feat: add version 6 config

.github/workflows/test_e2e.yaml

@@ -82,7 +82,7 @@ jobs:
           test -f example.py
           grep -q 'name = "${{ env.E2E_REPO_NAME }}"' build.toml
           grep -q 'repo-id = "${{ env.E2E_REPO_ID }}"' build.toml
-          grep -q 'backend = "cuda"' build.toml
+          grep -q 'language = "cpp-cuda"' build.toml
 
       - name: Patch flake.nix to use local nix-builder
         run: |

docs/source/builder/agents-guide.md

@@ -97,7 +97,7 @@ version = 1
 src = ["torch-ext/torch_binding.cpp"]
 
 [kernel.rmsnorm]
-backend = "cuda"
+language = "cpp-cuda"
 src = ["kernel_src/rmsnorm.cu"]
 depends = ["torch"]
 cuda-capabilities = ["9.0"]  # H100

docs/source/builder/triton-autotune.md

@@ -37,7 +37,7 @@ files are picked up from the kenel's Python source directory, so add `json` to t
 [general]
 name = "gemm-triton-autotune"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = ["cuda", "rocm", "xpu"]
 

docs/source/builder/writing-kernels.md

@@ -167,7 +167,7 @@ backends = [
 ]
 name = "mykernel"
 version = 1
-edition = 5
+edition = 6
 
 [general.hub]
 repo-id = "myorg/mykernel"
@@ -179,7 +179,7 @@ src = [
 ]
 
 [kernel.mykernel]
-backend = "cuda"
+language = "cpp-cuda"
 depends = ["torch"]
 src = ["mykernel_cuda/mykernel.cu"]
 # If the kernel is only supported on specific capabilities, set the
@@ -199,7 +199,7 @@ The following sections enumerate all supported options for `build.toml`.
   by the `kernel-builder upload` command to upload the kernel to a version
   branch named `v<version>`.
 - `edition` (required): the `build.toml` format edition. The current
-  edition is `5`. Older `build.toml` files can be migrated with
+  edition is `6`. Older `build.toml` files can be migrated with
   `kernel-builder update-build`.
 - `backends` (required): a list of supported backends. Must be one or
   more of `cpu`, `cuda`, `metal`, `rocm`, or `xpu`.
@@ -316,8 +316,9 @@ for an example with multiple kernel sections.
 
 The following options can be set for a kernel:
 
-- `backend` (required): the compute backend of the kernel. The currently
-  supported backends are `cpu`, `cuda`, `metal`, `rocm`, and `xpu`.
+- `language` (required): the language and compute backend of the kernel,
+  as `<language>-<backend>`. The currently supported values are `cpp-cpu`,
+  `cpp-cuda`, `cpp-metal`, `cpp-rocm`, and `cpp-xpu`.
   **The `cpu` backend is currently experimental and might still change.**
 - `depends` (required): a list of dependencies. The supported dependencies
   are listed in [`cpp-deps.nix`](https://github.com/huggingface/kernels/blob/main/builder/lib/cpp-deps.nix).

examples/kernels/cpp20-symbols/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "cpp20-symbols"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = ["cpu"]
 
@@ -15,6 +15,6 @@ src = [
 ]
 
 [kernel.cpp20_symbols_cpu]
-backend = "cpu"
+language = "cpp-cpu"
 depends = ["torch"]
 src = ["cpu/cpu.cpp"]

examples/kernels/cutlass-gemm-tvm-ffi/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "cutlass-gemm-tvm-ffi"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",
@@ -15,15 +15,15 @@ repo-id = "kernels-test/cutlass-gemm-tvm-ffi"
 src = ["tvm-ffi-ext/tvm_ffi_binding.cpp"]
 
 [kernel.gemm]
-backend = "cuda"
+language = "cpp-cuda"
 depends = ["cutlass_3_6"]
 src = [
     "gemm.cu",
     "util.hh",
 ]
 
 [kernel.gemm_xpu]
-backend = "xpu"
+language = "cpp-xpu"
 depends = ["sycl_tla"]
 src = [
     "gemm_sycl.cpp",

examples/kernels/cutlass-gemm/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "cutlass-gemm"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",
@@ -18,15 +18,15 @@ src = [
 ]
 
 [kernel.gemm_xpu]
-backend = "xpu"
+language = "cpp-xpu"
 depends = [
     "torch",
     "sycl_tla",
 ]
 src = ["gemm_sycl.cpp"]
 
 [kernel.gemm]
-backend = "cuda"
+language = "cpp-cuda"
 depends = [
     "torch",
     "cutlass_3_6",

examples/kernels/gemm-triton-autotune/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "gemm-triton-autotune"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",

examples/kernels/kernel-deps/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "kernel-deps"
 version = 1
-edition = 5
+edition = 6
 license = "apache-2.0"
 backends = [
     "cpu",

examples/kernels/relu-archs-subset/build.toml

@@ -1,6 +1,7 @@
 [general]
 name = "relu-archs-subset"
 version = 1
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",
@@ -23,7 +24,7 @@ src = [
 # supported sets.
 
 [kernel.relu_rocm]
-backend = "rocm"
+language = "cpp-rocm"
 depends = ["torch"]
 rocm-archs = [
     "gfx90a",
@@ -34,7 +35,7 @@ rocm-archs = [
 src = ["relu_cuda/relu.cu"]
 
 [kernel.relu]
-backend = "cuda"
+language = "cpp-cuda"
 depends = ["torch"]
 cuda-capabilities = [
     "5.0",

examples/kernels/relu-backprop-compile/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu-backprop-compile"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",
@@ -18,12 +18,12 @@ src = [
 ]
 
 [kernel.relu]
-backend = "cuda"
+language = "cpp-cuda"
 depends = ["torch"]
 src = ["relu_cuda/relu.cu"]
 
 [kernel.relu_rocm]
-backend = "rocm"
+language = "cpp-rocm"
 depends = ["torch"]
 rocm-archs = [
     "gfx906",

examples/kernels/relu-compiler-flags/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu-compiler-flags"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",
@@ -20,19 +20,19 @@ src = [
 cxx-flags = ["-DCANARY_IN_THE_KERNEL"]
 
 [kernel.activation_xpu]
-backend = "xpu"
+language = "cpp-xpu"
 depends = ["torch"]
 sycl-flags = ["-DWHO_AM_I_IF_NOT_THE_CANARY"]
 src = ["relu_xpu/relu.cpp"]
 
 [kernel.activation]
-backend = "cuda"
+language = "cpp-cuda"
 cuda-flags = ["-DWHO_AM_I_IF_NOT_THE_CANARY"]
 depends = ["torch"]
 src = ["relu_cuda/relu.cu"]
 
 [kernel.activation_rocm]
-backend = "rocm"
+language = "cpp-rocm"
 depends = ["torch"]
 rocm-archs = [
     "gfx906",

examples/kernels/relu-invalid-capability/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu-invalid-capability"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cpu",
@@ -21,13 +21,13 @@ src = [
 ]
 
 [kernel.relu_rocm]
-backend = "rocm"
+language = "cpp-rocm"
 depends = ["torch"]
 rocm-archs = [ "gfx99999" ]
 src = ["relu_cuda/relu.cu"]
 
 [kernel.relu]
-backend = "cuda"
+language = "cpp-cuda"
 depends = ["torch"]
 cuda-capabilities = [ "99999.0" ]
 src = ["relu_cuda/relu.cu"]

examples/kernels/relu-metal-cpp/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = ["metal"]
 
@@ -15,7 +15,7 @@ src = [
 ]
 
 [kernel.relu_metal]
-backend = "metal"
+language = "cpp-metal"
 depends = [
     "torch",
     "metal-cpp",

examples/kernels/relu-nki/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu-nki"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = ["neuron"]
 

examples/kernels/relu-specific-torch/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu-specific-torch"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",
@@ -18,12 +18,12 @@ src = [
 ]
 
 [kernel.relu]
-backend = "cuda"
+language = "cpp-cuda"
 depends = ["torch"]
 src = ["relu_cuda/relu.cu"]
 
 [kernel.relu_rocm]
-backend = "rocm"
+language = "cpp-rocm"
 depends = ["torch"]
 rocm-archs = [
     "gfx906",

examples/kernels/relu-torch-bounds/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",
@@ -20,12 +20,12 @@ src = [
 ]
 
 [kernel.relu]
-backend = "cuda"
+language = "cpp-cuda"
 depends = ["torch"]
 src = ["relu_cuda/relu.cu"]
 
 [kernel.relu_rocm]
-backend = "rocm"
+language = "cpp-rocm"
 depends = ["torch"]
 rocm-archs = [
     "gfx906",

examples/kernels/relu-torch-stable-abi/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu-torch-stable-abi"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cpu",
@@ -28,7 +28,7 @@ rocm = "2.11"
 xpu = "2.11"
 
 [kernel.relu_xpu]
-backend = "xpu"
+language = "cpp-xpu"
 depends = ["torch"]
 src = ["relu_xpu/relu.cpp"]
 
@@ -46,7 +46,7 @@ src = ["relu_xpu/relu.cpp"]
 #]
 
 [kernel.relu_rocm]
-backend = "rocm"
+language = "cpp-rocm"
 depends = ["torch"]
 rocm-archs = [
     "gfx906",
@@ -62,11 +62,11 @@ rocm-archs = [
 src = ["relu_cuda/relu.cu"]
 
 [kernel.relu_cpu]
-backend = "cpu"
+language = "cpp-cpu"
 depends = ["torch"]
 src = ["relu_cpu/relu_cpu.cpp"]
 
 [kernel.relu]
-backend = "cuda"
+language = "cpp-cuda"
 depends = ["torch"]
 src = ["relu_cuda/relu.cu"]

examples/kernels/relu-tpu/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu-tpu"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = ["tpu"]
 

examples/kernels/relu-triton/build.toml

@@ -1,7 +1,7 @@
 [general]
 name = "relu-triton"
 version = 1
-edition = 5
+edition = 6
 license = "Apache-2.0"
 backends = [
     "cuda",