[GPU] Remove register reinterpret_casts + WAIT_REG_MEM volatility

Hopefully prevents some potential #1971-like situations.

WAIT_REG_MEM's implementation also allowed the compiler to load the value
only once, which caused an infinite loop with the other changes in the
commit (even in debug builds), so it's now accessed as volatile. Possibly
it would be even better to replace it with some (acquire/release?) atomic
load/store some day at least for the registers actually seen as
participating in those waits.

Also fixes the endianness being handled only on the first wait iteration in
WAIT_REG_MEM.

src/xenia/gpu/shader_interpreter.h

@@ -11,6 +11,7 @@
 #define XENIA_GPU_SHADER_INTERPRETER_H_
 
 #include <algorithm>
+#include <array>
 #include <cstddef>
 #include <cstdint>
 
@@ -120,8 +121,8 @@ class ShaderInterpreter {
   float* GetTempRegister(uint32_t address, bool is_relative) {
     return temp_registers_[GetTempRegisterIndex(address, is_relative)];
   }
-  const float* GetFloatConstant(uint32_t address, bool is_relative,
-                                bool relative_address_is_a0) const;
+  const std::array<float, 4> GetFloatConstant(
+      uint32_t address, bool is_relative, bool relative_address_is_a0) const;
 
   void ExecuteAluInstruction(ucode::AluInstruction instr);
   void StoreFetchResult(uint32_t dest, bool is_dest_relative, uint32_t swizzle,