fix: prevent OpenCV from overriding our threading configuration (#700)

## Description

We observed activity on all CPU cores despite manually configuring the
thread pool. OpenCV's internal threading was activating all available
cores, overriding our optimized thread configuration and resulting in
worse performance.

### Introduces a breaking change?

- [ ] Yes
- [x] No

### Type of change

- [x] Bug fix (change which fixes an issue)
- [ ] New feature (change which adds functionality)
- [ ] Documentation update (improves or adds clarity to existing
documentation)
- [ ] Other (chores, tests, code style improvements etc.)

### Tested on

- [x] iOS
- [x] Android

### Testing instructions

<!-- Provide step-by-step instructions on how to test your changes.
Include setup details if necessary. -->

### Screenshots

<!-- Add screenshots here, if applicable -->

### Related issues

<!-- Link related issues here using #issue-number -->

### Checklist

- [ ] I have performed a self-review of my code
- [ ] I have commented my code, particularly in hard-to-understand areas
- [ ] I have updated the documentation accordingly
- [ ] My changes generate no new warnings

### Additional notes

<!-- Include any additional information, assumptions, or context that
reviewers might need to understand this PR. -->

Author: chmjkb

packages/react-native-executorch/common/rnexecutorch/RnExecutorchInstaller.cpp

@@ -108,22 +108,6 @@ void RnExecutorchInstaller::injectJSIBindings(
 
   threads::utils::unsafeSetupThreadPool();
   threads::GlobalThreadPool::initialize();
-
-#if defined(__ANDROID__) && defined(__aarch64__)
-  auto num_of_perf_cores =
-      ::executorch::extension::cpuinfo::get_num_performant_cores();
-  log(LOG_LEVEL::Info, "Detected ", num_of_perf_cores, " performant cores");
-  // setting num_of_cores to floor(num_of_perf_cores / 2) + 1) because depending
-  // on cpu arch as when possible we want to leave at least 2 performant cores
-  // for other tasks (setting more actually results in drop of performance). For
-  // older devices (i.e. samsung s22) resolves to 3 cores, and for newer ones
-  // (like OnePlus 12) resolves to 4, which when benchamrked gives highest
-  // throughput.
-  auto num_of_cores = static_cast<uint32_t>(num_of_perf_cores / 2) + 1;
-  ::executorch::extension::threadpool::get_threadpool()
-      ->_unsafe_reset_threadpool(num_of_cores);
-  log(LOG_LEVEL::Info, "Configuring xnnpack for ", num_of_cores, " threads");
-#endif
 }
 
 } // namespace rnexecutorch

packages/react-native-executorch/common/rnexecutorch/threads/GlobalThreadPool.h

@@ -4,6 +4,7 @@
 #include <executorch/extension/threadpool/cpuinfo_utils.h>
 #include <memory>
 #include <mutex>
+#include <opencv2/opencv.hpp>
 #include <optional>
 #include <rnexecutorch/Log.h>
 #include <rnexecutorch/threads/HighPerformanceThreadPool.h>
@@ -38,6 +39,9 @@ class GlobalThreadPool {
           numThreads, "threads");
       instance = std::make_unique<HighPerformanceThreadPool>(numThreads.value(),
                                                              config);
+      // Disable OpenCV's internal threading to prevent it from overriding our
+      // thread pool configuration, which would cause degraded performance
+      cv::setNumThreads(0);
     });
   }