493     auto [n2, outputShapes, timing] = executor->computeOnCpuFallback();  in cpuFallbackPartial()
494     return {n2, std::move(outputShapes), timing, executor};  in cpuFallbackPartial()
507     Timing timing = kNoTiming;  in asyncStartComputePartitioned()  local
531             executionCallback->notify(ErrorStatus::NONE, outputShapes, timing);  in asyncStartComputePartitioned()
549             timing = stepTiming;  in asyncStartComputePartitioned()
592             timing = fallbackTiming;  in asyncStartComputePartitioned()
1092     const auto [n, outputShapes, timing] = mPreparedModel->execute(  in computeWithMemories()
1094     mExecutionBuilder->reportTimingWithoutFencedExecutionCallback(timing);  in computeWithMemories()
1096     return {n, std::move(outputShapes), timing};  in computeWithMemories()
1116     const auto [n, syncFence, computeFencedCallback, timing] = mPreparedModel->executeFenced(  in computeFenced()
1120         mExecutionBuilder->reportTimingWithoutFencedExecutionCallback(timing);  in computeFenced()
1190     auto [nCompute, outputShapes, timing] = computeWithMemories({}, memories);  in computeOnCpuFallback()
1192         return {nCompute, std::move(outputShapes), timing};  in computeOnCpuFallback()
1205     return {ANEURALNETWORKS_NO_ERROR, std::move(outputShapes), timing};  in computeOnCpuFallback()