On Sat, Feb 16, 2013 at 12:00 AM, Johannes Pfau <johannesp...@gmail.com> wrote: > Ian Lance Taylor <iant <at> google.com> writes: > >> >> Why is that? decl_replaceable_p is supposed to be true for a function >> that may be replaced by an entirely different function at runtime. >> This is mainly to implement the correct semantics for weak functions. >> You can't inline a weak function, but at runtime the runtime linker >> might have resolved the weak function to something completely >> different. > > AFAICS make_decl_one_only also sets DECL_WEAK which ultimately > causes decl_replaceable_p to return false. I understand why weak > functions are replaceable but is this also true for DECL_ONE_ONLY > functions? Couldn'tdecl_replaceable_p check for DECL_ONE_ONLY in > addition to DECL_COMDAT?
Sorry, you're right, make_decl_one_only does set DECL_WEAK. I managed to confuse myself. >> Set DECL_COMDAT. You said that didn't work but you didn't fully >> explain why. A DECL_COMDAT function should be output in every object >> file in which it is referenced. > > I wasn't sure if that's the correct approach. If it is, some > further investigation will be necessary why it doesn't work. We > should of course find the places where template instances are > used in the frontend and not put out, but that's too much effort > right now only to make inlining work. Yes, setting the DECL_COMDAT flag is supposed to do the right thing. That is what the C++ frontend does; see comdat_linkage. > How does this approach work for explicitly instantiated templates? > Those might seem to be not referenced but they should still be always put > out? The C++ frontend does not set the DECL_COMDAT flag on an explicit function instantiation. See mark_decl_instantiated. Ian
