[None][feat] DSv4 prep: compressor and mHC primitives

cpp/tensorrt_llm/CMakeLists.txt

@@ -200,6 +200,8 @@ set(TRTLLM_LINK_LIBS
     layers_src
     runtime_src
     testing_src
+    compressorKernels_src
+    mhcKernels_src
     userbuffers_src
     ${DECODER_SHARED_TARGET_0}
     ${DECODER_SHARED_TARGET_1})

cpp/tensorrt_llm/kernels/CMakeLists.txt

@@ -30,6 +30,8 @@ add_subdirectory(dsv3MinLatencyKernels)
 add_subdirectory(causalConv1d)
 add_subdirectory(fusedGatedRMSNormQuant)
 add_subdirectory(mamba2MTPSSMCache)
+add_subdirectory(mhcKernels)
+add_subdirectory(compressorKernels)
 
 file(GLOB_RECURSE SRC_CPP *.cpp)
 file(GLOB_RECURSE SRC_CU *.cu)
@@ -55,6 +57,10 @@ list(FILTER SRC_CU EXCLUDE REGEX "userbuffers/.*")
 list(FILTER SRC_CU EXCLUDE REGEX "fusedLayernormKernels/.*")
 list(FILTER SRC_CU EXCLUDE REGEX "fusedGatedRMSNormQuant/.*")
 list(FILTER SRC_CU EXCLUDE REGEX "mamba2MTPSSMCache/.*")
+list(FILTER SRC_CPP EXCLUDE REGEX "mhcKernels/.*")
+list(FILTER SRC_CU EXCLUDE REGEX "mhcKernels/.*")
+list(FILTER SRC_CPP EXCLUDE REGEX "compressorKernels/.*")
+list(FILTER SRC_CU EXCLUDE REGEX "compressorKernels/.*")
 
 if(NOT ENABLE_MULTI_DEVICE)
   list(FILTER SRC_CU EXCLUDE REGEX "customAllReduceKernels*.*cu$")

cpp/tensorrt_llm/kernels/compressorKernels/CMakeLists.txt

@@ -0,0 +1,25 @@
+#
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES.
+# All rights reserved. SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License"); you may not
+# use this file except in compliance with the License. You may obtain a copy of
+# the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
+# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
+# License for the specific language governing permissions and limitations under
+# the License.
+#
+
+set(SRC_CU compressorKernels.cu)
+
+add_library(compressorKernels_src OBJECT ${SRC_CU})
+set_property(TARGET compressorKernels_src PROPERTY POSITION_INDEPENDENT_CODE ON)
+set_property(TARGET compressorKernels_src PROPERTY CUDA_RESOLVE_DEVICE_SYMBOLS
+                                                   ON)
+target_compile_options(compressorKernels_src
+                       PRIVATE $<$<COMPILE_LANGUAGE:CUDA>:--use_fast_math>)