Merge pull request #3 from fishingfly/main

Support github runner

Author: yeahdongcn

.github/Dockerfile

@@ -0,0 +1,48 @@
+FROM ubuntu:22.04
+
+LABEL name="Moore Threads Mtml Python Library CI Base Image"
+LABEL summary="Python Library CI Base Image for Moore Threads Mtml"
+LABEL description="See summary"
+LABEL vendor="MOORE THREADS CORPORATION"
+LABEL maintainer="MOORE THREADS CORPORATION <mt-sw-cloudcomputing@mthreads.com>"
+
+
+ENV DEBIAN_FRONTEND=noninteractive
+
+RUN cp /etc/apt/sources.list /etc/apt/sources.list.bak \
+    && sed -i 's@//.*archive.ubuntu.com@//mirrors.aliyun.com@g' /etc/apt/sources.list \
+    && sed -i 's@//.*security.ubuntu.com@//mirrors.aliyun.com@g' /etc/apt/sources.list
+
+RUN apt-get update \
+    && apt-get install -y --no-install-recommends \
+        python3 python3-pip python3-dev build-essential ca-certificates \
+        git curl \
+        libicu-dev \
+        sudo \
+    && apt-get clean \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN ln -s /usr/bin/python3 /usr/bin/python
+
+RUN python3 -m pip install --upgrade pip -i https://pypi.tuna.tsinghua.edu.cn/simple \
+    && pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple
+
+RUN groupdel render 2>/dev/null || true \
+    && sed -i '/:109:/d' /etc/group
+
+RUN groupadd -g 109 render
+
+RUN groupadd -g 1001 mccxadmin
+
+RUN useradd -l -u 1000 -g 1001 -G video,render -s /bin/bash -m mccxadmin
+
+RUN echo "mccxadmin ALL=(ALL) NOPASSWD:ALL" >/etc/sudoers.d/mccxadmin \
+    && chmod 0440 /etc/sudoers.d/mccxadmin
+
+WORKDIR /app
+
+RUN chown -R mccxadmin:mccxadmin /app
+
+USER mccxadmin
+
+CMD ["/bin/bash"]

.github/README.md

@@ -0,0 +1,4 @@
+# Start ci container
+```
+docker container create --name ci-mtml-py --env MTHREADS_VISIBLE_DEVICES=all --env MTHREADS_DRIVER_CAPABILITIES=all  -v $PWD:/app/test -it  registry.cn-hangzhou.aliyuncs.com/fishingfly/mthreads-ml-py-ci:latest bash
+```

.github/workflows/main.yml

@@ -0,0 +1,21 @@
+name: mthreads-ml-py local runner
+
+on: [push, workflow_dispatch]
+
+jobs:
+  Unit-test-job:
+    runs-on: self-hosted
+    steps:
+      - name: Print device info
+        run: mthreads-gmi
+      - name: Checkout code
+        uses: actions/checkout@v4
+      - name: Test mtml sglang compatibility
+        run: python test_sglang_compat.py
+      - name: Test mtml
+        run: |
+          export CI_MTML_SKIP_ECC=1
+          python test_pymtml.py
+      - name: Test pynvml
+        run: |
+          python test_pynvml.py

.gitignore

@@ -182,9 +182,9 @@ cython_debug/
 .abstra/
 
 # Visual Studio Code
-#  Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore 
+#  Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore
 #  that can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
-#  and can be added to the global gitignore or merged into this file. However, if you prefer, 
+#  and can be added to the global gitignore or merged into this file. However, if you prefer,
 #  you could uncomment the following to ignore the entire vscode folder
 # .vscode/
 
@@ -205,3 +205,6 @@ cython_debug/
 marimo/_static/
 marimo/_lsp/
 __marimo__/
+
+# Vscode
+.vscode/

test_pymtml.py

@@ -6,6 +6,7 @@
 
 import sys
 import traceback
+import os
 
 from pymtml import *
 
@@ -81,7 +82,9 @@ def test_device_pci_apis(self, device, device_idx):
         if pci_info.busId and pci_info.busId[0].isalnum():
             print_result("PCI Info busId populated", pci_info.busId)
         else:
-            print_result("PCI Info busId populated", "[FAIL: busId is empty or invalid]")
+            print_result(
+                "PCI Info busId populated", "[FAIL: busId is empty or invalid]"
+            )
         test_error("PCIe Slot Info", lambda: mtmlDeviceGetPcieSlotInfo(device))
 
     def test_device_fan_apis(self, device, device_idx):
@@ -198,7 +201,8 @@ def test_ecc_apis(self, device, device_idx):
 
         try:
             memory = mtmlDeviceInitMemory(device)
-            test_error("ECC Mode", lambda: mtmlMemoryGetEccMode(memory))
+            if not os.getenv("CI_MTML_SKIP_ECC") == "1":
+                test_error("ECC Mode", lambda: mtmlMemoryGetEccMode(memory))
             test_error(
                 "Retired Pages Count", lambda: mtmlMemoryGetRetiredPagesCount(memory)
             )