full cg support

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>

csrc/attention/mla/cutlass_sm100_mla/kernel/sm100_fmha_mla_tma_warpspecialized.hpp

@@ -582,9 +582,7 @@ struct Sm100FmhaMlaKernelTmaWarpspecialized {
         auto problem_shape = params.problem_shape;
 	auto local_split_kv = params.split_kv;
         if (params.mainloop.ptr_seq != nullptr) {
-          auto seqlen = params.mainloop.ptr_seq[get<2>(blk_coord)];
-          if (seqlen == 0) continue;
-          get<1>(problem_shape) = seqlen;
+          get<1>(problem_shape) = params.mainloop.ptr_seq[get<2>(blk_coord)];
 	  if (params.ptr_split_kv != nullptr) {
             local_split_kv = params.ptr_split_kv[get<2>(blk_coord)];
           }
@@ -609,9 +607,7 @@ struct Sm100FmhaMlaKernelTmaWarpspecialized {
 	  auto problem_shape = params.problem_shape;
 	  auto local_split_kv = params.split_kv;
           if (params.mainloop.ptr_seq != nullptr) {
-          auto seqlen = params.mainloop.ptr_seq[get<2>(blk_coord)];
-          if (seqlen == 0) continue;
-          get<1>(problem_shape) = seqlen;
+            get<1>(problem_shape) = params.mainloop.ptr_seq[get<2>(blk_coord)];
 	    if (params.ptr_split_kv != nullptr) {
               local_split_kv = params.ptr_split_kv[get<2>(blk_coord)];
             }
@@ -640,9 +636,7 @@ struct Sm100FmhaMlaKernelTmaWarpspecialized {
 	    auto problem_shape = params.problem_shape;
 	    auto local_split_kv = params.split_kv;
             if (params.mainloop.ptr_seq != nullptr) {
-            auto seqlen = params.mainloop.ptr_seq[get<2>(blk_coord)];
-            if (seqlen == 0) continue;
-            get<1>(problem_shape) = seqlen;
+              get<1>(problem_shape) = params.mainloop.ptr_seq[get<2>(blk_coord)];
 	      if (params.ptr_split_kv != nullptr) {
 	        local_split_kv = params.ptr_split_kv[get<2>(blk_coord)];
 	      }

docs/getting_started/installation/cpu/x86.inc.md

@@ -20,7 +20,80 @@ vLLM supports basic model inferencing and serving on x86 CPU platform, with data
 # --8<-- [end:pre-built-wheels]
 # --8<-- [start:build-wheel-from-source]
 
---8<-- "docs/getting_started/installation/cpu/build.inc.md"
+Install recommended compiler. We recommend to use `gcc/g++ >= 12.3.0` as the default compiler to avoid potential problems. For example, on Ubuntu 22.4, you can run:
+
+```bash
+sudo apt-get update -y
+sudo apt-get install -y gcc-12 g++-12 libnuma-dev python3-dev
+sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-12 10 --slave /usr/bin/g++ g++ /usr/bin/g++-12
+```
+
+Clone the vLLM project:
+
+```bash
+git clone https://github.com/vllm-project/vllm.git vllm_source
+cd vllm_source
+```
+
+Install the required dependencies:
+
+```bash
+uv pip install -r requirements/cpu-build.txt --torch-backend cpu
+uv pip install -r requirements/cpu.txt --torch-backend cpu
+```
+
+??? console "pip"
+    ```bash
+    pip install --upgrade pip
+    pip install -v -r requirements/cpu-build.txt --extra-index-url https://download.pytorch.org/whl/cpu
+    pip install -v -r requirements/cpu.txt --extra-index-url https://download.pytorch.org/whl/cpu
+    ```
+
+Build and install vLLM:
+
+```bash
+VLLM_TARGET_DEVICE=cpu uv pip install . --no-build-isolation
+```
+
+If you want to develop vLLM, install it in editable mode instead.
+
+```bash
+VLLM_TARGET_DEVICE=cpu uv pip install -e . --no-build-isolation
+```
+
+Optionally, build a portable wheel which you can then install elsewhere:
+
+```bash
+VLLM_TARGET_DEVICE=cpu uv build --wheel
+```
+
+```bash
+uv pip install dist/*.whl
+```
+
+??? console "pip"
+    ```bash
+    VLLM_TARGET_DEVICE=cpu python -m build --wheel --no-isolation
+    ```
+
+    ```bash
+    pip install dist/*.whl
+    ```
+
+!!! example "Troubleshooting"
+    - **NumPy ≥2.0 error**: Downgrade using `pip install "numpy<2.0"`.
+    - **CMake picks up CUDA**: Add `CMAKE_DISABLE_FIND_PACKAGE_CUDA=ON` to prevent CUDA detection during CPU builds, even if CUDA is installed.
+    - `AMD` requies at least 4th gen processors (Zen 4/Genoa) or higher to support [AVX512](https://www.phoronix.com/review/amd-zen4-avx512) to run vLLM on CPU.
+    - If you receive an error such as: `Could not find a version that satisfies the requirement torch==X.Y.Z+cpu+cpu`, consider updating [pyproject.toml](https://github.com/vllm-project/vllm/blob/main/pyproject.toml) to help pip resolve the dependency.
+    ```toml title="pyproject.toml"
+    [build-system]
+    requires = [
+      "cmake>=3.26.1",
+      ...
+      "torch==X.Y.Z+cpu"   # <-------
+    ]
+    ```
+    - If you are building vLLM from source and not using the pre-built images, remember to set `LD_PRELOAD="/usr/lib/x86_64-linux-gnu/libtcmalloc_minimal.so.4:$LD_PRELOAD"` on x86 machines before running vLLM.
 
 # --8<-- [end:build-wheel-from-source]
 # --8<-- [start:pre-built-images]
@@ -57,4 +130,4 @@ docker run --rm \
 
 # --8<-- [end:build-image-from-source]
 # --8<-- [start:extra-information]
-# --8<-- [end:extra-information]
+# --8<-- [end:extra-information]

vllm/v1/attention/backends/mla/flashinfer_mla.py

@@ -1,16 +1,18 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-from typing import Optional, Union
+from typing import ClassVar, Optional, Union
 
 import torch
 from flashinfer.decode import trtllm_batch_decode_with_kv_cache_mla
 
 from vllm.attention.backends.abstract import AttentionLayer, AttentionType
 from vllm.logger import init_logger
+from vllm.v1.attention.backends.utils import AttentionCGSupport
 from vllm.v1.attention.backends.mla.common import (MLACommonBackend,
                                                    MLACommonImpl,
-                                                   MLACommonMetadata)
+                                                   MLACommonMetadata,
+                                                   MLACommonMetadataBuilder)
 
 logger = init_logger(__name__)
 
@@ -23,6 +25,10 @@ class FlashInferMLABackend(MLACommonBackend):
     def get_name() -> str:
         return "FLASHINFER_MLA"
 
+    @staticmethod
+    def get_builder_cls() -> type["FlashInferMLAMetadataBuilder"]:
+        return FlashInferMLAMetadataBuilder
+
     @staticmethod
     def get_impl_cls() -> type["FlashInferMLAImpl"]:
         return FlashInferMLAImpl
@@ -34,6 +40,11 @@ g_fi_workspace = torch.zeros(
     device="cuda",
 )
 
+class FlashInferMLAMetadataBuilder(MLACommonMetadataBuilder[MLACommonMetadata]):
+    cudagraph_support: ClassVar[
+        AttentionCGSupport] = AttentionCGSupport.UNIFORM_SINGLE_TOKEN_DECODE
+    pass
+
 
 class FlashInferMLAImpl(MLACommonImpl[MLACommonMetadata]):
 

vllm/v1/attention/backends/mla/triton_mla.py

@@ -1,7 +1,7 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-from typing import Optional, Union
+from typing import ClassVar, Optional, Union
 
 import torch
 
@@ -13,9 +13,11 @@ from vllm.attention.ops.triton_flash_attention import triton_attention
 from vllm.logger import init_logger
 from vllm.platforms import current_platform
 from vllm.triton_utils import HAS_TRITON
+from vllm.v1.attention.backends.utils import AttentionCGSupport
 from vllm.v1.attention.backends.mla.common import (MLACommonBackend,
                                                    MLACommonImpl,
-                                                   MLACommonMetadata)
+                                                   MLACommonMetadata,
+                                                   MLACommonMetadataBuilder)
 
 logger = init_logger(__name__)
 
@@ -24,12 +26,21 @@ class TritonMLABackend(MLACommonBackend):
 
     @staticmethod
     def get_name() -> str:
-        return "TRITON_MLA"
+        return "TRITON_MLA_VLLM_V1"
+
+    @staticmethod
+    def get_builder_cls() -> type["TritonMLAMetadataBuilder"]:
+        return TritonMLAMetadataBuilder
 
     @staticmethod
     def get_impl_cls() -> type["TritonMLAImpl"]:
         return TritonMLAImpl
-
+    
+    
+class TritonMLAMetadataBuilder(MLACommonMetadataBuilder[MLACommonMetadata]):
+    cudagraph_support: ClassVar[AttentionCGSupport] = \
+        AttentionCGSupport.UNIFORM_BATCH
+    pass
 
 class TritonMLAImpl(MLACommonImpl[MLACommonMetadata]):
     can_return_lse_for_decode: bool = True