Re: ups cores when compiling with 3.2.3

In message <[email protected]>
        [email protected] wrote:

> I configured with the --enable_dwarf

Good.

> I also tried compiling with -gstabs+ (using both ups 3.38-beta2 and ups
> 3.37)

I'd recommend against that, it's only like to make things worse.

> At this point I've tried every combination I can think of on both Linux
> and Solaris.

Well I use it on Linux every day without this problem, so it must be
something to do with the debugging data that gcc is generating for
your program.

> I've run the ups core under a debugger (thankfully I compiled ups with gcc
> 2.95.3).  It looks like it is having problems in ao_symload.c:
> get_fi_ypes().

That's in the stabs reading code so I don't understand how it can be
there if you compiled with DWARF debugging...

> Here's my test program.  Put a break point anywhere, start up and try to
> view "buf".  (On linux it doesn't even get this far, it cores
> immediately.)

A minimal test case is very helpful, except that in this case I know
that it isn't going to fail for me as I use ups daily on much bigger
programs compiled with a very similar version of gcc.

If you can compile that test case on a linux box and send me the
binary then I should be able to work out what is going wrong.

Tom

-- 
Tom Hughes ([email protected])
Software Engineer, Cyberscience Corporation
http://www.cyberscience.com/