This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: failure to construct global object



Folks:

I have discovered the source of my error.  My apologies to each and
every one of you for wasting your time, especially you, Martin.

Early in the process I was asked whether I used g++ to link my
executable, because if this was not the case, all bets would be off.
I responded that I was indeed using g++ to link.  What I failed to
mention/notice was that I build shared objects from each of my modules
before linking them and I was using ld to do this!  

For what it is worth, here is why I took a second look at this.  This
evening, while doing the homework assigned to me by Martin, I noticed
that the debugger was putting my breakpoints in wildly incorrect
locations.  E.g., when I requested a breakpoint in main() I was
rewarded with a breakpoint in stl_vector, a red flag if there ever was
one.  I cannot imagine why I didn't notice this before.

Anyway, let me repeat my apologies and thank you for your work.  You
people are beautiful!

Gisli



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]