From ee500220baa1a46f3d7ca1486fa43dc1587f20d0 Mon Sep 17 00:00:00 2001
From: WeiN76LQh <WeiN76LQh@github.com>
Date: Fri, 3 Jan 2025 17:55:10 +0000
Subject: [ObjC] Create a shared ObjC processor for Macho and DSC views

Both the Macho and DSC views need to process Objective-C but have separate processor classes. It would appear that the DSC version was largely a copy and paste of the Macho view one, with some modifications. The majority of code overlaps between the 2 so it doesn't make sense to maintain 2 and copy and paste improvements/fixes between them.

This commit fixes that by creating a base Objective-C processor that contains the shared code. View specific code is implemented in the respective subclasses for the views. Although there is very little view specific code for each.
---
 view/sharedcache/core/ObjC.h | 248 ++++++-------------------------------------
 1 file changed, 31 insertions(+), 217 deletions(-)

(limited to 'view/sharedcache/core/ObjC.h')

diff --git a/view/sharedcache/core/ObjC.h b/view/sharedcache/core/ObjC.h
index a308c23a..13790364 100644
--- a/view/sharedcache/core/ObjC.h
+++ b/view/sharedcache/core/ObjC.h
@@ -6,233 +6,47 @@
 #define SHAREDCACHE_OBJC_H
 
 #include <binaryninjaapi.h>
-#include "VM.h"
+#include <objectivec/objc.h>
 #include "SharedCache.h"
 
-using namespace BinaryNinja;
-
-
 namespace DSCObjC {
+	class DSCObjCReader : public ObjCReader {
+	private:
+		VMReader m_reader;
+		size_t m_addressSize;
 
-	// This set of structs is based on the objc4 source,
-	// 		however pointers have been replaced with view_ptr_t
-
-	// Used for pointers within BinaryView, primarily to make it far more clear in typedefs
-	// 		whether the size of a field can vary between architectures.
-	// These should _not_ be used in sizeof or direct Read() calls.
-	typedef uint64_t view_ptr_t;
-
-	typedef struct {
-		view_ptr_t name;
-		view_ptr_t types;
-		view_ptr_t imp;
-	} method_t;
-	typedef struct {
-		uint32_t name;
-		uint32_t types;
-		uint32_t imp;
-	} method_entry_t;
-	typedef struct {
-		view_ptr_t offset;
-		view_ptr_t name;
-		view_ptr_t type;
-		uint32_t alignmentRaw;
-		uint32_t size;
-	} ivar_t;
-	typedef struct {
-		view_ptr_t name;
-		view_ptr_t attributes;
-	} property_t;
-	typedef struct {
-		uint32_t entsizeAndFlags;
-		uint32_t count;
-	} method_list_t;
-	typedef struct {
-		uint32_t entsizeAndFlags;
-		uint32_t count;
-	} ivar_list_t;
-	typedef struct {
-		uint32_t entsizeAndFlags;
-		uint32_t count;
-	} property_list_t;
-	typedef struct {
-		uint64_t count;
-	} protocol_list_t;
-	struct relative_list_list_entry_t {
-	    uint64_t imageIndex: 16;
-	    int64_t listOffset: 48;
-	};
-	typedef struct {
-		view_ptr_t isa;
-		view_ptr_t mangledName;
-		view_ptr_t protocols;
-		view_ptr_t instanceMethods;
-		view_ptr_t classMethods;
-		view_ptr_t optionalInstanceMethods;
-		view_ptr_t optionalClassMethods;
-		view_ptr_t instanceProperties;
-		uint32_t size;
-		uint32_t flags;
-	} protocol_t;
-	typedef struct {
-		uint32_t flags;
-		uint32_t instanceStart;
-		uint32_t instanceSize;
-		uint32_t reserved;
-		view_ptr_t ivarLayout;
-		view_ptr_t name;
-		view_ptr_t baseMethods;
-		view_ptr_t baseProtocols;
-		view_ptr_t ivars;
-		view_ptr_t weakIvarLayout;
-		view_ptr_t baseProperties;
-	} class_ro_t;
-	typedef struct {
-		view_ptr_t isa;
-		view_ptr_t super;
-		view_ptr_t cache;
-		view_ptr_t vtable;
-		view_ptr_t data;
-	} class_t;
-	typedef struct {
-		view_ptr_t name;
-		view_ptr_t cls;
-		view_ptr_t instanceMethods;
-		view_ptr_t classMethods;
-		view_ptr_t protocols;
-		view_ptr_t instanceProperties;
-	} category_t;
-	typedef struct {
-		view_ptr_t receiver;
-		view_ptr_t current_class;
-	} objc_super2;
-	typedef struct {
-		view_ptr_t imp;
-		view_ptr_t sel;
-	} message_ref_t;
-
-	struct Method {
-		std::string name;
-		std::string types;
-		view_ptr_t imp;
-	};
-
-	struct Ivar {
-		uint32_t offset;
-		std::string name;
-		std::string type;
-		uint32_t alignment;
-		uint32_t size;
-	};
-
-	struct Property {
-		std::string name;
-		std::string attributes;
-	};
-
-	struct ClassBase {
-		std::map<uint64_t, Method> methodList;
-		std::map<uint64_t, Ivar> ivarList;
-	};
-
-	struct Class {
-		std::string name;
-		ClassBase instanceClass;
-		ClassBase metaClass;
-
-		// Loaded by type processing
-		QualifiedName associatedName;
-	};
-
-	class Protocol {
 	public:
-		std::string name;
-		std::vector<QualifiedName> protocols;
-		ClassBase instanceMethods;
-		ClassBase classMethods;
-		ClassBase optionalInstanceMethods;
-		ClassBase optionalClassMethods;
-	};
-
-	struct QualifiedNameOrType {
-		BinaryNinja::Ref<BinaryNinja::Type> type = nullptr;
-		BinaryNinja::QualifiedName name;
-		size_t ptrCount = 0;
-	};
-
-	class DSCObjCProcessor {
-		struct Types {
-			// QualifiedName relativePtr;
-			// QualifiedName id;
-			// QualifiedName sel;
-			// QualifiedName BOOL;
-			// QualifiedName nsInteger;
-			// QualifiedName nsuInteger;
-			// QualifiedName cgFloat;
-			QualifiedName cfStringFlag;
-			QualifiedName cfString;
-			QualifiedName cfStringUTF16;
-			QualifiedName imageInfoFlags;
-			QualifiedName imageInfoSwiftVersion;
-			QualifiedName imageInfo;
-			QualifiedName methodEntry;
-			QualifiedName method;
-			QualifiedName methodList;
-			QualifiedName classRO;
-			QualifiedName cls;
-			QualifiedName category;
-			QualifiedName protocol;
-			QualifiedName protocolList;
-			QualifiedName ivar;
-			QualifiedName ivarList;
-		} m_typeNames;
-
-		bool m_isBackedByDatabase;
-
-		BinaryView* m_data;
-		SymbolQueue* m_symbolQueue = nullptr;
-		Ref<Logger> m_logger;
-		std::map<uint64_t, Class> m_classes;
-		std::map<uint64_t, Class> m_categories;
-		std::map<uint64_t, Protocol> m_protocols;
-		std::unordered_map<uint64_t, std::string> m_selectorCache;
-		std::unordered_map<uint64_t, Method> m_localMethods;
-
-		// Required for workflow_objc type heuristics, should be removed when that is no longer a thing.
-		std::map<uint64_t, std::string> m_selRefToName;
-		std::map<uint64_t, std::vector<uint64_t>> m_selRefToImplementations;
-		std::map<uint64_t, std::vector<uint64_t>> m_selToImplementations;
-		// --
-
-
+		void Read(void* dest, size_t len) override;
+		std::string ReadCString() override;
+		uint8_t Read8() override;
+		uint16_t Read16() override;
+		uint32_t Read32() override;
+		uint64_t Read64() override;
+		int8_t ReadS8() override;
+		int16_t ReadS16() override;
+		int32_t ReadS32() override;
+		int64_t ReadS64() override;
+		uint64_t ReadPointer() override;
+		uint64_t GetOffset() const override;
+		void Seek(uint64_t offset) override;
+		void SeekRelative(int64_t offset) override;
+
+		VMReader& GetVMReader();
+
+		DSCObjCReader(SharedCacheCore::SharedCache* cache, size_t addressSize);
+	};
+
+	class DSCObjCProcessor : public ObjCProcessor {
 		std::optional<uint64_t> m_customRelativeMethodSelectorBase = std::nullopt;
 		SharedCacheCore::SharedCache* m_cache;
 
-		uint64_t ReadPointerAccountingForRelocations(VMReader* reader);
-		std::unordered_map<uint64_t, uint64_t> m_relocationPointerRewrites;
-
-		static Ref<Metadata> SerializeMethod(uint64_t loc, const Method& method);
-		static Ref<Metadata> SerializeClass(uint64_t loc, const Class& cls);
-
-		Ref<Metadata> SerializeMetadata();
-		std::vector<QualifiedNameOrType> ParseEncodedType(const std::string& type);
-		void DefineObjCSymbol(BNSymbolType symbolType, QualifiedName typeName, const std::string& name, uint64_t addr, bool deferred);
-		void DefineObjCSymbol(BNSymbolType symbolType, Ref<Type> type, const std::string& name, uint64_t addr, bool deferred);
-		void ReadIvarList(VMReader* reader, ClassBase& cls, std::string_view name, view_ptr_t start);
-		void ReadMethodList(VMReader* reader, ClassBase& cls, std::string_view name, view_ptr_t start);
-		void ReadListOfMethodLists(VMReader* reader, ClassBase& cls, std::string_view name, view_ptr_t start);
-		void LoadClasses(VMReader* reader, Ref<Section> listSection);
-		void LoadCategories(VMReader* reader, Ref<Section> listSection);
-		void LoadProtocols(VMReader* reader, Ref<Section> listSection);
-		void GenerateClassTypes();
-		bool ApplyMethodType(Class& cls, Method& method, bool isInstanceMethod);
-		void ApplyMethodTypes(Class& cls);
-		void PostProcessObjCSections(VMReader* reader, std::string baseName);
+		std::shared_ptr<ObjCReader> GetReader() override;
+		void GetRelativeMethod(ObjCReader* reader, method_t& meth) override;
+	
 	public:
 		DSCObjCProcessor(BinaryView* data, SharedCacheCore::SharedCache* cache, bool isBackedByDatabase);
-		void ProcessObjCData(std::shared_ptr<VM> vm, std::string baseName);
-		void ProcessCFStrings(std::shared_ptr<VM> vm, std::string baseName);
-		void AddRelocatedPointer(uint64_t location, uint64_t rewrite);
+		
+		uint64_t GetObjCRelativeMethodBaseAddress(ObjCReader* reader) override;
 	};
 }
 #endif //SHAREDCACHE_OBJC_H
-- 
cgit v1.3.1