> gcc-3.4 is emitting them. I know because because the BFD backend > was crashing due to some bugs in my sparc changes as I was > working on them. There were BFD_*_VT* relocations, .vtable_* was > in the generated assembler, and at link time BFD was trying to work > with these relocations. Do you have a C++ testcase at hand? -- Eric Botcazou