Wrap op (de)serialization methods in anonymous namespace

It's a known bug that older GCC is not happy with method specialization in
the enclosing (global) namespace:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56480

This CL wraps the generated specialization methods in the anonymous namespace
to make sure the specialization is in the same namespace as the class.

PiperOrigin-RevId: 258983181

mlir/lib/Dialect/SPIRV/Serialization/Deserializer.cpp

@@ -462,10 +462,12 @@ LogicalResult Deserializer::processInstruction(spirv::Opcode opcode,
   return dispatchToAutogenDeserialization(opcode, operands);
 }
 
+namespace {
 // Pull in auto-generated Deserializer::dispatchToAutogenDeserialization() and
 // various processOpImpl specializations.
 #define GET_DESERIALIZATION_FNS
 #include "mlir/Dialect/SPIRV/SPIRVSerialization.inc"
+} // namespace
 
 Optional<spirv::ModuleOp> spirv::deserialize(ArrayRef<uint32_t> binary,
                                              MLIRContext *context) {

mlir/lib/Dialect/SPIRV/Serialization/Serializer.cpp

@@ -409,10 +409,12 @@ LogicalResult Serializer::processOperation(Operation *op) {
   return dispatchToAutogenSerialization(op);
 }
 
+namespace {
 // Pull in auto-generated Serializer::dispatchToAutogenSerialization() and
 // various processOpImpl specializations.
 #define GET_SERIALIZATION_FNS
 #include "mlir/Dialect/SPIRV/SPIRVSerialization.inc"
+} // namespace
 
 LogicalResult spirv::serialize(spirv::ModuleOp module,
                                SmallVectorImpl<uint32_t> &binary) {