[ci] Add benchmark to postsubmit workflow (#3220)

* [ci] Add benchmark to postsubmit workflow

* remove label 'rtx2060'

* using buildbot's installation process

* remove needs

* remove benchmark result

.github/workflows/postsubmit.yml

@@ -222,3 +222,29 @@ jobs:
           TI_LIB_DIR=`python3 -c "import taichi;print(taichi.__path__[0])" | tail -1`
           TI_LIB_DIR="$TI_LIB_DIR/lib" ./build/taichi_cpp_tests
           ti test -vr2 -t4 -x
+    
+  performance_monitoring:
+    name: Performance monitoring (NVGPU)
+    timeout-minutes: 60
+    runs-on: [self-hosted, x64, cuda, linux, benchmark]
+    steps:
+      - uses: actions/checkout@v2
+        with:
+          submodules: 'recursive'
+
+      - name: Build & Install
+        run: |
+          export PATH=$PATH:/usr/local/cuda/bin
+          .github/workflows/scripts/unix_build.sh
+        env:
+          LLVM_LIB_ROOT_DIR: /opt/taichi-llvm-10.0.0
+          LLVM_PATH: /opt/taichi-llvm-10.0.0/bin
+          LLVM_DIR: /opt/taichi-llvm-10.0.0/lib/cmake/llvm
+          CUDA_TOOLKIT_ROOT_DIR: /usr/local/cuda/
+          CI_SETUP_CMAKE_ARGS: -DTI_WITH_CUDA_TOOLKIT:BOOL=ON
+          BUILD_NUM_THREADS: 8
+          CXX: clang++-10
+
+      - name: Run benchmark
+        run: |
+          python3 benchmarks/misc/run.py /home/benchmarkbot/benchmark/

benchmarks/misc/membound.py

@@ -8,7 +8,7 @@
 test_cases = [fill, saxpy, reduction]
 test_archs = [ti.cuda]
 test_dtype = [ti.i32, ti.i64, ti.f32, ti.f64]
-test_dsize = [(4**i) * kibibyte for i in range(1, 11)]  #[4KB,16KB...1GB]
+test_dsize = [(4**i) * kibibyte for i in range(1, 10)]  #[4KB,16KB...256MB]
 test_repeat = 10
 results_evaluation = [geometric_mean]
 

benchmarks/misc/run.py

@@ -1,4 +1,9 @@
+import datetime
+import os
+import sys
+
 from membound import Membound
+from taichi.core import ti_core as _ti_core
 
 import taichi as ti
 
@@ -18,16 +23,30 @@ def run(self):
         for s in self.suites:
             s.run()
 
-    def write_md(self):
-        filename = f'performance_result.md'
-        with open(filename, 'w') as f:
+    def store_to_path(self, path_with_file_name='./performance_result.md'):
+        with open(path_with_file_name, 'w') as f:
             for arch in test_archs:
                 for s in self.suites:
                     lines = s.mdlines(arch)
                     for line in lines:
                         print(line, file=f)
 
+    def store_with_date_and_commit_id(self, file_dir='./'):
+        current_time = datetime.datetime.now().strftime("%Y%m%dd%Hh%Mm%Ss")
+        commit_hash = _ti_core.get_commit_hash()[:8]
+        file_name = f'perfresult_{current_time}_{commit_hash}.md'
+        path = os.path.join(file_dir, file_name)
+        print('Storing benchmark result to: ' + path)
+        self.store_to_path(path)
+
+
+def main():
+    file_dir = sys.argv[1] if len(sys.argv) > 1 else './'
+    p = PerformanceMonitoring()
+    p.run()
+    p.store_to_path()  # for /benchmark
+    p.store_with_date_and_commit_id(file_dir)  #for postsubmit
+
 
-p = PerformanceMonitoring()
-p.run()
-p.write_md()
+if __name__ == '__main__':
+    main()