https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123565
--- Comment #9 from Michal Jireš <mjires at gcc dot gnu.org> --- (In reply to Roland McGrath from comment #8) > > A 15 bug tracks when a bug is fixed _in 15_. This is a 15 feature and it's > reasonable to expect that its fundamental bugs will be fixed on the 15 > branch. > The support of the syntax is preparation for LTO. If you rewrite already existing basic toplevel asm to new extended asm, it will work in GCC 15 and will work with LTO in GCC 16. That the new syntax is useful without LTO and require additional handling might be an oversight, but that part is implemented with the LTO support in GCC 16. > Moreover, nobody has addressed the original point that this test case > demonstrates multiple issues and it's not at all clear that making the test > case in the other bug work will indeed fix all these issues. AFAICT nobody > has actually looked at my test case for real. All the issues are fixed and the test is working without issues with GCC 16.
