diff --git a/python/pyproject.toml b/python/pyproject.toml
index eaf584c3b..5bd244d7c 100644
--- a/python/pyproject.toml
+++ b/python/pyproject.toml
@@ -54,7 +54,7 @@ srt = [
 
 # HIP (Heterogeneous-computing Interface for Portability) for AMD
 # => base docker rocm/vllm-dev:20250114, not from public vllm whl
-srt_hip = ["sglang[runtime_common]", "sgl-kernel==0.0.3.post6", "torch", "vllm==0.6.7.dev2", "outlines==0.1.11"]
+srt_hip = ["sglang[runtime_common]", "torch", "vllm==0.6.7.dev2", "outlines==0.1.11"]
 
 # xpu is not enabled in public vllm and torch whl,
 # need to follow https://docs.vllm.ai/en/latest/getting_started/xpu-installation.htmlinstall vllm
diff --git a/sgl-kernel/csrc/torch_extension_rocm.cc b/sgl-kernel/csrc/torch_extension_rocm.cc
index 014e311cf..ade9a6d44 100644
--- a/sgl-kernel/csrc/torch_extension_rocm.cc
+++ b/sgl-kernel/csrc/torch_extension_rocm.cc
@@ -63,4 +63,4 @@ TORCH_LIBRARY_EXPAND(sgl_kernel, m) {
   m.impl("moe_align_block_size", torch::kCUDA, &moe_align_block_size);
 }
 
-REGISTER_EXTENSION(_kernels)
+REGISTER_EXTENSION(common_ops)
diff --git a/sgl-kernel/setup_rocm.py b/sgl-kernel/setup_rocm.py
index 25484ae7a..55bc37266 100644
--- a/sgl-kernel/setup_rocm.py
+++ b/sgl-kernel/setup_rocm.py
@@ -79,7 +79,7 @@ ext_modules = [
 setup(
     name="sgl-kernel",
     version=_get_version(),
-    packages=find_packages(),
+    packages=find_packages(where="python"),
     package_dir={"": "python"},
     ext_modules=ext_modules,
     cmdclass={"build_ext": BuildExtension.with_options(use_ninja=True)},