Now we can distinguish from reg to mem (AMD64_OP_ADDR_REG) and from mem to reg (AMD64_OP_REG_ADDR) operations.

There is no node left that uses it.

We use lower_CopyB anyway so there are no copybs left when lowering.
They are all lowered to load/store sequences or memcpy calls.

This avoids AnyVal/additional Keep constructins.

I'm not proud of the 4 restriction but for now this works.

There is plain no need to.

arm_cmp_attr_t contains an arm_shifter_operand_t as base, but this was not checked.
This should correct langshootout/bintree.c.

This uses the newly introduced be_Relocation node to perform most PIC
transformations in ia32_pic.c. This introduces the elf and elf-no-plt
PIC modes. The "efl" mode does not work yet because the additional ebx
input makes calls register pressure unfaithful.

This node will be used as a helper for PIC lowering passes which want to
force a specific relocation getting used for an entities address.

One Conv is sufficient.

The modes did not match in several places.

We only need to force a specific frame entity size if we are actually
going to store one for the bitcast. We must not do that when we load from
an existing memory location.

This fixes backend/frame_entity.c

Simplify the creation of instrinsic/runtime function entities a bit and avoid creating duplicate types.

Only IA32_PIC_NONE and IA32_PIC_MACHO so far, but more will come.

This is not really a new entity type but merely a relocation type.
Handle it with a x86_immediate_kind.

Handle this similar to the ia32 backend.

This way we don't need a separate frame_entity field in the attribute
struct anymore. We can also use the immediate type to determine whether
we need to assign a fresh stackentity to a node.
This commit does not change the frame_use field, which is now a
duplicate way to indicate whether we need to assign a stackentity. This
currently requires a lot of asserts to ensure the two stay in sync. In
the future I'd like to use frame_use only for the special case when a
32bit or 64bit entity needs to be forced.

No need to adjust Pop which cannot have frame offset destinations,
determine whether the fixup is necessary by checking the base register
used instead of relying on generic frame offset stuff.