Gahow Wang 445e491123 Add vLLM v0.18.1 source tree with KV transfer abort fix ...

third_party/vllm/ now tracked in git for direct patch management.
Based on vLLM v0.18.1 release with one patch applied:

  vllm/v1/core/sched/scheduler.py:
    Replace fatal assert with graceful skip when KV transfer callback
    arrives for an already-aborted request during PD disaggregated serving.

Future vLLM modifications should be made directly in third_party/vllm/
and committed normally. The patches/ directory is kept as documentation
of what changed from upstream.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-05-22 00:30:38 +08:00

..

attention

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

core

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cpu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cutlass_extensions

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

mamba/mamba_ssm

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

moe

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

quantization

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

quickreduce

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

rocm

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

sparse/cutlass

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

activation_kernels.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cache_kernels_fused.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cache_kernels.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cache.h

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

concat_mla_q.cuh

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cub_helpers.h

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cuda_compat.h

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cuda_utils_kernels.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cuda_utils.h

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cuda_vec_utils.cuh

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cuda_view.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cumem_allocator_compat.h

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

cumem_allocator.cpp

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

custom_all_reduce_test.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

custom_all_reduce.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

custom_all_reduce.cuh

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

custom_quickreduce.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

dispatch_utils.h

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

dsv3_fused_a_gemm.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

fused_qknorm_rope_kernel.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

launch_bounds_utils.h

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

layernorm_kernels.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

layernorm_quant_kernels.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

ops.h

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

permute_cols.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

pos_encoding_kernels.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

sampler.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

topk.cu

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

torch_bindings.cpp

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00

type_convert.cuh

Add vLLM v0.18.1 source tree with KV transfer abort fix

2026-05-22 00:30:38 +08:00