Documantation: nanobench

site/dev/testing/tests.md

-Writing Unit and Rendering Tests
-================================
+Writing Skia Tests
+==================
 
++   [Unit Tests](#test)
++   [Rendering Tests](#gm)
++   [Benchmark Tests](#bench)
+
+<span id="test"></span>
+
 Writing a Unit Test
 -------------------
 
 1.  Add a file `tests/NewUnitTest.cpp`:
 
     <!--?prettify lang=cc?-->
 
         /*
          * Copyright ........
          *
          * Use of this source code is governed by a BSD-style license
          * that can be found in the LICENSE file.
          */
         #include "Test.h"
         DEF_TEST(NewUnitTest, reporter) {
             if (1 + 1 != 2) {
                 ERRORF(reporter, "%d + %d != %d", 1, 1, 2);
             }
             bool lifeIsGood = true;
             REPORTER_ASSERT(reporter, lifeIsGood);
         }
 
 2.  Recompile and run test:
 
         python bin/sync-and-gyp
         ninja -C out/Debug dm
         out/Debug/dm --match NewUnitTest
 
+<span id="gm"></span>
+
 Writing a Rendering Test
 ------------------------
 
 1.  Add a file `gm/newgmtest.cpp`:
 
     <!--?prettify lang=cc?-->
 
         /*
          * Copyright ........
          *
@@ skipping 16 matching lines @@
 
 3.  Run the GM inside SampleApp:
 
         python bin/sync-and-gyp
         ninja -C out/Debug SampleApp
         out/Debug/SampleApp --slide GM:newgmtest
 
     On MacOS, try this:
 
         out/Debug/SampleApp.app/Contents/MacOS/SampleApp --slide GM:newgmtest
+
+<span id="bench"></span>
+
+Writing a Benchmark Test
+------------------------
+
+1.  Add a file `bench/FooBench.cpp`:
+
+    <!--?prettify lang=cc?-->
+
+        /*
+         * Copyright ........
+         *
+         * Use of this source code is governed by a BSD-style license
+         * that can be found in the LICENSE file.
+         */
+        #include "Benchmark.h"
+        #include "SkCanvas.h"
+        #include "SkRandom.h"
+        namespace {
+        class FooBench : public Benchmark {
+        public:
+            FooBench() {}
+            virtual ~FooBench() {}
+        protected:
+            const char* onGetName() override { return "Foo"; }
+            SkIPoint onGetSize() override { return SkIPoint{100, 100}; }
+            void onDraw(int loops, SkCanvas* canvas) override {
+                SkRandom r;
+                while (loops-- > 0) {
+                    canvas->drawLine(r.nextRangeF(0.0f, 100.0f),

[mtklein, 2016/02/12 15:59:32]

Oh boy, this is also a bad idea.  Those calls to nextRangeF() are not ordered.

[hal.canary, 2016/02/12 16:05:03]

how's this?

+                                     r.nextRangeF(0.0f, 100.0f),
+                                     r.nextRangeF(0.0f, 100.0f),
+                                     r.nextRangeF(0.0f, 100.0f),
+                                     SkPaint());
+                }
+            }
+        };
+        }  // namespace
+        DEF_BENCH(return new FooBench;)
+
+
+2.  Recompile and run nanobench:
+
+        python bin/sync-and-gyp
+        ninja -C out/Release nanobench
+        out/Release/nanobench --match Foo