[TRTLLM-14810][chore] Pre-flight: sync sa_worker.py with main (#16759); require packaging>=24.2

docs/source/deployment-guide/deployment-guide-for-kimi-k3-on-trtllm.md

@@ -56,13 +56,12 @@ Kimi K3 additionally depends on `fla` and `einops`, installed into the same in-p
 To use the optimized CuTeDSL MLA kernel, install the FlashInfer revision used by the Kimi K3 example into the same in-place environment:
 
 ```bash
-.venv-3.12/bin/python -m pip install -U 'packaging>=24.2'  # required by the FlashInfer source build
 .venv-3.12/bin/python -u -m pip install --force-reinstall --no-deps \
     --no-build-isolation \
     "flashinfer-python[cu13] @ git+https://github.com/PerkzZheng/flashinfer-k3.git@b6cc594918baf76c40c3a6236fd53f0f8fb9d2dc"
 ```
 
-The TensorRT LLM environment already provides FlashInfer's runtime dependencies. The `--no-deps` option prevents `pip` from replacing the pinned PyTorch, Triton, CUDA, and CuTeDSL packages. Install FlashInfer after TensorRT LLM because a later dependency-resolving TensorRT LLM installation can replace this source revision with the currently pinned `flashinfer-python==0.6.14`.
+The `packaging>=24.2` requirement of this source build is already satisfied by `requirements.txt`. The TensorRT LLM environment already provides FlashInfer's runtime dependencies. The `--no-deps` option prevents `pip` from replacing the pinned PyTorch, Triton, CUDA, and CuTeDSL packages. Install FlashInfer after TensorRT LLM because a later dependency-resolving TensorRT LLM installation can replace this source revision with the currently pinned `flashinfer-python==0.6.14`.
 
 For general build-from-source instructions see [https://nvidia.github.io/TensorRT-LLM/latest/installation/build-from-source.html](https://nvidia.github.io/TensorRT-LLM/latest/installation/build-from-source.html).
 

examples/kimi_k3/README.md

@@ -41,13 +41,14 @@ other GPU architectures may be added in a future release.
   revision into the same in-place environment after installing TensorRT-LLM:
 
   ```bash
-  .venv-3.12/bin/python -m pip install -U 'packaging>=24.2'  # required by the FlashInfer source build
   .venv-3.12/bin/python -u -m pip install --force-reinstall --no-deps \
       --no-build-isolation \
       "flashinfer-python[cu13] @ git+https://github.com/PerkzZheng/flashinfer-k3.git@b6cc594918baf76c40c3a6236fd53f0f8fb9d2dc"
   ```
 
-  The TensorRT-LLM environment already provides FlashInfer's runtime
+  The `packaging>=24.2` requirement of this source build is already
+  satisfied by `requirements.txt`. The TensorRT-LLM environment already
+  provides FlashInfer's runtime
   dependencies; `--no-deps` prevents pip from replacing its pinned PyTorch,
   Triton, CUDA, and CuTeDSL packages. Install FlashInfer last: TensorRT-LLM
   currently pins `flashinfer-python==0.6.14`, so a later

requirements.txt

@@ -52,6 +52,8 @@ fastapi>=0.120.1,<=0.121.3
 starlette>=0.49.1
 uvicorn
 setuptools<80
+# FlashInfer source builds (--no-build-isolation) need packaging>=24.2 for setuptools license-expression validation
+packaging>=24.2
 ordered-set
 peft>=0.18.1,<0.19.0
 patchelf

tensorrt_llm/_torch/speculative/sa_worker.py

@@ -188,9 +188,7 @@ def forward(
         # one-engine workers (dflash/eagle3); no-op for pure-attention
         # models via the isinstance gate.
         num_gens = batch_size - num_contexts
-        if num_gens > 0 and isinstance(
-            attn_metadata.kv_cache_manager, MambaHybridCacheManager
-        ):
+        if num_gens > 0 and isinstance(attn_metadata.kv_cache_manager, MambaHybridCacheManager):
             attn_metadata.kv_cache_manager.update_mamba_states(
                 attn_metadata=attn_metadata,
                 num_accepted_tokens=num_accepted_tokens,