New category "gnu-libstdc++" provides summary for std::string and synthetic children for types std::map, std::list and std::vector

The category is enabled by default. If you run into issues with it, disable it and the previous behavior of LLDB is restored
 ** This is a temporary solution. The general solution to having formatters pulled in at startup should involve going through the Platform.
Fixed an issue in type synthetic list where a category with synthetic providers in it was not shown if all the providers were regex-based

llvm-svn: 137850

lldb/include/lldb/Core/Debugger.h

@@ -270,6 +270,12 @@ public:
     static void
     SettingsTerminate ();
     
+    static void
+    FormatManagerInitialize();
+    
+    static void
+    FormatManagerTerminate();
+    
     static void
     Destroy (lldb::DebuggerSP &debugger_sp);
 

lldb/lldb.xcodeproj/project.pbxproj

@@ -654,7 +654,6 @@
 		2669605E1199F4230075C61A /* lldb.swig */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = lldb.swig; sourceTree = "<group>"; };
 		266960601199F4230075C61A /* build-swig-Python.sh */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text.script.sh; path = "build-swig-Python.sh"; sourceTree = "<group>"; };
 		266960611199F4230075C61A /* edit-swig-python-wrapper-file.py */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text.script.python; path = "edit-swig-python-wrapper-file.py"; sourceTree = "<group>"; };
-		266960621199F4230075C61A /* finish-swig-Python-lldb.sh */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text.script.sh; path = "finish-swig-Python-lldb.sh"; sourceTree = "<group>"; };
 		266960631199F4230075C61A /* sed-sources */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text.script.perl; path = "sed-sources"; sourceTree = "<group>"; };
 		266A42D5128E3FFB0090CF7C /* ClangNamespaceDecl.cpp */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.cpp.cpp; name = ClangNamespaceDecl.cpp; path = source/Symbol/ClangNamespaceDecl.cpp; sourceTree = "<group>"; };
 		266A42D7128E40040090CF7C /* ClangNamespaceDecl.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; name = ClangNamespaceDecl.h; path = include/lldb/Symbol/ClangNamespaceDecl.h; sourceTree = "<group>"; };
@@ -1179,6 +1178,7 @@
 		94B6E76013D8833C005F417F /* ValueObjectSyntheticFilter.h */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.c.h; name = ValueObjectSyntheticFilter.h; path = include/lldb/Core/ValueObjectSyntheticFilter.h; sourceTree = "<group>"; };
 		94B6E76113D88362005F417F /* ValueObjectSyntheticFilter.cpp */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.cpp.cpp; name = ValueObjectSyntheticFilter.cpp; path = source/Core/ValueObjectSyntheticFilter.cpp; sourceTree = "<group>"; };
 		94EBAC8313D9EE26009BA64E /* PythonPointer.h */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.c.h; name = PythonPointer.h; path = include/lldb/Utility/PythonPointer.h; sourceTree = "<group>"; };
+		94FE476613FC1DA8001F8475 /* finish-swig-Python-LLDB.sh */ = {isa = PBXFileReference; lastKnownFileType = text.script.sh; path = "finish-swig-Python-LLDB.sh"; sourceTree = "<group>"; };
 		961FABB81235DE1600F93A47 /* FuncUnwinders.cpp */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.cpp.cpp; name = FuncUnwinders.cpp; path = source/Symbol/FuncUnwinders.cpp; sourceTree = "<group>"; };
 		961FABB91235DE1600F93A47 /* UnwindPlan.cpp */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.cpp.cpp; name = UnwindPlan.cpp; path = source/Symbol/UnwindPlan.cpp; sourceTree = "<group>"; };
 		961FABBA1235DE1600F93A47 /* UnwindTable.cpp */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.cpp.cpp; name = UnwindTable.cpp; path = source/Symbol/UnwindTable.cpp; sourceTree = "<group>"; };
@@ -1734,7 +1734,7 @@
 			children = (
 				266960601199F4230075C61A /* build-swig-Python.sh */,
 				266960611199F4230075C61A /* edit-swig-python-wrapper-file.py */,
-				266960621199F4230075C61A /* finish-swig-Python-lldb.sh */,
+				94FE476613FC1DA8001F8475 /* finish-swig-Python-LLDB.sh */,
 				9A48A3A7124AAA5A00922451 /* python-extensions.swig */,
 				94005E0313F438DF001EF42D /* python-wrapper.swig */,
 			);

lldb/scripts/Python/finish-swig-Python-LLDB.sh

@@ -165,5 +165,20 @@ else
     fi
 fi
 
+# Copy the OSX C++ STL formatters over to the framework Python directory
+if [ -f "${SRC_ROOT}/source/osxcpp.py" ]
+then
+    if [ $Debug == 1 ]
+    then
+        echo "Copying osxcpp.py to ${framework_python_dir}"
+    fi
+    cp "${SRC_ROOT}/source/osxcpp.py" "${framework_python_dir}"
+else
+    if [ $Debug == 1 ]
+    then
+        echo "Unable to find ${SRC_ROOT}/source/osxcpp.py"
+    fi
+fi
+
 exit 0
 

lldb/source/Commands/CommandObjectType.cpp

@@ -1538,7 +1538,7 @@ private:
         CommandReturnObject* result = param->result;
         
         // if the category is disabled or empty and there is no regex, just skip it
-        if ((cate->IsEnabled() == false || cate->GetCount() == 0) && param->cate_regex == NULL)
+        if ((cate->IsEnabled() == false || cate->GetCount(FormatCategory::eSummary | FormatCategory::eRegexSummary) == 0) && param->cate_regex == NULL)
             return true;
         
         // if we have a regex and this category does not match it, just skip it
@@ -2032,7 +2032,7 @@ private:
         CommandReturnObject* result = param->result;
         
         // if the category is disabled or empty and there is no regex, just skip it
-        if ((cate->IsEnabled() == false || cate->Filter()->GetCount() == 0) && param->cate_regex == NULL)
+        if ((cate->IsEnabled() == false || cate->GetCount(FormatCategory::eFilter | FormatCategory::eRegexFilter) == 0) && param->cate_regex == NULL)
             return true;
         
         // if we have a regex and this category does not match it, just skip it
@@ -2240,7 +2240,7 @@ private:
         CommandReturnObject* result = param->result;
         
         // if the category is disabled or empty and there is no regex, just skip it
-        if ((cate->IsEnabled() == false || cate->Synth()->GetCount() == 0) && param->cate_regex == NULL)
+        if ((cate->IsEnabled() == false || cate->GetCount(FormatCategory::eSynth | FormatCategory::eRegexSynth) == 0) && param->cate_regex == NULL)
             return true;
         
         // if we have a regex and this category does not match it, just skip it

lldb/source/Core/Debugger.cpp

@@ -82,6 +82,7 @@ Debugger::Initialize ()
     if (g_shared_debugger_refcount == 0)
     {
         lldb_private::Initialize();
+        Debugger::FormatManagerInitialize();
     }
     g_shared_debugger_refcount++;
 
@@ -95,6 +96,7 @@ Debugger::Terminate ()
         g_shared_debugger_refcount--;
         if (g_shared_debugger_refcount == 0)
         {
+            Debugger::FormatManagerTerminate();
             lldb_private::WillTerminate();
             lldb_private::Terminate();
 
@@ -1781,6 +1783,72 @@ GetFormatManager() {
     return g_format_manager;
 }
 
+// The platform should be responsible for initializing its own formatters
+// (e.g. to handle versioning, different runtime libraries, ...)
+// Currently, basic formatters for std:: objects as implemented by
+// the GNU libstdc++ are defined regardless, and enabled by default
+// This is going to be moved to some platform-dependent location
+// (in the meanwhile, these formatters should work for Mac OS X & Linux)
+void
+Debugger::FormatManagerInitialize()
+{
+    static bool g_initialized = false;
+    
+    if (!g_initialized)
+    {
+        g_initialized = true;
+        ConstString gnulib("gnu-libstdc++");
+        FormatManager& format_mgr = GetFormatManager();
+        lldb::FormatCategorySP osxcpp = format_mgr.Category(gnulib.AsCString());
+        osxcpp->Summary()->Add(ConstString("std::string").AsCString(),
+                               SummaryFormatSP(new StringSummaryFormat(true,
+                                                                       false,
+                                                                       false,
+                                                                       true,
+                                                                       true,
+                                                                       false,
+                                                                       "${var._M_dataplus._M_p}")));
+        osxcpp->Summary()->Add(ConstString("std::basic_string<char>").AsCString(),
+                               SummaryFormatSP(new StringSummaryFormat(true,
+                                                                       false,
+                                                                       false,
+                                                                       true,
+                                                                       true,
+                                                                       false,
+                                                                       "${var._M_dataplus._M_p}")));
+        osxcpp->Summary()->Add(ConstString("std::basic_string<char,std::char_traits<char>,std::allocator<char> >").AsCString(),
+                               SummaryFormatSP(new StringSummaryFormat(true,
+                                                                       false,
+                                                                       false,
+                                                                       true,
+                                                                       true,
+                                                                       false,
+                                                                       "${var._M_dataplus._M_p}")));
+        osxcpp->RegexSynth()->Add(RegularExpressionSP(new RegularExpression("std::vector<")),
+                                    SyntheticChildrenSP(new SyntheticScriptProvider(true,
+                                                                                    false,
+                                                                                    false,
+                                                                                    "StdVectorSynthProvider")));
+        osxcpp->RegexSynth()->Add(RegularExpressionSP(new RegularExpression("std::map<")),
+                                    SyntheticChildrenSP(new SyntheticScriptProvider(true,
+                                                                                    false,
+                                                                                    false,
+                                                                                    "StdMapSynthProvider")));
+        osxcpp->RegexSynth()->Add(RegularExpressionSP(new RegularExpression("std::list<")),
+                                    SyntheticChildrenSP(new SyntheticScriptProvider(true,
+                                                                                    false,
+                                                                                    false,
+                                                                                    "StdListSynthProvider")));
+        
+        format_mgr.EnableCategory(gnulib.AsCString());
+        
+    }
+}
+
+void
+Debugger::FormatManagerTerminate()
+{}
+
 void
 Debugger::Formatting::ForceUpdate()
 {

lldb/source/Interpreter/ScriptInterpreterPython.cpp

@@ -173,6 +173,11 @@ ScriptInterpreterPython::ScriptInterpreterPython (CommandInterpreter &interprete
                        interpreter.GetDebugger().GetID());
     PyRun_SimpleString (run_string.GetData());
     
+    run_string.Clear();
+    run_string.Printf ("run_one_line (%s, 'from osxcpp import *')", m_dictionary_name.c_str(),
+                       interpreter.GetDebugger().GetID());
+    PyRun_SimpleString (run_string.GetData());
+    
     if (m_dbg_stdout != NULL)
     {
         m_new_sysout = PyFile_FromFile (m_dbg_stdout, (char *) "", (char *) "w", _check_and_flush);

lldb/source/Plugins/LanguageRuntime/ObjC/AppleObjCRuntime/AppleObjCRuntimeV2.cpp

@@ -656,10 +656,7 @@ AppleObjCRuntimeV2::GetActualTypeName(lldb_private::ObjCLanguageRuntime::ObjCISA
      struct class_t *superclass;
      Cache cache;
      IMP *vtable;
--->     uintptr_t data_NEVER_USE;
-     WARNING: this data_NEVER_USE pointer might one day contain flags in the least-significant bits
-     currently, rdar://problem/8955342 prevents the runtime from doing so
-     it presently is just a pointer to a class_rw_t
+-->     class_rw_t data;
      */
     
     lldb::addr_t rw_pointer = isa + (4 * pointer_size);

lldb/source/osxcpp.py

@@ -0,0 +1,231 @@
+import re
+
+# C++ STL formatters for LLDB
+# These formatters are based upon the version of the STL that ships
+# with Mac OS X Snow Leopard 10.6.8 and OS X Lion 10.7.0
+# The STL implementation *might* change on other releases of Apple's
+# operating system and library infrastructure, and might be different on
+# other operating systems
+
+class StdListSynthProvider:
+
+	def __init__(self, valobj, dict):
+		self.valobj = valobj
+		self.update()
+
+	def num_children(self):
+		next_val = self.next.GetValueAsUnsigned(0)
+		prev_val = self.prev.GetValueAsUnsigned(0)
+		# After a std::list has been initialized, both next and prev will be non-NULL
+		if next_val == 0 or prev_val == 0:
+			return 0
+		if next_val == self.node_address:
+			return 0
+		if next_val == prev_val:
+			return 1
+		size = 2
+		current = self.next
+		while current.GetChildMemberWithName('_M_next').GetValueAsUnsigned(0) != self.node_address:
+			size = size + 1
+			current = current.GetChildMemberWithName('_M_next')
+		return (size - 1)
+
+	def get_child_index(self,name):
+		return int(name.lstrip('[').rstrip(']'))
+
+	def get_child_at_index(self,index):
+		if index >= self.num_children():
+			return None;
+		offset = index
+		current = self.next
+		while offset > 0:
+			current = current.GetChildMemberWithName('_M_next')
+			offset = offset - 1
+		return current.CreateChildAtOffset('['+str(index)+']',2*current.GetType().GetByteSize(),self.data_type)
+
+	def extract_type_name(self,name):
+		self.type_name = name[16:]
+		index = 2
+		count_of_template = 1
+		while index < len(self.type_name):
+			if self.type_name[index] == '<':
+				count_of_template = count_of_template + 1
+			elif self.type_name[index] == '>':
+				count_of_template = count_of_template - 1
+			elif self.type_name[index] == ',' and count_of_template == 1:
+				self.type_name = self.type_name[:index]
+				break
+			index = index + 1
+		self.type_name_nospaces = self.type_name.replace(", ", ",")
+
+	def update(self):
+		impl = self.valobj.GetChildMemberWithName('_M_impl')
+		node = impl.GetChildMemberWithName('_M_node')
+		self.extract_type_name(impl.GetType().GetName())
+		self.node_address = self.valobj.AddressOf().GetValueAsUnsigned(0)
+		self.next = node.GetChildMemberWithName('_M_next')
+		self.prev = node.GetChildMemberWithName('_M_prev')
+		self.data_type = node.GetTarget().FindFirstType(self.type_name)
+		# tries to fight against a difference in formatting type names between gcc and clang
+		if self.data_type.IsValid() == False:
+			self.data_type = node.GetTarget().FindFirstType(self.type_name_nospaces)
+		self.data_size = self.data_type.GetByteSize()
+
+class StdVectorSynthProvider:
+
+	def __init__(self, valobj, dict):
+		self.valobj = valobj;
+		self.update()
+
+	def num_children(self):
+		start_val = self.start.GetValueAsUnsigned(0)
+		finish_val = self.finish.GetValueAsUnsigned(0)
+		end_val  = self.end.GetValueAsUnsigned(0)
+		# Before a vector has been constructed, it will contain bad values
+		# so we really need to be careful about the length we return since
+		# unitialized data can cause us to return a huge number. We need
+		# to also check for any of the start, finish or end of storage values
+		# being zero (NULL). If any are, then this vector has not been 
+		# initialized yet and we should return zero
+
+		# Make sure nothing is NULL
+		if start_val == 0 or finish_val == 0 or end_val == 0:
+			return 0
+		# Make sure start is less than finish
+		if start_val >= finish_val:
+			return 0
+		# Make sure finish is less than or equal to end of storage
+		if finish_val > end_val:
+			return 0
+
+		num_children = (finish_val-start_val)/self.data_size
+		return num_children
+
+	def get_child_index(self,name):
+		return int(name.lstrip('[').rstrip(']'))
+
+	def get_child_at_index(self,index):
+		if index >= self.num_children():
+			return None;
+		offset = index * self.data_size
+		return self.start.CreateChildAtOffset('['+str(index)+']',offset,self.data_type)
+
+	def update(self):
+		impl = self.valobj.GetChildMemberWithName('_M_impl')
+		self.start = impl.GetChildMemberWithName('_M_start')
+		self.finish = impl.GetChildMemberWithName('_M_finish')
+		self.end = impl.GetChildMemberWithName('_M_end_of_storage')
+		self.data_type = self.start.GetType().GetPointeeType()
+		self.data_size = self.data_type.GetByteSize()
+
+
+class StdMapSynthProvider:
+
+	def __init__(self, valobj, dict):
+		self.valobj = valobj;
+		self.update()
+
+	def update(self):
+		self.Mt = self.valobj.GetChildMemberWithName('_M_t')
+		self.Mimpl = self.Mt.GetChildMemberWithName('_M_impl')
+		self.Mheader = self.Mimpl.GetChildMemberWithName('_M_header')
+		# from libstdc++ implementation of _M_root for rbtree
+		self.Mroot = self.Mheader.GetChildMemberWithName('_M_parent')
+		# the stuff into the tree is actually a std::pair<const key, value>
+		# life would be much easier if gcc had a coherent way to print out
+		# template names in debug info
+		self.expand_clang_type_name()
+		self.expand_gcc_type_name()
+		self.data_type = self.Mt.GetTarget().FindFirstType(self.clang_type_name)
+		if self.data_type.IsValid() == False:
+			self.data_type = self.Mt.GetTarget().FindFirstType(self.gcc_type_name)
+		self.data_size = self.data_type.GetByteSize()
+		self.skip_size = self.Mheader.GetType().GetByteSize()
+
+	def expand_clang_type_name(self):
+		type_name = self.Mimpl.GetType().GetName()
+		index = type_name.find("std::pair<")
+		type_name = type_name[index+5:]
+		index = 6
+		template_count = 1
+		while index < len(type_name):
+			if type_name[index] == '<':
+				template_count = template_count + 1
+			elif type_name[index] == '>' and template_count == 1:
+				type_name = type_name[:index+1]
+				break
+			elif type_name[index] == '>':
+				template_count = template_count - 1
+			index = index + 1;
+		self.clang_type_name = type_name
+
+	def expand_gcc_type_name(self):
+		type_name = self.Mt.GetType().GetName()
+		index = type_name.find("std::pair<")
+		type_name = type_name[index+5:]
+		index = 6
+		template_count = 1
+		while index < len(type_name):
+			if type_name[index] == '<':
+				template_count = template_count + 1
+			elif type_name[index] == '>' and template_count == 1:
+				type_name = type_name[:index+1]
+				break
+			elif type_name[index] == '>':
+				template_count = template_count - 1
+			elif type_name[index] == ' ' and template_count == 1 and type_name[index-1] == ',':
+			    type_name = type_name[0:index] + type_name[index+1:]
+			    index = index - 1
+			index = index + 1;
+		self.gcc_type_name = type_name
+
+	def num_children(self):
+		root_ptr_val = self.node_ptr_value(self.Mroot)
+		if root_ptr_val == 0:
+			return 0;
+		return self.Mimpl.GetChildMemberWithName('_M_node_count').GetValueAsUnsigned(0)
+
+	def get_child_index(self,name):
+		return int(name.lstrip('[').rstrip(']'))
+
+	def get_child_at_index(self,index):
+		if index >= self.num_children():
+			return None;
+		offset = index
+		current = self.left(self.Mheader);
+		while offset > 0:
+			current = self.increment_node(current)
+			offset = offset - 1;
+		# skip all the base stuff and get at the data
+		return current.CreateChildAtOffset('['+str(index)+']',self.skip_size,self.data_type)
+
+	# utility functions
+	def node_ptr_value(self,node):
+		return node.GetValueAsUnsigned(0)
+
+	def right(self,node):
+		return node.GetChildMemberWithName("_M_right");
+
+	def left(self,node):
+		return node.GetChildMemberWithName("_M_left");
+
+	def parent(self,node):
+		return node.GetChildMemberWithName("_M_parent");
+
+	# from libstdc++ implementation of iterator for rbtree
+	def increment_node(self,node):
+		if self.node_ptr_value(self.right(node)) != 0:
+			x = self.right(node);
+			while self.node_ptr_value(self.left(x)) != 0:
+				x = self.left(x);
+			return x;
+		else:
+			x = node;
+			y = self.parent(x)
+			while(self.node_ptr_value(x) == self.node_ptr_value(self.right(y))):
+				x = y;
+				y = self.parent(y);
+			if self.node_ptr_value(self.right(x)) != self.node_ptr_value(y):
+				x = y;
+			return x;
+

lldb/test/functionalities/data-formatter/data-formatter-python-synth/TestDataFormatterPythonSynth.py

@@ -222,9 +222,9 @@ class DataFormatterTestCase(TestBase):
         
         
         # std::vector
-        self.runCmd("script from StdVectorSynthProvider import *")
-        self.runCmd("type synth add -l StdVectorSynthProvider std::int_vect int_vect")
-        self.runCmd("type synth add -l StdVectorSynthProvider std::string_vect string_vect")
+        #self.runCmd("script from StdVectorSynthProvider import *")
+        #self.runCmd("type synth add -l StdVectorSynthProvider std::int_vect int_vect")
+        #self.runCmd("type synth add -l StdVectorSynthProvider std::string_vect string_vect")
 
         self.runCmd("n")
 
@@ -351,12 +351,12 @@ class DataFormatterTestCase(TestBase):
             substrs = ['vector has 0 items'])
 
         # now test std::list
-        self.runCmd("script from StdListSynthProvider import *")
+        #self.runCmd("script from StdListSynthProvider import *")
 
         self.runCmd("n")
 
         self.runCmd("frame variable numbers_list -T")
-        self.runCmd("type synth add std::int_list std::string_list int_list string_list -l StdListSynthProvider")
+        #self.runCmd("type synth add std::int_list std::string_list int_list string_list -l StdListSynthProvider")
         self.runCmd("type summary add std::int_list std::string_list int_list string_list -f \"list has ${svar%#} items\" -e")
         self.runCmd("type format add -f hex int")
 
@@ -467,13 +467,13 @@ class DataFormatterTestCase(TestBase):
         self.runCmd("n")
         self.runCmd("frame variable ii -T")
         
-        self.runCmd("script from StdMapSynthProvider import *")
+        #self.runCmd("script from StdMapSynthProvider import *")
         self.runCmd("type summary add -x \"std::map<\" -f \"map has ${svar%#} items\" -e") 
         
         #import time
         #time.sleep(30)
         
-        self.runCmd("type synth add -x \"std::map<\" -l StdMapSynthProvider")
+        #self.runCmd("type synth add -x \"std::map<\" -l StdMapSynthProvider")
 
 
         self.expect('frame variable ii',