Merge branch 'main' into shard-python-bindings

Author: Svoch

.ci/generate_test_report_github.py

@@ -8,10 +8,15 @@
 
 import generate_test_report_lib
 
-PLATFORM_TITLES = {
-    "Windows": ":window: Windows x64 Test Results",
-    "Linux": ":penguin: Linux x64 Test Results",
-}
+def compute_platform_title() -> str:
+    logo = ":window:" if platform.system() == "Windows" else ":penguin:"
+    # On Linux the machine value is x86_64 on Windows it is AMD64.
+    if platform.machine() == "x86_64" or platform.machine() == "AMD64":
+        arch = "x64"
+    else:
+        arch = platform.machine()
+    return f"{logo} {platform.system()} {arch} Test Results"
+
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
@@ -22,7 +27,7 @@
     args = parser.parse_args()
 
     report = generate_test_report_lib.generate_report_from_files(
-        PLATFORM_TITLES[platform.system()], args.return_code, args.build_test_logs
+        compute_platform_title(), args.return_code, args.build_test_logs
     )
 
     print(report)

.ci/generate_test_report_lib.py

@@ -98,6 +98,23 @@ def _format_ninja_failures(ninja_failures: list[tuple[str, str]]) -> list[str]:
         )
     return output
 
+def get_failures(junit_objects) -> dict[str, list[tuple[str, str]]]:
+    failures = {}
+    for results in junit_objects:
+        for testsuite in results:
+            for test in testsuite:
+                if (
+                    not test.is_passed
+                    and test.result
+                    and isinstance(test.result[0], Failure)
+                ):
+                    if failures.get(testsuite.name) is None:
+                        failures[testsuite.name] = []
+                    failures[testsuite.name].append(
+                        (test.classname + "/" + test.name, test.result[0].text)
+                    )
+    return failures
+
 
 # Set size_limit to limit the byte size of the report. The default is 1MB as this
 # is the most that can be put into an annotation. If the generated report exceeds
@@ -113,7 +130,7 @@ def generate_report(
     size_limit=1024 * 1024,
     list_failures=True,
 ):
-    failures = {}
+    failures = get_failures(junit_objects)
     tests_run = 0
     tests_skipped = 0
     tests_failed = 0
@@ -124,18 +141,6 @@ def generate_report(
             tests_skipped += testsuite.skipped
             tests_failed += testsuite.failures
 
-            for test in testsuite:
-                if (
-                    not test.is_passed
-                    and test.result
-                    and isinstance(test.result[0], Failure)
-                ):
-                    if failures.get(testsuite.name) is None:
-                        failures[testsuite.name] = []
-                    failures[testsuite.name].append(
-                        (test.classname + "/" + test.name, test.result[0].text)
-                    )
-
     report = [f"# {title}", ""]
 
     if tests_run == 0:
@@ -258,7 +263,7 @@ def plural(num_tests):
     return report
 
 
-def generate_report_from_files(title, return_code, build_log_files):
+def load_info_from_files(build_log_files):
     junit_files = [
         junit_file for junit_file in build_log_files if junit_file.endswith(".xml")
     ]
@@ -271,6 +276,9 @@ def generate_report_from_files(title, return_code, build_log_files):
             ninja_logs.append(
                 [log_line.strip() for log_line in ninja_log_file_handle.readlines()]
             )
-    return generate_report(
-        title, return_code, [JUnitXml.fromfile(p) for p in junit_files], ninja_logs
-    )
+    return [JUnitXml.fromfile(p) for p in junit_files], ninja_logs
+
+
+def generate_report_from_files(title, return_code, build_log_files):
+    junit_objects, ninja_logs = load_info_from_files(build_log_files)
+    return generate_report(title, return_code, junit_objects, ninja_logs)

.ci/metrics/metrics.py

@@ -40,6 +40,7 @@
 GITHUB_JOB_TO_TRACK = {
     "github_llvm_premerge_checks": {
         "Build and Test Linux": "premerge_linux",
+        "Build and Test Linux AArch64": "premerge_linux_aarch64",
         "Build and Test Windows": "premerge_windows",
     },
     "github_libcxx_premerge_checks": {

.ci/monolithic-linux.sh

@@ -66,11 +66,13 @@ start-group "ninja"
 
 # Targets are not escaped as they are passed as separate arguments.
 ninja -C "${BUILD_DIR}" -k 0 ${targets} |& tee ninja.log
+cp ${BUILD_DIR}/.ninja_log ninja.ninja_log
 
 if [[ "${runtime_targets}" != "" ]]; then
   start-group "ninja Runtimes"
 
   ninja -C "${BUILD_DIR}" ${runtime_targets} |& tee ninja_runtimes.log
+  cp ${BUILD_DIR}/.ninja_log ninja_runtimes.ninja_log
 fi
 
 # Compiling runtimes with just-built Clang and running their tests
@@ -87,6 +89,7 @@ if [[ "${runtime_targets_needs_reconfig}" != "" ]]; then
 
   ninja -C "${BUILD_DIR}" ${runtime_targets_needs_reconfig} \
     |& tee ninja_runtimes_needs_reconfig1.log
+  cp ${BUILD_DIR}/.ninja_log ninja_runtimes_needs_reconig.ninja_log
 
   start-group "CMake Runtimes Clang Modules"
 
@@ -99,4 +102,5 @@ if [[ "${runtime_targets_needs_reconfig}" != "" ]]; then
 
   ninja -C "${BUILD_DIR}" ${runtime_targets_needs_reconfig} \
     |& tee ninja_runtimes_needs_reconfig2.log
+  cp ${BUILD_DIR}/.ninja_log ninja_runtimes_needs_reconfig2.ninja_log
 fi

.ci/monolithic-windows.sh

@@ -55,9 +55,11 @@ start-group "ninja"
 
 # Targets are not escaped as they are passed as separate arguments.
 ninja -C "${BUILD_DIR}" -k 0 ${targets} |& tee ninja.log
+cp ${BUILD_DIR}/.ninja_log ninja.ninja_log
 
 if [[ "${runtime_targets}" != "" ]]; then
   start-group "ninja runtimes"
 
   ninja -C "${BUILD_DIR}" -k 0 ${runtimes_targets} |& tee ninja_runtimes.log
+  cp ${BUILD_DIR}/.ninja_log ninja_runtimes.ninja_log
 fi

.ci/premerge_advisor_explain.py

@@ -0,0 +1,63 @@
+# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+# See https://llvm.org/LICENSE.txt for license information.
+# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+"""Script for getting explanations from the premerge advisor."""
+
+import argparse
+import os
+import platform
+import sys
+
+import requests
+
+import generate_test_report_lib
+
+PREMERGE_ADVISOR_URL = (
+    "http://premerge-advisor.premerge-advisor.svc.cluster.local:5000/explain"
+)
+
+
+def main(commit_sha: str, build_log_files: list[str]):
+    junit_objects, ninja_logs = generate_test_report_lib.load_info_from_files(
+        build_log_files
+    )
+    test_failures = generate_test_report_lib.get_failures(junit_objects)
+    current_platform = f"{platform.system()}-{platform.machine()}".lower()
+    explanation_request = {
+        "base_commit_sha": commit_sha,
+        "platform": current_platform,
+        "failures": [],
+    }
+    if test_failures:
+        for _, failures in test_failures.items():
+            for name, failure_messsage in failures:
+                explanation_request["failures"].append(
+                    {"name": name, "message": failure_messsage}
+                )
+    else:
+        ninja_failures = generate_test_report_lib.find_failure_in_ninja_logs(ninja_logs)
+        for name, failure_message in ninja_failures:
+            explanation_request["failures"].append(
+                {"name": name, "message": failure_message}
+            )
+    advisor_response = requests.get(PREMERGE_ADVISOR_URL, json=explanation_request)
+    if advisor_response.status_code == 200:
+        print(advisor_response.json())
+    else:
+        print(advisor_response.reason)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("commit_sha", help="The base commit SHA for the test.")
+    parser.add_argument(
+        "build_log_files", help="Paths to JUnit report files and ninja logs.", nargs="*"
+    )
+    args = parser.parse_args()
+
+    # Skip looking for results on AArch64 for now because the premerge advisor
+    # service is not available on AWS currently.
+    if platform.machine() == "arm64":
+        sys.exit(0)
+
+    main(args.commit_sha, args.build_log_files)

.ci/premerge_advisor_upload.py

@@ -0,0 +1,61 @@
+# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+# See https://llvm.org/LICENSE.txt for license information.
+# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+"""Script for uploading results to the premerge advisor."""
+
+import argparse
+import os
+import platform
+import sys
+
+import requests
+
+import generate_test_report_lib
+
+PREMERGE_ADVISOR_URL = (
+    "http://premerge-advisor.premerge-advisor.svc.cluster.local:5000/upload"
+)
+
+
+def main(commit_sha, workflow_run_number, build_log_files):
+    junit_objects, ninja_logs = generate_test_report_lib.load_info_from_files(
+        build_log_files
+    )
+    test_failures = generate_test_report_lib.get_failures(junit_objects)
+    source = "pull_request" if "GITHUB_ACTIONS" in os.environ else "postcommit"
+    current_platform = f"{platform.system()}-{platform.machine()}".lower()
+    failure_info = {
+        "source_type": source,
+        "base_commit_sha": commit_sha,
+        "source_id": workflow_run_number,
+        "failures": [],
+        "platform": current_platform,
+    }
+    if test_failures:
+        for _, failures in test_failures.items():
+            for name, failure_message in failures:
+                failure_info["failures"].append(
+                    {"name": name, "message": failure_message}
+                )
+    else:
+        ninja_failures = generate_test_report_lib.find_failure_in_ninja_logs(ninja_logs)
+        for name, failure_message in ninja_failures:
+            failure_info["failures"].append({"name": name, "message": failure_message})
+    requests.post(PREMERGE_ADVISOR_URL, json=failure_info)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("commit_sha", help="The base commit SHA for the test.")
+    parser.add_argument("workflow_run_number", help="The run number from GHA.")
+    parser.add_argument(
+        "build_log_files", help="Paths to JUnit report files and ninja logs.", nargs="*"
+    )
+    args = parser.parse_args()
+
+    # Skip uploading results on AArch64 for now because the premerge advisor
+    # service is not available on AWS currently.
+    if platform.machine() == "arm64":
+        sys.exit(0)
+
+    main(args.commit_sha, args.workflow_run_number, args.build_log_files)

.ci/utils.sh

@@ -26,7 +26,7 @@ function at-exit {
   mkdir -p artifacts
   sccache --show-stats
   sccache --show-stats >> artifacts/sccache_stats.txt
-  cp "${BUILD_DIR}"/.ninja_log artifacts/.ninja_log
+  cp "${MONOREPO_ROOT}"/*.ninja_log artifacts/ || :
   cp "${MONOREPO_ROOT}"/*.log artifacts/ || :
   cp "${BUILD_DIR}"/test-results.*.xml artifacts/ || :
 
@@ -38,6 +38,17 @@ function at-exit {
       $retcode "${BUILD_DIR}"/test-results.*.xml "${MONOREPO_ROOT}"/ninja*.log \
       >> $GITHUB_STEP_SUMMARY
   fi
+
+  if [[ "$retcode" != "0" ]]; then
+    python "${MONOREPO_ROOT}"/.ci/premerge_advisor_upload.py \
+      $(git rev-parse HEAD~1) $GITHUB_RUN_NUMBER \
+      "${BUILD_DIR}"/test-results.*.xml "${MONOREPO_ROOT}"/ninja*.log
+    if [[ "$GITHUB_ACTIONS" != "" ]]; then
+      python "${MONOREPO_ROOT}"/.ci/premerge_advisor_explain.py \
+        $(git rev-parse HEAD~1) "${BUILD_DIR}"/test-results.*.xml \
+        "${MONOREPO_ROOT}"/ninja*.log
+    fi
+  fi
 }
 trap at-exit EXIT
 
@@ -56,6 +67,7 @@ function start-group {
 export PIP_BREAK_SYSTEM_PACKAGES=1
 pip install -q -r "${MONOREPO_ROOT}"/.ci/all_requirements.txt
 
-if [[ "$GITHUB_ACTIONS" != "" ]]; then
+# The ARM64 builders run on AWS and don't have access to the GCS cache.
+if [[ "$GITHUB_ACTIONS" != "" ]] && [[ "$RUNNER_ARCH" != "ARM64" ]]; then
   python .ci/cache_lit_timing_files.py download
 fi

.github/CODEOWNERS

@@ -53,6 +53,13 @@
 /mlir/include/mlir/Interfaces/DestinationStyleOpInterface.* @matthias-springer
 /mlir/lib/Interfaces/DestinationStyleOpInterface.* @matthias-springer
 
+# AMDGPU and ROCDL dialects in MLIR.
+/mlir/include/mlir/Dialect/AMDGPU @krzysz00 @kuhar
+/mlir/lib/Dialect/AMDGPU @krzysz00 @kuhar
+/mlir/lib/Conversion/*AMDGPU* @krzysz00 @kuhar
+/mlir/lib/Conversion/*ToROCDL @krzysz00 @kuhar
+/mlir/include/mlir/Dialect/LLVMIR/ROCDL* @krzysz00 @kuhar
+
 # Bufferization Dialect in MLIR.
 /mlir/include/mlir/Dialect/Bufferization @matthias-springer
 /mlir/lib/Dialect/Bufferization @matthias-springer
@@ -131,6 +138,7 @@
 /mlir/test/python/ @ftynse @makslevental @stellaraccident @rolfmorel
 /mlir/python/ @ftynse @makslevental @stellaraccident @rolfmorel
 /mlir/lib/Bindings/Python @makslevental @rolfmorel
+/mlir/include/Bindings/Python @makslevental @rolfmorel
 
 # MLIR Mem2Reg/SROA
 /mlir/**/Transforms/Mem2Reg.* @moxinilian

.github/new-prs-labeler.yml

@@ -1096,8 +1096,8 @@ clang:openmp:
   - llvm/test/Transforms/OpenMP/**
 
 clang:temporal-safety:
-  - clang/include/clang/Analysis/Analyses/LifetimeSafety*
-  - clang/lib/Analysis/LifetimeSafety*
+  - clang/include/clang/Analysis/Analyses/LifetimeSafety/**
+  - clang/lib/Analysis/LifetimeSafety/**
   - clang/unittests/Analysis/LifetimeSafety*
   - clang/test/Sema/*lifetime-safety*
   - clang/test/Sema/*lifetime-analysis*