[PATCH] D49341: [test-suite] Added Image Processing Kernels Using Benchmark Library: Blur Algorithms

[Pankaj via Phabricator via llvm-commits]

proton added inline comments.


================
Comment at: MicroBenchmarks/ImageProcessing/Blur/main.cpp:167
+  int width = state.range(1);
+  gaussianBlurKernel(*inputImage, *outputImage, height, width);
+
----------------
dberris wrote:
> You probably want a `::benchmark::DoNotOptimize(...)` around this and other calls to the kernel.
I don't think I can use "::benchmark::DoNotOptimize(...)" here as the functions return void and DoNotOptimize forces the compiler to flush pending writes to memory.


https://reviews.llvm.org/D49341