[inductor] getting AOT inductor to treat None args correctly

linter

remove import

ghstack-source-id: 72ceaf4a8e8c5bb2c465cf293c1e436876186645
Pull Request resolved: https://github.com/pytorch/pytorch/pull/138910

lint

address feedback

lint

nit

.bazelversion

@@ -1 +1 @@
-6.5.0
+6.1.1

.buckconfig.oss

@@ -0,0 +1,23 @@
+[pt]
+  is_oss=1
+
+[buildfile]
+  name = BUCK.oss
+  includes = //tools/build_defs/select.bzl
+
+[repositories]
+  bazel_skylib = third_party/bazel-skylib/
+  ovr_config = .
+
+[download]
+  in_build = true
+
+[cxx]
+  cxxflags = -std=c++17
+  ldflags = -Wl,--no-undefined
+  should_remap_host_platform = true
+  cpp = /usr/bin/clang
+  cc = /usr/bin/clang
+  cxx = /usr/bin/clang++
+  cxxpp = /usr/bin/clang++
+  ld = /usr/bin/clang++

.ci/aarch64_linux/README.md

@@ -1,19 +0,0 @@
-# Aarch64 (ARM/Graviton) Support Scripts
-Scripts for building aarch64 PyTorch PIP Wheels. These scripts build the following wheels:
-* torch
-* torchvision
-* torchaudio
-* torchtext
-* torchdata
-## Aarch64_ci_build.sh
-This script is design to support CD operations within PyPi manylinux aarch64 container, and be executed in the container. It prepares the container and then executes __aarch64_wheel_ci_build.py__ to build the wheels. The script "assumes" the PyTorch repo is located at: ```/pytorch``` and will put the wheels into ```/artifacts```.
-### Usage
-```DESIRED_PYTHON=<PythonVersion> aarch64_ci_build.sh```
-
-__NOTE:__ CI build is currently __EXPERMINTAL__
-
-## Build_aarch64_wheel.py
-This app allows a person to build using AWS EC3 resources and requires AWS-CLI and Boto3 with AWS credentials to support building EC2 instances for the wheel builds. Can be used in a codebuild CD or from a local system.
-
-### Usage
-```build_aarch64_wheel.py --key-name <YourPemKey> --use-docker --python 3.8 --branch <RCtag>```

.ci/aarch64_linux/aarch64_ci_build.sh

@@ -1,29 +0,0 @@
-#!/bin/bash
-set -eux -o pipefail
-
-GPU_ARCH_VERSION=${GPU_ARCH_VERSION:-}
-
-# cuda arm build for Grace Hopper solely
-export TORCH_CUDA_ARCH_LIST="9.0"
-
-SCRIPTPATH="$( cd -- "$(dirname "$0")" >/dev/null 2>&1 ; pwd -P )"
-source $SCRIPTPATH/aarch64_ci_setup.sh
-
-###############################################################################
-# Run aarch64 builder python
-###############################################################################
-cd /
-# adding safe directory for git as the permissions will be
-# on the mounted pytorch repo
-git config --global --add safe.directory /pytorch
-pip install -r /pytorch/requirements.txt
-pip install auditwheel
-if [ "$DESIRED_CUDA" = "cpu" ]; then
-    echo "BASE_CUDA_VERSION is not set. Building cpu wheel."
-    #USE_PRIORITIZED_TEXT_FOR_LD for enable linker script optimization https://github.com/pytorch/pytorch/pull/121975/files
-    USE_PRIORITIZED_TEXT_FOR_LD=1 python /pytorch/.ci/aarch64_linux/aarch64_wheel_ci_build.py --enable-mkldnn
-else
-    echo "BASE_CUDA_VERSION is set to: $DESIRED_CUDA"
-    #USE_PRIORITIZED_TEXT_FOR_LD for enable linker script optimization https://github.com/pytorch/pytorch/pull/121975/files
-    USE_PRIORITIZED_TEXT_FOR_LD=1 python /pytorch/.ci/aarch64_linux/aarch64_wheel_ci_build.py --enable-mkldnn --enable-cuda
-fi

.ci/aarch64_linux/aarch64_ci_setup.sh

@@ -1,23 +0,0 @@
-#!/bin/bash
-set -eux -o pipefail
-
-# This script is used to prepare the Docker container for aarch64_ci_wheel_build.py python script
-# By creating symlinks from desired /opt/python to /usr/local/bin/
-
-NUMPY_VERSION=2.0.2
-PYGIT2_VERSION=1.15.1
-if [[ "$DESIRED_PYTHON"  == "3.13" ]]; then
-    NUMPY_VERSION=2.1.2
-    PYGIT2_VERSION=1.16.0
-fi
-
-SCRIPTPATH="$( cd "$(dirname "$0")" ; pwd -P )"
-source $SCRIPTPATH/../manywheel/set_desired_python.sh
-
-pip install -q numpy==${NUMPY_VERSION} pyyaml==6.0.2 scons==4.7.0 ninja==1.11.1 patchelf==0.17.2 pygit2==${PYGIT2_VERSION}
-
-for tool in python python3 pip pip3 ninja scons patchelf; do
-    ln -sf ${DESIRED_PYTHON_BIN_DIR}/${tool} /usr/local/bin;
-done
-
-python --version

.ci/aarch64_linux/aarch64_wheel_ci_build.py

@@ -1,230 +0,0 @@
-#!/usr/bin/env python3
-# encoding: UTF-8
-
-import os
-import shutil
-from subprocess import check_call, check_output
-from typing import List
-
-from pygit2 import Repository
-
-
-def list_dir(path: str) -> List[str]:
-    """'
-    Helper for getting paths for Python
-    """
-    return check_output(["ls", "-1", path]).decode().split("\n")
-
-
-def build_ArmComputeLibrary() -> None:
-    """
-    Using ArmComputeLibrary for aarch64 PyTorch
-    """
-    print("Building Arm Compute Library")
-    acl_build_flags = [
-        "debug=0",
-        "neon=1",
-        "opencl=0",
-        "os=linux",
-        "openmp=1",
-        "cppthreads=0",
-        "arch=armv8a",
-        "multi_isa=1",
-        "fixed_format_kernels=1",
-        "build=native",
-    ]
-    acl_install_dir = "/acl"
-    acl_checkout_dir = "ComputeLibrary"
-    os.makedirs(acl_install_dir)
-    check_call(
-        [
-            "git",
-            "clone",
-            "https://github.com/ARM-software/ComputeLibrary.git",
-            "-b",
-            "v24.09",
-            "--depth",
-            "1",
-            "--shallow-submodules",
-        ]
-    )
-
-    check_call(
-        ["scons", "Werror=1", "-j8", f"build_dir=/{acl_install_dir}/build"]
-        + acl_build_flags,
-        cwd=acl_checkout_dir,
-    )
-    for d in ["arm_compute", "include", "utils", "support", "src"]:
-        shutil.copytree(f"{acl_checkout_dir}/{d}", f"{acl_install_dir}/{d}")
-
-
-def update_wheel(wheel_path) -> None:
-    """
-    Update the cuda wheel libraries
-    """
-    folder = os.path.dirname(wheel_path)
-    wheelname = os.path.basename(wheel_path)
-    os.mkdir(f"{folder}/tmp")
-    os.system(f"unzip {wheel_path} -d {folder}/tmp")
-    libs_to_copy = [
-        "/usr/local/cuda/extras/CUPTI/lib64/libcupti.so.12",
-        "/usr/local/cuda/lib64/libcudnn.so.9",
-        "/usr/local/cuda/lib64/libcublas.so.12",
-        "/usr/local/cuda/lib64/libcublasLt.so.12",
-        "/usr/local/cuda/lib64/libcudart.so.12",
-        "/usr/local/cuda/lib64/libcufft.so.11",
-        "/usr/local/cuda/lib64/libcusparse.so.12",
-        "/usr/local/cuda/lib64/libcusparseLt.so.0",
-        "/usr/local/cuda/lib64/libcusolver.so.11",
-        "/usr/local/cuda/lib64/libcurand.so.10",
-        "/usr/local/cuda/lib64/libnvToolsExt.so.1",
-        "/usr/local/cuda/lib64/libnvJitLink.so.12",
-        "/usr/local/cuda/lib64/libnvrtc.so.12",
-        "/usr/local/cuda/lib64/libnvrtc-builtins.so.12.6",
-        "/usr/local/cuda/lib64/libcudnn_adv.so.9",
-        "/usr/local/cuda/lib64/libcudnn_cnn.so.9",
-        "/usr/local/cuda/lib64/libcudnn_graph.so.9",
-        "/usr/local/cuda/lib64/libcudnn_ops.so.9",
-        "/usr/local/cuda/lib64/libcudnn_engines_runtime_compiled.so.9",
-        "/usr/local/cuda/lib64/libcudnn_engines_precompiled.so.9",
-        "/usr/local/cuda/lib64/libcudnn_heuristic.so.9",
-        "/lib64/libgomp.so.1",
-        "/usr/lib64/libgfortran.so.5",
-        "/acl/build/libarm_compute.so",
-        "/acl/build/libarm_compute_graph.so",
-    ]
-    if enable_cuda:
-        libs_to_copy += [
-            "/usr/local/lib/libnvpl_lapack_lp64_gomp.so.0",
-            "/usr/local/lib/libnvpl_blas_lp64_gomp.so.0",
-            "/usr/local/lib/libnvpl_lapack_core.so.0",
-            "/usr/local/lib/libnvpl_blas_core.so.0",
-        ]
-    else:
-        libs_to_copy += [
-            "/opt/OpenBLAS/lib/libopenblas.so.0",
-        ]
-    # Copy libraries to unzipped_folder/a/lib
-    for lib_path in libs_to_copy:
-        lib_name = os.path.basename(lib_path)
-        shutil.copy2(lib_path, f"{folder}/tmp/torch/lib/{lib_name}")
-        os.system(
-            f"cd {folder}/tmp/torch/lib/; "
-            f"patchelf --set-rpath '$ORIGIN' --force-rpath {folder}/tmp/torch/lib/{lib_name}"
-        )
-    os.mkdir(f"{folder}/cuda_wheel")
-    os.system(f"cd {folder}/tmp/; zip -r {folder}/cuda_wheel/{wheelname} *")
-    shutil.move(
-        f"{folder}/cuda_wheel/{wheelname}",
-        f"{folder}/{wheelname}",
-        copy_function=shutil.copy2,
-    )
-    os.system(f"rm -rf {folder}/tmp/ {folder}/cuda_wheel/")
-
-
-def complete_wheel(folder: str) -> str:
-    """
-    Complete wheel build and put in artifact location
-    """
-    wheel_name = list_dir(f"/{folder}/dist")[0]
-
-    if "pytorch" in folder and not enable_cuda:
-        print("Repairing Wheel with AuditWheel")
-        check_call(["auditwheel", "repair", f"dist/{wheel_name}"], cwd=folder)
-        repaired_wheel_name = list_dir(f"/{folder}/wheelhouse")[0]
-
-        print(f"Moving {repaired_wheel_name} wheel to /{folder}/dist")
-        os.rename(
-            f"/{folder}/wheelhouse/{repaired_wheel_name}",
-            f"/{folder}/dist/{repaired_wheel_name}",
-        )
-    else:
-        repaired_wheel_name = wheel_name
-
-    print(f"Copying {repaired_wheel_name} to artifacts")
-    shutil.copy2(
-        f"/{folder}/dist/{repaired_wheel_name}", f"/artifacts/{repaired_wheel_name}"
-    )
-
-    return repaired_wheel_name
-
-
-def parse_arguments():
-    """
-    Parse inline arguments
-    """
-    from argparse import ArgumentParser
-
-    parser = ArgumentParser("AARCH64 wheels python CD")
-    parser.add_argument("--debug", action="store_true")
-    parser.add_argument("--build-only", action="store_true")
-    parser.add_argument("--test-only", type=str)
-    parser.add_argument("--enable-mkldnn", action="store_true")
-    parser.add_argument("--enable-cuda", action="store_true")
-    return parser.parse_args()
-
-
-if __name__ == "__main__":
-    """
-    Entry Point
-    """
-    args = parse_arguments()
-    enable_mkldnn = args.enable_mkldnn
-    enable_cuda = args.enable_cuda
-    repo = Repository("/pytorch")
-    branch = repo.head.name
-    if branch == "HEAD":
-        branch = "master"
-
-    print("Building PyTorch wheel")
-    build_vars = "MAX_JOBS=5 CMAKE_SHARED_LINKER_FLAGS=-Wl,-z,max-page-size=0x10000 "
-    os.system("cd /pytorch; python setup.py clean")
-
-    override_package_version = os.getenv("OVERRIDE_PACKAGE_VERSION")
-    if override_package_version is not None:
-        version = override_package_version
-        build_vars += (
-            f"BUILD_TEST=0 PYTORCH_BUILD_VERSION={version} PYTORCH_BUILD_NUMBER=1 "
-        )
-    elif branch in ["nightly", "master"]:
-        build_date = (
-            check_output(["git", "log", "--pretty=format:%cs", "-1"], cwd="/pytorch")
-            .decode()
-            .replace("-", "")
-        )
-        version = (
-            check_output(["cat", "version.txt"], cwd="/pytorch").decode().strip()[:-2]
-        )
-        if enable_cuda:
-            desired_cuda = os.getenv("DESIRED_CUDA")
-            build_vars += f"BUILD_TEST=0 PYTORCH_BUILD_VERSION={version}.dev{build_date}+{desired_cuda} PYTORCH_BUILD_NUMBER=1 "
-        else:
-            build_vars += f"BUILD_TEST=0 PYTORCH_BUILD_VERSION={version}.dev{build_date} PYTORCH_BUILD_NUMBER=1 "
-    elif branch.startswith(("v1.", "v2.")):
-        build_vars += f"BUILD_TEST=0 PYTORCH_BUILD_VERSION={branch[1:branch.find('-')]} PYTORCH_BUILD_NUMBER=1 "
-
-    if enable_mkldnn:
-        build_ArmComputeLibrary()
-        print("build pytorch with mkldnn+acl backend")
-        build_vars += (
-            "USE_MKLDNN=ON USE_MKLDNN_ACL=ON "
-            "ACL_ROOT_DIR=/acl "
-            "LD_LIBRARY_PATH=/pytorch/build/lib:/acl/build:$LD_LIBRARY_PATH "
-            "ACL_INCLUDE_DIR=/acl/build "
-            "ACL_LIBRARY=/acl/build "
-        )
-        if enable_cuda:
-            build_vars += "BLAS=NVPL "
-        else:
-            build_vars += "BLAS=OpenBLAS OpenBLAS_HOME=/OpenBLAS "
-    else:
-        print("build pytorch without mkldnn backend")
-
-    os.system(f"cd /pytorch; {build_vars} python3 setup.py bdist_wheel")
-    if enable_cuda:
-        print("Updating Cuda Dependency")
-        filename = os.listdir("/pytorch/dist/")
-        wheel_path = f"/pytorch/dist/{filename[0]}"
-        update_wheel(wheel_path)
-    pytorch_wheel_name = complete_wheel("/pytorch/")
-    print(f"Build Complete. Created {pytorch_wheel_name}..")

.ci/aarch64_linux/embed_library.py

@@ -1,87 +0,0 @@
-#!/usr/bin/env python3
-
-import os
-import shutil
-import sys
-from subprocess import check_call
-from tempfile import TemporaryDirectory
-
-from auditwheel.elfutils import elf_file_filter
-from auditwheel.lddtree import lddtree
-from auditwheel.patcher import Patchelf
-from auditwheel.repair import copylib
-from auditwheel.wheeltools import InWheelCtx
-
-
-def replace_tag(filename):
-    with open(filename) as f:
-        lines = f.read().split("\\n")
-    for i, line in enumerate(lines):
-        if not line.startswith("Tag: "):
-            continue
-        lines[i] = line.replace("-linux_", "-manylinux2014_")
-        print(f"Updated tag from {line} to {lines[i]}")
-
-    with open(filename, "w") as f:
-        f.write("\\n".join(lines))
-
-
-class AlignedPatchelf(Patchelf):
-    def set_soname(self, file_name: str, new_soname: str) -> None:
-        check_call(
-            ["patchelf", "--page-size", "65536", "--set-soname", new_soname, file_name]
-        )
-
-    def replace_needed(self, file_name: str, soname: str, new_soname: str) -> None:
-        check_call(
-            [
-                "patchelf",
-                "--page-size",
-                "65536",
-                "--replace-needed",
-                soname,
-                new_soname,
-                file_name,
-            ]
-        )
-
-
-def embed_library(whl_path, lib_soname, update_tag=False):
-    patcher = AlignedPatchelf()
-    out_dir = TemporaryDirectory()
-    whl_name = os.path.basename(whl_path)
-    tmp_whl_name = os.path.join(out_dir.name, whl_name)
-    with InWheelCtx(whl_path) as ctx:
-        torchlib_path = os.path.join(ctx._tmpdir.name, "torch", "lib")
-        ctx.out_wheel = tmp_whl_name
-        new_lib_path, new_lib_soname = None, None
-        for filename, _ in elf_file_filter(ctx.iter_files()):
-            if not filename.startswith("torch/lib"):
-                continue
-            libtree = lddtree(filename)
-            if lib_soname not in libtree["needed"]:
-                continue
-            lib_path = libtree["libs"][lib_soname]["path"]
-            if lib_path is None:
-                print(f"Can't embed {lib_soname} as it could not be found")
-                break
-            if lib_path.startswith(torchlib_path):
-                continue
-
-            if new_lib_path is None:
-                new_lib_soname, new_lib_path = copylib(lib_path, torchlib_path, patcher)
-            patcher.replace_needed(filename, lib_soname, new_lib_soname)
-            print(f"Replacing {lib_soname} with {new_lib_soname} for {filename}")
-        if update_tag:
-            # Add manylinux2014 tag
-            for filename in ctx.iter_files():
-                if os.path.basename(filename) != "WHEEL":
-                    continue
-                replace_tag(filename)
-    shutil.move(tmp_whl_name, whl_path)
-
-
-if __name__ == "__main__":
-    embed_library(
-        sys.argv[1], "libgomp.so.1", len(sys.argv) > 2 and sys.argv[2] == "--update-tag"
-    )

.ci/docker/android/AndroidManifest.xml

@@ -0,0 +1 @@
+<manifest package="org.pytorch.deps" />

.ci/docker/android/build.gradle

@@ -0,0 +1,66 @@
+buildscript {
+    ext {
+        minSdkVersion = 21
+        targetSdkVersion = 28
+        compileSdkVersion = 28
+        buildToolsVersion = '28.0.3'
+
+        coreVersion = "1.2.0"
+        extJUnitVersion = "1.1.1"
+        runnerVersion = "1.2.0"
+        rulesVersion = "1.2.0"
+        junitVersion = "4.12"
+    }
+
+    repositories {
+        google()
+        mavenLocal()
+        mavenCentral()
+        jcenter()
+    }
+
+    dependencies {
+        classpath 'com.android.tools.build:gradle:4.1.2'
+        classpath 'com.vanniktech:gradle-maven-publish-plugin:0.14.2'
+    }
+}
+
+repositories {
+    google()
+    jcenter()
+}
+
+apply plugin: 'com.android.library'
+
+android {
+    compileSdkVersion rootProject.compileSdkVersion
+    buildToolsVersion rootProject.buildToolsVersion
+
+    defaultConfig {
+        minSdkVersion minSdkVersion
+        targetSdkVersion targetSdkVersion
+    }
+
+    sourceSets {
+        main {
+            manifest.srcFile 'AndroidManifest.xml'
+        }
+    }
+}
+
+dependencies {
+    implementation 'com.android.support:appcompat-v7:28.0.0'
+    implementation 'androidx.appcompat:appcompat:1.0.0'
+    implementation 'com.facebook.fbjni:fbjni-java-only:0.2.2'
+    implementation 'com.google.code.findbugs:jsr305:3.0.1'
+    implementation 'com.facebook.soloader:nativeloader:0.10.5'
+
+    implementation 'junit:junit:' + rootProject.junitVersion
+    implementation 'androidx.test:core:' + rootProject.coreVersion
+
+    implementation 'junit:junit:' + rootProject.junitVersion
+    implementation 'androidx.test:core:' + rootProject.coreVersion
+    implementation 'androidx.test.ext:junit:' + rootProject.extJUnitVersion
+    implementation 'androidx.test:rules:' + rootProject.rulesVersion
+    implementation 'androidx.test:runner:' + rootProject.runnerVersion
+}

.ci/docker/aotriton_version.txt

@@ -0,0 +1,5 @@
+0.7b
+manylinux_2_17
+rocm6.2
+9be04068c3c0857a4cfd17d7e39e71d0423ebac2
+3e9e1959d23b93d78a08fcc5f868125dc3854dece32fd9458be9ef4467982291

.ci/docker/build.sh

@@ -179,10 +179,10 @@ case "$image" in
     TRITON=yes
     INDUCTOR_BENCHMARKS=yes
     ;;
-  pytorch-linux-focal-cuda12.4-cudnn9-py3.13-gcc9-inductor-benchmarks)
-    CUDA_VERSION=12.4.1
+  pytorch-linux-focal-cuda11.8-cudnn9-py3-gcc9)
+    CUDA_VERSION=11.8.0
     CUDNN_VERSION=9
-    ANACONDA_PYTHON_VERSION=3.13
+    ANACONDA_PYTHON_VERSION=3.10
     GCC_VERSION=9
     PROTOBUF=yes
     DB=yes
@@ -192,10 +192,9 @@ case "$image" in
     UCC_COMMIT=${_UCC_COMMIT}
     CONDA_CMAKE=yes
     TRITON=yes
-    INDUCTOR_BENCHMARKS=yes
     ;;
-  pytorch-linux-focal-cuda11.8-cudnn9-py3-gcc9)
-    CUDA_VERSION=11.8.0
+  pytorch-linux-focal-cuda12.4-cudnn9-py3-gcc9)
+    CUDA_VERSION=12.4.1
     CUDNN_VERSION=9
     ANACONDA_PYTHON_VERSION=3.10
     GCC_VERSION=9
@@ -222,6 +221,20 @@ case "$image" in
     CONDA_CMAKE=yes
     TRITON=yes
     ;;
+  pytorch-linux-focal-cuda12.4-cudnn9-py3-gcc9)
+    CUDA_VERSION=12.4.1
+    CUDNN_VERSION=9
+    ANACONDA_PYTHON_VERSION=3.10
+    GCC_VERSION=9
+    PROTOBUF=yes
+    DB=yes
+    VISION=yes
+    KATEX=yes
+    UCX_COMMIT=${_UCX_COMMIT}
+    UCC_COMMIT=${_UCC_COMMIT}
+    CONDA_CMAKE=yes
+    TRITON=yes
+    ;;
   pytorch-linux-focal-py3-clang10-onnx)
     ANACONDA_PYTHON_VERSION=3.9
     CLANG_VERSION=10
@@ -231,6 +244,16 @@ case "$image" in
     CONDA_CMAKE=yes
     ONNX=yes
     ;;
+  pytorch-linux-focal-py3-clang9-android-ndk-r21e)
+    ANACONDA_PYTHON_VERSION=3.9
+    CLANG_VERSION=9
+    LLVMDEV=yes
+    PROTOBUF=yes
+    ANDROID=yes
+    ANDROID_NDK_VERSION=r21e
+    GRADLE_VERSION=6.8.3
+    NINJA_VERSION=1.9.0
+    ;;
   pytorch-linux-focal-py3.9-clang10)
     ANACONDA_PYTHON_VERSION=3.9
     CLANG_VERSION=10
@@ -268,7 +291,7 @@ case "$image" in
     PROTOBUF=yes
     DB=yes
     VISION=yes
-    ROCM_VERSION=6.2.4
+    ROCM_VERSION=6.1
     NINJA_VERSION=1.9.0
     CONDA_CMAKE=yes
     TRITON=yes
@@ -279,7 +302,7 @@ case "$image" in
     PROTOBUF=yes
     DB=yes
     VISION=yes
-    ROCM_VERSION=6.3
+    ROCM_VERSION=6.2
     NINJA_VERSION=1.9.0
     CONDA_CMAKE=yes
     TRITON=yes
@@ -295,17 +318,6 @@ case "$image" in
     CONDA_CMAKE=yes
     TRITON=yes
     ;;
-  pytorch-linux-jammy-xpu-2025.0-py3)
-    ANACONDA_PYTHON_VERSION=3.9
-    GCC_VERSION=11
-    PROTOBUF=yes
-    DB=yes
-    VISION=yes
-    XPU_VERSION=2025.0
-    NINJA_VERSION=1.9.0
-    CONDA_CMAKE=yes
-    TRITON=yes
-    ;;
     pytorch-linux-jammy-py3.9-gcc11-inductor-benchmarks)
     ANACONDA_PYTHON_VERSION=3.9
     GCC_VERSION=11
@@ -402,6 +414,9 @@ case "$image" in
     DB=yes
     VISION=yes
     CONDA_CMAKE=yes
+    # snadampal: skipping sccache due to the following issue
+    # https://github.com/pytorch/pytorch/issues/121559
+    SKIP_SCCACHE_INSTALL=yes
     # snadampal: skipping llvm src build install because the current version
     # from pytorch/llvm:9.0.1 is x86 specific
     SKIP_LLVM_SRC_BUILD_INSTALL=yes
@@ -414,6 +429,9 @@ case "$image" in
     DB=yes
     VISION=yes
     CONDA_CMAKE=yes
+    # snadampal: skipping sccache due to the following issue
+    # https://github.com/pytorch/pytorch/issues/121559
+    SKIP_SCCACHE_INSTALL=yes
     # snadampal: skipping llvm src build install because the current version
     # from pytorch/llvm:9.0.1 is x86 specific
     SKIP_LLVM_SRC_BUILD_INSTALL=yes
@@ -490,6 +508,8 @@ docker build \
        --build-arg "CUDA_VERSION=${CUDA_VERSION}" \
        --build-arg "CUDNN_VERSION=${CUDNN_VERSION}" \
        --build-arg "TENSORRT_VERSION=${TENSORRT_VERSION}" \
+       --build-arg "ANDROID=${ANDROID}" \
+       --build-arg "ANDROID_NDK=${ANDROID_NDK_VERSION}" \
        --build-arg "GRADLE_VERSION=${GRADLE_VERSION}" \
        --build-arg "VULKAN_SDK_VERSION=${VULKAN_SDK_VERSION}" \
        --build-arg "SWIFTSHADER=${SWIFTSHADER}" \
@@ -497,7 +517,7 @@ docker build \
        --build-arg "NINJA_VERSION=${NINJA_VERSION:-}" \
        --build-arg "KATEX=${KATEX:-}" \
        --build-arg "ROCM_VERSION=${ROCM_VERSION:-}" \
-       --build-arg "PYTORCH_ROCM_ARCH=${PYTORCH_ROCM_ARCH:-gfx90a;gfx942}" \
+       --build-arg "PYTORCH_ROCM_ARCH=${PYTORCH_ROCM_ARCH:-gfx906;gfx90a}" \
        --build-arg "IMAGE_NAME=${IMAGE_NAME}" \
        --build-arg "UCX_COMMIT=${UCX_COMMIT}" \
        --build-arg "UCC_COMMIT=${UCC_COMMIT}" \

.ci/docker/centos-rocm/Dockerfile

@@ -113,6 +113,13 @@ COPY triton_version.txt triton_version.txt
 RUN if [ -n "${TRITON}" ]; then bash ./install_triton.sh; fi
 RUN rm install_triton.sh common_utils.sh triton.txt triton_version.txt
 
+# Install AOTriton (Early fail)
+COPY ./aotriton_version.txt aotriton_version.txt
+COPY ./common/common_utils.sh common_utils.sh
+COPY ./common/install_aotriton.sh install_aotriton.sh
+RUN ["/bin/bash", "-c", "./install_aotriton.sh /opt/rocm && rm -rf install_aotriton.sh aotriton_version.txt common_utils.sh"]
+ENV AOTRITON_INSTALLED_PREFIX /opt/rocm/aotriton
+
 # Install ccache/sccache (do this last, so we get priority in PATH)
 COPY ./common/install_cache.sh install_cache.sh
 ENV PATH /opt/cache/bin:$PATH

.ci/docker/ci_commit_pins/executorch.txt

@@ -1 +1 @@
-a29b208a06ab378bb29ab1aa68932e412f8e09f1
+16b633b4daa7f3d3442be62a3589bd60b2f7fdc7

.ci/docker/ci_commit_pins/triton-xpu.txt

@@ -1 +1 @@
-e98b6fcb8df5b44eb0d0addb6767c573d37ba024
+91b14bf5593cf58a8541f3e6b9125600a867d4ef

.ci/docker/ci_commit_pins/triton.txt

@@ -1 +1 @@
-0d4682f073ded4d1a8260dd4208a43d735ae3a2b
+cf34004b8a67d290a962da166f5aa2fc66751326

.ci/docker/common/install_acl.sh

@@ -1,7 +1,7 @@
 set -euo pipefail
 
 readonly version=v24.04
-readonly src_host=https://github.com/ARM-software
+readonly src_host=https://review.mlplatform.org/ml
 readonly src_repo=ComputeLibrary
 
 # Clone ACL

.ci/docker/common/install_android.sh

@@ -0,0 +1,112 @@
+#!/bin/bash
+
+set -ex
+
+[ -n "${ANDROID_NDK}" ]
+
+_https_amazon_aws=https://ossci-android.s3.amazonaws.com
+
+apt-get update
+apt-get install -y --no-install-recommends autotools-dev autoconf unzip
+apt-get autoclean && apt-get clean
+rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
+
+pushd /tmp
+curl -Os --retry 3 $_https_amazon_aws/android-ndk-${ANDROID_NDK}-linux-x86_64.zip
+popd
+_ndk_dir=/opt/ndk
+mkdir -p "$_ndk_dir"
+unzip -qo /tmp/android*.zip -d "$_ndk_dir"
+_versioned_dir=$(find "$_ndk_dir/" -mindepth 1 -maxdepth 1 -type d)
+mv "$_versioned_dir"/* "$_ndk_dir"/
+rmdir "$_versioned_dir"
+rm -rf /tmp/*
+
+# Install OpenJDK
+# https://hub.docker.com/r/picoded/ubuntu-openjdk-8-jdk/dockerfile/
+
+sudo apt-get update && \
+    apt-get install -y openjdk-8-jdk && \
+    apt-get install -y ant && \
+    apt-get clean && \
+    rm -rf /var/lib/apt/lists/* && \
+    rm -rf /var/cache/oracle-jdk8-installer;
+
+# Fix certificate issues, found as of
+# https://bugs.launchpad.net/ubuntu/+source/ca-certificates-java/+bug/983302
+
+sudo apt-get update && \
+    apt-get install -y ca-certificates-java && \
+    apt-get clean && \
+    update-ca-certificates -f && \
+    rm -rf /var/lib/apt/lists/* && \
+    rm -rf /var/cache/oracle-jdk8-installer;
+
+export JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64/
+
+# Installing android sdk
+# https://github.com/circleci/circleci-images/blob/staging/android/Dockerfile.m4
+
+_tmp_sdk_zip=/tmp/android-sdk-linux.zip
+_android_home=/opt/android/sdk
+
+rm -rf $_android_home
+sudo mkdir -p $_android_home
+curl --silent --show-error --location --fail --retry 3 --output /tmp/android-sdk-linux.zip $_https_amazon_aws/android-sdk-linux-tools3859397-build-tools2803-2902-platforms28-29.zip
+sudo unzip -q $_tmp_sdk_zip -d $_android_home
+rm $_tmp_sdk_zip
+
+sudo chmod -R 777 $_android_home
+
+export ANDROID_HOME=$_android_home
+export ADB_INSTALL_TIMEOUT=120
+
+export PATH="${ANDROID_HOME}/tools:${ANDROID_HOME}/tools/bin:${ANDROID_HOME}/platform-tools:${PATH}"
+echo "PATH:${PATH}"
+
+# Installing Gradle
+echo "GRADLE_VERSION:${GRADLE_VERSION}"
+_gradle_home=/opt/gradle
+sudo rm -rf $gradle_home
+sudo mkdir -p $_gradle_home
+
+curl --silent --output /tmp/gradle.zip --retry 3 $_https_amazon_aws/gradle-${GRADLE_VERSION}-bin.zip
+
+sudo unzip -q /tmp/gradle.zip -d $_gradle_home
+rm /tmp/gradle.zip
+
+sudo chmod -R 777 $_gradle_home
+
+export GRADLE_HOME=$_gradle_home/gradle-$GRADLE_VERSION
+alias gradle="${GRADLE_HOME}/bin/gradle"
+
+export PATH="${GRADLE_HOME}/bin/:${PATH}"
+echo "PATH:${PATH}"
+
+gradle --version
+
+mkdir /var/lib/jenkins/gradledeps
+cp build.gradle /var/lib/jenkins/gradledeps
+cp AndroidManifest.xml /var/lib/jenkins/gradledeps
+
+pushd /var/lib/jenkins
+
+export GRADLE_LOCAL_PROPERTIES=gradledeps/local.properties
+rm -f $GRADLE_LOCAL_PROPERTIES
+echo "sdk.dir=/opt/android/sdk" >> $GRADLE_LOCAL_PROPERTIES
+echo "ndk.dir=/opt/ndk" >> $GRADLE_LOCAL_PROPERTIES
+
+chown -R jenkins /var/lib/jenkins/gradledeps
+chgrp -R jenkins /var/lib/jenkins/gradledeps
+
+sudo -H -u jenkins $GRADLE_HOME/bin/gradle -Pandroid.useAndroidX=true -p /var/lib/jenkins/gradledeps -g /var/lib/jenkins/.gradle --refresh-dependencies --debug --stacktrace assemble
+
+chown -R jenkins /var/lib/jenkins/.gradle
+chgrp -R jenkins /var/lib/jenkins/.gradle
+
+popd
+
+rm -rf /var/lib/jenkins/.gradle/daemon
+
+# Cache vision models used by the test
+source "$(dirname "${BASH_SOURCE[0]}")/cache_vision_models.sh"

.ci/docker/common/install_aotriton.sh

@@ -0,0 +1,23 @@
+#!/bin/bash
+
+set -ex
+
+source "$(dirname "${BASH_SOURCE[0]}")/common_utils.sh"
+
+TARBALL='aotriton.tar.gz'
+# This read command alwasy returns with exit code 1
+read -d "\n" VER MANYLINUX ROCMBASE PINNED_COMMIT SHA256 < aotriton_version.txt || true
+ARCH=$(uname -m)
+AOTRITON_INSTALL_PREFIX="$1"
+AOTRITON_URL="https://github.com/ROCm/aotriton/releases/download/${VER}/aotriton-${VER}-${MANYLINUX}_${ARCH}-${ROCMBASE}-shared.tar.gz"
+
+cd "${AOTRITON_INSTALL_PREFIX}"
+# Must use -L to follow redirects
+curl -L --retry 3 -o "${TARBALL}" "${AOTRITON_URL}"
+ACTUAL_SHA256=$(sha256sum "${TARBALL}" | cut -d " " -f 1)
+if [ "${SHA256}" != "${ACTUAL_SHA256}" ]; then
+  echo -n "Error: The SHA256 of downloaded tarball is ${ACTUAL_SHA256},"
+  echo " which does not match the expected value ${SHA256}."
+  exit
+fi
+tar xf "${TARBALL}" && rm -rf "${TARBALL}"

.ci/docker/common/install_base.sh

@@ -76,8 +76,7 @@ install_ubuntu() {
     vim \
     unzip \
     gpg-agent \
-    gdb \
-    bc
+    gdb
 
   # Should resolve issues related to various apt package repository cert issues
   # see: https://github.com/pytorch/pytorch/issues/65931

.ci/docker/common/install_cache.sh

@@ -9,7 +9,7 @@ install_ubuntu() {
   # Instead use lib and headers from OpenSSL1.1 installed in `install_openssl.sh``
   apt-get install -y cargo
   echo "Checking out sccache repo"
-  git clone https://github.com/mozilla/sccache -b v0.9.0
+  git clone https://github.com/pytorch/sccache
   cd sccache
   echo "Building sccache"
   cargo build --release
@@ -19,10 +19,6 @@ install_ubuntu() {
   rm -rf sccache
   apt-get remove -y cargo rustc
   apt-get autoclean && apt-get clean
-
-  echo "Downloading old sccache binary from S3 repo for PCH builds"
-  curl --retry 3 https://s3.amazonaws.com/ossci-linux/sccache -o /opt/cache/bin/sccache-0.2.14a
-  chmod 755 /opt/cache/bin/sccache-0.2.14a
 }
 
 install_binary() {
@@ -36,42 +32,22 @@ sed -e 's|PATH="\(.*\)"|PATH="/opt/cache/bin:\1"|g' -i /etc/environment
 export PATH="/opt/cache/bin:$PATH"
 
 # Setup compiler cache
-install_ubuntu
+if [ -n "$ROCM_VERSION" ]; then
+  curl --retry 3 http://repo.radeon.com/misc/.sccache_amd/sccache -o /opt/cache/bin/sccache
+else
+  ID=$(grep -oP '(?<=^ID=).+' /etc/os-release | tr -d '"')
+  # TODO: Install the pre-built binary from S3 as building from source
+  # https://github.com/pytorch/sccache has started failing mysteriously
+  # in which sccache server couldn't start with the following error:
+  #   sccache: error: Invalid argument (os error 22)
+  install_binary
+fi
 chmod a+x /opt/cache/bin/sccache
 
 function write_sccache_stub() {
   # Unset LD_PRELOAD for ps because of asan + ps issues
   # https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90589
-  if [ $1 == "gcc" ]; then
-    # Do not call sccache recursively when dumping preprocessor argument
-    # For some reason it's very important for the first cached nvcc invocation
-    cat >"/opt/cache/bin/$1" <<EOF
-#!/bin/sh
-
-# sccache does not support -E flag, so we need to call the original compiler directly in order to avoid calling this wrapper recursively
-for arg in "\$@"; do
-  if [ "\$arg" = "-E" ]; then
-    exec $(which $1) "\$@"
-  fi
-done
-
-if [ \$(env -u LD_PRELOAD ps -p \$PPID -o comm=) != sccache ]; then
-  exec sccache $(which $1) "\$@"
-else
-  exec $(which $1) "\$@"
-fi
-EOF
-  else
-    cat >"/opt/cache/bin/$1" <<EOF
-#!/bin/sh
-
-if [ \$(env -u LD_PRELOAD ps -p \$PPID -o comm=) != sccache ]; then
-  exec sccache $(which $1) "\$@"
-else
-  exec $(which $1) "\$@"
-fi
-EOF
-  fi
+  printf "#!/bin/sh\nif [ \$(env -u LD_PRELOAD ps -p \$PPID -o comm=) != sccache ]; then\n  exec sccache $(which $1) \"\$@\"\nelse\n  exec $(which $1) \"\$@\"\nfi" > "/opt/cache/bin/$1"
   chmod a+x "/opt/cache/bin/$1"
 }
 
@@ -112,7 +88,7 @@ if [ -n "$ROCM_VERSION" ]; then
     TOPDIR=$(dirname $OLDCOMP)
     WRAPPED="$TOPDIR/original/$COMPNAME"
     mv "$OLDCOMP" "$WRAPPED"
-    printf "#!/bin/sh\nexec sccache $WRAPPED \"\$@\"" >"$OLDCOMP"
+    printf "#!/bin/sh\nexec sccache $WRAPPED \"\$@\"" > "$OLDCOMP"
     chmod a+x "$OLDCOMP"
   }
 

.ci/docker/common/install_conda.sh

@@ -25,8 +25,7 @@ if [ -n "$ANACONDA_PYTHON_VERSION" ]; then
   mkdir -p /opt/conda
   chown jenkins:jenkins /opt/conda
 
-  SCRIPT_FOLDER="$( cd "$(dirname "$0")" ; pwd -P )"
-  source "${SCRIPT_FOLDER}/common_utils.sh"
+  source "$(dirname "${BASH_SOURCE[0]}")/common_utils.sh"
 
   pushd /tmp
   wget -q "${BASE_URL}/${CONDA_FILE}"
@@ -66,7 +65,7 @@ if [ -n "$ANACONDA_PYTHON_VERSION" ]; then
 
   # Install PyTorch conda deps, as per https://github.com/pytorch/pytorch README
   if [[ $(uname -m) == "aarch64" ]]; then
-    conda_install "openblas==0.3.28=*openmp*"
+    conda_install "openblas==0.3.25=*openmp*"
   else
     conda_install "mkl=2021.4.0 mkl-include=2021.4.0"
   fi
@@ -85,9 +84,8 @@ if [ -n "$ANACONDA_PYTHON_VERSION" ]; then
 
   # Magma package names are concatenation of CUDA major and minor ignoring revision
   # I.e. magma-cuda102 package corresponds to CUDA_VERSION=10.2 and CUDA_VERSION=10.2.89
-  # Magma is installed from a tarball in the ossci-linux bucket into the conda env
   if [ -n "$CUDA_VERSION" ]; then
-    ${SCRIPT_FOLDER}/install_magma_conda.sh $(cut -f1-2 -d'.' <<< ${CUDA_VERSION}) ${ANACONDA_PYTHON_VERSION}
+    conda_install magma-cuda$(TMP=${CUDA_VERSION/./};echo ${TMP%.*[0-9]}) -c pytorch
   fi
 
   # Install some other packages, including those needed for Python test reporting

.ci/docker/common/install_cpython.sh

@@ -70,7 +70,7 @@ function do_cpython_build {
     # install setuptools since python 3.12 is required to use distutils
     ${prefix}/bin/pip install wheel==0.34.2 setuptools==68.2.2
     local abi_tag=$(${prefix}/bin/python -c "from wheel.pep425tags import get_abbr_impl, get_impl_ver, get_abi_tag; print('{0}{1}-{2}'.format(get_abbr_impl(), get_impl_ver(), get_abi_tag()))")
-    ln -sf ${prefix} /opt/python/${abi_tag}
+    ln -s ${prefix} /opt/python/${abi_tag}
 }
 
 function build_cpython {

.ci/docker/common/install_cuda.sh

@@ -3,7 +3,7 @@
 set -ex
 
 NCCL_VERSION=v2.21.5-1
-CUDNN_VERSION=9.5.1.17
+CUDNN_VERSION=9.1.0.70
 
 function install_cusparselt_040 {
     # cuSparseLt license: https://docs.nvidia.com/cuda/cusparselt/license.html
@@ -38,19 +38,7 @@ function install_cusparselt_062 {
     rm -rf tmp_cusparselt
 }
 
-function install_cusparselt_063 {
-    # cuSparseLt license: https://docs.nvidia.com/cuda/cusparselt/license.html
-    mkdir tmp_cusparselt && pushd tmp_cusparselt
-    wget -q https://developer.download.nvidia.com/compute/cusparselt/redist/libcusparse_lt/linux-x86_64/libcusparse_lt-linux-x86_64-0.6.3.2-archive.tar.xz
-    tar xf libcusparse_lt-linux-x86_64-0.6.3.2-archive.tar.xz
-    cp -a libcusparse_lt-linux-x86_64-0.6.3.2-archive/include/* /usr/local/cuda/include/
-    cp -a libcusparse_lt-linux-x86_64-0.6.3.2-archive/lib/* /usr/local/cuda/lib64/
-    popd
-    rm -rf tmp_cusparselt
-}
-
 function install_118 {
-    CUDNN_VERSION=9.1.0.70
     echo "Installing CUDA 11.8 and cuDNN ${CUDNN_VERSION} and NCCL ${NCCL_VERSION} and cuSparseLt-0.4.0"
     rm -rf /usr/local/cuda-11.8 /usr/local/cuda
     # install CUDA 11.8.0 in the same container
@@ -117,7 +105,6 @@ function install_121 {
 }
 
 function install_124 {
-  CUDNN_VERSION=9.1.0.70
   echo "Installing CUDA 12.4.1 and cuDNN ${CUDNN_VERSION} and NCCL ${NCCL_VERSION} and cuSparseLt-0.6.2"
   rm -rf /usr/local/cuda-12.4 /usr/local/cuda
   # install CUDA 12.4.1 in the same container
@@ -151,13 +138,13 @@ function install_124 {
 }
 
 function install_126 {
-  echo "Installing CUDA 12.6.3 and cuDNN ${CUDNN_VERSION} and NCCL ${NCCL_VERSION} and cuSparseLt-0.6.3"
+  echo "Installing CUDA 12.6.2 and cuDNN ${CUDNN_VERSION} and NCCL ${NCCL_VERSION} and cuSparseLt-0.6.2"
   rm -rf /usr/local/cuda-12.6 /usr/local/cuda
-  # install CUDA 12.6.3 in the same container
-  wget -q https://developer.download.nvidia.com/compute/cuda/12.6.3/local_installers/cuda_12.6.3_560.35.05_linux.run
-  chmod +x cuda_12.6.3_560.35.05_linux.run
-  ./cuda_12.6.3_560.35.05_linux.run --toolkit --silent
-  rm -f cuda_12.6.3_560.35.05_linux.run
+  # install CUDA 12.6.2 in the same container
+  wget -q https://developer.download.nvidia.com/compute/cuda/12.6.2/local_installers/cuda_12.6.2_560.35.03_linux.run
+  chmod +x cuda_12.6.2_560.35.03_linux.run
+  ./cuda_12.6.2_560.35.03_linux.run --toolkit --silent
+  rm -f cuda_12.6.2_560.35.03_linux.run
   rm -f /usr/local/cuda && ln -s /usr/local/cuda-12.6 /usr/local/cuda
 
   # cuDNN license: https://developer.nvidia.com/cudnn/license_agreement
@@ -178,7 +165,7 @@ function install_126 {
   cd ..
   rm -rf nccl
 
-  install_cusparselt_063
+  install_cusparselt_062
 
   ldconfig
 }

.ci/docker/common/install_cuda_aarch64.sh

@@ -4,7 +4,6 @@
 set -ex
 
 NCCL_VERSION=v2.21.5-1
-CUDNN_VERSION=9.5.1.17
 
 function install_cusparselt_062 {
     # cuSparseLt license: https://docs.nvidia.com/cuda/cusparselt/license.html
@@ -17,20 +16,8 @@ function install_cusparselt_062 {
     rm -rf tmp_cusparselt
 }
 
-function install_cusparselt_063 {
-    # cuSparseLt license: https://docs.nvidia.com/cuda/cusparselt/license.html
-    mkdir tmp_cusparselt && pushd tmp_cusparselt
-    wget -q https://developer.download.nvidia.com/compute/cusparselt/redist/libcusparse_lt/linux-sbsa/libcusparse_lt-linux-sbsa-0.6.3.2-archive.tar.xz
-    tar xf libcusparse_lt-linux-sbsa-0.6.3.2-archive.tar.xz
-    cp -a libcusparse_lt-linux-sbsa-0.6.3.2-archive/include/* /usr/local/cuda/include/
-    cp -a libcusparse_lt-linux-sbsa-0.6.3.2-archive/lib/* /usr/local/cuda/lib64/
-    popd
-    rm -rf tmp_cusparselt
-}
-
 function install_124 {
-  CUDNN_VERSION=9.1.0.70
-  echo "Installing CUDA 12.4.1 and cuDNN ${CUDNN_VERSION} and NCCL ${NCCL_VERSION} and cuSparseLt-0.6.2"
+  echo "Installing CUDA 12.4.1 and cuDNN 9.1 and NCCL ${NCCL_VERSION} and cuSparseLt-0.6.2"
   rm -rf /usr/local/cuda-12.4 /usr/local/cuda
   # install CUDA 12.4.1 in the same container
   wget -q https://developer.download.nvidia.com/compute/cuda/12.4.1/local_installers/cuda_12.4.1_550.54.15_linux_sbsa.run
@@ -41,10 +28,10 @@ function install_124 {
 
   # cuDNN license: https://developer.nvidia.com/cudnn/license_agreement
   mkdir tmp_cudnn && cd tmp_cudnn
-  wget -q https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/linux-sbsa/cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive.tar.xz -O cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive.tar.xz
-  tar xf cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive.tar.xz
-  cp -a cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive/include/* /usr/local/cuda/include/
-  cp -a cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive/lib/* /usr/local/cuda/lib64/
+  wget -q https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/linux-sbsa/cudnn-linux-sbsa-9.1.0.70_cuda12-archive.tar.xz -O cudnn-linux-sbsa-9.1.0.70_cuda12-archive.tar.xz
+  tar xf cudnn-linux-sbsa-9.1.0.70_cuda12-archive.tar.xz
+  cp -a cudnn-linux-sbsa-9.1.0.70_cuda12-archive/include/* /usr/local/cuda/include/
+  cp -a cudnn-linux-sbsa-9.1.0.70_cuda12-archive/lib/* /usr/local/cuda/lib64/
   cd ..
   rm -rf tmp_cudnn
 
@@ -87,87 +74,18 @@ function prune_124 {
   $NVPRUNE $GENCODE_CUDNN $CUDA_LIB_DIR/libcublasLt_static.a -o $CUDA_LIB_DIR/libcublasLt_static.a
 
   #####################################################################################
-  # CUDA 12.4 prune visual tools
+  # CUDA 12.1 prune visual tools
   #####################################################################################
   export CUDA_BASE="/usr/local/cuda-12.4/"
   rm -rf $CUDA_BASE/libnvvp $CUDA_BASE/nsightee_plugins $CUDA_BASE/nsight-compute-2024.1.0 $CUDA_BASE/nsight-systems-2023.4.4/
 }
 
-function install_126 {
-  echo "Installing CUDA 12.6.3 and cuDNN ${CUDNN_VERSION} and NCCL ${NCCL_VERSION} and cuSparseLt-0.6.3"
-  rm -rf /usr/local/cuda-12.6 /usr/local/cuda
-  # install CUDA 12.6.3 in the same container
-  wget -q https://developer.download.nvidia.com/compute/cuda/12.6.3/local_installers/cuda_12.6.3_560.35.05_linux_sbsa.run
-  chmod +x cuda_12.6.3_560.35.05_linux_sbsa.run
-  ./cuda_12.6.3_560.35.05_linux_sbsa.run --toolkit --silent
-  rm -f cuda_12.6.3_560.35.05_linux_sbsa.run
-  rm -f /usr/local/cuda && ln -s /usr/local/cuda-12.6 /usr/local/cuda
-
-  # cuDNN license: https://developer.nvidia.com/cudnn/license_agreement
-  mkdir tmp_cudnn && cd tmp_cudnn
-  wget -q https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/linux-sbsa/cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive.tar.xz -O cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive.tar.xz
-  tar xf cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive.tar.xz
-  cp -a cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive/include/* /usr/local/cuda/include/
-  cp -a cudnn-linux-sbsa-${CUDNN_VERSION}_cuda12-archive/lib/* /usr/local/cuda/lib64/
-  cd ..
-  rm -rf tmp_cudnn
-
-  # NCCL license: https://docs.nvidia.com/deeplearning/nccl/#licenses
-  # Follow build: https://github.com/NVIDIA/nccl/tree/master?tab=readme-ov-file#build
-  git clone -b ${NCCL_VERSION} --depth 1 https://github.com/NVIDIA/nccl.git
-  cd nccl && make -j src.build
-  cp -a build/include/* /usr/local/cuda/include/
-  cp -a build/lib/* /usr/local/cuda/lib64/
-  cd ..
-  rm -rf nccl
-
-  install_cusparselt_063
-
-  ldconfig
-}
-
-function prune_126 {
-  echo "Pruning CUDA 12.6"
-  #####################################################################################
-  # CUDA 12.6 prune static libs
-  #####################################################################################
-  export NVPRUNE="/usr/local/cuda-12.6/bin/nvprune"
-  export CUDA_LIB_DIR="/usr/local/cuda-12.6/lib64"
-
-  export GENCODE="-gencode arch=compute_50,code=sm_50 -gencode arch=compute_60,code=sm_60 -gencode arch=compute_70,code=sm_70 -gencode arch=compute_75,code=sm_75 -gencode arch=compute_80,code=sm_80 -gencode arch=compute_86,code=sm_86 -gencode arch=compute_90,code=sm_90"
-  export GENCODE_CUDNN="-gencode arch=compute_50,code=sm_50 -gencode arch=compute_60,code=sm_60 -gencode arch=compute_61,code=sm_61 -gencode arch=compute_70,code=sm_70 -gencode arch=compute_75,code=sm_75 -gencode arch=compute_80,code=sm_80 -gencode arch=compute_86,code=sm_86 -gencode arch=compute_90,code=sm_90"
-
-  if [[ -n "$OVERRIDE_GENCODE" ]]; then
-      export GENCODE=$OVERRIDE_GENCODE
-  fi
-  if [[ -n "$OVERRIDE_GENCODE_CUDNN" ]]; then
-      export GENCODE_CUDNN=$OVERRIDE_GENCODE_CUDNN
-  fi
-
-  # all CUDA libs except CuDNN and CuBLAS
-  ls $CUDA_LIB_DIR/ | grep "\.a" | grep -v "culibos" | grep -v "cudart" | grep -v "cudnn" | grep -v "cublas" | grep -v "metis"  \
-      | xargs -I {} bash -c \
-                "echo {} && $NVPRUNE $GENCODE $CUDA_LIB_DIR/{} -o $CUDA_LIB_DIR/{}"
-
-  # prune CuDNN and CuBLAS
-  $NVPRUNE $GENCODE_CUDNN $CUDA_LIB_DIR/libcublas_static.a -o $CUDA_LIB_DIR/libcublas_static.a
-  $NVPRUNE $GENCODE_CUDNN $CUDA_LIB_DIR/libcublasLt_static.a -o $CUDA_LIB_DIR/libcublasLt_static.a
-
-  #####################################################################################
-  # CUDA 12.6 prune visual tools
-  #####################################################################################
-  export CUDA_BASE="/usr/local/cuda-12.6/"
-  rm -rf $CUDA_BASE/libnvvp $CUDA_BASE/nsightee_plugins $CUDA_BASE/nsight-compute-2024.3.2 $CUDA_BASE/nsight-systems-2024.5.1/
-}
-
 # idiomatic parameter and option handling in sh
 while test $# -gt 0
 do
     case "$1" in
     12.4) install_124; prune_124
         ;;
-    12.6) install_126; prune_126
-        ;;
     *) echo "bad argument $1"; exit 1
         ;;
     esac

.ci/docker/common/install_cudnn.sh

@@ -4,9 +4,7 @@ if [[ -n "${CUDNN_VERSION}" ]]; then
     # cuDNN license: https://developer.nvidia.com/cudnn/license_agreement
     mkdir tmp_cudnn
     pushd tmp_cudnn
-    if [[ ${CUDA_VERSION:0:4} == "12.6" ]]; then
-        CUDNN_NAME="cudnn-linux-x86_64-9.5.1.17_cuda12-archive"
-    elif [[ ${CUDA_VERSION:0:2} == "12" ]]; then
+    if [[ ${CUDA_VERSION:0:2} == "12" ]]; then
         CUDNN_NAME="cudnn-linux-x86_64-9.1.0.70_cuda12-archive"
     elif [[ ${CUDA_VERSION:0:2} == "11" ]]; then
         CUDNN_NAME="cudnn-linux-x86_64-9.1.0.70_cuda11-archive"

.ci/docker/common/install_executorch.sh

@@ -36,19 +36,25 @@ install_conda_dependencies() {
 }
 
 install_pip_dependencies() {
-  pushd executorch
-  as_jenkins bash install_requirements.sh --pybind xnnpack
+  pushd executorch/.ci/docker
+  # Install PyTorch CPU build beforehand to avoid installing the much bigger CUDA
+  # binaries later, ExecuTorch only needs CPU
+  pip_install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
+  # Install all Python dependencies
+  pip_install -r requirements-ci.txt
   popd
 }
 
 setup_executorch() {
   pushd executorch
+  # Setup swiftshader and Vulkan SDK which are required to build the Vulkan delegate
+  as_jenkins bash .ci/scripts/setup-vulkan-linux-deps.sh
 
   export PYTHON_EXECUTABLE=python
   export EXECUTORCH_BUILD_PYBIND=ON
   export CMAKE_ARGS="-DEXECUTORCH_BUILD_XNNPACK=ON -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON"
 
-  as_jenkins .ci/scripts/setup-linux.sh cmake || true
+  as_jenkins .ci/scripts/setup-linux.sh cmake
   popd
 }
 

.ci/docker/common/install_inductor_benchmark_deps.sh

@@ -7,20 +7,14 @@ source "$(dirname "${BASH_SOURCE[0]}")/common_utils.sh"
 function install_huggingface() {
   local version
   commit=$(get_pinned_commit huggingface)
+  pip_install pandas==2.0.3
   pip_install "git+https://github.com/huggingface/transformers@${commit}"
 }
 
 function install_timm() {
   local commit
   commit=$(get_pinned_commit timm)
-
-  # TODO (huydhn): There is no torchvision release on 3.13 when I write this, so
-  # I'm using nightly here instead. We just need to package to be able to install
-  # TIMM. Removing this once vision has a release on 3.13
-  if [[ "${ANACONDA_PYTHON_VERSION}" == "3.13" ]]; then
-    pip_install --pre torch torchvision --index-url https://download.pytorch.org/whl/nightly/cu124
-  fi
-
+  pip_install pandas==2.0.3
   pip_install "git+https://github.com/huggingface/pytorch-image-models@${commit}"
   # Clean up
   conda_run pip uninstall -y cmake torch torchvision triton

.ci/docker/common/install_magma.sh

@@ -3,6 +3,8 @@
 
 set -eou pipefail
 
+MAGMA_VERSION="2.5.2"
+
 function do_install() {
     cuda_version=$1
     cuda_version_nodot=${1/./}
@@ -15,7 +17,7 @@ function do_install() {
         set -x
         tmp_dir=$(mktemp -d)
         pushd ${tmp_dir}
-        curl -OLs https://ossci-linux.s3.us-east-1.amazonaws.com/${magma_archive}
+        curl -OLs https://anaconda.org/pytorch/magma-cuda${cuda_version_nodot}/${MAGMA_VERSION}/download/linux-64/${magma_archive}
         tar -xvf "${magma_archive}"
         mkdir -p "${cuda_dir}/magma"
         mv include "${cuda_dir}/magma/include"

.ci/docker/common/install_magma_conda.sh

@@ -1,26 +0,0 @@
-#!/usr/bin/env bash
-# Script that replaces the magma install from a conda package
-
-set -eou pipefail
-
-function do_install() {
-    cuda_version_nodot=${1/./}
-    anaconda_python_version=$2
-
-    MAGMA_VERSION="2.6.1"
-    magma_archive="magma-cuda${cuda_version_nodot}-${MAGMA_VERSION}-1.tar.bz2"
-
-    anaconda_dir="/opt/conda/envs/py_${anaconda_python_version}"
-    (
-        set -x
-        tmp_dir=$(mktemp -d)
-        pushd ${tmp_dir}
-        curl -OLs https://ossci-linux.s3.us-east-1.amazonaws.com/${magma_archive}
-        tar -xvf "${magma_archive}"
-        mv include/* "${anaconda_dir}/include/"
-        mv lib/* "${anaconda_dir}/lib"
-        popd
-    )
-}
-
-do_install $1 $2

.ci/docker/common/install_miopen.sh

@@ -16,7 +16,7 @@ case "$ID" in
   ubuntu)
     IS_UBUNTU=1
     ;;
-  centos|almalinux)
+  centos)
     IS_UBUNTU=0
     ;;
   *)
@@ -43,6 +43,12 @@ else
 fi
 ROCM_INT=$(($ROCM_VERSION_MAJOR * 10000 + $ROCM_VERSION_MINOR * 100 + $ROCM_VERSION_PATCH))
 
+# Install custom MIOpen + COMgr for ROCm >= 4.0.1
+if [[ $ROCM_INT -lt 40001 ]]; then
+    echo "ROCm version < 4.0.1; will not install custom MIOpen"
+    exit 0
+fi
+
 # Function to retry functions that sometimes timeout or have flaky failures
 retry () {
     $*  || (sleep 1 && $*) || (sleep 2 && $*) || (sleep 4 && $*) || (sleep 8 && $*)
@@ -60,27 +66,55 @@ else
     ROCM_INSTALL_PATH="/opt/rocm-${ROCM_VERSION}"
 fi
 
+# MIOPEN_USE_HIP_KERNELS is a Workaround for COMgr issues
 MIOPEN_CMAKE_COMMON_FLAGS="
 -DMIOPEN_USE_COMGR=ON
 -DMIOPEN_BUILD_DRIVER=OFF
 "
-if [[ $ROCM_INT -ge 60200 ]] && [[ $ROCM_INT -lt 60204 ]]; then
-    MIOPEN_BRANCH="release/rocm-rel-6.2-staging"
-else
-    echo "ROCm ${ROCM_VERSION} does not need any patches, do not build from source"
+# Pull MIOpen repo and set DMIOPEN_EMBED_DB based on ROCm version
+if [[ $ROCM_INT -ge 60300 ]]; then
+    echo "ROCm 6.3+ MIOpen does not need any patches, do not build from source"
     exit 0
+elif [[ $ROCM_INT -ge 60200 ]] && [[ $ROCM_INT -lt 60300 ]]; then
+    MIOPEN_BRANCH="release/rocm-rel-6.2-staging"
+elif [[ $ROCM_INT -ge 60100 ]] && [[ $ROCM_INT -lt 60200 ]]; then
+    echo "ROCm 6.1 MIOpen does not need any patches, do not build from source"
+    exit 0
+elif [[ $ROCM_INT -ge 60000 ]] && [[ $ROCM_INT -lt 60100 ]]; then
+    echo "ROCm 6.0 MIOpen does not need any patches, do not build from source"
+    exit 0
+elif [[ $ROCM_INT -ge 50700 ]] && [[ $ROCM_INT -lt 60000 ]]; then
+    echo "ROCm 5.7 MIOpen does not need any patches, do not build from source"
+    exit 0
+elif [[ $ROCM_INT -ge 50600 ]] && [[ $ROCM_INT -lt 50700 ]]; then
+    MIOPEN_BRANCH="release/rocm-rel-5.6-staging"
+elif [[ $ROCM_INT -ge 50500 ]] && [[ $ROCM_INT -lt 50600 ]]; then
+    MIOPEN_BRANCH="release/rocm-rel-5.5-gfx11"
+elif [[ $ROCM_INT -ge 50400 ]] && [[ $ROCM_INT -lt 50500 ]]; then
+    MIOPEN_CMAKE_DB_FLAGS="-DMIOPEN_EMBED_DB=gfx900_56;gfx906_60;gfx90878;gfx90a6e;gfx1030_36 -DMIOPEN_USE_MLIR=Off"
+    MIOPEN_BRANCH="release/rocm-rel-5.4-staging"
+elif [[ $ROCM_INT -ge 50300 ]] && [[ $ROCM_INT -lt 50400 ]]; then
+    MIOPEN_CMAKE_DB_FLAGS="-DMIOPEN_EMBED_DB=gfx900_56;gfx906_60;gfx90878;gfx90a6e;gfx1030_36 -DMIOPEN_USE_MLIR=Off"
+    MIOPEN_BRANCH="release/rocm-rel-5.3-staging"
+elif [[ $ROCM_INT -ge 50200 ]] && [[ $ROCM_INT -lt 50300 ]]; then
+    MIOPEN_CMAKE_DB_FLAGS="-DMIOPEN_EMBED_DB=gfx900_56;gfx906_60;gfx90878;gfx90a6e;gfx1030_36 -DMIOPEN_USE_MLIR=Off"
+    MIOPEN_BRANCH="release/rocm-rel-5.2-staging"
+elif [[ $ROCM_INT -ge 50100 ]] && [[ $ROCM_INT -lt 50200 ]]; then
+    MIOPEN_CMAKE_DB_FLAGS="-DMIOPEN_EMBED_DB=gfx900_56;gfx906_60;gfx90878;gfx90a6e;gfx1030_36"
+    MIOPEN_BRANCH="release/rocm-rel-5.1-staging"
+elif [[ $ROCM_INT -ge 50000 ]] && [[ $ROCM_INT -lt 50100 ]]; then
+    MIOPEN_CMAKE_DB_FLAGS="-DMIOPEN_EMBED_DB=gfx900_56;gfx906_60;gfx90878;gfx90a6e;gfx1030_36"
+    MIOPEN_BRANCH="release/rocm-rel-5.0-staging"
+else
+    echo "Unhandled ROCM_VERSION ${ROCM_VERSION}"
+    exit 1
 fi
 
 
 if [[ ${IS_UBUNTU} == 1 ]]; then
   apt-get remove -y miopen-hip
 else
-  # Workaround since almalinux manylinux image already has this and cget doesn't like that
-  rm -rf /usr/local/lib/pkgconfig/sqlite3.pc
-
-  # Versioned package name needs regex match
-  # Use --noautoremove to prevent other rocm packages from being uninstalled
-  yum remove -y miopen-hip* --noautoremove
+  yum remove -y miopen-hip
 fi
 
 git clone https://github.com/ROCm/MIOpen -b ${MIOPEN_BRANCH}
@@ -88,7 +122,16 @@ pushd MIOpen
 # remove .git to save disk space since CI runner was running out
 rm -rf .git
 # Don't build CK to save docker build time
-sed -i '/composable_kernel/d' requirements.txt
+if [[ $ROCM_INT -ge 60200 ]]; then
+    sed -i '/composable_kernel/d' requirements.txt
+fi
+# Don't build MLIR to save docker build time
+# since we are disabling MLIR backend for MIOpen anyway
+if [[ $ROCM_INT -ge 50400 ]] && [[ $ROCM_INT -lt 50500 ]]; then
+    sed -i '/rocMLIR/d' requirements.txt
+elif [[ $ROCM_INT -ge 50200 ]] && [[ $ROCM_INT -lt 50400 ]]; then
+    sed -i '/llvm-project-mlir/d' requirements.txt
+fi
 ## MIOpen minimum requirements
 cmake -P install_deps.cmake --minimum
 
@@ -110,7 +153,7 @@ cd build
 PKG_CONFIG_PATH=/usr/local/lib/pkgconfig CXX=${ROCM_INSTALL_PATH}/llvm/bin/clang++ cmake .. \
     ${MIOPEN_CMAKE_COMMON_FLAGS} \
     ${MIOPEN_CMAKE_DB_FLAGS} \
-    -DCMAKE_PREFIX_PATH="${ROCM_INSTALL_PATH}"
+    -DCMAKE_PREFIX_PATH="${ROCM_INSTALL_PATH}/hip;${ROCM_INSTALL_PATH}"
 make MIOpen -j $(nproc)
 
 # Build MIOpen package

.ci/docker/common/install_onnx.sh

@@ -32,7 +32,7 @@ pip_install coloredlogs packaging
 
 pip_install onnxruntime==1.18.1
 pip_install onnx==1.16.2
-pip_install onnxscript==0.1.0.dev20241124 --no-deps
+pip_install onnxscript==0.1.0.dev20241009 --no-deps
 # required by onnxscript
 pip_install ml_dtypes
 

.ci/docker/common/install_openblas.sh

@@ -4,7 +4,7 @@
 set -ex
 
 cd /
-git clone https://github.com/OpenMathLib/OpenBLAS.git -b v0.3.28 --depth 1 --shallow-submodules
+git clone https://github.com/OpenMathLib/OpenBLAS.git -b v0.3.25 --depth 1 --shallow-submodules
 
 
 OPENBLAS_BUILD_FLAGS="

.ci/docker/common/install_rocm.sh

@@ -62,22 +62,6 @@ install_ubuntu() {
         sqlite3 $kdb "PRAGMA journal_mode=off; PRAGMA VACUUM;"
     done
 
-    # ROCm 6.3 had a regression where initializing static code objects had significant overhead
-    if [[ $(ver $ROCM_VERSION) -eq $(ver 6.3) ]]; then
-        # clr build needs CppHeaderParser but can only find it using conda's python
-        /opt/conda/bin/python -m pip install CppHeaderParser
-        git clone https://github.com/ROCm/HIP -b rocm-6.3.x
-        HIP_COMMON_DIR=$(readlink -f HIP)
-        git clone https://github.com/jeffdaily/clr -b release/rocm-rel-6.3-statco-hotfix
-        mkdir -p clr/build
-        pushd clr/build
-        cmake .. -DCLR_BUILD_HIP=ON -DHIP_COMMON_DIR=$HIP_COMMON_DIR
-        make -j
-        cp hipamd/lib/libamdhip64.so.6.3.* /opt/rocm/lib/libamdhip64.so.6.3.*
-        popd
-        rm -rf HIP clr
-    fi
-
     # Cleanup
     apt-get autoclean && apt-get clean
     rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*

.ci/docker/common/install_rocm_drm.sh

@@ -12,7 +12,7 @@ case "$ID" in
     apt-get install -y libpciaccess-dev pkg-config
     apt-get clean
     ;;
-  centos|almalinux)
+  centos)
     yum install -y libpciaccess-devel pkgconfig
     ;;
   *)

.ci/docker/common/install_rocm_magma.sh

@@ -3,18 +3,6 @@
 
 set -ex
 
-# Magma build scripts need `python`
-ln -sf /usr/bin/python3 /usr/bin/python
-
-ID=$(grep -oP '(?<=^ID=).+' /etc/os-release | tr -d '"')
-case "$ID" in
-  almalinux)
-    yum install -y gcc-gfortran
-    ;;
-  *)
-    echo "No preinstalls to build magma..."
-    ;;
-esac
 
 MKLROOT=${MKLROOT:-/opt/conda/envs/py_$ANACONDA_PYTHON_VERSION}
 

.ci/docker/common/install_user.sh

@@ -2,13 +2,6 @@
 
 set -ex
 
-# Since version 24 the system ships with user 'ubuntu' that has id 1000
-# We need a work-around to enable id 1000 usage for this script
-if [[ $UBUNTU_VERSION == 24.04 ]]; then
-    # touch is used to disable harmless error message
-    touch /var/mail/ubuntu && chown ubuntu /var/mail/ubuntu && userdel -r ubuntu
-fi
-
 # Mirror jenkins user in container
 # jenkins user as ec2-user should have the same user-id
 echo "jenkins:x:1000:1000::/var/lib/jenkins:" >> /etc/passwd

.ci/docker/common/install_xpu.sh

@@ -24,10 +24,10 @@ function install_ubuntu() {
         | tee /etc/apt/sources.list.d/intel-gpu-${VERSION_CODENAME}.list
     # To add the online network network package repository for the Intel Support Packages
     wget -O- https://apt.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB \
-        | gpg --dearmor > /usr/share/keyrings/oneapi-archive-keyring.gpg.gpg
-    echo "deb [signed-by=/usr/share/keyrings/oneapi-archive-keyring.gpg.gpg] \
-        https://apt.repos.intel.com/${XPU_REPO_NAME} all main" \
-        | tee /etc/apt/sources.list.d/oneAPI.list
+        | gpg --dearmor > /usr/share/keyrings/intel-for-pytorch-gpu-dev-keyring.gpg
+    echo "deb [signed-by=/usr/share/keyrings/intel-for-pytorch-gpu-dev-keyring.gpg] \
+        https://apt.repos.intel.com/intel-for-pytorch-gpu-dev all main" \
+        | tee /etc/apt/sources.list.d/intel-for-pytorch-gpu-dev.list
 
     # Update the packages list and repository index
     apt-get update
@@ -47,7 +47,11 @@ function install_ubuntu() {
     # Development Packages
     apt-get install -y libigc-dev intel-igc-cm libigdfcl-dev libigfxcmrt-dev level-zero-dev
     # Install Intel Support Packages
-    apt-get install -y ${XPU_PACKAGES}
+    if [ -n "$XPU_VERSION" ]; then
+        apt-get install -y intel-for-pytorch-gpu-dev-${XPU_VERSION} intel-pti-dev-0.9
+    else
+        apt-get install -y intel-for-pytorch-gpu-dev-0.5 intel-pti-dev-0.9
+    fi
 
     # Cleanup
     apt-get autoclean && apt-get clean
@@ -57,13 +61,13 @@ function install_ubuntu() {
 function install_rhel() {
     . /etc/os-release
     if [[ "${ID}" == "rhel" ]]; then
-        if [[ ! " 8.8 8.9 9.0 9.2 9.3 " =~ " ${VERSION_ID} " ]]; then
+        if [[ ! " 8.6 8.8 8.9 9.0 9.2 9.3 " =~ " ${VERSION_ID} " ]]; then
             echo "RHEL version ${VERSION_ID} not supported"
             exit
         fi
     elif [[ "${ID}" == "almalinux" ]]; then
         # Workaround for almalinux8 which used by quay.io/pypa/manylinux_2_28_x86_64
-        VERSION_ID="8.8"
+        VERSION_ID="8.6"
     fi
 
     dnf install -y 'dnf-command(config-manager)'
@@ -71,18 +75,16 @@ function install_rhel() {
     dnf config-manager --add-repo \
         https://repositories.intel.com/gpu/rhel/${VERSION_ID}${XPU_DRIVER_VERSION}/unified/intel-gpu-${VERSION_ID}.repo
     # To add the online network network package repository for the Intel Support Packages
-    tee > /etc/yum.repos.d/oneAPI.repo << EOF
-[oneAPI]
+    tee > /etc/yum.repos.d/intel-for-pytorch-gpu-dev.repo << EOF
+[intel-for-pytorch-gpu-dev]
 name=Intel for Pytorch GPU dev repository
-baseurl=https://yum.repos.intel.com/${XPU_REPO_NAME}
+baseurl=https://yum.repos.intel.com/intel-for-pytorch-gpu-dev
 enabled=1
 gpgcheck=1
 repo_gpgcheck=1
 gpgkey=https://yum.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB
 EOF
 
-    # Install Intel Support Packages
-    yum install -y ${XPU_PACKAGES}
     # The xpu-smi packages
     dnf install -y xpu-smi
     # Compute and Media Runtimes
@@ -97,6 +99,8 @@ EOF
     dnf install -y --refresh \
         intel-igc-opencl-devel level-zero-devel intel-gsc-devel libmetee-devel \
         level-zero-devel
+    # Install Intel Support Packages
+    yum install -y intel-for-pytorch-gpu-dev-0.5 intel-pti-dev-0.9
 
     # Cleanup
     dnf clean all
@@ -118,7 +122,7 @@ function install_sles() {
         https://repositories.intel.com/gpu/sles/${VERSION_SP}${XPU_DRIVER_VERSION}/unified/intel-gpu-${VERSION_SP}.repo
     rpm --import https://repositories.intel.com/gpu/intel-graphics.key
     # To add the online network network package repository for the Intel Support Packages
-    zypper addrepo https://yum.repos.intel.com/${XPU_REPO_NAME} oneAPI
+    zypper addrepo https://yum.repos.intel.com/intel-for-pytorch-gpu-dev intel-for-pytorch-gpu-dev
     rpm --import https://yum.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB
 
     # The xpu-smi packages
@@ -130,7 +134,7 @@ function install_sles() {
     zypper install -y libigdfcl-devel intel-igc-cm libigfxcmrt-devel level-zero-devel
 
     # Install Intel Support Packages
-    zypper install -y ${XPU_PACKAGES}
+    zypper install -y intel-for-pytorch-gpu-dev-0.5 intel-pti-dev-0.9
 
 }
 
@@ -141,13 +145,6 @@ if [[ "${XPU_DRIVER_TYPE,,}" == "rolling" ]]; then
     XPU_DRIVER_VERSION=""
 fi
 
-XPU_REPO_NAME="intel-for-pytorch-gpu-dev"
-XPU_PACKAGES="intel-for-pytorch-gpu-dev-0.5 intel-pti-dev-0.9"
-if [[ "$XPU_VERSION" == "2025.0" ]]; then
-    XPU_REPO_NAME="oneapi"
-    XPU_PACKAGES="intel-deep-learning-essentials-2025.0"
-fi
-
 # The installation depends on the base OS
 ID=$(grep -oP '(?<=^ID=).+' /etc/os-release | tr -d '"')
 case "$ID" in

.ci/docker/conda/Dockerfile

@@ -1,39 +1,47 @@
-ARG CUDA_VERSION=12.4
+ARG CUDA_VERSION=10.2
 ARG BASE_TARGET=cuda${CUDA_VERSION}
-FROM amd64/almalinux:8 as base
+FROM centos:7 as base
 
 ENV LC_ALL en_US.UTF-8
 ENV LANG en_US.UTF-8
 ENV LANGUAGE en_US.UTF-8
 
-ARG DEVTOOLSET_VERSION=11
-
-ENV LC_ALL en_US.UTF-8
-ENV LANG en_US.UTF-8
-ENV LANGUAGE en_US.UTF-8
-
-RUN yum -y update
-RUN yum -y install epel-release
-RUN yum install -y sudo wget curl perl util-linux xz bzip2 git patch which perl zlib-devel openssl-devel yum-utils autoconf automake make gcc-toolset-${DEVTOOLSET_VERSION}-toolchain
+ARG DEVTOOLSET_VERSION=9
+RUN sed -i s/mirror.centos.org/vault.centos.org/g /etc/yum.repos.d/*.repo
+RUN sed -i s/^#.*baseurl=http/baseurl=http/g /etc/yum.repos.d/*.repo
+RUN sed -i s/^mirrorlist=http/#mirrorlist=http/g /etc/yum.repos.d/*.repo
+RUN yum update -y
+RUN yum install -y wget curl perl util-linux xz bzip2 git patch which unzip
 # Just add everything as a safe.directory for git since these will be used in multiple places with git
 RUN git config --global --add safe.directory '*'
-ENV PATH=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/bin:$PATH
+RUN yum install -y yum-utils centos-release-scl
+RUN yum-config-manager --enable rhel-server-rhscl-7-rpms
+RUN sed -i s/mirror.centos.org/vault.centos.org/g /etc/yum.repos.d/*.repo
+RUN sed -i s/^#.*baseurl=http/baseurl=http/g /etc/yum.repos.d/*.repo
+RUN sed -i s/^mirrorlist=http/#mirrorlist=http/g /etc/yum.repos.d/*.repo
+RUN yum install -y devtoolset-${DEVTOOLSET_VERSION}-gcc devtoolset-${DEVTOOLSET_VERSION}-gcc-c++ devtoolset-${DEVTOOLSET_VERSION}-gcc-gfortran devtoolset-${DEVTOOLSET_VERSION}-binutils
+# EPEL for cmake
+RUN yum --enablerepo=extras install -y epel-release
 
-# cmake-3.18.4 from pip
-RUN yum install -y python3-pip && \
-    python3 -mpip install cmake==3.18.4 && \
-    ln -s /usr/local/bin/cmake /usr/bin/cmake3
+# cmake
+RUN yum install -y cmake3 && \
+    ln -s /usr/bin/cmake3 /usr/bin/cmake
+ENV PATH=/opt/rh/devtoolset-${DEVTOOLSET_VERSION}/root/usr/bin:$PATH
+ENV LD_LIBRARY_PATH=/opt/rh/devtoolset-${DEVTOOLSET_VERSION}/root/usr/lib64:/opt/rh/devtoolset-${DEVTOOLSET_VERSION}/root/usr/lib:$LD_LIBRARY_PATH
+
+RUN yum install -y autoconf aclocal automake make sudo
 RUN rm -rf /usr/local/cuda-*
 
-FROM base as openssl
-ADD ./common/install_openssl.sh install_openssl.sh
-RUN bash ./install_openssl.sh && rm install_openssl.sh
-
 FROM base as patchelf
 # Install patchelf
 ADD ./common/install_patchelf.sh install_patchelf.sh
 RUN bash ./install_patchelf.sh && rm install_patchelf.sh && cp $(which patchelf) /patchelf
 
+FROM base as openssl
+# Install openssl
+ADD ./common/install_openssl.sh install_openssl.sh
+RUN bash ./install_openssl.sh && rm install_openssl.sh
+
 FROM base as conda
 # Install Anaconda
 ADD ./common/install_conda_docker.sh install_conda.sh
@@ -41,7 +49,7 @@ RUN bash ./install_conda.sh && rm install_conda.sh
 
 # Install CUDA
 FROM base as cuda
-ARG CUDA_VERSION=12.4
+ARG CUDA_VERSION=10.2
 RUN rm -rf /usr/local/cuda-*
 ADD ./common/install_cuda.sh install_cuda.sh
 ENV CUDA_HOME=/usr/local/cuda-${CUDA_VERSION}
@@ -88,8 +96,7 @@ COPY ./common/install_jni.sh install_jni.sh
 COPY ./java/jni.h jni.h
 RUN bash ./install_jni.sh && rm install_jni.sh
 
-ENV PATH /opt/conda/bin:$PATH
-ENV LD_LIBRARY_PATH=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib64:/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib:$LD_LIBRARY_PATH
+ENV  PATH /opt/conda/bin:$PATH
 COPY --from=mnist  /usr/local/mnist /usr/local/mnist
 RUN rm -rf /usr/local/cuda
 RUN chmod o+rw /usr/local

.ci/docker/conda/build.sh

@@ -48,10 +48,10 @@ esac
     --progress plain \
     --build-arg "BASE_TARGET=${BASE_TARGET}" \
     --build-arg "CUDA_VERSION=${CUDA_VERSION}" \
-    --build-arg "DEVTOOLSET_VERSION=11" \
+    --build-arg "DEVTOOLSET_VERSION=9" \
     -t ${DOCKER_IMAGE_NAME} \
     $@ \
-    -f "${TOPDIR}/.ci/docker/almalinux/Dockerfile" \
+    -f "${TOPDIR}/.ci/docker/conda/Dockerfile" \
     ${TOPDIR}/.ci/docker/
 )
 

.ci/docker/libtorch/Dockerfile

@@ -92,6 +92,13 @@ RUN apt-get update -y && \
 RUN bash ./install_rocm_drm.sh && rm install_rocm_drm.sh
 RUN bash ./install_rocm_magma.sh && rm install_rocm_magma.sh
 
+# Install AOTriton
+COPY ./common/common_utils.sh common_utils.sh
+COPY ./aotriton_version.txt aotriton_version.txt
+COPY ./common/install_aotriton.sh install_aotriton.sh
+RUN bash ./install_aotriton.sh /opt/rocm && rm install_aotriton.sh aotriton_version.txt
+ENV AOTRITON_INSTALLED_PREFIX /opt/rocm/aotriton
+
 FROM ${BASE_TARGET} as final
 COPY --from=openssl            /opt/openssl           /opt/openssl
 # Install patchelf

.ci/docker/libtorch/build.sh

@@ -39,7 +39,17 @@ case ${GPU_ARCH_TYPE} in
         BASE_TARGET=rocm
         DOCKER_TAG=rocm${GPU_ARCH_VERSION}
         GPU_IMAGE=rocm/dev-ubuntu-20.04:${GPU_ARCH_VERSION}-complete
-        PYTORCH_ROCM_ARCH="gfx900;gfx906;gfx908;gfx90a;gfx1030;gfx1100;gfx1101;gfx942"
+        PYTORCH_ROCM_ARCH="gfx900;gfx906;gfx908;gfx90a;gfx1030;gfx1100"
+        ROCM_REGEX="([0-9]+)\.([0-9]+)[\.]?([0-9]*)"
+        if [[ $GPU_ARCH_VERSION =~ $ROCM_REGEX ]]; then
+            ROCM_VERSION_INT=$((${BASH_REMATCH[1]}*10000 + ${BASH_REMATCH[2]}*100 + ${BASH_REMATCH[3]:-0}))
+        else
+            echo "ERROR: rocm regex failed"
+            exit 1
+        fi
+        if [[ $ROCM_VERSION_INT -ge 60000 ]]; then
+            PYTORCH_ROCM_ARCH+=";gfx942"
+        fi
         DOCKER_GPU_BUILD_ARG="--build-arg PYTORCH_ROCM_ARCH=${PYTORCH_ROCM_ARCH}"
         ;;
     *)

.ci/docker/linter-cuda/Dockerfile

@@ -25,8 +25,7 @@ ENV PATH /opt/conda/envs/py_$ANACONDA_PYTHON_VERSION/bin:/opt/conda/bin:$PATH
 COPY requirements-ci.txt /opt/conda/requirements-ci.txt
 COPY ./common/install_conda.sh install_conda.sh
 COPY ./common/common_utils.sh common_utils.sh
-COPY ./common/install_magma_conda.sh install_magma_conda.sh
-RUN bash ./install_conda.sh && rm install_conda.sh install_magma_conda.sh common_utils.sh /opt/conda/requirements-ci.txt
+RUN bash ./install_conda.sh && rm install_conda.sh common_utils.sh /opt/conda/requirements-ci.txt
 
 # Install cuda and cudnn
 ARG CUDA_VERSION

.ci/docker/manywheel/Dockerfile

@@ -144,10 +144,6 @@ COPY --from=libpng             /usr/local/lib/pkgconfig              /usr/local/
 FROM common as cpu_final
 ARG BASE_CUDA_VERSION=10.1
 ARG DEVTOOLSET_VERSION=9
-# Install Anaconda
-ADD ./common/install_conda_docker.sh install_conda.sh
-RUN bash ./install_conda.sh && rm install_conda.sh
-ENV PATH /opt/conda/bin:$PATH
 RUN sed -i s/mirror.centos.org/vault.centos.org/g /etc/yum.repos.d/*.repo
 RUN sed -i s/^#.*baseurl=http/baseurl=http/g /etc/yum.repos.d/*.repo
 RUN sed -i s/^mirrorlist=http/#mirrorlist=http/g /etc/yum.repos.d/*.repo
@@ -198,3 +194,10 @@ ADD ./common/install_rocm_magma.sh install_rocm_magma.sh
 RUN bash ./install_rocm_magma.sh && rm install_rocm_magma.sh
 ADD ./common/install_miopen.sh install_miopen.sh
 RUN bash ./install_miopen.sh ${ROCM_VERSION} && rm install_miopen.sh
+
+# Install AOTriton
+COPY ./common/common_utils.sh common_utils.sh
+COPY ./aotriton_version.txt aotriton_version.txt
+COPY ./common/install_aotriton.sh install_aotriton.sh
+RUN bash ./install_aotriton.sh /opt/rocm && rm install_aotriton.sh aotriton_version.txt
+ENV AOTRITON_INSTALLED_PREFIX /opt/rocm/aotriton

.ci/docker/manywheel/Dockerfile_2_28

@@ -1,4 +1,5 @@
 # syntax = docker/dockerfile:experimental
+ARG ROCM_VERSION=3.7
 ARG BASE_CUDA_VERSION=11.8
 ARG GPU_IMAGE=amd64/almalinux:8
 FROM quay.io/pypa/manylinux_2_28_x86_64 as base
@@ -116,49 +117,30 @@ COPY --from=jni                /usr/local/include/jni.h              /usr/local/
 FROM common as cpu_final
 ARG BASE_CUDA_VERSION=11.8
 ARG DEVTOOLSET_VERSION=11
-# Install Anaconda
-ADD ./common/install_conda_docker.sh install_conda.sh
-RUN bash ./install_conda.sh && rm install_conda.sh
-ENV PATH /opt/conda/bin:$PATH
 # Ensure the expected devtoolset is used
 ENV PATH=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/bin:$PATH
 ENV LD_LIBRARY_PATH=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib64:/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib:$LD_LIBRARY_PATH
-# Install setuptools and wheel for python 3.12/3.13
-RUN for cpython_version in "cp312-cp312" "cp313-cp313" "cp313-cp313t"; do \
-    /opt/python/${cpython_version}/bin/python -m pip install setuptools wheel; \
-    done;
 
-
-# cmake-3.18.4 from pip; force in case cmake3 already exists
+# cmake-3.18.4 from pip
 RUN yum install -y python3-pip && \
     python3 -mpip install cmake==3.18.4 && \
-    ln -sf /usr/local/bin/cmake /usr/bin/cmake3
+    ln -s /usr/local/bin/cmake /usr/bin/cmake3
 
 FROM cpu_final as cuda_final
 RUN rm -rf /usr/local/cuda-${BASE_CUDA_VERSION}
 COPY --from=cuda     /usr/local/cuda-${BASE_CUDA_VERSION}  /usr/local/cuda-${BASE_CUDA_VERSION}
 COPY --from=magma    /usr/local/cuda-${BASE_CUDA_VERSION}  /usr/local/cuda-${BASE_CUDA_VERSION}
-RUN ln -sf /usr/local/cuda-${BASE_CUDA_VERSION} /usr/local/cuda
-ENV PATH=/usr/local/cuda/bin:$PATH
 
-FROM cpu_final as rocm_final
-ARG ROCM_VERSION=6.0
-ARG PYTORCH_ROCM_ARCH
-ENV PYTORCH_ROCM_ARCH ${PYTORCH_ROCM_ARCH}
-ARG DEVTOOLSET_VERSION=11
-ENV LDFLAGS="-Wl,-rpath=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib64 -Wl,-rpath=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib"
-# Somewhere in ROCm stack, we still use non-existing /opt/rocm/hip path,
-# below workaround helps avoid error
-ENV ROCM_PATH /opt/rocm
-# cmake-3.28.4 from pip to get enable_language(HIP)
-# and avoid 3.21.0 cmake+ninja issues with ninja inserting "-Wl,--no-as-needed" in LINK_FLAGS for static linker
-RUN python3 -m pip install --upgrade pip && \
-    python3 -mpip install cmake==3.28.4
-ADD ./common/install_rocm_drm.sh install_rocm_drm.sh
-RUN bash ./install_rocm_drm.sh && rm install_rocm_drm.sh
-ENV MKLROOT /opt/intel
-ADD ./common/install_rocm_magma.sh install_rocm_magma.sh
-RUN bash ./install_rocm_magma.sh && rm install_rocm_magma.sh
+FROM common as rocm_final
+ARG ROCM_VERSION=3.7
+# Install ROCm
+ADD ./common/install_rocm.sh install_rocm.sh
+RUN bash ./install_rocm.sh ${ROCM_VERSION} && rm install_rocm.sh
+# cmake is already installed inside the rocm base image, but both 2 and 3 exist
+# cmake3 is needed for the later MIOpen custom build, so that step is last.
+RUN yum install -y cmake3 && \
+    rm -f /usr/bin/cmake && \
+    ln -s /usr/bin/cmake3 /usr/bin/cmake
 ADD ./common/install_miopen.sh install_miopen.sh
 RUN bash ./install_miopen.sh ${ROCM_VERSION} && rm install_miopen.sh
 
@@ -168,7 +150,8 @@ ENV XPU_DRIVER_TYPE ROLLING
 # cmake-3.28.4 from pip
 RUN python3 -m pip install --upgrade pip && \
     python3 -mpip install cmake==3.28.4
+# Install setuptools and wheel for python 3.13
+RUN /opt/python/cp313-cp313/bin/python -m pip install setuptools wheel
 ADD ./common/install_xpu.sh install_xpu.sh
-ENV XPU_VERSION 2025.0
 RUN bash ./install_xpu.sh && rm install_xpu.sh
 RUN pushd /opt/_internal && tar -xJf static-libs-for-embedding-only.tar.xz && popd

.ci/docker/manywheel/Dockerfile_2_28_aarch64

@@ -48,11 +48,6 @@ ENV LD_LIBRARY_PATH=/opt/rh/gcc-toolset-${GCCTOOLSET_VERSION}/root/usr/lib64:/op
 # For more details see https://github.com/pytorch/pytorch/issues/78659#issuecomment-1144107327
 RUN git config --global --add safe.directory "*"
 
-FROM base as openblas
-# Install openblas
-ADD ./common/install_openblas.sh install_openblas.sh
-RUN bash ./install_openblas.sh && rm install_openblas.sh
-
 FROM base as final
 
 # remove unncessary python versions
@@ -60,5 +55,3 @@ RUN rm -rf /opt/python/cp26-cp26m /opt/_internal/cpython-2.6.9-ucs2
 RUN rm -rf /opt/python/cp26-cp26mu /opt/_internal/cpython-2.6.9-ucs4
 RUN rm -rf /opt/python/cp33-cp33m /opt/_internal/cpython-3.3.6
 RUN rm -rf /opt/python/cp34-cp34m /opt/_internal/cpython-3.4.6
-COPY --from=openblas     /opt/OpenBLAS/  /opt/OpenBLAS/
-ENV LD_LIBRARY_PATH=/opt/OpenBLAS/lib:$LD_LIBRARY_PATH

.ci/docker/manywheel/Dockerfile_aarch64

@@ -61,7 +61,7 @@ RUN git config --global --add safe.directory "*"
 # NOTE: Need a better way to get this library as Ubuntu's package can be removed by the vender, or changed
 ###############################################################################
 RUN cd ~/ \
-  && curl -L -o ~/libgfortran-10-dev.deb http://ports.ubuntu.com/ubuntu-ports/pool/universe/g/gcc-10/libgfortran-10-dev_10.5.0-4ubuntu2_arm64.deb \
+  && curl -L -o ~/libgfortran-10-dev.deb http://ports.ubuntu.com/ubuntu-ports/pool/universe/g/gcc-10/libgfortran-10-dev_10.5.0-1ubuntu1_arm64.deb \
   && ar x ~/libgfortran-10-dev.deb \
   && tar --use-compress-program=unzstd -xvf data.tar.zst -C ~/ \
   && cp -f ~/usr/lib/gcc/aarch64-linux-gnu/10/libgfortran.a /opt/rh/devtoolset-10/root/usr/lib/gcc/aarch64-redhat-linux/10/

.ci/docker/manywheel/Dockerfile_s390x

@@ -1,20 +1,17 @@
-FROM quay.io/pypa/manylinux_2_28_s390x as base
+FROM --platform=linux/s390x docker.io/ubuntu:24.04 as base
 
 # Language variables
 ENV LC_ALL=C.UTF-8
 ENV LANG=C.UTF-8
 ENV LANGUAGE=C.UTF-8
 
-ARG DEVTOOLSET_VERSION=13
 # Installed needed OS packages. This is to support all
 # the binary builds (torch, vision, audio, text, data)
-RUN yum -y install epel-release
-RUN yum -y update
-RUN yum install -y \
-  sudo \
+RUN apt update ; apt upgrade -y
+RUN apt install -y \
+  build-essential \
   autoconf \
   automake \
-  bison \
   bzip2 \
   curl \
   diffutils \
@@ -27,40 +24,19 @@ RUN yum install -y \
   util-linux \
   wget \
   which \
-  xz \
-  yasm \
+  xz-utils \
   less \
   zstd \
-  libgomp \
-  gcc-toolset-${DEVTOOLSET_VERSION}-gcc \
-  gcc-toolset-${DEVTOOLSET_VERSION}-gcc-c++ \
-  gcc-toolset-${DEVTOOLSET_VERSION}-binutils \
-  gcc-toolset-${DEVTOOLSET_VERSION}-gcc-gfortran \
   cmake \
-  rust \
-  cargo \
-  llvm-devel \
-  libzstd-devel \
-  python3.12-devel \
-  python3.12-setuptools \
-  python3.12-pip \
-  python3-virtualenv \
-  python3.12-pyyaml \
-  python3.12-numpy \
-  python3.12-wheel \
-  python3.12-cryptography \
-  blas-devel \
-  openblas-devel \
-  lapack-devel \
-  atlas-devel \
-  libjpeg-devel \
-  libxslt-devel \
-  libxml2-devel \
-  openssl-devel \
-  valgrind
-
-ENV PATH=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/bin:$PATH
-ENV LD_LIBRARY_PATH=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib64:/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib:$LD_LIBRARY_PATH
+  python3 \
+  python3-dev \
+  python3-setuptools \
+  python3-yaml \
+  python3-typing-extensions \
+  libblas-dev \
+  libopenblas-dev \
+  liblapack-dev \
+  libatlas-base-dev
 
 # git236+ would refuse to run git commands in repos owned by other users
 # Which causes version check to fail, as pytorch repo is bind-mounted into the image
@@ -68,8 +44,14 @@ ENV LD_LIBRARY_PATH=/opt/rh/gcc-toolset-${DEVTOOLSET_VERSION}/root/usr/lib64:/op
 # For more details see https://github.com/pytorch/pytorch/issues/78659#issuecomment-1144107327
 RUN git config --global --add safe.directory "*"
 
-# installed python doesn't have development parts. Rebuild it from scratch
-RUN /bin/rm -rf /opt/_internal /opt/python /usr/local/*/*
+FROM base as openssl
+# Install openssl (this must precede `build python` step)
+# (In order to have a proper SSL module, Python is compiled
+# against a recent openssl [see env vars above], which is linked
+# statically. We delete openssl afterwards.)
+ADD ./common/install_openssl.sh install_openssl.sh
+RUN bash ./install_openssl.sh && rm install_openssl.sh
+ENV SSL_CERT_FILE=/opt/_internal/certs.pem
 
 # EPEL for cmake
 FROM base as patchelf
@@ -82,43 +64,10 @@ FROM patchelf as python
 # build python
 COPY manywheel/build_scripts /build_scripts
 ADD ./common/install_cpython.sh /build_scripts/install_cpython.sh
-ENV SSL_CERT_FILE=
 RUN bash build_scripts/build.sh && rm -r build_scripts
 
-FROM base as final
+FROM openssl as final
 COPY --from=python             /opt/python                           /opt/python
 COPY --from=python             /opt/_internal                        /opt/_internal
-COPY --from=python             /opt/python/cp39-cp39/bin/auditwheel  /usr/local/bin/auditwheel
+COPY --from=python             /opt/python/cp39-cp39/bin/auditwheel /usr/local/bin/auditwheel
 COPY --from=patchelf           /usr/local/bin/patchelf               /usr/local/bin/patchelf
-
-RUN alternatives --set python /usr/bin/python3.12
-RUN alternatives --set python3 /usr/bin/python3.12
-
-RUN pip-3.12 install typing_extensions
-
-ENTRYPOINT []
-CMD ["/bin/bash"]
-
-# install test dependencies:
-# - grpcio requires system openssl, bundled crypto fails to build
-# - ml_dtypes 0.4.0 requires some fixes provided in later commits to build
-RUN dnf install -y \
-  protobuf-devel \
-  protobuf-c-devel \
-  protobuf-lite-devel \
-  wget \
-  patch
-
-RUN env GRPC_PYTHON_BUILD_SYSTEM_OPENSSL=True pip3 install grpcio==1.65.4
-RUN cd ~ && \
-  git clone https://github.com/jax-ml/ml_dtypes && \
-  cd ml_dtypes && \
-  git checkout v0.4.0 && \
-  git submodule update --init --recursive && \
-  wget https://github.com/jax-ml/ml_dtypes/commit/b969f76914d6b30676721bc92bf0f6021a0d1321.patch && \
-  wget https://github.com/jax-ml/ml_dtypes/commit/d4e6d035ecda073eab8bcf60f4eef572ee7087e6.patch && \
-  patch -p1 < b969f76914d6b30676721bc92bf0f6021a0d1321.patch && \
-  patch -p1 < d4e6d035ecda073eab8bcf60f4eef572ee7087e6.patch && \
-  python3 setup.py bdist_wheel && \
-  pip3 install dist/*.whl && \
-  rm -rf ml_dtypes

.ci/docker/manywheel/build.sh

@@ -61,7 +61,7 @@ case ${GPU_ARCH_TYPE} in
     cpu-s390x)
         TARGET=final
         DOCKER_TAG=cpu-s390x
-        GPU_IMAGE=s390x/almalinux:8
+        GPU_IMAGE=redhat/ubi9
         DOCKER_GPU_BUILD_ARG=""
         MANY_LINUX_VERSION="s390x"
         ;;
@@ -87,18 +87,22 @@ case ${GPU_ARCH_TYPE} in
         MANY_LINUX_VERSION="aarch64"
         DOCKERFILE_SUFFIX="_cuda_aarch64"
         ;;
-    rocm|rocm-manylinux_2_28)
+    rocm)
         TARGET=rocm_final
         DOCKER_TAG=rocm${GPU_ARCH_VERSION}
         GPU_IMAGE=rocm/dev-centos-7:${GPU_ARCH_VERSION}-complete
-        DEVTOOLSET_VERSION="9"
-        if [ ${GPU_ARCH_TYPE} == "rocm-manylinux_2_28" ]; then
-            MANY_LINUX_VERSION="2_28"
-            DEVTOOLSET_VERSION="11"
-            GPU_IMAGE=rocm/dev-almalinux-8:${GPU_ARCH_VERSION}-complete
+        PYTORCH_ROCM_ARCH="gfx900;gfx906;gfx908;gfx90a;gfx1030;gfx1100"
+        ROCM_REGEX="([0-9]+)\.([0-9]+)[\.]?([0-9]*)"
+        if [[ $GPU_ARCH_VERSION =~ $ROCM_REGEX ]]; then
+            ROCM_VERSION_INT=$((${BASH_REMATCH[1]}*10000 + ${BASH_REMATCH[2]}*100 + ${BASH_REMATCH[3]:-0}))
+        else
+            echo "ERROR: rocm regex failed"
+            exit 1
         fi
-        PYTORCH_ROCM_ARCH="gfx900;gfx906;gfx908;gfx90a;gfx942;gfx1030;gfx1100;gfx1101"
-        DOCKER_GPU_BUILD_ARG="--build-arg ROCM_VERSION=${GPU_ARCH_VERSION} --build-arg PYTORCH_ROCM_ARCH=${PYTORCH_ROCM_ARCH} --build-arg DEVTOOLSET_VERSION=${DEVTOOLSET_VERSION}"
+        if [[ $ROCM_VERSION_INT -ge 60000 ]]; then
+            PYTORCH_ROCM_ARCH+=";gfx942"
+        fi
+        DOCKER_GPU_BUILD_ARG="--build-arg ROCM_VERSION=${GPU_ARCH_VERSION} --build-arg PYTORCH_ROCM_ARCH=${PYTORCH_ROCM_ARCH} --build-arg DEVTOOLSET_VERSION=9"
         ;;
     xpu)
         TARGET=xpu_final
@@ -121,13 +125,11 @@ fi
 (
     set -x
 
-    if [ "$(uname -m)" != "s390x" ]; then
-        # TODO: Remove LimitNOFILE=1048576 patch once https://github.com/pytorch/test-infra/issues/5712
-        # is resolved. This patch is required in order to fix timing out of Docker build on Amazon Linux 2023.
-        sudo sed -i s/LimitNOFILE=infinity/LimitNOFILE=1048576/ /usr/lib/systemd/system/docker.service
-        sudo systemctl daemon-reload
-        sudo systemctl restart docker
-    fi
+    # TODO: Remove LimitNOFILE=1048576 patch once https://github.com/pytorch/test-infra/issues/5712
+    # is resolved. This patch is required in order to fix timing out of Docker build on Amazon Linux 2023.
+    sudo sed -i s/LimitNOFILE=infinity/LimitNOFILE=1048576/ /usr/lib/systemd/system/docker.service
+    sudo systemctl daemon-reload
+    sudo systemctl restart docker
 
     DOCKER_BUILDKIT=1 docker build  \
         ${DOCKER_GPU_BUILD_ARG} \

.ci/docker/manywheel/build_scripts/build.sh

@@ -16,27 +16,37 @@ CURL_HASH=cf34fe0b07b800f1c01a499a6e8b2af548f6d0e044dca4a29d88a4bee146d131
 AUTOCONF_ROOT=autoconf-2.69
 AUTOCONF_HASH=954bd69b391edc12d6a4a51a2dd1476543da5c6bbf05a95b59dc0dd6fd4c2969
 
-# Dependencies for compiling Python that we want to remove from
-# the final image after compiling Python
-PYTHON_COMPILE_DEPS="zlib-devel bzip2-devel ncurses-devel sqlite-devel readline-devel tk-devel gdbm-devel libpcap-devel xz-devel libffi-devel"
-
-if [ "$(uname -m)" != "s390x" ] ; then
-    PYTHON_COMPILE_DEPS="${PYTHON_COMPILE_DEPS} db4-devel"
-else
-    PYTHON_COMPILE_DEPS="${PYTHON_COMPILE_DEPS} libdb-devel"
-fi
-
-# Libraries that are allowed as part of the manylinux1 profile
-MANYLINUX1_DEPS="glibc-devel libstdc++-devel glib2-devel libX11-devel libXext-devel libXrender-devel  mesa-libGL-devel libICE-devel libSM-devel ncurses-devel"
-
 # Get build utilities
 MY_DIR=$(dirname "${BASH_SOURCE[0]}")
 source $MY_DIR/build_utils.sh
 
-# Development tools and libraries
-yum -y install bzip2 make git patch unzip bison yasm diffutils \
-    automake which file \
-    ${PYTHON_COMPILE_DEPS}
+if [ "$(uname -m)" != "s390x" ] ; then
+    # Dependencies for compiling Python that we want to remove from
+    # the final image after compiling Python
+    PYTHON_COMPILE_DEPS="zlib-devel bzip2-devel ncurses-devel sqlite-devel readline-devel tk-devel gdbm-devel db4-devel libpcap-devel xz-devel libffi-devel"
+
+    # Libraries that are allowed as part of the manylinux1 profile
+    MANYLINUX1_DEPS="glibc-devel libstdc++-devel glib2-devel libX11-devel libXext-devel libXrender-devel  mesa-libGL-devel libICE-devel libSM-devel ncurses-devel"
+
+    # Development tools and libraries
+    yum -y install bzip2 make git patch unzip bison yasm diffutils \
+        automake which file cmake28 \
+        kernel-devel-`uname -r` \
+        ${PYTHON_COMPILE_DEPS}
+else
+    # Dependencies for compiling Python that we want to remove from
+    # the final image after compiling Python
+    PYTHON_COMPILE_DEPS="zlib1g-dev libbz2-dev libncurses-dev libsqlite3-dev libdb-dev libpcap-dev liblzma-dev libffi-dev"
+
+    # Libraries that are allowed as part of the manylinux1 profile
+    MANYLINUX1_DEPS="libglib2.0-dev libX11-dev libncurses-dev"
+
+    # Development tools and libraries
+    apt install -y bzip2 make git patch unzip diffutils \
+        automake which file cmake \
+        linux-headers-virtual \
+        ${PYTHON_COMPILE_DEPS}
+fi
 
 # Install newest autoconf
 build_autoconf $AUTOCONF_ROOT $AUTOCONF_HASH
@@ -82,13 +92,16 @@ ln -s $PY39_BIN/auditwheel /usr/local/bin/auditwheel
 
 # Clean up development headers and other unnecessary stuff for
 # final image
-yum -y erase wireless-tools gtk2 libX11 hicolor-icon-theme \
-    avahi freetype bitstream-vera-fonts \
-    ${PYTHON_COMPILE_DEPS} || true > /dev/null 2>&1
-yum -y install ${MANYLINUX1_DEPS}
-yum -y clean all > /dev/null 2>&1
-yum list installed
-
+if [ "$(uname -m)" != "s390x" ] ; then
+    yum -y erase wireless-tools gtk2 libX11 hicolor-icon-theme \
+        avahi freetype bitstream-vera-fonts \
+        ${PYTHON_COMPILE_DEPS} || true > /dev/null 2>&1
+    yum -y install ${MANYLINUX1_DEPS}
+    yum -y clean all > /dev/null 2>&1
+    yum list installed
+else
+    apt purge -y ${PYTHON_COMPILE_DEPS} || true > /dev/null 2>&1
+fi
 # we don't need libpython*.a, and they're many megabytes
 find /opt/_internal -name '*.a' -print0 | xargs -0 rm -f
 # Strip what we can -- and ignore errors, because this just attempts to strip

.ci/docker/requirements-ci.txt

@@ -30,13 +30,13 @@ dill==0.3.7
 #Pinned versions: 0.3.7
 #test that import: dynamo/test_replay_record.py test_dataloader.py test_datapipe.py test_serialization.py
 
-expecttest==0.3.0
+expecttest==0.2.1
 #Description: method for writing tests where test framework auto populates
 # the expected output based on previous runs
-#Pinned versions: 0.3.0
+#Pinned versions: 0.2.1
 #test that import:
 
-fbscribelogger==0.1.7
+fbscribelogger==0.1.6
 #Description: write to scribe from authenticated jobs on CI
 #Pinned versions: 0.1.6
 #test that import:
@@ -90,7 +90,7 @@ librosa>=0.6.2 ; python_version < "3.11"
 #Pinned versions:
 #test that import:
 
-mypy==1.13.0
+mypy==1.11.2
 # Pin MyPy version because new errors are likely to appear with each release
 #Description: linter
 #Pinned versions: 1.10.0
@@ -128,13 +128,11 @@ numba==0.55.2 ; python_version == "3.10"
 #test_nn.py, test_namedtensor.py, test_linalg.py, test_jit_cuda_fuser.py,
 #test_jit.py, test_indexing.py, test_datapipe.py, test_dataloader.py,
 #test_binary_ufuncs.py
-numpy==1.22.4; python_version == "3.9" or python_version == "3.10"
+numpy==1.21.2; python_version == "3.9"
+numpy==1.22.4; python_version == "3.10"
 numpy==1.26.2; python_version == "3.11" or python_version == "3.12"
 numpy==2.1.2; python_version >= "3.13"
 
-pandas==2.0.3; python_version < "3.13"
-pandas==2.2.3; python_version >= "3.13"
-
 #onnxruntime
 #Description: scoring engine for Open Neural Network Exchange (ONNX) models
 #Pinned versions: 1.9.0
@@ -158,7 +156,7 @@ optree==0.13.0
 #test_pointwise_ops.py, test_dtensor_ops.py, test_torchinductor.py, test_fx.py,
 #test_fake_tensor.py, test_mps.py
 
-pillow==11.0.0
+pillow==10.3.0
 #Description:  Python Imaging Library fork
 #Pinned versions: 10.3.0
 #test that import:
@@ -193,11 +191,6 @@ pytest-rerunfailures>=10.3
 #Pinned versions:
 #test that import:
 
-pytest-subtests==0.13.1
-#Description: plugin for subtest support
-#Pinned versions:
-#test that import:
-
 #pytest-benchmark
 #Description: fixture for benchmarking code
 #Pinned versions: 3.2.3
@@ -245,7 +238,7 @@ scikit-image==0.22.0 ; python_version >= "3.10"
 #test that import:
 
 scipy==1.10.1 ; python_version <= "3.11"
-scipy==1.14.1 ; python_version >= "3.12"
+scipy==1.12.0 ; python_version == "3.12"
 # Pin SciPy because of failing distribution tests (see #60347)
 #Description: scientific python
 #Pinned versions: 1.10.1
@@ -280,21 +273,26 @@ unittest-xml-reporting<=3.2.0,>=2.0.0
 #test that import:
 
 #lintrunner is supported on aarch64-linux only from 0.12.4 version
-lintrunner==0.12.7
+lintrunner==0.12.5
 #Description: all about linters!
-#Pinned versions: 0.12.7
+#Pinned versions: 0.12.5
 #test that import:
 
 redis>=4.0.0
 #Description: redis database
 #test that import: anything that tests OSS caching/mocking (inductor/test_codecache.py, inductor/test_max_autotune.py)
 
+rockset==1.0.3
+#Description: queries Rockset
+#Pinned versions: 1.0.3
+#test that import:
+
 ghstack==0.8.0
 #Description: ghstack tool
 #Pinned versions: 0.8.0
 #test that import:
 
-jinja2==3.1.5
+jinja2==3.1.4
 #Description: jinja2 template engine
 #Pinned versions: 3.1.4
 #test that import:
@@ -304,25 +302,24 @@ pytest-cpp==2.3.0
 #Pinned versions: 2.3.0
 #test that import:
 
-z3-solver==4.12.6.0
+z3-solver==4.12.2.0
 #Description: The Z3 Theorem Prover Project
 #Pinned versions:
 #test that import:
 
-tensorboard==2.13.0 ; python_version < "3.13"
-tensorboard==2.18.0 ; python_version >= "3.13"
+tensorboard==2.13.0
 #Description: Also included in .ci/docker/requirements-docs.txt
 #Pinned versions:
 #test that import: test_tensorboard
 
 pywavelets==1.4.1 ; python_version < "3.12"
-pywavelets==1.7.0 ; python_version >= "3.12"
+pywavelets==1.5.0 ; python_version >= "3.12"
 #Description: This is a requirement of scikit-image, we need to pin
 # it here because 1.5.0 conflicts with numpy 1.21.2 used in CI
 #Pinned versions: 1.4.1
 #test that import:
 
-lxml==5.3.0
+lxml==5.0.0
 #Description: This is a requirement of unittest-xml-reporting
 
 # Python-3.9 binaries

.ci/docker/requirements-docs.txt

@@ -14,8 +14,7 @@ matplotlib==3.5.3
 #Description: This is used to generate PyTorch docs
 #Pinned versions: 3.5.3
 
-tensorboard==2.13.0 ; python_version < "3.13"
-tensorboard==2.18.0 ; python_version >= "3.13"
+tensorboard==2.13.0
 #Description: This is used to generate PyTorch docs
 #Pinned versions: 2.13.0
 

.ci/docker/triton_version.txt

@@ -1 +1 @@
-3.2.0
+3.1.0

.ci/docker/ubuntu-cuda/Dockerfile

@@ -30,8 +30,7 @@ ARG CONDA_CMAKE
 COPY requirements-ci.txt /opt/conda/requirements-ci.txt
 COPY ./common/install_conda.sh install_conda.sh
 COPY ./common/common_utils.sh common_utils.sh
-COPY ./common/install_magma_conda.sh install_magma_conda.sh
-RUN bash ./install_conda.sh && rm install_conda.sh install_magma_conda.sh common_utils.sh /opt/conda/requirements-ci.txt
+RUN bash ./install_conda.sh && rm install_conda.sh common_utils.sh /opt/conda/requirements-ci.txt
 
 # Install gcc
 ARG GCC_VERSION
@@ -81,8 +80,6 @@ RUN bash ./install_openssl.sh
 ENV OPENSSL_DIR /opt/openssl
 
 ARG INDUCTOR_BENCHMARKS
-ARG ANACONDA_PYTHON_VERSION
-ENV ANACONDA_PYTHON_VERSION=$ANACONDA_PYTHON_VERSION
 COPY ./common/install_inductor_benchmark_deps.sh install_inductor_benchmark_deps.sh
 COPY ./common/common_utils.sh common_utils.sh
 COPY ci_commit_pins/huggingface.txt huggingface.txt

.ci/docker/ubuntu-rocm/Dockerfile

@@ -107,11 +107,12 @@ COPY triton_version.txt triton_version.txt
 RUN if [ -n "${TRITON}" ]; then bash ./install_triton.sh; fi
 RUN rm install_triton.sh common_utils.sh triton.txt triton_version.txt
 
-# This is needed by sccache
-COPY ./common/install_openssl.sh install_openssl.sh
-ENV OPENSSL_ROOT_DIR /opt/openssl
-RUN bash ./install_openssl.sh
-ENV OPENSSL_DIR /opt/openssl
+# Install AOTriton
+COPY ./aotriton_version.txt aotriton_version.txt
+COPY ./common/common_utils.sh common_utils.sh
+COPY ./common/install_aotriton.sh install_aotriton.sh
+RUN ["/bin/bash", "-c", "./install_aotriton.sh /opt/rocm && rm -rf install_aotriton.sh aotriton_version.txt common_utils.sh"]
+ENV AOTRITON_INSTALLED_PREFIX /opt/rocm/aotriton
 
 # Install ccache/sccache (do this last, so we get priority in PATH)
 COPY ./common/install_cache.sh install_cache.sh

.ci/docker/ubuntu/Dockerfile

@@ -36,8 +36,7 @@ ENV DOCS=$DOCS
 COPY requirements-ci.txt requirements-docs.txt /opt/conda/
 COPY ./common/install_conda.sh install_conda.sh
 COPY ./common/common_utils.sh common_utils.sh
-COPY ./common/install_magma_conda.sh install_magma_conda.sh
-RUN bash ./install_conda.sh && rm install_conda.sh install_magma_conda.sh common_utils.sh /opt/conda/requirements-ci.txt /opt/conda/requirements-docs.txt
+RUN bash ./install_conda.sh && rm install_conda.sh common_utils.sh /opt/conda/requirements-ci.txt /opt/conda/requirements-docs.txt
 RUN if [ -n "${UNINSTALL_DILL}" ]; then pip uninstall -y dill; fi
 
 # Install gcc
@@ -88,6 +87,19 @@ RUN if [ -n "${VISION}" ]; then bash ./install_vision.sh; fi
 RUN rm install_vision.sh cache_vision_models.sh common_utils.sh
 ENV INSTALLED_VISION ${VISION}
 
+# (optional) Install Android NDK
+ARG ANDROID
+ARG ANDROID_NDK
+ARG GRADLE_VERSION
+COPY ./common/install_android.sh ./common/cache_vision_models.sh ./common/common_utils.sh ./
+COPY ./android/AndroidManifest.xml AndroidManifest.xml
+COPY ./android/build.gradle build.gradle
+RUN if [ -n "${ANDROID}" ]; then bash ./install_android.sh; fi
+RUN rm install_android.sh cache_vision_models.sh common_utils.sh
+RUN rm AndroidManifest.xml
+RUN rm build.gradle
+ENV INSTALLED_ANDROID ${ANDROID}
+
 # (optional) Install Vulkan SDK
 ARG VULKAN_SDK_VERSION
 COPY ./common/install_vulkan_sdk.sh install_vulkan_sdk.sh

.ci/magma/.gitignore

@@ -1,2 +0,0 @@
-output/
-magma-cuda*/

.ci/magma/Makefile

@@ -1,48 +0,0 @@
-SHELL=/usr/bin/env bash
-
-DOCKER_CMD ?= docker
-DESIRED_CUDA ?= 11.8
-DESIRED_CUDA_SHORT = $(subst .,,$(DESIRED_CUDA))
-PACKAGE_NAME = magma-cuda
-CUDA_ARCH_LIST ?= -gencode arch=compute_50,code=sm_50 -gencode arch=compute_60,code=sm_60 -gencode arch=compute_70,code=sm_70 -gencode arch=compute_80,code=sm_80 -gencode arch=compute_86,code=sm_86 -gencode arch=compute_90,code=sm_90
-
-DOCKER_RUN = set -eou pipefail; ${DOCKER_CMD} run --rm -i \
-	-v $(shell git rev-parse --show-toplevel)/.ci:/builder \
-	-w /builder \
-	-e PACKAGE_NAME=${PACKAGE_NAME}${DESIRED_CUDA_SHORT} \
-	-e DESIRED_CUDA=${DESIRED_CUDA} \
-	-e CUDA_ARCH_LIST="${CUDA_ARCH_LIST}" \
-	"pytorch/manylinux-builder:cuda${DESIRED_CUDA}-main" \
-	magma/build_magma.sh
-
-.PHONY: all
-all: magma-cuda126
-all: magma-cuda124
-all: magma-cuda121
-all: magma-cuda118
-
-.PHONY:
-clean:
-	$(RM) -r magma-*
-	$(RM) -r output
-
-.PHONY: magma-cuda126
-magma-cuda126: DESIRED_CUDA := 12.6
-magma-cuda126:
-	$(DOCKER_RUN)
-
-.PHONY: magma-cuda124
-magma-cuda124: DESIRED_CUDA := 12.4
-magma-cuda124:
-	$(DOCKER_RUN)
-
-.PHONY: magma-cuda121
-magma-cuda121: DESIRED_CUDA := 12.1
-magma-cuda121:
-	$(DOCKER_RUN)
-
-.PHONY: magma-cuda118
-magma-cuda118: DESIRED_CUDA := 11.8
-magma-cuda118: CUDA_ARCH_LIST += -gencode arch=compute_37,code=sm_37
-magma-cuda118:
-	$(DOCKER_RUN)

.ci/magma/README.md

@@ -1,50 +0,0 @@
-# Magma
-
-This folder contains the scripts and configurations to build magma, statically linked for various versions of CUDA.
-
-## Building
-
-Look in the `Makefile` for available targets to build. To build any target, for example `magma-cuda118`, run
-
-```
-# Using `docker`
-make magma-cuda118
-
-# Using `podman`
-DOCKER_CMD=podman make magma-cuda118
-```
-
-This spawns a `pytorch/manylinux-cuda<version>` docker image, which has the required `devtoolset` and CUDA versions installed.
-Within the docker image, it runs `build_magma.sh` with the correct environment variables set, which package the necessary files
-into a tarball, with the following structure:
-
-```
-.
-├── include       # header files
-├── lib           # libmagma.a
-├── info
-│   ├── licenses  # license file
-│   └── recipe    # build script and patches
-```
-
-More specifically, `build_magma.sh` copies over the relevant files from the `package_files` directory depending on the CUDA version.
-Outputted binaries should be in the `output` folder.
-
-
-## Pushing
-
-Packages can be uploaded to an S3 bucket using:
-
-```
-aws s3 cp output/*/magma-cuda*.bz2 <bucket-with-path>
-```
-
-If you do not have upload permissions, please ping @seemethere or @soumith to gain access
-
-## New versions
-
-New CUDA versions can be added by creating a new make target with the next desired version. For CUDA version NN.n, the target should be named `magma-cudaNNn`.
-
-Make sure to edit the appropriate environment variables (e.g., DESIRED_CUDA, CUDA_ARCH_LIST) in the `Makefile` accordingly. Remember also to check `build_magma.sh` to ensure the logic for copying over the files remains correct.
-
-New patches can be added by editing `Makefile` and`build_magma.sh` the same way `getrf_nbparam.patch` is implemented.

.ci/magma/build_magma.sh

@@ -1,50 +0,0 @@
-#!/usr/bin/env bash
-
-set -eou pipefail
-
-# Environment variables
-# The script expects DESIRED_CUDA and PACKAGE_NAME to be set
-ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
-MAGMA_VERSION=2.6.1
-
-# Folders for the build
-PACKAGE_FILES=${ROOT_DIR}/magma/package_files # source patches and metadata
-PACKAGE_DIR=${ROOT_DIR}/magma/${PACKAGE_NAME} # build workspace
-PACKAGE_OUTPUT=${ROOT_DIR}/magma/output # where tarballs are stored
-PACKAGE_BUILD=${PACKAGE_DIR}/build # where the content of the tarball is prepared
-PACKAGE_RECIPE=${PACKAGE_BUILD}/info/recipe
-PACKAGE_LICENSE=${PACKAGE_BUILD}/info/licenses
-mkdir -p ${PACKAGE_DIR} ${PACKAGE_OUTPUT}/linux-64 ${PACKAGE_BUILD} ${PACKAGE_RECIPE} ${PACKAGE_LICENSE}
-
-# Fetch magma sources and verify checksum
-pushd ${PACKAGE_DIR}
-curl -LO http://icl.utk.edu/projectsfiles/magma/downloads/magma-${MAGMA_VERSION}.tar.gz
-tar zxf magma-${MAGMA_VERSION}.tar.gz
-sha256sum --check < ${PACKAGE_FILES}/magma-${MAGMA_VERSION}.sha256
-popd
-
-# Apply patches and build
-pushd ${PACKAGE_DIR}/magma-${MAGMA_VERSION}
-patch < ${PACKAGE_FILES}/CMake.patch
-patch < ${PACKAGE_FILES}/cmakelists.patch
-patch -p0 < ${PACKAGE_FILES}/thread_queue.patch
-patch -p1 < ${PACKAGE_FILES}/getrf_shfl.patch
-patch -p1 < ${PACKAGE_FILES}/getrf_nbparam.patch
-# The build.sh script expects to be executed from the sources root folder
-INSTALL_DIR=${PACKAGE_BUILD} ${PACKAGE_FILES}/build.sh
-popd
-
-# Package recipe, license and tarball
-# Folder and package name are backward compatible for the build workflow
-cp ${PACKAGE_FILES}/build.sh ${PACKAGE_RECIPE}/build.sh
-cp ${PACKAGE_FILES}/thread_queue.patch ${PACKAGE_RECIPE}/thread_queue.patch
-cp ${PACKAGE_FILES}/cmakelists.patch ${PACKAGE_RECIPE}/cmakelists.patch
-cp ${PACKAGE_FILES}/getrf_shfl.patch ${PACKAGE_RECIPE}/getrf_shfl.patch
-cp ${PACKAGE_FILES}/getrf_nbparam.patch ${PACKAGE_RECIPE}/getrf_nbparam.patch
-cp ${PACKAGE_FILES}/CMake.patch ${PACKAGE_RECIPE}/CMake.patch
-cp ${PACKAGE_FILES}/magma-${MAGMA_VERSION}.sha256 ${PACKAGE_RECIPE}/magma-${MAGMA_VERSION}.sha256
-cp ${PACKAGE_DIR}/magma-${MAGMA_VERSION}/COPYRIGHT ${PACKAGE_LICENSE}/COPYRIGHT
-pushd ${PACKAGE_BUILD}
-tar cjf ${PACKAGE_OUTPUT}/linux-64/${PACKAGE_NAME}-${MAGMA_VERSION}-1.tar.bz2 include lib info
-echo Built in ${PACKAGE_OUTPUT}/linux-64/${PACKAGE_NAME}-${MAGMA_VERSION}-1.tar.bz2
-popd

.ci/magma/package_files/CMake.patch

@@ -1,40 +0,0 @@
---- CMake.src.cuda	2023-03-29 10:05:32.136954140 +0000
-+++ CMake.src.cuda	2023-03-29 10:05:50.281318043 +0000
-@@ -283,10 +283,10 @@
- magmablas/zgeadd.cu
- magmablas/zgeadd2.cu
- magmablas/zgeam.cu
--magmablas/zgemm_fermi.cu
-+#magmablas/zgemm_fermi.cu
- magmablas/zgemm_reduce.cu
- magmablas/zgemv_conj.cu
--magmablas/zgemv_fermi.cu
-+#magmablas/zgemv_fermi.cu
- magmablas/zgerbt.cu
- magmablas/zgerbt_kernels.cu
- magmablas/zgetmatrix_transpose.cpp
-@@ -1009,18 +1009,18 @@
- magmablas/sgeam.cu
- magmablas/dgeam.cu
- magmablas/cgeam.cu
--magmablas/sgemm_fermi.cu
--magmablas/dgemm_fermi.cu
--magmablas/cgemm_fermi.cu
-+#magmablas/sgemm_fermi.cu
-+#magmablas/dgemm_fermi.cu
-+#magmablas/cgemm_fermi.cu
- magmablas/sgemm_reduce.cu
- magmablas/dgemm_reduce.cu
- magmablas/cgemm_reduce.cu
- magmablas/sgemv_conj.cu
- magmablas/dgemv_conj.cu
- magmablas/cgemv_conj.cu
--magmablas/sgemv_fermi.cu
--magmablas/dgemv_fermi.cu
--magmablas/cgemv_fermi.cu
-+#magmablas/sgemv_fermi.cu
-+#magmablas/dgemv_fermi.cu
-+#magmablas/cgemv_fermi.cu
- magmablas/sgerbt.cu
- magmablas/dgerbt.cu
- magmablas/cgerbt.cu

.ci/magma/package_files/build.sh

@@ -1,12 +0,0 @@
-CUDA__VERSION=$(nvcc --version|sed -n 4p|cut -f5 -d" "|cut -f1 -d",")
-if [ "$CUDA__VERSION" != "$DESIRED_CUDA" ]; then
-    echo "CUDA Version is not $DESIRED_CUDA. CUDA Version found: $CUDA__VERSION"
-    exit 1
-fi
-
-mkdir build
-cd build
-cmake .. -DUSE_FORTRAN=OFF -DGPU_TARGET="All" -DCMAKE_INSTALL_PREFIX="$INSTALL_DIR" -DCUDA_ARCH_LIST="$CUDA_ARCH_LIST"
-make -j$(getconf _NPROCESSORS_CONF)
-make install
-cd ..

.ci/magma/package_files/cmakelists.patch

@@ -1,388 +0,0 @@
-diff --git a/CMakeLists.txt b/CMakeLists.txt
-index d5d8d87d..8a507334 100644
---- a/CMakeLists.txt
-+++ b/CMakeLists.txt
-@@ -3,7 +3,7 @@ cmake_minimum_required( VERSION 2.8.1 )
- # ----------------------------------------
- # to disable Fortran, set this to "off"
- # see also -DADD_ below
--option( USE_FORTRAN "Fortran is required for some tester checks, but can be disabled with reduced functionality" ON )
-+option( USE_FORTRAN "Fortran is required for some tester checks, but can be disabled with reduced functionality" OFF )
-
- if (USE_FORTRAN)
-     project( MAGMA C CXX Fortran )
-@@ -75,6 +75,8 @@ else()
-     message( WARNING "The compiler ${CMAKE_CXX_COMPILER} doesn't support the -std=c++11 flag. Some code may not compile.")
- endif()
-
-+set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -static-libstdc++ -fno-exceptions")
-+
- CHECK_C_COMPILER_FLAG("-std=c99" COMPILER_SUPPORTS_C99)
- if (COMPILER_SUPPORTS_C99)
-     set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -std=c99")
-@@ -101,15 +103,15 @@ endif()
-
-
- # ----------------------------------------
--# locate OpenMP
--find_package( OpenMP )
--if (OPENMP_FOUND)
--    message( STATUS "Found OpenMP" )
--    message( STATUS "    OpenMP_C_FLAGS   ${OpenMP_C_FLAGS}" )
--    message( STATUS "    OpenMP_CXX_FLAGS ${OpenMP_CXX_FLAGS}" )
--    set( CMAKE_C_FLAGS   "${CMAKE_C_FLAGS} ${OpenMP_C_FLAGS}" )
--    set( CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${OpenMP_CXX_FLAGS}" )
--endif()
-+# # locate OpenMP
-+# find_package( OpenMP )
-+# if (OPENMP_FOUND)
-+#     message( STATUS "Found OpenMP" )
-+#     message( STATUS "    OpenMP_C_FLAGS   ${OpenMP_C_FLAGS}" )
-+#     message( STATUS "    OpenMP_CXX_FLAGS ${OpenMP_CXX_FLAGS}" )
-+#     set( CMAKE_C_FLAGS   "${CMAKE_C_FLAGS} ${OpenMP_C_FLAGS}" )
-+#     set( CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${OpenMP_CXX_FLAGS}" )
-+# endif()
-
- if (MAGMA_ENABLE_CUDA)
-   # ----------------------------------------
-@@ -132,7 +134,7 @@ if (MAGMA_ENABLE_CUDA)
-     set( NV_SM    "" )
-     set( NV_COMP  "" )
-
--    set(CUDA_SEPARABLE_COMPILATION ON)
-+    set(CUDA_SEPARABLE_COMPILATION OFF)
-
-     # nvcc >= 6.5 supports -std=c++11, so propagate CXXFLAGS to NVCCFLAGS.
-     # Older nvcc didn't support -std=c++11, so previously we disabled propagation.
-@@ -294,11 +296,18 @@ if (MAGMA_ENABLE_CUDA)
-         message( STATUS "    compile for CUDA arch 8.0 (Ampere)" )
-     endif()
-
-+    if ( ${GPU_TARGET} MATCHES "All")
-+        set( MIN_ARCH 370)
-+        SET( NV_SM ${CUDA_ARCH_LIST})
-+        SET( NV_COMP "")
-+    endif()
-+
-     if (NOT MIN_ARCH)
-         message( FATAL_ERROR "GPU_TARGET must contain one or more of Fermi, Kepler, Maxwell, Pascal, Volta, Turing, Ampere, or valid sm_[0-9][0-9]" )
-     endif()
-
--    set( CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS} -Xcompiler -fPIC ${NV_SM} ${NV_COMP} ${FORTRAN_CONVENTION} )
-+    set( CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS} -DHAVE_CUBLAS -Xfatbin -compress-all -Xcompiler -fPIC -std=c++11 ${NV_SM} ${NV_COMP} ${FORTRAN_CONVENTION} )
-+    MESSAGE(STATUS "CUDA_NVCC_FLAGS: ${CUDA_NVCC_FLAGS}")
-     #add_definitions( "-DMAGMA_HAVE_CUDA -DMAGMA_CUDA_ARCH_MIN=${MIN_ARCH}" )
-     set(MAGMA_HAVE_CUDA "1")
-     set(MAGMA_CUDA_ARCH_MIN "${MIN_ARCH}")
-@@ -413,7 +422,7 @@ set_property(CACHE BLA_VENDOR PROPERTY STRINGS
- set( LAPACK_LIBRARIES "" CACHE STRING "Libraries for LAPACK and BLAS, to manually override search" )
- if (LAPACK_LIBRARIES STREQUAL "")
-     message( STATUS "Searching for BLAS and LAPACK. To override, set LAPACK_LIBRARIES using ccmake." )
--    find_package( LAPACK )
-+    # find_package( LAPACK )
-     # force showing updated LAPACK_LIBRARIES in ccmake / cmake-gui.
-     set( LAPACK_LIBRARIES ${LAPACK_LIBRARIES} CACHE STRING "Libraries for LAPACK and BLAS, to manually override search" FORCE )
- else()
-@@ -552,12 +561,12 @@ if (WIN32)
-     #message( "libmagma_all_f   ${libmagma_all_f}"   )
-
-     # on Windows, Fortran files aren't compiled if listed here...
--    cuda_add_library( magma ${libmagma_all_cpp} )
-+    cuda_add_library( magma STATIC ${libmagma_all_cpp} OPTIONS --compiler-options "-fPIC")
-     target_link_libraries( magma
-         ${LAPACK_LIBRARIES}
-         ${CUDA_CUDART_LIBRARY}
-         ${CUDA_CUBLAS_LIBRARIES}
--        ${CUDA_cusparse_LIBRARY}
-+        # ${CUDA_cusparse_LIBRARY}
-     )
-
-     # no Fortran files at the moment (how to test libmagma_all_f is not empty?),
-@@ -575,13 +584,13 @@ if (WIN32)
- else()
-     # Unix doesn't seem to have a problem with mixing C, CUDA, and Fortran files
-     if (MAGMA_ENABLE_CUDA)
--      cuda_add_library( magma ${libmagma_all} )
-+      cuda_add_library( magma STATIC ${libmagma_all} OPTIONS --compiler-options "-fPIC")
-       target_link_libraries( magma
-         ${blas_fix}
-         ${LAPACK_LIBRARIES}
-         ${CUDA_CUDART_LIBRARY}
-         ${CUDA_CUBLAS_LIBRARIES}
--        ${CUDA_cusparse_LIBRARY}
-+        # ${CUDA_cusparse_LIBRARY}
- 	)
-     else()
-       find_package( hipBLAS )
-@@ -614,138 +623,139 @@ else()
-     endif()
- endif()
- add_custom_target( lib DEPENDS magma )
--
--
--# ----------------------------------------
--# compile lapacktest library
--# If use fortran, compile only Fortran files, not magma_[sdcz]_no_fortran.cpp
--# else,           compile only C++     files, not Fortran files
--if (USE_FORTRAN)
--    foreach( filename ${liblapacktest_all} )
--        if (filename MATCHES "\\.(f|f90|F90)$")
--            list( APPEND liblapacktest_all_f ${filename} )
--        endif()
--    endforeach()
--    add_library( lapacktest ${liblapacktest_all_f} )
--else()
--    # alternatively, use only C/C++/CUDA files, including magma_[sdcz]_no_fortran.cpp
--    foreach( filename ${liblapacktest_all} )
--        if (filename MATCHES "\\.(c|cu|cpp)$")
--            list( APPEND liblapacktest_all_cpp ${filename} )
--        endif()
--    endforeach()
--    add_library( lapacktest ${liblapacktest_all_cpp} )
--endif()
--target_link_libraries( lapacktest
--    ${blas_fix}
--    ${LAPACK_LIBRARIES}
--)
--
--
--# ----------------------------------------
--# compile tester library
--add_library( tester ${libtest_all} )
--target_link_libraries( tester
--    magma
--    lapacktest
--    ${blas_fix}
--    ${LAPACK_LIBRARIES}
--)
-+set_target_properties(magma PROPERTIES POSITION_INDEPENDENT_CODE ON)
-+
-+
-+# # ----------------------------------------
-+# # compile lapacktest library
-+# # If use fortran, compile only Fortran files, not magma_[sdcz]_no_fortran.cpp
-+# # else,           compile only C++     files, not Fortran files
-+# if (USE_FORTRAN)
-+#     foreach( filename ${liblapacktest_all} )
-+#         if (filename MATCHES "\\.(f|f90|F90)$")
-+#             list( APPEND liblapacktest_all_f ${filename} )
-+#         endif()
-+#     endforeach()
-+#     add_library( lapacktest ${liblapacktest_all_f} )
-+# else()
-+#     # alternatively, use only C/C++/CUDA files, including magma_[sdcz]_no_fortran.cpp
-+#     foreach( filename ${liblapacktest_all} )
-+#         if (filename MATCHES "\\.(c|cu|cpp)$")
-+#             list( APPEND liblapacktest_all_cpp ${filename} )
-+#         endif()
-+#     endforeach()
-+#     add_library( lapacktest ${liblapacktest_all_cpp} )
-+# endif()
-+# target_link_libraries( lapacktest
-+#     ${blas_fix}
-+#     ${LAPACK_LIBRARIES}
-+# )
-+
-+
-+# # ----------------------------------------
-+# # compile tester library
-+# add_library( tester ${libtest_all} )
-+# target_link_libraries( tester
-+#     magma
-+#     lapacktest
-+#     ${blas_fix}
-+#     ${LAPACK_LIBRARIES}
-+# )
-
-
- # ----------------------------------------
- # compile MAGMA sparse library
-
- # sparse doesn't have Fortran at the moment, so no need for above shenanigans
--if (MAGMA_ENABLE_CUDA)
--  include_directories( sparse/include )
--  include_directories( sparse/control )
--else()
--  include_directories( sparse_hip/include )
--  include_directories( sparse_hip/control )
--endif()
--include_directories( testing )
--
--if (MAGMA_ENABLE_CUDA)
--  cuda_add_library( magma_sparse ${libsparse_all} )
--  target_link_libraries( magma_sparse
--    magma
--    ${blas_fix}
--    ${LAPACK_LIBRARIES}
--    ${CUDA_CUDART_LIBRARY}
--    ${CUDA_CUBLAS_LIBRARIES}
--    ${CUDA_cusparse_LIBRARY}
--    )
--else()
--  add_library( magma_sparse ${libsparse_all} )
--  target_link_libraries( magma_sparse
--    magma
--    ${blas_fix}
--    ${LAPACK_LIBRARIES}
--    hip::device
--    roc::hipblas
--    roc::hipsparse
--    )
--endif()
--add_custom_target( sparse-lib DEPENDS magma_sparse )
--
--
--# ----------------------------------------
--# compile each tester
--
--# save testers to testing/
--# save tester lib files to testing_lib/ to avoid cluttering lib/
--set( CMAKE_RUNTIME_OUTPUT_DIRECTORY testing )
--set( CMAKE_ARCHIVE_OUTPUT_DIRECTORY testing_lib )
--set( CMAKE_LIBRARY_OUTPUT_DIRECTORY testing_lib )
--
--# skip Fortran testers, which require an extra file from CUDA
--foreach( filename ${testing_all} )
--    if (filename MATCHES "\\.(c|cu|cpp)$")
--        list( APPEND testing_all_cpp ${filename} )
--    endif()
--endforeach()
--foreach( TEST ${testing_all_cpp} )
--    string( REGEX REPLACE "\\.(cpp|f90|F90)" "" EXE ${TEST} )
--    string( REGEX REPLACE "testing/" "" EXE ${EXE} )
--    #message( "${TEST} --> ${EXE}" )
--    add_executable( ${EXE} ${TEST} )
--    target_link_libraries( ${EXE} tester lapacktest magma )
--    list( APPEND testing ${EXE} )
--endforeach()
--add_custom_target( testing DEPENDS ${testing} )
--
--
--# ----------------------------------------
--# compile each sparse tester
--
--if (MAGMA_ENABLE_CUDA)
--  set(SPARSE_TEST_DIR "sparse/testing")
--else()
--  set(SPARSE_TEST_DIR "sparse_hip/testing")
--endif()
--
--
--set( CMAKE_RUNTIME_OUTPUT_DIRECTORY "${SPARSE_TEST_DIR}" )
--cmake_policy( SET CMP0037 OLD)
--foreach( TEST ${sparse_testing_all} )
--    string( REGEX REPLACE "\\.(cpp|f90|F90)"     "" EXE ${TEST} )
--    string( REGEX REPLACE "${SPARSE_TEST_DIR}/" "" EXE ${EXE} )
--    #message( "${TEST} --> ${EXE}" )
--    add_executable( ${EXE} ${TEST} )
--    target_link_libraries( ${EXE} magma_sparse magma )
--    list( APPEND sparse-testing ${EXE} )
--endforeach()
--add_custom_target( sparse-testing DEPENDS ${sparse-testing} )
-+# if (MAGMA_ENABLE_CUDA)
-+#   include_directories( sparse/include )
-+#   include_directories( sparse/control )
-+# else()
-+#   include_directories( sparse_hip/include )
-+#   include_directories( sparse_hip/control )
-+# endif()
-+# include_directories( testing )
-+
-+# if (MAGMA_ENABLE_CUDA)
-+#   cuda_add_library( magma_sparse ${libsparse_all} )
-+#   target_link_libraries( magma_sparse
-+#     magma
-+#     ${blas_fix}
-+#     ${LAPACK_LIBRARIES}
-+#     ${CUDA_CUDART_LIBRARY}
-+#     ${CUDA_CUBLAS_LIBRARIES}
-+#     ${CUDA_cusparse_LIBRARY}
-+#     )
-+# else()
-+#   add_library( magma_sparse ${libsparse_all} )
-+#   target_link_libraries( magma_sparse
-+#     magma
-+#     ${blas_fix}
-+#     ${LAPACK_LIBRARIES}
-+#     hip::device
-+#     roc::hipblas
-+#     roc::hipsparse
-+#     )
-+# endif()
-+# add_custom_target( sparse-lib DEPENDS magma_sparse )
-+
-+
-+# # ----------------------------------------
-+# # compile each tester
-+
-+# # save testers to testing/
-+# # save tester lib files to testing_lib/ to avoid cluttering lib/
-+# set( CMAKE_RUNTIME_OUTPUT_DIRECTORY testing )
-+# set( CMAKE_ARCHIVE_OUTPUT_DIRECTORY testing_lib )
-+# set( CMAKE_LIBRARY_OUTPUT_DIRECTORY testing_lib )
-+
-+# # skip Fortran testers, which require an extra file from CUDA
-+# foreach( filename ${testing_all} )
-+#     if (filename MATCHES "\\.(c|cu|cpp)$")
-+#         list( APPEND testing_all_cpp ${filename} )
-+#     endif()
-+# endforeach()
-+# foreach( TEST ${testing_all_cpp} )
-+#     string( REGEX REPLACE "\\.(cpp|f90|F90)" "" EXE ${TEST} )
-+#     string( REGEX REPLACE "testing/" "" EXE ${EXE} )
-+#     #message( "${TEST} --> ${EXE}" )
-+#     add_executable( ${EXE} ${TEST} )
-+#     target_link_libraries( ${EXE} tester lapacktest magma )
-+#     list( APPEND testing ${EXE} )
-+# endforeach()
-+# add_custom_target( testing DEPENDS ${testing} )
-+
-+
-+# # ----------------------------------------
-+# # compile each sparse tester
-+
-+# if (MAGMA_ENABLE_CUDA)
-+#   set(SPARSE_TEST_DIR "sparse/testing")
-+# else()
-+#   set(SPARSE_TEST_DIR "sparse_hip/testing")
-+# endif()
-+
-+
-+# set( CMAKE_RUNTIME_OUTPUT_DIRECTORY "${SPARSE_TEST_DIR}" )
-+# cmake_policy( SET CMP0037 OLD)
-+# foreach( TEST ${sparse_testing_all} )
-+#     string( REGEX REPLACE "\\.(cpp|f90|F90)"     "" EXE ${TEST} )
-+#     string( REGEX REPLACE "${SPARSE_TEST_DIR}/" "" EXE ${EXE} )
-+#     #message( "${TEST} --> ${EXE}" )
-+#     add_executable( ${EXE} ${TEST} )
-+#     target_link_libraries( ${EXE} magma_sparse magma )
-+#     list( APPEND sparse-testing ${EXE} )
-+# endforeach()
-+# add_custom_target( sparse-testing DEPENDS ${sparse-testing} )
-
-
- # ----------------------------------------
- # what to install
--install( TARGETS magma magma_sparse ${blas_fix}
-+install( TARGETS magma ${blas_fix}
-          RUNTIME DESTINATION bin
-          LIBRARY DESTINATION lib
-          ARCHIVE DESTINATION lib )
--file( GLOB headers include/*.h sparse/include/*.h "${CMAKE_BINARY_DIR}/include/*.h" )
-+file( GLOB headers include/*.h "${CMAKE_BINARY_DIR}/include/*.h" )
- if (USE_FORTRAN)
-     install( FILES ${headers} ${modules}
-              DESTINATION include )
-@@ -769,9 +779,9 @@ else()
-     "${blas_fix_lib} ${LAPACK_LIBS} hip::device roc::hipblas roc::hipsparse" )
- endif()
- set( MAGMA_REQUIRED "" )
--configure_file( "${pkgconfig}.in" "${pkgconfig}" @ONLY )
--install( FILES "${CMAKE_BINARY_DIR}/${pkgconfig}"
--         DESTINATION lib/pkgconfig )
-+# configure_file( "${pkgconfig}.in" "${pkgconfig}" @ONLY )
-+# install( FILES "${CMAKE_BINARY_DIR}/${pkgconfig}"
-+#          DESTINATION lib/pkgconfig )
-
- # ----------------------------------------
- get_directory_property( compile_definitions COMPILE_DEFINITIONS )

.ci/magma/package_files/getrf_nbparam.patch

@@ -1,40 +0,0 @@
-diff --git a/control/get_batched_crossover.cpp b/control/get_batched_crossover.cpp
-index 4ec57306..912f8608 100644
---- a/control/get_batched_crossover.cpp
-+++ b/control/get_batched_crossover.cpp
-@@ -119,7 +119,7 @@ void magma_get_spotrf_batched_nbparam(magma_int_t n, magma_int_t *nb, magma_int_
- void magma_get_zgetrf_batched_nbparam(magma_int_t n, magma_int_t *nb, magma_int_t *recnb)
- {
-     *nb    = 64;
--    *recnb = 32;
-+    *recnb = 16;
-     return;
- }
- 
-@@ -127,7 +127,7 @@ void magma_get_zgetrf_batched_nbparam(magma_int_t n, magma_int_t *nb, magma_int_
- void magma_get_cgetrf_batched_nbparam(magma_int_t n, magma_int_t *nb, magma_int_t *recnb)
- {
-     *nb    = 128;
--    *recnb =  32;
-+    *recnb =  16;
-     return;
- }
- 
-@@ -135,7 +135,7 @@ void magma_get_cgetrf_batched_nbparam(magma_int_t n, magma_int_t *nb, magma_int_
- void magma_get_dgetrf_batched_nbparam(magma_int_t n, magma_int_t *nb, magma_int_t *recnb)
- {
-     *nb    = 128;
--    *recnb =  32;
-+    *recnb =  16;
-     return;
- }
- 
-@@ -143,7 +143,7 @@ void magma_get_dgetrf_batched_nbparam(magma_int_t n, magma_int_t *nb, magma_int_
- void magma_get_sgetrf_batched_nbparam(magma_int_t n, magma_int_t *nb, magma_int_t *recnb)
- {
-     *nb    = 128;
--    *recnb =  32;
-+    *recnb =  16;
-     return;
- }
- 

.ci/magma/package_files/getrf_shfl.patch

@@ -1,15 +0,0 @@
-diff --git a/src/zgetrf_batched.cpp b/src/zgetrf_batched.cpp
-index 24a65a90..884d9352 100644
---- a/src/zgetrf_batched.cpp
-+++ b/src/zgetrf_batched.cpp
-@@ -116,7 +116,9 @@ magma_zgetrf_batched(
-             return magma_zgetrf_batched_smallsq_noshfl( m, dA_array, ldda, ipiv_array, info_array, batchCount, queue );
-         }
-         else{
--            return magma_zgetrf_batched_smallsq_shfl( m, dA_array, ldda, ipiv_array, info_array, batchCount, queue );
-+            // magma_cgetrf_batched_smallsq_shfl is broken, therefore let's call noshfl version for arch < 700
-+            // return magma_zgetrf_batched_smallsq_shfl( m, dA_array, ldda, ipiv_array, info_array, batchCount, queue );
-+            return magma_zgetrf_batched_smallsq_noshfl( m, dA_array, ldda, ipiv_array, info_array, batchCount, queue );
-         }
-         #else
-         return magma_zgetrf_batched_smallsq_noshfl( m, dA_array, ldda, ipiv_array, info_array, batchCount, queue );

.ci/magma/package_files/magma-2.6.1.sha256

@@ -1 +0,0 @@
-6cd83808c6e8bc7a44028e05112b3ab4e579bcc73202ed14733f66661127e213  magma-2.6.1.tar.gz

.ci/magma/package_files/thread_queue.patch

@@ -1,20 +0,0 @@
---- control/thread_queue.cpp	2016-08-30 06:37:49.000000000 -0700
-+++ control/thread_queue.cpp	2016-10-10 19:47:28.911580965 -0700
-@@ -15,7 +15,7 @@
- {
-     if ( err != 0 ) {
-         fprintf( stderr, "Error: %s (%d)\n", strerror(err), err );
--        throw std::exception();
-+        // throw std::exception();
-     }
- }
- 
-@@ -172,7 +172,7 @@
-     check( pthread_mutex_lock( &mutex ));
-     if ( quit_flag ) {
-         fprintf( stderr, "Error: push_task() called after quit()\n" );
--        throw std::exception();
-+        // throw std::exception();
-     }
-     q.push( task );
-     ntask += 1;

.ci/manywheel/build.sh

@@ -15,12 +15,9 @@ case "${GPU_ARCH_TYPE:-BLANK}" in
     rocm)
         bash "${SCRIPTPATH}/build_rocm.sh"
         ;;
-    cpu | cpu-cxx11-abi | cpu-s390x)
+    cpu | cpu-cxx11-abi | cpu-s390x | xpu)
         bash "${SCRIPTPATH}/build_cpu.sh"
         ;;
-    xpu)
-        bash "${SCRIPTPATH}/build_xpu.sh"
-        ;;
     *)
         echo "Un-recognized GPU_ARCH_TYPE '${GPU_ARCH_TYPE}', exiting..."
         exit 1

.ci/manywheel/build_common.sh

@@ -4,9 +4,12 @@
 set -ex
 SOURCE_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null && pwd )"
 
-source ${SOURCE_DIR}/set_desired_python.sh
-
 
+# Require only one python installation
+if [[ -z "$DESIRED_PYTHON" ]]; then
+    echo "Need to set DESIRED_PYTHON env variable"
+    exit 1
+fi
 if [[ -n "$BUILD_PYTHONLESS" && -z "$LIBTORCH_VARIANT" ]]; then
     echo "BUILD_PYTHONLESS is set, so need LIBTORCH_VARIANT to also be set"
     echo "LIBTORCH_VARIANT should be one of shared-with-deps shared-without-deps static-with-deps static-without-deps"
@@ -18,14 +21,12 @@ retry () {
     $*  || (sleep 1 && $*) || (sleep 2 && $*) || (sleep 4 && $*) || (sleep 8 && $*)
 }
 
-PLATFORM="manylinux2014_x86_64"
 # TODO move this into the Docker images
 OS_NAME=$(awk -F= '/^NAME/{print $2}' /etc/os-release)
 if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then
     retry yum install -q -y zip openssl
 elif [[ "$OS_NAME" == *"AlmaLinux"* ]]; then
     retry yum install -q -y zip openssl
-    PLATFORM="manylinux_2_28_x86_64"
 elif [[ "$OS_NAME" == *"Red Hat Enterprise Linux"* ]]; then
     retry dnf install -q -y zip openssl
 elif [[ "$OS_NAME" == *"Ubuntu"* ]]; then
@@ -79,7 +80,27 @@ if [[ -e /opt/openssl ]]; then
     export CMAKE_INCLUDE_PATH="/opt/openssl/include":$CMAKE_INCLUDE_PATH
 fi
 
+# If given a python version like 3.6m or 2.7mu, convert this to the format we
+# expect. The binary CI jobs pass in python versions like this; they also only
+# ever pass one python version, so we assume that DESIRED_PYTHON is not a list
+# in this case
+if [[ -n "$DESIRED_PYTHON" && $DESIRED_PYTHON =~ ([0-9].[0-9]+)t ]]; then
+    python_digits="$(echo $DESIRED_PYTHON | tr -cd [:digit:])"
+    py_majmin="${DESIRED_PYTHON}"
+    DESIRED_PYTHON="cp${python_digits}-cp${python_digits}t"
+elif [[ -n "$DESIRED_PYTHON" && "$DESIRED_PYTHON" != cp* ]]; then
+    python_nodot="$(echo $DESIRED_PYTHON | tr -d m.u)"
+    DESIRED_PYTHON="cp${python_nodot}-cp${python_nodot}"
+    if [[ ${python_nodot} -ge 310 ]]; then
+        py_majmin="${DESIRED_PYTHON:2:1}.${DESIRED_PYTHON:3:2}"
+    else
+        py_majmin="${DESIRED_PYTHON:2:1}.${DESIRED_PYTHON:3:1}"
+    fi
+fi
 
+pydir="/opt/python/$DESIRED_PYTHON"
+export PATH="$pydir/bin:$PATH"
+echo "Will build for Python version: ${DESIRED_PYTHON} with ${python_installation}"
 
 mkdir -p /tmp/$WHEELHOUSE_DIR
 
@@ -255,11 +276,11 @@ make_wheel_record() {
     FPATH=$1
     if echo $FPATH | grep RECORD >/dev/null 2>&1; then
         # if the RECORD file, then
-        echo "\"$FPATH\",,"
+        echo "$FPATH,,"
     else
         HASH=$(openssl dgst -sha256 -binary $FPATH | openssl base64 | sed -e 's/+/-/g' | sed -e 's/\//_/g' | sed -e 's/=//g')
         FSIZE=$(ls -nl $FPATH | awk '{print $5}')
-        echo "\"$FPATH\",sha256=$HASH,$FSIZE"
+        echo "$FPATH,sha256=$HASH,$FSIZE"
     fi
 }
 
@@ -379,12 +400,6 @@ for pkg in /$WHEELHOUSE_DIR/torch_no_python*.whl /$WHEELHOUSE_DIR/torch*linux*.w
         $PATCHELF_BIN --print-rpath $sofile
     done
 
-    # create Manylinux 2_28 tag this needs to happen before regenerate the RECORD
-    if [[ $PLATFORM == "manylinux_2_28_x86_64" && $GPU_ARCH_TYPE != "cpu-s390x" && $GPU_ARCH_TYPE != "xpu" ]]; then
-        wheel_file=$(echo $(basename $pkg) | sed -e 's/-cp.*$/.dist-info\/WHEEL/g')
-        sed -i -e s#linux_x86_64#"${PLATFORM}"# $wheel_file;
-    fi
-
     # regenerate the RECORD file with new hashes
     record_file=$(echo $(basename $pkg) | sed -e 's/-cp.*$/.dist-info\/RECORD/g')
     if [[ -e $record_file ]]; then
@@ -424,20 +439,12 @@ for pkg in /$WHEELHOUSE_DIR/torch_no_python*.whl /$WHEELHOUSE_DIR/torch*linux*.w
         popd
     fi
 
-    # Rename wheel for Manylinux 2_28
-    if [[ $PLATFORM == "manylinux_2_28_x86_64" && $GPU_ARCH_TYPE != "cpu-s390x" && $GPU_ARCH_TYPE != "xpu" ]]; then
-        pkg_name=$(echo $(basename $pkg) | sed -e s#linux_x86_64#"${PLATFORM}"#)
-        zip -rq $pkg_name $PREIX*
-        rm -f $pkg
-        mv $pkg_name $(dirname $pkg)/$pkg_name
-    else
-        # zip up the wheel back
-        zip -rq $(basename $pkg) $PREIX*
-        # remove original wheel
-        rm -f $pkg
-        mv $(basename $pkg) $pkg
-    fi
+    # zip up the wheel back
+    zip -rq $(basename $pkg) $PREIX*
 
+    # replace original wheel
+    rm -f $pkg
+    mv $(basename $pkg) $pkg
     cd ..
     rm -rf tmp
 done
@@ -490,9 +497,9 @@ if [[ -z "$BUILD_PYTHONLESS" ]]; then
   echo "$(date) :: Running tests"
   pushd "$PYTORCH_ROOT"
 
-
+  #TODO: run_tests.sh and check_binary.sh should be moved to pytorch/pytorch project
   LD_LIBRARY_PATH=/usr/local/nvidia/lib64 \
-          "${PYTORCH_ROOT}/.ci/pytorch/run_tests.sh" manywheel "${py_majmin}" "$DESIRED_CUDA"
+          "/builder/run_tests.sh" manywheel "${py_majmin}" "$DESIRED_CUDA"
   popd
   echo "$(date) :: Finished tests"
 fi

.ci/manywheel/build_cpu.sh

@@ -2,6 +2,8 @@
 
 set -ex
 
+GPU_ARCH_TYPE=${GPU_ARCH_TYPE:-cpu}
+
 export TH_BINARY_BUILD=1
 export USE_CUDA=0
 
@@ -15,13 +17,22 @@ if [[ -z "$EXTRA_CAFFE2_CMAKE_FLAGS" ]]; then
     EXTRA_CAFFE2_CMAKE_FLAGS=()
 fi
 
-WHEELHOUSE_DIR="wheelhousecpu"
-LIBTORCH_HOUSE_DIR="libtorch_housecpu"
+DIR_SUFFIX=cpu
+if [[ "$GPU_ARCH_TYPE" == "xpu" ]]; then
+    DIR_SUFFIX=xpu
+    # Refer https://www.intel.com/content/www/us/en/developer/articles/tool/pytorch-prerequisites-for-intel-gpu/2-5.html
+    source /opt/intel/oneapi/pytorch-gpu-dev-0.5/oneapi-vars.sh
+    source /opt/intel/oneapi/pti/latest/env/vars.sh
+    export USE_STATIC_MKL=1
+fi
+
+WHEELHOUSE_DIR="wheelhouse$DIR_SUFFIX"
+LIBTORCH_HOUSE_DIR="libtorch_house$DIR_SUFFIX"
 if [[ -z "$PYTORCH_FINAL_PACKAGE_DIR" ]]; then
     if [[ -z "$BUILD_PYTHONLESS" ]]; then
-        PYTORCH_FINAL_PACKAGE_DIR="/remote/wheelhousecpu"
+        PYTORCH_FINAL_PACKAGE_DIR="/remote/wheelhouse$DIR_SUFFIX"
     else
-        PYTORCH_FINAL_PACKAGE_DIR="/remote/libtorch_housecpu"
+        PYTORCH_FINAL_PACKAGE_DIR="/remote/libtorch_house$DIR_SUFFIX"
     fi
 fi
 mkdir -p "$PYTORCH_FINAL_PACKAGE_DIR" || true
@@ -49,6 +60,34 @@ DEPS_SONAME=(
     "libgomp.so.1"
 )
 
+if [[ "$GPU_ARCH_TYPE" == "xpu" ]]; then
+    echo "Bundling with xpu support package libs."
+    DEPS_LIST+=(
+        "/opt/intel/oneapi/compiler/latest/lib/libsycl-preview.so.7"
+        "/opt/intel/oneapi/compiler/latest/lib/libOpenCL.so.1"
+        "/opt/intel/oneapi/compiler/latest/lib/libxptifw.so"
+        "/opt/intel/oneapi/compiler/latest/lib/libsvml.so"
+        "/opt/intel/oneapi/compiler/latest/lib/libirng.so"
+        "/opt/intel/oneapi/compiler/latest/lib/libimf.so"
+        "/opt/intel/oneapi/compiler/latest/lib/libintlc.so.5"
+        "/opt/intel/oneapi/compiler/latest/lib/libpi_level_zero.so"
+        "/opt/intel/oneapi/pti/latest/lib/libpti_view.so.0.9"
+        "/opt/intel/oneapi/pti/latest/lib/libpti.so.0.9"
+    )
+    DEPS_SONAME+=(
+        "libsycl-preview.so.7"
+        "libOpenCL.so.1"
+        "libxptifw.so"
+        "libsvml.so"
+        "libirng.so"
+        "libimf.so"
+        "libintlc.so.5"
+        "libpi_level_zero.so"
+        "libpti_view.so.0.9"
+        "libpti.so.0.9"
+    )
+fi
+
 rm -rf /usr/local/cuda*
 
 SOURCE_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null && pwd )"

.ci/manywheel/build_cuda.sh

@@ -43,6 +43,13 @@ if [[ -n "$DESIRED_CUDA" ]]; then
         fi
     fi
     echo "Using CUDA $CUDA_VERSION as determined by DESIRED_CUDA"
+
+    # There really has to be a better way to do this - eli
+    # Possibly limiting builds to specific cuda versions be delimiting images would be a choice
+    if [[ "$OS_NAME" == *"Ubuntu"* ]]; then
+        echo "Switching to CUDA version ${DESIRED_CUDA}"
+        /builder/conda/switch_cuda_version.sh "${DESIRED_CUDA}"
+    fi
 else
     CUDA_VERSION=$(nvcc --version|grep release|cut -f5 -d" "|cut -f1 -d",")
     echo "CUDA $CUDA_VERSION Detected"
@@ -52,11 +59,15 @@ cuda_version_nodot=$(echo $CUDA_VERSION | tr -d '.')
 
 TORCH_CUDA_ARCH_LIST="5.0;6.0;7.0;7.5;8.0;8.6"
 case ${CUDA_VERSION} in
-    12.6)
-        TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST};9.0+PTX"
+    12.4)
+        if [[ "$GPU_ARCH_TYPE" = "cuda-aarch64" ]]; then
+            TORCH_CUDA_ARCH_LIST="9.0"
+        else
+            TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST};9.0+PTX"
+        fi
         EXTRA_CAFFE2_CMAKE_FLAGS+=("-DATEN_NO_TEST=ON")
         ;;
-    12.4)
+    12.1)
         TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST};9.0"
         EXTRA_CAFFE2_CMAKE_FLAGS+=("-DATEN_NO_TEST=ON")
         ;;
@@ -64,6 +75,10 @@ case ${CUDA_VERSION} in
         TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST};3.7;9.0"
         EXTRA_CAFFE2_CMAKE_FLAGS+=("-DATEN_NO_TEST=ON")
         ;;
+    11.[67])
+        TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST};3.7"
+        EXTRA_CAFFE2_CMAKE_FLAGS+=("-DATEN_NO_TEST=ON")
+        ;;
     *)
         echo "unknown cuda version $CUDA_VERSION"
         exit 1
@@ -103,9 +118,7 @@ DEPS_SONAME=(
     "libgomp.so.1"
 )
 
-# CUDA 11.8 have to ship the libcusparseLt.so.0 with the binary
-# since nvidia-cusparselt-cu11 is not available in PYPI
-if [[ $USE_CUSPARSELT == "1" && $CUDA_VERSION == "11.8" ]]; then
+if [[ $USE_CUSPARSELT == "1" ]]; then
         DEPS_SONAME+=(
             "libcusparseLt.so.0"
         )
@@ -114,7 +127,7 @@ if [[ $USE_CUSPARSELT == "1" && $CUDA_VERSION == "11.8" ]]; then
         )
 fi
 
-if [[ $CUDA_VERSION == "12.4" || $CUDA_VERSION == "12.6" ]]; then
+if [[ $CUDA_VERSION == "12.1" || $CUDA_VERSION == "12.4" ]]; then
     export USE_STATIC_CUDNN=0
     # Try parallelizing nvcc as well
     export TORCH_NVCC_FLAGS="-Xfatbin -compress-all --threads 2"
@@ -132,7 +145,6 @@ if [[ $CUDA_VERSION == "12.4" || $CUDA_VERSION == "12.6" ]]; then
             "/usr/local/cuda/lib64/libcudnn.so.9"
             "/usr/local/cuda/lib64/libcublas.so.12"
             "/usr/local/cuda/lib64/libcublasLt.so.12"
-            "/usr/local/cuda/lib64/libcusparseLt.so.0"
             "/usr/local/cuda/lib64/libcudart.so.12"
             "/usr/local/cuda/lib64/libnvToolsExt.so.1"
             "/usr/local/cuda/lib64/libnvrtc.so.12"
@@ -149,7 +161,6 @@ if [[ $CUDA_VERSION == "12.4" || $CUDA_VERSION == "12.6" ]]; then
             "libcudnn.so.9"
             "libcublas.so.12"
             "libcublasLt.so.12"
-            "libcusparseLt.so.0"
             "libcudart.so.12"
             "libnvToolsExt.so.1"
             "libnvrtc.so.12"
@@ -167,7 +178,6 @@ if [[ $CUDA_VERSION == "12.4" || $CUDA_VERSION == "12.6" ]]; then
             '$ORIGIN/../../nvidia/curand/lib'
             '$ORIGIN/../../nvidia/cusolver/lib'
             '$ORIGIN/../../nvidia/cusparse/lib'
-            '$ORIGIN/../../cusparselt/lib'
             '$ORIGIN/../../nvidia/nccl/lib'
             '$ORIGIN/../../nvidia/nvtx/lib'
         )
@@ -256,7 +266,7 @@ else
     exit 1
 fi
 
-# run_tests.sh requires DESIRED_CUDA to know what tests to exclude
+# builder/test.sh requires DESIRED_CUDA to know what tests to exclude
 export DESIRED_CUDA="$cuda_version_nodot"
 
 # Switch `/usr/local/cuda` to the desired CUDA version

.ci/manywheel/build_libtorch.sh

@@ -225,11 +225,11 @@ make_wheel_record() {
     FPATH=$1
     if echo $FPATH | grep RECORD >/dev/null 2>&1; then
         # if the RECORD file, then
-        echo "\"$FPATH\",,"
+        echo "$FPATH,,"
     else
         HASH=$(openssl dgst -sha256 -binary $FPATH | openssl base64 | sed -e 's/+/-/g' | sed -e 's/\//_/g' | sed -e 's/=//g')
         FSIZE=$(ls -nl $FPATH | awk '{print $5}')
-        echo "\"$FPATH\",sha256=$HASH,$FSIZE"
+        echo "$FPATH,sha256=$HASH,$FSIZE"
     fi
 }
 

.ci/manywheel/build_rocm.sh

@@ -107,29 +107,17 @@ if [[ $ROCM_INT -ge 60200 ]]; then
 fi
 
 OS_NAME=`awk -F= '/^NAME/{print $2}' /etc/os-release`
-if [[ "$OS_NAME" == *"CentOS Linux"* || "$OS_NAME" == *"AlmaLinux"* ]]; then
+if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then
     LIBGOMP_PATH="/usr/lib64/libgomp.so.1"
     LIBNUMA_PATH="/usr/lib64/libnuma.so.1"
     LIBELF_PATH="/usr/lib64/libelf.so.1"
-    if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then
-        LIBTINFO_PATH="/usr/lib64/libtinfo.so.5"
-    else
-        LIBTINFO_PATH="/usr/lib64/libtinfo.so.6"
-    fi
+    LIBTINFO_PATH="/usr/lib64/libtinfo.so.5"
     LIBDRM_PATH="/opt/amdgpu/lib64/libdrm.so.2"
     LIBDRM_AMDGPU_PATH="/opt/amdgpu/lib64/libdrm_amdgpu.so.1"
-    if [[ $ROCM_INT -ge 60100 && $ROCM_INT -lt 60300 ]]; then
+    if [[ $ROCM_INT -ge 60100 ]]; then
         # Below libs are direct dependencies of libhipsolver
         LIBSUITESPARSE_CONFIG_PATH="/lib64/libsuitesparseconfig.so.4"
-        if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then
-            LIBCHOLMOD_PATH="/lib64/libcholmod.so.2"
-            # Below libs are direct dependencies of libsatlas
-            LIBGFORTRAN_PATH="/lib64/libgfortran.so.3"
-        else
-            LIBCHOLMOD_PATH="/lib64/libcholmod.so.3"
-            # Below libs are direct dependencies of libsatlas
-            LIBGFORTRAN_PATH="/lib64/libgfortran.so.5"
-        fi
+        LIBCHOLMOD_PATH="/lib64/libcholmod.so.2"
         # Below libs are direct dependencies of libcholmod
         LIBAMD_PATH="/lib64/libamd.so.2"
         LIBCAMD_PATH="/lib64/libcamd.so.2"
@@ -137,6 +125,7 @@ if [[ "$OS_NAME" == *"CentOS Linux"* || "$OS_NAME" == *"AlmaLinux"* ]]; then
         LIBCOLAMD_PATH="/lib64/libcolamd.so.2"
         LIBSATLAS_PATH="/lib64/atlas/libsatlas.so.3"
         # Below libs are direct dependencies of libsatlas
+        LIBGFORTRAN_PATH="/lib64/libgfortran.so.3"
         LIBQUADMATH_PATH="/lib64/libquadmath.so.0"
     fi
     MAYBE_LIB64=lib64
@@ -151,7 +140,7 @@ elif [[ "$OS_NAME" == *"Ubuntu"* ]]; then
     fi
     LIBDRM_PATH="/usr/lib/x86_64-linux-gnu/libdrm.so.2"
     LIBDRM_AMDGPU_PATH="/usr/lib/x86_64-linux-gnu/libdrm_amdgpu.so.1"
-    if [[ $ROCM_INT -ge 60100 && $ROCM_INT -lt 60300 ]]; then
+    if [[ $ROCM_INT -ge 60100 ]]; then
         # Below libs are direct dependencies of libhipsolver
         LIBCHOLMOD_PATH="/lib/x86_64-linux-gnu/libcholmod.so.3"
         # Below libs are direct dependencies of libcholmod
@@ -186,6 +175,12 @@ do
     OS_SO_FILES[${#OS_SO_FILES[@]}]=$file_name # Append lib to array
 done
 
+# PyTorch-version specific
+# AOTriton dependency only for PyTorch >= 2.4
+if (( $(echo "${PYTORCH_VERSION} 2.4" | awk '{print ($1 >= $2)}') )); then
+    ROCM_SO_FILES+=("libaotriton_v2.so")
+fi
+
 # rocBLAS library files
 ROCBLAS_LIB_SRC=$ROCM_HOME/lib/rocblas/library
 ROCBLAS_LIB_DST=lib/rocblas/library

.ci/manywheel/build_xpu.sh

@@ -1,108 +0,0 @@
-#!/usr/bin/env bash
-
-set -ex
-
-export TH_BINARY_BUILD=1
-export USE_CUDA=0
-
-# Keep an array of cmake variables to add to
-if [[ -z "$CMAKE_ARGS" ]]; then
-    # These are passed to tools/build_pytorch_libs.sh::build()
-    CMAKE_ARGS=()
-fi
-if [[ -z "$EXTRA_CAFFE2_CMAKE_FLAGS" ]]; then
-    # These are passed to tools/build_pytorch_libs.sh::build_caffe2()
-    EXTRA_CAFFE2_CMAKE_FLAGS=()
-fi
-
-
-# Refer https://www.intel.com/content/www/us/en/developer/articles/tool/pytorch-prerequisites-for-intel-gpus.html
-source /opt/intel/oneapi/compiler/latest/env/vars.sh
-source /opt/intel/oneapi/pti/latest/env/vars.sh
-source /opt/intel/oneapi/umf/latest/env/vars.sh
-export USE_STATIC_MKL=1
-
-WHEELHOUSE_DIR="wheelhousexpu"
-LIBTORCH_HOUSE_DIR="libtorch_housexpu"
-if [[ -z "$PYTORCH_FINAL_PACKAGE_DIR" ]]; then
-    if [[ -z "$BUILD_PYTHONLESS" ]]; then
-        PYTORCH_FINAL_PACKAGE_DIR="/remote/wheelhousexpu"
-    else
-        PYTORCH_FINAL_PACKAGE_DIR="/remote/libtorch_housexpu"
-    fi
-fi
-mkdir -p "$PYTORCH_FINAL_PACKAGE_DIR" || true
-
-OS_NAME=$(awk -F= '/^NAME/{print $2}' /etc/os-release)
-if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then
-    LIBGOMP_PATH="/usr/lib64/libgomp.so.1"
-elif [[ "$OS_NAME" == *"Red Hat Enterprise Linux"* ]]; then
-    LIBGOMP_PATH="/usr/lib64/libgomp.so.1"
-elif [[ "$OS_NAME" == *"AlmaLinux"* ]]; then
-    LIBGOMP_PATH="/usr/lib64/libgomp.so.1"
-elif [[ "$OS_NAME" == *"Ubuntu"* ]]; then
-    if [[ "$(uname -m)" == "s390x" ]]; then
-        LIBGOMP_PATH="/usr/lib/s390x-linux-gnu/libgomp.so.1"
-    else
-        LIBGOMP_PATH="/usr/lib/x86_64-linux-gnu/libgomp.so.1"
-    fi
-fi
-
-DEPS_LIST=(
-    "$LIBGOMP_PATH"
-    "/opt/intel/oneapi/compiler/latest/lib/libOpenCL.so.1"
-)
-
-DEPS_SONAME=(
-    "libgomp.so.1"
-    "libOpenCL.so.1"
-)
-
-if [[ -z "$PYTORCH_EXTRA_INSTALL_REQUIREMENTS" ]]; then
-    echo "Bundling with xpu support package libs."
-    DEPS_LIST+=(
-        "/opt/intel/oneapi/compiler/latest/lib/libsycl.so.8"
-        "/opt/intel/oneapi/compiler/latest/lib/libur_loader.so.0"
-        "/opt/intel/oneapi/compiler/latest/lib/libur_adapter_level_zero.so.0"
-        "/opt/intel/oneapi/compiler/latest/lib/libur_adapter_opencl.so.0"
-        "/opt/intel/oneapi/compiler/latest/lib/libsvml.so"
-        "/opt/intel/oneapi/compiler/latest/lib/libirng.so"
-        "/opt/intel/oneapi/compiler/latest/lib/libimf.so"
-        "/opt/intel/oneapi/compiler/latest/lib/libintlc.so.5"
-        "/opt/intel/oneapi/pti/latest/lib/libpti_view.so.0.10"
-        "/opt/intel/oneapi/umf/latest/lib/libumf.so.0"
-        "/opt/intel/oneapi/tcm/latest/lib/libhwloc.so.15"
-    )
-    DEPS_SONAME+=(
-        "libsycl.so.8"
-        "libur_loader.so.0"
-        "libur_adapter_level_zero.so.0"
-        "libur_adapter_opencl.so.0"
-        "libsvml.so"
-        "libirng.so"
-        "libimf.so"
-        "libintlc.so.5"
-        "libpti_view.so.0.10"
-        "libumf.so.0"
-        "libhwloc.so.15"
-    )
-else
-    echo "Using xpu runtime libs from pypi."
-    XPU_RPATHS=(
-        '$ORIGIN/../../../..'
-    )
-    XPU_RPATHS=$(IFS=: ; echo "${XPU_RPATHS[*]}")
-    export C_SO_RPATH=$XPU_RPATHS':$ORIGIN:$ORIGIN/lib'
-    export LIB_SO_RPATH=$XPU_RPATHS':$ORIGIN'
-    export FORCE_RPATH="--force-rpath"
-fi
-
-rm -rf /usr/local/cuda*
-
-SOURCE_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null && pwd )"
-if [[ -z "$BUILD_PYTHONLESS" ]]; then
-    BUILD_SCRIPT=build_common.sh
-else
-    BUILD_SCRIPT=build_libtorch.sh
-fi
-source ${SOURCE_DIR}/${BUILD_SCRIPT}

.ci/manywheel/set_desired_python.sh

@@ -1,30 +0,0 @@
-#!/usr/bin/env bash
-
-# Require only one python installation
-if [[ -z "$DESIRED_PYTHON" ]]; then
-    echo "Need to set DESIRED_PYTHON env variable"
-    exit 1
-fi
-
-# If given a python version like 3.6m or 2.7mu, convert this to the format we
-# expect. The binary CI jobs pass in python versions like this; they also only
-# ever pass one python version, so we assume that DESIRED_PYTHON is not a list
-# in this case
-if [[ -n "$DESIRED_PYTHON" && $DESIRED_PYTHON =~ ([0-9].[0-9]+)t ]]; then
-    python_digits="$(echo $DESIRED_PYTHON | tr -cd [:digit:])"
-    py_majmin="${DESIRED_PYTHON}"
-    DESIRED_PYTHON="cp${python_digits}-cp${python_digits}t"
-elif [[ -n "$DESIRED_PYTHON" && "$DESIRED_PYTHON" != cp* ]]; then
-    python_nodot="$(echo $DESIRED_PYTHON | tr -d m.u)"
-    DESIRED_PYTHON="cp${python_nodot}-cp${python_nodot}"
-    if [[ ${python_nodot} -ge 310 ]]; then
-        py_majmin="${DESIRED_PYTHON:2:1}.${DESIRED_PYTHON:3:2}"
-    else
-        py_majmin="${DESIRED_PYTHON:2:1}.${DESIRED_PYTHON:3:1}"
-    fi
-fi
-
-pydir="/opt/python/$DESIRED_PYTHON"
-export DESIRED_PYTHON_BIN_DIR="${pydir}/bin"
-export PATH="$DESIRED_PYTHON_BIN_DIR:$PATH"
-echo "Will build for Python version: ${DESIRED_PYTHON}"

.ci/pytorch/build.sh

@@ -1,6 +1,6 @@
 #!/bin/bash
 
-set -ex -o pipefail
+set -ex
 
 # Required environment variable: $BUILD_ENVIRONMENT
 # (This is set by default in the Docker images we build, so you don't
@@ -87,7 +87,7 @@ else
 
   # Workaround required for MKL library linkage
   # https://github.com/pytorch/pytorch/issues/119557
-  if [[ "$ANACONDA_PYTHON_VERSION" = "3.12" || "$ANACONDA_PYTHON_VERSION" = "3.13" ]]; then
+  if [ "$ANACONDA_PYTHON_VERSION" = "3.12" ]; then
     export CMAKE_LIBRARY_PATH="/opt/conda/envs/py_$ANACONDA_PYTHON_VERSION/lib/"
     export CMAKE_INCLUDE_PATH="/opt/conda/envs/py_$ANACONDA_PYTHON_VERSION/include/"
   fi
@@ -191,7 +191,7 @@ fi
 
 # We only build FlashAttention files for CUDA 8.0+, and they require large amounts of
 # memory to build and will OOM
-if [[ "$BUILD_ENVIRONMENT" == *cuda* ]] && [[ 1 -eq $(echo "${TORCH_CUDA_ARCH_LIST} >= 8.0" | bc) ]]; then
+if [[ "$BUILD_ENVIRONMENT" == *cuda* ]] && [[ "$TORCH_CUDA_ARCH_LIST" == *"8.6"* || "$TORCH_CUDA_ARCH_LIST" == *"8.0"* ]]; then
   echo "WARNING: FlashAttention files require large amounts of memory to build and will OOM"
   echo "Setting MAX_JOBS=(nproc-2)/3 to reduce memory usage"
   export MAX_JOBS="$(( $(nproc --ignore=2) / 3 ))"
@@ -228,9 +228,9 @@ if [[ "$BUILD_ENVIRONMENT" == *-debug* ]]; then
   export CMAKE_BUILD_TYPE=RelWithAssert
 fi
 
-# Do not change workspace permissions for ROCm and s390x CI jobs
+# Do not change workspace permissions for ROCm CI jobs
 # as it can leave workspace with bad permissions for cancelled jobs
-if [[ "$BUILD_ENVIRONMENT" != *rocm* && "$BUILD_ENVIRONMENT" != *s390x* && -d /var/lib/jenkins/workspace ]]; then
+if [[ "$BUILD_ENVIRONMENT" != *rocm* && "$BUILD_ENVIRONMENT" != *s390x* ]]; then
   # Workaround for dind-rootless userid mapping (https://github.com/pytorch/ci-infra/issues/96)
   WORKSPACE_ORIGINAL_OWNER_ID=$(stat -c '%u' "/var/lib/jenkins/workspace")
   cleanup_workspace() {
@@ -247,9 +247,10 @@ if [[ "$BUILD_ENVIRONMENT" != *rocm* && "$BUILD_ENVIRONMENT" != *s390x* && -d /v
 fi
 
 if [[ "$BUILD_ENVIRONMENT" == *-bazel-* ]]; then
-  set -e -o pipefail
+  set -e
 
   get_bazel
+  install_sccache_nvcc_for_bazel
 
   # Leave 1 CPU free and use only up to 80% of memory to reduce the change of crashing
   # the runner
@@ -278,13 +279,14 @@ else
           "$BUILD_ENVIRONMENT" != *xla* ]]; then
       if [[ "$BUILD_ENVIRONMENT" != *py3.8* ]]; then
         # Install numpy-2.0.2 for builds which are backward compatible with 1.X
-        python -mpip install numpy==2.0.2
+        python -mpip install --pre numpy==2.0.2
       fi
 
       WERROR=1 python setup.py clean
 
       if [[ "$USE_SPLIT_BUILD" == "true" ]]; then
-        python3 tools/packaging/split_wheel.py bdist_wheel
+        BUILD_LIBTORCH_WHL=1 BUILD_PYTHON_ONLY=0 python setup.py bdist_wheel
+        BUILD_LIBTORCH_WHL=0 BUILD_PYTHON_ONLY=1 python setup.py bdist_wheel --cmake
       else
         WERROR=1 python setup.py bdist_wheel
       fi
@@ -395,7 +397,9 @@ if [[ "$BUILD_ENVIRONMENT" != *libtorch* && "$BUILD_ENVIRONMENT" != *bazel* ]];
   # don't do this for libtorch as libtorch is C++ only and thus won't have python tests run on its build
   python tools/stats/export_test_times.py
 fi
-# don't do this for bazel or s390x as they don't use sccache
-if [[ "$BUILD_ENVIRONMENT" != *s390x* && "$BUILD_ENVIRONMENT" != *-bazel-* ]]; then
+
+# snadampal: skipping it till sccache support added for aarch64
+# https://github.com/pytorch/pytorch/issues/121559
+if [[ "$BUILD_ENVIRONMENT" != *aarch64* &&  "$BUILD_ENVIRONMENT" != *s390x* ]]; then
   print_sccache_stats
 fi

.ci/pytorch/check_binary.sh

@@ -1,394 +0,0 @@
-#!/bin/bash
-
-# shellcheck disable=SC2086,SC2006,SC2207,SC2076,SC2155,SC2046,SC1091,SC2143
-# TODO: Re-enable shellchecks above
-
-set -eux -o pipefail
-
-# This script checks the following things on binaries
-# 1. The gcc abi matches DESIRED_DEVTOOLSET
-# 2. MacOS binaries do not link against OpenBLAS
-# 3. There are no protobuf symbols of any sort anywhere (turned off, because
-#    this is currently not true)
-# 4. Standard Python imports work
-# 5. MKL is available everywhere except for MacOS wheels
-# 6. XNNPACK is available everywhere except for MacOS wheels
-# 7. CUDA is setup correctly and does not hang
-# 8. Magma is available for CUDA builds
-# 9. CuDNN is available for CUDA builds
-#
-# This script needs the env variables DESIRED_PYTHON, DESIRED_CUDA,
-# DESIRED_DEVTOOLSET and PACKAGE_TYPE
-#
-# This script expects PyTorch to be installed into the active Python (the
-# Python returned by `which python`). Or, if this is testing a libtorch
-# Pythonless binary, then it expects to be in the root folder of the unzipped
-# libtorch package.
-
-
-if [[ -z ${DESIRED_PYTHON:-} ]]; then
-  export DESIRED_PYTHON=${MATRIX_PYTHON_VERSION:-}
-fi
-if [[ -z ${DESIRED_CUDA:-} ]]; then
-  export DESIRED_CUDA=${MATRIX_DESIRED_CUDA:-}
-fi
-if [[ -z ${DESIRED_DEVTOOLSET:-} ]]; then
-  export DESIRED_DEVTOOLSET=${MATRIX_DESIRED_DEVTOOLSET:-}
-fi
-if [[ -z ${PACKAGE_TYPE:-} ]]; then
-  export PACKAGE_TYPE=${MATRIX_PACKAGE_TYPE:-}
-fi
-
-# The install root depends on both the package type and the os
-# All MacOS packages use conda, even for the wheel packages.
-if [[ "$PACKAGE_TYPE" == libtorch ]]; then
-  # NOTE: Only $PWD works on both CentOS and Ubuntu
-  export install_root="$PWD"
-else
-
-  if [[ $DESIRED_PYTHON =~ ([0-9].[0-9]+)t ]]; then
-    # For python that is maj.mint keep original version
-    py_dot="$DESIRED_PYTHON"
-  elif [[ $DESIRED_PYTHON =~ ([0-9].[0-9]+) ]];  then
-    # Strip everything but major.minor from DESIRED_PYTHON version
-    py_dot="${BASH_REMATCH[0]}"
-  else
-    echo "Unexpected ${DESIRED_PYTHON} format"
-    exit 1
-  fi
-  export install_root="$(dirname $(which python))/../lib/python${py_dot}/site-packages/torch/"
-fi
-
-###############################################################################
-# Setup XPU ENV
-###############################################################################
-if [[ "$DESIRED_CUDA" == 'xpu' ]]; then
-  set +u
-  # Refer https://www.intel.com/content/www/us/en/developer/articles/tool/pytorch-prerequisites-for-intel-gpus.html
-  source /opt/intel/oneapi/compiler/latest/env/vars.sh
-  source /opt/intel/oneapi/pti/latest/env/vars.sh
-fi
-
-###############################################################################
-# Check GCC ABI
-###############################################################################
-
-# NOTE [ Building libtorch with old vs. new gcc ABI ]
-#
-# Packages built with one version of ABI could not be linked against by client
-# C++ libraries that were compiled using the other version of ABI. Since both
-# gcc ABIs are still common in the wild, we need to support both ABIs. Currently:
-#
-# - All the nightlies built on CentOS 7 + devtoolset7 use the old gcc ABI.
-# - All the nightlies built on Ubuntu 16.04 + gcc 5.4 use the new gcc ABI.
-
-echo "Checking that the gcc ABI is what we expect"
-if [[ "$(uname)" != 'Darwin' ]]; then
-  function is_expected() {
-    if [[ "$DESIRED_DEVTOOLSET" == *"cxx11-abi"* || "$DESIRED_CUDA" == *"rocm"* ]]; then
-      if [[ "$1" -gt 0 || "$1" == "ON " ]]; then
-        echo 1
-      fi
-    else
-      if [[ -z "$1" || "$1" == 0 || "$1" == "OFF" ]]; then
-        echo 1
-      fi
-    fi
-  }
-
-  # First we check that the env var in TorchConfig.cmake is correct
-
-  # We search for D_GLIBCXX_USE_CXX11_ABI=1 in torch/TorchConfig.cmake
-  torch_config="${install_root}/share/cmake/Torch/TorchConfig.cmake"
-  if [[ ! -f "$torch_config" ]]; then
-    echo "No TorchConfig.cmake found!"
-    ls -lah "$install_root/share/cmake/Torch"
-    exit 1
-  fi
-  echo "Checking the TorchConfig.cmake"
-  cat "$torch_config"
-
-  # The sed call below is
-  #   don't print lines by default (only print the line we want)
-  # -n
-  #   execute the following expression
-  # e
-  #   replace lines that match with the first capture group and print
-  # s/.*D_GLIBCXX_USE_CXX11_ABI=\(.\)".*/\1/p
-  #   any characters, D_GLIBCXX_USE_CXX11_ABI=, exactly one any character, a
-  #   quote, any characters
-  #   Note the exactly one single character after the '='. In the case that the
-  #     variable is not set the '=' will be followed by a '"' immediately and the
-  #     line will fail the match and nothing will be printed; this is what we
-  #     want.  Otherwise it will capture the 0 or 1 after the '='.
-  # /.*D_GLIBCXX_USE_CXX11_ABI=\(.\)".*/
-  #   replace the matched line with the capture group and print
-  # /\1/p
-  actual_gcc_abi="$(sed -ne 's/.*D_GLIBCXX_USE_CXX11_ABI=\(.\)".*/\1/p' < "$torch_config")"
-  if [[ "$(is_expected "$actual_gcc_abi")" != 1 ]]; then
-    echo "gcc ABI $actual_gcc_abi not as expected."
-    exit 1
-  fi
-
-  # We also check that there are [not] cxx11 symbols in libtorch
-  #
-  echo "Checking that symbols in libtorch.so have the right gcc abi"
-  python3 "$(dirname ${BASH_SOURCE[0]})/smoke_test/check_binary_symbols.py"
-
-  echo "cxx11 symbols seem to be in order"
-fi # if on Darwin
-
-###############################################################################
-# Check for no OpenBLAS
-# TODO Check for no Protobuf symbols (not finished)
-# Print *all* runtime dependencies
-###############################################################################
-# We have to loop through all shared libraries for this
-if [[ "$(uname)" == 'Darwin' ]]; then
-  all_dylibs=($(find "$install_root" -name '*.dylib'))
-  for dylib in "${all_dylibs[@]}"; do
-    echo "All dependencies of $dylib are $(otool -L $dylib) with rpath $(otool -l $dylib | grep LC_RPATH -A2)"
-
-    # Check that OpenBlas is not linked to on Macs
-    echo "Checking the OpenBLAS is not linked to"
-    if [[ -n "$(otool -L $dylib | grep -i openblas)" ]]; then
-      echo "ERROR: Found openblas as a dependency of $dylib"
-      echo "Full dependencies is: $(otool -L $dylib)"
-      exit 1
-    fi
-
-    # Check for protobuf symbols
-    #proto_symbols="$(nm $dylib | grep protobuf)" || true
-    #if [[ -n "$proto_symbols" ]]; then
-    #  echo "ERROR: Detected protobuf symbols in $dylib"
-    #  echo "Symbols are $proto_symbols"
-    #  exit 1
-    #fi
-  done
-else
-  all_libs=($(find "$install_root" -name '*.so'))
-  for lib in "${all_libs[@]}"; do
-    echo "All dependencies of $lib are $(ldd $lib) with runpath $(objdump -p $lib | grep RUNPATH)"
-
-    # Check for protobuf symbols
-    #proto_symbols=$(nm $lib | grep protobuf) || true
-    #if [[ -n "$proto_symbols" ]]; then
-    #  echo "ERROR: Detected protobuf symbols in $lib"
-    #  echo "Symbols are $proto_symbols"
-    #  exit 1
-    #fi
-  done
-fi
-
-setup_link_flags () {
-  REF_LIB="-Wl,-R${install_root}/lib"
-  if [[ "$(uname)" == 'Darwin' ]]; then
-    REF_LIB="-Wl,-rpath ${install_root}/lib"
-  fi
-  ADDITIONAL_LINKER_FLAGS=""
-  if [[ "$(uname)" == 'Linux' ]]; then
-    ADDITIONAL_LINKER_FLAGS="-Wl,--no-as-needed"
-  fi
-  C10_LINK_FLAGS=""
-  if [ -f "${install_root}/lib/libc10.so" ] || [ -f "${install_root}/lib/libc10.dylib" ]; then
-    C10_LINK_FLAGS="-lc10"
-  fi
-  TORCH_CPU_LINK_FLAGS=""
-  if [ -f "${install_root}/lib/libtorch_cpu.so" ] || [ -f "${install_root}/lib/libtorch_cpu.dylib" ]; then
-    TORCH_CPU_LINK_FLAGS="-ltorch_cpu"
-  fi
-  TORCH_CUDA_LINK_FLAGS=""
-  if [ -f "${install_root}/lib/libtorch_cuda.so" ] || [ -f "${install_root}/lib/libtorch_cuda.dylib" ]; then
-    TORCH_CUDA_LINK_FLAGS="-ltorch_cuda"
-  elif [ -f "${install_root}/lib/libtorch_cuda_cpp.so" ] && [ -f "${install_root}/lib/libtorch_cuda_cpp.so" ] || \
-    [ -f "${install_root}/lib/libtorch_cuda_cu.dylib" ] && [ -f "${install_root}/lib/libtorch_cuda_cu.dylib" ]; then
-    TORCH_CUDA_LINK_FLAGS="-ltorch_cuda_cpp -ltorch_cuda_cu"
-  fi
-}
-
-TEST_CODE_DIR="$(dirname $(realpath ${BASH_SOURCE[0]}))/test_example_code"
-build_and_run_example_cpp () {
-  if [[ "$DESIRED_DEVTOOLSET" == *"cxx11-abi"* ]]; then
-    GLIBCXX_USE_CXX11_ABI=1
-  else
-    GLIBCXX_USE_CXX11_ABI=0
-  fi
-  setup_link_flags
-  g++ ${TEST_CODE_DIR}/$1.cpp -I${install_root}/include -I${install_root}/include/torch/csrc/api/include -D_GLIBCXX_USE_CXX11_ABI=$GLIBCXX_USE_CXX11_ABI -std=gnu++17 -L${install_root}/lib ${REF_LIB} ${ADDITIONAL_LINKER_FLAGS} -ltorch $TORCH_CPU_LINK_FLAGS $TORCH_CUDA_LINK_FLAGS $C10_LINK_FLAGS -o $1
-  ./$1
-}
-
-build_example_cpp_with_incorrect_abi () {
-  if [[ "$DESIRED_DEVTOOLSET" == *"cxx11-abi"* ]]; then
-    GLIBCXX_USE_CXX11_ABI=0
-  else
-    GLIBCXX_USE_CXX11_ABI=1
-  fi
-  set +e
-  setup_link_flags
-  g++ ${TEST_CODE_DIR}/$1.cpp -I${install_root}/include -I${install_root}/include/torch/csrc/api/include -D_GLIBCXX_USE_CXX11_ABI=$GLIBCXX_USE_CXX11_ABI -std=gnu++17 -L${install_root}/lib ${REF_LIB} ${ADDITIONAL_LINKER_FLAGS} -ltorch $TORCH_CPU_LINK_FLAGS $TORCH_CUDA_LINK_FLAGS $C10_LINK_FLAGS -o $1
-  ERRCODE=$?
-  set -e
-  if [ "$ERRCODE" -eq "0" ]; then
-    echo "Building example with incorrect ABI didn't throw error. Aborting."
-    exit 1
-  else
-    echo "Building example with incorrect ABI throws expected error. Proceeding."
-  fi
-}
-
-###############################################################################
-# Check simple Python/C++ calls
-###############################################################################
-if [[ "$PACKAGE_TYPE" == 'libtorch' ]]; then
-  # NS: Set LD_LIBRARY_PATH for CUDA builds, but perhaps it should be removed
-  if [[ "$DESIRED_CUDA" == "cu"* ]]; then
-    export LD_LIBRARY_PATH=/usr/local/cuda/lib64
-  fi
-  build_and_run_example_cpp simple-torch-test
-  # `_GLIBCXX_USE_CXX11_ABI` is always ignored by gcc in devtoolset7, so we test
-  # the expected failure case for Ubuntu 16.04 + gcc 5.4 only.
-  if [[ "$DESIRED_DEVTOOLSET" == *"cxx11-abi"* ]]; then
-    build_example_cpp_with_incorrect_abi simple-torch-test
-  fi
-else
-  pushd /tmp
-  python -c 'import torch'
-  popd
-fi
-
-###############################################################################
-# Check torch.git_version
-###############################################################################
-if [[ "$PACKAGE_TYPE" != 'libtorch' ]]; then
-  pushd /tmp
-  python -c 'import torch; assert torch.version.git_version != "Unknown"'
-  python -c 'import torch; assert torch.version.git_version != None'
-  popd
-fi
-
-
-###############################################################################
-# Check for MKL
-###############################################################################
-
-if [[ "$PACKAGE_TYPE" == 'libtorch' ]]; then
-  echo "Checking that MKL is available"
-  build_and_run_example_cpp check-torch-mkl
-elif [[ "$(uname -m)" != "arm64" && "$(uname -m)" != "s390x" ]]; then
-  if [[ "$(uname)" != 'Darwin' || "$PACKAGE_TYPE" != *wheel ]]; then
-    if [[ "$(uname -m)" == "aarch64" ]]; then
-      echo "Checking that MKLDNN is available on aarch64"
-      pushd /tmp
-      python -c 'import torch; exit(0 if torch.backends.mkldnn.is_available() else 1)'
-      popd
-    else
-      echo "Checking that MKL is available"
-      pushd /tmp
-      python -c 'import torch; exit(0 if torch.backends.mkl.is_available() else 1)'
-      popd
-    fi
-  fi
-fi
-
-###############################################################################
-# Check for XNNPACK
-###############################################################################
-
-if [[ "$PACKAGE_TYPE" == 'libtorch' ]]; then
-  echo "Checking that XNNPACK is available"
-  build_and_run_example_cpp check-torch-xnnpack
-else
-  if [[ "$(uname)" != 'Darwin' || "$PACKAGE_TYPE" != *wheel ]] && [[ "$(uname -m)" != "s390x"  ]]; then
-    echo "Checking that XNNPACK is available"
-    pushd /tmp
-    python -c 'import torch.backends.xnnpack; exit(0 if torch.backends.xnnpack.enabled else 1)'
-    popd
-  fi
-fi
-
-###############################################################################
-# Check CUDA configured correctly
-###############################################################################
-# Skip these for Windows machines without GPUs
-if [[ "$OSTYPE" == "msys" ]]; then
-    GPUS=$(wmic path win32_VideoController get name)
-    if [[ ! "$GPUS" == *NVIDIA* ]]; then
-        echo "Skip CUDA tests for machines without a Nvidia GPU card"
-        exit 0
-    fi
-fi
-
-# Test that CUDA builds are setup correctly
-if [[ "$DESIRED_CUDA" != 'cpu' && "$DESIRED_CUDA" != 'xpu' && "$DESIRED_CUDA" != 'cpu-cxx11-abi' && "$DESIRED_CUDA" != *"rocm"* && "$(uname -m)" != "s390x" ]]; then
-  if [[ "$PACKAGE_TYPE" == 'libtorch' ]]; then
-    build_and_run_example_cpp check-torch-cuda
-  else
-    pushd /tmp
-    echo "Checking that CUDA archs are setup correctly"
-    timeout 20 python -c 'import torch; torch.randn([3,5]).cuda()'
-
-    # These have to run after CUDA is initialized
-
-    echo "Checking that magma is available"
-    python -c 'import torch; torch.rand(1).cuda(); exit(0 if torch.cuda.has_magma else 1)'
-
-    echo "Checking that CuDNN is available"
-    python -c 'import torch; exit(0 if torch.backends.cudnn.is_available() else 1)'
-
-    # Validates builds is free of linker regressions reported in https://github.com/pytorch/pytorch/issues/57744
-    echo "Checking that exception handling works"
-    python -c "import torch; from unittest import TestCase;TestCase().assertRaises(RuntimeError, lambda:torch.eye(7, 7, device='cuda:7'))"
-
-    echo "Checking that basic RNN works"
-    python ${TEST_CODE_DIR}/rnn_smoke.py
-
-    echo "Checking that basic CNN works"
-    python "${TEST_CODE_DIR}/cnn_smoke.py"
-
-    echo "Test that linalg works"
-    python -c "import torch;x=torch.rand(3,3,device='cuda');print(torch.linalg.svd(torch.mm(x.t(), x)))"
-
-    popd
-  fi # if libtorch
-fi # if cuda
-
-##########################
-# Run parts of smoke tests
-##########################
-if [[ "$PACKAGE_TYPE" != 'libtorch' ]]; then
-  pushd "$(dirname ${BASH_SOURCE[0]})/smoke_test"
-  python -c "from smoke_test import test_linalg; test_linalg()"
-  if [[ "$DESIRED_CUDA" == *cuda* ]]; then
-    python -c "from smoke_test import test_linalg; test_linalg('cuda')"
-  fi
-  popd
-fi
-
-###############################################################################
-# Check PyTorch supports TCP_TLS gloo transport
-###############################################################################
-
-if [[ "$(uname)" == 'Linux' && "$PACKAGE_TYPE" != 'libtorch' ]]; then
-  GLOO_CHECK="import torch.distributed as dist
-try:
-    dist.init_process_group('gloo', rank=0, world_size=1)
-except RuntimeError as e:
-    print(e)
-"
-  RESULT=`GLOO_DEVICE_TRANSPORT=TCP_TLS MASTER_ADDR=localhost MASTER_PORT=63945 python -c "$GLOO_CHECK"`
-  GLOO_TRANSPORT_IS_NOT_SUPPORTED='gloo transport is not supported'
-  if [[ "$RESULT" =~ "$GLOO_TRANSPORT_IS_NOT_SUPPORTED" ]]; then
-    echo "PyTorch doesn't support TLS_TCP transport, please build with USE_GLOO_WITH_OPENSSL=1"
-    exit 1
-  fi
-fi
-
-###############################################################################
-# Check for C++ ABI compatibility between gcc7 and gcc9 compiled binaries
-###############################################################################
-if [[ "$(uname)" == 'Linux' && ("$PACKAGE_TYPE" == 'conda' || "$PACKAGE_TYPE" == 'manywheel')]]; then
-  pushd /tmp
-  python -c "import torch; exit(0 if torch.compiled_with_cxx11_abi() else (0 if torch._C._PYBIND11_BUILD_ABI == '_cxxabi1011' else 1))"
-  popd
-fi

.ci/pytorch/common-build.sh

@@ -6,12 +6,6 @@ if [[ "$BUILD_ENVIRONMENT" != *win-* ]]; then
     # Save the absolute path in case later we chdir (as occurs in the gpu perf test)
     script_dir="$( cd "$(dirname "${BASH_SOURCE[0]}")" || exit ; pwd -P )"
 
-    if [[ "${BUILD_ENVIRONMENT}" == *-pch* ]]; then
-        # This is really weird, but newer sccache somehow produces broken binary
-        # see https://github.com/pytorch/pytorch/issues/139188
-        sudo mv /opt/cache/bin/sccache-0.2.14a /opt/cache/bin/sccache
-    fi
-
     if which sccache > /dev/null; then
         # Save sccache logs to file
         sccache --stop-server > /dev/null  2>&1 || true

.ci/pytorch/common.sh

@@ -3,7 +3,7 @@
 # Common setup for all Jenkins scripts
 # shellcheck source=./common_utils.sh
 source "$(dirname "${BASH_SOURCE[0]}")/common_utils.sh"
-set -ex -o pipefail
+set -ex
 
 # Required environment variables:
 #   $BUILD_ENVIRONMENT (should be set by your Docker image)

.ci/pytorch/common_utils.sh

@@ -81,15 +81,14 @@ function pip_install_whl() {
 
 function pip_install() {
   # retry 3 times
-  pip_install_pkg="python3 -m pip install --progress-bar off"
-  ${pip_install_pkg} "$@" || \
-    ${pip_install_pkg} "$@" || \
-    ${pip_install_pkg} "$@"
+  # old versions of pip don't have the "--progress-bar" flag
+  pip install --progress-bar off "$@" || pip install --progress-bar off "$@" || pip install --progress-bar off "$@" ||\
+  pip install "$@" || pip install "$@" || pip install "$@"
 }
 
 function pip_uninstall() {
   # uninstall 2 times
-  pip3 uninstall -y "$@" || pip3 uninstall -y "$@"
+  pip uninstall -y "$@" || pip uninstall -y "$@"
 }
 
 function get_exit_code() {
@@ -105,12 +104,32 @@ function get_bazel() {
   # version of Bazelisk to fetch the platform specific version of
   # Bazel to use from .bazelversion.
   retry curl --location --output tools/bazel \
-    https://raw.githubusercontent.com/bazelbuild/bazelisk/v1.23.0/bazelisk.py
+    https://raw.githubusercontent.com/bazelbuild/bazelisk/v1.16.0/bazelisk.py
   shasum --algorithm=1 --check \
-    <(echo '01df9cf7f08dd80d83979ed0d0666a99349ae93c  tools/bazel')
+    <(echo 'd4369c3d293814d3188019c9f7527a948972d9f8  tools/bazel')
   chmod u+x tools/bazel
 }
 
+# This function is bazel specific because of the bug
+# in the bazel that requires some special paths massaging
+# as a workaround. See
+# https://github.com/bazelbuild/bazel/issues/10167
+function install_sccache_nvcc_for_bazel() {
+  sudo mv /usr/local/cuda/bin/nvcc /usr/local/cuda/bin/nvcc-real
+
+  # Write the `/usr/local/cuda/bin/nvcc`
+  cat << EOF | sudo tee /usr/local/cuda/bin/nvcc
+#!/bin/sh
+if [ \$(env -u LD_PRELOAD ps -p \$PPID -o comm=) != sccache ]; then
+  exec sccache /usr/local/cuda/bin/nvcc "\$@"
+else
+  exec external/local_cuda/cuda/bin/nvcc-real "\$@"
+fi
+EOF
+
+  sudo chmod +x /usr/local/cuda/bin/nvcc
+}
+
 function install_monkeytype {
   # Install MonkeyType
   pip_install MonkeyType
@@ -160,7 +179,7 @@ function install_torchvision() {
 }
 
 function install_tlparse() {
-  pip_install --user "tlparse==0.3.30"
+  pip_install --user "tlparse==0.3.25"
   PATH="$(python -m site --user-base)/bin:$PATH"
 }
 
@@ -221,12 +240,6 @@ function checkout_install_torchbench() {
   popd
 }
 
-function install_torchao() {
-  local commit
-  commit=$(get_pinned_commit torchao)
-  pip_install --no-use-pep517 --user "git+https://github.com/pytorch/ao.git@${commit}"
-}
-
 function print_sccache_stats() {
   echo 'PyTorch Build Statistics'
   sccache --show-stats

.ci/pytorch/cpp_doc_push_script.sh

@@ -40,7 +40,7 @@ echo "Building PyTorch C++ API docs..."
 rm -rf cppdocs
 git clone https://github.com/pytorch/cppdocs
 
-set -ex -o pipefail
+set -ex
 
 # Generate ATen files
 pushd "${pt_checkout}"

.ci/pytorch/functorch_doc_push_script.sh

@@ -5,7 +5,7 @@ pt_checkout="/var/lib/jenkins/workspace"
 source "$pt_checkout/.ci/pytorch/common_utils.sh"
 echo "functorch_doc_push_script.sh: Invoked with $*"
 
-set -ex -o pipefail
+set -ex
 
 version=${DOCS_VERSION:-nightly}
 echo "version: $version"

.ci/pytorch/install_cache_xla.sh

@@ -6,7 +6,7 @@
 # return the same thing, ex checks for for rocm, CUDA, and changing the path
 # where sccache is installed, and not changing /etc/environment.
 
-set -ex -o pipefail
+set -ex
 
 install_binary() {
   echo "Downloading sccache binary from S3 repo"

.ci/pytorch/macos-test.sh

@@ -1,5 +1,4 @@
 #!/bin/bash
-set -x
 
 # shellcheck disable=SC2034
 # shellcheck source=./macos-common.sh
@@ -149,146 +148,9 @@ test_jit_hooks() {
   assert_git_not_dirty
 }
 
-torchbench_setup_macos() {
-  git clone --recursive https://github.com/pytorch/vision torchvision
-  git clone --recursive https://github.com/pytorch/audio torchaudio
-
-  pushd torchvision
-  git fetch
-  git checkout "$(cat ../.github/ci_commit_pins/vision.txt)"
-  git submodule update --init --recursive
-  python setup.py clean
-  python setup.py develop
-  popd
-
-  pushd torchaudio
-  git fetch
-  git checkout "$(cat ../.github/ci_commit_pins/audio.txt)"
-  git submodule update --init --recursive
-  python setup.py clean
-  python setup.py develop
-  popd
-
-  # Shellcheck doesn't like it when you pass no arguments to a function that can take args. See https://www.shellcheck.net/wiki/SC2120
-  # shellcheck disable=SC2119,SC2120
-  checkout_install_torchbench
-}
-
-conda_benchmark_deps() {
-  conda install -y astunparse numpy scipy ninja pyyaml setuptools cmake typing-extensions requests protobuf numba cython scikit-learn
-  conda install -y -c conda-forge librosa
-}
-
-
-test_torchbench_perf() {
-  print_cmake_info
-
-  echo "Launching torchbench setup"
-  conda_benchmark_deps
-  torchbench_setup_macos
-
-  TEST_REPORTS_DIR=$(pwd)/test/test-reports
-  mkdir -p "$TEST_REPORTS_DIR"
-
-  local backend=eager
-  local dtype=notset
-  local device=mps
-
-  echo "Setup complete, launching torchbench training performance run"
-  PYTHONPATH="$(pwd)"/torchbench python benchmarks/dynamo/torchbench.py \
-    --performance --backend "$backend" --training --devices "$device" \
-    --output "$TEST_REPORTS_DIR/inductor_${backend}_torchbench_${dtype}_training_${device}_performance.csv"
-
-  echo "Launching torchbench inference performance run"
-  PYTHONPATH="$(pwd)"/torchbench python benchmarks/dynamo/torchbench.py \
-    --performance --backend "$backend" --inference --devices "$device" \
-    --output "$TEST_REPORTS_DIR/inductor_${backend}_torchbench_${dtype}_inference_${device}_performance.csv"
-
-  echo "Pytorch benchmark on mps device completed"
-}
-
-test_torchbench_smoketest() {
-  print_cmake_info
-
-  echo "Launching torchbench setup"
-  conda_benchmark_deps
-  # shellcheck disable=SC2119,SC2120
-  torchbench_setup_macos
-
-  TEST_REPORTS_DIR=$(pwd)/test/test-reports
-  mkdir -p "$TEST_REPORTS_DIR"
-
-  local backend=eager
-  local dtype=notset
-  local device=mps
-
-  touch "$TEST_REPORTS_DIR/inductor_${backend}_torchbench_${dtype}_training_${device}_performance.csv"
-  touch "$TEST_REPORTS_DIR/inductor_${backend}_torchbench_${dtype}_inference_${device}_performance.csv"
-
-  echo "Setup complete, launching torchbench training performance run"
-  for model in hf_T5 llama BERT_pytorch dcgan hf_GPT2 yolov3 resnet152; do
-    PYTHONPATH="$(pwd)"/torchbench python benchmarks/dynamo/torchbench.py \
-      --performance --only "$model" --backend "$backend" --training --devices "$device" \
-      --output "$TEST_REPORTS_DIR/inductor_${backend}_torchbench_${dtype}_training_${device}_performance.csv"
-  done
-
-  echo "Launching torchbench inference performance run"
-  for model in hf_T5 llama BERT_pytorch dcgan hf_GPT2 yolov3 resnet152; do
-    PYTHONPATH="$(pwd)"/torchbench python benchmarks/dynamo/torchbench.py \
-      --performance --only "$model" --backend "$backend" --inference --devices "$device" \
-      --output "$TEST_REPORTS_DIR/inductor_${backend}_torchbench_${dtype}_inference_${device}_performance.csv"
-  done
-
-  echo "Pytorch benchmark on mps device completed"
-}
-
-test_hf_perf() {
-  print_cmake_info
-  TEST_REPORTS_DIR=$(pwd)/test/test-reports
-  mkdir -p "$TEST_REPORTS_DIR"
-  conda_benchmark_deps
-  torchbench_setup_macos
-
-  echo "Launching HuggingFace training perf run"
-  python "$(pwd)"/benchmarks/dynamo/huggingface.py --backend eager --device mps --performance --training --output="${TEST_REPORTS_DIR}"/hf_training.csv
-
-  echo "Launching HuggingFace inference perf run"
-  python "$(pwd)"/benchmarks/dynamo/huggingface.py --backend eager --device mps --performance --training --output="${TEST_REPORTS_DIR}"/hf_inference.csv
-
-  echo "HuggingFace benchmark on mps device completed"
-}
-
-test_timm_perf() {
-  print_cmake_info
-  TEST_REPORTS_DIR=$(pwd)/test/test-reports
-  mkdir -p "$TEST_REPORTS_DIR"
-  conda_benchmark_deps
-  torchbench_setup_macos
-
-  echo "Launching timm training perf run"
-  python "$(pwd)"/benchmarks/dynamo/timm_models.py --backend eager --device mps --performance --training --output="${TEST_REPORTS_DIR}"/timm_training.csv
-
-  echo "Launching timm inference perf run"
-  python "$(pwd)"/benchmarks/dynamo/timm_models.py --backend eager --device mps --performance --training --output="${TEST_REPORTS_DIR}"/timm_inference.csv
-
-  echo "timm benchmark on mps device completed"
-}
-
 install_tlparse
 
-if [[ $TEST_CONFIG == *"perf_all"* ]]; then
-  test_torchbench_perf
-  test_hf_perf
-  test_timm_perf
-elif [[ $TEST_CONFIG == *"perf_torchbench"* ]]; then
-  test_torchbench_perf
-elif [[ $TEST_CONFIG == *"perf_hf"* ]]; then
-  test_hf_perf
-elif [[ $TEST_CONFIG == *"perf_timm"* ]]; then
-  test_timm_perf
-elif [[ $TEST_CONFIG == *"perf_smoketest"* ]]; then
-  test_torchbench_smoketest
-elif [[ $NUM_TEST_SHARDS -gt 1 ]]; then
+if [[ $NUM_TEST_SHARDS -gt 1 ]]; then
   test_python_shard "${SHARD_NUMBER}"
   if [[ "${SHARD_NUMBER}" == 1 ]]; then
     test_libtorch

.ci/pytorch/multigpu-test.sh

@@ -8,62 +8,55 @@
 source "$(dirname "${BASH_SOURCE[0]}")/common.sh"
 
 echo "Testing pytorch"
-# When adding more tests, please use HUD to see which shard is shorter
-if [[ "${SHARD_NUMBER:-1}" == "1" ]]; then
-    # FSDP tests
-    for f in test/distributed/fsdp/*.py ; do time python test/run_test.py --verbose -i "${f#*/}" ; done
-fi
+time python test/run_test.py --include test_cuda_multigpu test_cuda_primary_ctx --verbose
 
-if [[ "${SHARD_NUMBER:-2}" == "2" ]]; then
-    time python test/run_test.py --include test_cuda_multigpu test_cuda_primary_ctx --verbose
+# Disabling tests to see if they solve timeout issues; see https://github.com/pytorch/pytorch/issues/70015
+# python tools/download_mnist.py --quiet -d test/cpp/api/mnist
+# OMP_NUM_THREADS=2 TORCH_CPP_TEST_MNIST_PATH="test/cpp/api/mnist" build/bin/test_api
+time python test/run_test.py --verbose -i distributed/test_c10d_common
+time python test/run_test.py --verbose -i distributed/test_c10d_gloo
+time python test/run_test.py --verbose -i distributed/test_c10d_nccl
+time python test/run_test.py --verbose -i distributed/test_c10d_spawn_gloo
+time python test/run_test.py --verbose -i distributed/test_c10d_spawn_nccl
+time python test/run_test.py --verbose -i distributed/test_compute_comm_reordering
+time python test/run_test.py --verbose -i distributed/test_store
+time python test/run_test.py --verbose -i distributed/test_symmetric_memory
+time python test/run_test.py --verbose -i distributed/test_pg_wrapper
+time python test/run_test.py --verbose -i distributed/rpc/cuda/test_tensorpipe_agent
+# FSDP tests
+for f in test/distributed/fsdp/*.py ; do time python test/run_test.py --verbose -i "${f#*/}" ; done
+# ShardedTensor tests
+time python test/run_test.py --verbose -i distributed/checkpoint/test_checkpoint
+time python test/run_test.py --verbose -i distributed/checkpoint/test_file_system_checkpoint
+time python test/run_test.py --verbose -i distributed/_shard/sharding_spec/test_sharding_spec
+time python test/run_test.py --verbose -i distributed/_shard/sharding_plan/test_sharding_plan
+time python test/run_test.py --verbose -i distributed/_shard/sharded_tensor/test_sharded_tensor
+time python test/run_test.py --verbose -i distributed/_shard/sharded_tensor/test_sharded_tensor_reshard
 
-    # Disabling tests to see if they solve timeout issues; see https://github.com/pytorch/pytorch/issues/70015
-    # python tools/download_mnist.py --quiet -d test/cpp/api/mnist
-    # OMP_NUM_THREADS=2 TORCH_CPP_TEST_MNIST_PATH="test/cpp/api/mnist" build/bin/test_api
-    time python test/run_test.py --verbose -i distributed/test_c10d_common
-    time python test/run_test.py --verbose -i distributed/test_c10d_gloo
-    time python test/run_test.py --verbose -i distributed/test_c10d_nccl
-    time python test/run_test.py --verbose -i distributed/test_c10d_spawn_gloo
-    time python test/run_test.py --verbose -i distributed/test_c10d_spawn_nccl
-    time python test/run_test.py --verbose -i distributed/test_compute_comm_reordering
-    time python test/run_test.py --verbose -i distributed/test_store
-    time python test/run_test.py --verbose -i distributed/test_symmetric_memory
-    time python test/run_test.py --verbose -i distributed/test_pg_wrapper
-    time python test/run_test.py --verbose -i distributed/rpc/cuda/test_tensorpipe_agent
+# functional collective tests
+time python test/run_test.py --verbose -i distributed/test_functional_api
 
-    # ShardedTensor tests
-    time python test/run_test.py --verbose -i distributed/checkpoint/test_checkpoint
-    time python test/run_test.py --verbose -i distributed/checkpoint/test_file_system_checkpoint
-    time python test/run_test.py --verbose -i distributed/_shard/sharding_spec/test_sharding_spec
-    time python test/run_test.py --verbose -i distributed/_shard/sharding_plan/test_sharding_plan
-    time python test/run_test.py --verbose -i distributed/_shard/sharded_tensor/test_sharded_tensor
-    time python test/run_test.py --verbose -i distributed/_shard/sharded_tensor/test_sharded_tensor_reshard
+# DTensor tests
+time python test/run_test.py --verbose -i distributed/_tensor/test_random_ops
+time python test/run_test.py --verbose -i distributed/_tensor/test_dtensor_compile
 
-    # functional collective tests
-    time python test/run_test.py --verbose -i distributed/test_functional_api
+# DeviceMesh test
+time python test/run_test.py --verbose -i distributed/test_device_mesh
 
-    # DTensor tests
-    time python test/run_test.py --verbose -i distributed/tensor/test_random_ops
-    time python test/run_test.py --verbose -i distributed/tensor/test_dtensor_compile
+# DTensor/TP tests
+time python test/run_test.py --verbose -i distributed/tensor/parallel/test_tp_examples
+time python test/run_test.py --verbose -i distributed/tensor/parallel/test_tp_random_state
 
-    # DeviceMesh test
-    time python test/run_test.py --verbose -i distributed/test_device_mesh
+# FSDP2 tests
+time python test/run_test.py --verbose -i distributed/_composable/fsdp/test_fully_shard_training -- -k test_2d_mlp_with_nd_mesh
 
-    # DTensor/TP tests
-    time python test/run_test.py --verbose -i distributed/tensor/parallel/test_tp_examples
-    time python test/run_test.py --verbose -i distributed/tensor/parallel/test_tp_random_state
+# ND composability tests
+time python test/run_test.py --verbose -i distributed/_composable/test_composability/test_2d_composability
+time python test/run_test.py --verbose -i distributed/_composable/test_composability/test_pp_composability
 
-    # FSDP2 tests
-    time python test/run_test.py --verbose -i distributed/_composable/fsdp/test_fully_shard_training -- -k test_2d_mlp_with_nd_mesh
-
-    # ND composability tests
-    time python test/run_test.py --verbose -i distributed/_composable/test_composability/test_2d_composability
-    time python test/run_test.py --verbose -i distributed/_composable/test_composability/test_pp_composability
-
-    # Other tests
-    time python test/run_test.py --verbose -i test_cuda_primary_ctx
-    time python test/run_test.py --verbose -i test_optim -- -k test_forloop_goes_right_direction_multigpu
-    time python test/run_test.py --verbose -i test_optim -- -k test_mixed_device_dtype
-    time python test/run_test.py --verbose -i test_foreach -- -k test_tensors_grouping
-fi
+# Other tests
+time python test/run_test.py --verbose -i test_cuda_primary_ctx
+time python test/run_test.py --verbose -i test_optim -- -k test_forloop_goes_right_direction_multigpu
+time python test/run_test.py --verbose -i test_optim -- -k test_mixed_device_dtype
+time python test/run_test.py --verbose -i test_foreach -- -k test_tensors_grouping
 assert_git_not_dirty

.ci/pytorch/python_doc_push_script.sh

@@ -7,7 +7,7 @@ source "$pt_checkout/.ci/pytorch/common_utils.sh"
 
 echo "python_doc_push_script.sh: Invoked with $*"
 
-set -ex -o pipefail
+set -ex
 
 # for statements like ${1:-${DOCS_INSTALL_PATH:-docs/}}
 # the order of operations goes:
@@ -63,7 +63,7 @@ build_docs () {
     echo "(tried to echo the WARNINGS above the ==== line)"
     echo =========================
   fi
-  set -ex -o pipefail
+  set -ex
   return $code
 }
 

.ci/pytorch/run_tests.sh

@@ -1,436 +0,0 @@
-#!/bin/bash
-# shellcheck disable=SC2086,SC2048,SC2068,SC2145,SC2034,SC2207,SC2143
-# TODO: Re-enable shellchecks above
-
-set -eux -o pipefail
-
-# Essentially runs pytorch/test/run_test.py, but keeps track of which tests to
-# skip in a centralized place.
-#
-# TODO Except for a few tests, this entire file is a giant TODO. Why are these
-# tests # failing?
-# TODO deal with Windows
-
-# This script expects to be in the pytorch root folder
-if [[ ! -d 'test' || ! -f 'test/run_test.py' ]]; then
-    echo "run_tests.sh expects to be run from the Pytorch root directory " \
-         "but I'm actually in $(pwd)"
-    exit 2
-fi
-
-# Allow master skip of all tests
-if [[ -n "${SKIP_ALL_TESTS:-}" ]]; then
-    exit 0
-fi
-
-# If given specific test params then just run those
-if [[ -n "${RUN_TEST_PARAMS:-}" ]]; then
-    echo "$(date) :: Calling user-command $(pwd)/test/run_test.py ${RUN_TEST_PARAMS[@]}"
-    python test/run_test.py ${RUN_TEST_PARAMS[@]}
-    exit 0
-fi
-
-# Function to retry functions that sometimes timeout or have flaky failures
-retry () {
-    $*  || (sleep 1 && $*) || (sleep 2 && $*) || (sleep 4 && $*) || (sleep 8 && $*)
-}
-
-# Parameters
-##############################################################################
-if [[ "$#" != 3 ]]; then
-  if [[ -z "${DESIRED_PYTHON:-}" || -z "${DESIRED_CUDA:-}" || -z "${PACKAGE_TYPE:-}" ]]; then
-    echo "USAGE: run_tests.sh  PACKAGE_TYPE  DESIRED_PYTHON  DESIRED_CUDA"
-    echo "The env variable PACKAGE_TYPE must be set to 'conda' or 'manywheel' or 'libtorch'"
-    echo "The env variable DESIRED_PYTHON must be set like '2.7mu' or '3.6m' etc"
-    echo "The env variable DESIRED_CUDA must be set like 'cpu' or 'cu80' etc"
-    exit 1
-  fi
-  package_type="$PACKAGE_TYPE"
-  py_ver="$DESIRED_PYTHON"
-  cuda_ver="$DESIRED_CUDA"
-else
-  package_type="$1"
-  py_ver="$2"
-  cuda_ver="$3"
-fi
-
-if [[ "$cuda_ver" == 'cpu-cxx11-abi' ]]; then
-    cuda_ver="cpu"
-fi
-
-# cu80, cu90, cu100, cpu
-if [[ ${#cuda_ver} -eq 4 ]]; then
-    cuda_ver_majmin="${cuda_ver:2:1}.${cuda_ver:3:1}"
-elif [[ ${#cuda_ver} -eq 5 ]]; then
-    cuda_ver_majmin="${cuda_ver:2:2}.${cuda_ver:4:1}"
-fi
-
-NUMPY_PACKAGE=""
-if [[ ${py_ver} == "3.10" ]]; then
-    PROTOBUF_PACKAGE="protobuf>=3.17.2"
-    NUMPY_PACKAGE="numpy>=1.21.2"
-else
-    PROTOBUF_PACKAGE="protobuf=3.14.0"
-fi
-
-# Environment initialization
-if [[ "$(uname)" == Darwin ]]; then
-    # Install the testing dependencies
-    retry conda install -yq future hypothesis ${NUMPY_PACKAGE} ${PROTOBUF_PACKAGE} pytest setuptools six typing_extensions pyyaml
-else
-    retry pip install -qr requirements.txt || true
-    retry pip install -q hypothesis protobuf pytest setuptools || true
-    numpy_ver=1.15
-    case "$(python --version 2>&1)" in
-      *2* | *3.5* | *3.6*)
-        numpy_ver=1.11
-        ;;
-    esac
-    retry pip install -q "numpy==${numpy_ver}" || true
-fi
-
-echo "Testing with:"
-pip freeze
-conda list || true
-
-##############################################################################
-# Smoke tests
-##############################################################################
-# TODO use check_binary.sh, which requires making sure it runs on Windows
-pushd /
-echo "Smoke testing imports"
-python -c 'import torch'
-
-# Test that MKL is there
-if [[ "$(uname)" == 'Darwin' && "$package_type" == *wheel ]]; then
-    echo 'Not checking for MKL on Darwin wheel packages'
-else
-    echo "Checking that MKL is available"
-    python -c 'import torch; exit(0 if torch.backends.mkl.is_available() else 1)'
-fi
-
-if [[ "$OSTYPE" == "msys" ]]; then
-    GPUS=$(wmic path win32_VideoController get name)
-    if [[ ! "$GPUS" == *NVIDIA* ]]; then
-        echo "Skip CUDA tests for machines without a Nvidia GPU card"
-        exit 0
-    fi
-fi
-
-# Test that the version number is consistent during building and testing
-if [[ "$PYTORCH_BUILD_NUMBER" -gt 1 ]]; then
-    expected_version="${PYTORCH_BUILD_VERSION}.post${PYTORCH_BUILD_NUMBER}"
-else
-    expected_version="${PYTORCH_BUILD_VERSION}"
-fi
-echo "Checking that we are testing the package that is just built"
-python -c "import torch; exit(0 if torch.__version__ == '$expected_version' else 1)"
-
-# Test that CUDA builds are setup correctly
-if [[ "$cuda_ver" != 'cpu' ]]; then
-    cuda_installed=1
-    nvidia-smi || cuda_installed=0
-    if [[ "$cuda_installed" == 0 ]]; then
-      echo "Skip CUDA tests for machines without a Nvidia GPU card"
-    else
-      # Test CUDA archs
-      echo "Checking that CUDA archs are setup correctly"
-      timeout 20 python -c 'import torch; torch.randn([3,5]).cuda()'
-
-      # These have to run after CUDA is initialized
-      echo "Checking that magma is available"
-      python -c 'import torch; torch.rand(1).cuda(); exit(0 if torch.cuda.has_magma else 1)'
-      echo "Checking that CuDNN is available"
-      python -c 'import torch; exit(0 if torch.backends.cudnn.is_available() else 1)'
-    fi
-fi
-
-# Check that OpenBlas is not linked to on MacOS
-if [[ "$(uname)" == 'Darwin' ]]; then
-    echo "Checking the OpenBLAS is not linked to"
-    all_dylibs=($(find "$(python -c "import site; print(site.getsitepackages()[0])")"/torch -name '*.dylib'))
-    for dylib in "${all_dylibs[@]}"; do
-        if [[ -n "$(otool -L $dylib | grep -i openblas)" ]]; then
-            echo "Found openblas as a dependency of $dylib"
-            echo "Full dependencies is: $(otool -L $dylib)"
-            exit 1
-        fi
-    done
-
-    echo "Checking that OpenMP is available"
-    python -c "import torch; exit(0 if torch.backends.openmp.is_available() else 1)"
-fi
-
-popd
-
-# TODO re-enable the other tests after the nightlies are moved to CI. This is
-# because the binaries keep breaking, often from additional tests, that aren't
-# real problems. Once these are on circleci and a smoke-binary-build is added
-# to PRs then this should stop happening and these can be re-enabled.
-echo "Not running unit tests. Hopefully these problems are caught by CI"
-exit 0
-
-
-##############################################################################
-# Running unit tests (except not right now)
-##############################################################################
-echo "$(date) :: Starting tests for $package_type package for python$py_ver and $cuda_ver"
-
-# We keep track of exact tests to skip, as otherwise we would be hardly running
-# any tests. But b/c of issues working with pytest/normal-python-test/ and b/c
-# of special snowflake tests in test/run_test.py we also take special care of
-# those
-tests_to_skip=()
-
-#
-# Entire file exclusions
-##############################################################################
-entire_file_exclusions=("-x")
-
-# cpp_extensions doesn't work with pytest, so we exclude it from the pytest run
-# here and then manually run it later. Note that this is only because this
-# entire_fil_exclusions flag is only passed to the pytest run
-entire_file_exclusions+=("cpp_extensions")
-
-# TODO temporary line to fix next days nightlies, but should be removed when
-# issue is fixed
-entire_file_exclusions+=('type_info')
-
-if [[ "$cuda_ver" == 'cpu' ]]; then
-    # test/test_cuda.py exits early if the installed torch is not built with
-    # CUDA, but the exit doesn't work when running with pytest, so pytest will
-    # still try to run all the CUDA tests and then fail
-    entire_file_exclusions+=("cuda")
-    entire_file_exclusions+=("nccl")
-fi
-
-if [[ "$(uname)" == 'Darwin' || "$OSTYPE" == "msys" ]]; then
-    # pytest on Mac doesn't like the exits in these files
-    entire_file_exclusions+=('c10d')
-    entire_file_exclusions+=('distributed')
-
-    # pytest doesn't mind the exit but fails the tests. On Mac we run this
-    # later without pytest
-    entire_file_exclusions+=('thd_distributed')
-fi
-
-
-#
-# Universal flaky tests
-##############################################################################
-
-# RendezvousEnvTest sometimes hangs forever
-# Otherwise it will fail on CUDA with
-#   Traceback (most recent call last):
-#     File "test_c10d.py", line 179, in test_common_errors
-#       next(gen)
-#   AssertionError: ValueError not raised
-tests_to_skip+=('RendezvousEnvTest and test_common_errors')
-
-# This hung forever once on conda_3.5_cu92
-tests_to_skip+=('TestTorch and test_sum_dim')
-
-# test_trace_warn isn't actually flaky, but it doesn't work with pytest so we
-# just skip it
-tests_to_skip+=('TestJit and test_trace_warn')
-#
-# Python specific flaky tests
-##############################################################################
-
-# test_dataloader.py:721: AssertionError
-# looks like a timeout, but interestingly only appears on python 3
-if [[ "$py_ver" == 3* ]]; then
-    tests_to_skip+=('TestDataLoader and test_proper_exit')
-fi
-
-#
-# CUDA flaky tests, all package types
-##############################################################################
-if [[ "$cuda_ver" != 'cpu' ]]; then
-
-    #
-    # DistributedDataParallelTest
-    # All of these seem to fail
-    tests_to_skip+=('DistributedDataParallelTest')
-
-    #
-    # RendezvousEnvTest
-    # Traceback (most recent call last):
-    #   File "test_c10d.py", line 201, in test_nominal
-    #     store0, rank0, size0 = next(gen0)
-    #   File "/opt/python/cp36-cp36m/lib/python3.6/site-packages/torch/distributed/rendezvous.py", line 131, in _env_rendezvous_handler
-    #     store = TCPStore(master_addr, master_port, start_daemon)
-    # RuntimeError: Address already in use
-    tests_to_skip+=('RendezvousEnvTest and test_nominal')
-
-    #
-    # TestCppExtension
-    #
-    # Traceback (most recent call last):
-    #   File "test_cpp_extensions.py", line 134, in test_jit_cudnn_extension
-    #     with_cuda=True)
-    #   File "/opt/python/cp35-cp35m/lib/python3.5/site-packages/torch/utils/cpp_extension.py", line 552, in load
-    #     with_cuda)
-    #   File "/opt/python/cp35-cp35m/lib/python3.5/site-packages/torch/utils/cpp_extension.py", line 729, in _jit_compile
-    #     return _import_module_from_library(name, build_directory)
-    #   File "/opt/python/cp35-cp35m/lib/python3.5/site-packages/torch/utils/cpp_extension.py", line 867, in _import_module_from_library
-    #     return imp.load_module(module_name, file, path, description)
-    #   File "/opt/python/cp35-cp35m/lib/python3.5/imp.py", line 243, in load_module
-    #     return load_dynamic(name, filename, file)
-    #   File "/opt/python/cp35-cp35m/lib/python3.5/imp.py", line 343, in load_dynamic
-    #     return _load(spec)
-    #   File "<frozen importlib._bootstrap>", line 693, in _load
-    #   File "<frozen importlib._bootstrap>", line 666, in _load_unlocked
-    #   File "<frozen importlib._bootstrap>", line 577, in module_from_spec
-    #   File "<frozen importlib._bootstrap_external>", line 938, in create_module
-    #   File "<frozen importlib._bootstrap>", line 222, in _call_with_frames_removed
-    # ImportError: libcudnn.so.7: cannot open shared object file: No such file or directory
-    tests_to_skip+=('TestCppExtension and test_jit_cudnn_extension')
-
-    #
-    # TestCuda
-    #
-
-    # 3.7_cu80
-    #  RuntimeError: CUDA error: out of memory
-    tests_to_skip+=('TestCuda and test_arithmetic_large_tensor')
-
-    # 3.7_cu80
-    # RuntimeError: cuda runtime error (2) : out of memory at /opt/conda/conda-bld/pytorch-nightly_1538097262541/work/aten/src/THC/THCTensorCopy.cu:205
-    tests_to_skip+=('TestCuda and test_autogpu')
-
-    #
-    # TestDistBackend
-    #
-
-    # Traceback (most recent call last):
-    #   File "test_thd_distributed.py", line 1046, in wrapper
-    #     self._join_and_reduce(fn)
-    #   File "test_thd_distributed.py", line 1108, in _join_and_reduce
-    #     self.assertEqual(p.exitcode, first_process.exitcode)
-    #   File "/pytorch/test/common.py", line 399, in assertEqual
-    #     super(TestCase, self).assertEqual(x, y, message)
-    # AssertionError: None != 77 :
-    tests_to_skip+=('TestDistBackend and test_all_gather_group')
-    tests_to_skip+=('TestDistBackend and test_all_reduce_group_max')
-    tests_to_skip+=('TestDistBackend and test_all_reduce_group_min')
-    tests_to_skip+=('TestDistBackend and test_all_reduce_group_sum')
-    tests_to_skip+=('TestDistBackend and test_all_reduce_group_product')
-    tests_to_skip+=('TestDistBackend and test_barrier_group')
-    tests_to_skip+=('TestDistBackend and test_broadcast_group')
-
-    # Traceback (most recent call last):
-    #   File "test_thd_distributed.py", line 1046, in wrapper
-    #     self._join_and_reduce(fn)
-    #   File "test_thd_distributed.py", line 1108, in _join_and_reduce
-    #     self.assertEqual(p.exitcode, first_process.exitcode)
-    #   File "/pytorch/test/common.py", line 397, in assertEqual
-    #     super(TestCase, self).assertLessEqual(abs(x - y), prec, message)
-    # AssertionError: 12 not less than or equal to 1e-05
-    tests_to_skip+=('TestDistBackend and test_barrier')
-
-    # Traceback (most recent call last):
-    #   File "test_distributed.py", line 1267, in wrapper
-    #     self._join_and_reduce(fn)
-    #   File "test_distributed.py", line 1350, in _join_and_reduce
-    #     self.assertEqual(p.exitcode, first_process.exitcode)
-    #   File "/pytorch/test/common.py", line 399, in assertEqual
-    #     super(TestCase, self).assertEqual(x, y, message)
-    # AssertionError: None != 1
-    tests_to_skip+=('TestDistBackend and test_broadcast')
-
-    # Memory leak very similar to all the conda ones below, but appears on manywheel
-    # 3.6m_cu80
-    # AssertionError: 1605632 not less than or equal to 1e-05 : __main__.TestEndToEndHybridFrontendModels.test_vae_cuda leaked 1605632 bytes CUDA memory on device 0
-    tests_to_skip+=('TestEndToEndHybridFrontendModels and test_vae_cuda')
-
-    # ________________________ TestNN.test_embedding_bag_cuda ________________________
-    #
-    # self = <test_nn.TestNN testMethod=test_embedding_bag_cuda>
-    # dtype = torch.float32
-    #
-    #     @unittest.skipIf(not TEST_CUDA, "CUDA unavailable")
-    #     @repeat_test_for_types(ALL_TENSORTYPES)
-    #     @skipIfRocm
-    #     def test_embedding_bag_cuda(self, dtype=torch.float):
-    #         self._test_EmbeddingBag(True, 'sum', False, dtype)
-    #         self._test_EmbeddingBag(True, 'mean', False, dtype)
-    #         self._test_EmbeddingBag(True, 'max', False, dtype)
-    #         if dtype != torch.half:
-    #             # torch.cuda.sparse.HalfTensor is not enabled.
-    #             self._test_EmbeddingBag(True, 'sum', True, dtype)
-    # >           self._test_EmbeddingBag(True, 'mean', True, dtype)
-    #
-    # test_nn.py:2144:
-    # _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
-    # test_nn.py:2062: in _test_EmbeddingBag
-    #     _test_vs_Embedding(N, D, B, L)
-    # test_nn.py:2059: in _test_vs_Embedding
-    #     self.assertEqual(es_weight_grad, e.weight.grad, needed_prec)
-    # common.py:373: in assertEqual
-    #     assertTensorsEqual(x, y)
-    # common.py:365: in assertTensorsEqual
-    #     self.assertLessEqual(max_err, prec, message)
-    # E   AssertionError: tensor(0.0000, device='cuda:0', dtype=torch.float32) not less than or equal to 2e-05 :
-    #  1 failed, 1202 passed, 19 skipped, 2 xfailed, 796 warnings in 1166.73 seconds =
-    # Traceback (most recent call last):
-    #   File "test/run_test.py", line 391, in <module>
-    #     main()
-    #   File "test/run_test.py", line 383, in main
-    #     raise RuntimeError(message)
-    tests_to_skip+=('TestNN and test_embedding_bag_cuda')
-fi
-
-##############################################################################
-# MacOS specific flaky tests
-##############################################################################
-
-if [[ "$(uname)" == 'Darwin' ]]; then
-    # TestCppExtensions by default uses a temp folder in /tmp. This doesn't
-    # work for this Mac machine cause there is only one machine and /tmp is
-    # shared. (All the linux builds are on docker so have their own /tmp).
-    tests_to_skip+=('TestCppExtension')
-fi
-
-# Turn the set of tests to skip into an invocation that pytest understands
-excluded_tests_logic=''
-for exclusion in "${tests_to_skip[@]}"; do
-    if [[ -z "$excluded_tests_logic" ]]; then
-        # Only true for i==0
-        excluded_tests_logic="not ($exclusion)"
-    else
-        excluded_tests_logic="$excluded_tests_logic and not ($exclusion)"
-    fi
-done
-
-
-##############################################################################
-# Run the tests
-##############################################################################
-echo
-echo "$(date) :: Calling 'python test/run_test.py -v -p pytest ${entire_file_exclusions[@]} -- --disable-pytest-warnings -k '$excluded_tests_logic'"
-
-python test/run_test.py -v -p pytest ${entire_file_exclusions[@]} -- --disable-pytest-warnings -k "'" "$excluded_tests_logic" "'"
-
-echo
-echo "$(date) :: Finished 'python test/run_test.py -v -p pytest ${entire_file_exclusions[@]} -- --disable-pytest-warnings -k '$excluded_tests_logic'"
-
-# cpp_extensions don't work with pytest, so we run them without pytest here,
-# except there's a failure on CUDA builds (documented above), and
-# cpp_extensions doesn't work on a shared mac machine (also documented above)
-if [[ "$cuda_ver" == 'cpu' && "$(uname)" != 'Darwin' ]]; then
-    echo
-    echo "$(date) :: Calling 'python test/run_test.py -v -i cpp_extensions'"
-    python test/run_test.py -v -i cpp_extensions
-    echo
-    echo "$(date) :: Finished 'python test/run_test.py -v -i cpp_extensions'"
-fi
-
-# thd_distributed can run on Mac but not in pytest
-if [[ "$(uname)" == 'Darwin' ]]; then
-    echo
-    echo "$(date) :: Calling 'python test/run_test.py -v -i thd_distributed'"
-    python test/run_test.py -v -i thd_distributed
-    echo
-    echo "$(date) :: Finished 'python test/run_test.py -v -i thd_distributed'"
-fi

.ci/pytorch/smoke_test/check_binary_symbols.py

@@ -1,130 +0,0 @@
-#!/usr/bin/env python3
-import concurrent.futures
-import distutils.sysconfig
-import functools
-import itertools
-import os
-import re
-from pathlib import Path
-from typing import Any, List, Tuple
-
-
-# We also check that there are [not] cxx11 symbols in libtorch
-#
-# To check whether it is using cxx11 ABI, check non-existence of symbol:
-PRE_CXX11_SYMBOLS = (
-    "std::basic_string<",
-    "std::list",
-)
-# To check whether it is using pre-cxx11 ABI, check non-existence of symbol:
-CXX11_SYMBOLS = (
-    "std::__cxx11::basic_string",
-    "std::__cxx11::list",
-)
-# NOTE: Checking the above symbols in all namespaces doesn't work, because
-# devtoolset7 always produces some cxx11 symbols even if we build with old ABI,
-# and CuDNN always has pre-cxx11 symbols even if we build with new ABI using gcc 5.4.
-# Instead, we *only* check the above symbols in the following namespaces:
-LIBTORCH_NAMESPACE_LIST = (
-    "c10::",
-    "at::",
-    "caffe2::",
-    "torch::",
-)
-
-
-def _apply_libtorch_symbols(symbols):
-    return [
-        re.compile(f"{x}.*{y}")
-        for (x, y) in itertools.product(LIBTORCH_NAMESPACE_LIST, symbols)
-    ]
-
-
-LIBTORCH_CXX11_PATTERNS = _apply_libtorch_symbols(CXX11_SYMBOLS)
-
-LIBTORCH_PRE_CXX11_PATTERNS = _apply_libtorch_symbols(PRE_CXX11_SYMBOLS)
-
-
-@functools.lru_cache(100)
-def get_symbols(lib: str) -> List[Tuple[str, str, str]]:
-    from subprocess import check_output
-
-    lines = check_output(f'nm "{lib}"|c++filt', shell=True)
-    return [x.split(" ", 2) for x in lines.decode("latin1").split("\n")[:-1]]
-
-
-def grep_symbols(lib: str, patterns: List[Any]) -> List[str]:
-    def _grep_symbols(
-        symbols: List[Tuple[str, str, str]], patterns: List[Any]
-    ) -> List[str]:
-        rc = []
-        for _s_addr, _s_type, s_name in symbols:
-            for pattern in patterns:
-                if pattern.match(s_name):
-                    rc.append(s_name)
-                    continue
-        return rc
-
-    all_symbols = get_symbols(lib)
-    num_workers = 32
-    chunk_size = (len(all_symbols) + num_workers - 1) // num_workers
-
-    def _get_symbols_chunk(i):
-        return all_symbols[i * chunk_size : (i + 1) * chunk_size]
-
-    with concurrent.futures.ThreadPoolExecutor(max_workers=32) as executor:
-        tasks = [
-            executor.submit(_grep_symbols, _get_symbols_chunk(i), patterns)
-            for i in range(num_workers)
-        ]
-        return functools.reduce(list.__add__, (x.result() for x in tasks), [])
-
-
-def check_lib_symbols_for_abi_correctness(lib: str, pre_cxx11_abi: bool = True) -> None:
-    print(f"lib: {lib}")
-    cxx11_symbols = grep_symbols(lib, LIBTORCH_CXX11_PATTERNS)
-    pre_cxx11_symbols = grep_symbols(lib, LIBTORCH_PRE_CXX11_PATTERNS)
-    num_cxx11_symbols = len(cxx11_symbols)
-    num_pre_cxx11_symbols = len(pre_cxx11_symbols)
-    print(f"num_cxx11_symbols: {num_cxx11_symbols}")
-    print(f"num_pre_cxx11_symbols: {num_pre_cxx11_symbols}")
-    if pre_cxx11_abi:
-        if num_cxx11_symbols > 0:
-            raise RuntimeError(
-                f"Found cxx11 symbols, but there shouldn't be any, see: {cxx11_symbols[:100]}"
-            )
-        if num_pre_cxx11_symbols < 1000:
-            raise RuntimeError("Didn't find enough pre-cxx11 symbols.")
-        # Check for no recursive iterators, regression test for https://github.com/pytorch/pytorch/issues/133437
-        rec_iter_symbols = grep_symbols(
-            lib, [re.compile("std::filesystem::recursive_directory_iterator.*")]
-        )
-        if len(rec_iter_symbols) > 0:
-            raise RuntimeError(
-                f"recursive_directory_iterator in used pre-CXX11 binaries, see; {rec_iter_symbols}"
-            )
-    else:
-        if num_pre_cxx11_symbols > 0:
-            raise RuntimeError(
-                f"Found pre-cxx11 symbols, but there shouldn't be any, see: {pre_cxx11_symbols[:100]}"
-            )
-        if num_cxx11_symbols < 100:
-            raise RuntimeError("Didn't find enought cxx11 symbols")
-
-
-def main() -> None:
-    if "install_root" in os.environ:
-        install_root = Path(os.getenv("install_root"))  # noqa: SIM112
-    else:
-        if os.getenv("PACKAGE_TYPE") == "libtorch":
-            install_root = Path(os.getcwd())
-        else:
-            install_root = Path(distutils.sysconfig.get_python_lib()) / "torch"
-
-    libtorch_cpu_path = install_root / "lib" / "libtorch_cpu.so"
-    pre_cxx11_abi = "cxx11-abi" not in os.getenv("DESIRED_DEVTOOLSET", "")
-    check_lib_symbols_for_abi_correctness(libtorch_cpu_path, pre_cxx11_abi)
-
-
-if __name__ == "__main__":
-    main()

.ci/pytorch/smoke_test/max_autotune.py

@@ -1,205 +0,0 @@
-import argparse
-
-from torchvision import datasets, transforms
-
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-import torch.optim as optim
-from torch.optim.lr_scheduler import StepLR
-
-
-class Net(nn.Module):
-    def __init__(self):
-        super(Net, self).__init__()  # noqa: UP008
-        self.conv1 = nn.Conv2d(1, 32, 3, 1)
-        self.conv2 = nn.Conv2d(32, 64, 3, 1)
-        self.dropout1 = nn.Dropout(0.25)
-        self.dropout2 = nn.Dropout(0.5)
-        self.fc1 = nn.Linear(9216, 128)
-        self.fc2 = nn.Linear(128, 10)
-
-    def forward(self, x):
-        x = self.conv1(x)
-        x = F.relu(x)
-        x = self.conv2(x)
-        x = F.relu(x)
-        x = F.max_pool2d(x, 2)
-        x = self.dropout1(x)
-        x = torch.flatten(x, 1)
-        x = self.fc1(x)
-        x = F.relu(x)
-        x = self.dropout2(x)
-        x = self.fc2(x)
-        output = F.log_softmax(x, dim=1)
-        return output
-
-
-def train(args, model, device, train_loader, optimizer, epoch):
-    model.train()
-    for batch_idx, (data, target) in enumerate(train_loader):
-        data, target = data.to(device), target.to(device)
-        optimizer.zero_grad()
-        output = model(data)
-        loss = F.nll_loss(output, target)
-        loss.backward()
-        optimizer.step()
-        if batch_idx % args.log_interval == 0:
-            print(
-                f"Train Epoch: {epoch} [{batch_idx * len(data)}/{len(train_loader.dataset)} ({100. * batch_idx / len(train_loader):.0f}%)]\tLoss: {loss.item():.6f}"  # noqa: B950
-            )
-            if args.dry_run:
-                break
-
-
-def test(model, device, test_loader):
-    model.eval()
-    test_loss = 0
-    correct = 0
-    with torch.no_grad():
-        for data, target in test_loader:
-            data, target = data.to(device), target.to(device)
-            output = model(data)
-            test_loss += F.nll_loss(
-                output, target, reduction="sum"
-            ).item()  # sum up batch loss
-            pred = output.argmax(
-                dim=1, keepdim=True
-            )  # get the index of the max log-probability
-            correct += pred.eq(target.view_as(pred)).sum().item()
-
-    test_loss /= len(test_loader.dataset)
-
-    print(
-        f"\nTest set: Average loss: {test_loss:.4f}, Accuracy: {correct}/{len(test_loader.dataset)} ({100. * correct / len(test_loader.dataset):.0f}%)\n"  # noqa: B950
-    )
-
-
-def timed(fn):
-    start = torch.cuda.Event(enable_timing=True)
-    end = torch.cuda.Event(enable_timing=True)
-    start.record()
-    result = fn()
-    end.record()
-    torch.cuda.synchronize()
-    return result, start.elapsed_time(end) / 1000
-
-
-def main():
-    # Training settings
-    parser = argparse.ArgumentParser(description="PyTorch MNIST Example")
-    parser.add_argument(
-        "--batch-size",
-        type=int,
-        default=64,
-        metavar="N",
-        help="input batch size for training (default: 64)",
-    )
-    parser.add_argument(
-        "--test-batch-size",
-        type=int,
-        default=1000,
-        metavar="N",
-        help="input batch size for testing (default: 1000)",
-    )
-    parser.add_argument(
-        "--epochs",
-        type=int,
-        default=4,
-        metavar="N",
-        help="number of epochs to train (default: 14)",
-    )
-    parser.add_argument(
-        "--lr",
-        type=float,
-        default=1.0,
-        metavar="LR",
-        help="learning rate (default: 1.0)",
-    )
-    parser.add_argument(
-        "--gamma",
-        type=float,
-        default=0.7,
-        metavar="M",
-        help="Learning rate step gamma (default: 0.7)",
-    )
-    parser.add_argument(
-        "--no-cuda", action="store_true", default=False, help="disables CUDA training"
-    )
-    parser.add_argument(
-        "--no-mps",
-        action="store_true",
-        default=False,
-        help="disables macOS GPU training",
-    )
-    parser.add_argument(
-        "--dry-run",
-        action="store_true",
-        default=False,
-        help="quickly check a single pass",
-    )
-    parser.add_argument(
-        "--seed", type=int, default=1, metavar="S", help="random seed (default: 1)"
-    )
-    parser.add_argument(
-        "--log-interval",
-        type=int,
-        default=100,
-        metavar="N",
-        help="how many batches to wait before logging training status",
-    )
-    parser.add_argument(
-        "--save-model",
-        action="store_true",
-        default=False,
-        help="For Saving the current Model",
-    )
-    args = parser.parse_args()
-    use_cuda = not args.no_cuda and torch.cuda.is_available()
-    use_mps = not args.no_mps and torch.backends.mps.is_available()
-
-    torch.manual_seed(args.seed)
-    torch.backends.cuda.matmul.allow_tf32 = True
-
-    if use_cuda:
-        device = torch.device("cuda")
-    elif use_mps:
-        device = torch.device("mps")
-    else:
-        device = torch.device("cpu")
-
-    train_kwargs = {"batch_size": args.batch_size}
-    test_kwargs = {"batch_size": args.test_batch_size}
-    if use_cuda:
-        cuda_kwargs = {"num_workers": 1, "pin_memory": True, "shuffle": True}
-        train_kwargs.update(cuda_kwargs)
-        test_kwargs.update(cuda_kwargs)
-
-    transform = transforms.Compose(
-        [transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))]
-    )
-    dataset1 = datasets.MNIST("../data", train=True, download=True, transform=transform)
-    dataset2 = datasets.MNIST("../data", train=False, transform=transform)
-    train_loader = torch.utils.data.DataLoader(dataset1, **train_kwargs)
-    test_loader = torch.utils.data.DataLoader(dataset2, **test_kwargs)
-
-    model = Net().to(device)
-    opt_model = torch.compile(model, mode="max-autotune")
-    optimizer = optim.Adadelta(opt_model.parameters(), lr=args.lr)
-
-    scheduler = StepLR(optimizer, step_size=1, gamma=args.gamma)
-    for epoch in range(1, args.epochs + 1):
-        print(
-            f"Training Time: {timed(lambda: train(args, opt_model, device, train_loader, optimizer, epoch))[1]}"
-        )
-        print(
-            f"Evaluation Time: {timed(lambda: test(opt_model, device, test_loader))[1]}"
-        )
-        scheduler.step()
-
-    if args.save_model:
-        torch.save(opt_model.state_dict(), "mnist_cnn.pt")
-
-
-if __name__ == "__main__":
-    main()

.ci/pytorch/smoke_test/smoke_test.py

@@ -1,394 +0,0 @@
-import argparse
-import importlib
-import json
-import os
-import re
-import subprocess
-import sys
-from pathlib import Path
-
-import torch
-import torch._dynamo
-import torch.nn as nn
-import torch.nn.functional as F
-
-
-if "MATRIX_GPU_ARCH_VERSION" in os.environ:
-    gpu_arch_ver = os.getenv("MATRIX_GPU_ARCH_VERSION")
-else:
-    gpu_arch_ver = os.getenv("GPU_ARCH_VERSION")  # Use fallback if available
-gpu_arch_type = os.getenv("MATRIX_GPU_ARCH_TYPE")
-channel = os.getenv("MATRIX_CHANNEL")
-package_type = os.getenv("MATRIX_PACKAGE_TYPE")
-target_os = os.getenv("TARGET_OS", sys.platform)
-BASE_DIR = Path(__file__).parent.parent.parent
-
-is_cuda_system = gpu_arch_type == "cuda"
-NIGHTLY_ALLOWED_DELTA = 3
-
-MODULES = [
-    {
-        "name": "torchvision",
-        "repo": "https://github.com/pytorch/vision.git",
-        "smoke_test": "./vision/test/smoke_test.py",
-        "extension": "extension",
-        "repo_name": "vision",
-    },
-    {
-        "name": "torchaudio",
-        "repo": "https://github.com/pytorch/audio.git",
-        "smoke_test": "./audio/test/smoke_test/smoke_test.py --no-ffmpeg",
-        "extension": "_extension",
-        "repo_name": "audio",
-    },
-]
-
-
-class Net(nn.Module):
-    def __init__(self):
-        super().__init__()
-        self.conv1 = nn.Conv2d(1, 32, 3, 1)
-        self.conv2 = nn.Conv2d(32, 64, 3, 1)
-        self.fc1 = nn.Linear(9216, 1)
-
-    def forward(self, x):
-        x = self.conv1(x)
-        x = self.conv2(x)
-        x = F.max_pool2d(x, 2)
-        x = torch.flatten(x, 1)
-        output = self.fc1(x)
-        return output
-
-
-def load_json_from_basedir(filename: str):
-    try:
-        with open(BASE_DIR / filename) as fptr:
-            return json.load(fptr)
-    except FileNotFoundError as exc:
-        raise ImportError(f"File {filename} not found error: {exc.strerror}") from exc
-    except json.JSONDecodeError as exc:
-        raise ImportError(f"Invalid JSON {filename}") from exc
-
-
-def read_release_matrix():
-    return load_json_from_basedir("release_matrix.json")
-
-
-def test_numpy():
-    import numpy as np
-
-    x = np.arange(5)
-    torch.tensor(x)
-
-
-def check_version(package: str) -> None:
-    release_version = os.getenv("RELEASE_VERSION")
-    # if release_version is specified, use it to validate the packages
-    if release_version:
-        release_matrix = read_release_matrix()
-        stable_version = release_matrix["torch"]
-    else:
-        stable_version = os.getenv("MATRIX_STABLE_VERSION")
-
-    # only makes sense to check nightly package where dates are known
-    if channel == "nightly":
-        check_nightly_binaries_date(package)
-    elif stable_version is not None:
-        if not torch.__version__.startswith(stable_version):
-            raise RuntimeError(
-                f"Torch version mismatch, expected {stable_version} for channel {channel}. But its {torch.__version__}"
-            )
-
-        if release_version and package == "all":
-            for module in MODULES:
-                imported_module = importlib.import_module(module["name"])
-                module_version = imported_module.__version__
-                if not module_version.startswith(release_matrix[module["name"]]):
-                    raise RuntimeError(
-                        f"{module['name']} version mismatch, expected: \
-                            {release_matrix[module['name']]} for channel {channel}. But its {module_version}"
-                    )
-                else:
-                    print(
-                        f"{module['name']} version actual: {module_version} expected: \
-                        {release_matrix[module['name']]} for channel {channel}."
-                    )
-
-    else:
-        print(f"Skip version check for channel {channel} as stable version is None")
-
-
-def check_nightly_binaries_date(package: str) -> None:
-    from datetime import datetime
-
-    format_dt = "%Y%m%d"
-
-    date_t_str = re.findall("dev\\d+", torch.__version__)
-    date_t_delta = datetime.now() - datetime.strptime(date_t_str[0][3:], format_dt)
-    if date_t_delta.days >= NIGHTLY_ALLOWED_DELTA:
-        raise RuntimeError(
-            f"the binaries are from {date_t_str} and are more than {NIGHTLY_ALLOWED_DELTA} days old!"
-        )
-
-    if package == "all":
-        for module in MODULES:
-            imported_module = importlib.import_module(module["name"])
-            module_version = imported_module.__version__
-            date_m_str = re.findall("dev\\d+", module_version)
-            date_m_delta = datetime.now() - datetime.strptime(
-                date_m_str[0][3:], format_dt
-            )
-            print(f"Nightly date check for {module['name']} version {module_version}")
-            if date_m_delta.days > NIGHTLY_ALLOWED_DELTA:
-                raise RuntimeError(
-                    f"Expected {module['name']} to be less then {NIGHTLY_ALLOWED_DELTA} days. But its {date_m_delta}"
-                )
-
-
-def test_cuda_runtime_errors_captured() -> None:
-    cuda_exception_missed = True
-    try:
-        print("Testing test_cuda_runtime_errors_captured")
-        torch._assert_async(torch.tensor(0, device="cuda"))
-        torch._assert_async(torch.tensor(0 + 0j, device="cuda"))
-    except RuntimeError as e:
-        if re.search("CUDA", f"{e}"):
-            print(f"Caught CUDA exception with success: {e}")
-            cuda_exception_missed = False
-        else:
-            raise e
-    if cuda_exception_missed:
-        raise RuntimeError("Expected CUDA RuntimeError but have not received!")
-
-
-def smoke_test_cuda(
-    package: str, runtime_error_check: str, torch_compile_check: str
-) -> None:
-    if not torch.cuda.is_available() and is_cuda_system:
-        raise RuntimeError(f"Expected CUDA {gpu_arch_ver}. However CUDA is not loaded.")
-
-    if package == "all" and is_cuda_system:
-        for module in MODULES:
-            imported_module = importlib.import_module(module["name"])
-            # TBD for vision move extension module to private so it will
-            # be _extention.
-            version = "N/A"
-            if module["extension"] == "extension":
-                version = imported_module.extension._check_cuda_version()
-            else:
-                version = imported_module._extension._check_cuda_version()
-            print(f"{module['name']} CUDA: {version}")
-
-    # torch.compile is available on macos-arm64 and Linux for python 3.8-3.13
-    if (
-        torch_compile_check == "enabled"
-        and sys.version_info < (3, 14, 0)
-        and target_os in ["linux", "linux-aarch64", "macos-arm64", "darwin"]
-    ):
-        smoke_test_compile("cuda" if torch.cuda.is_available() else "cpu")
-
-    if torch.cuda.is_available():
-        if torch.version.cuda != gpu_arch_ver:
-            raise RuntimeError(
-                f"Wrong CUDA version. Loaded: {torch.version.cuda} Expected: {gpu_arch_ver}"
-            )
-        print(f"torch cuda: {torch.version.cuda}")
-        # todo add cudnn version validation
-        print(f"torch cudnn: {torch.backends.cudnn.version()}")
-        print(f"cuDNN enabled? {torch.backends.cudnn.enabled}")
-
-        torch.cuda.init()
-        print("CUDA initialized successfully")
-        print(f"Number of CUDA devices: {torch.cuda.device_count()}")
-        for i in range(torch.cuda.device_count()):
-            print(f"Device {i}: {torch.cuda.get_device_name(i)}")
-
-        # nccl is availbale only on Linux
-        if sys.platform in ["linux", "linux2"]:
-            print(f"torch nccl version: {torch.cuda.nccl.version()}")
-
-        if runtime_error_check == "enabled":
-            test_cuda_runtime_errors_captured()
-
-
-def smoke_test_conv2d() -> None:
-    import torch.nn as nn
-
-    print("Testing smoke_test_conv2d")
-    # With square kernels and equal stride
-    m = nn.Conv2d(16, 33, 3, stride=2)
-    # non-square kernels and unequal stride and with padding
-    m = nn.Conv2d(16, 33, (3, 5), stride=(2, 1), padding=(4, 2))
-    assert m is not None
-    # non-square kernels and unequal stride and with padding and dilation
-    basic_conv = nn.Conv2d(
-        16, 33, (3, 5), stride=(2, 1), padding=(4, 2), dilation=(3, 1)
-    )
-    input = torch.randn(20, 16, 50, 100)
-    output = basic_conv(input)
-
-    if is_cuda_system:
-        print("Testing smoke_test_conv2d with cuda")
-        conv = nn.Conv2d(3, 3, 3).cuda()
-        x = torch.randn(1, 3, 24, 24, device="cuda")
-        with torch.cuda.amp.autocast():
-            out = conv(x)
-        assert out is not None
-
-        supported_dtypes = [torch.float16, torch.float32, torch.float64]
-        for dtype in supported_dtypes:
-            print(f"Testing smoke_test_conv2d with cuda for {dtype}")
-            conv = basic_conv.to(dtype).cuda()
-            input = torch.randn(20, 16, 50, 100, device="cuda").type(dtype)
-            output = conv(input)
-            assert output is not None
-
-
-def test_linalg(device="cpu") -> None:
-    print(f"Testing smoke_test_linalg on {device}")
-    A = torch.randn(5, 3, device=device)
-    U, S, Vh = torch.linalg.svd(A, full_matrices=False)
-    assert (
-        U.shape == A.shape
-        and S.shape == torch.Size([3])
-        and Vh.shape == torch.Size([3, 3])
-    )
-    torch.dist(A, U @ torch.diag(S) @ Vh)
-
-    U, S, Vh = torch.linalg.svd(A)
-    assert (
-        U.shape == torch.Size([5, 5])
-        and S.shape == torch.Size([3])
-        and Vh.shape == torch.Size([3, 3])
-    )
-    torch.dist(A, U[:, :3] @ torch.diag(S) @ Vh)
-
-    A = torch.randn(7, 5, 3, device=device)
-    U, S, Vh = torch.linalg.svd(A, full_matrices=False)
-    torch.dist(A, U @ torch.diag_embed(S) @ Vh)
-
-    if device == "cuda":
-        supported_dtypes = [torch.float32, torch.float64]
-        for dtype in supported_dtypes:
-            print(f"Testing smoke_test_linalg with cuda for {dtype}")
-            A = torch.randn(20, 16, 50, 100, device=device, dtype=dtype)
-            torch.linalg.svd(A)
-
-
-def smoke_test_compile(device: str = "cpu") -> None:
-    supported_dtypes = [torch.float16, torch.float32, torch.float64]
-
-    def foo(x: torch.Tensor) -> torch.Tensor:
-        return torch.sin(x) + torch.cos(x)
-
-    for dtype in supported_dtypes:
-        print(f"Testing smoke_test_compile for {device} and {dtype}")
-        x = torch.rand(3, 3, device=device).type(dtype)
-        x_eager = foo(x)
-        x_pt2 = torch.compile(foo)(x)
-        torch.testing.assert_close(x_eager, x_pt2)
-
-    # Check that SIMD were detected for the architecture
-    if device == "cpu":
-        from torch._inductor.codecache import pick_vec_isa
-
-        isa = pick_vec_isa()
-        if not isa:
-            raise RuntimeError("Can't detect vectorized ISA for CPU")
-        print(f"Picked CPU ISA {type(isa).__name__} bit width {isa.bit_width()}")
-
-    # Reset torch dynamo since we are changing mode
-    torch._dynamo.reset()
-    dtype = torch.float32
-    torch.set_float32_matmul_precision("high")
-    print(f"Testing smoke_test_compile with mode 'max-autotune' for {dtype}")
-    x = torch.rand(64, 1, 28, 28, device=device).type(torch.float32)
-    model = Net().to(device=device)
-    x_pt2 = torch.compile(model, mode="max-autotune")(x)
-
-
-def smoke_test_modules():
-    cwd = os.getcwd()
-    for module in MODULES:
-        if module["repo"]:
-            if not os.path.exists(f"{cwd}/{module['repo_name']}"):
-                print(f"Path does not exist: {cwd}/{module['repo_name']}")
-                try:
-                    subprocess.check_output(
-                        f"git clone --depth 1 {module['repo']}",
-                        stderr=subprocess.STDOUT,
-                        shell=True,
-                    )
-                except subprocess.CalledProcessError as exc:
-                    raise RuntimeError(
-                        f"Cloning {module['repo']} FAIL: {exc.returncode} Output: {exc.output}"
-                    ) from exc
-            try:
-                smoke_test_command = f"python3 {module['smoke_test']}"
-                if target_os == "windows":
-                    smoke_test_command = f"python {module['smoke_test']}"
-                output = subprocess.check_output(
-                    smoke_test_command,
-                    stderr=subprocess.STDOUT,
-                    shell=True,
-                    universal_newlines=True,
-                )
-            except subprocess.CalledProcessError as exc:
-                raise RuntimeError(
-                    f"Module {module['name']} FAIL: {exc.returncode} Output: {exc.output}"
-                ) from exc
-            else:
-                print(f"Output: \n{output}\n")
-
-
-def parse_args():
-    parser = argparse.ArgumentParser()
-    parser.add_argument(
-        "--package",
-        help="Package to include in smoke testing",
-        type=str,
-        choices=["all", "torchonly"],
-        default="all",
-    )
-    parser.add_argument(
-        "--runtime-error-check",
-        help="No Runtime Error check",
-        type=str,
-        choices=["enabled", "disabled"],
-        default="enabled",
-    )
-    parser.add_argument(
-        "--torch-compile-check",
-        help="Check torch compile",
-        type=str,
-        choices=["enabled", "disabled"],
-        default="enabled",
-    )
-    return parser.parse_args()
-
-
-def main() -> None:
-    options = parse_args()
-    print(f"torch: {torch.__version__}")
-    print(torch.__config__.parallel_info())
-    # All PyTorch binary builds should be built with OpenMP
-    if not torch.backends.openmp.is_available():
-        raise RuntimeError("PyTorch must be built with OpenMP support")
-
-    check_version(options.package)
-    smoke_test_conv2d()
-    test_linalg()
-    test_numpy()
-    if is_cuda_system:
-        test_linalg("cuda")
-
-    if options.package == "all":
-        smoke_test_modules()
-
-    smoke_test_cuda(
-        options.package, options.runtime_error_check, options.torch_compile_check
-    )
-
-
-if __name__ == "__main__":
-    main()

.ci/pytorch/test.sh

@@ -4,7 +4,7 @@
 # (This is set by default in the Docker images we build, so you don't
 # need to set it yourself.
 
-set -ex -o pipefail
+set -ex
 
 # Suppress ANSI color escape sequences
 export TERM=vt100
@@ -12,9 +12,9 @@ export TERM=vt100
 # shellcheck source=./common.sh
 source "$(dirname "${BASH_SOURCE[0]}")/common.sh"
 
-# Do not change workspace permissions for ROCm and s390x CI jobs
+# Do not change workspace permissions for ROCm CI jobs
 # as it can leave workspace with bad permissions for cancelled jobs
-if [[ "$BUILD_ENVIRONMENT" != *rocm* && "$BUILD_ENVIRONMENT" != *s390x* && -d /var/lib/jenkins/workspace ]]; then
+if [[ "$BUILD_ENVIRONMENT" != *rocm* ]]; then
   # Workaround for dind-rootless userid mapping (https://github.com/pytorch/ci-infra/issues/96)
   WORKSPACE_ORIGINAL_OWNER_ID=$(stat -c '%u' "/var/lib/jenkins/workspace")
   cleanup_workspace() {
@@ -48,7 +48,7 @@ NUM_TEST_SHARDS="${NUM_TEST_SHARDS:=1}"
 
 export VALGRIND=ON
 # export TORCH_INDUCTOR_INSTALL_GXX=ON
-if [[ "$BUILD_ENVIRONMENT" == *clang9* || "$BUILD_ENVIRONMENT" == *xpu* ]]; then
+if [[ "$BUILD_ENVIRONMENT" == *clang9* ]]; then
   # clang9 appears to miscompile code involving std::optional<c10::SymInt>,
   # such that valgrind complains along these lines:
   #
@@ -86,13 +86,6 @@ if [[ "$BUILD_ENVIRONMENT" == *clang9* || "$BUILD_ENVIRONMENT" == *xpu* ]]; then
   export VALGRIND=OFF
 fi
 
-
-if [[ "$BUILD_ENVIRONMENT" == *s390x* ]]; then
-  # There are additional warnings on s390x, maybe due to newer gcc.
-  # Skip this check for now
-  export VALGRIND=OFF
-fi
-
 if [[ "${PYTORCH_TEST_RERUN_DISABLED_TESTS}" == "1" ]] || [[ "${CONTINUE_THROUGH_ERROR}" == "1" ]]; then
   # When rerunning disable tests, do not generate core dumps as it could consume
   # the runner disk space when crashed tests are run multiple times. Running out
@@ -136,7 +129,7 @@ if [[ "$TEST_CONFIG" == 'default' ]]; then
 fi
 
 if [[ "$TEST_CONFIG" == 'distributed' ]] && [[ "$BUILD_ENVIRONMENT" == *rocm* ]]; then
-  export HIP_VISIBLE_DEVICES=0,1,2,3
+  export HIP_VISIBLE_DEVICES=0,1
 fi
 
 if [[ "$TEST_CONFIG" == 'slow' ]]; then
@@ -160,8 +153,6 @@ elif [[ "$BUILD_ENVIRONMENT" == *xpu* ]]; then
   export PYTORCH_TESTING_DEVICE_ONLY_FOR="xpu"
   # setting PYTHON_TEST_EXTRA_OPTION
   export PYTHON_TEST_EXTRA_OPTION="--xpu"
-  # Disable sccache for xpu test due to flaky issue https://github.com/pytorch/pytorch/issues/143585
-  sudo rm -rf /opt/cache
 fi
 
 if [[ "$TEST_CONFIG" == *crossref* ]]; then
@@ -178,13 +169,9 @@ fi
 
 if [[ "$BUILD_ENVIRONMENT" == *xpu* ]]; then
   # Source Intel oneAPI envrioment script to enable xpu runtime related libraries
-  # refer to https://www.intel.com/content/www/us/en/developer/articles/tool/pytorch-prerequisites-for-intel-gpus.html
+  # refer to https://www.intel.com/content/www/us/en/developer/articles/tool/pytorch-prerequisites-for-intel-gpu/2-5.html
   # shellcheck disable=SC1091
   source /opt/intel/oneapi/compiler/latest/env/vars.sh
-  if [ -f /opt/intel/oneapi/umf/latest/env/vars.sh ]; then
-    # shellcheck disable=SC1091
-    source /opt/intel/oneapi/umf/latest/env/vars.sh
-  fi
   # Check XPU status before testing
   xpu-smi discovery
 fi
@@ -309,7 +296,7 @@ test_python() {
 }
 
 
-test_dynamo_wrapped_shard() {
+test_dynamo_shard() {
   if [[ -z "$NUM_TEST_SHARDS" ]]; then
     echo "NUM_TEST_SHARDS must be defined to run a Python test shard"
     exit 1
@@ -322,7 +309,6 @@ test_dynamo_wrapped_shard() {
     --exclude-jit-executor \
     --exclude-distributed-tests \
     --exclude-torch-export-tests \
-    --exclude-aot-dispatch-tests \
     --shard "$1" "$NUM_TEST_SHARDS" \
     --verbose \
     --upload-artifacts-while-running
@@ -336,7 +322,7 @@ test_inductor_distributed() {
   python test/run_test.py -i inductor/test_aot_inductor.py -k test_non_default_cuda_device --verbose
   python test/run_test.py -i inductor/test_aot_inductor.py -k test_replicate_on_devices --verbose
   python test/run_test.py -i distributed/test_c10d_functional_native.py --verbose
-  python test/run_test.py -i distributed/tensor/test_dtensor_compile.py --verbose
+  python test/run_test.py -i distributed/_tensor/test_dtensor_compile.py --verbose
   python test/run_test.py -i distributed/tensor/parallel/test_micro_pipeline_tp.py --verbose
   python test/run_test.py -i distributed/_composable/test_replicate_with_compiler.py --verbose
   python test/run_test.py -i distributed/_composable/fsdp/test_fully_shard_comm.py --verbose
@@ -389,53 +375,27 @@ test_inductor_aoti() {
   CPP_TESTS_DIR="${BUILD_BIN_DIR}" LD_LIBRARY_PATH="${TORCH_LIB_DIR}" python test/run_test.py --cpp --verbose -i cpp/test_aoti_abi_check cpp/test_aoti_inference
 }
 
-test_inductor_cpp_wrapper_shard() {
-  if [[ -z "$NUM_TEST_SHARDS" ]]; then
-    echo "NUM_TEST_SHARDS must be defined to run a Python test shard"
-    exit 1
-  fi
-
+test_inductor_cpp_wrapper() {
   export TORCHINDUCTOR_CPP_WRAPPER=1
   TEST_REPORTS_DIR=$(pwd)/test/test-reports
   mkdir -p "$TEST_REPORTS_DIR"
 
-  if [[ "$1" -eq "2" ]]; then
-    # For now, manually put the opinfo tests in shard 2, and all other tests in
-    # shard 1.  Test specific things triggering past bugs, for now.
-    python test/run_test.py \
-      --include inductor/test_torchinductor_opinfo \
-      -k 'linalg or to_sparse' \
-      --verbose
-    exit
-  fi
-
-  # Run certain inductor unit tests with cpp wrapper. In the end state, we
-  # should be able to run all the inductor unit tests with cpp_wrapper.
-  python test/run_test.py --include inductor/test_torchinductor --verbose
-
-  # Run inductor benchmark tests with cpp wrapper.
-  # Skip benchmark tests if it's in rerun-disabled-mode.
-  if [[ "${PYTORCH_TEST_RERUN_DISABLED_TESTS}" == "1" ]]; then
-    echo "skip dynamo benchmark tests for rerun-disabled-test"
-  else
-    echo "run dynamo benchmark tests with cpp wrapper"
-    python benchmarks/dynamo/timm_models.py --device cuda --accuracy --amp \
+  python benchmarks/dynamo/timm_models.py --device cuda --accuracy --amp \
     --training --inductor --disable-cudagraphs --only vit_base_patch16_224 \
     --output "$TEST_REPORTS_DIR/inductor_cpp_wrapper_training.csv"
-    python benchmarks/dynamo/check_accuracy.py \
-      --actual "$TEST_REPORTS_DIR/inductor_cpp_wrapper_training.csv" \
-      --expected "benchmarks/dynamo/ci_expected_accuracy/inductor_timm_training.csv"
+  python benchmarks/dynamo/check_accuracy.py \
+    --actual "$TEST_REPORTS_DIR/inductor_cpp_wrapper_training.csv" \
+    --expected "benchmarks/dynamo/ci_expected_accuracy/inductor_timm_training.csv"
 
-    python benchmarks/dynamo/torchbench.py --device cuda --accuracy \
-      --bfloat16 --inference --inductor --only hf_T5 --output "$TEST_REPORTS_DIR/inductor_cpp_wrapper_inference.csv"
-    python benchmarks/dynamo/torchbench.py --device cuda --accuracy \
-      --bfloat16 --inference --inductor --only llama --output "$TEST_REPORTS_DIR/inductor_cpp_wrapper_inference.csv"
-    python benchmarks/dynamo/torchbench.py --device cuda --accuracy \
-      --bfloat16 --inference --inductor --only moco --output "$TEST_REPORTS_DIR/inductor_cpp_wrapper_inference.csv"
-    python benchmarks/dynamo/check_accuracy.py \
-      --actual "$TEST_REPORTS_DIR/inductor_cpp_wrapper_inference.csv" \
-      --expected "benchmarks/dynamo/ci_expected_accuracy/inductor_torchbench_inference.csv"
-  fi
+  python benchmarks/dynamo/torchbench.py --device cuda --accuracy \
+    --bfloat16 --inference --inductor --only hf_T5 --output "$TEST_REPORTS_DIR/inductor_cpp_wrapper_inference.csv"
+  python benchmarks/dynamo/torchbench.py --device cuda --accuracy \
+    --bfloat16 --inference --inductor --only llama --output "$TEST_REPORTS_DIR/inductor_cpp_wrapper_inference.csv"
+  python benchmarks/dynamo/torchbench.py --device cuda --accuracy \
+    --bfloat16 --inference --inductor --only moco --output "$TEST_REPORTS_DIR/inductor_cpp_wrapper_inference.csv"
+  python benchmarks/dynamo/check_accuracy.py \
+    --actual "$TEST_REPORTS_DIR/inductor_cpp_wrapper_inference.csv" \
+    --expected "benchmarks/dynamo/ci_expected_accuracy/inductor_torchbench_inference.csv"
 }
 
 # "Global" flags for inductor benchmarking controlled by TEST_CONFIG
@@ -541,7 +501,7 @@ test_perf_for_dashboard() {
             --dynamic-batch-only "$@" \
             --output "$TEST_REPORTS_DIR/${backend}_dynamic_${suite}_${dtype}_${mode}_${device}_${target}.csv"
       fi
-      if [[ "$DASHBOARD_TAG" == *cppwrapper-true* ]]; then
+      if [[ "$DASHBOARD_TAG" == *cppwrapper-true* ]] && [[ "$mode" == "inference" ]]; then
         TORCHINDUCTOR_CPP_WRAPPER=1 $TASKSET python "benchmarks/dynamo/$suite.py" \
             "${target_flag[@]}" --"$mode" --"$dtype" --backend "$backend" --disable-cudagraphs "$@" \
             --output "$TEST_REPORTS_DIR/${backend}_cpp_wrapper_${suite}_${dtype}_${mode}_${device}_${target}.csv"
@@ -639,11 +599,6 @@ test_single_dynamo_benchmark() {
 }
 
 test_inductor_micro_benchmark() {
-  # torchao requires cuda 8.0 or above for bfloat16 support
-  if [[ "$BUILD_ENVIRONMENT" == *cuda* ]]; then
-    export TORCH_CUDA_ARCH_LIST="8.0;8.6"
-  fi
-  install_torchao
   TEST_REPORTS_DIR=$(pwd)/test/test-reports
   if [[ "${TEST_CONFIG}" == *cpu* ]]; then
     test_inductor_set_cpu_affinity
@@ -847,7 +802,7 @@ test_without_numpy() {
   # Regression test for https://github.com/pytorch/pytorch/issues/66353
   python -c "import sys;sys.path.insert(0, 'fake_numpy');import torch;print(torch.tensor([torch.tensor(0.), torch.tensor(1.)]))"
   # Regression test for https://github.com/pytorch/pytorch/issues/109387
-  if [[ "${TEST_CONFIG}" == *dynamo_wrapped* ]]; then
+  if [[ "${TEST_CONFIG}" == *dynamo* ]]; then
     python -c "import sys;sys.path.insert(0, 'fake_numpy');import torch;torch.compile(lambda x:print(x))('Hello World')"
   fi
   popd
@@ -917,20 +872,10 @@ test_libtorch_api() {
   else
     # Exclude IMethodTest that relies on torch::deploy, which will instead be ran in test_deploy
     OMP_NUM_THREADS=2 TORCH_CPP_TEST_MNIST_PATH="${MNIST_DIR}" python test/run_test.py --cpp --verbose -i cpp/test_api -k "not IMethodTest"
-
-    # On s390x, pytorch is built without llvm.
-    # Even if it would be built with llvm, llvm currently doesn't support used features on s390x and
-    # test fails with errors like:
-    # JIT session error: Unsupported target machine architecture in ELF object pytorch-jitted-objectbuffer
-    # unknown file: Failure
-    # C++ exception with description "valOrErr INTERNAL ASSERT FAILED at "/var/lib/jenkins/workspace/torch/csrc/jit/tensorexpr/llvm_jit.h":34, please report a bug to PyTorch. Unexpected failure in LLVM JIT: Failed to materialize symbols: { (main, { func }) }
-    if [[ "${BUILD_ENVIRONMENT}" != *s390x* ]]; then
-      python test/run_test.py --cpp --verbose -i cpp/test_tensorexpr
-    fi
+    python test/run_test.py --cpp --verbose -i cpp/test_tensorexpr
   fi
 
-  # quantization is not fully supported on s390x yet
-  if [[ "${BUILD_ENVIRONMENT}" != *android* && "${BUILD_ENVIRONMENT}" != *cuda* && "${BUILD_ENVIRONMENT}" != *asan* && "${BUILD_ENVIRONMENT}" != *s390x* ]]; then
+  if [[ "${BUILD_ENVIRONMENT}" != *android* && "${BUILD_ENVIRONMENT}" != *cuda* && "${BUILD_ENVIRONMENT}" != *asan* ]]; then
     # NB: This test is not under TORCH_BIN_DIR but under BUILD_BIN_DIR
     export CPP_TESTS_DIR="${BUILD_BIN_DIR}"
     python test/run_test.py --cpp --verbose -i cpp/static_runtime_test
@@ -991,9 +936,6 @@ test_distributed() {
     python test/run_test.py --cpp --verbose -i cpp/HashStoreTest
     python test/run_test.py --cpp --verbose -i cpp/TCPStoreTest
 
-    echo "Testing multi-GPU linalg tests"
-    python test/run_test.py -i test_linalg.py -k test_matmul_offline_mgpu_tunable --verbose
-
     if [[ "$BUILD_ENVIRONMENT" == *cuda* ]]; then
       MPIEXEC=$(command -v mpiexec)
       if [[ -n "$MPIEXEC" ]]; then
@@ -1243,7 +1185,7 @@ EOF
   git reset --hard "${SHA_TO_COMPARE}"
   git submodule sync && git submodule update --init --recursive
   echo "::group::Installing Torch From Base Commit"
-  pip3 install -r requirements.txt
+  pip install -r requirements.txt
   # shellcheck source=./common-build.sh
   source "$(dirname "${BASH_SOURCE[0]}")/common-build.sh"
   python setup.py bdist_wheel --bdist-dir="base_bdist_tmp" --dist-dir="base_dist"
@@ -1277,7 +1219,7 @@ EOF
 }
 
 test_bazel() {
-  set -e -o pipefail
+  set -e
 
   # bazel test needs sccache setup.
   # shellcheck source=./common-build.sh
@@ -1400,11 +1342,10 @@ test_executorch() {
   export EXECUTORCH_BUILD_PYBIND=ON
   export CMAKE_ARGS="-DEXECUTORCH_BUILD_XNNPACK=ON -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON"
 
-  # For llama3
-  bash examples/models/llama3_2_vision/install_requirements.sh
   # NB: We need to rebuild ExecuTorch runner here because it depends on PyTorch
   # from the PR
-  bash .ci/scripts/setup-linux.sh cmake
+  # shellcheck disable=SC1091
+  source .ci/scripts/setup-linux.sh cmake
 
   echo "Run ExecuTorch unit tests"
   pytest -v -n auto
@@ -1427,8 +1368,7 @@ test_executorch() {
 
 test_linux_aarch64() {
   python test/run_test.py --include test_modules test_mkldnn test_mkldnn_fusion test_openmp test_torch test_dynamic_shapes \
-        test_transformers test_multiprocessing test_numpy_interop test_autograd test_binary_ufuncs test_complex test_spectral_ops \
-        test_foreach test_reductions test_unary_ufuncs \
+        test_transformers test_multiprocessing test_numpy_interop \
         --shard "$SHARD_NUMBER" "$NUM_TEST_SHARDS" --verbose
 
   # Dynamo tests
@@ -1446,7 +1386,6 @@ test_linux_aarch64() {
        inductor/test_pattern_matcher inductor/test_perf inductor/test_profiler inductor/test_select_algorithm inductor/test_smoke \
        inductor/test_split_cat_fx_passes inductor/test_standalone_compile inductor/test_torchinductor \
        inductor/test_torchinductor_codegen_dynamic_shapes inductor/test_torchinductor_dynamic_shapes inductor/test_memory \
-       inductor/test_triton_cpu_backend inductor/test_triton_extension_backend inductor/test_mkldnn_pattern_matcher inductor/test_cpu_cpp_wrapper \
        --shard "$SHARD_NUMBER" "$NUM_TEST_SHARDS" --verbose
 }
 
@@ -1454,11 +1393,7 @@ if ! [[ "${BUILD_ENVIRONMENT}" == *libtorch* || "${BUILD_ENVIRONMENT}" == *-baze
   (cd test && python -c "import torch; print(torch.__config__.show())")
   (cd test && python -c "import torch; print(torch.__config__.parallel_info())")
 fi
-if [[ "${TEST_CONFIG}" == *numpy_2* ]]; then
-  # Install numpy-2.0.2 and compatible scipy & numba versions
-  python -mpip install --pre numpy==2.0.2 scipy==1.13.1 numba==0.60.0
-  python test/run_test.py --include dynamo/test_functions.py dynamo/test_unspec.py test_binary_ufuncs.py test_fake_tensor.py test_linalg.py test_numpy_interop.py test_tensor_creation_ops.py test_torch.py torch_np/test_basic.py
-elif [[ "${BUILD_ENVIRONMENT}" == *aarch64* && "${TEST_CONFIG}" != *perf_cpu_aarch64* ]]; then
+if [[ "${BUILD_ENVIRONMENT}" == *aarch64* && "${TEST_CONFIG}" != *perf_cpu_aarch64* ]]; then
   test_linux_aarch64
 elif [[ "${TEST_CONFIG}" == *backward* ]]; then
   test_forward_backward_compatibility
@@ -1531,7 +1466,7 @@ elif [[ "${TEST_CONFIG}" == *inductor_cpp_wrapper* ]]; then
   install_torchaudio cuda
   install_torchvision
   checkout_install_torchbench hf_T5 llama moco
-  PYTHONPATH=$(pwd)/torchbench test_inductor_cpp_wrapper_shard "$SHARD_NUMBER"
+  PYTHONPATH=$(pwd)/torchbench test_inductor_cpp_wrapper
 elif [[ "${TEST_CONFIG}" == *inductor* ]]; then
   install_torchvision
   test_inductor_shard "${SHARD_NUMBER}"
@@ -1540,9 +1475,9 @@ elif [[ "${TEST_CONFIG}" == *inductor* ]]; then
       test_inductor_distributed
     fi
   fi
-elif [[ "${TEST_CONFIG}" == *dynamo_wrapped* ]]; then
+elif [[ "${TEST_CONFIG}" == *dynamo* ]]; then
   install_torchvision
-  test_dynamo_wrapped_shard "${SHARD_NUMBER}"
+  test_dynamo_shard "${SHARD_NUMBER}"
   if [[ "${SHARD_NUMBER}" == 1 ]]; then
     test_aten
   fi

.ci/pytorch/test_example_code/CMakeLists.txt

@@ -1,26 +0,0 @@
-cmake_minimum_required(VERSION 3.0 FATAL_ERROR)
-project(simple-torch-test)
-
-find_package(Torch REQUIRED)
-set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${TORCH_CXX_FLAGS}")
-
-
-add_executable(simple-torch-test simple-torch-test.cpp)
-target_include_directories(simple-torch-test PRIVATE  ${TORCH_INCLUDE_DIRS})
-target_link_libraries(simple-torch-test "${TORCH_LIBRARIES}")
-set_property(TARGET simple-torch-test PROPERTY CXX_STANDARD 17)
-
-find_package(CUDAToolkit 11.8)
-
-target_link_libraries(simple-torch-test CUDA::cudart CUDA::cufft CUDA::cusparse CUDA::cublas CUDA::cusolver)
-find_library(CUDNN_LIBRARY NAMES cudnn)
-target_link_libraries(simple-torch-test  ${CUDNN_LIBRARY} )
-if(MSVC)
-  file(GLOB TORCH_DLLS  "$ENV{CUDA_PATH}/bin/cudnn64_8.dll" "$ENV{NVTOOLSEXT_PATH}/bin/x64/*.dll")
-  message("dlls to copy "  ${TORCH_DLLS})
-  add_custom_command(TARGET simple-torch-test
-                     POST_BUILD
-                     COMMAND ${CMAKE_COMMAND} -E copy_if_different
-                     ${TORCH_DLLS}
-                     $<TARGET_FILE_DIR:simple-torch-test>)
-endif(MSVC)

.ci/pytorch/test_example_code/check-torch-cuda.cpp

@@ -1,15 +0,0 @@
-#include <torch/torch.h>
-
-int main(int argc, const char* argv[]) {
-    std::cout << "Checking that CUDA archs are setup correctly" << std::endl;
-    TORCH_CHECK(torch::rand({ 3, 5 }, torch::Device(torch::kCUDA)).defined(), "CUDA archs are not setup correctly");
-
-    // These have to run after CUDA is initialized
-
-    std::cout << "Checking that magma is available" << std::endl;
-    TORCH_CHECK(torch::hasMAGMA(), "MAGMA is not available");
-
-    std::cout << "Checking that CuDNN is available" << std::endl;
-    TORCH_CHECK(torch::cuda::cudnn_is_available(), "CuDNN is not available");
-    return 0;
-}

.ci/pytorch/test_example_code/check-torch-mkl.cpp

@@ -1,6 +0,0 @@
-#include <torch/torch.h>
-
-int main(int argc, const char* argv[]) {
-    TORCH_CHECK(torch::hasMKL(), "MKL is not available");
-    return 0;
-}

.ci/pytorch/test_example_code/check-torch-xnnpack.cpp

@@ -1,7 +0,0 @@
-#include <ATen/ATen.h>
-#include <torch/torch.h>
-
-int main(int argc, const char* argv[]) {
-    TORCH_CHECK(at::globalContext().isXNNPACKAvailable(), "XNNPACK is not available");
-    return 0;
-}

.ci/pytorch/test_example_code/cnn_smoke.py

@@ -1,38 +0,0 @@
-r"""
-It's used to check basic rnn features with cuda.
-For example, it would throw exception if some components are missing
-"""
-
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-import torch.optim as optim
-
-
-class SimpleCNN(nn.Module):
-    def __init__(self):
-        super().__init__()
-        self.conv = nn.Conv2d(1, 1, 3)
-        self.pool = nn.MaxPool2d(2, 2)
-
-    def forward(self, inputs):
-        output = self.pool(F.relu(self.conv(inputs)))
-        output = output.view(1)
-        return output
-
-
-# Mock one infer
-device = torch.device("cuda:0")
-net = SimpleCNN().to(device)
-net_inputs = torch.rand((1, 1, 5, 5), device=device)
-outputs = net(net_inputs)
-print(outputs)
-
-criterion = nn.MSELoss()
-optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.1)
-
-# Mock one step training
-label = torch.full((1,), 1.0, dtype=torch.float, device=device)
-loss = criterion(outputs, label)
-loss.backward()
-optimizer.step()

.ci/pytorch/test_example_code/rnn_smoke.py

@@ -1,14 +0,0 @@
-r"""
-It's used to check basic rnn features with cuda.
-For example, it would throw exception if missing some components are missing
-"""
-
-import torch
-import torch.nn as nn
-
-
-device = torch.device("cuda:0")
-rnn = nn.RNN(10, 20, 2).to(device)
-inputs = torch.randn(5, 3, 10).to(device)
-h0 = torch.randn(2, 3, 20).to(device)
-output, hn = rnn(inputs, h0)

.ci/pytorch/test_example_code/simple-torch-test.cpp

@@ -1,6 +0,0 @@
-#include <torch/torch.h>
-
-int main(int argc, const char* argv[]) {
-    TORCH_WARN("Simple test passed!");
-    return 0;
-}

.ci/pytorch/win-build.sh

@@ -38,7 +38,7 @@ if [[ $PYLONG_API_CHECK == 0 ]]; then
   echo "PyLong_AsUnsignedLong -> THPUtils_unpackUInt32 / THPUtils_unpackUInt64"
   exit 1
 fi
-set -ex -o pipefail
+set -ex
 
 "$SCRIPT_HELPERS_DIR"/build_pytorch.bat
 

.ci/pytorch/win-test-helpers/build_pytorch.bat

@@ -26,8 +26,7 @@ if not errorlevel 0 goto fail
 
 if "%USE_XPU%"=="1" (
   :: Install xpu support packages
-  set CUDA_VERSION=xpu
-  call %SCRIPT_HELPERS_DIR%\..\windows\internal\xpu_install.bat
+  call %INSTALLER_DIR%\install_xpu.bat
   if errorlevel 1 exit /b 1
 )
 
@@ -53,8 +52,7 @@ if not errorlevel 0 goto fail
 
 if "%USE_XPU%"=="1" (
   :: Activate xpu environment - VS env is required for xpu
-  call "C:\Program Files (x86)\Intel\oneAPI\compiler\latest\env\vars.bat"
-  call "C:\Program Files (x86)\Intel\oneAPI\ocloc\latest\env\vars.bat"
+  call "C:\Program Files (x86)\Intel\oneAPI\setvars.bat"
   if errorlevel 1 exit /b 1
   :: Reduce build time. Only have MTL self-hosted runner now
   SET TORCH_XPU_ARCH_LIST=xe-lpg