https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69077
--- Comment #12 from Jakub Jelinek <jakub at gcc dot gnu.org> --- (In reply to Richard Biener from comment #11) > The machinery in dwarf2out is somewhat fragile (as discovered from LTO early > debug work). Note that I changed lto_symtab_merge to OR > DECL_POSSIBLY_INLINED > during that work (committed to trunk). But I remember I still hit that very > same assert. I see that code in lto_symtab_merge. But on this testcase that function is not ever called, in either the first or second lto1 invocation. But I can confirm that when compiling t2.cc DECL_POSSIBLY_INLINED equal to 1 is streamed for the mean function, and I can see in the first lto1 invocation that the first mean FUNCTION_DECL is read with DECL_POSSIBLY_INLINED set to 0 and the second one with DECL_POSSIBLY_INLINED set to 1. So, the question is where else is the mean from t2.o replaced with mean from t1.o or where else is that bit lost.
