This also implies that the entire ld/ subdirectory can go away after this is in.

Someone double check this, but as far as I can tell, we want NEITHER a MK_BINUTILS nor a MK_LLD_IS_LD check at this point -- the existence of this file should be based purely on whether or not base/binutils is installed.

Yep. IMO you could also do it in the same commit (because ld/ is pretty much trivial), fine with me either way.

I believe you are correct, paging @jhb
That said the other WITH_PORT_BASE_BINUTILS cases are in the top-level ObsoleteFiles.inc.

Yes, this should probably move to the top-level at some point. I think it would be fine to do that separately as part of removing support for in-tree ld.bfd and perhaps not part of this change? I would also put removing ld in that change as well. They may be small, but I think it's worth splitting those up.

Yep, fine with me either way

will take this comment back out, powerpc is lowest common denominator and should work fine even for stuff for use on book-E.

Shouldn't this be LD="${LD} -m elf32ppc_fbsd"? It seems like this would pass -m elf32ppc_fbsd to make and make ignores it since the directory doesn't exist?

-m directory
             Specify a directory in which to search for sys.mk and makefiles included via the <file>-style include statement.

The -m flag doesn't do much for lld since it will just set infer everything from the first .o file if not present. The only thing that the _fbsd suffix does is set the OSABI field (which might not be required since the crt files add the freebsd note section).

You're right, committed rS359373 to fix this.