GitLab

This meant initializing the address mode variant all over the ia32
backend, on the other hand this will be useful if we ever switch the
ia32 backend to a style without NoReg nodes (like amd64).

We previously left the amd64_addr_t attribute uninitialized for OP_REG,
OP_REG_REG and OP_REG_IMM. Make the code more consistent by always
initializing it to the newly introduced X86_ADDR_REG variant, as a bonus
we can also use base_input member to indicate the register input instead
of hardcoding it to 0.

The following get_tarval_long(tv) has the same assertion internally.

PC relative relocations need %rip as base register so we cannot match
anything else for base.

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.

It is now required to set a (relocation) kind when entity != NULL.
Adds IMM_ADDR, PICBASE_REL kinds to differentiate between absolute
addresses and picbase relative addresses. Maintain the no_picbase_adjust
flag in parallel for testing. We should be able to remove it in a
subsequent commit.

Now eat_immediate() simply returns, whether it was successful.
So there is less code (duplication) and only one pass is necessary instead of two.

Either it is not supported (-Address) or should not occur due to normalisation (Sub(x, Const)).

* No \n at end; panic() adds one
* No punctuation mark at end
* Lowercase first letter

- This should fix some cases where we would not reassociate constants
  because of multiple users and missed address mode patterns for these
  cases.

This is a common pattern for x % c.