chore: bump sgl-kernel version to 0.3.21 (#17075)

Co-authored-by: sglang-bot <sglang-bot@users.noreply.github.com>

docker/Dockerfile

@@ -11,7 +11,7 @@ ARG GRACE_BLACKWELL_DEEPEP_BRANCH=gb200_blog_part_2
 ARG HOPPER_SBO_DEEPEP_COMMIT=9f2fc4b3182a51044ae7ecb6610f7c9c3258c4d6
 ARG DEEPEP_COMMIT=9af0e0d0e74f3577af1979c9b9e1ac2cad0104ee
 ARG BUILD_AND_DOWNLOAD_PARALLEL=8
-ARG SGL_KERNEL_VERSION=0.3.20
+ARG SGL_KERNEL_VERSION=0.3.21
 ARG SGL_VERSION
 ARG USE_LATEST_SGLANG=0
 ARG GDRCOPY_VERSION=2.5.1

python/pyproject.toml

@@ -17,6 +17,7 @@ classifiers = [
 dependencies = [
   "IPython",
   "aiohttp",
+  "apache-tvm-ffi>=0.1.5,<0.2",
   "anthropic>=0.20.0",
   "av ; sys_platform == 'linux' and (platform_machine == 'aarch64' or platform_machine == 'arm64' and platform_machine == 'armv7l')",
   "blobfile==3.0.0",
@@ -38,7 +39,7 @@ dependencies = [
   "msgspec",
   "ninja",
   "numpy",
-  "nvidia-cutlass-dsl==4.3.1",
+  "nvidia-cutlass-dsl>=4.3.4",
   "nvidia-ml-py",
   "openai-harmony==0.0.4",
   "openai==2.6.1",
@@ -54,11 +55,12 @@ dependencies = [
   "pydantic",
   "python-multipart",
   "pyzmq>=25.1.2",
+  "quack-kernels==0.2.4",
   "requests",
   "scipy",
   "sentencepiece",
   "setproctitle",
-  "sgl-kernel==0.3.20",
+  "sgl-kernel==0.3.21",
   "soundfile==0.13.1",
   "tiktoken",
   "timm==1.0.16",

python/sglang/srt/entrypoints/engine.py

@@ -808,7 +808,7 @@ def _set_envs_and_config(server_args: ServerArgs):
         if _is_cuda:
             assert_pkg_version(
                 "sgl-kernel",
-                "0.3.20",
+                "0.3.21",
                 "Please reinstall the latest version with `pip install sgl-kernel --force-reinstall`",
             )