Address ANAPIC review comments -- runtime.
- Document that IPreparedModel::execute*WithConfig are for single-time
usages, and IPreparedModel::createReusableExecution is for reusable
usages.
- Make PrepareModelConfig::cacheToken a fixed sized array.
Fixes: 215566186
Test: VtsHalNeuralnetworksTargetTest
Test: NNT_static
Change-Id: I706808eaa2554d69549e08b136a049c0c623d2dd
diff --git a/shim_and_sl/ShimDevice.cpp b/shim_and_sl/ShimDevice.cpp
index 53ee761..c94a8bc 100644
--- a/shim_and_sl/ShimDevice.cpp
+++ b/shim_and_sl/ShimDevice.cpp
@@ -602,7 +602,8 @@
const std::shared_ptr<IPreparedModelCallback>& callback) {
// TODO(b/205898101): Pass the compilation hints properly.
return prepareModel(model, config.preference, config.priority, config.deadlineNs,
- config.modelCache, config.dataCache, config.cacheToken, callback);
+ config.modelCache, config.dataCache, utils::toVec(config.cacheToken),
+ callback);
}
ndk::ScopedAStatus ShimDevice::prepareModelFromCache(