Merge branch 'main' into flang_cuda_build_fixes

Author: vzakhari

.ci/generate_test_report_lib.py

@@ -92,7 +92,7 @@ def plural(num_tests):
             ]
         )
     elif failures:
-        report.extend(["", "## Failed Tests", "(click to see output)"])
+        report.extend(["", "## Failed Tests", "(click on a test name to see its output)"])
 
         for testsuite_name, failures in failures.items():
             report.extend(["", f"### {testsuite_name}"])

.ci/generate_test_report_lib_test.py

@@ -2,6 +2,9 @@
 # See https://llvm.org/LICENSE.txt for license information.
 # SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 
+# To run these tests:
+# python -m unittest generate_test_report_lib_test.py
+
 import unittest
 from io import StringIO
 from textwrap import dedent
@@ -150,7 +153,7 @@ def test_report_single_file_single_testsuite(self):
           * 2 tests failed
 
           ## Failed Tests
-          (click to see output)
+          (click on a test name to see its output)
 
           ### Bar
           <details>
@@ -182,7 +185,7 @@ def test_report_single_file_single_testsuite(self):
         * 2 tests failed
 
         ## Failed Tests
-        (click to see output)
+        (click on a test name to see its output)
 
         ### ABC
         <details>

.ci/metrics/metrics.py

@@ -282,6 +282,13 @@ def github_get_metrics(
     queued_count = collections.Counter()
     running_count = collections.Counter()
 
+    # Initialize all the counters to 0 so we report 0 when no job is queued
+    # or running.
+    for wf_name, wf_metric_name in GITHUB_WORKFLOW_TO_TRACK.items():
+        for job_name, job_metric_name in GITHUB_JOB_TO_TRACK[wf_metric_name].items():
+            queued_count[wf_metric_name + "_" + job_metric_name] = 0
+            running_count[wf_metric_name + "_" + job_metric_name] = 0
+
     # The list of workflows this iteration will process.
     # MaxSize = GITHUB_WORKFLOWS_MAX_PROCESS_COUNT
     workflow_seen_as_completed = set()

.ci/monolithic-linux.sh

@@ -53,6 +53,7 @@ targets="${2}"
 lit_args="-v --xunit-xml-output ${BUILD_DIR}/test-results.xml --use-unique-output-file-name --timeout=1200 --time-tests"
 
 echo "--- cmake"
+export PIP_BREAK_SYSTEM_PACKAGES=1
 pip install -q -r "${MONOREPO_ROOT}"/mlir/python/requirements.txt
 pip install -q -r "${MONOREPO_ROOT}"/lldb/test/requirements.txt
 pip install -q -r "${MONOREPO_ROOT}"/.ci/requirements.txt

.github/new-prs-labeler.yml

@@ -660,6 +660,12 @@ backend:DirectX:
   - '**/*dxil*/**'
   - '**/*DXContainer*'
   - '**/*DXContainer*/**'
+  - clang/lib/Sema/SemaDirectX.cpp
+  - clang/include/clang/Sema/SemaDirectX.h
+  - clang/include/clang/Basic/BuiltinsDirectX.td
+  - clang/lib/CodeGen/TargetBuiltins/DirectX.cpp
+  - clang/test/CodeGenDirectX/**
+  - clang/test/SemaDirectX/**
 
 backend:SPIR-V:
   - clang/lib/Driver/ToolChains/SPIRV.*

.github/workflows/containers/github-action-ci-windows/Dockerfile

@@ -108,7 +108,7 @@ RUN choco install -y handle
 
 RUN pip3 install pywin32 buildbot-worker==2.8.4
 
-ARG RUNNER_VERSION=2.322.0
+ARG RUNNER_VERSION=2.323.0
 ENV RUNNER_VERSION=$RUNNER_VERSION
 
 RUN powershell -Command \

.github/workflows/containers/github-action-ci/Dockerfile

@@ -95,7 +95,7 @@ WORKDIR /home/gha
 
 FROM ci-container as ci-container-agent
 
-ENV GITHUB_RUNNER_VERSION=2.322.0
+ENV GITHUB_RUNNER_VERSION=2.323.0
 
 RUN mkdir actions-runner && \
     cd actions-runner && \

.github/workflows/libcxx-build-and-test.yaml

@@ -197,10 +197,20 @@ jobs:
           os: macos-15
         - config: apple-configuration
           os: macos-15
+        # TODO: These jobs are intended to test back-deployment (building against ToT libc++ but running against an
+        #       older system-provided libc++.dylib). Doing this properly would require building the test suite on a
+        #       recent macOS using a recent Clang (hence recent Xcode), and then running the actual test suite on an
+        #       older mac. We could do that by e.g. sharing artifacts between the two jobs.
+        #
+        #       However, our Lit configuration currently doesn't provide a good way to do that in a batch, so our only
+        #       alternative is to actually build on the same host that we're going to run on. Sadly, that doesn't work
+        #       since older macOSes don't support newer Xcodes. For now, we run the "backdeployment" jobs on recent
+        #       macOS versions as a way to avoid rotting that configuration, but it doesn't provide a lot of additional
+        #       coverage.
         - config: apple-system
-          os: macos-13
+          os: macos-15
         - config: apple-system-hardened
-          os: macos-13
+          os: macos-15
     runs-on: ${{ matrix.os }}
     steps:
       - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2

bolt/include/bolt/Core/MCPlusBuilder.h

@@ -577,16 +577,51 @@ class MCPlusBuilder {
     return getNoRegister();
   }
 
-  /// Returns the register used as call destination, or no-register, if not
-  /// an indirect call. Sets IsAuthenticatedInternally if the instruction
-  /// accepts a signed pointer as its operand and authenticates it internally.
+  /// Returns the register used as the destination of an indirect branch or call
+  /// instruction. Sets IsAuthenticatedInternally if the instruction accepts
+  /// a signed pointer as its operand and authenticates it internally.
   virtual MCPhysReg
-  getRegUsedAsCallDest(const MCInst &Inst,
-                       bool &IsAuthenticatedInternally) const {
+  getRegUsedAsIndirectBranchDest(const MCInst &Inst,
+                                 bool &IsAuthenticatedInternally) const {
     llvm_unreachable("not implemented");
     return getNoRegister();
   }
 
+  /// Returns the register containing an address safely materialized by `Inst`
+  /// under the Pointer Authentication threat model.
+  ///
+  /// Returns the register `Inst` writes to if:
+  /// 1. the register is a materialized address, and
+  /// 2. the register has been materialized safely, i.e. cannot be attacker-
+  ///    controlled, under the Pointer Authentication threat model.
+  ///
+  /// If the instruction does not write to any register satisfying the above
+  /// two conditions, NoRegister is returned.
+  ///
+  /// The Pointer Authentication threat model assumes an attacker is able to
+  /// modify any writable memory, but not executable code (due to W^X).
+  virtual MCPhysReg
+  getMaterializedAddressRegForPtrAuth(const MCInst &Inst) const {
+    llvm_unreachable("not implemented");
+    return getNoRegister();
+  }
+
+  /// Analyzes if this instruction can safely perform address arithmetics
+  /// under Pointer Authentication threat model.
+  ///
+  /// If an (OutReg, InReg) pair is returned, then after Inst is executed,
+  /// OutReg is as trusted as InReg is.
+  ///
+  /// The arithmetic instruction is considered safe if OutReg is not attacker-
+  /// controlled, provided InReg and executable code are not. Please note that
+  /// registers other than InReg as well as the contents of memory which is
+  /// writable by the process should be considered attacker-controlled.
+  virtual std::optional<std::pair<MCPhysReg, MCPhysReg>>
+  analyzeAddressArithmeticsForPtrAuth(const MCInst &Inst) const {
+    llvm_unreachable("not implemented");
+    return std::make_pair(getNoRegister(), getNoRegister());
+  }
+
   virtual bool isTerminator(const MCInst &Inst) const;
 
   virtual bool isNoop(const MCInst &Inst) const {

bolt/include/bolt/Passes/DataflowAnalysis.h

@@ -292,14 +292,17 @@ class DataflowAnalysis {
   /// Relies on a ptr map to fetch the previous instruction and then retrieve
   /// state. WARNING: Watch out for invalidated pointers. Do not use this
   /// function if you invalidated pointers after the analysis has been completed
-  ErrorOr<const StateTy &> getStateBefore(const MCInst &Point) {
-    return getStateAt(PrevPoint[&Point]);
+  ErrorOr<const StateTy &> getStateBefore(const MCInst &Point) const {
+    auto It = PrevPoint.find(&Point);
+    if (It == PrevPoint.end())
+      return make_error_code(std::errc::result_out_of_range);
+    return getStateAt(It->getSecond());
   }
 
-  ErrorOr<const StateTy &> getStateBefore(ProgramPoint Point) {
+  ErrorOr<const StateTy &> getStateBefore(ProgramPoint Point) const {
     if (Point.isBB())
       return getStateAt(*Point.getBB());
-    return getStateAt(PrevPoint[Point.getInst()]);
+    return getStateBefore(*Point.getInst());
   }
 
   /// Remove any state annotations left by this analysis