Migrate header files to torch stable abi (#44013)

This commit is contained in:
Chris Leonard
2026-06-02 08:09:52 -07:00
committed by GitHub
parent 586201ebdc
commit 4d93bc35c9
18 changed files with 13 additions and 16 deletions
+1 -1
View File
@@ -21,7 +21,7 @@ repos:
rev: v21.1.2 rev: v21.1.2
hooks: hooks:
- id: clang-format - id: clang-format
exclude: 'csrc/(moe/topk_softmax_kernels.cu|quantization/gguf/(ggml-common.h|dequantize.cuh|vecdotq.cuh|mmq.cuh|mmvq.cuh))|vllm/third_party/.*' exclude: 'csrc/(moe/topk_softmax_kernels.cu|libtorch_stable/quantization/gguf/(ggml-common.h|dequantize.cuh|vecdotq.cuh|mmq.cuh|mmvq.cuh))|vllm/third_party/.*'
types_or: [c++, cuda] types_or: [c++, cuda]
args: [--style=file, --verbose] args: [--style=file, --verbose]
- repo: https://github.com/DavidAnson/markdownlint-cli2 - repo: https://github.com/DavidAnson/markdownlint-cli2
@@ -2,7 +2,7 @@
#include <torch/csrc/stable/tensor.h> #include <torch/csrc/stable/tensor.h>
#include "cutlass_extensions/epilogue/broadcast_load_epilogue_c2x.hpp" #include "broadcast_load_epilogue_c2x.hpp"
/* /*
This file defines custom epilogues for fusing channel scales, token scales, This file defines custom epilogues for fusing channel scales, token scales,
@@ -20,7 +20,7 @@
#include "torch_utils.h" #include "torch_utils.h"
#include "../async_util.cuh" #include "async_util.cuh"
#include "../cuda_compat.h" #include "../cuda_compat.h"
#include "../type_convert.cuh" #include "../type_convert.cuh"
#include "dispatch_utils.h" #include "dispatch_utils.h"
@@ -25,7 +25,7 @@
#include <cuda_fp8.h> #include <cuda_fp8.h>
#include "cuda_utils.h" #include "cuda_utils.h"
#include "launch_bounds_utils.h" #include "libtorch_stable/launch_bounds_utils.h"
// Define before including nvfp4_utils.cuh so the header // Define before including nvfp4_utils.cuh so the header
// can use this macro during compilation. // can use this macro during compilation.
@@ -34,7 +34,7 @@
static_assert(CVT_FP4_ELTS_PER_THREAD == 16, static_assert(CVT_FP4_ELTS_PER_THREAD == 16,
"MXFP4 experts quant requires PACK16 mode (CUDA >= 12.9)"); "MXFP4 experts quant requires PACK16 mode (CUDA >= 12.9)");
#include "launch_bounds_utils.h" #include "libtorch_stable/launch_bounds_utils.h"
namespace vllm { namespace vllm {
@@ -26,7 +26,7 @@
#include "cuda_utils.h" #include "cuda_utils.h"
#include "nvfp4_utils.cuh" #include "nvfp4_utils.cuh"
#include "launch_bounds_utils.h" #include "libtorch_stable/launch_bounds_utils.h"
namespace vllm { namespace vllm {
@@ -26,7 +26,7 @@
#include "../../cuda_vec_utils.cuh" #include "../../cuda_vec_utils.cuh"
#include "cuda_utils.h" #include "cuda_utils.h"
#include "launch_bounds_utils.h" #include "libtorch_stable/launch_bounds_utils.h"
// Define before including nvfp4_utils.cuh so the header // Define before including nvfp4_utils.cuh so the header
// can use this macro during compilation. // can use this macro during compilation.
@@ -7,14 +7,11 @@
#include <torch/csrc/stable/ops.h> #include <torch/csrc/stable/ops.h>
// NOTE: These headers are intentionally kept in csrc/quantization/gguf/ (not #include "ggml-common.h"
// moved to libtorch_stable) to avoid unnecessary reformatting that would break #include "vecdotq.cuh"
// git rename detection and pollute blame history. #include "dequantize.cuh"
#include "../../../quantization/gguf/ggml-common.h" #include "mmvq.cuh"
#include "../../../quantization/gguf/vecdotq.cuh" #include "mmq.cuh"
#include "../../../quantization/gguf/dequantize.cuh"
#include "../../../quantization/gguf/mmvq.cuh"
#include "../../../quantization/gguf/mmq.cuh"
#include "moe.cuh" #include "moe.cuh"
#include "moe_vec.cuh" #include "moe_vec.cuh"
+1 -1
View File
@@ -7,7 +7,7 @@
#include "torch_utils.h" #include "torch_utils.h"
#ifndef USE_ROCM #ifndef USE_ROCM
#include "../persistent_topk.cuh" #include "persistent_topk.cuh"
#endif #endif
namespace { namespace {