[ECOS] MIPS TX-39 and Debug Information ---

Rosimildo daSilva rosimildo@hotmail.com
Fri Apr 7 08:19:00 GMT 2000

>From: Jonathan Larmour <jlarmour@redhat.co.uk>
>To: Rosimildo daSilva <rosimildo@hotmail.com>
>CC: ecos-discuss@sourceware.cygnus.com
>Subject: Re: [ECOS] MIPS TX-39 and Debug Information ---
>Date: Fri, 07 Apr 2000 04:53:20 +0100
>Rosimildo daSilva wrote:
> >
> > Hi,
> >
> > I am trying to check the TX-39 toolset. I have built everything.
> > Compiled eCOS 1.3.1 and built the "hello" example. Now I am
> > trying to debug it, and GDB reports that there is no debug info.
> >
> > Using objdump, I see no .mdebug section, or something like that.
> > I can see by the size of hello.exe that there is no debug info
> > in it.
>The tx39 tools output dwarf-2 debugging information by default, not stabs
>(which would be when the mdebug section is used). There should be no
>.mdebug section - in fact we explicitly remove it for the tx39 because we
>found sometimes GDB got confused when both stabs and dwarf-2 debug
>information was in the same executable.
>Which gcc and gdb versions are you using? And did you compile hello.exe
>(and your app) with -g and not -ggdb or -gstabs or anything? If you just
>used the makefile, then that would mean just -g.

I followed the instructions from the eCos documentation to
generate the toolset for TX39. I used GCC 2.95.2.

For gdb, I have used the tarball from ( 2000, Mar 13 ).
I have tried with GDB and Insight. Both fail. I am almost
cartain that it is the "flags" passed to the link. I just
cannot see what it is wrong. :-).

yes, I used the makefile for hello. I have tried -g, -ggdb. It did not work.

BTW, if I check the hello.o, I can see that the debug info it is
there. So, it seems to be something related with the link phase.
The '-g' flag is included with the flags passed to the linker.

I have tried this under Linux, and CygWin 20.1. Same results.

Thanks for your help.

Regards, Rosimildo.

Get Your Private, Free Email at http://www.hotmail.com

More information about the Ecos-discuss mailing list