Merged main:82ebd333a889d2372c8445dc3d5d527ec48537db into amd-gfx:0e8…

…d0cecbbff

Local branch amd-gfx 0e8d0ce Merged main:438ad9f2bf25575c474313de4ad85a5da6f69e4c into amd-gfx:b82566cedfb0
Remote branch main 82ebd33 [LLDB][Minidumps] Read x64 registers as 64b and handle truncation in the file builder (llvm#106473)

clang-tools-extra/clangd/CollectMacros.cpp

@@ -32,6 +32,7 @@ void CollectMainFileMacros::add(const Token &MacroNameTok, const MacroInfo *MI,
   if (Loc.isInvalid() || Loc.isMacroID())
     return;
 
+  assert(isInsideMainFile(Loc, SM));
   auto Name = MacroNameTok.getIdentifierInfo()->getName();
   Out.Names.insert(Name);
   size_t Start = SM.getFileOffset(Loc);

clang-tools-extra/clangd/CollectMacros.h

@@ -82,6 +82,14 @@ class CollectMainFileMacros : public PPCallbacks {
 
   void SourceRangeSkipped(SourceRange R, SourceLocation EndifLoc) override;
 
+  // Called when the AST build is done to disable further recording
+  // of macros by this class. This is needed because some clang-tidy
+  // checks can trigger PP callbacks by calling directly into the
+  // preprocessor. Such calls are not interleaved with FileChanged()
+  // in the expected way, leading this class to erroneously process
+  // macros that are not in the main file.
+  void doneParse() { InMainFile = false; }
+
 private:
   void add(const Token &MacroNameTok, const MacroInfo *MI,
            bool IsDefinition = false, bool InConditionalDirective = false);

clang-tools-extra/clangd/ParsedAST.cpp

@@ -688,7 +688,9 @@ ParsedAST::build(llvm::StringRef Filename, const ParseInputs &Inputs,
     Marks = Patch->marks();
   }
   auto &PP = Clang->getPreprocessor();
-  PP.addPPCallbacks(std::make_unique<CollectMainFileMacros>(PP, Macros));
+  auto MacroCollector = std::make_unique<CollectMainFileMacros>(PP, Macros);
+  auto *MacroCollectorPtr = MacroCollector.get(); // so we can call doneParse()
+  PP.addPPCallbacks(std::move(MacroCollector));
 
   PP.addPPCallbacks(
       collectPragmaMarksCallback(Clang->getSourceManager(), Marks));
@@ -709,6 +711,10 @@ ParsedAST::build(llvm::StringRef Filename, const ParseInputs &Inputs,
     log("Execute() failed when building AST for {0}: {1}", MainInput.getFile(),
         toString(std::move(Err)));
 
+  // Disable the macro collector for the remainder of this function, e.g.
+  // clang-tidy checkers.
+  MacroCollectorPtr->doneParse();
+
   // We have to consume the tokens before running clang-tidy to avoid collecting
   // tokens from running the preprocessor inside the checks (only
   // modernize-use-trailing-return-type does that today).

clang-tools-extra/clangd/unittests/DiagnosticsTests.cpp

@@ -940,6 +940,23 @@ TEST(DiagnosticTest, ClangTidySelfContainedDiagsFormatting) {
                         withFix(equalToFix(ExpectedFix2))))));
 }
 
+TEST(DiagnosticsTest, ClangTidyCallingIntoPreprocessor) {
+  std::string Main = R"cpp(
+    extern "C" {
+    #include "b.h"
+    }
+  )cpp";
+  std::string Header = R"cpp(
+    #define EXTERN extern
+    EXTERN int waldo();
+  )cpp";
+  auto TU = TestTU::withCode(Main);
+  TU.AdditionalFiles["b.h"] = Header;
+  TU.ClangTidyProvider = addTidyChecks("modernize-use-trailing-return-type");
+  // Check that no assertion failures occur during the build
+  TU.build();
+}
+
 TEST(DiagnosticsTest, Preprocessor) {
   // This looks like a preamble, but there's an #else in the middle!
   // Check that:

lldb/source/Plugins/ObjectFile/Minidump/MinidumpFileBuilder.cpp

@@ -491,12 +491,14 @@ GetThreadContext_x86_64(RegisterContext *reg_ctx) {
   thread_context.r14 = read_register_u64(reg_ctx, "r14");
   thread_context.r15 = read_register_u64(reg_ctx, "r15");
   thread_context.rip = read_register_u64(reg_ctx, "rip");
-  thread_context.eflags = read_register_u32(reg_ctx, "rflags");
-  thread_context.cs = read_register_u16(reg_ctx, "cs");
-  thread_context.fs = read_register_u16(reg_ctx, "fs");
-  thread_context.gs = read_register_u16(reg_ctx, "gs");
-  thread_context.ss = read_register_u16(reg_ctx, "ss");
-  thread_context.ds = read_register_u16(reg_ctx, "ds");
+  // To make our code agnostic to whatever type the register value identifies
+  // itself as, we read as a u64 and truncate to u32/u16 ourselves.
+  thread_context.eflags = read_register_u64(reg_ctx, "rflags");
+  thread_context.cs = read_register_u64(reg_ctx, "cs");
+  thread_context.fs = read_register_u64(reg_ctx, "fs");
+  thread_context.gs = read_register_u64(reg_ctx, "gs");
+  thread_context.ss = read_register_u64(reg_ctx, "ss");
+  thread_context.ds = read_register_u64(reg_ctx, "ds");
   return thread_context;
 }
 

lldb/test/API/functionalities/process_save_core_minidump/TestProcessSaveCoreMinidump.py

@@ -17,6 +17,7 @@ def verify_core_file(
         expected_modules,
         expected_threads,
         stacks_to_sps_map,
+        stacks_to_registers_map,
     ):
         # To verify, we'll launch with the mini dump
         target = self.dbg.CreateTarget(None)
@@ -62,6 +63,17 @@ def verify_core_file(
             # Try to read just past the red zone and fail
             process.ReadMemory(sp - red_zone - 1, 1, error)
             self.assertTrue(error.Fail(), "No failure when reading past the red zone")
+            # Verify the registers are the same
+            self.assertIn(thread_id, stacks_to_registers_map)
+            register_val_list = stacks_to_registers_map[thread_id]
+            frame_register_list = frame.GetRegisters()
+            for x in register_val_list:
+                self.assertEqual(
+                    x.GetValueAsUnsigned(),
+                    frame_register_list.GetFirstValueByName(
+                        x.GetName()
+                    ).GetValueAsUnsigned(),
+                )
 
         self.dbg.DeleteTarget(target)
 
@@ -93,12 +105,16 @@ def test_save_linux_mini_dump(self):
             expected_number_of_threads = process.GetNumThreads()
             expected_threads = []
             stacks_to_sp_map = {}
+            stakcs_to_registers_map = {}
 
             for thread_idx in range(process.GetNumThreads()):
                 thread = process.GetThreadAtIndex(thread_idx)
                 thread_id = thread.GetThreadID()
                 expected_threads.append(thread_id)
                 stacks_to_sp_map[thread_id] = thread.GetFrameAtIndex(0).GetSP()
+                stakcs_to_registers_map[thread_id] = thread.GetFrameAtIndex(
+                    0
+                ).GetRegisters()
 
             # save core and, kill process and verify corefile existence
             base_command = "process save-core --plugin-name=minidump "
@@ -110,6 +126,7 @@ def test_save_linux_mini_dump(self):
                 expected_modules,
                 expected_threads,
                 stacks_to_sp_map,
+                stakcs_to_registers_map,
             )
 
             self.runCmd(base_command + " --style=modified-memory '%s'" % (core_dirty))
@@ -120,6 +137,7 @@ def test_save_linux_mini_dump(self):
                 expected_modules,
                 expected_threads,
                 stacks_to_sp_map,
+                stakcs_to_registers_map,
             )
 
             self.runCmd(base_command + " --style=full '%s'" % (core_full))
@@ -130,6 +148,7 @@ def test_save_linux_mini_dump(self):
                 expected_modules,
                 expected_threads,
                 stacks_to_sp_map,
+                stakcs_to_registers_map,
             )
 
             options = lldb.SBSaveCoreOptions()
@@ -147,6 +166,7 @@ def test_save_linux_mini_dump(self):
                 expected_modules,
                 expected_threads,
                 stacks_to_sp_map,
+                stakcs_to_registers_map,
             )
 
             options = lldb.SBSaveCoreOptions()
@@ -163,6 +183,7 @@ def test_save_linux_mini_dump(self):
                 expected_modules,
                 expected_threads,
                 stacks_to_sp_map,
+                stakcs_to_registers_map,
             )
 
             # Minidump can now save full core files, but they will be huge and
@@ -181,6 +202,7 @@ def test_save_linux_mini_dump(self):
                 expected_modules,
                 expected_threads,
                 stacks_to_sp_map,
+                stakcs_to_registers_map,
             )
 
             self.assertSuccess(process.Kill())
@@ -276,13 +298,16 @@ def test_save_linux_mini_dump_default_options(self):
             expected_threads = []
             stacks_to_sp_map = {}
             expected_pid = process.GetProcessInfo().GetProcessID()
+            stacks_to_registers_map = {}
 
             for thread_idx in range(process.GetNumThreads()):
                 thread = process.GetThreadAtIndex(thread_idx)
                 thread_id = thread.GetThreadID()
                 expected_threads.append(thread_id)
                 stacks_to_sp_map[thread_id] = thread.GetFrameAtIndex(0).GetSP()
-
+                stacks_to_registers_map[thread_id] = thread.GetFrameAtIndex(
+                    0
+                ).GetRegisters()
 
             # This is almost identical to the single thread test case because
             # minidump defaults to stacks only, so we want to see if the
@@ -294,7 +319,14 @@ def test_save_linux_mini_dump_default_options(self):
             error = process.SaveCore(options)
             self.assertTrue(error.Success())
 
-            self.verify_core_file(default_value_file, expected_pid, expected_modules, expected_threads, stacks_to_sp_map)
+            self.verify_core_file(
+                default_value_file,
+                expected_pid,
+                expected_modules,
+                expected_threads,
+                stacks_to_sp_map,
+                stacks_to_registers_map,
+            )
 
         finally:
             self.assertTrue(self.dbg.DeleteTarget(target))

llvm/include/llvm/Config/llvm-config.h.cmake

@@ -16,7 +16,7 @@
 
 /* Indicate that this is LLVM compiled from the amd-gfx branch. */
 #define LLVM_HAVE_BRANCH_AMD_GFX
-#define LLVM_MAIN_REVISION 509964
+#define LLVM_MAIN_REVISION 509968
 
 /* Define if LLVM_ENABLE_DUMP is enabled */
 #cmakedefine LLVM_ENABLE_DUMP

llvm/include/llvm/Transforms/Utils/SimplifyCFGOptions.h

@@ -27,6 +27,7 @@ struct SimplifyCFGOptions {
   bool ConvertSwitchToLookupTable = false;
   bool NeedCanonicalLoop = true;
   bool HoistCommonInsts = false;
+  bool HoistLoadsStoresWithCondFaulting = false;
   bool SinkCommonInsts = false;
   bool SimplifyCondBranch = true;
   bool SpeculateBlocks = true;
@@ -59,6 +60,10 @@ struct SimplifyCFGOptions {
     HoistCommonInsts = B;
     return *this;
   }
+  SimplifyCFGOptions &hoistLoadsStoresWithCondFaulting(bool B) {
+    HoistLoadsStoresWithCondFaulting = B;
+    return *this;
+  }
   SimplifyCFGOptions &sinkCommonInsts(bool B) {
     SinkCommonInsts = B;
     return *this;

llvm/lib/Passes/PassBuilder.cpp

@@ -848,6 +848,8 @@ Expected<SimplifyCFGOptions> parseSimplifyCFGOptions(StringRef Params) {
       Result.needCanonicalLoops(Enable);
     } else if (ParamName == "hoist-common-insts") {
       Result.hoistCommonInsts(Enable);
+    } else if (ParamName == "hoist-loads-stores-with-cond-faulting") {
+      Result.hoistLoadsStoresWithCondFaulting(Enable);
     } else if (ParamName == "sink-common-insts") {
       Result.sinkCommonInsts(Enable);
     } else if (ParamName == "speculate-unpredictables") {

llvm/lib/Passes/PassBuilderPipelines.cpp

@@ -1534,9 +1534,11 @@ PassBuilder::buildModuleOptimizationPipeline(OptimizationLevel Level,
 
   // LoopSink (and other loop passes since the last simplifyCFG) might have
   // resulted in single-entry-single-exit or empty blocks. Clean up the CFG.
-  OptimizePM.addPass(SimplifyCFGPass(SimplifyCFGOptions()
-                                         .convertSwitchRangeToICmp(true)
-                                         .speculateUnpredictables(true)));
+  OptimizePM.addPass(
+      SimplifyCFGPass(SimplifyCFGOptions()
+                          .convertSwitchRangeToICmp(true)
+                          .speculateUnpredictables(true)
+                          .hoistLoadsStoresWithCondFaulting(true)));
 
   // Add the core optimizing pipeline.
   MPM.addPass(createModuleToFunctionPassAdaptor(std::move(OptimizePM),

llvm/lib/Transforms/Scalar/SimplifyCFGPass.cpp

@@ -73,6 +73,11 @@ static cl::opt<bool> UserHoistCommonInsts(
     "hoist-common-insts", cl::Hidden, cl::init(false),
     cl::desc("hoist common instructions (default = false)"));
 
+static cl::opt<bool> UserHoistLoadsStoresWithCondFaulting(
+    "hoist-loads-stores-with-cond-faulting", cl::Hidden, cl::init(false),
+    cl::desc("Hoist loads/stores if the target supports conditional faulting "
+             "(default = false)"));
+
 static cl::opt<bool> UserSinkCommonInsts(
     "sink-common-insts", cl::Hidden, cl::init(false),
     cl::desc("Sink common instructions (default = false)"));
@@ -326,6 +331,9 @@ static void applyCommandLineOverridesToOptions(SimplifyCFGOptions &Options) {
     Options.NeedCanonicalLoop = UserKeepLoops;
   if (UserHoistCommonInsts.getNumOccurrences())
     Options.HoistCommonInsts = UserHoistCommonInsts;
+  if (UserHoistLoadsStoresWithCondFaulting.getNumOccurrences())
+    Options.HoistLoadsStoresWithCondFaulting =
+        UserHoistLoadsStoresWithCondFaulting;
   if (UserSinkCommonInsts.getNumOccurrences())
     Options.SinkCommonInsts = UserSinkCommonInsts;
   if (UserSpeculateUnpredictables.getNumOccurrences())
@@ -354,6 +362,8 @@ void SimplifyCFGPass::printPipeline(
      << "switch-to-lookup;";
   OS << (Options.NeedCanonicalLoop ? "" : "no-") << "keep-loops;";
   OS << (Options.HoistCommonInsts ? "" : "no-") << "hoist-common-insts;";
+  OS << (Options.HoistLoadsStoresWithCondFaulting ? "" : "no-")
+     << "hoist-loads-stores-with-cond-faulting;";
   OS << (Options.SinkCommonInsts ? "" : "no-") << "sink-common-insts;";
   OS << (Options.SpeculateBlocks ? "" : "no-") << "speculate-blocks;";
   OS << (Options.SimplifyCondBranch ? "" : "no-") << "simplify-cond-branch;";