Merged main:ef112833e11e94ea049f98bec4a29b4fe96a25dd into amd-gfx:d621799ca22b

Local branch amd-gfx d621799 Merged main:0d87e2577914a6384f4ad5952b8fa9b0d8e48da8 into amd-gfx:a22e10e4323c
Remote branch main ef11283 [MLIR][SCF] Add support for pipelining dynamic loops (llvm#74350)

Author: SC llvm team

Diff for: .git-blame-ignore-revs

@@ -64,3 +64,9 @@ f6d557ee34b6bbdb1dc32f29e34b4a4a8ad35e81
 
 # [libc++][NFC] clang-format <shared_mutex>
 2d7eb9c9ea1a146412a83603d5c0c6339a5d8284
+
+# [libc++] Rename _LIBCPP_INLINE_VISIBILITY to _LIBCPP_HIDE_FROM_ABI
+4c198542226223f6a5c5511a1f89b37d15ee10b9
+
+# [libc++] Replace uses of _VSTD:: by std:: (#74331)
+77a00c0d546cd4aa8311b5b9031ae9ea8cdb050c

Diff for: .github/CODEOWNERS

@@ -33,26 +33,42 @@
 
 /lldb/ @JDevlieghere
 
-/mlir/include/mlir/Interfaces/TilingInterface.* @MaheshRavishankar
+# Linalg in MLIR.
+/mlir/include/mlir/Dialect/Linalg @dcaballe @nicolasvasilache
+/mlir/lib/Dialect/Linalg @dcaballe @nicolasvasilache
 
-/mlir/lib/Dialect/Linalg/Transforms/DecomposeLinalgOps.cpp @MaheshRavishankar
-/mlir/lib/Dialect/Linalg/Transforms/DropUnitDims.cpp @MaheshRavishankar
-/mlir/lib/Dialect/Linalg/Transforms/ElementwiseOpFusion.cpp @MaheshRavishankar
-/mlir/lib/Dialect/MemRef/Transforms/EmulateNarrowType.cpp @MaheshRavishankar
-/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp @MaheshRavishankar
-/mlir/lib/Interfaces/TilingInterface.* @MaheshRavishankar
+# Vector in MLIR.
+/mlir/**/*AMX* @dcaballe
+/mlir/**/*Neon* @banach-space @dcaballe @nicolasvasilache
+/mlir/**/*SME* @banach-space @dcaballe @nicolasvasilache
+/mlir/**/*SVE* @banach-space @dcaballe @nicolasvasilache
+/mlir/**/*VectorInterfaces* @dcaballe @nicolasvasilache
+/mlir/**/*VectorToSCF* @banach-space @dcaballe @nicolasvasilache
+/mlir/**/*VectorToLLVM* @banach-space @dcaballe @nicolasvasilache
+/mlir/**/*X86Vector* @dcaballe @nicolasvasilache
+/mlir/include/mlir/Dialect/Vector @dcaballe @nicolasvasilache
+/mlir/lib/Dialect/Vector @dcaballe @nicolasvasilache
+
+/mlir/include/mlir/Interfaces/TilingInterface.* @MaheshRavishankar @nicolasvasilache
+
+/mlir/lib/Dialect/Linalg/Transforms/DecomposeLinalgOps.cpp @MaheshRavishankar @nicolasvasilache
+/mlir/lib/Dialect/Linalg/Transforms/DropUnitDims.cpp @MaheshRavishankar @nicolasvasilache
+/mlir/lib/Dialect/Linalg/Transforms/ElementwiseOpFusion.cpp @MaheshRavishankar @nicolasvasilache
+/mlir/lib/Dialect/MemRef/Transforms/EmulateNarrowType.cpp @MaheshRavishankar @nicolasvasilache
+/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp @MaheshRavishankar @nicolasvasilache
+/mlir/lib/Interfaces/TilingInterface.* @MaheshRavishankar @nicolasvasilache
 
 /mlir/**/*EmulateNarrowType* @hanhanW
-/mlir/lib/Dialect/Linalg/Transforms/DataLayoutPropagation.cpp @hanhanW
-/mlir/lib/Dialect/Linalg/Transforms/Transforms.cpp @hanhanW
-/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp @hanhanW
-/mlir/lib/Dialect/Tensor/IR/TensorTilingInterfaceImpl.cpp @hanhanW
-/mlir/lib/Dialect/Tensor/Transforms/FoldIntoPackAndUnpackPatterns.cpp @hanhanW
-/mlir/lib/Dialect/Vector/Transforms/* @hanhanW
+/mlir/lib/Dialect/Linalg/Transforms/DataLayoutPropagation.cpp @hanhanW @nicolasvasilache
+/mlir/lib/Dialect/Linalg/Transforms/Transforms.cpp @hanhanW @nicolasvasilache
+/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp @hanhanW @nicolasvasilache
+/mlir/lib/Dialect/Tensor/IR/TensorTilingInterfaceImpl.cpp @hanhanW @nicolasvasilache
+/mlir/lib/Dialect/Tensor/Transforms/FoldIntoPackAndUnpackPatterns.cpp @hanhanW @nicolasvasilache
+/mlir/lib/Dialect/Vector/Transforms/* @hanhanW @nicolasvasilache
 
 # Transform Dialect in MLIR.
-/mlir/include/mlir/Dialect/Transform/* @ftynse
-/mlir/lib/Dialect/Transform/* @ftynse
+/mlir/include/mlir/Dialect/Transform/* @ftynse @nicolasvasilache
+/mlir/lib/Dialect/Transform/* @ftynse @nicolasvasilache
 
 # SPIR-V in MLIR.
 /mlir/**/SPIRV/ @antiagainst @kuhar

Diff for: .github/workflows/libcxx-build-and-test.yaml

@@ -50,11 +50,10 @@ env:
 jobs:
   stage1:
     if: github.repository_owner == 'llvm'
-    runs-on:
-      group: libcxx-runners-8
+    runs-on: libcxx-runners-8-set
     continue-on-error: false
     strategy:
-      fail-fast: true
+      fail-fast: false
       matrix:
         config: [
           'generic-cxx03',
@@ -89,12 +88,11 @@ jobs:
             **/crash_diagnostics/*
   stage2:
     if: github.repository_owner == 'llvm'
-    runs-on:
-      group: libcxx-runners-8
+    runs-on: libcxx-runners-8-set
     needs: [ stage1 ]
     continue-on-error: false
     strategy:
-      fail-fast: true
+      fail-fast: false
       matrix:
         config: [
           'generic-cxx11',
@@ -169,27 +167,26 @@ jobs:
           'benchmarks',
           'bootstrapping-build'
         ]
-        machine: [ 'libcxx-runners-8' ]
+        machine: [ 'libcxx-runners-8-set' ]
         std_modules: [ 'OFF' ]
         include:
         - config: 'generic-cxx26'
-          machine: libcxx-runners-8
+          machine: libcxx-runners-8-set
           std_modules: 'ON'
         - config: 'generic-asan'
-          machine: libcxx-runners-8
+          machine: libcxx-runners-8-set
           std_modules: 'OFF'
         - config: 'generic-tsan'
-          machine: libcxx-runners-8
+          machine: libcxx-runners-8-set
           std_modules: 'OFF'
         - config: 'generic-ubsan'
-          machine: libcxx-runners-8
+          machine: libcxx-runners-8-set
           std_modules: 'OFF'
         # Use a larger machine for MSAN to avoid timeout and memory allocation issues.
         - config: 'generic-msan'
-          machine: libcxx-runners-32
+          machine: libcxx-runners-32-set
           std_modules: 'OFF'
-    runs-on:
-      group: ${{ matrix.machine }}
+    runs-on: ${{ matrix.machine }}
     steps:
       - uses: actions/checkout@v4
       - name: ${{ matrix.config }}

Diff for: .github/workflows/llvm-project-tests.yml

@@ -10,6 +10,11 @@ on:
         required: false
       projects:
         required: false
+      extra_cmake_args:
+        required: false
+      os_list:
+        required: false
+        default: '["ubuntu-latest", "windows-2019", "macOS-11"]'
   workflow_call:
     inputs:
       build_target:
@@ -20,6 +25,19 @@ on:
         required: true
         type: string
 
+      extra_cmake_args:
+        required: false
+        type: string
+
+      os_list:
+        required: false
+        type: string
+        # Use windows-2019 due to:
+        # https://developercommunity.visualstudio.com/t/Prev-Issue---with-__assume-isnan-/1597317
+        # We're using a specific version of macOS due to:
+        # https://github.com/actions/virtual-environments/issues/5900
+        default: '["ubuntu-latest", "windows-2019", "macOS-11"]'
+
 concurrency:
   # Skip intermediate builds: always.
   # Cancel intermediate builds: only if it is a pull request build.
@@ -35,14 +53,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        os:
-          - ubuntu-latest
-          # Use windows-2019 due to:
-          # https://developercommunity.visualstudio.com/t/Prev-Issue---with-__assume-isnan-/1597317
-          - windows-2019
-          # We're using a specific version of macOS due to:
-          # https://github.com/actions/virtual-environments/issues/5900
-          - macOS-11
+        os: ${{ fromJSON(inputs.os_list) }}
     steps:
       - name: Setup Windows
         if: startsWith(matrix.os, 'windows')
@@ -85,7 +96,7 @@ jobs:
           # This should be a no-op for non-mac OSes
           PKG_CONFIG_PATH: /usr/local/Homebrew/Library/Homebrew/os/mac/pkgconfig//12
         with:
-          cmake_args: '-GNinja -DLLVM_ENABLE_PROJECTS="${{ inputs.projects }}" -DCMAKE_BUILD_TYPE=Release -DLLDB_INCLUDE_TESTS=OFF -DCMAKE_C_COMPILER_LAUNCHER=sccache -DCMAKE_CXX_COMPILER_LAUNCHER=sccache'
+          cmake_args: '-GNinja -DLLVM_ENABLE_PROJECTS="${{ inputs.projects }}" -DCMAKE_BUILD_TYPE=Release -DLLDB_INCLUDE_TESTS=OFF -DCMAKE_C_COMPILER_LAUNCHER=sccache -DCMAKE_CXX_COMPILER_LAUNCHER=sccache ${{ inputs.extra_cmake_args }}'
           build_target: '${{ inputs.build_target }}'
 
       - name: Build and Test libclc

Diff for: .github/workflows/new-prs.yml

@@ -15,16 +15,43 @@ on:
       - synchronize
 
 jobs:
-  automate-prs-labels:
+  greeter:
+    runs-on: ubuntu-latest
     permissions:
       pull-requests: write
+    # Only comment on PRs that have been opened for the first time, by someone
+    # new to LLVM or to GitHub as a whole.
+    if: >-
+      (github.repository == 'llvm/llvm-project') &&
+      (github.event.action == 'opened') &&
+      (github.event.pull_request.author_association == 'FIRST_TIME_CONTRIBUTOR' ||
+       github.event.pull_request.author_association == 'FIRST_TIMER')
+    steps:
+      - name: Setup Automation Script
+        run: |
+          curl -O -L --fail https://raw.githubusercontent.com/"$GITHUB_REPOSITORY"/main/llvm/utils/git/github-automation.py
+          curl -O -L --fail https://raw.githubusercontent.com/"$GITHUB_REPOSITORY"/main/llvm/utils/git/requirements.txt
+          chmod a+x github-automation.py
+          pip install -r requirements.txt
+
+      - name: Greet Author
+        run: |
+          ./github-automation.py \
+            --token '${{ secrets.GITHUB_TOKEN }}' \
+            pr-greeter \
+            --issue-number "${{ github.event.pull_request.number }}"
+
+  automate-prs-labels:
+    # Greet first so that only the author gets that notification.
+    needs: greeter
     runs-on: ubuntu-latest
     # Ignore PRs with more than 10 commits.  Pull requests with a lot of
     # commits tend to be accidents usually when someone made a mistake while trying
     # to rebase.  We want to ignore these pull requests to avoid excessive
     # notifications.
+    # always() means that even if greeter is skipped, this job will run.
     if: >
-      github.repository == 'llvm/llvm-project' &&
+      always() && github.repository == 'llvm/llvm-project' &&
       github.event.pull_request.draft == false &&
       github.event.pull_request.commits < 10
     steps:

Diff for: .github/workflows/spirv-tests.yml

@@ -0,0 +1,29 @@
+name: SPIR-V Tests
+
+permissions:
+  contents: read
+
+on:
+  workflow_dispatch:
+  pull_request:
+    paths:
+      - 'llvm/lib/Target/SPIRV/**'
+      - 'llvm/test/CodeGen/SPIRV/**'
+      - '.github/workflows/spirv-tests.yml'
+
+concurrency:
+  # Skip intermediate builds: always.
+  # Cancel intermediate builds: only if it is a pull request build.
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: ${{ startsWith(github.ref, 'refs/pull/') }}
+
+jobs:
+  check_spirv:
+    if: github.repository_owner == 'llvm'
+    name: Test SPIR-V
+    uses: ./.github/workflows/llvm-project-tests.yml
+    with:
+      build_target: check-llvm-codegen-spirv
+      projects:
+      extra_cmake_args: '-DLLVM_TARGETS_TO_BUILD="" -DLLVM_EXPERIMENTAL_TARGETS_TO_BUILD="SPIRV"'
+      os_list: '["ubuntu-latest"]'

Diff for: .mailmap

@@ -24,6 +24,7 @@
 
 <compnerd@compnerd.org> <abdulras@fb.com>
 <compnerd@compnerd.org> <abdulras@google.com>
+<compnerd@compnerd.org> <abdulras@thebrowser.company>
 <git@xen0n.name> <xen0n@gentoo.org>
 <hans@hanshq.net> <hans@chromium.org>
 <i@maskray.me> <maskray@google.com>
@@ -32,6 +33,8 @@
 <qiucofan@cn.ibm.com> <qiucf@cn.ibm.com>
 <rnk@google.com> <reid@kleckner.net>
 <thakis@chromium.org> <nicolasweber@gmx.de>
+Jianjian GUAN <jacquesguan@me.com>
+Jianjian GUAN <jacquesguan@me.com> <Jianjian.Guan@streamcomputing.com>
 Jon Roelofs <jonathan_roelofs@apple.com> <jonathan@codesourcery.com>
 Jon Roelofs <jonathan_roelofs@apple.com> <jroelofs@jroelofs.com>
 Jonathan Thackray <jonathan.thackray@arm.com> <jthackray@users.noreply.github.com>
@@ -42,3 +45,6 @@ Med Ismail Bennani <ismail@bennani.ma> <medismail.bennani@gmail.com>
 Ramkumar Ramachandra <r@artagnon.com> <artagnon@gmail.com>
 Saleem Abdulrasool <compnerd@compnerd.org>
 Tommy Chiang <ty1208chiang@gmail.com> <oToToT@users.noreply.github.com>
+Wang Pengcheng <wangpengcheng.pp@bytedance.com>
+Wang Pengcheng <wangpengcheng.pp@bytedance.com> <137158460+wangpc-pp@users.noreply.github.com>
+Wang Pengcheng <wangpengcheng.pp@bytedance.com> <pc.wang@linux.alibaba.com>

Diff for: bolt/lib/Core/Exceptions.cpp

@@ -108,7 +108,8 @@ void BinaryFunction::parseLSDA(ArrayRef<uint8_t> LSDASectionData,
   DWARFDataExtractor Data(
       StringRef(reinterpret_cast<const char *>(LSDASectionData.data()),
                 LSDASectionData.size()),
-      BC.DwCtx->getDWARFObj().isLittleEndian(), 8);
+      BC.DwCtx->getDWARFObj().isLittleEndian(),
+      BC.DwCtx->getDWARFObj().getAddressSize());
   uint64_t Offset = getLSDAAddress() - LSDASectionAddress;
   assert(Data.isValidOffset(Offset) && "wrong LSDA address");
 

Diff for: bolt/lib/Passes/IndirectCallPromotion.cpp

@@ -158,6 +158,7 @@ static cl::opt<bool> ICPPeelForInline(
 
 } // namespace opts
 
+#ifndef NDEBUG
 static bool verifyProfile(std::map<uint64_t, BinaryFunction> &BFs) {
   bool IsValid = true;
   for (auto &BFI : BFs) {
@@ -182,6 +183,7 @@ static bool verifyProfile(std::map<uint64_t, BinaryFunction> &BFs) {
   }
   return IsValid;
 }
+#endif
 
 namespace llvm {
 namespace bolt {
@@ -1467,7 +1469,6 @@ void IndirectCallPromotion::runOnFunctions(BinaryContext &BC) {
                                std::max<uint64_t>(TotalIndexBasedCandidates, 1))
          << "%\n";
 
-  (void)verifyProfile;
 #ifndef NDEBUG
   verifyProfile(BFs);
 #endif

Diff for: bolt/lib/Passes/LongJmp.cpp

@@ -202,10 +202,23 @@ LongJmpPass::replaceTargetWithStub(BinaryBasicBlock &BB, MCInst &Inst,
     }
   } else if (LocalStubsIter != Stubs.end() &&
              LocalStubsIter->second.count(TgtBB)) {
-    // If we are replacing a local stub (because it is now out of range),
-    // use its target instead of creating a stub to jump to another stub
+    // The TgtBB and TgtSym now are the local out-of-range stub and its label.
+    // So, we are attempting to restore BB to its previous state without using
+    // this stub.
     TgtSym = BC.MIB->getTargetSymbol(*TgtBB->begin());
-    TgtBB = BB.getSuccessor(TgtSym, BI);
+    assert(TgtSym &&
+           "First instruction is expected to contain a target symbol.");
+    BinaryBasicBlock *TgtBBSucc = TgtBB->getSuccessor(TgtSym, BI);
+
+    // TgtBB might have no successor. e.g. a stub for a function call.
+    if (TgtBBSucc) {
+      BB.replaceSuccessor(TgtBB, TgtBBSucc, BI.Count, BI.MispredictedCount);
+      assert(TgtBB->getExecutionCount() >= BI.Count &&
+             "At least equal or greater than the branch count.");
+      TgtBB->setExecutionCount(TgtBB->getExecutionCount() - BI.Count);
+    }
+
+    TgtBB = TgtBBSucc;
   }
 
   BinaryBasicBlock *StubBB = lookupLocalStub(BB, Inst, TgtSym, DotAddress);

Diff for: bolt/lib/Passes/ReorderAlgorithm.cpp

@@ -20,7 +20,7 @@
 #include <random>
 #include <stack>
 
-#undef  DEBUG_TYPE
+#undef DEBUG_TYPE
 #define DEBUG_TYPE "bolt"
 
 using namespace llvm;
@@ -425,7 +425,7 @@ void TSPReorderAlgorithm::reorderBasicBlocks(BinaryFunction &BF,
   }
 
   std::vector<std::vector<int64_t>> DP;
-  DP.resize(1 << N);
+  DP.resize(static_cast<size_t>(1) << N);
   for (std::vector<int64_t> &Elmt : DP)
     Elmt.resize(N, -1);