chore: import upstream snapshot with attribution

2026-07-13 13:33:03 +08:00
commit 5b57521aa1
8226 changed files with 3425766 additions and 0 deletions
@@ -0,0 +1,62 @@
+//
+//  ConvolutionWinogradBridge.cpp
+//  MNN
+//
+//  Created by MNN on 2022/01/20.
+//  Copyright © 2018 - 2022, Alibaba Group Holding Limited
+//
+
+#include "backend/cpu/CPUConvolution.hpp"
+#include "backend/cpu/compute/ConvolutionWinogradImpl.hpp"
+#include "backend/cpu/compute/ConvolutionWinogradBridge.hpp"
+#include "backend/cpu/compute/ConvolutionPackFreeWinograd.hpp"
+#include "backend/cpu/compute/ConvolutionPackWinograd.hpp"
+
+namespace MNN {
+
+
+WinogradConfig ConvolutionWinogradBridge::bestWinogradUnit(const Convolution2DCommon *common, const Tensor *inputTensor,
+                                          const Tensor *outputTensor, int threadNumber, Backend* b, const PerfConfig& denseConfig) {
+
+//  Currently packfree is only used in x86 architecture
+#ifdef MNN_USE_SSE
+    auto core = static_cast<CPUBackend*>(b)->functions();
+    if (16 == core->pack) { // avx512
+        return ConvolutionPackFreeWinograd::bestWinogradUnit(common, inputTensor, outputTensor, threadNumber, b, denseConfig);
+    } else {
+#endif
+        return ConvolutionPackWinograd::bestWinogradUnit(common, inputTensor, outputTensor, threadNumber, b, denseConfig);
+
+#ifdef MNN_USE_SSE
+    }
+#endif
+
+}
+
+bool ConvolutionWinogradBridge::canUseWinograd(const Convolution2DCommon *common) {
+    return ConvolutionPackWinograd::canUseWinograd(common);
+}
+
+ConvolutionWinogradImpl *ConvolutionWinogradBridge::createWinogradImpl(const Convolution2DCommon *common,
+                                                                       const Tensor *input, const Tensor *output,
+                                                                       Backend *b, const float *originWeight,
+                                                                       size_t originWeightSize, const float *bias,
+                                                                       size_t biasSize, WinogradConfig config) {
+
+#ifdef MNN_USE_SSE
+    auto core = static_cast<CPUBackend*>(b)->functions();
+    // Adopt different algorithm for x86 and arm
+    if (16 == core->pack) { // avx512
+        return new ConvolutionPackFreeWinograd(common, input, output, b, originWeight, originWeightSize, bias, biasSize,
+                                   config);
+    } else {
+#endif
+
+        return new ConvolutionPackWinograd(common, input, output, b, originWeight, originWeightSize, bias, biasSize,
+                                   config);
+#ifdef MNN_USE_SSE
+    }
+#endif
+}
+
+} // namespace MNN