x64 load-op-store imm folding

Right now the given CLIF

```
v1 = load.i32 v0+32
v2 = iconst.i32 2
v3 = iadd v1, v2
store v3, v0+32
```

generates the following assembly

```
movl $2, %eax
addl %eax, 32(%rdi)
```

However, x64 allows the following encoding

```
addl $2, 32(%rdi)
```

I tried to fix that, by:

1. Change the `src2` of `AluRM` from `Gpr` to `GprMemImm`,
2. Add the corresponding encoding.
3. Change the `alu_rm` constructor to take `GprMemImm`
4. Change the `x64_add_mem`&co constructors to take `GprMemImm`.

This is a bit stupid since `AluRM` already takes a memory operand, so there is no way to encode the `Mem` part of `GprMemImm`. I guess maybe we could introduce `RegImm` and `GprImm` types. But that feels wrong and redundant.

Related issue where this first was brought up https://github.com/bytecodealliance/wasmtime/issues/1925

Start by reading the related issue 1925 and inspecting the named AluRM type, alu_rm constructor, and x64_add_mem constructors. Determine how an immediate source operand can be represented without duplicating memory handling. Done means the given load/add/store CLIF produces the immediate-memory add encoding on x64, with coverage for the new encoding.

Written by the indexing model from the issue text.