Use 0 as load address for Mach-O libraries

[fmagin]

No description provided.

[rhelmot]

There seem to be three testcase failures - one spurious and two which assert that the thing you changed has not change. fix these testcases and I shall merge at your command!

[fmagin]

The Mach-O tests for the old base address are now fixed.

By now I also understand what was going on that required the workaround of loading at 0: There is an entire blob with rebase bytecode that isn't parsed yet, that should have fixed up some pointers. I'll be working on that in the next few days, and then Mach-O apps and libraries should have proper rebasing support