Re: [PATCH, X86] Test for MACH-O before HAVE_AS_GOTOFF_IN_DATA

On August 18, 2018 2:40:01 PM GMT+02:00, Iain Sandoe wrote: >Hi, > >If we use an assembler which supports HAVE_AS_GOTOFF_IN_DATA, >(e.g. a modern GAS) on Darwin, we produce wrong code because the >gotoff-in-data test is conducted before the mach-o case. > >This should be a no-op on non-Darwin targ

[PATCH, X86] Test for MACH-O before HAVE_AS_GOTOFF_IN_DATA

Hi, If we use an assembler which supports HAVE_AS_GOTOFF_IN_DATA, (e.g. a modern GAS) on Darwin, we produce wrong code because the gotoff-in-data test is conducted before the mach-o case. This should be a no-op on non-Darwin targets, since the Darwin test is guarded on #ifdef TARGET_MACHO. Boots