On Fri, 22 Jul 2016, Bernd Schmidt wrote: > What's the motivation for supporting -fno-toplevel-reorder anyway? That's > practically just a legacy mode as far as I know.
I've made the prerequisite middle-end patch after noticing that libgcc explicitly enables -fno-toplevel-reorder (and that was breaking the build on my branch). The patch in the opening message of this thread is just a small cleanup for the backend that could be made as a followup. (plus, nothing in the nvptx backend fundamentally depends on -ftoplevel-reorder, so the override was rather confusing) Can you comment on the question about ASM_OUTPUT_EXTERNAL, did you avoid it for any particular reason? Thanks. Alexander