[llvm] [LLVM][MC][DecoderEmitter] Add support to specialize decoder per bitwidth (PR #154865)

[Rahul Joshi via llvm-commits]

================
@@ -700,6 +700,12 @@ static constexpr DecoderListEntry DecoderList32[]{
     {DecoderTableZdinxRV32Only32, {}, "RV32-only Zdinx (Double in Integer)"},
 };
 
+// Define bitwidths for various types used to instantiate the decoder.
+template <> static constexpr uint32_t llvm::MCD::InsnBitWidth<uint16_t> = 16;
+template <> static constexpr uint32_t llvm::MCD::InsnBitWidth<uint32_t> = 32;
+// Use uint64_t to represent 48 bit instructions.
+template <> static constexpr uint32_t llvm::MCD::InsnBitWidth<uint64_t> = 48;
----------------
jurahul wrote:

So something like?

```
template <>
struct TypeForBitWidth<32> {
  using T = uint32_t;
};
template <>
struct TypeForBitWidth<64> {
  using T = uint64_t;
};

int decodeToMCInst(TypeForBitWidth<32>::T Insn) {
  return 32;
}

int decodeToMCInst(TypeForBitWidth<64>::T Insn) {
  return 64;
}

int main() {
  uint32_t x = 0;
  std::cout << decodeToMCInst(x) << '\n';
  uint64_t y = 0;
  std::cout << decodeToMCInst(y) << '\n';
}
```
@topperc and @s-barannikov WDYT? It conveys the same information, but I agree its more directly expressing things. Note that with this though it won't directly work if/when we want to support > 1 bitwidth for a given type (which I did attempt in this PR but gave up because MSVC broke). 

https://github.com/llvm/llvm-project/pull/154865