diff --git a/.devcontainer/cuda12.6-llvm14/devcontainer.json b/.devcontainer/cuda12.6-llvm14/devcontainer.json deleted file mode 100644 index 16d79694534..00000000000 --- a/.devcontainer/cuda12.6-llvm14/devcontainer.json +++ /dev/null @@ -1,54 +0,0 @@ -{ - "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-llvm14-cuda12.6", - "hostRequirements": { - "gpu": "optional" - }, - "initializeCommand": [ - "/bin/bash", - "-c", - "mkdir -m 0755 -p ${localWorkspaceFolder}/.{aws,cache,config}; mkdir -m 0755 -p ${localWorkspaceFolder}/build;", - "if [[ -n ${WSLENV+set} ]]; then docker volume create cccl-build; else docker volume create --driver local --opt type=none --opt device=${localWorkspaceFolder}/build --opt o=bind cccl-build fi;" - ], - "containerEnv": { - "SCCACHE_REGION": "us-east-2", - "SCCACHE_BUCKET": "rapids-sccache-devs", - "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", - "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-llvm14", - "CCCL_CUDA_VERSION": "12.6", - "CCCL_HOST_COMPILER": "llvm", - "CCCL_HOST_COMPILER_VERSION": "14", - "CCCL_BUILD_INFIX": "cuda12.6-llvm14", - "CCCL_CUDA_EXTENDED": "false" - }, - "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", - "workspaceMount": "source=${localWorkspaceFolder},target=/home/coder/${localWorkspaceFolderBasename},type=bind,consistency=consistent", - "mounts": [ - "source=${localWorkspaceFolder}/.aws,target=/home/coder/.aws,type=bind,consistency=consistent", - "source=${localWorkspaceFolder}/.cache,target=/home/coder/.cache,type=bind,consistency=consistent", - "source=${localWorkspaceFolder}/.config,target=/home/coder/.config,type=bind,consistency=consistent", - "source=cccl-build,target=/home/coder/cccl/build" - ], - "customizations": { - "vscode": { - "extensions": [ - "llvm-vs-code-extensions.vscode-clangd", - "xaver.clang-format", - "nvidia.nsight-vscode-edition", - "ms-vscode.cmake-tools" - ], - "settings": { - "editor.defaultFormatter": "xaver.clang-format", - "editor.formatOnSave": true, - "clang-format.executable": "/usr/bin/clang-format", - "clangd.arguments": [ - "--compile-commands-dir=${workspaceFolder}" - ], - "files.eol": "\n", - "files.trimTrailingWhitespace": true - } - } - }, - "name": "cuda12.6-llvm14" -} diff --git a/.devcontainer/cuda12.6-gcc10/devcontainer.json b/.devcontainer/cuda12.8-gcc10/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-gcc10/devcontainer.json rename to .devcontainer/cuda12.8-gcc10/devcontainer.json index 8ecf9dfa038..5fa59ac052f 100644 --- a/.devcontainer/cuda12.6-gcc10/devcontainer.json +++ b/.devcontainer/cuda12.8-gcc10/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc10-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-gcc10-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-gcc10", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-gcc10", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "10", - "CCCL_BUILD_INFIX": "cuda12.6-gcc10", + "CCCL_BUILD_INFIX": "cuda12.8-gcc10", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-gcc10" + "name": "cuda12.8-gcc10" } diff --git a/.devcontainer/cuda12.6-gcc11/devcontainer.json b/.devcontainer/cuda12.8-gcc11/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-gcc11/devcontainer.json rename to .devcontainer/cuda12.8-gcc11/devcontainer.json index 2f26a6e4d4f..e6408c727e9 100644 --- a/.devcontainer/cuda12.6-gcc11/devcontainer.json +++ b/.devcontainer/cuda12.8-gcc11/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc11-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-gcc11-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-gcc11", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-gcc11", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "11", - "CCCL_BUILD_INFIX": "cuda12.6-gcc11", + "CCCL_BUILD_INFIX": "cuda12.8-gcc11", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-gcc11" + "name": "cuda12.8-gcc11" } diff --git a/.devcontainer/cuda12.6-gcc12/devcontainer.json b/.devcontainer/cuda12.8-gcc12/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-gcc12/devcontainer.json rename to .devcontainer/cuda12.8-gcc12/devcontainer.json index 208dd6f76e1..1538817d17b 100644 --- a/.devcontainer/cuda12.6-gcc12/devcontainer.json +++ b/.devcontainer/cuda12.8-gcc12/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc12-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-gcc12-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-gcc12", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-gcc12", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "12", - "CCCL_BUILD_INFIX": "cuda12.6-gcc12", + "CCCL_BUILD_INFIX": "cuda12.8-gcc12", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-gcc12" + "name": "cuda12.8-gcc12" } diff --git a/.devcontainer/cuda12.6-gcc13/devcontainer.json b/.devcontainer/cuda12.8-gcc13/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-gcc13/devcontainer.json rename to .devcontainer/cuda12.8-gcc13/devcontainer.json index 1c7203a1a3c..0cf38c8b80b 100644 --- a/.devcontainer/cuda12.6-gcc13/devcontainer.json +++ b/.devcontainer/cuda12.8-gcc13/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc13-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-gcc13-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-gcc13", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-gcc13", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "13", - "CCCL_BUILD_INFIX": "cuda12.6-gcc13", + "CCCL_BUILD_INFIX": "cuda12.8-gcc13", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-gcc13" + "name": "cuda12.8-gcc13" } diff --git a/.devcontainer/cuda12.6-gcc7/devcontainer.json b/.devcontainer/cuda12.8-gcc7/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-gcc7/devcontainer.json rename to .devcontainer/cuda12.8-gcc7/devcontainer.json index 6e9aa206662..5ca28834c3b 100644 --- a/.devcontainer/cuda12.6-gcc7/devcontainer.json +++ b/.devcontainer/cuda12.8-gcc7/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc7-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-gcc7-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-gcc7", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-gcc7", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "7", - "CCCL_BUILD_INFIX": "cuda12.6-gcc7", + "CCCL_BUILD_INFIX": "cuda12.8-gcc7", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-gcc7" + "name": "cuda12.8-gcc7" } diff --git a/.devcontainer/cuda12.6-gcc8/devcontainer.json b/.devcontainer/cuda12.8-gcc8/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-gcc8/devcontainer.json rename to .devcontainer/cuda12.8-gcc8/devcontainer.json index a862b1656e5..eefd8226a3e 100644 --- a/.devcontainer/cuda12.6-gcc8/devcontainer.json +++ b/.devcontainer/cuda12.8-gcc8/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc8-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-gcc8-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-gcc8", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-gcc8", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "8", - "CCCL_BUILD_INFIX": "cuda12.6-gcc8", + "CCCL_BUILD_INFIX": "cuda12.8-gcc8", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-gcc8" + "name": "cuda12.8-gcc8" } diff --git a/.devcontainer/cuda12.6-gcc9/devcontainer.json b/.devcontainer/cuda12.8-gcc9/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-gcc9/devcontainer.json rename to .devcontainer/cuda12.8-gcc9/devcontainer.json index 39f2d58b97e..abad6fcb3b5 100644 --- a/.devcontainer/cuda12.6-gcc9/devcontainer.json +++ b/.devcontainer/cuda12.8-gcc9/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc9-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-gcc9-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-gcc9", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-gcc9", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "9", - "CCCL_BUILD_INFIX": "cuda12.6-gcc9", + "CCCL_BUILD_INFIX": "cuda12.8-gcc9", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-gcc9" + "name": "cuda12.8-gcc9" } diff --git a/.devcontainer/cuda12.6-llvm18/devcontainer.json b/.devcontainer/cuda12.8-llvm14/devcontainer.json similarity index 91% rename from .devcontainer/cuda12.6-llvm18/devcontainer.json rename to .devcontainer/cuda12.8-llvm14/devcontainer.json index 077066d1727..98e36c7f538 100644 --- a/.devcontainer/cuda12.6-llvm18/devcontainer.json +++ b/.devcontainer/cuda12.8-llvm14/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-llvm18-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-llvm14-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-llvm18", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-llvm14", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "llvm", - "CCCL_HOST_COMPILER_VERSION": "18", - "CCCL_BUILD_INFIX": "cuda12.6-llvm18", + "CCCL_HOST_COMPILER_VERSION": "14", + "CCCL_BUILD_INFIX": "cuda12.8-llvm14", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-llvm18" + "name": "cuda12.8-llvm14" } diff --git a/.devcontainer/cuda12.6-llvm15/devcontainer.json b/.devcontainer/cuda12.8-llvm15/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-llvm15/devcontainer.json rename to .devcontainer/cuda12.8-llvm15/devcontainer.json index 4cda886b473..cee653c0c00 100644 --- a/.devcontainer/cuda12.6-llvm15/devcontainer.json +++ b/.devcontainer/cuda12.8-llvm15/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-llvm15-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-llvm15-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-llvm15", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-llvm15", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "llvm", "CCCL_HOST_COMPILER_VERSION": "15", - "CCCL_BUILD_INFIX": "cuda12.6-llvm15", + "CCCL_BUILD_INFIX": "cuda12.8-llvm15", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-llvm15" + "name": "cuda12.8-llvm15" } diff --git a/.devcontainer/cuda12.6-llvm16/devcontainer.json b/.devcontainer/cuda12.8-llvm16/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-llvm16/devcontainer.json rename to .devcontainer/cuda12.8-llvm16/devcontainer.json index d85655f22cf..d7f0dce7566 100644 --- a/.devcontainer/cuda12.6-llvm16/devcontainer.json +++ b/.devcontainer/cuda12.8-llvm16/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-llvm16-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-llvm16-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-llvm16", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-llvm16", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "llvm", "CCCL_HOST_COMPILER_VERSION": "16", - "CCCL_BUILD_INFIX": "cuda12.6-llvm16", + "CCCL_BUILD_INFIX": "cuda12.8-llvm16", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-llvm16" + "name": "cuda12.8-llvm16" } diff --git a/.devcontainer/cuda12.6-llvm17/devcontainer.json b/.devcontainer/cuda12.8-llvm17/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6-llvm17/devcontainer.json rename to .devcontainer/cuda12.8-llvm17/devcontainer.json index 43d7a88e4bb..d4d3851dd85 100644 --- a/.devcontainer/cuda12.6-llvm17/devcontainer.json +++ b/.devcontainer/cuda12.8-llvm17/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-llvm17-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-llvm17-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-llvm17", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-llvm17", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "llvm", "CCCL_HOST_COMPILER_VERSION": "17", - "CCCL_BUILD_INFIX": "cuda12.6-llvm17", + "CCCL_BUILD_INFIX": "cuda12.8-llvm17", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-llvm17" + "name": "cuda12.8-llvm17" } diff --git a/.devcontainer/cuda12.8-llvm18/devcontainer.json b/.devcontainer/cuda12.8-llvm18/devcontainer.json new file mode 100644 index 00000000000..e2e27cbc08c --- /dev/null +++ b/.devcontainer/cuda12.8-llvm18/devcontainer.json @@ -0,0 +1,54 @@ +{ + "shutdownAction": "stopContainer", + "image": "rapidsai/devcontainers:25.02-cpp-llvm18-cuda12.8", + "hostRequirements": { + "gpu": "optional" + }, + "initializeCommand": [ + "/bin/bash", + "-c", + "mkdir -m 0755 -p ${localWorkspaceFolder}/.{aws,cache,config}; mkdir -m 0755 -p ${localWorkspaceFolder}/build;", + "if [[ -n ${WSLENV+set} ]]; then docker volume create cccl-build; else docker volume create --driver local --opt type=none --opt device=${localWorkspaceFolder}/build --opt o=bind cccl-build fi;" + ], + "containerEnv": { + "SCCACHE_REGION": "us-east-2", + "SCCACHE_BUCKET": "rapids-sccache-devs", + "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", + "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", + "DEVCONTAINER_NAME": "cuda12.8-llvm18", + "CCCL_CUDA_VERSION": "12.8", + "CCCL_HOST_COMPILER": "llvm", + "CCCL_HOST_COMPILER_VERSION": "18", + "CCCL_BUILD_INFIX": "cuda12.8-llvm18", + "CCCL_CUDA_EXTENDED": "false" + }, + "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", + "workspaceMount": "source=${localWorkspaceFolder},target=/home/coder/${localWorkspaceFolderBasename},type=bind,consistency=consistent", + "mounts": [ + "source=${localWorkspaceFolder}/.aws,target=/home/coder/.aws,type=bind,consistency=consistent", + "source=${localWorkspaceFolder}/.cache,target=/home/coder/.cache,type=bind,consistency=consistent", + "source=${localWorkspaceFolder}/.config,target=/home/coder/.config,type=bind,consistency=consistent", + "source=cccl-build,target=/home/coder/cccl/build" + ], + "customizations": { + "vscode": { + "extensions": [ + "llvm-vs-code-extensions.vscode-clangd", + "xaver.clang-format", + "nvidia.nsight-vscode-edition", + "ms-vscode.cmake-tools" + ], + "settings": { + "editor.defaultFormatter": "xaver.clang-format", + "editor.formatOnSave": true, + "clang-format.executable": "/usr/bin/clang-format", + "clangd.arguments": [ + "--compile-commands-dir=${workspaceFolder}" + ], + "files.eol": "\n", + "files.trimTrailingWhitespace": true + } + } + }, + "name": "cuda12.8-llvm18" +} diff --git a/.devcontainer/cuda12.6ext-gcc13/devcontainer.json b/.devcontainer/cuda12.8ext-gcc13/devcontainer.json similarity index 90% rename from .devcontainer/cuda12.6ext-gcc13/devcontainer.json rename to .devcontainer/cuda12.8ext-gcc13/devcontainer.json index 3c83bc62d95..fe118768428 100644 --- a/.devcontainer/cuda12.6ext-gcc13/devcontainer.json +++ b/.devcontainer/cuda12.8ext-gcc13/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc13-cuda12.6ext", + "image": "rapidsai/devcontainers:25.02-cpp-gcc13-cuda12.8ext", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6ext-gcc13", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8ext-gcc13", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "13", - "CCCL_BUILD_INFIX": "cuda12.6ext-gcc13", + "CCCL_BUILD_INFIX": "cuda12.8ext-gcc13", "CCCL_CUDA_EXTENDED": "true" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6ext-gcc13" + "name": "cuda12.8ext-gcc13" } diff --git a/.devcontainer/devcontainer.json b/.devcontainer/devcontainer.json index 1c7203a1a3c..0cf38c8b80b 100644 --- a/.devcontainer/devcontainer.json +++ b/.devcontainer/devcontainer.json @@ -1,6 +1,6 @@ { "shutdownAction": "stopContainer", - "image": "rapidsai/devcontainers:25.02-cpp-gcc13-cuda12.6", + "image": "rapidsai/devcontainers:25.02-cpp-gcc13-cuda12.8", "hostRequirements": { "gpu": "optional" }, @@ -15,11 +15,11 @@ "SCCACHE_BUCKET": "rapids-sccache-devs", "AWS_ROLE_ARN": "arn:aws:iam::279114543810:role/nv-gha-token-sccache-devs", "HISTFILE": "${containerWorkspaceFolder}/.cache/._bash_history", - "DEVCONTAINER_NAME": "cuda12.6-gcc13", - "CCCL_CUDA_VERSION": "12.6", + "DEVCONTAINER_NAME": "cuda12.8-gcc13", + "CCCL_CUDA_VERSION": "12.8", "CCCL_HOST_COMPILER": "gcc", "CCCL_HOST_COMPILER_VERSION": "13", - "CCCL_BUILD_INFIX": "cuda12.6-gcc13", + "CCCL_BUILD_INFIX": "cuda12.8-gcc13", "CCCL_CUDA_EXTENDED": "false" }, "workspaceFolder": "/home/coder/${localWorkspaceFolderBasename}", @@ -50,5 +50,5 @@ } } }, - "name": "cuda12.6-gcc13" + "name": "cuda12.8-gcc13" } diff --git a/.github/actions/workflow-build/build-workflow.py b/.github/actions/workflow-build/build-workflow.py index 62b90f1472d..41ea47a96a2 100755 --- a/.github/actions/workflow-build/build-workflow.py +++ b/.github/actions/workflow-build/build-workflow.py @@ -930,12 +930,19 @@ def validate_tags(matrix_job, ignore_required=False): error_message_with_matrix_job(matrix_job, f"Unknown tag '{tag}'") ) - if "gpu" in matrix_job and matrix_job["gpu"] not in matrix_yaml["gpus"].keys(): - raise Exception( - error_message_with_matrix_job( - matrix_job, f"Unknown gpu '{matrix_job['gpu']}'" - ) + if "gpu" in matrix_job: + gpus = ( + matrix_job["gpu"] + if isinstance(matrix_job["gpu"], list) + else [matrix_job["gpu"]] ) + for gpu in gpus: + if gpu not in matrix_yaml["gpus"].keys(): + raise Exception( + error_message_with_matrix_job( + matrix_job, f"Unknown gpu '{matrix_job['gpu']}'" + ) + ) def set_default_tags(matrix_job): diff --git a/.github/workflows/build-matx.yml b/.github/workflows/build-matx.yml index 0c6ef548555..3544b8823c5 100644 --- a/.github/workflows/build-matx.yml +++ b/.github/workflows/build-matx.yml @@ -61,7 +61,7 @@ jobs: .devcontainer/launch.sh \ --docker \ - --cuda 12.6 \ + --cuda 12.8 \ --host gcc13 \ --cuda-ext \ --env VAULT_HOST= \ diff --git a/ci/matrix.yaml b/ci/matrix.yaml index 5ec715fb59b..dff3b75df8e 100644 --- a/ci/matrix.yaml +++ b/ci/matrix.yaml @@ -19,26 +19,24 @@ workflows: - {jobs: ['build'], std: 'max', cxx: ['msvc2019']} - {jobs: ['build'], std: 'all', cxx: ['gcc', 'clang', 'msvc']} # Current CTK testing: - - {jobs: ['test'], project: ['thrust'], std: 'max', cxx: ['gcc', 'clang'], gpu: 'rtx4090'} - - {jobs: ['test'], project: ['libcudacxx'], std: 'max', cxx: ['gcc', 'clang'], gpu: 'rtx2080'} - # Disabled until we figure out the issue with the TBB dll - #- {jobs: ['test'], project: ['thrust'], std: 'max', cxx: ['msvc'], gpu: 'rtx4090'} - - {jobs: ['test'], project: ['libcudacxx'], std: 'max', cxx: ['msvc'], gpu: 'rtx2080'} + - {jobs: ['test'], project: ['thrust'], std: 'max', cxx: ['gcc', 'clang', 'msvc'], gpu: 'rtx4090'} + - {jobs: ['test'], project: ['libcudacxx'], std: 'max', cxx: ['gcc', 'clang', 'msvc'], gpu: 'rtx2080'} # Split up cub tests: - {jobs: ['test_nolid', 'test_lid0'], project: ['cub'], std: 'max', cxx: ['gcc'], gpu: 'rtxa6000'} - {jobs: ['test_lid1', 'test_lid2'], project: ['cub'], std: 'max', cxx: ['gcc'], gpu: 'rtxa6000'} - {jobs: ['test_nolid', 'test_lid0'], project: ['cub'], std: 'max', cxx: ['clang', 'msvc'], gpu: 'rtxa6000'} - - {jobs: ['test_lid0'], project: ['cub'], std: 'max', cxx: 'gcc12', gpu: 'h100', sm: 'gpu' } # Modded builds: - {jobs: ['build'], std: 'all', ctk: '12.5', cxx: 'nvhpc'} - {jobs: ['build'], std: 'max', cxx: ['gcc', 'clang'], cpu: 'arm64'} - - {jobs: ['build'], std: 'max', cxx: ['gcc'], sm: '90a'} + - {jobs: ['build'], std: 'max', cxx: 'gcc', sm: ['90;90a;100']} + - {jobs: ['test_lid0'], project: 'cub', std: 'max', cxx: 'gcc', gpu: 'h100', sm: 'gpu' } # Test Thrust 32-bit-only dispatch here, since it's most likely to break. 64-bit-only is tested in nightly. - {jobs: ['test_gpu'], project: 'thrust', cmake_options: '-DTHRUST_DISPATCH_TYPE=Force32bit', gpu: 'rtx4090'} # default_projects: clang-cuda - {jobs: ['build'], std: 'all', cudacxx: 'clang', cxx: 'clang'} - - {jobs: ['build'], project: 'libcudacxx', std: 'max', cudacxx: 'clang', cxx: 'clang', sm: '90'} - - {jobs: ['build'], project: 'libcudacxx', std: 'max', cudacxx: 'clang', cxx: 'clang', sm: '90a'} + # Disabled; see discussion on #3633. Should be fixed in clang-20. +# - {jobs: ['build'], project: 'libcudacxx', std: 'max', cudacxx: 'clang', cxx: 'clang', sm: '90'} +# - {jobs: ['build'], project: 'libcudacxx', std: 'max', cudacxx: 'clang', cxx: 'clang', sm: '90a'} # nvrtc: - {jobs: ['nvrtc'], project: 'libcudacxx', std: 'all', gpu: 'rtx2080', sm: 'gpu'} # verify-codegen: @@ -54,7 +52,7 @@ workflows: - {jobs: ['build'], project: 'cudax', ctk: ['curr'], std: 'all', cxx: ['gcc', 'clang'], cpu: 'arm64'} - {jobs: ['test'], project: 'cudax', ctk: ['curr'], std: 20, cxx: ['gcc12', 'clang', 'msvc'], gpu: 'rtx2080'} # Python and c/parallel jobs: - - {jobs: ['test'], project: ['cccl_c_parallel', 'python'], ctk: '12.6', gpu: 'rtx2080'} + - {jobs: ['test'], project: ['cccl_c_parallel', 'python'], gpu: 'rtx2080'} # cccl-infra: - {jobs: ['infra'], project: 'cccl', ctk: '12.0', cxx: ['gcc12', 'clang14'], gpu: 'rtx2080'} - {jobs: ['infra'], project: 'cccl', ctk: 'curr', cxx: ['gcc', 'clang'], gpu: 'rtx2080'} @@ -73,18 +71,18 @@ workflows: - {jobs: ['build'], std: 'all', cxx: ['msvc2019']} # Test current CTK - {jobs: ['test'], project: 'cub', std: 'all', cxx: ['gcc', 'clang', 'msvc'], gpu: 'rtxa6000'} - - {jobs: ['test_lid0'], project: 'cub', std: 'max', cxx: 'gcc', gpu: 'v100'} - - {jobs: ['test_lid0'], project: 'cub', std: 'max', cxx: 'gcc', gpu: 'h100', sm: 'gpu' } - {jobs: ['test'], project: 'thrust', std: 'all', cxx: ['gcc', 'clang', 'msvc'], gpu: 'rtx4090'} - {jobs: ['test'], project: 'libcudacxx', std: 'all', cxx: ['gcc', 'clang', 'msvc'], gpu: 'rtx2080'} # Modded builds: - {jobs: ['build'], std: 'all', ctk: '12.5', cxx: 'nvhpc'} - {jobs: ['build'], std: 'all', cxx: ['gcc', 'clang'], cpu: 'arm64'} - {jobs: ['build'], std: 'all', cxx: ['gcc'], sm: '90a'} + - {jobs: ['test_lid0'], project: 'cub', std: 'max', cxx: 'gcc', gpu: ['v100', 't4', 'l4', 'h100'], sm: '70;75;89;90;100'} # default_projects: clang-cuda - {jobs: ['build'], std: 'all', cudacxx: 'clang', cxx: 'clang'} - - {jobs: ['build'], project: 'libcudacxx', std: 'all', cudacxx: 'clang', cxx: 'clang', sm: '90'} - - {jobs: ['build'], project: 'libcudacxx', std: 'all', cudacxx: 'clang', cxx: 'clang', sm: '90a'} + # Disabled; see discussion on #3633. Should be fixed in clang-20. + # - {jobs: ['build'], project: 'libcudacxx', std: 'all', cudacxx: 'clang', cxx: 'clang', sm: '90'} + # - {jobs: ['build'], project: 'libcudacxx', std: 'all', cudacxx: 'clang', cxx: 'clang', sm: '90a'} # cudax - {jobs: ['build'], project: 'cudax', ctk: ['12.0', 'curr'], std: 'all', cxx: ['gcc9', 'gcc10', 'gcc11']} - {jobs: ['build'], project: 'cudax', ctk: [ 'curr'], std: 'all', cxx: ['clang14', 'clang15', 'clang16', 'clang17']} @@ -116,7 +114,7 @@ all_stds: [17, 20] ctk_versions: 12.0: { stds: [17, 20] } 12.5: { stds: [17, 20] } - 12.6: { stds: [17, 20], aka: 'curr' } + 12.8: { stds: [17, 20], aka: 'curr' } device_compilers: nvcc: # Version / stds are taken from CTK @@ -269,7 +267,7 @@ tags: # CPU architecture cpu: { default: 'amd64' } # GPU model - gpu: { default: 'v100' } + gpu: { default: 'rtx2080' } # Host compiler {name, version, exe} # See the `host_compilers` map. cxx: { default: 'gcc' } diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/nvrtc_workaround.h b/libcudacxx/test/libcudacxx/cuda/ptx/nvrtc_workaround.h index fef34f25ef4..2e049bc6217 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/nvrtc_workaround.h +++ b/libcudacxx/test/libcudacxx/cuda/ptx/nvrtc_workaround.h @@ -13,6 +13,25 @@ // required for the PTX tests, we define them here outside the header guards. // TODO(bgruber): limit this workaround to NVRTC versions older than the first one shipping those macros #ifdef __CUDACC_RTC__ + +// missing SM_100 +# define _NV_TARGET_VAL_SM_100 1000 +# define _NV_TARGET___NV_PROVIDES_SM_100 (_NV_TARGET_PROVIDES(_NV_TARGET_VAL_SM_100)) +# define _NV_TARGET___NV_IS_EXACTLY_SM_100 (_NV_TARGET_IS_EXACTLY(_NV_TARGET_VAL_SM_100)) +# define NV_PROVIDES_SM_100 __NV_PROVIDES_SM_100 +# define NV_IS_EXACTLY_SM_100 __NV_IS_EXACTLY_SM_100 +# if (_NV_TARGET___NV_IS_EXACTLY_SM_100) +# define _NV_TARGET_BOOL___NV_IS_EXACTLY_SM_100 1 +# else +# define _NV_TARGET_BOOL___NV_IS_EXACTLY_SM_100 0 +# endif +# if (_NV_TARGET___NV_PROVIDES_SM_100) +# define _NV_TARGET_BOOL___NV_PROVIDES_SM_100 1 +# else +# define _NV_TARGET_BOOL___NV_PROVIDES_SM_100 0 +# endif + +// missing SM_100a # ifndef NV_HAS_FEATURE_SM_100a # define NV_HAS_FEATURE_SM_100a __NV_HAS_FEATURE_SM_100a # if (defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 1000) && defined(__CUDA_ARCH_FEAT_SM100_ALL)) @@ -22,7 +41,7 @@ # endif # endif // NV_HAS_FEATURE_SM_100a -// Re-enable sm_101a support in nvcc. +// missing SM_101a # ifndef NV_HAS_FEATURE_SM_101a # define NV_HAS_FEATURE_SM_101a __NV_HAS_FEATURE_SM_101a # if (defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 1010) && defined(__CUDA_ARCH_FEAT_SM101_ALL)) @@ -31,4 +50,5 @@ # define _NV_TARGET_BOOL___NV_HAS_FEATURE_SM_101a 0 # endif # endif // NV_HAS_FEATURE_SM_101a + #endif // __CUDACC_RTC__ diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.clusterlaunchcontrol.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.clusterlaunchcontrol.compile.pass.cpp index 212414c4535..bc4e3b1b9d4 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.clusterlaunchcontrol.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.clusterlaunchcontrol.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/clusterlaunchcontrol.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.st.bulk.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.st.bulk.compile.pass.cpp index 951e1a9f513..56c9c9027ec 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.st.bulk.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.st.bulk.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/st_bulk.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.alloc.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.alloc.compile.pass.cpp index 49f9df928e9..6ba6bb390b9 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.alloc.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.alloc.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_alloc.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.commit.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.commit.compile.pass.cpp index 73ea1851bec..1408b505381 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.commit.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.commit.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_commit.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.cp.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.cp.compile.pass.cpp index 85ddc17efe4..133e47b9d03 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.cp.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.cp.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_cp.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.fence.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.fence.compile.pass.cpp index fda57b348de..1f0f830572b 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.fence.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.fence.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_fence.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.ld.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.ld.compile.pass.cpp index 8da8e54f18d..5b407140162 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.ld.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.ld.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_ld.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.mma.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.mma.compile.pass.cpp index 098cbbfa896..082bc0ceb4b 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.mma.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.mma.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_mma.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.mma.ws.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.mma.ws.compile.pass.cpp index 350c964d749..272710c3601 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.mma.ws.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.mma.ws.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_mma_ws.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.shift.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.shift.compile.pass.cpp index 5ecfff7ff3b..5079b71d6ab 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.shift.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.shift.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_shift.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.st.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.st.compile.pass.cpp index 92a49224f0e..fa214acb37e 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.st.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.st.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_st.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.wait.compile.pass.cpp b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.wait.compile.pass.cpp index 4bb3156ed12..ef368bf0a02 100644 --- a/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.wait.compile.pass.cpp +++ b/libcudacxx/test/libcudacxx/cuda/ptx/ptx.tcgen05.wait.compile.pass.cpp @@ -14,6 +14,8 @@ #include #include +#include "nvrtc_workaround.h" +// above header needs to be included before the generated test header #include "generated/tcgen05_wait.h" int main(int, char**) diff --git a/libcudacxx/test/libcudacxx/std/utilities/meta/meta.unary/meta.unary.prop/is_constructible.pass.cpp b/libcudacxx/test/libcudacxx/std/utilities/meta/meta.unary/meta.unary.prop/is_constructible.pass.cpp index c9e544789be..a766e4ad038 100644 --- a/libcudacxx/test/libcudacxx/std/utilities/meta/meta.unary/meta.unary.prop/is_constructible.pass.cpp +++ b/libcudacxx/test/libcudacxx/std/utilities/meta/meta.unary/meta.unary.prop/is_constructible.pass.cpp @@ -146,8 +146,8 @@ template __host__ __device__ void test_is_not_constructible() { static_assert((!cuda::std::is_constructible::value), ""); -#if !defined(TEST_COMPILER_MSVC) && !(defined(TEST_COMPILER_CLANG) && __clang_major__ >= 16) - // The fallback SFINAE version doesn't work reliable with MSVC, and we don't +#if !defined(TEST_COMPILER_MSVC) && !defined(TEST_COMPILER_CLANG) && !defined(TEST_COMPILER_NVRTC) + // The fallback SFINAE version doesn't work reliable with Clang/MSVC/NVRTC, and we don't // use it, so waive it. static_assert((!cuda::std::__cccl_is_constructible::type::value), ""); #endif