> Thanks for these patches!   I have integrated them into the tree.
> They will go into 3.3.10, when it is released.
>
> I'm concerned that the HAVE_SUNDBX won't be set unless the person
> building the code knows to do it.
>
> - Will these fixes break other dbx versions if they are not
>    protected by HAVE_SUNDBX?  (I know that Tru64 and Irix use dbx,
>    and perhaps other Unixes too.)
> - If so, then perhaps an autoconf setting would ease the build process:
>    "--enable-sun-dbx" or something similar.
> - If not, then we can remove them.
>
> I'm reluctant to have this setting detected by autoconf, since
> many ddd users will run it on one platform via rsh or ssh to their
> target platform where the inferior debugger runs.  Thus if a non-
> Solaris machine runs ddd and connects to a Solaris machine with
> dbx, then it has no way of knowing...
>

A better design is to have a "ddd --sun-dbx" option and replace the
#ifdef HAVE_SUNDBX by "if (isSunDbx)" style code.
Shouldn't be too hard to do ? Any volunteer ?

Regards,

> Anyway, these are just some thoughts.
>
> Thanks for your help improving ddd!
>
> Regards,
> Andrew.