Add rocm720 base image

.github/workflows/pr-test-amd.yml

@@ -14,7 +14,7 @@ on:
       - ".github/workflows/pr-test-amd.yml"
       - "docker/rocm.Dockerfile"
   pull_request:
-    branches: [ main ]
+    branches: [ dont-trigger-this-one-anyway ]
     paths:
       - "python/**"
       - "scripts/ci/**"

.github/workflows/release-docker-amd-rocm720-preview.yml

@@ -0,0 +1,83 @@
+name: Release Docker Images ROCm 7.2.0 Preview (AMD)
+on:
+  pull_request:
+    branches:
+      - main
+    paths:
+      - "docker/rocm720.Dockerfile"
+      - ".github/workflows/release-docker-amd-rocm720-preview.yml"
+  push:
+    tags:
+      - 'v[0-9]+.*'
+  workflow_dispatch:
+    inputs:
+      version:
+        description: 'Version to build (without v prefix, e.g., 0.5.7)'
+        required: true
+
+concurrency:
+  # A PR number if a pull request and otherwise the commit hash. This cancels
+  # queued and in-progress runs for the same PR (presubmit) or commit
+  # (postsubmit). The workflow name is prepended to avoid conflicts between
+  # different workflows.
+  group: ${{ github.workflow }}-${{ github.event.number || github.sha }}
+  cancel-in-progress: true
+
+jobs:
+  publish:
+    if: github.repository == 'sgl-project/sglang'
+    runs-on: amd-docker-scale
+    environment: 'prod'
+    strategy:
+      fail-fast: false
+      matrix:
+        gpu_arch: ['gfx942-rocm720', 'gfx950-rocm720']
+        build_type: ['all']
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0  # Required for git describe to find tags
+
+      - name: "Set Date"
+        run: |
+          echo "DATE=$(date +%Y%m%d)" >> $GITHUB_ENV
+
+      - name: Get version from latest tag
+        id: version
+        run: |
+          # Get the latest version tag sorted by version number (e.g., v0.5.7 -> 0.5.7)
+          VERSION=$(git tag -l 'v[0-9]*' --sort=-v:refname | head -1 | sed 's/^v//')
+
+          if [ -z "$VERSION" ]; then
+            echo "::error::Could not determine version from git tags"
+            exit 1
+          fi
+
+          echo "version=${VERSION}" >> $GITHUB_OUTPUT
+          echo "Detected version: ${VERSION}"
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v2
+        with:
+          username: ${{ secrets.DOCKERHUB_AMD_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_AMD_TOKEN }}
+
+      - name: Build and Push
+        run: |
+          version=${{ steps.version.outputs.version }}
+          echo "Version: ${version}"
+
+          if [ "${{ matrix.gpu_arch }}" = "gfx942-rocm720" ]; then
+            rocm_tag="rocm720-mi30x"
+          elif [ "${{ matrix.gpu_arch }}" = "gfx950-rocm720" ]; then
+            rocm_tag="rocm720-mi35x"
+          else
+            echo "Unsupported gfx arch"
+            exit 1
+          fi
+
+          tag=v${version}-${rocm_tag}
+
+          docker build . -f docker/rocm720.Dockerfile --build-arg BUILD_TYPE=${{ matrix.build_type }} --build-arg GPU_ARCH=${{ matrix.gpu_arch }} -t rocm/sgl-dev:${tag}-${{ env.DATE }}-preview --no-cache
+          docker push rocm/sgl-dev:${tag}-${{ env.DATE }}-preview

docker/aiter.patch

@@ -0,0 +1,48 @@
+diff --git a/csrc/py_itfs_cu/asm_mla.cu b/csrc/py_itfs_cu/asm_mla.cu
+index 995364105..0adab889e 100644
+--- a/csrc/py_itfs_cu/asm_mla.cu
++++ b/csrc/py_itfs_cu/asm_mla.cu
+@@ -283,14 +283,14 @@ void mla_decode_stage1_asm_fwd(
+                 else if(max_seqlen_q <= 4)
+                 {
+                     // assert(false);
+-                    //sub_Q = 128;
+-                    //static AiterAsmKernel impl_fp8(
+-                    //    "_ZN5aiter36mla_a8w8_qh16_qseqlen4_gqaratio16_psE",
+-                    //    "/mla/mla_a8w8_qh16_qseqlen4_gqaratio16_ps.co");
+-                    sub_Q = 64;
++                    sub_Q = 128;
+                     static AiterAsmKernel impl_fp8(
+-                        "_ZN5aiter36mla_a8w8_qh64_qseqlen4_gqaratio16_psE",
+-                        "/mla/mla_a8w8_qh64_qseqlen4_gqaratio16_ps.co");
++                        "_ZN5aiter36mla_a8w8_qh16_qseqlen4_gqaratio16_psE",
++                        "/mla/mla_a8w8_qh16_qseqlen4_gqaratio16_ps.co");
++                    //sub_Q = 64;
++                    //static AiterAsmKernel impl_fp8(
++                    //    "_ZN5aiter36mla_a8w8_qh64_qseqlen4_gqaratio16_psE",
++                    //    "/mla/mla_a8w8_qh64_qseqlen4_gqaratio16_ps.co");
+                     impl_ptr = &impl_fp8;
+                 }
+                 else
+@@ -319,14 +319,14 @@ void mla_decode_stage1_asm_fwd(
+                 else if(max_seqlen_q <= 4)
+                 {
+                     // assert(false);
+-                    //sub_Q = 128;
+-                    //static AiterAsmKernel impl_fp8(
+-                    //    "_ZN5aiter33mla_a8w8_qh16_qseqlen4_gqaratio16E",
+-                    //    "/mla/mla_a8w8_qh16_qseqlen4_gqaratio16.co");
+-                    sub_Q = 64;
++                    sub_Q = 128;
+                     static AiterAsmKernel impl_fp8(
+-                        "_ZN5aiter33mla_a8w8_qh64_qseqlen4_gqaratio16E",
+-                        "/mla/mla_a8w8_qh64_qseqlen4_gqaratio16.co");
++                        "_ZN5aiter33mla_a8w8_qh16_qseqlen4_gqaratio16E",
++                        "/mla/mla_a8w8_qh16_qseqlen4_gqaratio16.co");
++                    //sub_Q = 64;
++                    //static AiterAsmKernel impl_fp8(
++                    //    "_ZN5aiter33mla_a8w8_qh64_qseqlen4_gqaratio16E",
++                    //    "/mla/mla_a8w8_qh64_qseqlen4_gqaratio16.co");
+                     impl_ptr = &impl_fp8;
+                 }
+                 else