Merge branch 'main' into issue-104423

Author: spall

.ci/compute_projects_test.py

@@ -147,7 +147,7 @@ def test_flang(self):
 
     def test_invalid_subproject(self):
         env_variables = compute_projects.get_env_variables(
-            [".ci/compute_projects.py"], "Linux"
+            ["third-party/benchmark/CMakeLists.txt"], "Linux"
         )
         self.assertEqual(env_variables["projects_to_build"], "")
         self.assertEqual(env_variables["project_check_targets"], "")
@@ -163,7 +163,7 @@ def test_top_level_file(self):
 
     def test_exclude_runtiems_in_projects(self):
         env_variables = compute_projects.get_env_variables(
-            [".ci/compute_projects.py", "libcxx/CMakeLists.txt"], "Linux"
+            ["libcxx/CMakeLists.txt"], "Linux"
         )
         self.assertEqual(env_variables["projects_to_build"], "")
         self.assertEqual(env_variables["project_check_targets"], "")
@@ -192,10 +192,10 @@ def test_ci(self):
         env_variables = compute_projects.get_env_variables(
             [".ci/compute_projects.py"], "Linux"
         )
-        self.assertEqual(env_variables["projects_to_build"], "clang;lld;llvm;lldb")
+        self.assertEqual(env_variables["projects_to_build"], "clang;lld;lldb;llvm")
         self.assertEqual(
             env_variables["project_check_targets"],
-            "check-clang check-lld check-llvm check-lldb",
+            "check-clang check-lld check-lldb check-llvm",
         )
         self.assertEqual(
             env_variables["runtimes_to_build"], "libcxx;libcxxabi;libunwind"

.ci/generate_test_report_lib.py

@@ -92,7 +92,9 @@ def plural(num_tests):
             ]
         )
     elif failures:
-        report.extend(["", "## Failed Tests", "(click on a test name to see its 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}"])
@@ -120,6 +122,17 @@ def plural(num_tests):
             ]
         )
 
+    if failures or return_code != 0:
+        report.extend(
+            [
+                "",
+                "If these failures are unrelated to your changes (for example "
+                "tests are broken or flaky at HEAD), please open an issue at "
+                "https://github.com/llvm/llvm-project/issues and add the "
+                "`infrastructure` label.",
+            ]
+        )
+
     report = "\n".join(report)
     if len(report.encode("utf-8")) > size_limit:
         return generate_report(

.ci/generate_test_report_lib_test.py

@@ -109,7 +109,9 @@ def test_no_failures_build_failed(self):
 
               All tests passed but another part of the build **failed**.
 
-              [Download](https://buildkite.com/organizations/organization_slug/pipelines/pipeline_slug/builds/build_number/jobs/job_id/download.txt) the build's log file to see the details."""
+              [Download](https://buildkite.com/organizations/organization_slug/pipelines/pipeline_slug/builds/build_number/jobs/job_id/download.txt) the build's log file to see the details.
+              
+              If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label."""
                 ),
                 "error",
             ),
@@ -169,7 +171,9 @@ def test_report_single_file_single_testsuite(self):
           ```
           Other output goes here
           ```
-          </details>"""
+          </details>
+          
+          If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label."""
                 ),
                 "error",
             ),
@@ -203,7 +207,9 @@ def test_report_single_file_single_testsuite(self):
         ```
         DEF/test_2 output goes here
         ```
-        </details>"""
+        </details>
+        
+        If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label."""
         ),
         "error",
     )
@@ -311,7 +317,9 @@ def test_report_dont_list_failures(self):
 
           * 1 test failed
 
-          Failed tests and their output was too large to report. Download the build's log file to see the details."""
+          Failed tests and their output was too large to report. Download the build's log file to see the details.
+          
+          If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label."""
                 ),
                 "error",
             ),
@@ -352,13 +360,16 @@ def test_report_dont_list_failures_link_to_log(self):
 
           * 1 test failed
 
-          Failed tests and their output was too large to report. [Download](https://buildkite.com/organizations/organization_slug/pipelines/pipeline_slug/builds/build_number/jobs/job_id/download.txt) the build's log file to see the details."""
+          Failed tests and their output was too large to report. [Download](https://buildkite.com/organizations/organization_slug/pipelines/pipeline_slug/builds/build_number/jobs/job_id/download.txt) the build's log file to see the details.
+          
+          If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label."""
                 ),
                 "error",
             ),
         )
 
     def test_report_size_limit(self):
+        test_output = "f" * 1000
         self.assertEqual(
             generate_test_report_lib.generate_report(
                 "Foo",
@@ -371,14 +382,16 @@ def test_report_size_limit(self):
           <testsuites time="0.02">
           <testsuite name="Bar" tests="1" failures="1" skipped="0" time="0.02">
           <testcase classname="Bar/test_1" name="test_1" time="0.02">
-            <failure><![CDATA[Some long output goes here...]]></failure>
+            <failure><![CDATA[{output}]]></failure>
           </testcase>
           </testsuite>
-          </testsuites>"""
+          </testsuites>""".format(
+                                output=test_output
+                            )
                         )
                     )
                 ],
-                size_limit=128,
+                size_limit=512,
             ),
             (
                 dedent(
@@ -387,7 +400,9 @@ def test_report_size_limit(self):
 
           * 1 test failed
 
-          Failed tests and their output was too large to report. Download the build's log file to see the details."""
+          Failed tests and their output was too large to report. Download the build's log file to see the details.
+          
+          If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label."""
                 ),
                 "error",
             ),

.ci/metrics/metrics.py

@@ -29,8 +29,8 @@
 # name.
 GITHUB_JOB_TO_TRACK = {
     "github_llvm_premerge_checks": {
-        "Build and Test Linux (Test Only - Please Ignore Results)": "premerge_linux",
-        "Build and Test Windows (Test Only - Please Ignore Results)": "premerge_windows",
+        "Build and Test Linux": "premerge_linux",
+        "Build and Test Windows": "premerge_windows",
     }
 }
 
@@ -67,6 +67,7 @@
 # remain small.
 BUILDKITE_GRAPHQL_BUILDS_PER_PAGE = 50
 
+
 @dataclass
 class JobMetrics:
     job_name: str
@@ -77,6 +78,7 @@ class JobMetrics:
     workflow_id: int
     workflow_name: str
 
+
 @dataclass
 class GaugeMetric:
     name: str
@@ -258,6 +260,7 @@ def buildkite_get_metrics(
 
     return output, incomplete_now
 
+
 def github_get_metrics(
     github_repo: github.Repository, last_workflows_seen_as_completed: set[int]
 ) -> tuple[list[JobMetrics], int]:
@@ -346,19 +349,7 @@ def github_get_metrics(
                     running_count[metric_name] += 1
                 continue
 
-            job_result = int(job.conclusion == "success")
-            if job_result:
-                # We still might want to mark the job as a failure if one of the steps
-                # failed. This is required due to use setting continue-on-error in
-                # the premerge pipeline to prevent sending emails while we are
-                # testing the infrastructure.
-                # TODO(boomanaiden154): Remove this once the premerge pipeline is no
-                # longer in a testing state and we can directly assert the workflow
-                # result.
-                for step in job.steps:
-                    if step.conclusion != "success" and step.conclusion != "skipped":
-                        job_result = 0
-                        break
+            job_result = int(job.conclusion == "success" or job.conclusion == "skipped")
 
             created_at = job.created_at
             started_at = job.started_at

.ci/monolithic-linux.sh

@@ -38,6 +38,7 @@ function at-exit {
 
   ccache --print-stats > artifacts/ccache_stats.txt
   cp "${BUILD_DIR}"/.ninja_log artifacts/.ninja_log
+  cp "${BUILD_DIR}"/test-results.*.xml artifacts/ || :
 
   # If building fails there will be no results files.
   shopt -s nullglob

.ci/monolithic-windows.sh

@@ -33,6 +33,7 @@ function at-exit {
   mkdir -p artifacts
   sccache --show-stats >> artifacts/sccache_stats.txt
   cp "${BUILD_DIR}"/.ninja_log artifacts/.ninja_log
+  cp "${BUILD_DIR}"/test-results.*.xml artifacts/ || :
 
   # If building fails there will be no results files.
   shopt -s nullglob

.github/new-prs-labeler.yml

@@ -702,12 +702,20 @@ mlgo:
   - llvm/unittests/CodeGen/ML*
   - llvm/test/CodeGen/MLRegAlloc/**
   - llvm/utils/mlgo-utils/**
+  - llvm/docs/MLGO.rst
+  - llvm/include/llvm/Analysis/IR2Vec.h
+  - llvm/lib/Analysis/IR2Vec.cpp
+  - llvm/lib/Analysis/models/**
+  - llvm/test/Analysis/IR2Vec/**
 
 tools:llvm-exegesis:
   - llvm/tools/llvm-exegesis/**
   - llvm/test/tools/llvm-exegesis/**
   - llvm/unittests/tools/llvm-exegesis/**
 
+tools:llvm-reduce:
+  - llvm/tools/llvm-reduce/**
+
 platform:windows:
   - lld/COFF/**
   - clang/lib/Driver/MSVC.cpp

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

@@ -39,33 +39,15 @@ RUN regsvr32 /S "C:\BuildTools\DIA SDK\bin\amd64\msdia140.dll" & \
 
 # install tools as described in https://llvm.org/docs/GettingStartedVS.html
 # and a few more that were not documented...
-RUN choco install -y ninja git
+RUN choco install -y ninja git sccache
 # Pin an older version of Python; the current Python 3.10 fails when
 # doing "pip install" for the other dependencies, as it fails to find libxml
 # while compiling some package.
 RUN choco install -y python3 --version 3.9.7
 
-# ActivePerl is currently not installable via Chocolatey, see
-# http://disq.us/p/2ipditb. Install StrawberryPerl instead. Unfortunately,
-# StrawberryPerl not only installs Perl, but also a redundant C/C++ compiler
-# toolchain, and a copy of pkg-config which can cause misdetections for other
-# built products, see
-# https://github.com/StrawberryPerl/Perl-Dist-Strawberry/issues/11 for further
-# details. Remove the redundant and unnecessary parts of the StrawberryPerl
-# install.
-RUN choco install -y strawberryperl && \
-    rmdir /q /s c:\strawberry\c && \
-    del /q c:\strawberry\perl\bin\pkg-config*
-
-# libcxx requires clang(-cl) to be available
-RUN choco install -y sccache llvm
+# Testing requires psutil
 RUN pip install psutil
 
-RUN curl -LO https://github.com/mstorsjo/llvm-mingw/releases/download/20230320/llvm-mingw-20230320-ucrt-x86_64.zip && \
-    powershell Expand-Archive llvm-mingw-*-ucrt-x86_64.zip -DestinationPath . && \
-    del llvm-mingw-*-ucrt-x86_64.zip && \
-    ren llvm-mingw-20230320-ucrt-x86_64 llvm-mingw
-
 # configure Python encoding
 ENV PYTHONIOENCODING=UTF-8
 
@@ -103,11 +85,6 @@ RUN powershell -Command \
 RUN git config --system core.longpaths true & \
     git config --global core.autocrlf false
 
-# handle for debugging of files beeing locked by some processes.
-RUN choco install -y handle
-
-RUN pip3 install pywin32 buildbot-worker==2.8.4
-
 ARG RUNNER_VERSION=2.324.0
 ENV RUNNER_VERSION=$RUNNER_VERSION
 

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

@@ -37,7 +37,7 @@ jobs:
   stage1:
     if: github.repository_owner == 'llvm'
     runs-on: libcxx-self-hosted-linux
-    container: ghcr.io/llvm/libcxx-linux-builder:2b57ebb50b6d418e70382e655feaa619b558e254
+    container: ghcr.io/llvm/libcxx-linux-builder:b060022103f551d8ca1dad84122ef73927c86512
     continue-on-error: false
     strategy:
       fail-fast: false
@@ -184,7 +184,7 @@ jobs:
             **/crash_diagnostics/*
 
   macos:
-    needs: [ stage3 ]
+    needs: [ stage2 ]
     strategy:
       fail-fast: false
       matrix:

.github/workflows/pr-code-format.yml

@@ -55,7 +55,7 @@ jobs:
       - name: Install clang-format
         uses: aminya/setup-cpp@17c11551771948abc5752bbf3183482567c7caf0 # v1.1.1
         with:
-          clangformat: 19.1.6
+          clangformat: 20.1.5
 
       - name: Setup Python env
         uses: actions/setup-python@42375524e23c412d93fb67b49958b491fce71c38 # v5.4.0

.github/workflows/premerge.yaml

@@ -16,7 +16,6 @@ on:
       - closed
   push:
     branches:
-      - 'main'
       - 'release/**'
 
 concurrency:
@@ -25,7 +24,7 @@ concurrency:
 
 jobs:
   premerge-checks-linux:
-    name: Build and Test Linux (Test Only - Please Ignore Results)
+    name: Build and Test Linux
     if: >-
         github.repository_owner == 'llvm' &&
         (github.event_name != 'pull_request' || github.event.action != 'closed')
@@ -43,9 +42,6 @@ jobs:
         # Mark the job as a success even if the step fails so that people do
         # not get notified while the new premerge pipeline is in an
         # experimental state.
-        # TODO(boomanaiden154): Remove this once the pipeline is stable and we
-        # are ready for people to start recieving notifications.
-        continue-on-error: true
         run: |
           git config --global --add safe.directory '*'
 
@@ -74,7 +70,7 @@ jobs:
           include-hidden-files: 'true'
 
   premerge-checks-windows:
-    name: Build and Test Windows (Test Only - Please Ignore Results)
+    name: Build and Test Windows
     if: >-
         github.repository_owner == 'llvm' &&
         (github.event_name != 'pull_request' || github.event.action != 'closed')
@@ -110,9 +106,6 @@ jobs:
         # Mark the job as a success even if the step fails so that people do
         # not get notified while the new premerge pipeline is in an
         # experimental state.
-        # TODO(boomanaiden154): Remove this once the pipeline is stable and we
-        # are ready for people to start recieving notifications.
-        continue-on-error: true
         if: ${{ steps.vars.outputs.windows-projects != '' }}
         shell: cmd
         run: |

bolt/CMakeLists.txt

@@ -46,6 +46,10 @@ if(BOLT_BUILT_STANDALONE)
   set(LLVM_RUNTIME_OUTPUT_INTDIR ${CMAKE_BINARY_DIR}/${CMAKE_CFG_INTDIR}/bin)
   set(LLVM_LIBRARY_OUTPUT_INTDIR ${CMAKE_BINARY_DIR}/${CMAKE_CFG_INTDIR}/lib${LLVM_LIBDIR_SUFFIX})
 
+  separate_arguments(LLVM_DEFINITIONS_LIST NATIVE_COMMAND ${LLVM_DEFINITIONS})
+  add_definitions(${LLVM_DEFINITIONS_LIST})
+  list(APPEND CMAKE_REQUIRED_DEFINITIONS ${LLVM_DEFINITIONS_LIST})
+
   include(AddLLVM)
   include(TableGen)
   include_directories(${LLVM_INCLUDE_DIRS})

bolt/Maintainers.txt

@@ -20,6 +20,10 @@ N: Vladislav Khmelevsky
 E: [email protected]
 D: AArch64 backend
 
+N: Paschalis Mpeis
+E: [email protected]
+D: AArch64 backend
+
 N: Job Noorman
 E: [email protected]
 D: RISC-V backend