🚀 go-pugleaf

| I am having a problem with gcc 1.40 on a DECstation 5000. Whenever I
| compile a program with the -g option and try to use dbx or gdb,
| neither debugger knows anything at all about any local variables. The
| only variables I can access are global and static variables. Needless
| to say, this makes it very difficult to debug.
| 
| Any ideas?

The offical 1.xx MIPS/DECstation/SGI compiler port does not provide
full debugging support.  The reason for this is that the MIPS
assembler does not provide any support for describing local variables
and such because it's own compiler bypasses the first assembler stage.
I wrote a post processor that reads in the assembler source with the
debug information written as comments and rewrites the object adding
the debug information, and put it into my OSF modifications to GCC.
This post processor will be part of the GCC 2.00 release whenever it
occurs, but GCC 1.xx releases are now supposed to be just bug fixes.

If you want to get a copy of my changes (for 1.39, not 1.40, since
1.40 came out after my freeze date for internal OSF support), they can
be found on the machine foobar.cs.colorado.edu via anonymous FTP, in
the pub/Gnu-For-Pmax directory.  I maintain a mailing list to be
notified when there are new patches.  Note that OSF/1 now uses it's
own object file format (OSF/rose) and standard BSD stabs for
debugging, which means I rely on others to report bugs.

Given that these patches are not official FSF source at this time, you
should report bugs to me, or at least indicate that you are using a
patched version.  I hope to put a new release out today or tomorrow,
which includes among other things a fix to the object file munger.

In terms of bugs, I have had it reported to me, that the Ultrix 4.1
dbx segfaults, but gdb has no problem with the executables.

--
Michael Meissner	email: meissner@osf.org		phone: 617-621-8861
Open Software Foundation, 11 Cambridge Center, Cambridge, MA, 02142

You are in a twisty little passage of standards, all conflicting.