The following be_peephole_exchange() has a stricter check, i.e. same mode instead of only not mode_T.

It is like new_r_Proj(), but the mode is determined automatically from the out requirement of the predecessor.
This probably resolves some mode mismatches in untested code paths -- AMD64 backend and SPARC hard float, I'm looking at you!

Now register requirements always have have a register class.
This is more consistent and removes some null pointer checks.

Nodes without requirements are never operands of a {Copy,}Keep.

match_requirement() cannot be called with a requirement without a register class.

Now we can use things, which get initialized for the middle end, e.g. modes.

The caller of be_opt_register() calls it, too.

Now the template backend has a node, which can represent an adress with an offset.

We never used the distinction which flags get modified, so just simplify the specification a bit.

The next release will feature a cmake build.

It was a nice experiment but I do not plan to keep this up to date.

This is meant for the release because users often expect things like an
install target pkg-config scripts or project files for their favorite
IDE.

Main firm development will stay with the simple Makefile, we only plan
to update the cmake build for releases.

This fixes gcc/pr43784_2.c.

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.