bootstrap failure; tiny test case

Robert Lipe robertl@sco.com
Sun Feb 6 11:53:00 GMT 2000


I originally thought this was a fortran thing.  I was wrong.  It just
turned out that libg77 contained the C code that triggers failures in
both dwarfout.c and dwarf2out.c that has a note saying:

          /* When optimization is enabled (via -O) some parts of the compiler 
             (e.g. jump.c and cse.c) may try to delete CODE_LABEL insns which 
             represent source-level labels which were explicitly declared by
             the user.  This really shouldn't be happening though, so catch
             it if it ever does happen.  */

Contrary to the note, -O doesn't have to be present to trigger this.
I can reproduce it on GNU/Linux by building the following program with
either -gdwarf or -gdwarf-2.  



Test case:

void f_clos(int x)

{        
        switch(x) {
                default:
                mumble:
        }
}

This is one reason for my recent bootstrap failures on all the SCO
targets.  The failure was introduced sometime since Feb 2.

RJL


More information about the Gcc mailing list