[Ops][Refactor] Remove custom rotary_embedding operator (#6523)

### What this PR does / why we need it?
This PR removes the custom `rotary_embedding` operator and its
associated C++ kernel implementation, PyTorch bindings, and tests.

The codebase now falls back to using the native
`torch_npu._npu_rotary_embedding` implementation. This change simplifies
the codebase by removing custom, platform-specific kernel code and
relying on the standard NPU library implementation, which is presumably
more optimized and easier to maintain.

### Does this PR introduce _any_ user-facing change?
No. This is an internal refactoring and does not introduce any
user-facing changes.

### How was this patch tested?
The tests for the custom `rotary_embedding` operator have been removed
along with the operator itself. The correctness of the fallback to the
native `torch_npu` implementation is verified by existing CI tests for
attention layers and models that use rotary embeddings.

- vLLM version: v0.15.0
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.15.0

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2026-02-07 09:24:05 +08:00
committed by GitHub
parent 06aa6036f6
commit 6c49f95da2
8 changed files with 59 additions and 1392 deletions

View File

@@ -24,13 +24,6 @@
#include "torch_npu/csrc/aten/common/from_blob.h"
namespace vllm_ascend {
extern void rotary_embedding_impl(AscendType type, bool isNeox, void *stream, int64_t *positions, void *queryDst,
void *keyDst, void *query, void *key, void *cosSinCache, const int rotDim,
const int64_t queryStride, const int64_t keyStride, const int64_t dstQueryStride,
const int64_t dstKeyStride, const int numHeads, const int numKvHeads,
const int headSize, const int64_t numTokens, const uint32_t loopCnt,
uint32_t aivNum);
extern void get_masked_input_and_mask_impl(
void* stream,
void* input,