Merge pull request #576 from nicolasvasilache/pr/debug

Drop duplicate state in tuner / options cache which resulted in intermittent disagreements between tune and load from cache

Author: nicolasvasilache

python/benchmarks/python_overhead.py

@@ -111,7 +111,9 @@ def matmul_bgrad(float(M,N) A, float(M,K) d_C) -> (d_B) {
         tuner_config)
     cache = MappingOptionsCache(cache_file.name)
     top10 = cache.load(mm, "matmul", (A, B), 10)
-    assert top1.__str__() == top10[0].__str__()
+    assert top1.__str__() == top10[0].__str__(), (
+        "Expected {}\nGot {}\nTop10 are {}".format(
+            top1, top10[0], "\n".join(opt for opt in top10)))
 
     # Compile and run with the new options
     compilation_cache.compile("matmul", (A, B), top1)

tc/aten/aten_autotuner-inl.h

@@ -53,6 +53,7 @@ ATenAutotuner<Backend, Search>::tune(
     const std::string& tcName,
     const std::vector<at::Tensor>& inputs,
     const std::vector<typename Backend::MappingOptionsType>& baseMappings,
+    size_t topK,
     const tc::autotune::TuningParameterFixer& fixedParams) {
   // TODO: some checks that inputs memory lives on the proper Backend device
 
@@ -91,6 +92,7 @@ ATenAutotuner<Backend, Search>::tune(
       rawInputsPerDevice,
       rawOutputsPerDevice,
       baseMappings,
+      topK,
       fixedParams);
 }
 } // namespace aten

tc/aten/aten_autotuner.h

@@ -80,6 +80,7 @@ class ATenAutotuner : public tc::autotune::Autotuner<Backend, SearchStrategy> {
       const std::string& tcEntryPoint,
       const std::vector<at::Tensor>& inputs,
       const std::vector<MappingOptionsType>& baseMappings,
+      size_t topK = 1,
       const tc::autotune::TuningParameterFixer& fixedParams = {});
 
  protected:

tc/autotuner/autotuner-inl.h

@@ -15,6 +15,7 @@
  */
 #include <atomic>
 #include <chrono>
+#include <functional>
 #include <numeric>
 #include <thread>
 
@@ -48,8 +49,6 @@ TuningHarness<Backend>::TuningHarness(
       baseMapping_(baseMapping),
       inputs_(inputs),
       outputs_(outputs),
-      bestTime_(Duration::max()),
-      bestMappingOptions_(baseMapping),
       optionsCache_(optionsCache) {}
 
 template <typename Backend>
@@ -67,13 +66,6 @@ void TuningHarness<Backend>::stopAfterCurrentIteration() {
   stopRequested_ = true;
 }
 
-template <typename Backend>
-const typename Backend::MappingOptionsType&
-TuningHarness<Backend>::bestMappingOptions() const {
-  std::lock_guard<std::mutex> lock(bestTimeMutex_);
-  return bestMappingOptions_;
-}
-
 #define LOG_LINE_BY_LINE(GSTREAM, ISTREAM)               \
   for (std::string line; std::getline(ISTREAM, line);) { \
     LOG(GSTREAM) << line;                                \
@@ -180,11 +172,14 @@ void TuningHarness<Backend>::doEvaluate(
 
     std::vector<Duration> runtimes{Duration::max()};
     try {
-      Duration bestTimeSoFar(Duration::max());
-      {
-        std::lock_guard<std::mutex> lock(bestTimeMutex_);
-        bestTimeSoFar = bestTime_;
-      }
+      auto vBest = optionsCache_->getTopKEntries(
+          lang::canonicalTc(tcTree_),
+          makeTensorInfoVector(inputs),
+          makeTensorInfoVector(outputs),
+          Backend::backendString(),
+          1);
+      Duration bestTimeSoFar =
+          (vBest.size() > 0) ? vBest[0].second : Duration::max();
       auto prune = detail::skipExecutionOrWarmup<Backend>(
           *pExecutor, outputs, inputs, bestTimeSoFar);
       if (prune) {
@@ -234,15 +229,6 @@ void TuningHarness<Backend>::doEvaluate(
         Backend::backendString(),
         options,
         prof);
-
-    // Save best time under lock
-    {
-      std::lock_guard<std::mutex> lock(bestTimeMutex_);
-      if (prof < bestTime_) {
-        bestTime_ = prof;
-        bestMappingOptions_ = options;
-      }
-    }
   } // end while
 }
 
@@ -310,7 +296,14 @@ void TuningHarness<Backend>::runOneIteration(
     LOG(INFO) << "[TUNER][ITERATION LOG] best option so far:";
     std::stringstream ssInfo;
     typename Backend::MappingOptionsCppPrinter infoPrinter(ssInfo);
-    infoPrinter << bestMappingOptions();
+    auto vBest = optionsCache_->getTopKOptions(
+        lang::canonicalTc(tcTree_),
+        makeTensorInfoVector(inputs_.begin()->second),
+        makeTensorInfoVector(outputs_.begin()->second),
+        Backend::backendString(),
+        1);
+    CHECK_GT(vBest.size(), 0);
+    infoPrinter << vBest[0];
     LOG_LINE_BY_LINE(INFO, ssInfo);
   }
   searchStrategy.updateParameters();
@@ -426,6 +419,7 @@ Autotuner<Backend, SearchStrategy>::tune(
     const std::unordered_map<size_t, std::vector<const DLConstTensor*>>& inputs,
     std::unordered_map<size_t, std::vector<const DLTensor*>>& outputs,
     const std::vector<typename Backend::MappingOptionsType>& baseMappings,
+    size_t topK,
     const TuningParameterFixer& fixedParams) {
   std::map<std::string, lang::TreeRef> tcEntryPointMap(tc::detail::parse(tc));
   TC_CHECK_EQ(tcEntryPointMap.count(tcEntryPoint), 1u)
@@ -511,7 +505,12 @@ Autotuner<Backend, SearchStrategy>::tune(
     std::rethrow_exception(tuningHarnessThreadEx);
   }
 
-  return {tuningHarness.bestMappingOptions()};
+  return optionsCache->getTopKOptions(
+      lang::canonicalTc(tcEntryPointMap.at(tcEntryPoint)),
+      makeTensorInfoVector(inputs.begin()->second),
+      makeTensorInfoVector(outputs.begin()->second),
+      Backend::backendString(),
+      topK);
 }
 } // namespace autotune
 } // namespace tc

tc/autotuner/autotuner.h

@@ -67,9 +67,6 @@ class TuningHarness {
   /// TODO: we should detect when we come from python and exit properly in C++.
   void stopAfterCurrentIteration();
 
-  /// Under lock, returns the best mapping options found so far
-  const MappingOptionsType& bestMappingOptions() const;
-
  private:
   /// Traverse one iteration of candidates in parallel and evaluate their
   /// runtimes
@@ -92,7 +89,6 @@ class TuningHarness {
   /// This way it is easy to implement multi-threaded termination by just
   /// taking an atomic counter and pushing/popping the queues under lock until
   /// we have evaluated searchStrategy->population.size() compilation results.
-  mutable std::mutex bestTimeMutex_;
   std::mutex executorsMutex_;
   std::atomic_bool stopRequested_;
   std::atomic_size_t currentCompilationJob_;
@@ -112,10 +108,6 @@ class TuningHarness {
   const std::unordered_map<size_t, std::vector<const DLConstTensor*>> inputs_;
   std::unordered_map<size_t, std::vector<const DLTensor*>> outputs_;
 
-  // results
-  Duration bestTime_;
-  MappingOptionsType bestMappingOptions_;
-
   // backing options cache
   std::shared_ptr<OptionsCache<Backend>> optionsCache_;
 };
@@ -165,6 +157,7 @@ class Autotuner {
           inputs,
       std::unordered_map<size_t, std::vector<const DLTensor*>>& outputs,
       const std::vector<MappingOptionsType>& baseMapping,
+      size_t topK = 1,
       const TuningParameterFixer& fixedParams = TuningParameterFixer());
 
  public:

tc/autotuner/options_cache-inl.h

@@ -28,6 +28,7 @@
 
 #include "tc/core/check.h"
 #include "tc/core/compiler.h"
+#include "tc/core/functional.h"
 #include "tc/core/tensor.h"
 #include "tc/core/utils/math.h"
 #include "tc/core/utils/time.h"
@@ -235,8 +236,8 @@ std::vector<OptionsWithMedianAndRuntimes<Backend>> sortedOptions(
 } // namespace detail
 
 template <typename Backend>
-std::vector<typename Backend::MappingOptionsType>
-OptionsCache<Backend>::getTopKOptions(
+std::vector<std::pair<typename Backend::MappingOptionsType, Duration>>
+OptionsCache<Backend>::getTopKEntries(
     const lang::CanonicalTcString& tc,
     const std::vector<TensorInfo>& inputs,
     const std::vector<TensorInfo>& outputs,
@@ -249,15 +250,32 @@ OptionsCache<Backend>::getTopKOptions(
   if (sorted.size() == 0u) {
     return {};
   }
-  std::vector<typename Backend::MappingOptionsType> res;
-  res.reserve(K);
+  std::vector<std::pair<typename Backend::MappingOptionsType, Duration>> res;
+  res.reserve(std::min(K, sorted.size()));
   for (size_t i = 0; i < std::min(K, sorted.size()); ++i) {
-    res.push_back(sorted[i].mappingOptions);
+    res.push_back(std::make_pair(sorted[i].mappingOptions, sorted[i].median));
   }
   ++numberSuccessfulRetrievals;
   return res;
 }
 
+template <typename Backend>
+std::vector<typename Backend::MappingOptionsType>
+OptionsCache<Backend>::getTopKOptions(
+    const lang::CanonicalTcString& tc,
+    const std::vector<TensorInfo>& inputs,
+    const std::vector<TensorInfo>& outputs,
+    const std::string& backendStr,
+    size_t K) const {
+  auto vBest = getTopKEntries(tc, inputs, outputs, backendStr, K);
+  using ReturnType = typename Backend::MappingOptionsType;
+  using ValueType = typename decltype(vBest)::value_type;
+  std::function<ReturnType(ValueType)> map = [](ValueType in) {
+    return in.first;
+  };
+  return tc::functional::Map(map, vBest);
+}
+
 template <typename Backend>
 std::unordered_set<OptionsCacheKey, OptionsCacheKeyHash>
 OptionsCache<Backend>::getKeys() const {

tc/autotuner/options_cache.h

@@ -132,6 +132,19 @@ struct OptionsCache {
   /// particular TC/inputs/outputs/device. Note that the result may be empty
   /// (in particular if problem size is small and pruning threshold is too high
   /// for the problem size).
+  /// \returns a vector of pair<mapping options, Duration>
+  std::vector<std::pair<typename Backend::MappingOptionsType, Duration>>
+  getTopKEntries(
+      const lang::CanonicalTcString& tc,
+      const std::vector<TensorInfo>& inputs,
+      const std::vector<TensorInfo>& outputs,
+      const std::string& backendStr,
+      size_t K) const;
+
+  /// Returns the top-K mapping options that have the best median runtime for a
+  /// particular TC/inputs/outputs/device. Note that the result may be empty
+  /// (in particular if problem size is small and pruning threshold is too high
+  /// for the problem size).
   /// \returns a vector of mapping options
   std::vector<typename Backend::MappingOptionsType> getTopKOptions(
       const lang::CanonicalTcString& tc,

tc/benchmarks/benchmark_fixture.h

@@ -188,7 +188,7 @@ struct Benchmark : public ::testing::Test {
         geneticAutotuneATen(tc);
     auto bestOptions = [&]() {
       auto options = geneticAutotuneATen.tune(
-          kernelName, inputs, {baseMapping}, fixedParams);
+          kernelName, inputs, {baseMapping}, 1, fixedParams);
       TC_CHECK_GE(options.size(), 1u) << "Benchmark mode: at least one "
                                       << "options expected";
       return options[0];

tc/core/polyhedral/functional.h renamed to tc/core/functional.h

@@ -21,7 +21,6 @@
 #include <vector>
 
 namespace tc {
-namespace polyhedral {
 namespace functional {
 
 template <typename I>
@@ -178,5 +177,4 @@ R MapReduce(std::function<R(R, I, bool)> fun, const std::vector<I>& vec) {
 }
 
 } // namespace functional
-} // namespace polyhedral
 } // namespace tc

tc/core/polyhedral/cuda/mapped_scop.cc

@@ -26,13 +26,13 @@
 #include "tc/core/check.h"
 #include "tc/core/cuda/cuda_libraries.h"
 #include "tc/core/flags.h"
+#include "tc/core/functional.h"
 #include "tc/core/gpu.h"
 #include "tc/core/polyhedral/cuda/codegen.h"
 #include "tc/core/polyhedral/cuda/mapping_types.h"
 #include "tc/core/polyhedral/cuda/memory_promotion_heuristic.h"
 #include "tc/core/polyhedral/cuda/tighten_launch_bounds.h"
 #include "tc/core/polyhedral/exceptions.h"
-#include "tc/core/polyhedral/functional.h"
 #include "tc/core/polyhedral/schedule_transforms.h"
 #include "tc/core/polyhedral/schedule_tree_matcher.h"
 #include "tc/core/polyhedral/schedule_utils.h"

tc/core/polyhedral/schedule_transforms.cc

@@ -30,7 +30,7 @@
 
 #include "tc/core/check.h"
 #include "tc/core/constants.h"
-#include "tc/core/polyhedral/functional.h"
+#include "tc/core/functional.h"
 #include "tc/core/polyhedral/mapping_types.h"
 #include "tc/core/polyhedral/schedule_tree_elem.h"
 #include "tc/core/polyhedral/schedule_tree_matcher.h"

tc/core/polyhedral/schedule_transforms.h

@@ -21,7 +21,7 @@
 #include <unordered_set>
 #include <vector>
 
-#include "tc/core/polyhedral/functional.h"
+#include "tc/core/functional.h"
 #include "tc/core/polyhedral/mapping_types.h"
 #include "tc/core/polyhedral/options.h"
 #include "tc/core/polyhedral/schedule_tree.h"

tc/core/polyhedral/schedule_tree.cc

@@ -29,7 +29,7 @@
 
 #include "tc/core/check.h"
 #include "tc/core/constants.h"
-#include "tc/core/polyhedral/functional.h"
+#include "tc/core/functional.h"
 #include "tc/core/polyhedral/schedule_tree_elem.h"
 #include "tc/core/scope_guard.h"
 #include "tc/external/isl.h"

tc/core/polyhedral/scop.cc

@@ -24,9 +24,9 @@
 #include <vector>
 
 #include "tc/core/check.h"
+#include "tc/core/functional.h"
 #include "tc/core/halide2isl.h"
 #include "tc/core/polyhedral/body.h"
-#include "tc/core/polyhedral/functional.h"
 #include "tc/core/polyhedral/memory_promotion.h"
 #include "tc/core/polyhedral/schedule_isl_conversion.h"
 #include "tc/core/polyhedral/schedule_transforms.h"

tensor_comprehensions/pybinds/tclib.cc

@@ -33,6 +33,7 @@
 #include "tc/core/cuda/cuda_backend.h"
 #include "tc/core/cuda/cuda_tc_executor.h"
 #include "tc/core/flags.h"
+#include "tc/core/functional.h"
 #include "tc/core/tensor.h"
 #include "tc/lang/canonicalize.h"
 

test/cuda/test_autotuner.cc

@@ -72,7 +72,8 @@ struct ATenCompilationUnitTest : public ::testing::Test {
     if (FLAGS_no_memory_promotion) {
       fix.fixUseSharedMemory(false).fixUsePrivateMemory(false);
     }
-    auto options = geneticAutotuneATen.tune(name, inputs, {baseMapping}, fix);
+    auto options = geneticAutotuneATen.tune(
+        name, inputs, {baseMapping}, std::numeric_limits<size_t>::max(), fix);
     if (options.size() > 0) {
       return options[0];
     }

test/test_cuda_mapper.cc

@@ -27,9 +27,9 @@
 #include "tc/core/constants.h"
 #include "tc/core/cuda/cuda_libraries.h"
 #include "tc/core/cuda/cuda_mapping_options.h"
+#include "tc/core/functional.h"
 #include "tc/core/polyhedral/cuda/codegen.h"
 #include "tc/core/polyhedral/cuda/mapped_scop.h"
-#include "tc/core/polyhedral/functional.h"
 #include "tc/core/polyhedral/mapping_types.h"
 #include "tc/core/polyhedral/schedule_isl_conversion.h"
 #include "tc/core/polyhedral/schedule_transforms.h"