Merge branch 'kpu:master' into masterHEADmaster

10 files changed, 4485 insertions, 1432 deletions

diff --git a/CMakeLists.txt b/CMakeLists.txt
index c9f78fa..11d613e 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -22,6 +22,13 @@ else()
   endif()
 endif()
 
+if("${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU")
+    # GCC 12.1 erroneously detects a bunch unitialised values
+    if (CMAKE_CXX_COMPILER_VERSION VERSION_EQUAL 12.1)
+        add_compile_options(-Wno-uninitialized -Wno-maybe-uninitialized)
+    endif()
+endif()
+
 # Check if compiler supports AVX2 (this should only catch emscripten)
 try_compile(INTGEMM_COMPILER_SUPPORTS_AVX2
   ${CMAKE_CURRENT_BINARY_DIR}/compile_tests
diff --git a/README.md b/README.md
index b8388dc..05ac8c3 100644
--- a/README.md
+++ b/README.md
@@ -54,7 +54,7 @@ The last argument of `Multiply` is a callback which is usually used to performs
    - in [callbacks/implementations.inl](callbacks/implementations.inl) if you want to implement it for all architecturs at the same time.
    - in `callbacks/ARCHITECTURE.h` (e.g. [callbacks/sse2.h](callbacks/sse2.h)) if you want to implement it only for the specific architecture.
 
-For 8-bit, you can make use a of a slightly faster implementation, assuming you can determine tha quantization multipliers and prepare the biases offline:
+For 8-bit, you can make use a of a slightly faster implementation, assuming you can determine the quantization multipliers and prepare the biases offline:
 
 ```C++
 #include "intgemm/intgemm.h"
diff --git a/example.cc b/example.cc
index 5f558d0..b1ba745 100644
--- a/example.cc
+++ b/example.cc
@@ -33,10 +33,13 @@ int main() {
   }
 
   // Compute the top left corner of C as a sanity check.
+  // Since we are only checking this with debug builds, disable it by default as it causes compilation failure with recent compilers
+#ifndef NDEBUG
   float top_left_reference = 0.0f;
   for (Index w = 0; w < width; ++w) {
     top_left_reference += A[w] * B[w * B_cols];
   }
+#endif
 
   // 16-bit multiplication.
   {
diff --git a/intgemm/intgemm.cc b/intgemm/intgemm.cc
index d6c26b9..58e4bc5 100644
--- a/intgemm/intgemm.cc
+++ b/intgemm/intgemm.cc
@@ -141,33 +141,33 @@ MeanStd Unsupported_VectorMeanStd(const float * /*begin*/, const float * /*end*/
   return MeanStd();
 }
 
-void (*Int16::Quantize)(const float *input, int16_t *output, float quant_mult, Index size) = ChooseCPU(AVX512BW::Kernels16::Quantize, AVX512BW::Kernels16::Quantize, AVX2::Kernels16::Quantize, SSE2::Kernels16::Quantize, SSE2::Kernels16::Quantize, Unsupported_16bit::Quantize);
+void (*const Int16::Quantize)(const float *input, int16_t *output, float quant_mult, Index size) = ChooseCPU(AVX512BW::Kernels16::Quantize, AVX512BW::Kernels16::Quantize, AVX2::Kernels16::Quantize, SSE2::Kernels16::Quantize, SSE2::Kernels16::Quantize, Unsupported_16bit::Quantize);
 
-void (*Int16::PrepareB)(const float *input, int16_t *output, float quant_mult, Index rows, Index cols) = ChooseCPU(AVX512BW::Kernels16::PrepareB, AVX512BW::Kernels16::PrepareB, AVX2::Kernels16::PrepareB, SSE2::Kernels16::PrepareB, SSE2::Kernels16::PrepareB, Unsupported_16bit::PrepareB);
+void (*const Int16::PrepareB)(const float *input, int16_t *output, float quant_mult, Index rows, Index cols) = ChooseCPU(AVX512BW::Kernels16::PrepareB, AVX512BW::Kernels16::PrepareB, AVX2::Kernels16::PrepareB, SSE2::Kernels16::PrepareB, SSE2::Kernels16::PrepareB, Unsupported_16bit::PrepareB);
 
-void (*Int16::PrepareBQuantizedTransposed)(const int16_t *input, int16_t *output, Index inner, Index B_untransposed_cols) = ChooseCPU(AVX512BW::Kernels16::PrepareBQuantizedTransposed, AVX512BW::Kernels16::PrepareBQuantizedTransposed, AVX2::Kernels16::PrepareBQuantizedTransposed, SSE2::Kernels16::PrepareBQuantizedTransposed, SSE2::Kernels16::PrepareBQuantizedTransposed, Unsupported_16bit::PrepareBQuantizedTransposed);
+void (*const Int16::PrepareBQuantizedTransposed)(const int16_t *input, int16_t *output, Index inner, Index B_untransposed_cols) = ChooseCPU(AVX512BW::Kernels16::PrepareBQuantizedTransposed, AVX512BW::Kernels16::PrepareBQuantizedTransposed, AVX2::Kernels16::PrepareBQuantizedTransposed, SSE2::Kernels16::PrepareBQuantizedTransposed, SSE2::Kernels16::PrepareBQuantizedTransposed, Unsupported_16bit::PrepareBQuantizedTransposed);
 
-void (*Int16::PrepareBTransposed)(const float *input, int16_t *output, float quant_mult, Index inner, Index B_untransposed_cols) = ChooseCPU(AVX512BW::Kernels16::PrepareBTransposed, AVX512BW::Kernels16::PrepareBTransposed, AVX2::Kernels16::PrepareBTransposed, SSE2::Kernels16::PrepareBTransposed, SSE2::Kernels16::PrepareBTransposed, Unsupported_16bit::PrepareBTransposed);
+void (*const Int16::PrepareBTransposed)(const float *input, int16_t *output, float quant_mult, Index inner, Index B_untransposed_cols) = ChooseCPU(AVX512BW::Kernels16::PrepareBTransposed, AVX512BW::Kernels16::PrepareBTransposed, AVX2::Kernels16::PrepareBTransposed, SSE2::Kernels16::PrepareBTransposed, SSE2::Kernels16::PrepareBTransposed, Unsupported_16bit::PrepareBTransposed);
 
-void (*Int16::SelectColumnsB)(const int16_t *input, int16_t *output, Index rows, const Index *cols_begin, const Index *cols_end) = ChooseCPU(AVX512BW::Kernels16::SelectColumnsB, AVX512BW::Kernels16::SelectColumnsB, AVX2::Kernels16::SelectColumnsB, SSE2::Kernels16::SelectColumnsB, SSE2::Kernels16::SelectColumnsB, Unsupported_16bit::SelectColumnsB);
+void (*const Int16::SelectColumnsB)(const int16_t *input, int16_t *output, Index rows, const Index *cols_begin, const Index *cols_end) = ChooseCPU(AVX512BW::Kernels16::SelectColumnsB, AVX512BW::Kernels16::SelectColumnsB, AVX2::Kernels16::SelectColumnsB, SSE2::Kernels16::SelectColumnsB, SSE2::Kernels16::SelectColumnsB, Unsupported_16bit::SelectColumnsB);
 
 const char *const Int16::kName = ChooseCPU(AVX512BW::Kernels16::kName, AVX512BW::Kernels16::kName, AVX2::Kernels16::kName, SSE2::Kernels16::kName, SSE2::Kernels16::kName, Unsupported_16bit::kName);
 
-void (*Int8::Quantize)(const float *input, int8_t *output, float quant_mult, Index size) = ChooseCPU(AVX512VNNI::Kernels8::Quantize, AVX512BW::Kernels8::Quantize, AVX2::Kernels8::Quantize, SSSE3::Kernels8::Quantize, Unsupported_8bit::Quantize, Unsupported_8bit::Quantize);
+void (*const Int8::Quantize)(const float *input, int8_t *output, float quant_mult, Index size) = ChooseCPU(AVX512VNNI::Kernels8::Quantize, AVX512BW::Kernels8::Quantize, AVX2::Kernels8::Quantize, SSSE3::Kernels8::Quantize, Unsupported_8bit::Quantize, Unsupported_8bit::Quantize);
 
-void (*Int8::QuantizeU)(const float *input, uint8_t *output, float quant_mult, Index size) = ChooseCPU(AVX512VNNI::Kernels8::QuantizeU, AVX512BW::Kernels8::QuantizeU, AVX2::Kernels8::QuantizeU, SSSE3::Kernels8::QuantizeU, Unsupported_8bit::QuantizeU, Unsupported_8bit::QuantizeU);
+void (*const Int8::QuantizeU)(const float *input, uint8_t *output, float quant_mult, Index size) = ChooseCPU(AVX512VNNI::Kernels8::QuantizeU, AVX512BW::Kernels8::QuantizeU, AVX2::Kernels8::QuantizeU, SSSE3::Kernels8::QuantizeU, Unsupported_8bit::QuantizeU, Unsupported_8bit::QuantizeU);
 
-void (*Int8::PrepareB)(const float *input, int8_t *output, float quant_mult, Index rows, Index cols) = ChooseCPU(AVX512VNNI::Kernels8::PrepareB, AVX512BW::Kernels8::PrepareB, AVX2::Kernels8::PrepareB, SSSE3::Kernels8::PrepareB, Unsupported_8bit::PrepareB, Unsupported_8bit::PrepareB);
+void (*const Int8::PrepareB)(const float *input, int8_t *output, float quant_mult, Index rows, Index cols) = ChooseCPU(AVX512VNNI::Kernels8::PrepareB, AVX512BW::Kernels8::PrepareB, AVX2::Kernels8::PrepareB, SSSE3::Kernels8::PrepareB, Unsupported_8bit::PrepareB, Unsupported_8bit::PrepareB);
 
-void (*Int8::PrepareBQuantizedTransposed)(const int8_t *input, int8_t *output, Index inner, Index B_untransposed_cols) = ChooseCPU(AVX512BW::Kernels8::PrepareBQuantizedTransposed, AVX512BW::Kernels8::PrepareBQuantizedTransposed, AVX2::Kernels8::PrepareBQuantizedTransposed, SSSE3::Kernels8::PrepareBQuantizedTransposed, Unsupported_8bit::PrepareBQuantizedTransposed, Unsupported_8bit::PrepareBQuantizedTransposed);
+void (*const Int8::PrepareBQuantizedTransposed)(const int8_t *input, int8_t *output, Index inner, Index B_untransposed_cols) = ChooseCPU(AVX512BW::Kernels8::PrepareBQuantizedTransposed, AVX512BW::Kernels8::PrepareBQuantizedTransposed, AVX2::Kernels8::PrepareBQuantizedTransposed, SSSE3::Kernels8::PrepareBQuantizedTransposed, Unsupported_8bit::PrepareBQuantizedTransposed, Unsupported_8bit::PrepareBQuantizedTransposed);
 
-void (*Int8::PrepareBTransposed)(const float *input, int8_t *output, float quant_mult, Index inner, Index B_untransposed_cols) = ChooseCPU(AVX512BW::Kernels8::PrepareBTransposed, AVX512BW::Kernels8::PrepareBTransposed, AVX2::Kernels8::PrepareBTransposed, SSSE3::Kernels8::PrepareBTransposed, Unsupported_8bit::PrepareBTransposed, Unsupported_8bit::PrepareBTransposed);
+void (*const Int8::PrepareBTransposed)(const float *input, int8_t *output, float quant_mult, Index inner, Index B_untransposed_cols) = ChooseCPU(AVX512BW::Kernels8::PrepareBTransposed, AVX512BW::Kernels8::PrepareBTransposed, AVX2::Kernels8::PrepareBTransposed, SSSE3::Kernels8::PrepareBTransposed, Unsupported_8bit::PrepareBTransposed, Unsupported_8bit::PrepareBTransposed);
 
-void (*Int8::SelectColumnsB)(const int8_t *input, int8_t *output, Index rows, const Index *cols_begin, const Index *cols_end) = ChooseCPU(AVX512VNNI::Kernels8::SelectColumnsB, AVX512BW::Kernels8::SelectColumnsB, AVX2::Kernels8::SelectColumnsB, SSSE3::Kernels8::SelectColumnsB, Unsupported_8bit::SelectColumnsB, Unsupported_8bit::SelectColumnsB);
+void (*const Int8::SelectColumnsB)(const int8_t *input, int8_t *output, Index rows, const Index *cols_begin, const Index *cols_end) = ChooseCPU(AVX512VNNI::Kernels8::SelectColumnsB, AVX512BW::Kernels8::SelectColumnsB, AVX2::Kernels8::SelectColumnsB, SSSE3::Kernels8::SelectColumnsB, Unsupported_8bit::SelectColumnsB, Unsupported_8bit::SelectColumnsB);
 
 const char *const Int8::kName = ChooseCPU(AVX512VNNI::Kernels8::kName, AVX512BW::Kernels8::kName, AVX2::Kernels8::kName, SSSE3::Kernels8::kName, Unsupported_8bit::kName, Unsupported_8bit::kName);
 
-void (*Int8Shift::QuantizeU)(const float *input, uint8_t *output, float quant_mult, Index size) = ChooseCPU(AVX512VNNI::Kernels8::QuantizeU, AVX512BW::Kernels8::QuantizeU, AVX2::Kernels8::QuantizeU, SSSE3::Kernels8::QuantizeU, Unsupported_8bit::QuantizeU, Unsupported_8bit::QuantizeU);
+void (*const Int8Shift::QuantizeU)(const float *input, uint8_t *output, float quant_mult, Index size) = ChooseCPU(AVX512VNNI::Kernels8::QuantizeU, AVX512BW::Kernels8::QuantizeU, AVX2::Kernels8::QuantizeU, SSSE3::Kernels8::QuantizeU, Unsupported_8bit::QuantizeU, Unsupported_8bit::QuantizeU);
 
 const char *const Int8Shift::kName = ChooseCPU(AVX512VNNI::Kernels8::kName, AVX512BW::Kernels8::kName, AVX2::Kernels8::kName, SSSE3::Kernels8::kName, Unsupported_8bit::kName, Unsupported_8bit::kName);
 
@@ -184,9 +184,9 @@ using AVX2::VectorMeanStd;
 } // namespace AVX512BW
 #endif
 
-float (*MaxAbsolute)(const float *begin, const float *end) = ChooseCPU(AVX512BW::MaxAbsolute, AVX512BW::MaxAbsolute, AVX2::MaxAbsolute, SSE2::MaxAbsolute, SSE2::MaxAbsolute, Unsupported_MaxAbsolute);
+float (*const MaxAbsolute)(const float *begin, const float *end) = ChooseCPU(AVX512BW::MaxAbsolute, AVX512BW::MaxAbsolute, AVX2::MaxAbsolute, SSE2::MaxAbsolute, SSE2::MaxAbsolute, Unsupported_MaxAbsolute);
 
-MeanStd (*VectorMeanStd)(const float *begin, const float *end, bool absolute) = ChooseCPU(AVX512BW::VectorMeanStd, AVX512BW::VectorMeanStd, AVX2::VectorMeanStd, SSE2::VectorMeanStd, SSE2::VectorMeanStd, Unsupported_VectorMeanStd);
+MeanStd (*const VectorMeanStd)(const float *begin, const float *end, bool absolute) = ChooseCPU(AVX512BW::VectorMeanStd, AVX512BW::VectorMeanStd, AVX2::VectorMeanStd, SSE2::VectorMeanStd, SSE2::VectorMeanStd, Unsupported_VectorMeanStd);
 
 constexpr const char *const Unsupported_16bit::kName;
 constexpr const char *const Unsupported_8bit::kName;
diff --git a/intgemm/intgemm.h b/intgemm/intgemm.h
index 2528fdb..26febb5 100644
--- a/intgemm/intgemm.h
+++ b/intgemm/intgemm.h
@@ -178,29 +178,29 @@ struct Int8 {
   }
 
   // Multiply floats by quant_mult then convert to 8-bit integers with saturation.
-  static void (*Quantize)(const float *input, int8_t *output, float quant_mult, Index size);
+  static void (*const Quantize)(const float *input, int8_t *output, float quant_mult, Index size);
 
   // Multiply floats by quant_mult then convert to 8-bit integers with saturation.
   // A version that adds 127 to each number, making sure that all numbers are positive
-  static void (*QuantizeU)(const float *input, uint8_t *output, float quant_mult, Index size);
+  static void (*const QuantizeU)(const float *input, uint8_t *output, float quant_mult, Index size);
 
   // Warning: the output of PrepareB depends on the CPU.
   // It will match the Multiply function on the same CPU though.
-  static void (*PrepareB)(const float *input, int8_t *output, float quant_mult, Index rows, Index cols);
+  static void (*const PrepareB)(const float *input, int8_t *output, float quant_mult, Index rows, Index cols);
 
   // Convert from a B that was already transposed (routine not provided) and
   // quantized (e.g. with Quantize) to the CPU-dependent format used for
   // Multiply.  This is useful for storing a quantized model on disk then in a
   // CPU-independent fashion.
-  static void (*PrepareBQuantizedTransposed)(const int8_t *input, int8_t *output, Index inner, Index B_untransposed_cols);
+  static void (*const PrepareBQuantizedTransposed)(const int8_t *input, int8_t *output, Index inner, Index B_untransposed_cols);
 
   // Convert from a B that was already transposed (routine not provided) to
   // the CPU-dependent format used for Multiply.  This is useful for storing
   // a quantized model on disk then in a CPU-independent fashion.
-  static void (*PrepareBTransposed)(const float *input, int8_t *output, float quant_mul, Index inner, Index B_untransposed_cols);
+  static void (*const PrepareBTransposed)(const float *input, int8_t *output, float quant_mul, Index inner, Index B_untransposed_cols);
 
   // Select columns from a prepared B matrix.  The number of selected columns must be a multiple of 8.
-  static void (*SelectColumnsB)(const int8_t *input, int8_t *output, Index rows, const Index *cols_begin, const Index *cols_end);
+  static void (*const SelectColumnsB)(const int8_t *input, int8_t *output, Index rows, const Index *cols_begin, const Index *cols_end);
 
   // Multiply C = A * B, presuming A and B have been prepared.
   template <typename Callback>
@@ -213,12 +213,12 @@ struct Int8 {
 private:
   template <typename Callback>
   struct MultiplyImpl {
-    static void (*run)(const int8_t *A, const int8_t *B, Index A_rows, Index width, Index B_cols, Callback callback);
+    static void (*const run)(const int8_t *A, const int8_t *B, Index A_rows, Index width, Index B_cols, Callback callback);
   };
 };
 
 template <typename Callback>
-void (*Int8::MultiplyImpl<Callback>::run)(const int8_t *A, const int8_t *B, Index A_rows, Index width, Index B_cols, Callback callback) = ChooseCPU(OMPParallelWrap<Callback, AVX512VNNI::Kernels8>, OMPParallelWrap<Callback, AVX512BW::Kernels8>, OMPParallelWrap<Callback, AVX2::Kernels8>, OMPParallelWrap<Callback, SSSE3::Kernels8>, Unsupported_8bit::Multiply<Callback>, Unsupported_8bit::Multiply<Callback>);
+void (*const Int8::MultiplyImpl<Callback>::run)(const int8_t *A, const int8_t *B, Index A_rows, Index width, Index B_cols, Callback callback) = ChooseCPU(OMPParallelWrap<Callback, AVX512VNNI::Kernels8>, OMPParallelWrap<Callback, AVX512BW::Kernels8>, OMPParallelWrap<Callback, AVX2::Kernels8>, OMPParallelWrap<Callback, SSSE3::Kernels8>, Unsupported_8bit::Multiply<Callback>, Unsupported_8bit::Multiply<Callback>);
 
 /*
  * 8-bit matrix multiplication with shifting A by 127
@@ -236,7 +236,7 @@ struct Int8Shift {
 
   // Multiply floats by quant_mult then convert to 8-bit integers with saturation.
   // A version that adds 127 to each number, making sure that all numbers are positive
-  static void (*QuantizeU)(const float *input, uint8_t *output, float quant_mult, Index size);
+  static void (*const QuantizeU)(const float *input, uint8_t *output, float quant_mult, Index size);
   
   // Warning: the output of PrepareB depends on the CPU.
   // It will match the Multiply function on the same CPU though.
@@ -271,17 +271,17 @@ struct Int8Shift {
 private:
   template <typename Callback>
   struct MultiplyImpl {
-    static void (*run)(const uint8_t *A, const int8_t *B, Index A_rows, Index width, Index B_cols, Callback callback);
+    static void (*const run)(const uint8_t *A, const int8_t *B, Index A_rows, Index width, Index B_cols, Callback callback);
   };
 
   template <typename Callback>
   struct PrepareBiasImpl {
-    static void (*run)(const int8_t *B, Index width, Index B_cols, Callback callback);
+    static void (*const run)(const int8_t *B, Index width, Index B_cols, Callback callback);
   };
 };
 
 template <class Callback>
-void (*Int8Shift::MultiplyImpl<Callback>::run)(const uint8_t *A, const int8_t *B, Index A_rows, Index width, Index B_cols, Callback callback) = ChooseCPU(
+void (*const Int8Shift::MultiplyImpl<Callback>::run)(const uint8_t *A, const int8_t *B, Index A_rows, Index width, Index B_cols, Callback callback) = ChooseCPU(
     OMPParallelWrap8Shift<Callback, AVX512VNNI::Kernels8>,
     OMPParallelWrap8Shift<Callback, AVX512BW::Kernels8>,
     OMPParallelWrap8Shift<Callback, AVX2::Kernels8>,
@@ -289,7 +289,7 @@ void (*Int8Shift::MultiplyImpl<Callback>::run)(const uint8_t *A, const int8_t *B
     Unsupported_8bit::Multiply8Shift<Callback>, Unsupported_8bit::Multiply8Shift<Callback>);
 
 template <class Callback>
-void (*Int8Shift::PrepareBiasImpl<Callback>::run)(const int8_t *B, Index width, Index B_cols, Callback callback) = ChooseCPU(AVX512VNNI::Kernels8::PrepareBias<Callback>, AVX512BW::Kernels8::PrepareBias<Callback>, AVX2::Kernels8::PrepareBias<Callback>, SSSE3::Kernels8::PrepareBias<Callback>, SSSE3::Kernels8::PrepareBias<Callback>, Unsupported_8bit::PrepareBias);
+void (*const Int8Shift::PrepareBiasImpl<Callback>::run)(const int8_t *B, Index width, Index B_cols, Callback callback) = ChooseCPU(AVX512VNNI::Kernels8::PrepareBias<Callback>, AVX512BW::Kernels8::PrepareBias<Callback>, AVX2::Kernels8::PrepareBias<Callback>, SSSE3::Kernels8::PrepareBias<Callback>, SSSE3::Kernels8::PrepareBias<Callback>, Unsupported_8bit::PrepareBias);
 
 /*
  * 16-bit matrix multiplication
@@ -309,25 +309,25 @@ struct Int16 {
 
   // Multiply floats by quant_mult then convert to 16-bit integers with saturation.
   // input
-  static void (*Quantize)(const float *input, int16_t *output, float quant_mult, Index size);
+  static void (*const Quantize)(const float *input, int16_t *output, float quant_mult, Index size);
 
   // Warning: the output of PrepareB depends on the CPU.
   // It will match the Multiply function on the same CPU though.
-  static void (*PrepareB)(const float *input, int16_t *output, float quant_mult, Index rows, Index cols);
+  static void (*const PrepareB)(const float *input, int16_t *output, float quant_mult, Index rows, Index cols);
 
   // Convert from a B that was already transposed (routine not provided) and
   // quantized (e.g. with Quantize) to the CPU-dependent format used for
   // Multiply.  This is useful for storing a quantized model on disk then in a
   // CPU-independent fashion.
-  static void (*PrepareBQuantizedTransposed)(const int16_t *input, int16_t *output, Index inner, Index B_untransposed_cols);
+  static void (*const PrepareBQuantizedTransposed)(const int16_t *input, int16_t *output, Index inner, Index B_untransposed_cols);
 
   // Convert from a B that was already transposed (routine not provided) to
   // the CPU-dependent format used for Multiply.  This is useful for storing
   // a quantized model on disk then in a CPU-independent fashion.
-  static void (*PrepareBTransposed)(const float *input, int16_t *output, float quant_mul, Index inner, Index B_untransposed_cols);
+  static void (*const PrepareBTransposed)(const float *input, int16_t *output, float quant_mul, Index inner, Index B_untransposed_cols);
 
   // Select columns from a prepared B matrix.  The number of selected columns must be a multiple of 8. 
-  static void (*SelectColumnsB)(const int16_t *input, int16_t *output, Index rows, const Index *cols_begin, const Index *cols_end);
+  static void (*const SelectColumnsB)(const int16_t *input, int16_t *output, Index rows, const Index *cols_begin, const Index *cols_end);
 
   // Multiply C = A * B, presuming A and B have been prepared.
   template <typename Callback>
@@ -340,20 +340,20 @@ struct Int16 {
 private:
   template <typename Callback>
   struct MultiplyImpl {
-    static void (*run)(const int16_t *A, const int16_t *B, Index A_rows, Index width, Index B_cols, Callback callback);
+    static void (*const run)(const int16_t *A, const int16_t *B, Index A_rows, Index width, Index B_cols, Callback callback);
   };
 };
 
 template <typename Callback>
-void (*Int16::MultiplyImpl<Callback>::run)(const int16_t *A, const int16_t *B, Index A_rows, Index width, Index B_cols, Callback callback) = ChooseCPU(OMPParallelWrap<Callback, AVX512BW::Kernels16> /*TODO VNNI 16-bit. */, OMPParallelWrap<Callback, AVX512BW::Kernels16>, OMPParallelWrap<Callback, AVX2::Kernels16>, OMPParallelWrap<Callback, SSE2::Kernels16>, OMPParallelWrap<Callback, SSE2::Kernels16>, Unsupported_16bit::Multiply<Callback>);
+void (*const Int16::MultiplyImpl<Callback>::run)(const int16_t *A, const int16_t *B, Index A_rows, Index width, Index B_cols, Callback callback) = ChooseCPU(OMPParallelWrap<Callback, AVX512BW::Kernels16> /*TODO VNNI 16-bit. */, OMPParallelWrap<Callback, AVX512BW::Kernels16>, OMPParallelWrap<Callback, AVX2::Kernels16>, OMPParallelWrap<Callback, SSE2::Kernels16>, OMPParallelWrap<Callback, SSE2::Kernels16>, Unsupported_16bit::Multiply<Callback>);
 
 extern const CPUType kCPU;
 
 // Get the maximum absolute value of an array of floats. The number of floats must be a multiple of 16 and 64-byte aligned.
-extern float (*MaxAbsolute)(const float *begin, const float *end);
+extern float (*const MaxAbsolute)(const float *begin, const float *end);
 
 // Get a Quantization value that is equant to the mean of the data +N standard deviations. Use 2 by default
-extern MeanStd (*VectorMeanStd)(const float *begin, const float *end, bool);
+extern MeanStd (*const VectorMeanStd)(const float *begin, const float *end, bool);
 
 /* Returns the Mean and the Standard deviation of a vector. 
  * If "absolute" is set to true, it computes the mean and the standard deviation of the absolute values of the vector */
diff --git a/intgemm/intrinsics.h b/intgemm/intrinsics.h
index 9f370cd..8db234e 100644
--- a/intgemm/intrinsics.h
+++ b/intgemm/intrinsics.h
@@ -95,12 +95,7 @@ template <> INTGEMM_SSE2 inline __m128 loadu_ps(const float* mem_addr) {
   return _mm_loadu_ps(mem_addr);
 }
 INTGEMM_SSE2 static inline __m128i madd_epi16(__m128i first, __m128i second) {
-// https://bugzilla.mozilla.org/show_bug.cgi?id=1672160
-#ifdef INTGEMM_WORMHOLE
-  return wasm_v8x16_shuffle(first, second, 31, 0, 30, 2, 29, 4, 28, 6, 27, 8, 26, 10, 25, 12, 24, 2 /* PMADDWD */);
-#else
   return _mm_madd_epi16(first, second);
-#endif
 }
 INTGEMM_SSSE3 static inline __m128i maddubs_epi16(__m128i first, __m128i second) {
 // https://bugzilla.mozilla.org/show_bug.cgi?id=1672160
diff --git a/test/3rd_party/catch.hpp b/test/3rd_party/catch.hpp
index 1850fff..d2a1242 100644
--- a/test/3rd_party/catch.hpp
+++ b/test/3rd_party/catch.hpp
@@ -1,9 +1,9 @@
 /*
- *  Catch v2.7.0
- *  Generated: 2019-03-07 21:34:30.252164
+ *  Catch v2.13.9
+ *  Generated: 2022-04-12 22:37:23.260201
  *  ----------------------------------------------------------
  *  This file has been merged from multiple headers. Please don't edit it directly
- *  Copyright (c) 2019 Two Blue Cubes Ltd. All rights reserved.
+ *  Copyright (c) 2022 Two Blue Cubes Ltd. All rights reserved.
  *
  *  Distributed under the Boost Software License, Version 1.0. (See accompanying
  *  file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
@@ -14,8 +14,8 @@
 
 
 #define CATCH_VERSION_MAJOR 2
-#define CATCH_VERSION_MINOR 7
-#define CATCH_VERSION_PATCH 0
+#define CATCH_VERSION_MINOR 13
+#define CATCH_VERSION_PATCH 9
 
 #ifdef __clang__
 #    pragma clang system_header
@@ -66,13 +66,16 @@
 #if !defined(CATCH_CONFIG_IMPL_ONLY)
 // start catch_platform.h
 
+// See e.g.:
+// https://opensource.apple.com/source/CarbonHeaders/CarbonHeaders-18.1/TargetConditionals.h.auto.html
 #ifdef __APPLE__
-# include <TargetConditionals.h>
-# if TARGET_OS_OSX == 1
-#  define CATCH_PLATFORM_MAC
-# elif TARGET_OS_IPHONE == 1
-#  define CATCH_PLATFORM_IPHONE
-# endif
+#  include <TargetConditionals.h>
+#  if (defined(TARGET_OS_OSX) && TARGET_OS_OSX == 1) || \
+      (defined(TARGET_OS_MAC) && TARGET_OS_MAC == 1)
+#    define CATCH_PLATFORM_MAC
+#  elif (defined(TARGET_OS_IPHONE) && TARGET_OS_IPHONE == 1)
+#    define CATCH_PLATFORM_IPHONE
+#  endif
 
 #elif defined(linux) || defined(__linux) || defined(__linux__)
 #  define CATCH_PLATFORM_LINUX
@@ -132,30 +135,51 @@ namespace Catch {
 
 #endif
 
-#if defined(CATCH_CPP17_OR_GREATER)
-#  define CATCH_INTERNAL_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS
+// Only GCC compiler should be used in this block, so other compilers trying to
+// mask themselves as GCC should be ignored.
+#if defined(__GNUC__) && !defined(__clang__) && !defined(__ICC) && !defined(__CUDACC__) && !defined(__LCC__)
+#    define CATCH_INTERNAL_START_WARNINGS_SUPPRESSION _Pragma( "GCC diagnostic push" )
+#    define CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION  _Pragma( "GCC diagnostic pop" )
+
+#    define CATCH_INTERNAL_IGNORE_BUT_WARN(...) (void)__builtin_constant_p(__VA_ARGS__)
+
 #endif
 
-#ifdef __clang__
+#if defined(__clang__)
+
+#    define CATCH_INTERNAL_START_WARNINGS_SUPPRESSION _Pragma( "clang diagnostic push" )
+#    define CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION  _Pragma( "clang diagnostic pop" )
+
+// As of this writing, IBM XL's implementation of __builtin_constant_p has a bug
+// which results in calls to destructors being emitted for each temporary,
+// without a matching initialization. In practice, this can result in something
+// like `std::string::~string` being called on an uninitialized value.
+//
+// For example, this code will likely segfault under IBM XL:
+// ```
+// REQUIRE(std::string("12") + "34" == "1234")
+// ```
+//
+// Therefore, `CATCH_INTERNAL_IGNORE_BUT_WARN` is not implemented.
+#  if !defined(__ibmxl__) && !defined(__CUDACC__)
+#    define CATCH_INTERNAL_IGNORE_BUT_WARN(...) (void)__builtin_constant_p(__VA_ARGS__) /* NOLINT(cppcoreguidelines-pro-type-vararg, hicpp-vararg) */
+#  endif
+
+#    define CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
+         _Pragma( "clang diagnostic ignored \"-Wexit-time-destructors\"" ) \
+         _Pragma( "clang diagnostic ignored \"-Wglobal-constructors\"")
+
+#    define CATCH_INTERNAL_SUPPRESS_PARENTHESES_WARNINGS \
+         _Pragma( "clang diagnostic ignored \"-Wparentheses\"" )
+
+#    define CATCH_INTERNAL_SUPPRESS_UNUSED_WARNINGS \
+         _Pragma( "clang diagnostic ignored \"-Wunused-variable\"" )
 
-#       define CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
-            _Pragma( "clang diagnostic push" ) \
-            _Pragma( "clang diagnostic ignored \"-Wexit-time-destructors\"" ) \
-            _Pragma( "clang diagnostic ignored \"-Wglobal-constructors\"")
-#       define CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS \
-            _Pragma( "clang diagnostic pop" )
-
-#       define CATCH_INTERNAL_SUPPRESS_PARENTHESES_WARNINGS \
-            _Pragma( "clang diagnostic push" ) \
-            _Pragma( "clang diagnostic ignored \"-Wparentheses\"" )
-#       define CATCH_INTERNAL_UNSUPPRESS_PARENTHESES_WARNINGS \
-            _Pragma( "clang diagnostic pop" )
-
-#       define CATCH_INTERNAL_SUPPRESS_UNUSED_WARNINGS \
-            _Pragma( "clang diagnostic push" ) \
-            _Pragma( "clang diagnostic ignored \"-Wunused-variable\"" )
-#       define CATCH_INTERNAL_UNSUPPRESS_UNUSED_WARNINGS \
-            _Pragma( "clang diagnostic pop" )
+#    define CATCH_INTERNAL_SUPPRESS_ZERO_VARIADIC_WARNINGS \
+         _Pragma( "clang diagnostic ignored \"-Wgnu-zero-variadic-macro-arguments\"" )
+
+#    define CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS \
+         _Pragma( "clang diagnostic ignored \"-Wunused-template\"" )
 
 #endif // __clang__
 
@@ -180,6 +204,7 @@ namespace Catch {
 // Android somehow still does not support std::to_string
 #if defined(__ANDROID__)
 #    define CATCH_INTERNAL_CONFIG_NO_CPP11_TO_STRING
+#    define CATCH_INTERNAL_CONFIG_ANDROID_LOGWRITE
 #endif
 
 ////////////////////////////////////////////////////////////////////////////////
@@ -204,20 +229,16 @@ namespace Catch {
 // some versions of cygwin (most) do not support std::to_string. Use the libstd check.
 // https://gcc.gnu.org/onlinedocs/gcc-4.8.2/libstdc++/api/a01053_source.html line 2812-2813
 # if !((__cplusplus >= 201103L) && defined(_GLIBCXX_USE_C99) \
-	       && !defined(_GLIBCXX_HAVE_BROKEN_VSWPRINTF))
+           && !defined(_GLIBCXX_HAVE_BROKEN_VSWPRINTF))
 
-#	define CATCH_INTERNAL_CONFIG_NO_CPP11_TO_STRING
+#    define CATCH_INTERNAL_CONFIG_NO_CPP11_TO_STRING
 
 # endif
 #endif // __CYGWIN__
 
 ////////////////////////////////////////////////////////////////////////////////
 // Visual C++
-#ifdef _MSC_VER
-
-#  if _MSC_VER >= 1900 // Visual Studio 2015 or newer
-#    define CATCH_INTERNAL_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS
-#  endif
+#if defined(_MSC_VER)
 
 // Universal Windows platform does not support SEH
 // Or console colours (or console at all...)
@@ -227,15 +248,27 @@ namespace Catch {
 #    define CATCH_INTERNAL_CONFIG_WINDOWS_SEH
 #  endif
 
+#  if !defined(__clang__) // Handle Clang masquerading for msvc
+
 // MSVC traditional preprocessor needs some workaround for __VA_ARGS__
 // _MSVC_TRADITIONAL == 0 means new conformant preprocessor
 // _MSVC_TRADITIONAL == 1 means old traditional non-conformant preprocessor
-#  if !defined(_MSVC_TRADITIONAL) || (defined(_MSVC_TRADITIONAL) && _MSVC_TRADITIONAL)
-#    define CATCH_INTERNAL_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
-#  endif
+#    if !defined(_MSVC_TRADITIONAL) || (defined(_MSVC_TRADITIONAL) && _MSVC_TRADITIONAL)
+#      define CATCH_INTERNAL_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+#    endif // MSVC_TRADITIONAL
+
+// Only do this if we're not using clang on Windows, which uses `diagnostic push` & `diagnostic pop`
+#    define CATCH_INTERNAL_START_WARNINGS_SUPPRESSION __pragma( warning(push) )
+#    define CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION  __pragma( warning(pop) )
+#  endif // __clang__
 
 #endif // _MSC_VER
 
+#if defined(_REENTRANT) || defined(_MSC_VER)
+// Enable async processing, as -pthread is specified or no additional linking is required
+# define CATCH_INTERNAL_CONFIG_USE_ASYNC
+#endif // _MSC_VER
+
 ////////////////////////////////////////////////////////////////////////////////
 // Check if we are compiled with -fno-exceptions or equivalent
 #if defined(__EXCEPTIONS) || defined(__cpp_exceptions) || defined(_CPPUNWIND)
@@ -266,40 +299,56 @@ namespace Catch {
 #endif
 
 ////////////////////////////////////////////////////////////////////////////////
-// Check if string_view is available and usable
-// The check is split apart to work around v140 (VS2015) preprocessor issue...
-#if defined(__has_include)
-#if __has_include(<string_view>) && defined(CATCH_CPP17_OR_GREATER)
-#    define CATCH_INTERNAL_CONFIG_CPP17_STRING_VIEW
-#endif
+
+// RTX is a special version of Windows that is real time.
+// This means that it is detected as Windows, but does not provide
+// the same set of capabilities as real Windows does.
+#if defined(UNDER_RTSS) || defined(RTX64_BUILD)
+    #define CATCH_INTERNAL_CONFIG_NO_WINDOWS_SEH
+    #define CATCH_INTERNAL_CONFIG_NO_ASYNC
+    #define CATCH_CONFIG_COLOUR_NONE
 #endif
 
-////////////////////////////////////////////////////////////////////////////////
-// Check if optional is available and usable
-#if defined(__has_include)
-#  if __has_include(<optional>) && defined(CATCH_CPP17_OR_GREATER)
-#    define CATCH_INTERNAL_CONFIG_CPP17_OPTIONAL
-#  endif // __has_include(<optional>) && defined(CATCH_CPP17_OR_GREATER)
-#endif // __has_include
+#if !defined(_GLIBCXX_USE_C99_MATH_TR1)
+#define CATCH_INTERNAL_CONFIG_GLOBAL_NEXTAFTER
+#endif
 
-////////////////////////////////////////////////////////////////////////////////
-// Check if variant is available and usable
+// Various stdlib support checks that require __has_include
 #if defined(__has_include)
-#  if __has_include(<variant>) && defined(CATCH_CPP17_OR_GREATER)
-#    if defined(__clang__) && (__clang_major__ < 8)
-       // work around clang bug with libstdc++ https://bugs.llvm.org/show_bug.cgi?id=31852
-       // fix should be in clang 8, workaround in libstdc++ 8.2
-#      include <ciso646>
-#      if defined(__GLIBCXX__) && defined(_GLIBCXX_RELEASE) && (_GLIBCXX_RELEASE < 9)
-#        define CATCH_CONFIG_NO_CPP17_VARIANT
-#      else
-#        define CATCH_INTERNAL_CONFIG_CPP17_VARIANT
-#      endif // defined(__GLIBCXX__) && defined(_GLIBCXX_RELEASE) && (_GLIBCXX_RELEASE < 9)
-#    else
-#      define CATCH_INTERNAL_CONFIG_CPP17_VARIANT
-#    endif // defined(__clang__) && (__clang_major__ < 8)
-#  endif // __has_include(<variant>) && defined(CATCH_CPP17_OR_GREATER)
-#endif // __has_include
+  // Check if string_view is available and usable
+  #if __has_include(<string_view>) && defined(CATCH_CPP17_OR_GREATER)
+  #    define CATCH_INTERNAL_CONFIG_CPP17_STRING_VIEW
+  #endif
+
+  // Check if optional is available and usable
+  #  if __has_include(<optional>) && defined(CATCH_CPP17_OR_GREATER)
+  #    define CATCH_INTERNAL_CONFIG_CPP17_OPTIONAL
+  #  endif // __has_include(<optional>) && defined(CATCH_CPP17_OR_GREATER)
+
+  // Check if byte is available and usable
+  #  if __has_include(<cstddef>) && defined(CATCH_CPP17_OR_GREATER)
+  #    include <cstddef>
+  #    if defined(__cpp_lib_byte) && (__cpp_lib_byte > 0)
+  #      define CATCH_INTERNAL_CONFIG_CPP17_BYTE
+  #    endif
+  #  endif // __has_include(<cstddef>) && defined(CATCH_CPP17_OR_GREATER)
+
+  // Check if variant is available and usable
+  #  if __has_include(<variant>) && defined(CATCH_CPP17_OR_GREATER)
+  #    if defined(__clang__) && (__clang_major__ < 8)
+         // work around clang bug with libstdc++ https://bugs.llvm.org/show_bug.cgi?id=31852
+         // fix should be in clang 8, workaround in libstdc++ 8.2
+  #      include <ciso646>
+  #      if defined(__GLIBCXX__) && defined(_GLIBCXX_RELEASE) && (_GLIBCXX_RELEASE < 9)
+  #        define CATCH_CONFIG_NO_CPP17_VARIANT
+  #      else
+  #        define CATCH_INTERNAL_CONFIG_CPP17_VARIANT
+  #      endif // defined(__GLIBCXX__) && defined(_GLIBCXX_RELEASE) && (_GLIBCXX_RELEASE < 9)
+  #    else
+  #      define CATCH_INTERNAL_CONFIG_CPP17_VARIANT
+  #    endif // defined(__clang__) && (__clang_major__ < 8)
+  #  endif // __has_include(<variant>) && defined(CATCH_CPP17_OR_GREATER)
+#endif // defined(__has_include)
 
 #if defined(CATCH_INTERNAL_CONFIG_COUNTER) && !defined(CATCH_CONFIG_NO_COUNTER) && !defined(CATCH_CONFIG_COUNTER)
 #   define CATCH_CONFIG_COUNTER
@@ -324,10 +373,6 @@ namespace Catch {
 #  define CATCH_CONFIG_CPP17_OPTIONAL
 #endif
 
-#if defined(CATCH_INTERNAL_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS) && !defined(CATCH_CONFIG_NO_CPP17_UNCAUGHT_EXCEPTIONS) && !defined(CATCH_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS)
-#  define CATCH_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS
-#endif
-
 #if defined(CATCH_INTERNAL_CONFIG_CPP17_STRING_VIEW) && !defined(CATCH_CONFIG_NO_CPP17_STRING_VIEW) && !defined(CATCH_CONFIG_CPP17_STRING_VIEW)
 #  define CATCH_CONFIG_CPP17_STRING_VIEW
 #endif
@@ -336,6 +381,10 @@ namespace Catch {
 #  define CATCH_CONFIG_CPP17_VARIANT
 #endif
 
+#if defined(CATCH_INTERNAL_CONFIG_CPP17_BYTE) && !defined(CATCH_CONFIG_NO_CPP17_BYTE) && !defined(CATCH_CONFIG_CPP17_BYTE)
+#  define CATCH_CONFIG_CPP17_BYTE
+#endif
+
 #if defined(CATCH_CONFIG_EXPERIMENTAL_REDIRECT)
 #  define CATCH_INTERNAL_CONFIG_NEW_CAPTURE
 #endif
@@ -352,17 +401,53 @@ namespace Catch {
 #  define CATCH_CONFIG_POLYFILL_ISNAN
 #endif
 
+#if defined(CATCH_INTERNAL_CONFIG_USE_ASYNC)  && !defined(CATCH_INTERNAL_CONFIG_NO_ASYNC) && !defined(CATCH_CONFIG_NO_USE_ASYNC) && !defined(CATCH_CONFIG_USE_ASYNC)
+#  define CATCH_CONFIG_USE_ASYNC
+#endif
+
+#if defined(CATCH_INTERNAL_CONFIG_ANDROID_LOGWRITE) && !defined(CATCH_CONFIG_NO_ANDROID_LOGWRITE) && !defined(CATCH_CONFIG_ANDROID_LOGWRITE)
+#  define CATCH_CONFIG_ANDROID_LOGWRITE
+#endif
+
+#if defined(CATCH_INTERNAL_CONFIG_GLOBAL_NEXTAFTER) && !defined(CATCH_CONFIG_NO_GLOBAL_NEXTAFTER) && !defined(CATCH_CONFIG_GLOBAL_NEXTAFTER)
+#  define CATCH_CONFIG_GLOBAL_NEXTAFTER
+#endif
+
+// Even if we do not think the compiler has that warning, we still have
+// to provide a macro that can be used by the code.
+#if !defined(CATCH_INTERNAL_START_WARNINGS_SUPPRESSION)
+#   define CATCH_INTERNAL_START_WARNINGS_SUPPRESSION
+#endif
+#if !defined(CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION)
+#   define CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
+#endif
 #if !defined(CATCH_INTERNAL_SUPPRESS_PARENTHESES_WARNINGS)
 #   define CATCH_INTERNAL_SUPPRESS_PARENTHESES_WARNINGS
-#   define CATCH_INTERNAL_UNSUPPRESS_PARENTHESES_WARNINGS
 #endif
 #if !defined(CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS)
 #   define CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS
-#   define CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS
 #endif
 #if !defined(CATCH_INTERNAL_SUPPRESS_UNUSED_WARNINGS)
 #   define CATCH_INTERNAL_SUPPRESS_UNUSED_WARNINGS
-#   define CATCH_INTERNAL_UNSUPPRESS_UNUSED_WARNINGS
+#endif
+#if !defined(CATCH_INTERNAL_SUPPRESS_ZERO_VARIADIC_WARNINGS)
+#   define CATCH_INTERNAL_SUPPRESS_ZERO_VARIADIC_WARNINGS
+#endif
+
+// The goal of this macro is to avoid evaluation of the arguments, but
+// still have the compiler warn on problems inside...
+#if !defined(CATCH_INTERNAL_IGNORE_BUT_WARN)
+#   define CATCH_INTERNAL_IGNORE_BUT_WARN(...)
+#endif
+
+#if defined(__APPLE__) && defined(__apple_build_version__) && (__clang_major__ < 10)
+#   undef CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS
+#elif defined(__clang__) && (__clang_major__ < 5)
+#   undef CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS
+#endif
+
+#if !defined(CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS)
+#   define CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS
 #endif
 
 #if defined(CATCH_CONFIG_DISABLE_EXCEPTIONS)
@@ -427,7 +512,7 @@ namespace Catch {
         SourceLineInfo( SourceLineInfo&& )              noexcept = default;
         SourceLineInfo& operator = ( SourceLineInfo&& ) noexcept = default;
 
-        bool empty() const noexcept;
+        bool empty() const noexcept { return file[0] == '\0'; }
         bool operator == ( SourceLineInfo const& other ) const noexcept;
         bool operator < ( SourceLineInfo const& other ) const noexcept;
 
@@ -468,9 +553,10 @@ namespace Catch {
 } // end namespace Catch
 
 #define CATCH_REGISTER_TAG_ALIAS( alias, spec ) \
+    CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
     CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
     namespace{ Catch::RegistrarForTagAliases INTERNAL_CATCH_UNIQUE_NAME( AutoRegisterTagAlias )( alias, spec, CATCH_INTERNAL_LINEINFO ); } \
-    CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS
+    CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
 
 // end catch_tag_alias_autoregistrar.h
 // start catch_test_registry.h
@@ -497,6 +583,7 @@ namespace Catch {
         virtual std::vector<TestCase> const& getAllTestsSorted( IConfig const& config ) const = 0;
     };
 
+    bool isThrowSafe( TestCase const& testCase, IConfig const& config );
     bool matchTest( TestCase const& testCase, TestSpec const& testSpec, IConfig const& config );
     std::vector<TestCase> filterTests( std::vector<TestCase> const& testCases, TestSpec const& testSpec, IConfig const& config );
     std::vector<TestCase> const& getAllTestCasesSorted( IConfig const& config );
@@ -509,53 +596,30 @@ namespace Catch {
 #include <cstddef>
 #include <string>
 #include <iosfwd>
+#include <cassert>
 
 namespace Catch {
 
     /// A non-owning string class (similar to the forthcoming std::string_view)
     /// Note that, because a StringRef may be a substring of another string,
-    /// it may not be null terminated. c_str() must return a null terminated
-    /// string, however, and so the StringRef will internally take ownership
-    /// (taking a copy), if necessary. In theory this ownership is not externally
-    /// visible - but it does mean (substring) StringRefs should not be shared between
-    /// threads.
+    /// it may not be null terminated.
     class StringRef {
     public:
         using size_type = std::size_t;
+        using const_iterator = const char*;
 
     private:
-        friend struct StringRefTestAccess;
-
-        char const* m_start;
-        size_type m_size;
-
-        char* m_data = nullptr;
-
-        void takeOwnership();
-
         static constexpr char const* const s_empty = "";
 
-    public: // construction/ assignment
-        StringRef() noexcept
-        :   StringRef( s_empty, 0 )
-        {}
-
-        StringRef( StringRef const& other ) noexcept
-        :   m_start( other.m_start ),
-            m_size( other.m_size )
-        {}
+        char const* m_start = s_empty;
+        size_type m_size = 0;
 
-        StringRef( StringRef&& other ) noexcept
-        :   m_start( other.m_start ),
-            m_size( other.m_size ),
-            m_data( other.m_data )
-        {
-            other.m_data = nullptr;
-        }
+    public: // construction
+        constexpr StringRef() noexcept = default;
 
         StringRef( char const* rawChars ) noexcept;
 
-        StringRef( char const* rawChars, size_type size ) noexcept
+        constexpr StringRef( char const* rawChars, size_type size ) noexcept
         :   m_start( rawChars ),
             m_size( size )
         {}
@@ -565,101 +629,64 @@ namespace Catch {
             m_size( stdString.size() )
         {}
 
-        ~StringRef() noexcept {
-            delete[] m_data;
+        explicit operator std::string() const {
+            return std::string(m_start, m_size);
         }
 
-        auto operator = ( StringRef const &other ) noexcept -> StringRef& {
-            delete[] m_data;
-            m_data = nullptr;
-            m_start = other.m_start;
-            m_size = other.m_size;
-            return *this;
-        }
-
-        operator std::string() const;
-
-        void swap( StringRef& other ) noexcept;
-
     public: // operators
         auto operator == ( StringRef const& other ) const noexcept -> bool;
-        auto operator != ( StringRef const& other ) const noexcept -> bool;
+        auto operator != (StringRef const& other) const noexcept -> bool {
+            return !(*this == other);
+        }
 
-        auto operator[] ( size_type index ) const noexcept -> char;
+        auto operator[] ( size_type index ) const noexcept -> char {
+            assert(index < m_size);
+            return m_start[index];
+        }
 
     public: // named queries
-        auto empty() const noexcept -> bool {
+        constexpr auto empty() const noexcept -> bool {
             return m_size == 0;
         }
-        auto size() const noexcept -> size_type {
+        constexpr auto size() const noexcept -> size_type {
             return m_size;
         }
 
-        auto numberOfCharacters() const noexcept -> size_type;
+        // Returns the current start pointer. If the StringRef is not
+        // null-terminated, throws std::domain_exception
         auto c_str() const -> char const*;
 
     public: // substrings and searches
-        auto substr( size_type start, size_type size ) const noexcept -> StringRef;
+        // Returns a substring of [start, start + length).
+        // If start + length > size(), then the substring is [start, size()).
+        // If start > size(), then the substring is empty.
+        auto substr( size_type start, size_type length ) const noexcept -> StringRef;
 
-        // Returns the current start pointer.
-        // Note that the pointer can change when if the StringRef is a substring
-        auto currentData() const noexcept -> char const*;
+        // Returns the current start pointer. May not be null-terminated.
+        auto data() const noexcept -> char const*;
 
-    private: // ownership queries - may not be consistent between calls
-        auto isOwned() const noexcept -> bool;
-        auto isSubstring() const noexcept -> bool;
-    };
+        constexpr auto isNullTerminated() const noexcept -> bool {
+            return m_start[m_size] == '\0';
+        }
 
-    auto operator + ( StringRef const& lhs, StringRef const& rhs ) -> std::string;
-    auto operator + ( StringRef const& lhs, char const* rhs ) -> std::string;
-    auto operator + ( char const* lhs, StringRef const& rhs ) -> std::string;
+    public: // iterators
+        constexpr const_iterator begin() const { return m_start; }
+        constexpr const_iterator end() const { return m_start + m_size; }
+    };
 
     auto operator += ( std::string& lhs, StringRef const& sr ) -> std::string&;
     auto operator << ( std::ostream& os, StringRef const& sr ) -> std::ostream&;
 
-    inline auto operator "" _sr( char const* rawChars, std::size_t size ) noexcept -> StringRef {
+    constexpr auto operator "" _sr( char const* rawChars, std::size_t size ) noexcept -> StringRef {
         return StringRef( rawChars, size );
     }
-
 } // namespace Catch
 
-inline auto operator "" _catch_sr( char const* rawChars, std::size_t size ) noexcept -> Catch::StringRef {
+constexpr auto operator "" _catch_sr( char const* rawChars, std::size_t size ) noexcept -> Catch::StringRef {
     return Catch::StringRef( rawChars, size );
 }
 
 // end catch_stringref.h
-// start catch_type_traits.hpp
-
-
-#include <type_traits>
-
-namespace Catch{
-
-#ifdef CATCH_CPP17_OR_GREATER
-	template <typename...>
-	inline constexpr auto is_unique = std::true_type{};
-
-	template <typename T, typename... Rest>
-	inline constexpr auto is_unique<T, Rest...> = std::bool_constant<
-		(!std::is_same_v<T, Rest> && ...) && is_unique<Rest...>
-	>{};
-#else
-
-template <typename...>
-struct is_unique : std::true_type{};
-
-template <typename T0, typename T1, typename... Rest>
-struct is_unique<T0, T1, Rest...> : std::integral_constant
-<bool,
-     !std::is_same<T0, T1>::value
-     && is_unique<T0, Rest...>::value
-     && is_unique<T1, Rest...>::value
->{};
-
-#endif
-}
-
-// end catch_type_traits.hpp
 // start catch_preprocessor.hpp
 
 
@@ -722,23 +749,170 @@ struct is_unique<T0, T1, Rest...> : std::integral_constant
 #define INTERNAL_CATCH_STRINGIZE_WITHOUT_PARENS(param) (INTERNAL_CATCH_STRINGIZE(INTERNAL_CATCH_REMOVE_PARENS(param)) + 1)
 #endif
 
+#define INTERNAL_CATCH_MAKE_NAMESPACE2(...) ns_##__VA_ARGS__
+#define INTERNAL_CATCH_MAKE_NAMESPACE(name) INTERNAL_CATCH_MAKE_NAMESPACE2(name)
+
 #define INTERNAL_CATCH_REMOVE_PARENS(...) INTERNAL_CATCH_EXPAND1(INTERNAL_CATCH_DEF __VA_ARGS__)
 
-#define INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME2(Name, ...) INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME3(Name, __VA_ARGS__)
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
-#define INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME3(Name,...) Name " - " #__VA_ARGS__
-#define INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME(Name,...) INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME2(Name, INTERNAL_CATCH_REMOVE_PARENS(__VA_ARGS__))
+#define INTERNAL_CATCH_MAKE_TYPE_LIST2(...) decltype(get_wrapper<INTERNAL_CATCH_REMOVE_PARENS_GEN(__VA_ARGS__)>())
+#define INTERNAL_CATCH_MAKE_TYPE_LIST(...) INTERNAL_CATCH_MAKE_TYPE_LIST2(INTERNAL_CATCH_REMOVE_PARENS(__VA_ARGS__))
 #else
-// MSVC is adding extra space and needs more calls to properly remove ()
-#define INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME3(Name,...) Name " -" #__VA_ARGS__
-#define INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME1(Name, ...) INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME2(Name, __VA_ARGS__)
-#define INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME(Name, ...) INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME1(Name, INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_REMOVE_PARENS(__VA_ARGS__)))
+#define INTERNAL_CATCH_MAKE_TYPE_LIST2(...) INTERNAL_CATCH_EXPAND_VARGS(decltype(get_wrapper<INTERNAL_CATCH_REMOVE_PARENS_GEN(__VA_ARGS__)>()))
+#define INTERNAL_CATCH_MAKE_TYPE_LIST(...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_MAKE_TYPE_LIST2(INTERNAL_CATCH_REMOVE_PARENS(__VA_ARGS__)))
 #endif
 
-#define INTERNAL_CATCH_MAKE_TYPE_LIST(types) Catch::TypeList<INTERNAL_CATCH_REMOVE_PARENS(types)>
+#define INTERNAL_CATCH_MAKE_TYPE_LISTS_FROM_TYPES(...)\
+    CATCH_REC_LIST(INTERNAL_CATCH_MAKE_TYPE_LIST,__VA_ARGS__)
+
+#define INTERNAL_CATCH_REMOVE_PARENS_1_ARG(_0) INTERNAL_CATCH_REMOVE_PARENS(_0)
+#define INTERNAL_CATCH_REMOVE_PARENS_2_ARG(_0, _1) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_1_ARG(_1)
+#define INTERNAL_CATCH_REMOVE_PARENS_3_ARG(_0, _1, _2) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_2_ARG(_1, _2)
+#define INTERNAL_CATCH_REMOVE_PARENS_4_ARG(_0, _1, _2, _3) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_3_ARG(_1, _2, _3)
+#define INTERNAL_CATCH_REMOVE_PARENS_5_ARG(_0, _1, _2, _3, _4) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_4_ARG(_1, _2, _3, _4)
+#define INTERNAL_CATCH_REMOVE_PARENS_6_ARG(_0, _1, _2, _3, _4, _5) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_5_ARG(_1, _2, _3, _4, _5)
+#define INTERNAL_CATCH_REMOVE_PARENS_7_ARG(_0, _1, _2, _3, _4, _5, _6) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_6_ARG(_1, _2, _3, _4, _5, _6)
+#define INTERNAL_CATCH_REMOVE_PARENS_8_ARG(_0, _1, _2, _3, _4, _5, _6, _7) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_7_ARG(_1, _2, _3, _4, _5, _6, _7)
+#define INTERNAL_CATCH_REMOVE_PARENS_9_ARG(_0, _1, _2, _3, _4, _5, _6, _7, _8) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_8_ARG(_1, _2, _3, _4, _5, _6, _7, _8)
+#define INTERNAL_CATCH_REMOVE_PARENS_10_ARG(_0, _1, _2, _3, _4, _5, _6, _7, _8, _9) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_9_ARG(_1, _2, _3, _4, _5, _6, _7, _8, _9)
+#define INTERNAL_CATCH_REMOVE_PARENS_11_ARG(_0, _1, _2, _3, _4, _5, _6, _7, _8, _9, _10) INTERNAL_CATCH_REMOVE_PARENS(_0), INTERNAL_CATCH_REMOVE_PARENS_10_ARG(_1, _2, _3, _4, _5, _6, _7, _8, _9, _10)
+
+#define INTERNAL_CATCH_VA_NARGS_IMPL(_0, _1, _2, _3, _4, _5, _6, _7, _8, _9, _10, N, ...) N
+
+#define INTERNAL_CATCH_TYPE_GEN\
+    template<typename...> struct TypeList {};\
+    template<typename...Ts>\
+    constexpr auto get_wrapper() noexcept -> TypeList<Ts...> { return {}; }\
+    template<template<typename...> class...> struct TemplateTypeList{};\
+    template<template<typename...> class...Cs>\
+    constexpr auto get_wrapper() noexcept -> TemplateTypeList<Cs...> { return {}; }\
+    template<typename...>\
+    struct append;\
+    template<typename...>\
+    struct rewrap;\
+    template<template<typename...> class, typename...>\
+    struct create;\
+    template<template<typename...> class, typename>\
+    struct convert;\
+    \
+    template<typename T> \
+    struct append<T> { using type = T; };\
+    template< template<typename...> class L1, typename...E1, template<typename...> class L2, typename...E2, typename...Rest>\
+    struct append<L1<E1...>, L2<E2...>, Rest...> { using type = typename append<L1<E1...,E2...>, Rest...>::type; };\
+    template< template<typename...> class L1, typename...E1, typename...Rest>\
+    struct append<L1<E1...>, TypeList<mpl_::na>, Rest...> { using type = L1<E1...>; };\
+    \
+    template< template<typename...> class Container, template<typename...> class List, typename...elems>\
+    struct rewrap<TemplateTypeList<Container>, List<elems...>> { using type = TypeList<Container<elems...>>; };\
+    template< template<typename...> class Container, template<typename...> class List, class...Elems, typename...Elements>\
+    struct rewrap<TemplateTypeList<Container>, List<Elems...>, Elements...> { using type = typename append<TypeList<Container<Elems...>>, typename rewrap<TemplateTypeList<Container>, Elements...>::type>::type; };\
+    \
+    template<template <typename...> class Final, template< typename...> class...Containers, typename...Types>\
+    struct create<Final, TemplateTypeList<Containers...>, TypeList<Types...>> { using type = typename append<Final<>, typename rewrap<TemplateTypeList<Containers>, Types...>::type...>::type; };\
+    template<template <typename...> class Final, template <typename...> class List, typename...Ts>\
+    struct convert<Final, List<Ts...>> { using type = typename append<Final<>,TypeList<Ts>...>::type; };
+
+#define INTERNAL_CATCH_NTTP_1(signature, ...)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)> struct Nttp{};\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)>\
+    constexpr auto get_wrapper() noexcept -> Nttp<__VA_ARGS__> { return {}; } \
+    template<template<INTERNAL_CATCH_REMOVE_PARENS(signature)> class...> struct NttpTemplateTypeList{};\
+    template<template<INTERNAL_CATCH_REMOVE_PARENS(signature)> class...Cs>\
+    constexpr auto get_wrapper() noexcept -> NttpTemplateTypeList<Cs...> { return {}; } \
+    \
+    template< template<INTERNAL_CATCH_REMOVE_PARENS(signature)> class Container, template<INTERNAL_CATCH_REMOVE_PARENS(signature)> class List, INTERNAL_CATCH_REMOVE_PARENS(signature)>\
+    struct rewrap<NttpTemplateTypeList<Container>, List<__VA_ARGS__>> { using type = TypeList<Container<__VA_ARGS__>>; };\
+    template< template<INTERNAL_CATCH_REMOVE_PARENS(signature)> class Container, template<INTERNAL_CATCH_REMOVE_PARENS(signature)> class List, INTERNAL_CATCH_REMOVE_PARENS(signature), typename...Elements>\
+    struct rewrap<NttpTemplateTypeList<Container>, List<__VA_ARGS__>, Elements...> { using type = typename append<TypeList<Container<__VA_ARGS__>>, typename rewrap<NttpTemplateTypeList<Container>, Elements...>::type>::type; };\
+    template<template <typename...> class Final, template<INTERNAL_CATCH_REMOVE_PARENS(signature)> class...Containers, typename...Types>\
+    struct create<Final, NttpTemplateTypeList<Containers...>, TypeList<Types...>> { using type = typename append<Final<>, typename rewrap<NttpTemplateTypeList<Containers>, Types...>::type...>::type; };
+
+#define INTERNAL_CATCH_DECLARE_SIG_TEST0(TestName)
+#define INTERNAL_CATCH_DECLARE_SIG_TEST1(TestName, signature)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)>\
+    static void TestName()
+#define INTERNAL_CATCH_DECLARE_SIG_TEST_X(TestName, signature, ...)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)>\
+    static void TestName()
+
+#define INTERNAL_CATCH_DEFINE_SIG_TEST0(TestName)
+#define INTERNAL_CATCH_DEFINE_SIG_TEST1(TestName, signature)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)>\
+    static void TestName()
+#define INTERNAL_CATCH_DEFINE_SIG_TEST_X(TestName, signature,...)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)>\
+    static void TestName()
+
+#define INTERNAL_CATCH_NTTP_REGISTER0(TestFunc, signature)\
+    template<typename Type>\
+    void reg_test(TypeList<Type>, Catch::NameAndTags nameAndTags)\
+    {\
+        Catch::AutoReg( Catch::makeTestInvoker(&TestFunc<Type>), CATCH_INTERNAL_LINEINFO, Catch::StringRef(), nameAndTags);\
+    }
+
+#define INTERNAL_CATCH_NTTP_REGISTER(TestFunc, signature, ...)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)>\
+    void reg_test(Nttp<__VA_ARGS__>, Catch::NameAndTags nameAndTags)\
+    {\
+        Catch::AutoReg( Catch::makeTestInvoker(&TestFunc<__VA_ARGS__>), CATCH_INTERNAL_LINEINFO, Catch::StringRef(), nameAndTags);\
+    }
+
+#define INTERNAL_CATCH_NTTP_REGISTER_METHOD0(TestName, signature, ...)\
+    template<typename Type>\
+    void reg_test(TypeList<Type>, Catch::StringRef className, Catch::NameAndTags nameAndTags)\
+    {\
+        Catch::AutoReg( Catch::makeTestInvoker(&TestName<Type>::test), CATCH_INTERNAL_LINEINFO, className, nameAndTags);\
+    }
+
+#define INTERNAL_CATCH_NTTP_REGISTER_METHOD(TestName, signature, ...)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)>\
+    void reg_test(Nttp<__VA_ARGS__>, Catch::StringRef className, Catch::NameAndTags nameAndTags)\
+    {\
+        Catch::AutoReg( Catch::makeTestInvoker(&TestName<__VA_ARGS__>::test), CATCH_INTERNAL_LINEINFO, className, nameAndTags);\
+    }
+
+#define INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD0(TestName, ClassName)
+#define INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD1(TestName, ClassName, signature)\
+    template<typename TestType> \
+    struct TestName : INTERNAL_CATCH_REMOVE_PARENS(ClassName)<TestType> { \
+        void test();\
+    }
+
+#define INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X(TestName, ClassName, signature, ...)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)> \
+    struct TestName : INTERNAL_CATCH_REMOVE_PARENS(ClassName)<__VA_ARGS__> { \
+        void test();\
+    }
+
+#define INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD0(TestName)
+#define INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD1(TestName, signature)\
+    template<typename TestType> \
+    void INTERNAL_CATCH_MAKE_NAMESPACE(TestName)::TestName<TestType>::test()
+#define INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X(TestName, signature, ...)\
+    template<INTERNAL_CATCH_REMOVE_PARENS(signature)> \
+    void INTERNAL_CATCH_MAKE_NAMESPACE(TestName)::TestName<__VA_ARGS__>::test()
 
-#define INTERNAL_CATCH_MAKE_TYPE_LISTS_FROM_TYPES(types)\
-    CATCH_REC_LIST(INTERNAL_CATCH_MAKE_TYPE_LIST,INTERNAL_CATCH_REMOVE_PARENS(types))
+#ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+#define INTERNAL_CATCH_NTTP_0
+#define INTERNAL_CATCH_NTTP_GEN(...) INTERNAL_CATCH_VA_NARGS_IMPL(__VA_ARGS__, INTERNAL_CATCH_NTTP_1(__VA_ARGS__), INTERNAL_CATCH_NTTP_1(__VA_ARGS__), INTERNAL_CATCH_NTTP_1(__VA_ARGS__), INTERNAL_CATCH_NTTP_1(__VA_ARGS__), INTERNAL_CATCH_NTTP_1(__VA_ARGS__), INTERNAL_CATCH_NTTP_1( __VA_ARGS__), INTERNAL_CATCH_NTTP_1( __VA_ARGS__), INTERNAL_CATCH_NTTP_1( __VA_ARGS__), INTERNAL_CATCH_NTTP_1( __VA_ARGS__),INTERNAL_CATCH_NTTP_1( __VA_ARGS__), INTERNAL_CATCH_NTTP_0)
+#define INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD(TestName, ...) INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD1, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD0)(TestName, __VA_ARGS__)
+#define INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD(TestName, ClassName, ...) INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD1, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD0)(TestName, ClassName, __VA_ARGS__)
+#define INTERNAL_CATCH_NTTP_REG_METHOD_GEN(TestName, ...) INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD0, INTERNAL_CATCH_NTTP_REGISTER_METHOD0)(TestName, __VA_ARGS__)
+#define INTERNAL_CATCH_NTTP_REG_GEN(TestFunc, ...) INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER0, INTERNAL_CATCH_NTTP_REGISTER0)(TestFunc, __VA_ARGS__)
+#define INTERNAL_CATCH_DEFINE_SIG_TEST(TestName, ...) INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X,INTERNAL_CATCH_DEFINE_SIG_TEST_X,INTERNAL_CATCH_DEFINE_SIG_TEST1, INTERNAL_CATCH_DEFINE_SIG_TEST0)(TestName, __VA_ARGS__)
+#define INTERNAL_CATCH_DECLARE_SIG_TEST(TestName, ...) INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_DECLARE_SIG_TEST_X,INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X,INTERNAL_CATCH_DECLARE_SIG_TEST_X,INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST1, INTERNAL_CATCH_DECLARE_SIG_TEST0)(TestName, __VA_ARGS__)
+#define INTERNAL_CATCH_REMOVE_PARENS_GEN(...) INTERNAL_CATCH_VA_NARGS_IMPL(__VA_ARGS__, INTERNAL_CATCH_REMOVE_PARENS_11_ARG,INTERNAL_CATCH_REMOVE_PARENS_10_ARG,INTERNAL_CATCH_REMOVE_PARENS_9_ARG,INTERNAL_CATCH_REMOVE_PARENS_8_ARG,INTERNAL_CATCH_REMOVE_PARENS_7_ARG,INTERNAL_CATCH_REMOVE_PARENS_6_ARG,INTERNAL_CATCH_REMOVE_PARENS_5_ARG,INTERNAL_CATCH_REMOVE_PARENS_4_ARG,INTERNAL_CATCH_REMOVE_PARENS_3_ARG,INTERNAL_CATCH_REMOVE_PARENS_2_ARG,INTERNAL_CATCH_REMOVE_PARENS_1_ARG)(__VA_ARGS__)
+#else
+#define INTERNAL_CATCH_NTTP_0(signature)
+#define INTERNAL_CATCH_NTTP_GEN(...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_VA_NARGS_IMPL(__VA_ARGS__, INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_1,INTERNAL_CATCH_NTTP_1, INTERNAL_CATCH_NTTP_0)( __VA_ARGS__))
+#define INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD(TestName, ...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD1, INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD0)(TestName, __VA_ARGS__))
+#define INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD(TestName, ClassName, ...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X,INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD_X, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD1, INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD0)(TestName, ClassName, __VA_ARGS__))
+#define INTERNAL_CATCH_NTTP_REG_METHOD_GEN(TestName, ...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD, INTERNAL_CATCH_NTTP_REGISTER_METHOD0, INTERNAL_CATCH_NTTP_REGISTER_METHOD0)(TestName, __VA_ARGS__))
+#define INTERNAL_CATCH_NTTP_REG_GEN(TestFunc, ...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER, INTERNAL_CATCH_NTTP_REGISTER0, INTERNAL_CATCH_NTTP_REGISTER0)(TestFunc, __VA_ARGS__))
+#define INTERNAL_CATCH_DEFINE_SIG_TEST(TestName, ...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X,INTERNAL_CATCH_DEFINE_SIG_TEST_X,INTERNAL_CATCH_DEFINE_SIG_TEST1, INTERNAL_CATCH_DEFINE_SIG_TEST0)(TestName, __VA_ARGS__))
+#define INTERNAL_CATCH_DECLARE_SIG_TEST(TestName, ...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_VA_NARGS_IMPL( "dummy", __VA_ARGS__, INTERNAL_CATCH_DECLARE_SIG_TEST_X,INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DEFINE_SIG_TEST_X,INTERNAL_CATCH_DECLARE_SIG_TEST_X,INTERNAL_CATCH_DECLARE_SIG_TEST_X, INTERNAL_CATCH_DECLARE_SIG_TEST1, INTERNAL_CATCH_DECLARE_SIG_TEST0)(TestName, __VA_ARGS__))
+#define INTERNAL_CATCH_REMOVE_PARENS_GEN(...) INTERNAL_CATCH_EXPAND_VARGS(INTERNAL_CATCH_VA_NARGS_IMPL(__VA_ARGS__, INTERNAL_CATCH_REMOVE_PARENS_11_ARG,INTERNAL_CATCH_REMOVE_PARENS_10_ARG,INTERNAL_CATCH_REMOVE_PARENS_9_ARG,INTERNAL_CATCH_REMOVE_PARENS_8_ARG,INTERNAL_CATCH_REMOVE_PARENS_7_ARG,INTERNAL_CATCH_REMOVE_PARENS_6_ARG,INTERNAL_CATCH_REMOVE_PARENS_5_ARG,INTERNAL_CATCH_REMOVE_PARENS_4_ARG,INTERNAL_CATCH_REMOVE_PARENS_3_ARG,INTERNAL_CATCH_REMOVE_PARENS_2_ARG,INTERNAL_CATCH_REMOVE_PARENS_1_ARG)(__VA_ARGS__))
+#endif
 
 // end catch_preprocessor.hpp
 // start catch_meta.hpp
@@ -747,68 +921,40 @@ struct is_unique<T0, T1, Rest...> : std::integral_constant
 #include <type_traits>
 
 namespace Catch {
-template< typename... >
-struct TypeList {};
-
-template< typename... >
-struct append;
-
-template< template<typename...> class L1
-    , typename...E1
-    , template<typename...> class L2
-    , typename...E2
->
-struct append< L1<E1...>, L2<E2...> > {
-    using type = L1<E1..., E2...>;
-};
-
-template< template<typename...> class L1
-    , typename...E1
-    , template<typename...> class L2
-    , typename...E2
-    , typename...Rest
->
-struct append< L1<E1...>, L2<E2...>, Rest...> {
-    using type = typename append< L1<E1..., E2...>, Rest... >::type;
-};
-
-template< template<typename...> class
-    , typename...
->
-struct rewrap;
-
-template< template<typename...> class Container
-    , template<typename...> class List
-    , typename...elems
->
-struct rewrap<Container, List<elems...>> {
-    using type = TypeList< Container< elems... > >;
-};
+    template<typename T>
+    struct always_false : std::false_type {};
+
+    template <typename> struct true_given : std::true_type {};
+    struct is_callable_tester {
+        template <typename Fun, typename... Args>
+        true_given<decltype(std::declval<Fun>()(std::declval<Args>()...))> static test(int);
+        template <typename...>
+        std::false_type static test(...);
+    };
 
-template< template<typename...> class Container
-    , template<typename...> class List
-    , class...Elems
-    , typename...Elements>
-    struct rewrap<Container, List<Elems...>, Elements...> {
-    using type = typename append<TypeList<Container<Elems...>>, typename rewrap<Container, Elements...>::type>::type;
-};
+    template <typename T>
+    struct is_callable;
 
-template< template<typename...> class...Containers >
-struct combine {
-    template< typename...Types >
-    struct with_types {
-        template< template <typename...> class Final >
-        struct into {
-            using type = typename append<Final<>, typename rewrap<Containers, Types...>::type...>::type;
-        };
-    };
-};
+    template <typename Fun, typename... Args>
+    struct is_callable<Fun(Args...)> : decltype(is_callable_tester::test<Fun, Args...>(0)) {};
 
-template<typename T>
-struct always_false : std::false_type {};
+#if defined(__cpp_lib_is_invocable) && __cpp_lib_is_invocable >= 201703
+    // std::result_of is deprecated in C++17 and removed in C++20. Hence, it is
+    // replaced with std::invoke_result here.
+    template <typename Func, typename... U>
+    using FunctionReturnType = std::remove_reference_t<std::remove_cv_t<std::invoke_result_t<Func, U...>>>;
+#else
+    // Keep ::type here because we still support C++11
+    template <typename Func, typename... U>
+    using FunctionReturnType = typename std::remove_reference<typename std::remove_cv<typename std::result_of<Func(U...)>::type>::type>::type;
+#endif
 
 } // namespace Catch
 
+namespace mpl_{
+    struct na;
+}
+
 // end catch_meta.hpp
 namespace Catch {
 
@@ -854,38 +1000,70 @@ struct AutoReg : NonCopyable {
             };                            \
         }                                 \
         void TestName::test()
-    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION( TestName, ... )  \
-        template<typename TestType>                                             \
-        static void TestName()
-    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION( TestName, ClassName, ... )    \
+    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION_2( TestName, TestFunc, Name, Tags, Signature, ... )  \
+        INTERNAL_CATCH_DEFINE_SIG_TEST(TestFunc, INTERNAL_CATCH_REMOVE_PARENS(Signature))
+    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION_2( TestNameClass, TestName, ClassName, Name, Tags, Signature, ... )    \
         namespace{                                                                                  \
-            template<typename TestType>                                                             \
-            struct TestName : INTERNAL_CATCH_REMOVE_PARENS(ClassName <TestType>) {     \
-                void test();                                                                        \
-            };                                                                                      \
+            namespace INTERNAL_CATCH_MAKE_NAMESPACE(TestName) {                                      \
+            INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD(TestName, ClassName, INTERNAL_CATCH_REMOVE_PARENS(Signature));\
         }                                                                                           \
-        template<typename TestType>                                                                 \
-        void TestName::test()
+        }                                                                                           \
+        INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD(TestName, INTERNAL_CATCH_REMOVE_PARENS(Signature))
+
+    #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+        #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(Name, Tags, ...) \
+            INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, typename TestType, __VA_ARGS__ )
+    #else
+        #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(Name, Tags, ...) \
+            INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, typename TestType, __VA_ARGS__ ) )
+    #endif
+
+    #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+        #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG_NO_REGISTRATION(Name, Tags, Signature, ...) \
+            INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, Signature, __VA_ARGS__ )
+    #else
+        #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG_NO_REGISTRATION(Name, Tags, Signature, ...) \
+            INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, Signature, __VA_ARGS__ ) )
+    #endif
+
+    #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+        #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION( ClassName, Name, Tags,... ) \
+            INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_C_L_A_S_S_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ) , ClassName, Name, Tags, typename T, __VA_ARGS__ )
+    #else
+        #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION( ClassName, Name, Tags,... ) \
+            INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_C_L_A_S_S_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ) , ClassName, Name, Tags, typename T, __VA_ARGS__ ) )
+    #endif
+
+    #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+        #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG_NO_REGISTRATION( ClassName, Name, Tags, Signature, ... ) \
+            INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_C_L_A_S_S_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ) , ClassName, Name, Tags, Signature, __VA_ARGS__ )
+    #else
+        #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG_NO_REGISTRATION( ClassName, Name, Tags, Signature, ... ) \
+            INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_C_L_A_S_S_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ) , ClassName, Name, Tags, Signature, __VA_ARGS__ ) )
+    #endif
 #endif
 
     ///////////////////////////////////////////////////////////////////////////////
     #define INTERNAL_CATCH_TESTCASE2( TestName, ... ) \
         static void TestName(); \
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
         CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
         namespace{ Catch::AutoReg INTERNAL_CATCH_UNIQUE_NAME( autoRegistrar )( Catch::makeTestInvoker( &TestName ), CATCH_INTERNAL_LINEINFO, Catch::StringRef(), Catch::NameAndTags{ __VA_ARGS__ } ); } /* NOLINT */ \
-        CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS \
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION \
         static void TestName()
     #define INTERNAL_CATCH_TESTCASE( ... ) \
-        INTERNAL_CATCH_TESTCASE2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ), __VA_ARGS__ )
+        INTERNAL_CATCH_TESTCASE2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ), __VA_ARGS__ )
 
     ///////////////////////////////////////////////////////////////////////////////
     #define INTERNAL_CATCH_METHOD_AS_TEST_CASE( QualifiedMethod, ... ) \
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
         CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
         namespace{ Catch::AutoReg INTERNAL_CATCH_UNIQUE_NAME( autoRegistrar )( Catch::makeTestInvoker( &QualifiedMethod ), CATCH_INTERNAL_LINEINFO, "&" #QualifiedMethod, Catch::NameAndTags{ __VA_ARGS__ } ); } /* NOLINT */ \
-        CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
 
     ///////////////////////////////////////////////////////////////////////////////
     #define INTERNAL_CATCH_TEST_CASE_METHOD2( TestName, ClassName, ... )\
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
         CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
         namespace{ \
             struct TestName : INTERNAL_CATCH_REMOVE_PARENS(ClassName) { \
@@ -893,161 +1071,273 @@ struct AutoReg : NonCopyable {
             }; \
             Catch::AutoReg INTERNAL_CATCH_UNIQUE_NAME( autoRegistrar ) ( Catch::makeTestInvoker( &TestName::test ), CATCH_INTERNAL_LINEINFO, #ClassName, Catch::NameAndTags{ __VA_ARGS__ } ); /* NOLINT */ \
         } \
-        CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS \
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION \
         void TestName::test()
     #define INTERNAL_CATCH_TEST_CASE_METHOD( ClassName, ... ) \
-        INTERNAL_CATCH_TEST_CASE_METHOD2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ), ClassName, __VA_ARGS__ )
+        INTERNAL_CATCH_TEST_CASE_METHOD2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ), ClassName, __VA_ARGS__ )
 
     ///////////////////////////////////////////////////////////////////////////////
     #define INTERNAL_CATCH_REGISTER_TESTCASE( Function, ... ) \
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
         CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
         Catch::AutoReg INTERNAL_CATCH_UNIQUE_NAME( autoRegistrar )( Catch::makeTestInvoker( Function ), CATCH_INTERNAL_LINEINFO, Catch::StringRef(), Catch::NameAndTags{ __VA_ARGS__ } ); /* NOLINT */ \
-        CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
 
     ///////////////////////////////////////////////////////////////////////////////
-    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_2(TestName, TestFunc, Name, Tags, ... )\
+    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_2(TestName, TestFunc, Name, Tags, Signature, ... )\
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
         CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
-        template<typename TestType> \
-        static void TestFunc();\
+        CATCH_INTERNAL_SUPPRESS_ZERO_VARIADIC_WARNINGS \
+        CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS \
+        INTERNAL_CATCH_DECLARE_SIG_TEST(TestFunc, INTERNAL_CATCH_REMOVE_PARENS(Signature));\
         namespace {\
+        namespace INTERNAL_CATCH_MAKE_NAMESPACE(TestName){\
+            INTERNAL_CATCH_TYPE_GEN\
+            INTERNAL_CATCH_NTTP_GEN(INTERNAL_CATCH_REMOVE_PARENS(Signature))\
+            INTERNAL_CATCH_NTTP_REG_GEN(TestFunc,INTERNAL_CATCH_REMOVE_PARENS(Signature))\
             template<typename...Types> \
             struct TestName{\
-                template<typename...Ts> \
-                TestName(Ts...names){\
-                    CATCH_INTERNAL_CHECK_UNIQUE_TYPES(CATCH_REC_LIST(INTERNAL_CATCH_REMOVE_PARENS, __VA_ARGS__)) \
+                TestName(){\
+                    int index = 0;                                    \
+                    constexpr char const* tmpl_types[] = {CATCH_REC_LIST(INTERNAL_CATCH_STRINGIZE_WITHOUT_PARENS, __VA_ARGS__)};\
                     using expander = int[];\
-                    (void)expander{(Catch::AutoReg( Catch::makeTestInvoker( &TestFunc<Types> ), CATCH_INTERNAL_LINEINFO, Catch::StringRef(), Catch::NameAndTags{ names, Tags } ), 0)... };/* NOLINT */ \
+                    (void)expander{(reg_test(Types{}, Catch::NameAndTags{ Name " - " + std::string(tmpl_types[index]), Tags } ), index++)... };/* NOLINT */ \
                 }\
             };\
-            INTERNAL_CATCH_TEMPLATE_REGISTRY_INITIATE(TestName, Name, __VA_ARGS__) \
+            static int INTERNAL_CATCH_UNIQUE_NAME( globalRegistrar ) = [](){\
+            TestName<INTERNAL_CATCH_MAKE_TYPE_LISTS_FROM_TYPES(__VA_ARGS__)>();\
+            return 0;\
+        }();\
         }\
-        CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS \
-        template<typename TestType> \
-        static void TestFunc()
-
-#if defined(CATCH_CPP17_OR_GREATER)
-#define CATCH_INTERNAL_CHECK_UNIQUE_TYPES(...) static_assert(Catch::is_unique<__VA_ARGS__>,"Duplicate type detected in declaration of template test case");
-#else
-#define CATCH_INTERNAL_CHECK_UNIQUE_TYPES(...) static_assert(Catch::is_unique<__VA_ARGS__>::value,"Duplicate type detected in declaration of template test case");
-#endif
+        }\
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION \
+        INTERNAL_CATCH_DEFINE_SIG_TEST(TestFunc,INTERNAL_CATCH_REMOVE_PARENS(Signature))
 
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
     #define INTERNAL_CATCH_TEMPLATE_TEST_CASE(Name, Tags, ...) \
-        INTERNAL_CATCH_TEMPLATE_TEST_CASE_2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____F_U_N_C____ ), Name, Tags, __VA_ARGS__ )
+        INTERNAL_CATCH_TEMPLATE_TEST_CASE_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, typename TestType, __VA_ARGS__ )
 #else
     #define INTERNAL_CATCH_TEMPLATE_TEST_CASE(Name, Tags, ...) \
-        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____F_U_N_C____ ), Name, Tags, __VA_ARGS__ ) )
+        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, typename TestType, __VA_ARGS__ ) )
 #endif
 
-    #define INTERNAL_CATCH_TEMPLATE_REGISTRY_INITIATE(TestName, Name, ...)\
-        static int INTERNAL_CATCH_UNIQUE_NAME( globalRegistrar ) = [](){\
-            TestName<CATCH_REC_LIST(INTERNAL_CATCH_REMOVE_PARENS, __VA_ARGS__)>(CATCH_REC_LIST_UD(INTERNAL_CATCH_TEMPLATE_UNIQUE_NAME,Name, __VA_ARGS__));\
-            return 0;\
-        }();
+#ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG(Name, Tags, Signature, ...) \
+        INTERNAL_CATCH_TEMPLATE_TEST_CASE_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, Signature, __VA_ARGS__ )
+#else
+    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG(Name, Tags, Signature, ...) \
+        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, Signature, __VA_ARGS__ ) )
+#endif
 
-    #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE2(TestName, TestFuncName, Name, Tags, TmplTypes, TypesList) \
+    #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE2(TestName, TestFuncName, Name, Tags, Signature, TmplTypes, TypesList) \
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION                      \
         CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS                      \
+        CATCH_INTERNAL_SUPPRESS_ZERO_VARIADIC_WARNINGS                \
+        CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS              \
         template<typename TestType> static void TestFuncName();       \
-        namespace {                                                   \
+        namespace {\
+        namespace INTERNAL_CATCH_MAKE_NAMESPACE(TestName) {                                     \
+            INTERNAL_CATCH_TYPE_GEN                                                  \
+            INTERNAL_CATCH_NTTP_GEN(INTERNAL_CATCH_REMOVE_PARENS(Signature))         \
             template<typename... Types>                               \
             struct TestName {                                         \
-                TestName() {                                          \
-                    CATCH_INTERNAL_CHECK_UNIQUE_TYPES(Types...)       \
+                void reg_tests() {                                          \
                     int index = 0;                                    \
                     using expander = int[];                           \
                     constexpr char const* tmpl_types[] = {CATCH_REC_LIST(INTERNAL_CATCH_STRINGIZE_WITHOUT_PARENS, INTERNAL_CATCH_REMOVE_PARENS(TmplTypes))};\
                     constexpr char const* types_list[] = {CATCH_REC_LIST(INTERNAL_CATCH_STRINGIZE_WITHOUT_PARENS, INTERNAL_CATCH_REMOVE_PARENS(TypesList))};\
                     constexpr auto num_types = sizeof(types_list) / sizeof(types_list[0]);\
-                    (void)expander{(Catch::AutoReg( Catch::makeTestInvoker( &TestFuncName<Types> ), CATCH_INTERNAL_LINEINFO, Catch::StringRef(), Catch::NameAndTags{ Name " - " + std::string(tmpl_types[index / num_types]) + "<" + std::string(types_list[index % num_types]) + ">", Tags } ), index++, 0)... };/* NOLINT */\
+                    (void)expander{(Catch::AutoReg( Catch::makeTestInvoker( &TestFuncName<Types> ), CATCH_INTERNAL_LINEINFO, Catch::StringRef(), Catch::NameAndTags{ Name " - " + std::string(tmpl_types[index / num_types]) + "<" + std::string(types_list[index % num_types]) + ">", Tags } ), index++)... };/* NOLINT */\
                 }                                                     \
             };                                                        \
             static int INTERNAL_CATCH_UNIQUE_NAME( globalRegistrar ) = [](){ \
-                using TestInit = Catch::combine<INTERNAL_CATCH_REMOVE_PARENS(TmplTypes)> \
-                            ::with_types<INTERNAL_CATCH_MAKE_TYPE_LISTS_FROM_TYPES(TypesList)>::into<TestName>::type; \
-                TestInit();                                           \
+                using TestInit = typename create<TestName, decltype(get_wrapper<INTERNAL_CATCH_REMOVE_PARENS(TmplTypes)>()), TypeList<INTERNAL_CATCH_MAKE_TYPE_LISTS_FROM_TYPES(INTERNAL_CATCH_REMOVE_PARENS(TypesList))>>::type; \
+                TestInit t;                                           \
+                t.reg_tests();                                        \
                 return 0;                                             \
             }();                                                      \
         }                                                             \
-        CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS                    \
+        }                                                             \
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION                       \
         template<typename TestType>                                   \
         static void TestFuncName()
 
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
     #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE(Name, Tags, ...)\
-        INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE2(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____F_U_N_C____ ),Name,Tags,__VA_ARGS__)
+        INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE2(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, typename T,__VA_ARGS__)
 #else
     #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE(Name, Tags, ...)\
-        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____F_U_N_C____ ), Name, Tags, __VA_ARGS__ ) )
+        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, typename T, __VA_ARGS__ ) )
+#endif
+
+#ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+    #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG(Name, Tags, Signature, ...)\
+        INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE2(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, Signature, __VA_ARGS__)
+#else
+    #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG(Name, Tags, Signature, ...)\
+        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, Signature, __VA_ARGS__ ) )
 #endif
 
-    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_2( TestNameClass, TestName, ClassName, Name, Tags, ... ) \
+    #define INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE_2(TestName, TestFunc, Name, Tags, TmplList)\
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
         CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
-        namespace{ \
-            template<typename TestType> \
-            struct TestName : INTERNAL_CATCH_REMOVE_PARENS(ClassName <TestType>) { \
-                void test();\
-            };\
+        CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS \
+        template<typename TestType> static void TestFunc();       \
+        namespace {\
+        namespace INTERNAL_CATCH_MAKE_NAMESPACE(TestName){\
+        INTERNAL_CATCH_TYPE_GEN\
+        template<typename... Types>                               \
+        struct TestName {                                         \
+            void reg_tests() {                                          \
+                int index = 0;                                    \
+                using expander = int[];                           \
+                (void)expander{(Catch::AutoReg( Catch::makeTestInvoker( &TestFunc<Types> ), CATCH_INTERNAL_LINEINFO, Catch::StringRef(), Catch::NameAndTags{ Name " - " + std::string(INTERNAL_CATCH_STRINGIZE(TmplList)) + " - " + std::to_string(index), Tags } ), index++)... };/* NOLINT */\
+            }                                                     \
+        };\
+        static int INTERNAL_CATCH_UNIQUE_NAME( globalRegistrar ) = [](){ \
+                using TestInit = typename convert<TestName, TmplList>::type; \
+                TestInit t;                                           \
+                t.reg_tests();                                        \
+                return 0;                                             \
+            }();                                                      \
+        }}\
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION                       \
+        template<typename TestType>                                   \
+        static void TestFunc()
+
+    #define INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE(Name, Tags, TmplList) \
+        INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), Name, Tags, TmplList )
+
+    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_2( TestNameClass, TestName, ClassName, Name, Tags, Signature, ... ) \
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
+        CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
+        CATCH_INTERNAL_SUPPRESS_ZERO_VARIADIC_WARNINGS \
+        CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS \
+        namespace {\
+        namespace INTERNAL_CATCH_MAKE_NAMESPACE(TestName){ \
+            INTERNAL_CATCH_TYPE_GEN\
+            INTERNAL_CATCH_NTTP_GEN(INTERNAL_CATCH_REMOVE_PARENS(Signature))\
+            INTERNAL_CATCH_DECLARE_SIG_TEST_METHOD(TestName, ClassName, INTERNAL_CATCH_REMOVE_PARENS(Signature));\
+            INTERNAL_CATCH_NTTP_REG_METHOD_GEN(TestName, INTERNAL_CATCH_REMOVE_PARENS(Signature))\
             template<typename...Types> \
             struct TestNameClass{\
-                template<typename...Ts> \
-                TestNameClass(Ts...names){\
-                    CATCH_INTERNAL_CHECK_UNIQUE_TYPES(CATCH_REC_LIST(INTERNAL_CATCH_REMOVE_PARENS, __VA_ARGS__)) \
+                TestNameClass(){\
+                    int index = 0;                                    \
+                    constexpr char const* tmpl_types[] = {CATCH_REC_LIST(INTERNAL_CATCH_STRINGIZE_WITHOUT_PARENS, __VA_ARGS__)};\
                     using expander = int[];\
-                    (void)expander{(Catch::AutoReg( Catch::makeTestInvoker( &TestName<Types>::test ), CATCH_INTERNAL_LINEINFO, #ClassName, Catch::NameAndTags{ names, Tags } ), 0)... };/* NOLINT */ \
+                    (void)expander{(reg_test(Types{}, #ClassName, Catch::NameAndTags{ Name " - " + std::string(tmpl_types[index]), Tags } ), index++)... };/* NOLINT */ \
                 }\
             };\
-            INTERNAL_CATCH_TEMPLATE_REGISTRY_INITIATE(TestNameClass, Name, __VA_ARGS__)\
+            static int INTERNAL_CATCH_UNIQUE_NAME( globalRegistrar ) = [](){\
+                TestNameClass<INTERNAL_CATCH_MAKE_TYPE_LISTS_FROM_TYPES(__VA_ARGS__)>();\
+                return 0;\
+        }();\
         }\
-        CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS\
-        template<typename TestType> \
-        void TestName<TestType>::test()
+        }\
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION \
+        INTERNAL_CATCH_DEFINE_SIG_TEST_METHOD(TestName, INTERNAL_CATCH_REMOVE_PARENS(Signature))
 
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
     #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD( ClassName, Name, Tags,... ) \
-        INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____C_L_A_S_S____ ), INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ) , ClassName, Name, Tags, __VA_ARGS__ )
+        INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_C_L_A_S_S_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ) , ClassName, Name, Tags, typename T, __VA_ARGS__ )
 #else
     #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD( ClassName, Name, Tags,... ) \
-        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____C_L_A_S_S____ ), INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ) , ClassName, Name, Tags, __VA_ARGS__ ) )
+        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_C_L_A_S_S_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ) , ClassName, Name, Tags, typename T, __VA_ARGS__ ) )
+#endif
+
+#ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( ClassName, Name, Tags, Signature, ... ) \
+        INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_C_L_A_S_S_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ) , ClassName, Name, Tags, Signature, __VA_ARGS__ )
+#else
+    #define INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( ClassName, Name, Tags, Signature, ... ) \
+        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_C_L_A_S_S_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ) , ClassName, Name, Tags, Signature, __VA_ARGS__ ) )
 #endif
 
-    #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_2(TestNameClass, TestName, ClassName, Name, Tags, TmplTypes, TypesList)\
+    #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_2(TestNameClass, TestName, ClassName, Name, Tags, Signature, TmplTypes, TypesList)\
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
         CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
+        CATCH_INTERNAL_SUPPRESS_ZERO_VARIADIC_WARNINGS \
+        CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS \
         template<typename TestType> \
             struct TestName : INTERNAL_CATCH_REMOVE_PARENS(ClassName <TestType>) { \
                 void test();\
             };\
         namespace {\
+        namespace INTERNAL_CATCH_MAKE_NAMESPACE(TestNameClass) {\
+            INTERNAL_CATCH_TYPE_GEN                  \
+            INTERNAL_CATCH_NTTP_GEN(INTERNAL_CATCH_REMOVE_PARENS(Signature))\
             template<typename...Types>\
             struct TestNameClass{\
-                TestNameClass(){\
-                    CATCH_INTERNAL_CHECK_UNIQUE_TYPES(Types...)\
+                void reg_tests(){\
                     int index = 0;\
                     using expander = int[];\
                     constexpr char const* tmpl_types[] = {CATCH_REC_LIST(INTERNAL_CATCH_STRINGIZE_WITHOUT_PARENS, INTERNAL_CATCH_REMOVE_PARENS(TmplTypes))};\
                     constexpr char const* types_list[] = {CATCH_REC_LIST(INTERNAL_CATCH_STRINGIZE_WITHOUT_PARENS, INTERNAL_CATCH_REMOVE_PARENS(TypesList))};\
                     constexpr auto num_types = sizeof(types_list) / sizeof(types_list[0]);\
-                    (void)expander{(Catch::AutoReg( Catch::makeTestInvoker( &TestName<Types>::test ), CATCH_INTERNAL_LINEINFO, #ClassName, Catch::NameAndTags{ Name " - " + std::string(tmpl_types[index / num_types]) + "<" + std::string(types_list[index % num_types]) + ">", Tags } ), index++, 0)... };/* NOLINT */ \
+                    (void)expander{(Catch::AutoReg( Catch::makeTestInvoker( &TestName<Types>::test ), CATCH_INTERNAL_LINEINFO, #ClassName, Catch::NameAndTags{ Name " - " + std::string(tmpl_types[index / num_types]) + "<" + std::string(types_list[index % num_types]) + ">", Tags } ), index++)... };/* NOLINT */ \
                 }\
             };\
             static int INTERNAL_CATCH_UNIQUE_NAME( globalRegistrar ) = [](){\
-                using TestInit = Catch::combine<INTERNAL_CATCH_REMOVE_PARENS(TmplTypes)>\
-                            ::with_types<INTERNAL_CATCH_MAKE_TYPE_LISTS_FROM_TYPES(TypesList)>::into<TestNameClass>::type;\
-                TestInit();\
+                using TestInit = typename create<TestNameClass, decltype(get_wrapper<INTERNAL_CATCH_REMOVE_PARENS(TmplTypes)>()), TypeList<INTERNAL_CATCH_MAKE_TYPE_LISTS_FROM_TYPES(INTERNAL_CATCH_REMOVE_PARENS(TypesList))>>::type;\
+                TestInit t;\
+                t.reg_tests();\
                 return 0;\
             }(); \
         }\
-        CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS \
+        }\
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION \
         template<typename TestType> \
         void TestName<TestType>::test()
 
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
     #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( ClassName, Name, Tags, ... )\
-        INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____F_U_N_C____ ), ClassName, Name, Tags, __VA_ARGS__ )
+        INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), ClassName, Name, Tags, typename T, __VA_ARGS__ )
 #else
     #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( ClassName, Name, Tags, ... )\
-        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____F_U_N_C____ ), ClassName, Name, Tags, __VA_ARGS__ ) )
+        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), ClassName, Name, Tags, typename T,__VA_ARGS__ ) )
+#endif
+
+#ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
+    #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( ClassName, Name, Tags, Signature, ... )\
+        INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), ClassName, Name, Tags, Signature, __VA_ARGS__ )
+#else
+    #define INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( ClassName, Name, Tags, Signature, ... )\
+        INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), ClassName, Name, Tags, Signature,__VA_ARGS__ ) )
 #endif
 
+    #define INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE_METHOD_2( TestNameClass, TestName, ClassName, Name, Tags, TmplList) \
+        CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
+        CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
+        CATCH_INTERNAL_SUPPRESS_UNUSED_TEMPLATE_WARNINGS \
+        template<typename TestType> \
+        struct TestName : INTERNAL_CATCH_REMOVE_PARENS(ClassName <TestType>) { \
+            void test();\
+        };\
+        namespace {\
+        namespace INTERNAL_CATCH_MAKE_NAMESPACE(TestName){ \
+            INTERNAL_CATCH_TYPE_GEN\
+            template<typename...Types>\
+            struct TestNameClass{\
+                void reg_tests(){\
+                    int index = 0;\
+                    using expander = int[];\
+                    (void)expander{(Catch::AutoReg( Catch::makeTestInvoker( &TestName<Types>::test ), CATCH_INTERNAL_LINEINFO, #ClassName, Catch::NameAndTags{ Name " - " + std::string(INTERNAL_CATCH_STRINGIZE(TmplList)) + " - " + std::to_string(index), Tags } ), index++)... };/* NOLINT */ \
+                }\
+            };\
+            static int INTERNAL_CATCH_UNIQUE_NAME( globalRegistrar ) = [](){\
+                using TestInit = typename convert<TestNameClass, TmplList>::type;\
+                TestInit t;\
+                t.reg_tests();\
+                return 0;\
+            }(); \
+        }}\
+        CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION \
+        template<typename TestType> \
+        void TestName<TestType>::test()
+
+#define INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE_METHOD(ClassName, Name, Tags, TmplList) \
+        INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE_METHOD_2( INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_ ), INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_M_P_L_A_T_E_T_E_S_T_F_U_N_C_ ), ClassName, Name, Tags, TmplList )
+
 // end catch_test_registry.h
 // start catch_capture.hpp
 
@@ -1147,7 +1437,7 @@ namespace Catch {
 
     auto makeStream( StringRef const &filename ) -> IStream const*;
 
-    class ReusableStringStream {
+    class ReusableStringStream : NonCopyable {
         std::size_t m_index;
         std::ostream* m_oss;
     public:
@@ -1166,6 +1456,42 @@ namespace Catch {
 }
 
 // end catch_stream.h
+// start catch_interfaces_enum_values_registry.h
+
+#include <vector>
+
+namespace Catch {
+
+    namespace Detail {
+        struct EnumInfo {
+            StringRef m_name;
+            std::vector<std::pair<int, StringRef>> m_values;
+
+            ~EnumInfo();
+
+            StringRef lookup( int value ) const;
+        };
+    } // namespace Detail
+
+    struct IMutableEnumValuesRegistry {
+        virtual ~IMutableEnumValuesRegistry();
+
+        virtual Detail::EnumInfo const& registerEnum( StringRef enumName, StringRef allEnums, std::vector<int> const& values ) = 0;
+
+        template<typename E>
+        Detail::EnumInfo const& registerEnum( StringRef enumName, StringRef allEnums, std::initializer_list<E> values ) {
+            static_assert(sizeof(int) >= sizeof(E), "Cannot serialize enum to int");
+            std::vector<int> intValues;
+            intValues.reserve( values.size() );
+            for( auto enumValue : values )
+                intValues.push_back( static_cast<int>( enumValue ) );
+            return registerEnum( enumName, allEnums, intValues );
+        }
+    };
+
+} // Catch
+
+// end catch_interfaces_enum_values_registry.h
 
 #ifdef CATCH_CONFIG_CPP17_STRING_VIEW
 #include <string_view>
@@ -1236,9 +1562,9 @@ namespace Catch {
 
         template<typename T>
         class IsStreamInsertable {
-            template<typename SS, typename TT>
+            template<typename Stream, typename U>
             static auto test(int)
-                -> decltype(std::declval<SS&>() << std::declval<TT>(), std::true_type());
+                -> decltype(std::declval<Stream&>() << std::declval<U>(), std::true_type());
 
             template<typename, typename>
             static auto test(...)->std::false_type;
@@ -1400,6 +1726,12 @@ namespace Catch {
         }
     };
 
+#if defined(CATCH_CONFIG_CPP17_BYTE)
+    template<>
+    struct StringMaker<std::byte> {
+        static std::string convert(std::byte value);
+    };
+#endif // defined(CATCH_CONFIG_CPP17_BYTE)
     template<>
     struct StringMaker<int> {
         static std::string convert(int value);
@@ -1451,10 +1783,13 @@ namespace Catch {
     template<>
     struct StringMaker<float> {
         static std::string convert(float value);
+        static int precision;
     };
+
     template<>
     struct StringMaker<double> {
         static std::string convert(double value);
+        static int precision;
     };
 
     template <typename T>
@@ -1490,8 +1825,8 @@ namespace Catch {
 #endif
 
     namespace Detail {
-        template<typename InputIterator>
-        std::string rangeToString(InputIterator first, InputIterator last) {
+        template<typename InputIterator, typename Sentinel = InputIterator>
+        std::string rangeToString(InputIterator first, Sentinel last) {
             ReusableStringStream rss;
             rss << "{ ";
             if (first != last) {
@@ -1649,20 +1984,27 @@ namespace Catch {
 #endif // CATCH_CONFIG_ENABLE_VARIANT_STRINGMAKER
 
 namespace Catch {
-    struct not_this_one {}; // Tag type for detecting which begin/ end are being selected
-
-    // Import begin/ end from std here so they are considered alongside the fallback (...) overloads in this namespace
+    // Import begin/ end from std here
     using std::begin;
     using std::end;
 
-    not_this_one begin( ... );
-    not_this_one end( ... );
+    namespace detail {
+        template <typename...>
+        struct void_type {
+            using type = void;
+        };
+
+        template <typename T, typename = void>
+        struct is_range_impl : std::false_type {
+        };
+
+        template <typename T>
+        struct is_range_impl<T, typename void_type<decltype(begin(std::declval<T>()))>::type> : std::true_type {
+        };
+    } // namespace detail
 
     template <typename T>
-    struct is_range {
-        static const bool value =
-            !std::is_same<decltype(begin(std::declval<T>())), not_this_one>::value &&
-            !std::is_same<decltype(end(std::declval<T>())), not_this_one>::value;
+    struct is_range : detail::is_range_impl<T> {
     };
 
 #if defined(_MANAGED) // Managed types are never ranges
@@ -1827,6 +2169,18 @@ struct ratio_string<std::milli> {
 }
 #endif // CATCH_CONFIG_ENABLE_CHRONO_STRINGMAKER
 
+#define INTERNAL_CATCH_REGISTER_ENUM( enumName, ... ) \
+namespace Catch { \
+    template<> struct StringMaker<enumName> { \
+        static std::string convert( enumName value ) { \
+            static const auto& enumInfo = ::Catch::getMutableRegistryHub().getMutableEnumValuesRegistry().registerEnum( #enumName, #__VA_ARGS__, { __VA_ARGS__ } ); \
+            return static_cast<std::string>(enumInfo.lookup( static_cast<int>( value ) )); \
+        } \
+    }; \
+}
+
+#define CATCH_REGISTER_ENUM( enumName, ... ) INTERNAL_CATCH_REGISTER_ENUM( enumName, __VA_ARGS__ )
+
 #ifdef _MSC_VER
 #pragma warning(pop)
 #endif
@@ -2018,6 +2372,18 @@ namespace Catch {
         auto operator <= ( RhsT const& rhs ) -> BinaryExpr<LhsT, RhsT const&> const {
             return { static_cast<bool>(m_lhs <= rhs), m_lhs, "<=", rhs };
         }
+        template <typename RhsT>
+        auto operator | (RhsT const& rhs) -> BinaryExpr<LhsT, RhsT const&> const {
+            return { static_cast<bool>(m_lhs | rhs), m_lhs, "|", rhs };
+        }
+        template <typename RhsT>
+        auto operator & (RhsT const& rhs) -> BinaryExpr<LhsT, RhsT const&> const {
+            return { static_cast<bool>(m_lhs & rhs), m_lhs, "&", rhs };
+        }
+        template <typename RhsT>
+        auto operator ^ (RhsT const& rhs) -> BinaryExpr<LhsT, RhsT const&> const {
+            return { static_cast<bool>(m_lhs ^ rhs), m_lhs, "^", rhs };
+        }
 
         template<typename RhsT>
         auto operator && ( RhsT const& ) -> BinaryExpr<LhsT, RhsT const&> const {
@@ -2066,6 +2432,7 @@ namespace Catch {
 // start catch_interfaces_capture.h
 
 #include <string>
+#include <chrono>
 
 namespace Catch {
 
@@ -2076,14 +2443,18 @@ namespace Catch {
     struct MessageInfo;
     struct MessageBuilder;
     struct Counts;
-    struct BenchmarkInfo;
-    struct BenchmarkStats;
     struct AssertionReaction;
     struct SourceLineInfo;
 
     struct ITransientExpression;
     struct IGeneratorTracker;
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+    struct BenchmarkInfo;
+    template <typename Duration = std::chrono::duration<double, std::nano>>
+    struct BenchmarkStats;
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
+
     struct IResultCapture {
 
         virtual ~IResultCapture();
@@ -2093,10 +2464,14 @@ namespace Catch {
         virtual void sectionEnded( SectionEndInfo const& endInfo ) = 0;
         virtual void sectionEndedEarly( SectionEndInfo const& endInfo ) = 0;
 
-        virtual auto acquireGeneratorTracker( SourceLineInfo const& lineInfo ) -> IGeneratorTracker& = 0;
+        virtual auto acquireGeneratorTracker( StringRef generatorName, SourceLineInfo const& lineInfo ) -> IGeneratorTracker& = 0;
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+        virtual void benchmarkPreparing( std::string const& name ) = 0;
         virtual void benchmarkStarting( BenchmarkInfo const& info ) = 0;
-        virtual void benchmarkEnded( BenchmarkStats const& stats ) = 0;
+        virtual void benchmarkEnded( BenchmarkStats<> const& stats ) = 0;
+        virtual void benchmarkFailed( std::string const& error ) = 0;
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
         virtual void pushScopedMessage( MessageInfo const& message ) = 0;
         virtual void popScopedMessage( MessageInfo const& message ) = 0;
@@ -2327,15 +2702,16 @@ namespace Catch {
 ///////////////////////////////////////////////////////////////////////////////
 #define INTERNAL_CATCH_TEST( macroName, resultDisposition, ... ) \
     do { \
+        CATCH_INTERNAL_IGNORE_BUT_WARN(__VA_ARGS__); \
         Catch::AssertionHandler catchAssertionHandler( macroName##_catch_sr, CATCH_INTERNAL_LINEINFO, CATCH_INTERNAL_STRINGIFY(__VA_ARGS__), resultDisposition ); \
         INTERNAL_CATCH_TRY { \
+            CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
             CATCH_INTERNAL_SUPPRESS_PARENTHESES_WARNINGS \
             catchAssertionHandler.handleExpr( Catch::Decomposer() <= __VA_ARGS__ ); \
-            CATCH_INTERNAL_UNSUPPRESS_PARENTHESES_WARNINGS \
+            CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION \
         } INTERNAL_CATCH_CATCH( catchAssertionHandler ) \
         INTERNAL_CATCH_REACT( catchAssertionHandler ) \
-    } while( (void)0, (false) && static_cast<bool>( !!(__VA_ARGS__) ) ) // the expression here is never evaluated at runtime but it forces the compiler to give it a look
-    // The double negation silences MSVC's C4800 warning, the static_cast forces short-circuit evaluation if the type has overloaded &&.
+    } while( (void)0, (false) && static_cast<bool>( !!(__VA_ARGS__) ) )
 
 ///////////////////////////////////////////////////////////////////////////////
 #define INTERNAL_CATCH_IF( macroName, resultDisposition, ... ) \
@@ -2552,62 +2928,18 @@ namespace Catch {
 } // end namespace Catch
 
 #define INTERNAL_CATCH_SECTION( ... ) \
+    CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
     CATCH_INTERNAL_SUPPRESS_UNUSED_WARNINGS \
     if( Catch::Section const& INTERNAL_CATCH_UNIQUE_NAME( catch_internal_Section ) = Catch::SectionInfo( CATCH_INTERNAL_LINEINFO, __VA_ARGS__ ) ) \
-    CATCH_INTERNAL_UNSUPPRESS_UNUSED_WARNINGS
+    CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
 
 #define INTERNAL_CATCH_DYNAMIC_SECTION( ... ) \
+    CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
     CATCH_INTERNAL_SUPPRESS_UNUSED_WARNINGS \
     if( Catch::Section const& INTERNAL_CATCH_UNIQUE_NAME( catch_internal_Section ) = Catch::SectionInfo( CATCH_INTERNAL_LINEINFO, (Catch::ReusableStringStream() << __VA_ARGS__).str() ) ) \
-    CATCH_INTERNAL_UNSUPPRESS_UNUSED_WARNINGS
+    CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
 
 // end catch_section.h
-// start catch_benchmark.h
-
-#include <cstdint>
-#include <string>
-
-namespace Catch {
-
-    class BenchmarkLooper {
-
-        std::string m_name;
-        std::size_t m_count = 0;
-        std::size_t m_iterationsToRun = 1;
-        uint64_t m_resolution;
-        Timer m_timer;
-
-        static auto getResolution() -> uint64_t;
-    public:
-        // Keep most of this inline as it's on the code path that is being timed
-        BenchmarkLooper( StringRef name )
-        :   m_name( name ),
-            m_resolution( getResolution() )
-        {
-            reportStart();
-            m_timer.start();
-        }
-
-        explicit operator bool() {
-            if( m_count < m_iterationsToRun )
-                return true;
-            return needsMoreIterations();
-        }
-
-        void increment() {
-            ++m_count;
-        }
-
-        void reportStart();
-        auto needsMoreIterations() -> bool;
-    };
-
-} // end namespace Catch
-
-#define BENCHMARK( name ) \
-    for( Catch::BenchmarkLooper looper( name ); looper; looper.increment() )
-
-// end catch_benchmark.h
 // start catch_interfaces_exception.h
 
 // start catch_interfaces_registry_hub.h
@@ -2624,6 +2956,8 @@ namespace Catch {
     struct IReporterRegistry;
     struct IReporterFactory;
     struct ITagAliasRegistry;
+    struct IMutableEnumValuesRegistry;
+
     class StartupExceptionRegistry;
 
     using IReporterFactoryPtr = std::shared_ptr<IReporterFactory>;
@@ -2634,7 +2968,6 @@ namespace Catch {
         virtual IReporterRegistry const& getReporterRegistry() const = 0;
         virtual ITestCaseRegistry const& getTestCaseRegistry() const = 0;
         virtual ITagAliasRegistry const& getTagAliasRegistry() const = 0;
-
         virtual IExceptionTranslatorRegistry const& getExceptionTranslatorRegistry() const = 0;
 
         virtual StartupExceptionRegistry const& getStartupExceptionRegistry() const = 0;
@@ -2648,6 +2981,7 @@ namespace Catch {
         virtual void registerTranslator( const IExceptionTranslator* translator ) = 0;
         virtual void registerTagAlias( std::string const& alias, std::string const& tag, SourceLineInfo const& lineInfo ) = 0;
         virtual void registerStartupException() noexcept = 0;
+        virtual IMutableEnumValuesRegistry& getMutableEnumValuesRegistry() = 0;
     };
 
     IRegistryHub const& getRegistryHub();
@@ -2694,6 +3028,9 @@ namespace Catch {
             {}
 
             std::string translate( ExceptionTranslators::const_iterator it, ExceptionTranslators::const_iterator itEnd ) const override {
+#if defined(CATCH_CONFIG_DISABLE_EXCEPTIONS)
+                return "";
+#else
                 try {
                     if( it == itEnd )
                         std::rethrow_exception(std::current_exception());
@@ -2703,6 +3040,7 @@ namespace Catch {
                 catch( T& ex ) {
                     return m_translateFunction( ex );
                 }
+#endif
             }
 
         protected:
@@ -2721,9 +3059,10 @@ namespace Catch {
 ///////////////////////////////////////////////////////////////////////////////
 #define INTERNAL_CATCH_TRANSLATE_EXCEPTION2( translatorName, signature ) \
     static std::string translatorName( signature ); \
+    CATCH_INTERNAL_START_WARNINGS_SUPPRESSION \
     CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS \
     namespace{ Catch::ExceptionTranslatorRegistrar INTERNAL_CATCH_UNIQUE_NAME( catch_internal_ExceptionRegistrar )( &translatorName ); } \
-    CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS \
+    CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION \
     static std::string translatorName( signature )
 
 #define INTERNAL_CATCH_TRANSLATE_EXCEPTION( signature ) INTERNAL_CATCH_TRANSLATE_EXCEPTION2( INTERNAL_CATCH_UNIQUE_NAME( catch_internal_ExceptionTranslator ), signature )
@@ -2754,7 +3093,7 @@ namespace Detail {
         Approx operator-() const;
 
         template <typename T, typename = typename std::enable_if<std::is_constructible<double, T>::value>::type>
-        Approx operator()( T const& value ) {
+        Approx operator()( T const& value ) const {
             Approx approx( static_cast<double>(value) );
             approx.m_epsilon = m_epsilon;
             approx.m_margin = m_margin;
@@ -2854,6 +3193,7 @@ struct StringMaker<Catch::Detail::Approx> {
 
 #include <string>
 #include <iosfwd>
+#include <vector>
 
 namespace Catch {
 
@@ -2864,7 +3204,13 @@ namespace Catch {
     bool contains( std::string const& s, std::string const& infix );
     void toLowerInPlace( std::string& s );
     std::string toLower( std::string const& s );
+    //! Returns a new string without whitespace at the start/end
     std::string trim( std::string const& str );
+    //! Returns a substring of the original ref without whitespace. Beware lifetimes!
+    StringRef trim(StringRef ref);
+
+    // !!! Be aware, returns refs into original string - make sure original string outlives them
+    std::vector<StringRef> splitStringRef( StringRef str, char delimiter );
     bool replaceInPlace( std::string& str, std::string const& replaceThis, std::string const& withThis );
 
     struct pluralise {
@@ -2917,6 +3263,15 @@ namespace Matchers {
             virtual bool match( ObjectT const& arg ) const = 0;
         };
 
+#if defined(__OBJC__)
+        // Hack to fix Catch GH issue #1661. Could use id for generic Object support.
+        // use of const for Object pointers is very uncommon and under ARC it causes some kind of signature mismatch that breaks compilation
+        template<>
+        struct MatcherMethod<NSString*> {
+            virtual bool match( NSString* arg ) const = 0;
+        };
+#endif
+
 #ifdef __clang__
 #    pragma clang diagnostic pop
 #endif
@@ -2954,9 +3309,10 @@ namespace Matchers {
                 return description;
             }
 
-            MatchAllOf<ArgT>& operator && ( MatcherBase<ArgT> const& other ) {
-                m_matchers.push_back( &other );
-                return *this;
+            MatchAllOf<ArgT> operator && ( MatcherBase<ArgT> const& other ) {
+                auto copy(*this);
+                copy.m_matchers.push_back( &other );
+                return copy;
             }
 
             std::vector<MatcherBase<ArgT> const*> m_matchers;
@@ -2987,9 +3343,10 @@ namespace Matchers {
                 return description;
             }
 
-            MatchAnyOf<ArgT>& operator || ( MatcherBase<ArgT> const& other ) {
-                m_matchers.push_back( &other );
-                return *this;
+            MatchAnyOf<ArgT> operator || ( MatcherBase<ArgT> const& other ) {
+                auto copy(*this);
+                copy.m_matchers.push_back( &other );
+                return copy;
             }
 
             std::vector<MatcherBase<ArgT> const*> m_matchers;
@@ -3033,10 +3390,34 @@ using Matchers::Impl::MatcherBase;
 } // namespace Catch
 
 // end catch_matchers.h
-// start catch_matchers_floating.h
+// start catch_matchers_exception.hpp
 
-#include <type_traits>
-#include <cmath>
+namespace Catch {
+namespace Matchers {
+namespace Exception {
+
+class ExceptionMessageMatcher : public MatcherBase<std::exception> {
+    std::string m_message;
+public:
+
+    ExceptionMessageMatcher(std::string const& message):
+        m_message(message)
+    {}
+
+    bool match(std::exception const& ex) const override;
+
+    std::string describe() const override;
+};
+
+} // namespace Exception
+
+Exception::ExceptionMessageMatcher Message(std::string const& message);
+
+} // namespace Matchers
+} // namespace Catch
+
+// end catch_matchers_exception.hpp
+// start catch_matchers_floating.h
 
 namespace Catch {
 namespace Matchers {
@@ -3055,22 +3436,43 @@ namespace Matchers {
         };
 
         struct WithinUlpsMatcher : MatcherBase<double> {
-            WithinUlpsMatcher(double target, int ulps, FloatingPointKind baseType);
+            WithinUlpsMatcher(double target, uint64_t ulps, FloatingPointKind baseType);
             bool match(double const& matchee) const override;
             std::string describe() const override;
         private:
             double m_target;
-            int m_ulps;
+            uint64_t m_ulps;
             FloatingPointKind m_type;
         };
 
+        // Given IEEE-754 format for floats and doubles, we can assume
+        // that float -> double promotion is lossless. Given this, we can
+        // assume that if we do the standard relative comparison of
+        // |lhs - rhs| <= epsilon * max(fabs(lhs), fabs(rhs)), then we get
+        // the same result if we do this for floats, as if we do this for
+        // doubles that were promoted from floats.
+        struct WithinRelMatcher : MatcherBase<double> {
+            WithinRelMatcher(double target, double epsilon);
+            bool match(double const& matchee) const override;
+            std::string describe() const override;
+        private:
+            double m_target;
+            double m_epsilon;
+        };
+
     } // namespace Floating
 
     // The following functions create the actual matcher objects.
     // This allows the types to be inferred
-    Floating::WithinUlpsMatcher WithinULP(double target, int maxUlpDiff);
-    Floating::WithinUlpsMatcher WithinULP(float target, int maxUlpDiff);
+    Floating::WithinUlpsMatcher WithinULP(double target, uint64_t maxUlpDiff);
+    Floating::WithinUlpsMatcher WithinULP(float target, uint64_t maxUlpDiff);
     Floating::WithinAbsMatcher WithinAbs(double target, double margin);
+    Floating::WithinRelMatcher WithinRel(double target, double eps);
+    // defaults epsilon to 100*numeric_limits<double>::epsilon()
+    Floating::WithinRelMatcher WithinRel(double target);
+    Floating::WithinRelMatcher WithinRel(float target, float eps);
+    // defaults epsilon to 100*numeric_limits<float>::epsilon()
+    Floating::WithinRelMatcher WithinRel(float target);
 
 } // namespace Matchers
 } // namespace Catch
@@ -3113,7 +3515,7 @@ public:
 
     // The following functions create the actual matcher objects.
     // The user has to explicitly specify type to the function, because
-    // infering std::function<bool(T const&)> is hard (but possible) and
+    // inferring std::function<bool(T const&)> is hard (but possible) and
     // requires a lot of TMP.
     template<typename T>
     Generic::PredicateMatcher<T> Predicate(std::function<bool(T const&)> const& predicate, std::string const& description = "") {
@@ -3201,34 +3603,12 @@ namespace Catch {
 namespace Matchers {
 
     namespace Vector {
-        namespace Detail {
-            template <typename InputIterator, typename T>
-            size_t count(InputIterator first, InputIterator last, T const& item) {
-                size_t cnt = 0;
-                for (; first != last; ++first) {
-                    if (*first == item) {
-                        ++cnt;
-                    }
-                }
-                return cnt;
-            }
-            template <typename InputIterator, typename T>
-            bool contains(InputIterator first, InputIterator last, T const& item) {
-                for (; first != last; ++first) {
-                    if (*first == item) {
-                        return true;
-                    }
-                }
-                return false;
-            }
-        }
-
-        template<typename T>
-        struct ContainsElementMatcher : MatcherBase<std::vector<T>> {
+        template<typename T, typename Alloc>
+        struct ContainsElementMatcher : MatcherBase<std::vector<T, Alloc>> {
 
             ContainsElementMatcher(T const &comparator) : m_comparator( comparator) {}
 
-            bool match(std::vector<T> const &v) const override {
+            bool match(std::vector<T, Alloc> const &v) const override {
                 for (auto const& el : v) {
                     if (el == m_comparator) {
                         return true;
@@ -3244,12 +3624,12 @@ namespace Matchers {
             T const& m_comparator;
         };
 
-        template<typename T>
-        struct ContainsMatcher : MatcherBase<std::vector<T>> {
+        template<typename T, typename AllocComp, typename AllocMatch>
+        struct ContainsMatcher : MatcherBase<std::vector<T, AllocMatch>> {
 
-            ContainsMatcher(std::vector<T> const &comparator) : m_comparator( comparator ) {}
+            ContainsMatcher(std::vector<T, AllocComp> const &comparator) : m_comparator( comparator ) {}
 
-            bool match(std::vector<T> const &v) const override {
+            bool match(std::vector<T, AllocMatch> const &v) const override {
                 // !TBD: see note in EqualsMatcher
                 if (m_comparator.size() > v.size())
                     return false;
@@ -3271,18 +3651,18 @@ namespace Matchers {
                 return "Contains: " + ::Catch::Detail::stringify( m_comparator );
             }
 
-            std::vector<T> const& m_comparator;
+            std::vector<T, AllocComp> const& m_comparator;
         };
 
-        template<typename T>
-        struct EqualsMatcher : MatcherBase<std::vector<T>> {
+        template<typename T, typename AllocComp, typename AllocMatch>
+        struct EqualsMatcher : MatcherBase<std::vector<T, AllocMatch>> {
 
-            EqualsMatcher(std::vector<T> const &comparator) : m_comparator( comparator ) {}
+            EqualsMatcher(std::vector<T, AllocComp> const &comparator) : m_comparator( comparator ) {}
 
-            bool match(std::vector<T> const &v) const override {
+            bool match(std::vector<T, AllocMatch> const &v) const override {
                 // !TBD: This currently works if all elements can be compared using !=
                 // - a more general approach would be via a compare template that defaults
-                // to using !=. but could be specialised for, e.g. std::vector<T> etc
+                // to using !=. but could be specialised for, e.g. std::vector<T, Alloc> etc
                 // - then just call that directly
                 if (m_comparator.size() != v.size())
                     return false;
@@ -3294,47 +3674,60 @@ namespace Matchers {
             std::string describe() const override {
                 return "Equals: " + ::Catch::Detail::stringify( m_comparator );
             }
-            std::vector<T> const& m_comparator;
+            std::vector<T, AllocComp> const& m_comparator;
         };
 
-        template<typename T>
-        struct UnorderedEqualsMatcher : MatcherBase<std::vector<T>> {
-            UnorderedEqualsMatcher(std::vector<T> const& target) : m_target(target) {}
-            bool match(std::vector<T> const& vec) const override {
-                // Note: This is a reimplementation of std::is_permutation,
-                //       because I don't want to include <algorithm> inside the common path
-                if (m_target.size() != vec.size()) {
-                    return false;
-                }
-                auto lfirst = m_target.begin(), llast = m_target.end();
-                auto rfirst = vec.begin(), rlast = vec.end();
-                // Cut common prefix to optimize checking of permuted parts
-                while (lfirst != llast && *lfirst == *rfirst) {
-                    ++lfirst; ++rfirst;
-                }
-                if (lfirst == llast) {
-                    return true;
-                }
+        template<typename T, typename AllocComp, typename AllocMatch>
+        struct ApproxMatcher : MatcherBase<std::vector<T, AllocMatch>> {
 
-                for (auto mid = lfirst; mid != llast; ++mid) {
-                    // Skip already counted items
-                    if (Detail::contains(lfirst, mid, *mid)) {
-                        continue;
-                    }
-                    size_t num_vec = Detail::count(rfirst, rlast, *mid);
-                    if (num_vec == 0 || Detail::count(lfirst, llast, *mid) != num_vec) {
-                        return false;
-                    }
-                }
+            ApproxMatcher(std::vector<T, AllocComp> const& comparator) : m_comparator( comparator ) {}
 
+            bool match(std::vector<T, AllocMatch> const &v) const override {
+                if (m_comparator.size() != v.size())
+                    return false;
+                for (std::size_t i = 0; i < v.size(); ++i)
+                    if (m_comparator[i] != approx(v[i]))
+                        return false;
                 return true;
             }
+            std::string describe() const override {
+                return "is approx: " + ::Catch::Detail::stringify( m_comparator );
+            }
+            template <typename = typename std::enable_if<std::is_constructible<double, T>::value>::type>
+            ApproxMatcher& epsilon( T const& newEpsilon ) {
+                approx.epsilon(newEpsilon);
+                return *this;
+            }
+            template <typename = typename std::enable_if<std::is_constructible<double, T>::value>::type>
+            ApproxMatcher& margin( T const& newMargin ) {
+                approx.margin(newMargin);
+                return *this;
+            }
+            template <typename = typename std::enable_if<std::is_constructible<double, T>::value>::type>
+            ApproxMatcher& scale( T const& newScale ) {
+                approx.scale(newScale);
+                return *this;
+            }
+
+            std::vector<T, AllocComp> const& m_comparator;
+            mutable Catch::Detail::Approx approx = Catch::Detail::Approx::custom();
+        };
+
+        template<typename T, typename AllocComp, typename AllocMatch>
+        struct UnorderedEqualsMatcher : MatcherBase<std::vector<T, AllocMatch>> {
+            UnorderedEqualsMatcher(std::vector<T, AllocComp> const& target) : m_target(target) {}
+            bool match(std::vector<T, AllocMatch> const& vec) const override {
+                if (m_target.size() != vec.size()) {
+                    return false;
+                }
+                return std::is_permutation(m_target.begin(), m_target.end(), vec.begin());
+            }
 
             std::string describe() const override {
                 return "UnorderedEquals: " + ::Catch::Detail::stringify(m_target);
             }
         private:
-            std::vector<T> const& m_target;
+            std::vector<T, AllocComp> const& m_target;
         };
 
     } // namespace Vector
@@ -3342,24 +3735,29 @@ namespace Matchers {
     // The following functions create the actual matcher objects.
     // This allows the types to be inferred
 
-    template<typename T>
-    Vector::ContainsMatcher<T> Contains( std::vector<T> const& comparator ) {
-        return Vector::ContainsMatcher<T>( comparator );
+    template<typename T, typename AllocComp = std::allocator<T>, typename AllocMatch = AllocComp>
+    Vector::ContainsMatcher<T, AllocComp, AllocMatch> Contains( std::vector<T, AllocComp> const& comparator ) {
+        return Vector::ContainsMatcher<T, AllocComp, AllocMatch>( comparator );
     }
 
-    template<typename T>
-    Vector::ContainsElementMatcher<T> VectorContains( T const& comparator ) {
-        return Vector::ContainsElementMatcher<T>( comparator );
+    template<typename T, typename Alloc = std::allocator<T>>
+    Vector::ContainsElementMatcher<T, Alloc> VectorContains( T const& comparator ) {
+        return Vector::ContainsElementMatcher<T, Alloc>( comparator );
     }
 
-    template<typename T>
-    Vector::EqualsMatcher<T> Equals( std::vector<T> const& comparator ) {
-        return Vector::EqualsMatcher<T>( comparator );
+    template<typename T, typename AllocComp = std::allocator<T>, typename AllocMatch = AllocComp>
+    Vector::EqualsMatcher<T, AllocComp, AllocMatch> Equals( std::vector<T, AllocComp> const& comparator ) {
+        return Vector::EqualsMatcher<T, AllocComp, AllocMatch>( comparator );
     }
 
-    template<typename T>
-    Vector::UnorderedEqualsMatcher<T> UnorderedEquals(std::vector<T> const& target) {
-        return Vector::UnorderedEqualsMatcher<T>(target);
+    template<typename T, typename AllocComp = std::allocator<T>, typename AllocMatch = AllocComp>
+    Vector::ApproxMatcher<T, AllocComp, AllocMatch> Approx( std::vector<T, AllocComp> const& comparator ) {
+        return Vector::ApproxMatcher<T, AllocComp, AllocMatch>( comparator );
+    }
+
+    template<typename T, typename AllocComp = std::allocator<T>, typename AllocMatch = AllocComp>
+    Vector::UnorderedEqualsMatcher<T, AllocComp, AllocMatch> UnorderedEquals(std::vector<T, AllocComp> const& target) {
+        return Vector::UnorderedEqualsMatcher<T, AllocComp, AllocMatch>( target );
     }
 
 } // namespace Matchers
@@ -3470,7 +3868,7 @@ namespace Catch {
 // end catch_interfaces_generatortracker.h
 // start catch_enforce.h
 
-#include <stdexcept>
+#include <exception>
 
 namespace Catch {
 #if !defined(CATCH_CONFIG_DISABLE_EXCEPTIONS)
@@ -3483,18 +3881,30 @@ namespace Catch {
     [[noreturn]]
     void throw_exception(std::exception const& e);
 #endif
+
+    [[noreturn]]
+    void throw_logic_error(std::string const& msg);
+    [[noreturn]]
+    void throw_domain_error(std::string const& msg);
+    [[noreturn]]
+    void throw_runtime_error(std::string const& msg);
+
 } // namespace Catch;
 
-#define CATCH_PREPARE_EXCEPTION( type, msg ) \
-    type( ( Catch::ReusableStringStream() << msg ).str() )
-#define CATCH_INTERNAL_ERROR( msg ) \
-    Catch::throw_exception(CATCH_PREPARE_EXCEPTION( std::logic_error, CATCH_INTERNAL_LINEINFO << ": Internal Catch error: " << msg))
-#define CATCH_ERROR( msg ) \
-    Catch::throw_exception(CATCH_PREPARE_EXCEPTION( std::domain_error, msg ))
-#define CATCH_RUNTIME_ERROR( msg ) \
-    Catch::throw_exception(CATCH_PREPARE_EXCEPTION( std::runtime_error, msg ))
-#define CATCH_ENFORCE( condition, msg ) \
-    do{ if( !(condition) ) CATCH_ERROR( msg ); } while(false)
+#define CATCH_MAKE_MSG(...) \
+    (Catch::ReusableStringStream() << __VA_ARGS__).str()
+
+#define CATCH_INTERNAL_ERROR(...) \
+    Catch::throw_logic_error(CATCH_MAKE_MSG( CATCH_INTERNAL_LINEINFO << ": Internal Catch2 error: " << __VA_ARGS__))
+
+#define CATCH_ERROR(...) \
+    Catch::throw_domain_error(CATCH_MAKE_MSG( __VA_ARGS__ ))
+
+#define CATCH_RUNTIME_ERROR(...) \
+    Catch::throw_runtime_error(CATCH_MAKE_MSG( __VA_ARGS__ ))
+
+#define CATCH_ENFORCE( condition, ... ) \
+    do{ if( !(condition) ) CATCH_ERROR( __VA_ARGS__ ); } while(false)
 
 // end catch_enforce.h
 #include <memory>
@@ -3543,7 +3953,6 @@ namespace Generators {
     class SingleValueGenerator final : public IGenerator<T> {
         T m_value;
     public:
-        SingleValueGenerator(T const& value) : m_value( value ) {}
         SingleValueGenerator(T&& value) : m_value(std::move(value)) {}
 
         T const& get() const override {
@@ -3556,6 +3965,9 @@ namespace Generators {
 
     template<typename T>
     class FixedValuesGenerator final : public IGenerator<T> {
+        static_assert(!std::is_same<T, bool>::value,
+            "FixedValuesGenerator does not support bools because of std::vector<bool>"
+            "specialization, use SingleValue Generator instead.");
         std::vector<T> m_values;
         size_t m_idx = 0;
     public:
@@ -3603,21 +4015,21 @@ namespace Generators {
             m_generators.emplace_back(std::move(generator));
         }
         void populate(T&& val) {
-            m_generators.emplace_back(value(std::move(val)));
+            m_generators.emplace_back(value(std::forward<T>(val)));
         }
         template<typename U>
         void populate(U&& val) {
-            populate(T(std::move(val)));
+            populate(T(std::forward<U>(val)));
         }
         template<typename U, typename... Gs>
-        void populate(U&& valueOrGenerator, Gs... moreGenerators) {
+        void populate(U&& valueOrGenerator, Gs &&... moreGenerators) {
             populate(std::forward<U>(valueOrGenerator));
             populate(std::forward<Gs>(moreGenerators)...);
         }
 
     public:
         template <typename... Gs>
-        Generators(Gs... moreGenerators) {
+        Generators(Gs &&... moreGenerators) {
             m_generators.reserve(sizeof...(Gs));
             populate(std::forward<Gs>(moreGenerators)...);
         }
@@ -3648,7 +4060,7 @@ namespace Generators {
     struct as {};
 
     template<typename T, typename... Gs>
-    auto makeGenerators( GeneratorWrapper<T>&& generator, Gs... moreGenerators ) -> Generators<T> {
+    auto makeGenerators( GeneratorWrapper<T>&& generator, Gs &&... moreGenerators ) -> Generators<T> {
         return Generators<T>(std::move(generator), std::forward<Gs>(moreGenerators)...);
     }
     template<typename T>
@@ -3656,24 +4068,24 @@ namespace Generators {
         return Generators<T>(std::move(generator));
     }
     template<typename T, typename... Gs>
-    auto makeGenerators( T&& val, Gs... moreGenerators ) -> Generators<T> {
+    auto makeGenerators( T&& val, Gs &&... moreGenerators ) -> Generators<T> {
         return makeGenerators( value( std::forward<T>( val ) ), std::forward<Gs>( moreGenerators )... );
     }
     template<typename T, typename U, typename... Gs>
-    auto makeGenerators( as<T>, U&& val, Gs... moreGenerators ) -> Generators<T> {
+    auto makeGenerators( as<T>, U&& val, Gs &&... moreGenerators ) -> Generators<T> {
         return makeGenerators( value( T( std::forward<U>( val ) ) ), std::forward<Gs>( moreGenerators )... );
     }
 
-    auto acquireGeneratorTracker( SourceLineInfo const& lineInfo ) -> IGeneratorTracker&;
+    auto acquireGeneratorTracker( StringRef generatorName, SourceLineInfo const& lineInfo ) -> IGeneratorTracker&;
 
     template<typename L>
     // Note: The type after -> is weird, because VS2015 cannot parse
     //       the expression used in the typedef inside, when it is in
     //       return type. Yeah.
-    auto generate( SourceLineInfo const& lineInfo, L const& generatorExpression ) -> decltype(std::declval<decltype(generatorExpression())>().get()) {
+    auto generate( StringRef generatorName, SourceLineInfo const& lineInfo, L const& generatorExpression ) -> decltype(std::declval<decltype(generatorExpression())>().get()) {
         using UnderlyingType = typename decltype(generatorExpression())::type;
 
-        IGeneratorTracker& tracker = acquireGeneratorTracker( lineInfo );
+        IGeneratorTracker& tracker = acquireGeneratorTracker( generatorName, lineInfo );
         if (!tracker.hasGenerator()) {
             tracker.setGenerator(pf::make_unique<Generators<UnderlyingType>>(generatorExpression()));
         }
@@ -3686,7 +4098,17 @@ namespace Generators {
 } // namespace Catch
 
 #define GENERATE( ... ) \
-    Catch::Generators::generate( CATCH_INTERNAL_LINEINFO, []{ using namespace Catch::Generators; return makeGenerators( __VA_ARGS__ ); } )
+    Catch::Generators::generate( INTERNAL_CATCH_STRINGIZE(INTERNAL_CATCH_UNIQUE_NAME(generator)), \
+                                 CATCH_INTERNAL_LINEINFO, \
+                                 [ ]{ using namespace Catch::Generators; return makeGenerators( __VA_ARGS__ ); } ) //NOLINT(google-build-using-namespace)
+#define GENERATE_COPY( ... ) \
+    Catch::Generators::generate( INTERNAL_CATCH_STRINGIZE(INTERNAL_CATCH_UNIQUE_NAME(generator)), \
+                                 CATCH_INTERNAL_LINEINFO, \
+                                 [=]{ using namespace Catch::Generators; return makeGenerators( __VA_ARGS__ ); } ) //NOLINT(google-build-using-namespace)
+#define GENERATE_REF( ... ) \
+    Catch::Generators::generate( INTERNAL_CATCH_STRINGIZE(INTERNAL_CATCH_UNIQUE_NAME(generator)), \
+                                 CATCH_INTERNAL_LINEINFO, \
+                                 [&]{ using namespace Catch::Generators; return makeGenerators( __VA_ARGS__ ); } ) //NOLINT(google-build-using-namespace)
 
 // end catch_generators.hpp
 // start catch_generators_generic.hpp
@@ -3743,7 +4165,7 @@ namespace Generators {
             if (!m_predicate(m_generator.get())) {
                 // It might happen that there are no values that pass the
                 // filter. In that case we throw an exception.
-                auto has_initial_value = next();
+                auto has_initial_value = nextImpl();
                 if (!has_initial_value) {
                     Catch::throw_exception(GeneratorException("No valid value found in filtered generator"));
                 }
@@ -3755,6 +4177,11 @@ namespace Generators {
         }
 
         bool next() override {
+            return nextImpl();
+        }
+
+    private:
+        bool nextImpl() {
             bool success = m_generator.next();
             if (!success) {
                 return false;
@@ -3771,6 +4198,9 @@ namespace Generators {
 
     template <typename T>
     class RepeatGenerator : public IGenerator<T> {
+        static_assert(!std::is_same<T, bool>::value,
+            "RepeatGenerator currently does not support bools"
+            "because of std::vector<bool> specialization");
         GeneratorWrapper<T> m_generator;
         mutable std::vector<T> m_returned;
         size_t m_target_repeats;
@@ -3849,16 +4279,17 @@ namespace Generators {
         }
     };
 
-    template <typename T, typename U, typename Func>
+    template <typename Func, typename U, typename T = FunctionReturnType<Func, U>>
     GeneratorWrapper<T> map(Func&& function, GeneratorWrapper<U>&& generator) {
         return GeneratorWrapper<T>(
             pf::make_unique<MapGenerator<T, U, Func>>(std::forward<Func>(function), std::move(generator))
         );
     }
-    template <typename T, typename Func>
-    GeneratorWrapper<T> map(Func&& function, GeneratorWrapper<T>&& generator) {
+
+    template <typename T, typename U, typename Func>
+    GeneratorWrapper<T> map(Func&& function, GeneratorWrapper<U>&& generator) {
         return GeneratorWrapper<T>(
-            pf::make_unique<MapGenerator<T, T, Func>>(std::forward<Func>(function), std::move(generator))
+            pf::make_unique<MapGenerator<T, U, Func>>(std::forward<Func>(function), std::move(generator))
         );
     }
 
@@ -3873,12 +4304,14 @@ namespace Generators {
             m_chunk_size(size), m_generator(std::move(generator))
         {
             m_chunk.reserve(m_chunk_size);
-            m_chunk.push_back(m_generator.get());
-            for (size_t i = 1; i < m_chunk_size; ++i) {
-                if (!m_generator.next()) {
-                    Catch::throw_exception(GeneratorException("Not enough values to initialize the first chunk"));
-                }
+            if (m_chunk_size != 0) {
                 m_chunk.push_back(m_generator.get());
+                for (size_t i = 1; i < m_chunk_size; ++i) {
+                    if (!m_generator.next()) {
+                        Catch::throw_exception(GeneratorException("Not enough values to initialize the first chunk"));
+                    }
+                    m_chunk.push_back(m_generator.get());
+                }
             }
         }
         std::vector<T> const& get() const override {
@@ -3949,6 +4382,7 @@ namespace Catch {
     {
         if( !IMutableContext::currentContext )
             IMutableContext::createContext();
+        // NOLINTNEXTLINE(clang-analyzer-core.uninitialized.UndefReturn)
         return *IMutableContext::currentContext;
     }
 
@@ -3958,11 +4392,80 @@ namespace Catch {
     }
 
     void cleanUpContext();
+
+    class SimplePcg32;
+    SimplePcg32& rng();
 }
 
 // end catch_context.h
 // start catch_interfaces_config.h
 
+// start catch_option.hpp
+
+namespace Catch {
+
+    // An optional type
+    template<typename T>
+    class Option {
+    public:
+        Option() : nullableValue( nullptr ) {}
+        Option( T const& _value )
+        : nullableValue( new( storage ) T( _value ) )
+        {}
+        Option( Option const& _other )
+        : nullableValue( _other ? new( storage ) T( *_other ) : nullptr )
+        {}
+
+        ~Option() {
+            reset();
+        }
+
+        Option& operator= ( Option const& _other ) {
+            if( &_other != this ) {
+                reset();
+                if( _other )
+                    nullableValue = new( storage ) T( *_other );
+            }
+            return *this;
+        }
+        Option& operator = ( T const& _value ) {
+            reset();
+            nullableValue = new( storage ) T( _value );
+            return *this;
+        }
+
+        void reset() {
+            if( nullableValue )
+                nullableValue->~T();
+            nullableValue = nullptr;
+        }
+
+        T& operator*() { return *nullableValue; }
+        T const& operator*() const { return *nullableValue; }
+        T* operator->() { return nullableValue; }
+        const T* operator->() const { return nullableValue; }
+
+        T valueOr( T const& defaultValue ) const {
+            return nullableValue ? *nullableValue : defaultValue;
+        }
+
+        bool some() const { return nullableValue != nullptr; }
+        bool none() const { return nullableValue == nullptr; }
+
+        bool operator !() const { return nullableValue == nullptr; }
+        explicit operator bool() const {
+            return some();
+        }
+
+    private:
+        T *nullableValue;
+        alignas(alignof(T)) char storage[sizeof(T)];
+    };
+
+} // end namespace Catch
+
+// end catch_option.hpp
+#include <chrono>
 #include <iosfwd>
 #include <string>
 #include <vector>
@@ -4020,20 +4523,77 @@ namespace Catch {
         virtual int abortAfter() const = 0;
         virtual bool showInvisibles() const = 0;
         virtual ShowDurations::OrNot showDurations() const = 0;
+        virtual double minDuration() const = 0;
         virtual TestSpec const& testSpec() const = 0;
         virtual bool hasTestFilters() const = 0;
+        virtual std::vector<std::string> const& getTestsOrTags() const = 0;
         virtual RunTests::InWhatOrder runOrder() const = 0;
         virtual unsigned int rngSeed() const = 0;
-        virtual int benchmarkResolutionMultiple() const = 0;
         virtual UseColour::YesOrNo useColour() const = 0;
         virtual std::vector<std::string> const& getSectionsToRun() const = 0;
         virtual Verbosity verbosity() const = 0;
+
+        virtual bool benchmarkNoAnalysis() const = 0;
+        virtual int benchmarkSamples() const = 0;
+        virtual double benchmarkConfidenceInterval() const = 0;
+        virtual unsigned int benchmarkResamples() const = 0;
+        virtual std::chrono::milliseconds benchmarkWarmupTime() const = 0;
     };
 
     using IConfigPtr = std::shared_ptr<IConfig const>;
 }
 
 // end catch_interfaces_config.h
+// start catch_random_number_generator.h
+
+#include <cstdint>
+
+namespace Catch {
+
+    // This is a simple implementation of C++11 Uniform Random Number
+    // Generator. It does not provide all operators, because Catch2
+    // does not use it, but it should behave as expected inside stdlib's
+    // distributions.
+    // The implementation is based on the PCG family (http://pcg-random.org)
+    class SimplePcg32 {
+        using state_type = std::uint64_t;
+    public:
+        using result_type = std::uint32_t;
+        static constexpr result_type (min)() {
+            return 0;
+        }
+        static constexpr result_type (max)() {
+            return static_cast<result_type>(-1);
+        }
+
+        // Provide some default initial state for the default constructor
+        SimplePcg32():SimplePcg32(0xed743cc4U) {}
+
+        explicit SimplePcg32(result_type seed_);
+
+        void seed(result_type seed_);
+        void discard(uint64_t skip);
+
+        result_type operator()();
+
+    private:
+        friend bool operator==(SimplePcg32 const& lhs, SimplePcg32 const& rhs);
+        friend bool operator!=(SimplePcg32 const& lhs, SimplePcg32 const& rhs);
+
+        // In theory we also need operator<< and operator>>
+        // In practice we do not use them, so we will skip them for now
+
+        std::uint64_t m_state;
+        // This part of the state determines which "stream" of the numbers
+        // is chosen -- we take it as a constant for Catch2, so we only
+        // need to deal with seeding the main state.
+        // Picked by reading 8 bytes from `/dev/random` :-)
+        static const std::uint64_t s_inc = (0x13ed0cc53f939476ULL << 1ULL) | 1ULL;
+    };
+
+} // end namespace Catch
+
+// end catch_random_number_generator.h
 #include <random>
 
 namespace Catch {
@@ -4041,14 +4601,13 @@ namespace Generators {
 
 template <typename Float>
 class RandomFloatingGenerator final : public IGenerator<Float> {
-    // FIXME: What is the right seed?
-    std::minstd_rand m_rand;
+    Catch::SimplePcg32& m_rng;
     std::uniform_real_distribution<Float> m_dist;
     Float m_current_number;
 public:
 
     RandomFloatingGenerator(Float a, Float b):
-        m_rand(getCurrentContext().getConfig()->rngSeed()),
+        m_rng(rng()),
         m_dist(a, b) {
         static_cast<void>(next());
     }
@@ -4057,20 +4616,20 @@ public:
         return m_current_number;
     }
     bool next() override {
-        m_current_number = m_dist(m_rand);
+        m_current_number = m_dist(m_rng);
         return true;
     }
 };
 
 template <typename Integer>
 class RandomIntegerGenerator final : public IGenerator<Integer> {
-    std::minstd_rand m_rand;
+    Catch::SimplePcg32& m_rng;
     std::uniform_int_distribution<Integer> m_dist;
     Integer m_current_number;
 public:
 
     RandomIntegerGenerator(Integer a, Integer b):
-        m_rand(getCurrentContext().getConfig()->rngSeed()),
+        m_rng(rng()),
         m_dist(a, b) {
         static_cast<void>(next());
     }
@@ -4079,7 +4638,7 @@ public:
         return m_current_number;
     }
     bool next() override {
-        m_current_number = m_dist(m_rand);
+        m_current_number = m_dist(m_rng);
         return true;
     }
 };
@@ -4139,7 +4698,7 @@ public:
 
 template <typename T>
 GeneratorWrapper<T> range(T const& start, T const& end, T const& step) {
-    static_assert(std::is_integral<T>::value && !std::is_same<T, bool>::value, "Type must be an integer");
+    static_assert(std::is_arithmetic<T>::value && !std::is_same<T, bool>::value, "Type must be numeric");
     return GeneratorWrapper<T>(pf::make_unique<RangeGenerator<T>>(start, end, step));
 }
 
@@ -4149,6 +4708,45 @@ GeneratorWrapper<T> range(T const& start, T const& end) {
     return GeneratorWrapper<T>(pf::make_unique<RangeGenerator<T>>(start, end));
 }
 
+template <typename T>
+class IteratorGenerator final : public IGenerator<T> {
+    static_assert(!std::is_same<T, bool>::value,
+        "IteratorGenerator currently does not support bools"
+        "because of std::vector<bool> specialization");
+
+    std::vector<T> m_elems;
+    size_t m_current = 0;
+public:
+    template <typename InputIterator, typename InputSentinel>
+    IteratorGenerator(InputIterator first, InputSentinel last):m_elems(first, last) {
+        if (m_elems.empty()) {
+            Catch::throw_exception(GeneratorException("IteratorGenerator received no valid values"));
+        }
+    }
+
+    T const& get() const override {
+        return m_elems[m_current];
+    }
+
+    bool next() override {
+        ++m_current;
+        return m_current != m_elems.size();
+    }
+};
+
+template <typename InputIterator,
+          typename InputSentinel,
+          typename ResultType = typename std::iterator_traits<InputIterator>::value_type>
+GeneratorWrapper<ResultType> from_range(InputIterator from, InputSentinel to) {
+    return GeneratorWrapper<ResultType>(pf::make_unique<IteratorGenerator<ResultType>>(from, to));
+}
+
+template <typename Container,
+          typename ResultType = typename Container::value_type>
+GeneratorWrapper<ResultType> from_range(Container const& cnt) {
+    return GeneratorWrapper<ResultType>(pf::make_unique<IteratorGenerator<ResultType>>(cnt.begin(), cnt.end()));
+}
+
 } // namespace Generators
 } // namespace Catch
 
@@ -4352,7 +4950,7 @@ namespace Catch {
                     arcSafeRelease( m_substr );
                 }
 
-                bool match( NSString* arg ) const override {
+                bool match( NSString* str ) const override {
                     return false;
                 }
 
@@ -4375,7 +4973,7 @@ namespace Catch {
             struct Contains : StringHolder {
                 Contains( NSString* substr ) : StringHolder( substr ){}
 
-                bool match( NSString* str ) const {
+                bool match( NSString* str ) const override {
                     return  (str != nil || m_substr == nil ) &&
                             [str rangeOfString:m_substr].location != NSNotFound;
                 }
@@ -4451,7 +5049,8 @@ return @ desc; \
 // end catch_objc.hpp
 #endif
 
-#ifdef CATCH_CONFIG_EXTERNAL_INTERFACES
+// Benchmarking needs the externally-facing parts of reporters to work
+#if defined(CATCH_CONFIG_EXTERNAL_INTERFACES) || defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
 // start catch_external_interfaces.h
 
 // start catch_reporter_bases.hpp
@@ -4493,7 +5092,7 @@ namespace Catch
         virtual bool matches( std::string const& str ) const;
 
     private:
-        std::string adjustCase( std::string const& str ) const;
+        std::string normaliseString( std::string const& str ) const;
         CaseSensitive::Choice m_caseSensitivity;
         WildcardPosition m_wildcard = NoWildcard;
         std::string m_pattern;
@@ -4507,36 +5106,40 @@ namespace Catch
 
 namespace Catch {
 
+    struct IConfig;
+
     class TestSpec {
-        struct Pattern {
+        class Pattern {
+        public:
+            explicit Pattern( std::string const& name );
             virtual ~Pattern();
             virtual bool matches( TestCaseInfo const& testCase ) const = 0;
+            std::string const& name() const;
+        private:
+            std::string const m_name;
         };
         using PatternPtr = std::shared_ptr<Pattern>;
 
         class NamePattern : public Pattern {
         public:
-            NamePattern( std::string const& name );
-            virtual ~NamePattern();
-            virtual bool matches( TestCaseInfo const& testCase ) const override;
+            explicit NamePattern( std::string const& name, std::string const& filterString );
+            bool matches( TestCaseInfo const& testCase ) const override;
         private:
             WildcardPattern m_wildcardPattern;
         };
 
         class TagPattern : public Pattern {
         public:
-            TagPattern( std::string const& tag );
-            virtual ~TagPattern();
-            virtual bool matches( TestCaseInfo const& testCase ) const override;
+            explicit TagPattern( std::string const& tag, std::string const& filterString );
+            bool matches( TestCaseInfo const& testCase ) const override;
         private:
             std::string m_tag;
         };
 
         class ExcludedPattern : public Pattern {
         public:
-            ExcludedPattern( PatternPtr const& underlyingPattern );
-            virtual ~ExcludedPattern();
-            virtual bool matches( TestCaseInfo const& testCase ) const override;
+            explicit ExcludedPattern( PatternPtr const& underlyingPattern );
+            bool matches( TestCaseInfo const& testCase ) const override;
         private:
             PatternPtr m_underlyingPattern;
         };
@@ -4545,15 +5148,25 @@ namespace Catch {
             std::vector<PatternPtr> m_patterns;
 
             bool matches( TestCaseInfo const& testCase ) const;
+            std::string name() const;
         };
 
     public:
+        struct FilterMatch {
+            std::string name;
+            std::vector<TestCase const*> tests;
+        };
+        using Matches = std::vector<FilterMatch>;
+        using vectorStrings = std::vector<std::string>;
+
         bool hasFilters() const;
         bool matches( TestCaseInfo const& testCase ) const;
+        Matches matchesByFilter( std::vector<TestCase> const& testCases, IConfig const& config ) const;
+        const vectorStrings & getInvalidArgs() const;
 
     private:
         std::vector<Filter> m_filters;
-
+        std::vector<std::string> m_invalidArgs;
         friend class TestSpecParser;
     };
 }
@@ -4588,9 +5201,13 @@ namespace Catch {
     class TestSpecParser {
         enum Mode{ None, Name, QuotedName, Tag, EscapedName };
         Mode m_mode = None;
+        Mode lastMode = None;
         bool m_exclusion = false;
-        std::size_t m_start = std::string::npos, m_pos = 0;
+        std::size_t m_pos = 0;
+        std::size_t m_realPatternPos = 0;
         std::string m_arg;
+        std::string m_substring;
+        std::string m_patternName;
         std::vector<std::size_t> m_escapeChars;
         TestSpec::Filter m_currentFilter;
         TestSpec m_testSpec;
@@ -4603,32 +5220,32 @@ namespace Catch {
         TestSpec testSpec();
 
     private:
-        void visitChar( char c );
-        void startNewMode( Mode mode, std::size_t start );
+        bool visitChar( char c );
+        void startNewMode( Mode mode );
+        bool processNoneChar( char c );
+        void processNameChar( char c );
+        bool processOtherChar( char c );
+        void endMode();
         void escape();
-        std::string subString() const;
+        bool isControlChar( char c ) const;
+        void saveLastMode();
+        void revertBackToLastMode();
+        void addFilter();
+        bool separate();
 
-        template<typename T>
-        void addPattern() {
-            std::string token = subString();
-            for( std::size_t i = 0; i < m_escapeChars.size(); ++i )
-                token = token.substr( 0, m_escapeChars[i]-m_start-i ) + token.substr( m_escapeChars[i]-m_start-i+1 );
-            m_escapeChars.clear();
-            if( startsWith( token, "exclude:" ) ) {
-                m_exclusion = true;
-                token = token.substr( 8 );
-            }
-            if( !token.empty() ) {
-                TestSpec::PatternPtr pattern = std::make_shared<T>( token );
-                if( m_exclusion )
-                    pattern = std::make_shared<TestSpec::ExcludedPattern>( pattern );
-                m_currentFilter.m_patterns.push_back( pattern );
-            }
-            m_exclusion = false;
-            m_mode = None;
+        // Handles common preprocessing of the pattern for name/tag patterns
+        std::string preprocessPattern();
+        // Adds the current pattern as a test name
+        void addNamePattern();
+        // Adds the current pattern as a tag
+        void addTagPattern();
+
+        inline void addCharToPattern(char c) {
+            m_substring += c;
+            m_patternName += c;
+            m_realPatternPos++;
         }
 
-        void addFilter();
     };
     TestSpec parseTestSpec( std::string const& arg );
 
@@ -4669,11 +5286,17 @@ namespace Catch {
 
         int abortAfter = -1;
         unsigned int rngSeed = 0;
-        int benchmarkResolutionMultiple = 100;
+
+        bool benchmarkNoAnalysis = false;
+        unsigned int benchmarkSamples = 100;
+        double benchmarkConfidenceInterval = 0.95;
+        unsigned int benchmarkResamples = 100000;
+        std::chrono::milliseconds::rep benchmarkWarmupTime = 100;
 
         Verbosity verbosity = Verbosity::Normal;
         WarnAbout::What warnings = WarnAbout::Nothing;
         ShowDurations::OrNot showDurations = ShowDurations::DefaultForReporter;
+        double minDuration = -1;
         RunTests::InWhatOrder runOrder = RunTests::InDeclarationOrder;
         UseColour::YesOrNo useColour = UseColour::Auto;
         WaitForKeypress::When waitForKeypress = WaitForKeypress::Never;
@@ -4708,10 +5331,10 @@ namespace Catch {
         std::string getProcessName() const;
         std::string const& getReporterName() const;
 
-        std::vector<std::string> const& getTestsOrTags() const;
+        std::vector<std::string> const& getTestsOrTags() const override;
         std::vector<std::string> const& getSectionsToRun() const override;
 
-        virtual TestSpec const& testSpec() const override;
+        TestSpec const& testSpec() const override;
         bool hasTestFilters() const override;
 
         bool showHelp() const;
@@ -4724,14 +5347,19 @@ namespace Catch {
         bool warnAboutMissingAssertions() const override;
         bool warnAboutNoTests() const override;
         ShowDurations::OrNot showDurations() const override;
+        double minDuration() const override;
         RunTests::InWhatOrder runOrder() const override;
         unsigned int rngSeed() const override;
-        int benchmarkResolutionMultiple() const override;
         UseColour::YesOrNo useColour() const override;
         bool shouldDebugBreak() const override;
         int abortAfter() const override;
         bool showInvisibles() const override;
         Verbosity verbosity() const override;
+        bool benchmarkNoAnalysis() const override;
+        int benchmarkSamples() const override;
+        double benchmarkConfidenceInterval() const override;
+        unsigned int benchmarkResamples() const override;
+        std::chrono::milliseconds benchmarkWarmupTime() const override;
 
     private:
 
@@ -4792,76 +5420,61 @@ namespace Catch {
 } // end namespace Catch
 
 // end catch_assertionresult.h
-// start catch_option.hpp
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+// start catch_estimate.hpp
 
-namespace Catch {
+ // Statistics estimates
 
-    // An optional type
-    template<typename T>
-    class Option {
-    public:
-        Option() : nullableValue( nullptr ) {}
-        Option( T const& _value )
-        : nullableValue( new( storage ) T( _value ) )
-        {}
-        Option( Option const& _other )
-        : nullableValue( _other ? new( storage ) T( *_other ) : nullptr )
-        {}
 
-        ~Option() {
-            reset();
-        }
-
-        Option& operator= ( Option const& _other ) {
-            if( &_other != this ) {
-                reset();
-                if( _other )
-                    nullableValue = new( storage ) T( *_other );
+namespace Catch {
+    namespace Benchmark {
+        template <typename Duration>
+        struct Estimate {
+            Duration point;
+            Duration lower_bound;
+            Duration upper_bound;
+            double confidence_interval;
+
+            template <typename Duration2>
+            operator Estimate<Duration2>() const {
+                return { point, lower_bound, upper_bound, confidence_interval };
             }
-            return *this;
-        }
-        Option& operator = ( T const& _value ) {
-            reset();
-            nullableValue = new( storage ) T( _value );
-            return *this;
-        }
-
-        void reset() {
-            if( nullableValue )
-                nullableValue->~T();
-            nullableValue = nullptr;
-        }
-
-        T& operator*() { return *nullableValue; }
-        T const& operator*() const { return *nullableValue; }
-        T* operator->() { return nullableValue; }
-        const T* operator->() const { return nullableValue; }
+        };
+    } // namespace Benchmark
+} // namespace Catch
 
-        T valueOr( T const& defaultValue ) const {
-            return nullableValue ? *nullableValue : defaultValue;
-        }
+// end catch_estimate.hpp
+// start catch_outlier_classification.hpp
 
-        bool some() const { return nullableValue != nullptr; }
-        bool none() const { return nullableValue == nullptr; }
+// Outlier information
 
-        bool operator !() const { return nullableValue == nullptr; }
-        explicit operator bool() const {
-            return some();
-        }
+namespace Catch {
+    namespace Benchmark {
+        struct OutlierClassification {
+            int samples_seen = 0;
+            int low_severe = 0;     // more than 3 times IQR below Q1
+            int low_mild = 0;       // 1.5 to 3 times IQR below Q1
+            int high_mild = 0;      // 1.5 to 3 times IQR above Q3
+            int high_severe = 0;    // more than 3 times IQR above Q3
+
+            int total() const {
+                return low_severe + low_mild + high_mild + high_severe;
+            }
+        };
+    } // namespace Benchmark
+} // namespace Catch
 
-    private:
-        T *nullableValue;
-        alignas(alignof(T)) char storage[sizeof(T)];
-    };
+// end catch_outlier_classification.hpp
 
-} // end namespace Catch
+#include <iterator>
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
-// end catch_option.hpp
 #include <string>
 #include <iosfwd>
 #include <map>
 #include <set>
 #include <memory>
+#include <algorithm>
 
 namespace Catch {
 
@@ -4997,14 +5610,43 @@ namespace Catch {
         bool aborting;
     };
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
     struct BenchmarkInfo {
         std::string name;
+        double estimatedDuration;
+        int iterations;
+        int samples;
+        unsigned int resamples;
+        double clockResolution;
+        double clockCost;
     };
+
+    template <class Duration>
     struct BenchmarkStats {
         BenchmarkInfo info;
-        std::size_t iterations;
-        uint64_t elapsedTimeInNanoseconds;
+
+        std::vector<Duration> samples;
+        Benchmark::Estimate<Duration> mean;
+        Benchmark::Estimate<Duration> standardDeviation;
+        Benchmark::OutlierClassification outliers;
+        double outlierVariance;
+
+        template <typename Duration2>
+        operator BenchmarkStats<Duration2>() const {
+            std::vector<Duration2> samples2;
+            samples2.reserve(samples.size());
+            std::transform(samples.begin(), samples.end(), std::back_inserter(samples2), [](Duration d) { return Duration2(d); });
+            return {
+                info,
+                std::move(samples2),
+                mean,
+                standardDeviation,
+                outliers,
+                outlierVariance,
+            };
+        }
     };
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
     struct IStreamingReporter {
         virtual ~IStreamingReporter() = default;
@@ -5017,23 +5659,26 @@ namespace Catch {
 
         virtual void noMatchingTestCases( std::string const& spec ) = 0;
 
+        virtual void reportInvalidArguments(std::string const&) {}
+
         virtual void testRunStarting( TestRunInfo const& testRunInfo ) = 0;
         virtual void testGroupStarting( GroupInfo const& groupInfo ) = 0;
 
         virtual void testCaseStarting( TestCaseInfo const& testInfo ) = 0;
         virtual void sectionStarting( SectionInfo const& sectionInfo ) = 0;
 
-        // *** experimental ***
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+        virtual void benchmarkPreparing( std::string const& ) {}
         virtual void benchmarkStarting( BenchmarkInfo const& ) {}
+        virtual void benchmarkEnded( BenchmarkStats<> const& ) {}
+        virtual void benchmarkFailed( std::string const& ) {}
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
         virtual void assertionStarting( AssertionInfo const& assertionInfo ) = 0;
 
         // The return value indicates if the messages buffer should be cleared:
         virtual bool assertionEnded( AssertionStats const& assertionStats ) = 0;
 
-        // *** experimental ***
-        virtual void benchmarkEnded( BenchmarkStats const& ) {}
-
         virtual void sectionEnded( SectionStats const& sectionStats ) = 0;
         virtual void testCaseEnded( TestCaseStats const& testCaseStats ) = 0;
         virtual void testGroupEnded( TestGroupStats const& testGroupStats ) = 0;
@@ -5082,6 +5727,11 @@ namespace Catch {
     // Returns double formatted as %.3f (format expected on output)
     std::string getFormattedDuration( double duration );
 
+    //! Should the reporter show
+    bool shouldShowDuration( IConfig const& config, double duration );
+
+    std::string serializeFilters( std::vector<std::string> const& container );
+
     template<typename DerivedT>
     struct StreamingReporterBase : IStreamingReporter {
 
@@ -5106,9 +5756,12 @@ namespace Catch {
 
         void noMatchingTestCases(std::string const&) override {}
 
+        void reportInvalidArguments(std::string const&) override {}
+
         void testRunStarting(TestRunInfo const& _testRunInfo) override {
             currentTestRunInfo = _testRunInfo;
         }
+
         void testGroupStarting(GroupInfo const& _groupInfo) override {
             currentGroupInfo = _groupInfo;
         }
@@ -5399,11 +6052,11 @@ namespace Catch {
 
         class ReporterFactory : public IReporterFactory {
 
-            virtual IStreamingReporterPtr create( ReporterConfig const& config ) const override {
+            IStreamingReporterPtr create( ReporterConfig const& config ) const override {
                 return std::unique_ptr<T>( new T( config ) );
             }
 
-            virtual std::string getDescription() const override {
+            std::string getDescription() const override {
                 return T::getDescription();
             }
         };
@@ -5420,10 +6073,10 @@ namespace Catch {
 
         class ListenerFactory : public IReporterFactory {
 
-            virtual IStreamingReporterPtr create( ReporterConfig const& config ) const override {
+            IStreamingReporterPtr create( ReporterConfig const& config ) const override {
                 return std::unique_ptr<T>( new T( config ) );
             }
-            virtual std::string getDescription() const override {
+            std::string getDescription() const override {
                 return std::string();
             }
         };
@@ -5439,14 +6092,16 @@ namespace Catch {
 #if !defined(CATCH_CONFIG_DISABLE)
 
 #define CATCH_REGISTER_REPORTER( name, reporterType ) \
+    CATCH_INTERNAL_START_WARNINGS_SUPPRESSION         \
     CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS          \
     namespace{ Catch::ReporterRegistrar<reporterType> catch_internal_RegistrarFor##reporterType( name ); } \
-    CATCH_INTERNAL_UNSUPPRESS_GLOBALS_WARNINGS
+    CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
 
 #define CATCH_REGISTER_LISTENER( listenerType ) \
-     CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS   \
-     namespace{ Catch::ListenerRegistrar<listenerType> catch_internal_RegistrarFor##listenerType; } \
-     CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS
+    CATCH_INTERNAL_START_WARNINGS_SUPPRESSION   \
+    CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS    \
+    namespace{ Catch::ListenerRegistrar<listenerType> catch_internal_RegistrarFor##listenerType; } \
+    CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
 #else // CATCH_CONFIG_DISABLE
 
 #define CATCH_REGISTER_REPORTER(name, reporterType)
@@ -5468,8 +6123,6 @@ namespace Catch {
 
         static std::string getDescription();
 
-        ReporterPreferences getPreferences() const override;
-
         void noMatchingTestCases(std::string const& spec) override;
 
         void assertionStarting(AssertionInfo const&) override;
@@ -5508,6 +6161,8 @@ namespace Catch {
 
         void noMatchingTestCases(std::string const& spec) override;
 
+        void reportInvalidArguments(std::string const&arg) override;
+
         void assertionStarting(AssertionInfo const&) override;
 
         bool assertionEnded(AssertionStats const& _assertionStats) override;
@@ -5515,13 +6170,17 @@ namespace Catch {
         void sectionStarting(SectionInfo const& _sectionInfo) override;
         void sectionEnded(SectionStats const& _sectionStats) override;
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+        void benchmarkPreparing(std::string const& name) override;
         void benchmarkStarting(BenchmarkInfo const& info) override;
-        void benchmarkEnded(BenchmarkStats const& stats) override;
+        void benchmarkEnded(BenchmarkStats<> const& stats) override;
+        void benchmarkFailed(std::string const& error) override;
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
         void testCaseEnded(TestCaseStats const& _testCaseStats) override;
         void testGroupEnded(TestGroupStats const& _testGroupStats) override;
         void testRunEnded(TestRunStats const& _testRunStats) override;
-
+        void testRunStarting(TestRunInfo const& _testRunInfo) override;
     private:
 
         void lazyPrint();
@@ -5543,6 +6202,7 @@ namespace Catch {
 
         void printTotalsDivider(Totals const& totals);
         void printSummaryDivider();
+        void printTestFilters();
 
     private:
         bool m_headerPrinted = false;
@@ -5562,6 +6222,14 @@ namespace Catch {
 #include <vector>
 
 namespace Catch {
+    enum class XmlFormatting {
+        None = 0x00,
+        Indent = 0x01,
+        Newline = 0x02,
+    };
+
+    XmlFormatting operator | (XmlFormatting lhs, XmlFormatting rhs);
+    XmlFormatting operator & (XmlFormatting lhs, XmlFormatting rhs);
 
     class XmlEncode {
     public:
@@ -5583,14 +6251,14 @@ namespace Catch {
 
         class ScopedElement {
         public:
-            ScopedElement( XmlWriter* writer );
+            ScopedElement( XmlWriter* writer, XmlFormatting fmt );
 
             ScopedElement( ScopedElement&& other ) noexcept;
             ScopedElement& operator=( ScopedElement&& other ) noexcept;
 
             ~ScopedElement();
 
-            ScopedElement& writeText( std::string const& text, bool indent = true );
+            ScopedElement& writeText( std::string const& text, XmlFormatting fmt = XmlFormatting::Newline | XmlFormatting::Indent );
 
             template<typename T>
             ScopedElement& writeAttribute( std::string const& name, T const& attribute ) {
@@ -5600,6 +6268,7 @@ namespace Catch {
 
         private:
             mutable XmlWriter* m_writer = nullptr;
+            XmlFormatting m_fmt;
         };
 
         XmlWriter( std::ostream& os = Catch::cout() );
@@ -5608,11 +6277,11 @@ namespace Catch {
         XmlWriter( XmlWriter const& ) = delete;
         XmlWriter& operator=( XmlWriter const& ) = delete;
 
-        XmlWriter& startElement( std::string const& name );
+        XmlWriter& startElement( std::string const& name, XmlFormatting fmt = XmlFormatting::Newline | XmlFormatting::Indent);
 
-        ScopedElement scopedElement( std::string const& name );
+        ScopedElement scopedElement( std::string const& name, XmlFormatting fmt = XmlFormatting::Newline | XmlFormatting::Indent);
 
-        XmlWriter& endElement();
+        XmlWriter& endElement(XmlFormatting fmt = XmlFormatting::Newline | XmlFormatting::Indent);
 
         XmlWriter& writeAttribute( std::string const& name, std::string const& attribute );
 
@@ -5625,9 +6294,9 @@ namespace Catch {
             return writeAttribute( name, rss.str() );
         }
 
-        XmlWriter& writeText( std::string const& text, bool indent = true );
+        XmlWriter& writeText( std::string const& text, XmlFormatting fmt = XmlFormatting::Newline | XmlFormatting::Indent);
 
-        XmlWriter& writeComment( std::string const& text );
+        XmlWriter& writeComment(std::string const& text, XmlFormatting fmt = XmlFormatting::Newline | XmlFormatting::Indent);
 
         void writeStylesheetRef( std::string const& url );
 
@@ -5637,6 +6306,8 @@ namespace Catch {
 
     private:
 
+        void applyFormatting(XmlFormatting fmt);
+
         void writeDeclaration();
 
         void newlineIfNecessary();
@@ -5680,9 +6351,10 @@ namespace Catch {
 
         void writeTestCase(TestCaseNode const& testCaseNode);
 
-        void writeSection(std::string const& className,
-                          std::string const& rootName,
-                          SectionNode const& sectionNode);
+        void writeSection( std::string const& className,
+                           std::string const& rootName,
+                           SectionNode const& sectionNode,
+                           bool testOkToFail );
 
         void writeAssertions(SectionNode const& sectionNode);
         void writeAssertion(AssertionStats const& stats);
@@ -5737,6 +6409,13 @@ namespace Catch {
 
         void testRunEnded(TestRunStats const& testRunStats) override;
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+        void benchmarkPreparing(std::string const& name) override;
+        void benchmarkStarting(BenchmarkInfo const&) override;
+        void benchmarkEnded(BenchmarkStats<> const&) override;
+        void benchmarkFailed(std::string const&) override;
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
+
     private:
         Timer m_testCaseTimer;
         XmlWriter m_xml;
@@ -5750,6 +6429,1032 @@ namespace Catch {
 // end catch_external_interfaces.h
 #endif
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+// start catch_benchmarking_all.hpp
+
+// A proxy header that includes all of the benchmarking headers to allow
+// concise include of the benchmarking features. You should prefer the
+// individual includes in standard use.
+
+// start catch_benchmark.hpp
+
+ // Benchmark
+
+// start catch_chronometer.hpp
+
+// User-facing chronometer
+
+
+// start catch_clock.hpp
+
+// Clocks
+
+
+#include <chrono>
+#include <ratio>
+
+namespace Catch {
+    namespace Benchmark {
+        template <typename Clock>
+        using ClockDuration = typename Clock::duration;
+        template <typename Clock>
+        using FloatDuration = std::chrono::duration<double, typename Clock::period>;
+
+        template <typename Clock>
+        using TimePoint = typename Clock::time_point;
+
+        using default_clock = std::chrono::steady_clock;
+
+        template <typename Clock>
+        struct now {
+            TimePoint<Clock> operator()() const {
+                return Clock::now();
+            }
+        };
+
+        using fp_seconds = std::chrono::duration<double, std::ratio<1>>;
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_clock.hpp
+// start catch_optimizer.hpp
+
+ // Hinting the optimizer
+
+
+#if defined(_MSC_VER)
+#   include <atomic> // atomic_thread_fence
+#endif
+
+namespace Catch {
+    namespace Benchmark {
+#if defined(__GNUC__) || defined(__clang__)
+        template <typename T>
+        inline void keep_memory(T* p) {
+            asm volatile("" : : "g"(p) : "memory");
+        }
+        inline void keep_memory() {
+            asm volatile("" : : : "memory");
+        }
+
+        namespace Detail {
+            inline void optimizer_barrier() { keep_memory(); }
+        } // namespace Detail
+#elif defined(_MSC_VER)
+
+#pragma optimize("", off)
+        template <typename T>
+        inline void keep_memory(T* p) {
+            // thanks @milleniumbug
+            *reinterpret_cast<char volatile*>(p) = *reinterpret_cast<char const volatile*>(p);
+        }
+        // TODO equivalent keep_memory()
+#pragma optimize("", on)
+
+        namespace Detail {
+            inline void optimizer_barrier() {
+                std::atomic_thread_fence(std::memory_order_seq_cst);
+            }
+        } // namespace Detail
+
+#endif
+
+        template <typename T>
+        inline void deoptimize_value(T&& x) {
+            keep_memory(&x);
+        }
+
+        template <typename Fn, typename... Args>
+        inline auto invoke_deoptimized(Fn&& fn, Args&&... args) -> typename std::enable_if<!std::is_same<void, decltype(fn(args...))>::value>::type {
+            deoptimize_value(std::forward<Fn>(fn) (std::forward<Args...>(args...)));
+        }
+
+        template <typename Fn, typename... Args>
+        inline auto invoke_deoptimized(Fn&& fn, Args&&... args) -> typename std::enable_if<std::is_same<void, decltype(fn(args...))>::value>::type {
+            std::forward<Fn>(fn) (std::forward<Args...>(args...));
+        }
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_optimizer.hpp
+// start catch_complete_invoke.hpp
+
+// Invoke with a special case for void
+
+
+#include <type_traits>
+#include <utility>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            template <typename T>
+            struct CompleteType { using type = T; };
+            template <>
+            struct CompleteType<void> { struct type {}; };
+
+            template <typename T>
+            using CompleteType_t = typename CompleteType<T>::type;
+
+            template <typename Result>
+            struct CompleteInvoker {
+                template <typename Fun, typename... Args>
+                static Result invoke(Fun&& fun, Args&&... args) {
+                    return std::forward<Fun>(fun)(std::forward<Args>(args)...);
+                }
+            };
+            template <>
+            struct CompleteInvoker<void> {
+                template <typename Fun, typename... Args>
+                static CompleteType_t<void> invoke(Fun&& fun, Args&&... args) {
+                    std::forward<Fun>(fun)(std::forward<Args>(args)...);
+                    return {};
+                }
+            };
+
+            // invoke and not return void :(
+            template <typename Fun, typename... Args>
+            CompleteType_t<FunctionReturnType<Fun, Args...>> complete_invoke(Fun&& fun, Args&&... args) {
+                return CompleteInvoker<FunctionReturnType<Fun, Args...>>::invoke(std::forward<Fun>(fun), std::forward<Args>(args)...);
+            }
+
+            const std::string benchmarkErrorMsg = "a benchmark failed to run successfully";
+        } // namespace Detail
+
+        template <typename Fun>
+        Detail::CompleteType_t<FunctionReturnType<Fun>> user_code(Fun&& fun) {
+            CATCH_TRY{
+                return Detail::complete_invoke(std::forward<Fun>(fun));
+            } CATCH_CATCH_ALL{
+                getResultCapture().benchmarkFailed(translateActiveException());
+                CATCH_RUNTIME_ERROR(Detail::benchmarkErrorMsg);
+            }
+        }
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_complete_invoke.hpp
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            struct ChronometerConcept {
+                virtual void start() = 0;
+                virtual void finish() = 0;
+                virtual ~ChronometerConcept() = default;
+            };
+            template <typename Clock>
+            struct ChronometerModel final : public ChronometerConcept {
+                void start() override { started = Clock::now(); }
+                void finish() override { finished = Clock::now(); }
+
+                ClockDuration<Clock> elapsed() const { return finished - started; }
+
+                TimePoint<Clock> started;
+                TimePoint<Clock> finished;
+            };
+        } // namespace Detail
+
+        struct Chronometer {
+        public:
+            template <typename Fun>
+            void measure(Fun&& fun) { measure(std::forward<Fun>(fun), is_callable<Fun(int)>()); }
+
+            int runs() const { return k; }
+
+            Chronometer(Detail::ChronometerConcept& meter, int k)
+                : impl(&meter)
+                , k(k) {}
+
+        private:
+            template <typename Fun>
+            void measure(Fun&& fun, std::false_type) {
+                measure([&fun](int) { return fun(); }, std::true_type());
+            }
+
+            template <typename Fun>
+            void measure(Fun&& fun, std::true_type) {
+                Detail::optimizer_barrier();
+                impl->start();
+                for (int i = 0; i < k; ++i) invoke_deoptimized(fun, i);
+                impl->finish();
+                Detail::optimizer_barrier();
+            }
+
+            Detail::ChronometerConcept* impl;
+            int k;
+        };
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_chronometer.hpp
+// start catch_environment.hpp
+
+// Environment information
+
+
+namespace Catch {
+    namespace Benchmark {
+        template <typename Duration>
+        struct EnvironmentEstimate {
+            Duration mean;
+            OutlierClassification outliers;
+
+            template <typename Duration2>
+            operator EnvironmentEstimate<Duration2>() const {
+                return { mean, outliers };
+            }
+        };
+        template <typename Clock>
+        struct Environment {
+            using clock_type = Clock;
+            EnvironmentEstimate<FloatDuration<Clock>> clock_resolution;
+            EnvironmentEstimate<FloatDuration<Clock>> clock_cost;
+        };
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_environment.hpp
+// start catch_execution_plan.hpp
+
+ // Execution plan
+
+
+// start catch_benchmark_function.hpp
+
+ // Dumb std::function implementation for consistent call overhead
+
+
+#include <cassert>
+#include <type_traits>
+#include <utility>
+#include <memory>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            template <typename T>
+            using Decay = typename std::decay<T>::type;
+            template <typename T, typename U>
+            struct is_related
+                : std::is_same<Decay<T>, Decay<U>> {};
+
+            /// We need to reinvent std::function because every piece of code that might add overhead
+            /// in a measurement context needs to have consistent performance characteristics so that we
+            /// can account for it in the measurement.
+            /// Implementations of std::function with optimizations that aren't always applicable, like
+            /// small buffer optimizations, are not uncommon.
+            /// This is effectively an implementation of std::function without any such optimizations;
+            /// it may be slow, but it is consistently slow.
+            struct BenchmarkFunction {
+            private:
+                struct callable {
+                    virtual void call(Chronometer meter) const = 0;
+                    virtual callable* clone() const = 0;
+                    virtual ~callable() = default;
+                };
+                template <typename Fun>
+                struct model : public callable {
+                    model(Fun&& fun) : fun(std::move(fun)) {}
+                    model(Fun const& fun) : fun(fun) {}
+
+                    model<Fun>* clone() const override { return new model<Fun>(*this); }
+
+                    void call(Chronometer meter) const override {
+                        call(meter, is_callable<Fun(Chronometer)>());
+                    }
+                    void call(Chronometer meter, std::true_type) const {
+                        fun(meter);
+                    }
+                    void call(Chronometer meter, std::false_type) const {
+                        meter.measure(fun);
+                    }
+
+                    Fun fun;
+                };
+
+                struct do_nothing { void operator()() const {} };
+
+                template <typename T>
+                BenchmarkFunction(model<T>* c) : f(c) {}
+
+            public:
+                BenchmarkFunction()
+                    : f(new model<do_nothing>{ {} }) {}
+
+                template <typename Fun,
+                    typename std::enable_if<!is_related<Fun, BenchmarkFunction>::value, int>::type = 0>
+                    BenchmarkFunction(Fun&& fun)
+                    : f(new model<typename std::decay<Fun>::type>(std::forward<Fun>(fun))) {}
+
+                BenchmarkFunction(BenchmarkFunction&& that)
+                    : f(std::move(that.f)) {}
+
+                BenchmarkFunction(BenchmarkFunction const& that)
+                    : f(that.f->clone()) {}
+
+                BenchmarkFunction& operator=(BenchmarkFunction&& that) {
+                    f = std::move(that.f);
+                    return *this;
+                }
+
+                BenchmarkFunction& operator=(BenchmarkFunction const& that) {
+                    f.reset(that.f->clone());
+                    return *this;
+                }
+
+                void operator()(Chronometer meter) const { f->call(meter); }
+
+            private:
+                std::unique_ptr<callable> f;
+            };
+        } // namespace Detail
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_benchmark_function.hpp
+// start catch_repeat.hpp
+
+// repeat algorithm
+
+
+#include <type_traits>
+#include <utility>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            template <typename Fun>
+            struct repeater {
+                void operator()(int k) const {
+                    for (int i = 0; i < k; ++i) {
+                        fun();
+                    }
+                }
+                Fun fun;
+            };
+            template <typename Fun>
+            repeater<typename std::decay<Fun>::type> repeat(Fun&& fun) {
+                return { std::forward<Fun>(fun) };
+            }
+        } // namespace Detail
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_repeat.hpp
+// start catch_run_for_at_least.hpp
+
+// Run a function for a minimum amount of time
+
+
+// start catch_measure.hpp
+
+// Measure
+
+
+// start catch_timing.hpp
+
+// Timing
+
+
+#include <tuple>
+#include <type_traits>
+
+namespace Catch {
+    namespace Benchmark {
+        template <typename Duration, typename Result>
+        struct Timing {
+            Duration elapsed;
+            Result result;
+            int iterations;
+        };
+        template <typename Clock, typename Func, typename... Args>
+        using TimingOf = Timing<ClockDuration<Clock>, Detail::CompleteType_t<FunctionReturnType<Func, Args...>>>;
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_timing.hpp
+#include <utility>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            template <typename Clock, typename Fun, typename... Args>
+            TimingOf<Clock, Fun, Args...> measure(Fun&& fun, Args&&... args) {
+                auto start = Clock::now();
+                auto&& r = Detail::complete_invoke(fun, std::forward<Args>(args)...);
+                auto end = Clock::now();
+                auto delta = end - start;
+                return { delta, std::forward<decltype(r)>(r), 1 };
+            }
+        } // namespace Detail
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_measure.hpp
+#include <utility>
+#include <type_traits>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            template <typename Clock, typename Fun>
+            TimingOf<Clock, Fun, int> measure_one(Fun&& fun, int iters, std::false_type) {
+                return Detail::measure<Clock>(fun, iters);
+            }
+            template <typename Clock, typename Fun>
+            TimingOf<Clock, Fun, Chronometer> measure_one(Fun&& fun, int iters, std::true_type) {
+                Detail::ChronometerModel<Clock> meter;
+                auto&& result = Detail::complete_invoke(fun, Chronometer(meter, iters));
+
+                return { meter.elapsed(), std::move(result), iters };
+            }
+
+            template <typename Clock, typename Fun>
+            using run_for_at_least_argument_t = typename std::conditional<is_callable<Fun(Chronometer)>::value, Chronometer, int>::type;
+
+            struct optimized_away_error : std::exception {
+                const char* what() const noexcept override {
+                    return "could not measure benchmark, maybe it was optimized away";
+                }
+            };
+
+            template <typename Clock, typename Fun>
+            TimingOf<Clock, Fun, run_for_at_least_argument_t<Clock, Fun>> run_for_at_least(ClockDuration<Clock> how_long, int seed, Fun&& fun) {
+                auto iters = seed;
+                while (iters < (1 << 30)) {
+                    auto&& Timing = measure_one<Clock>(fun, iters, is_callable<Fun(Chronometer)>());
+
+                    if (Timing.elapsed >= how_long) {
+                        return { Timing.elapsed, std::move(Timing.result), iters };
+                    }
+                    iters *= 2;
+                }
+                Catch::throw_exception(optimized_away_error{});
+            }
+        } // namespace Detail
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_run_for_at_least.hpp
+#include <algorithm>
+#include <iterator>
+
+namespace Catch {
+    namespace Benchmark {
+        template <typename Duration>
+        struct ExecutionPlan {
+            int iterations_per_sample;
+            Duration estimated_duration;
+            Detail::BenchmarkFunction benchmark;
+            Duration warmup_time;
+            int warmup_iterations;
+
+            template <typename Duration2>
+            operator ExecutionPlan<Duration2>() const {
+                return { iterations_per_sample, estimated_duration, benchmark, warmup_time, warmup_iterations };
+            }
+
+            template <typename Clock>
+            std::vector<FloatDuration<Clock>> run(const IConfig &cfg, Environment<FloatDuration<Clock>> env) const {
+                // warmup a bit
+                Detail::run_for_at_least<Clock>(std::chrono::duration_cast<ClockDuration<Clock>>(warmup_time), warmup_iterations, Detail::repeat(now<Clock>{}));
+
+                std::vector<FloatDuration<Clock>> times;
+                times.reserve(cfg.benchmarkSamples());
+                std::generate_n(std::back_inserter(times), cfg.benchmarkSamples(), [this, env] {
+                    Detail::ChronometerModel<Clock> model;
+                    this->benchmark(Chronometer(model, iterations_per_sample));
+                    auto sample_time = model.elapsed() - env.clock_cost.mean;
+                    if (sample_time < FloatDuration<Clock>::zero()) sample_time = FloatDuration<Clock>::zero();
+                    return sample_time / iterations_per_sample;
+                });
+                return times;
+            }
+        };
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_execution_plan.hpp
+// start catch_estimate_clock.hpp
+
+ // Environment measurement
+
+
+// start catch_stats.hpp
+
+// Statistical analysis tools
+
+
+#include <algorithm>
+#include <functional>
+#include <vector>
+#include <iterator>
+#include <numeric>
+#include <tuple>
+#include <cmath>
+#include <utility>
+#include <cstddef>
+#include <random>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            using sample = std::vector<double>;
+
+            double weighted_average_quantile(int k, int q, std::vector<double>::iterator first, std::vector<double>::iterator last);
+
+            template <typename Iterator>
+            OutlierClassification classify_outliers(Iterator first, Iterator last) {
+                std::vector<double> copy(first, last);
+
+                auto q1 = weighted_average_quantile(1, 4, copy.begin(), copy.end());
+                auto q3 = weighted_average_quantile(3, 4, copy.begin(), copy.end());
+                auto iqr = q3 - q1;
+                auto los = q1 - (iqr * 3.);
+                auto lom = q1 - (iqr * 1.5);
+                auto him = q3 + (iqr * 1.5);
+                auto his = q3 + (iqr * 3.);
+
+                OutlierClassification o;
+                for (; first != last; ++first) {
+                    auto&& t = *first;
+                    if (t < los) ++o.low_severe;
+                    else if (t < lom) ++o.low_mild;
+                    else if (t > his) ++o.high_severe;
+                    else if (t > him) ++o.high_mild;
+                    ++o.samples_seen;
+                }
+                return o;
+            }
+
+            template <typename Iterator>
+            double mean(Iterator first, Iterator last) {
+                auto count = last - first;
+                double sum = std::accumulate(first, last, 0.);
+                return sum / count;
+            }
+
+            template <typename URng, typename Iterator, typename Estimator>
+            sample resample(URng& rng, int resamples, Iterator first, Iterator last, Estimator& estimator) {
+                auto n = last - first;
+                std::uniform_int_distribution<decltype(n)> dist(0, n - 1);
+
+                sample out;
+                out.reserve(resamples);
+                std::generate_n(std::back_inserter(out), resamples, [n, first, &estimator, &dist, &rng] {
+                    std::vector<double> resampled;
+                    resampled.reserve(n);
+                    std::generate_n(std::back_inserter(resampled), n, [first, &dist, &rng] { return first[dist(rng)]; });
+                    return estimator(resampled.begin(), resampled.end());
+                });
+                std::sort(out.begin(), out.end());
+                return out;
+            }
+
+            template <typename Estimator, typename Iterator>
+            sample jackknife(Estimator&& estimator, Iterator first, Iterator last) {
+                auto n = last - first;
+                auto second = std::next(first);
+                sample results;
+                results.reserve(n);
+
+                for (auto it = first; it != last; ++it) {
+                    std::iter_swap(it, first);
+                    results.push_back(estimator(second, last));
+                }
+
+                return results;
+            }
+
+            inline double normal_cdf(double x) {
+                return std::erfc(-x / std::sqrt(2.0)) / 2.0;
+            }
+
+            double erfc_inv(double x);
+
+            double normal_quantile(double p);
+
+            template <typename Iterator, typename Estimator>
+            Estimate<double> bootstrap(double confidence_level, Iterator first, Iterator last, sample const& resample, Estimator&& estimator) {
+                auto n_samples = last - first;
+
+                double point = estimator(first, last);
+                // Degenerate case with a single sample
+                if (n_samples == 1) return { point, point, point, confidence_level };
+
+                sample jack = jackknife(estimator, first, last);
+                double jack_mean = mean(jack.begin(), jack.end());
+                double sum_squares, sum_cubes;
+                std::tie(sum_squares, sum_cubes) = std::accumulate(jack.begin(), jack.end(), std::make_pair(0., 0.), [jack_mean](std::pair<double, double> sqcb, double x) -> std::pair<double, double> {
+                    auto d = jack_mean - x;
+                    auto d2 = d * d;
+                    auto d3 = d2 * d;
+                    return { sqcb.first + d2, sqcb.second + d3 };
+                });
+
+                double accel = sum_cubes / (6 * std::pow(sum_squares, 1.5));
+                int n = static_cast<int>(resample.size());
+                double prob_n = std::count_if(resample.begin(), resample.end(), [point](double x) { return x < point; }) / (double)n;
+                // degenerate case with uniform samples
+                if (prob_n == 0) return { point, point, point, confidence_level };
+
+                double bias = normal_quantile(prob_n);
+                double z1 = normal_quantile((1. - confidence_level) / 2.);
+
+                auto cumn = [n](double x) -> int {
+                    return std::lround(normal_cdf(x) * n); };
+                auto a = [bias, accel](double b) { return bias + b / (1. - accel * b); };
+                double b1 = bias + z1;
+                double b2 = bias - z1;
+                double a1 = a(b1);
+                double a2 = a(b2);
+                auto lo = (std::max)(cumn(a1), 0);
+                auto hi = (std::min)(cumn(a2), n - 1);
+
+                return { point, resample[lo], resample[hi], confidence_level };
+            }
+
+            double outlier_variance(Estimate<double> mean, Estimate<double> stddev, int n);
+
+            struct bootstrap_analysis {
+                Estimate<double> mean;
+                Estimate<double> standard_deviation;
+                double outlier_variance;
+            };
+
+            bootstrap_analysis analyse_samples(double confidence_level, int n_resamples, std::vector<double>::iterator first, std::vector<double>::iterator last);
+        } // namespace Detail
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_stats.hpp
+#include <algorithm>
+#include <iterator>
+#include <tuple>
+#include <vector>
+#include <cmath>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            template <typename Clock>
+            std::vector<double> resolution(int k) {
+                std::vector<TimePoint<Clock>> times;
+                times.reserve(k + 1);
+                std::generate_n(std::back_inserter(times), k + 1, now<Clock>{});
+
+                std::vector<double> deltas;
+                deltas.reserve(k);
+                std::transform(std::next(times.begin()), times.end(), times.begin(),
+                    std::back_inserter(deltas),
+                    [](TimePoint<Clock> a, TimePoint<Clock> b) { return static_cast<double>((a - b).count()); });
+
+                return deltas;
+            }
+
+            const auto warmup_iterations = 10000;
+            const auto warmup_time = std::chrono::milliseconds(100);
+            const auto minimum_ticks = 1000;
+            const auto warmup_seed = 10000;
+            const auto clock_resolution_estimation_time = std::chrono::milliseconds(500);
+            const auto clock_cost_estimation_time_limit = std::chrono::seconds(1);
+            const auto clock_cost_estimation_tick_limit = 100000;
+            const auto clock_cost_estimation_time = std::chrono::milliseconds(10);
+            const auto clock_cost_estimation_iterations = 10000;
+
+            template <typename Clock>
+            int warmup() {
+                return run_for_at_least<Clock>(std::chrono::duration_cast<ClockDuration<Clock>>(warmup_time), warmup_seed, &resolution<Clock>)
+                    .iterations;
+            }
+            template <typename Clock>
+            EnvironmentEstimate<FloatDuration<Clock>> estimate_clock_resolution(int iterations) {
+                auto r = run_for_at_least<Clock>(std::chrono::duration_cast<ClockDuration<Clock>>(clock_resolution_estimation_time), iterations, &resolution<Clock>)
+                    .result;
+                return {
+                    FloatDuration<Clock>(mean(r.begin(), r.end())),
+                    classify_outliers(r.begin(), r.end()),
+                };
+            }
+            template <typename Clock>
+            EnvironmentEstimate<FloatDuration<Clock>> estimate_clock_cost(FloatDuration<Clock> resolution) {
+                auto time_limit = (std::min)(
+                    resolution * clock_cost_estimation_tick_limit,
+                    FloatDuration<Clock>(clock_cost_estimation_time_limit));
+                auto time_clock = [](int k) {
+                    return Detail::measure<Clock>([k] {
+                        for (int i = 0; i < k; ++i) {
+                            volatile auto ignored = Clock::now();
+                            (void)ignored;
+                        }
+                    }).elapsed;
+                };
+                time_clock(1);
+                int iters = clock_cost_estimation_iterations;
+                auto&& r = run_for_at_least<Clock>(std::chrono::duration_cast<ClockDuration<Clock>>(clock_cost_estimation_time), iters, time_clock);
+                std::vector<double> times;
+                int nsamples = static_cast<int>(std::ceil(time_limit / r.elapsed));
+                times.reserve(nsamples);
+                std::generate_n(std::back_inserter(times), nsamples, [time_clock, &r] {
+                    return static_cast<double>((time_clock(r.iterations) / r.iterations).count());
+                });
+                return {
+                    FloatDuration<Clock>(mean(times.begin(), times.end())),
+                    classify_outliers(times.begin(), times.end()),
+                };
+            }
+
+            template <typename Clock>
+            Environment<FloatDuration<Clock>> measure_environment() {
+                static Environment<FloatDuration<Clock>>* env = nullptr;
+                if (env) {
+                    return *env;
+                }
+
+                auto iters = Detail::warmup<Clock>();
+                auto resolution = Detail::estimate_clock_resolution<Clock>(iters);
+                auto cost = Detail::estimate_clock_cost<Clock>(resolution.mean);
+
+                env = new Environment<FloatDuration<Clock>>{ resolution, cost };
+                return *env;
+            }
+        } // namespace Detail
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_estimate_clock.hpp
+// start catch_analyse.hpp
+
+ // Run and analyse one benchmark
+
+
+// start catch_sample_analysis.hpp
+
+// Benchmark results
+
+
+#include <algorithm>
+#include <vector>
+#include <string>
+#include <iterator>
+
+namespace Catch {
+    namespace Benchmark {
+        template <typename Duration>
+        struct SampleAnalysis {
+            std::vector<Duration> samples;
+            Estimate<Duration> mean;
+            Estimate<Duration> standard_deviation;
+            OutlierClassification outliers;
+            double outlier_variance;
+
+            template <typename Duration2>
+            operator SampleAnalysis<Duration2>() const {
+                std::vector<Duration2> samples2;
+                samples2.reserve(samples.size());
+                std::transform(samples.begin(), samples.end(), std::back_inserter(samples2), [](Duration d) { return Duration2(d); });
+                return {
+                    std::move(samples2),
+                    mean,
+                    standard_deviation,
+                    outliers,
+                    outlier_variance,
+                };
+            }
+        };
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_sample_analysis.hpp
+#include <algorithm>
+#include <iterator>
+#include <vector>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            template <typename Duration, typename Iterator>
+            SampleAnalysis<Duration> analyse(const IConfig &cfg, Environment<Duration>, Iterator first, Iterator last) {
+                if (!cfg.benchmarkNoAnalysis()) {
+                    std::vector<double> samples;
+                    samples.reserve(last - first);
+                    std::transform(first, last, std::back_inserter(samples), [](Duration d) { return d.count(); });
+
+                    auto analysis = Catch::Benchmark::Detail::analyse_samples(cfg.benchmarkConfidenceInterval(), cfg.benchmarkResamples(), samples.begin(), samples.end());
+                    auto outliers = Catch::Benchmark::Detail::classify_outliers(samples.begin(), samples.end());
+
+                    auto wrap_estimate = [](Estimate<double> e) {
+                        return Estimate<Duration> {
+                            Duration(e.point),
+                                Duration(e.lower_bound),
+                                Duration(e.upper_bound),
+                                e.confidence_interval,
+                        };
+                    };
+                    std::vector<Duration> samples2;
+                    samples2.reserve(samples.size());
+                    std::transform(samples.begin(), samples.end(), std::back_inserter(samples2), [](double d) { return Duration(d); });
+                    return {
+                        std::move(samples2),
+                        wrap_estimate(analysis.mean),
+                        wrap_estimate(analysis.standard_deviation),
+                        outliers,
+                        analysis.outlier_variance,
+                    };
+                } else {
+                    std::vector<Duration> samples;
+                    samples.reserve(last - first);
+
+                    Duration mean = Duration(0);
+                    int i = 0;
+                    for (auto it = first; it < last; ++it, ++i) {
+                        samples.push_back(Duration(*it));
+                        mean += Duration(*it);
+                    }
+                    mean /= i;
+
+                    return {
+                        std::move(samples),
+                        Estimate<Duration>{mean, mean, mean, 0.0},
+                        Estimate<Duration>{Duration(0), Duration(0), Duration(0), 0.0},
+                        OutlierClassification{},
+                        0.0
+                    };
+                }
+            }
+        } // namespace Detail
+    } // namespace Benchmark
+} // namespace Catch
+
+// end catch_analyse.hpp
+#include <algorithm>
+#include <functional>
+#include <string>
+#include <vector>
+#include <cmath>
+
+namespace Catch {
+    namespace Benchmark {
+        struct Benchmark {
+            Benchmark(std::string &&name)
+                : name(std::move(name)) {}
+
+            template <class FUN>
+            Benchmark(std::string &&name, FUN &&func)
+                : fun(std::move(func)), name(std::move(name)) {}
+
+            template <typename Clock>
+            ExecutionPlan<FloatDuration<Clock>> prepare(const IConfig &cfg, Environment<FloatDuration<Clock>> env) const {
+                auto min_time = env.clock_resolution.mean * Detail::minimum_ticks;
+                auto run_time = std::max(min_time, std::chrono::duration_cast<decltype(min_time)>(cfg.benchmarkWarmupTime()));
+                auto&& test = Detail::run_for_at_least<Clock>(std::chrono::duration_cast<ClockDuration<Clock>>(run_time), 1, fun);
+                int new_iters = static_cast<int>(std::ceil(min_time * test.iterations / test.elapsed));
+                return { new_iters, test.elapsed / test.iterations * new_iters * cfg.benchmarkSamples(), fun, std::chrono::duration_cast<FloatDuration<Clock>>(cfg.benchmarkWarmupTime()), Detail::warmup_iterations };
+            }
+
+            template <typename Clock = default_clock>
+            void run() {
+                IConfigPtr cfg = getCurrentContext().getConfig();
+
+                auto env = Detail::measure_environment<Clock>();
+
+                getResultCapture().benchmarkPreparing(name);
+                CATCH_TRY{
+                    auto plan = user_code([&] {
+                        return prepare<Clock>(*cfg, env);
+                    });
+
+                    BenchmarkInfo info {
+                        name,
+                        plan.estimated_duration.count(),
+                        plan.iterations_per_sample,
+                        cfg->benchmarkSamples(),
+                        cfg->benchmarkResamples(),
+                        env.clock_resolution.mean.count(),
+                        env.clock_cost.mean.count()
+                    };
+
+                    getResultCapture().benchmarkStarting(info);
+
+                    auto samples = user_code([&] {
+                        return plan.template run<Clock>(*cfg, env);
+                    });
+
+                    auto analysis = Detail::analyse(*cfg, env, samples.begin(), samples.end());
+                    BenchmarkStats<FloatDuration<Clock>> stats{ info, analysis.samples, analysis.mean, analysis.standard_deviation, analysis.outliers, analysis.outlier_variance };
+                    getResultCapture().benchmarkEnded(stats);
+
+                } CATCH_CATCH_ALL{
+                    if (translateActiveException() != Detail::benchmarkErrorMsg) // benchmark errors have been reported, otherwise rethrow.
+                        std::rethrow_exception(std::current_exception());
+                }
+            }
+
+            // sets lambda to be used in fun *and* executes benchmark!
+            template <typename Fun,
+                typename std::enable_if<!Detail::is_related<Fun, Benchmark>::value, int>::type = 0>
+                Benchmark & operator=(Fun func) {
+                fun = Detail::BenchmarkFunction(func);
+                run();
+                return *this;
+            }
+
+            explicit operator bool() {
+                return true;
+            }
+
+        private:
+            Detail::BenchmarkFunction fun;
+            std::string name;
+        };
+    }
+} // namespace Catch
+
+#define INTERNAL_CATCH_GET_1_ARG(arg1, arg2, ...) arg1
+#define INTERNAL_CATCH_GET_2_ARG(arg1, arg2, ...) arg2
+
+#define INTERNAL_CATCH_BENCHMARK(BenchmarkName, name, benchmarkIndex)\
+    if( Catch::Benchmark::Benchmark BenchmarkName{name} ) \
+        BenchmarkName = [&](int benchmarkIndex)
+
+#define INTERNAL_CATCH_BENCHMARK_ADVANCED(BenchmarkName, name)\
+    if( Catch::Benchmark::Benchmark BenchmarkName{name} ) \
+        BenchmarkName = [&]
+
+// end catch_benchmark.hpp
+// start catch_constructor.hpp
+
+// Constructor and destructor helpers
+
+
+#include <type_traits>
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+            template <typename T, bool Destruct>
+            struct ObjectStorage
+            {
+                using TStorage = typename std::aligned_storage<sizeof(T), std::alignment_of<T>::value>::type;
+
+                ObjectStorage() : data() {}
+
+                ObjectStorage(const ObjectStorage& other)
+                {
+                    new(&data) T(other.stored_object());
+                }
+
+                ObjectStorage(ObjectStorage&& other)
+                {
+                    new(&data) T(std::move(other.stored_object()));
+                }
+
+                ~ObjectStorage() { destruct_on_exit<T>(); }
+
+                template <typename... Args>
+                void construct(Args&&... args)
+                {
+                    new (&data) T(std::forward<Args>(args)...);
+                }
+
+                template <bool AllowManualDestruction = !Destruct>
+                typename std::enable_if<AllowManualDestruction>::type destruct()
+                {
+                    stored_object().~T();
+                }
+
+            private:
+                // If this is a constructor benchmark, destruct the underlying object
+                template <typename U>
+                void destruct_on_exit(typename std::enable_if<Destruct, U>::type* = 0) { destruct<true>(); }
+                // Otherwise, don't
+                template <typename U>
+                void destruct_on_exit(typename std::enable_if<!Destruct, U>::type* = 0) { }
+
+                T& stored_object() {
+                    return *static_cast<T*>(static_cast<void*>(&data));
+                }
+
+                T const& stored_object() const {
+                    return *static_cast<T*>(static_cast<void*>(&data));
+                }
+
+                TStorage data;
+            };
+        }
+
+        template <typename T>
+        using storage_for = Detail::ObjectStorage<T, true>;
+
+        template <typename T>
+        using destructable_object = Detail::ObjectStorage<T, false>;
+    }
+}
+
+// end catch_constructor.hpp
+// end catch_benchmarking_all.hpp
+#endif
+
 #endif // ! CATCH_CONFIG_IMPL_ONLY
 
 #ifdef CATCH_IMPL
@@ -5775,23 +7480,37 @@ namespace TestCaseTracking {
         SourceLineInfo location;
 
         NameAndLocation( std::string const& _name, SourceLineInfo const& _location );
+        friend bool operator==(NameAndLocation const& lhs, NameAndLocation const& rhs) {
+            return lhs.name == rhs.name
+                && lhs.location == rhs.location;
+        }
     };
 
-    struct ITracker;
+    class ITracker;
 
     using ITrackerPtr = std::shared_ptr<ITracker>;
 
-    struct ITracker {
-        virtual ~ITracker();
+    class  ITracker {
+        NameAndLocation m_nameAndLocation;
+
+    public:
+        ITracker(NameAndLocation const& nameAndLoc) :
+            m_nameAndLocation(nameAndLoc)
+        {}
 
         // static queries
-        virtual NameAndLocation const& nameAndLocation() const = 0;
+        NameAndLocation const& nameAndLocation() const {
+            return m_nameAndLocation;
+        }
+
+        virtual ~ITracker();
 
         // dynamic queries
         virtual bool isComplete() const = 0; // Successfully completed or failed
         virtual bool isSuccessfullyCompleted() const = 0;
         virtual bool isOpen() const = 0; // Started but not complete
         virtual bool hasChildren() const = 0;
+        virtual bool hasStarted() const = 0;
 
         virtual ITracker& parent() = 0;
 
@@ -5823,8 +7542,6 @@ namespace TestCaseTracking {
 
     public:
 
-        static TrackerContext& instance();
-
         ITracker& startRun();
         void endRun();
 
@@ -5848,7 +7565,6 @@ namespace TestCaseTracking {
         };
 
         using Children = std::vector<ITrackerPtr>;
-        NameAndLocation m_nameAndLocation;
         TrackerContext& m_ctx;
         ITracker* m_parent;
         Children m_children;
@@ -5857,11 +7573,13 @@ namespace TestCaseTracking {
     public:
         TrackerBase( NameAndLocation const& nameAndLocation, TrackerContext& ctx, ITracker* parent );
 
-        NameAndLocation const& nameAndLocation() const override;
         bool isComplete() const override;
         bool isSuccessfullyCompleted() const override;
         bool isOpen() const override;
         bool hasChildren() const override;
+        bool hasStarted() const override {
+            return m_runState != NotStarted;
+        }
 
         void addChild( ITrackerPtr const& child ) override;
 
@@ -5886,6 +7604,7 @@ namespace TestCaseTracking {
 
     class SectionTracker : public TrackerBase {
         std::vector<std::string> m_filters;
+        std::string m_trimmed_name;
     public:
         SectionTracker( NameAndLocation const& nameAndLocation, TrackerContext& ctx, ITracker* parent );
 
@@ -5899,6 +7618,10 @@ namespace TestCaseTracking {
 
         void addInitialFilters( std::vector<std::string> const& filters );
         void addNextFilters( std::vector<std::string> const& filters );
+        //! Returns filters active in this tracker
+        std::vector<std::string> const& getFilters() const;
+        //! Returns whitespace-trimmed name of the tracked section
+        std::string const& trimmedName() const;
     };
 
 } // namespace TestCaseTracking
@@ -5923,6 +7646,217 @@ namespace Catch {
 }
 // end catch_leak_detector.h
 // Cpp files will be included in the single-header file here
+// start catch_stats.cpp
+
+// Statistical analysis tools
+
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+
+#include <cassert>
+#include <random>
+
+#if defined(CATCH_CONFIG_USE_ASYNC)
+#include <future>
+#endif
+
+namespace {
+    double erf_inv(double x) {
+        // Code accompanying the article "Approximating the erfinv function" in GPU Computing Gems, Volume 2
+        double w, p;
+
+        w = -log((1.0 - x) * (1.0 + x));
+
+        if (w < 6.250000) {
+            w = w - 3.125000;
+            p = -3.6444120640178196996e-21;
+            p = -1.685059138182016589e-19 + p * w;
+            p = 1.2858480715256400167e-18 + p * w;
+            p = 1.115787767802518096e-17 + p * w;
+            p = -1.333171662854620906e-16 + p * w;
+            p = 2.0972767875968561637e-17 + p * w;
+            p = 6.6376381343583238325e-15 + p * w;
+            p = -4.0545662729752068639e-14 + p * w;
+            p = -8.1519341976054721522e-14 + p * w;
+            p = 2.6335093153082322977e-12 + p * w;
+            p = -1.2975133253453532498e-11 + p * w;
+            p = -5.4154120542946279317e-11 + p * w;
+            p = 1.051212273321532285e-09 + p * w;
+            p = -4.1126339803469836976e-09 + p * w;
+            p = -2.9070369957882005086e-08 + p * w;
+            p = 4.2347877827932403518e-07 + p * w;
+            p = -1.3654692000834678645e-06 + p * w;
+            p = -1.3882523362786468719e-05 + p * w;
+            p = 0.0001867342080340571352 + p * w;
+            p = -0.00074070253416626697512 + p * w;
+            p = -0.0060336708714301490533 + p * w;
+            p = 0.24015818242558961693 + p * w;
+            p = 1.6536545626831027356 + p * w;
+        } else if (w < 16.000000) {
+            w = sqrt(w) - 3.250000;
+            p = 2.2137376921775787049e-09;
+            p = 9.0756561938885390979e-08 + p * w;
+            p = -2.7517406297064545428e-07 + p * w;
+            p = 1.8239629214389227755e-08 + p * w;
+            p = 1.5027403968909827627e-06 + p * w;
+            p = -4.013867526981545969e-06 + p * w;
+            p = 2.9234449089955446044e-06 + p * w;
+            p = 1.2475304481671778723e-05 + p * w;
+            p = -4.7318229009055733981e-05 + p * w;
+            p = 6.8284851459573175448e-05 + p * w;
+            p = 2.4031110387097893999e-05 + p * w;
+            p = -0.0003550375203628474796 + p * w;
+            p = 0.00095328937973738049703 + p * w;
+            p = -0.0016882755560235047313 + p * w;
+            p = 0.0024914420961078508066 + p * w;
+            p = -0.0037512085075692412107 + p * w;
+            p = 0.005370914553590063617 + p * w;
+            p = 1.0052589676941592334 + p * w;
+            p = 3.0838856104922207635 + p * w;
+        } else {
+            w = sqrt(w) - 5.000000;
+            p = -2.7109920616438573243e-11;
+            p = -2.5556418169965252055e-10 + p * w;
+            p = 1.5076572693500548083e-09 + p * w;
+            p = -3.7894654401267369937e-09 + p * w;
+            p = 7.6157012080783393804e-09 + p * w;
+            p = -1.4960026627149240478e-08 + p * w;
+            p = 2.9147953450901080826e-08 + p * w;
+            p = -6.7711997758452339498e-08 + p * w;
+            p = 2.2900482228026654717e-07 + p * w;
+            p = -9.9298272942317002539e-07 + p * w;
+            p = 4.5260625972231537039e-06 + p * w;
+            p = -1.9681778105531670567e-05 + p * w;
+            p = 7.5995277030017761139e-05 + p * w;
+            p = -0.00021503011930044477347 + p * w;
+            p = -0.00013871931833623122026 + p * w;
+            p = 1.0103004648645343977 + p * w;
+            p = 4.8499064014085844221 + p * w;
+        }
+        return p * x;
+    }
+
+    double standard_deviation(std::vector<double>::iterator first, std::vector<double>::iterator last) {
+        auto m = Catch::Benchmark::Detail::mean(first, last);
+        double variance = std::accumulate(first, last, 0., [m](double a, double b) {
+            double diff = b - m;
+            return a + diff * diff;
+            }) / (last - first);
+            return std::sqrt(variance);
+    }
+
+}
+
+namespace Catch {
+    namespace Benchmark {
+        namespace Detail {
+
+            double weighted_average_quantile(int k, int q, std::vector<double>::iterator first, std::vector<double>::iterator last) {
+                auto count = last - first;
+                double idx = (count - 1) * k / static_cast<double>(q);
+                int j = static_cast<int>(idx);
+                double g = idx - j;
+                std::nth_element(first, first + j, last);
+                auto xj = first[j];
+                if (g == 0) return xj;
+
+                auto xj1 = *std::min_element(first + (j + 1), last);
+                return xj + g * (xj1 - xj);
+            }
+
+            double erfc_inv(double x) {
+                return erf_inv(1.0 - x);
+            }
+
+            double normal_quantile(double p) {
+                static const double ROOT_TWO = std::sqrt(2.0);
+
+                double result = 0.0;
+                assert(p >= 0 && p <= 1);
+                if (p < 0 || p > 1) {
+                    return result;
+                }
+
+                result = -erfc_inv(2.0 * p);
+                // result *= normal distribution standard deviation (1.0) * sqrt(2)
+                result *= /*sd * */ ROOT_TWO;
+                // result += normal disttribution mean (0)
+                return result;
+            }
+
+            double outlier_variance(Estimate<double> mean, Estimate<double> stddev, int n) {
+                double sb = stddev.point;
+                double mn = mean.point / n;
+                double mg_min = mn / 2.;
+                double sg = (std::min)(mg_min / 4., sb / std::sqrt(n));
+                double sg2 = sg * sg;
+                double sb2 = sb * sb;
+
+                auto c_max = [n, mn, sb2, sg2](double x) -> double {
+                    double k = mn - x;
+                    double d = k * k;
+                    double nd = n * d;
+                    double k0 = -n * nd;
+                    double k1 = sb2 - n * sg2 + nd;
+                    double det = k1 * k1 - 4 * sg2 * k0;
+                    return (int)(-2. * k0 / (k1 + std::sqrt(det)));
+                };
+
+                auto var_out = [n, sb2, sg2](double c) {
+                    double nc = n - c;
+                    return (nc / n) * (sb2 - nc * sg2);
+                };
+
+                return (std::min)(var_out(1), var_out((std::min)(c_max(0.), c_max(mg_min)))) / sb2;
+            }
+
+            bootstrap_analysis analyse_samples(double confidence_level, int n_resamples, std::vector<double>::iterator first, std::vector<double>::iterator last) {
+                CATCH_INTERNAL_START_WARNINGS_SUPPRESSION
+                CATCH_INTERNAL_SUPPRESS_GLOBALS_WARNINGS
+                static std::random_device entropy;
+                CATCH_INTERNAL_STOP_WARNINGS_SUPPRESSION
+
+                auto n = static_cast<int>(last - first); // seriously, one can't use integral types without hell in C++
+
+                auto mean = &Detail::mean<std::vector<double>::iterator>;
+                auto stddev = &standard_deviation;
+
+#if defined(CATCH_CONFIG_USE_ASYNC)
+                auto Estimate = [=](double(*f)(std::vector<double>::iterator, std::vector<double>::iterator)) {
+                    auto seed = entropy();
+                    return std::async(std::launch::async, [=] {
+                        std::mt19937 rng(seed);
+                        auto resampled = resample(rng, n_resamples, first, last, f);
+                        return bootstrap(confidence_level, first, last, resampled, f);
+                    });
+                };
+
+                auto mean_future = Estimate(mean);
+                auto stddev_future = Estimate(stddev);
+
+                auto mean_estimate = mean_future.get();
+                auto stddev_estimate = stddev_future.get();
+#else
+                auto Estimate = [=](double(*f)(std::vector<double>::iterator, std::vector<double>::iterator)) {
+                    auto seed = entropy();
+                    std::mt19937 rng(seed);
+                    auto resampled = resample(rng, n_resamples, first, last, f);
+                    return bootstrap(confidence_level, first, last, resampled, f);
+                };
+
+                auto mean_estimate = Estimate(mean);
+                auto stddev_estimate = Estimate(stddev);
+#endif // CATCH_USE_ASYNC
+
+                double outlier_variance = Detail::outlier_variance(mean_estimate, stddev_estimate, n);
+
+                return { mean_estimate, stddev_estimate, outlier_variance };
+            }
+        } // namespace Detail
+    } // namespace Benchmark
+} // namespace Catch
+
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
+// end catch_stats.cpp
 // start catch_approx.cpp
 
 #include <cmath>
@@ -5967,21 +7901,22 @@ namespace Detail {
     bool Approx::equalityComparisonImpl(const double other) const {
         // First try with fixed margin, then compute margin based on epsilon, scale and Approx's value
         // Thanks to Richard Harris for his help refining the scaled margin value
-        return marginComparison(m_value, other, m_margin) || marginComparison(m_value, other, m_epsilon * (m_scale + std::fabs(m_value)));
+        return marginComparison(m_value, other, m_margin)
+            || marginComparison(m_value, other, m_epsilon * (m_scale + std::fabs(std::isinf(m_value)? 0 : m_value)));
     }
 
-    void Approx::setMargin(double margin) {
-        CATCH_ENFORCE(margin >= 0,
-            "Invalid Approx::margin: " << margin << '.'
+    void Approx::setMargin(double newMargin) {
+        CATCH_ENFORCE(newMargin >= 0,
+            "Invalid Approx::margin: " << newMargin << '.'
             << " Approx::Margin has to be non-negative.");
-        m_margin = margin;
+        m_margin = newMargin;
     }
 
-    void Approx::setEpsilon(double epsilon) {
-        CATCH_ENFORCE(epsilon >= 0 && epsilon <= 1.0,
-            "Invalid Approx::epsilon: " << epsilon << '.'
+    void Approx::setEpsilon(double newEpsilon) {
+        CATCH_ENFORCE(newEpsilon >= 0 && newEpsilon <= 1.0,
+            "Invalid Approx::epsilon: " << newEpsilon << '.'
             << " Approx::epsilon has to be in [0, 1]");
-        m_epsilon = epsilon;
+        m_epsilon = newEpsilon;
     }
 
 } // end namespace Detail
@@ -6011,7 +7946,24 @@ namespace Catch {
 
 #ifdef CATCH_PLATFORM_MAC
 
-    #define CATCH_TRAP() __asm__("int $3\n" : : ) /* NOLINT */
+    #if defined(__i386__) || defined(__x86_64__)
+        #define CATCH_TRAP() __asm__("int $3\n" : : ) /* NOLINT */
+    #elif defined(__aarch64__)
+        #define CATCH_TRAP()  __asm__(".inst 0xd4200000")
+    #endif
+
+#elif defined(CATCH_PLATFORM_IPHONE)
+
+    // use inline assembler
+    #if defined(__i386__) || defined(__x86_64__)
+        #define CATCH_TRAP()  __asm__("int $3")
+    #elif defined(__aarch64__)
+        #define CATCH_TRAP()  __asm__(".inst 0xd4200000")
+    #elif defined(__arm__) && !defined(__thumb__)
+        #define CATCH_TRAP()  __asm__(".inst 0xe7f001f0")
+    #elif defined(__arm__) &&  defined(__thumb__)
+        #define CATCH_TRAP()  __asm__(".inst 0xde01")
+    #endif
 
 #elif defined(CATCH_PLATFORM_LINUX)
     // If we can use inline assembler, do it because this allows us to break
@@ -6031,10 +7983,12 @@ namespace Catch {
     #define CATCH_TRAP() DebugBreak()
 #endif
 
-#ifdef CATCH_TRAP
-    #define CATCH_BREAK_INTO_DEBUGGER() []{ if( Catch::isDebuggerActive() ) { CATCH_TRAP(); } }()
-#else
-    #define CATCH_BREAK_INTO_DEBUGGER() []{}()
+#ifndef CATCH_BREAK_INTO_DEBUGGER
+    #ifdef CATCH_TRAP
+        #define CATCH_BREAK_INTO_DEBUGGER() []{ if( Catch::isDebuggerActive() ) { CATCH_TRAP(); } }()
+    #else
+        #define CATCH_BREAK_INTO_DEBUGGER() []{}()
+    #endif
 #endif
 
 // end catch_debugger.h
@@ -6042,86 +7996,58 @@ namespace Catch {
 
 // start catch_fatal_condition.h
 
-// start catch_windows_h_proxy.h
-
-
-#if defined(CATCH_PLATFORM_WINDOWS)
-
-#if !defined(NOMINMAX) && !defined(CATCH_CONFIG_NO_NOMINMAX)
-#  define CATCH_DEFINED_NOMINMAX
-#  define NOMINMAX
-#endif
-#if !defined(WIN32_LEAN_AND_MEAN) && !defined(CATCH_CONFIG_NO_WIN32_LEAN_AND_MEAN)
-#  define CATCH_DEFINED_WIN32_LEAN_AND_MEAN
-#  define WIN32_LEAN_AND_MEAN
-#endif
-
-#ifdef __AFXDLL
-#include <AfxWin.h>
-#else
-#include <windows.h>
-#endif
-
-#ifdef CATCH_DEFINED_NOMINMAX
-#  undef NOMINMAX
-#endif
-#ifdef CATCH_DEFINED_WIN32_LEAN_AND_MEAN
-#  undef WIN32_LEAN_AND_MEAN
-#endif
-
-#endif // defined(CATCH_PLATFORM_WINDOWS)
-
-// end catch_windows_h_proxy.h
-#if defined( CATCH_CONFIG_WINDOWS_SEH )
+#include <cassert>
 
 namespace Catch {
 
-    struct FatalConditionHandler {
-
-        static LONG CALLBACK handleVectoredException(PEXCEPTION_POINTERS ExceptionInfo);
+    // Wrapper for platform-specific fatal error (signals/SEH) handlers
+    //
+    // Tries to be cooperative with other handlers, and not step over
+    // other handlers. This means that unknown structured exceptions
+    // are passed on, previous signal handlers are called, and so on.
+    //
+    // Can only be instantiated once, and assumes that once a signal
+    // is caught, the binary will end up terminating. Thus, there
+    class FatalConditionHandler {
+        bool m_started = false;
+
+        // Install/disengage implementation for specific platform.
+        // Should be if-defed to work on current platform, can assume
+        // engage-disengage 1:1 pairing.
+        void engage_platform();
+        void disengage_platform();
+    public:
+        // Should also have platform-specific implementations as needed
         FatalConditionHandler();
-        static void reset();
         ~FatalConditionHandler();
 
-    private:
-        static bool isSet;
-        static ULONG guaranteeSize;
-        static PVOID exceptionHandlerHandle;
-    };
-
-} // namespace Catch
-
-#elif defined ( CATCH_CONFIG_POSIX_SIGNALS )
-
-#include <signal.h>
-
-namespace Catch {
-
-    struct FatalConditionHandler {
-
-        static bool isSet;
-        static struct sigaction oldSigActions[];
-        static stack_t oldSigStack;
-        static char altStackMem[];
-
-        static void handleSignal( int sig );
+        void engage() {
+            assert(!m_started && "Handler cannot be installed twice.");
+            m_started = true;
+            engage_platform();
+        }
 
-        FatalConditionHandler();
-        ~FatalConditionHandler();
-        static void reset();
+        void disengage() {
+            assert(m_started && "Handler cannot be uninstalled without being installed first");
+            m_started = false;
+            disengage_platform();
+        }
     };
 
-} // namespace Catch
-
-#else
-
-namespace Catch {
-    struct FatalConditionHandler {
-        void reset();
+    //! Simple RAII guard for (dis)engaging the FatalConditionHandler
+    class FatalConditionHandlerGuard {
+        FatalConditionHandler* m_handler;
+    public:
+        FatalConditionHandlerGuard(FatalConditionHandler* handler):
+            m_handler(handler) {
+            m_handler->engage();
+        }
+        ~FatalConditionHandlerGuard() {
+            m_handler->disengage();
+        }
     };
-}
 
-#endif
+} // end namespace Catch
 
 // end catch_fatal_condition.h
 #include <string>
@@ -6181,10 +8107,14 @@ namespace Catch {
         void sectionEnded( SectionEndInfo const& endInfo ) override;
         void sectionEndedEarly( SectionEndInfo const& endInfo ) override;
 
-        auto acquireGeneratorTracker( SourceLineInfo const& lineInfo ) -> IGeneratorTracker& override;
+        auto acquireGeneratorTracker( StringRef generatorName, SourceLineInfo const& lineInfo ) -> IGeneratorTracker& override;
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+        void benchmarkPreparing( std::string const& name ) override;
         void benchmarkStarting( BenchmarkInfo const& info ) override;
-        void benchmarkEnded( BenchmarkStats const& stats ) override;
+        void benchmarkEnded( BenchmarkStats<> const& stats ) override;
+        void benchmarkFailed( std::string const& error ) override;
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
         void pushScopedMessage( MessageInfo const& message ) override;
         void popScopedMessage( MessageInfo const& message ) override;
@@ -6243,11 +8173,14 @@ namespace Catch {
         std::vector<SectionEndInfo> m_unfinishedSections;
         std::vector<ITracker*> m_activeSections;
         TrackerContext m_trackerContext;
+        FatalConditionHandler m_fatalConditionhandler;
         bool m_lastAssertionPassed = false;
         bool m_shouldReportUnexpected = true;
         bool m_includeSuccessfulResults;
     };
 
+    void seedRng(IConfig const& config);
+    unsigned int rngSeed();
 } // end namespace Catch
 
 // end catch_run_context.h
@@ -6394,7 +8327,7 @@ namespace Catch {
     }
 
     bool AssertionResult::hasExpression() const {
-        return m_info.capturedExpression[0] != 0;
+        return !m_info.capturedExpression.empty();
     }
 
     bool AssertionResult::hasMessage() const {
@@ -6402,16 +8335,22 @@ namespace Catch {
     }
 
     std::string AssertionResult::getExpression() const {
-        if( isFalseTest( m_info.resultDisposition ) )
-            return "!(" + m_info.capturedExpression + ")";
-        else
-            return m_info.capturedExpression;
+        // Possibly overallocating by 3 characters should be basically free
+        std::string expr; expr.reserve(m_info.capturedExpression.size() + 3);
+        if (isFalseTest(m_info.resultDisposition)) {
+            expr += "!(";
+        }
+        expr += m_info.capturedExpression;
+        if (isFalseTest(m_info.resultDisposition)) {
+            expr += ')';
+        }
+        return expr;
     }
 
     std::string AssertionResult::getExpressionInMacro() const {
         std::string expr;
-        if( m_info.macroName[0] == 0 )
-            expr = m_info.capturedExpression;
+        if( m_info.macroName.empty() )
+            expr = static_cast<std::string>(m_info.capturedExpression);
         else {
             expr.reserve( m_info.macroName.size() + m_info.capturedExpression.size() + 4 );
             expr += m_info.macroName;
@@ -6446,32 +8385,6 @@ namespace Catch {
 
 } // end namespace Catch
 // end catch_assertionresult.cpp
-// start catch_benchmark.cpp
-
-namespace Catch {
-
-    auto BenchmarkLooper::getResolution() -> uint64_t {
-        return getEstimatedClockResolution() * getCurrentContext().getConfig()->benchmarkResolutionMultiple();
-    }
-
-    void BenchmarkLooper::reportStart() {
-        getResultCapture().benchmarkStarting( { m_name } );
-    }
-    auto BenchmarkLooper::needsMoreIterations() -> bool {
-        auto elapsed = m_timer.getElapsedNanoseconds();
-
-        // Exponentially increasing iterations until we're confident in our timer resolution
-        if( elapsed < m_resolution ) {
-            m_iterationsToRun *= 10;
-            return true;
-        }
-
-        getResultCapture().benchmarkEnded( { { m_name }, m_count, elapsed } );
-        return false;
-    }
-
-} // end namespace Catch
-// end catch_benchmark.cpp
 // start catch_capture_matchers.cpp
 
 namespace Catch {
@@ -6617,6 +8530,9 @@ public:
 			m_suffix = false;
 			auto width = m_column.m_width - indent();
 			m_end = m_pos;
+			if (line()[m_pos] == '\n') {
+				++m_end;
+			}
 			while (m_end < line().size() && line()[m_end] != '\n')
 				++m_end;
 
@@ -7168,7 +9084,7 @@ namespace detail {
     }
     inline auto convertInto( std::string const &source, bool &target ) -> ParserResult {
         std::string srcLC = source;
-        std::transform( srcLC.begin(), srcLC.end(), srcLC.begin(), []( char c ) { return static_cast<char>( std::tolower(c) ); } );
+        std::transform( srcLC.begin(), srcLC.end(), srcLC.begin(), []( unsigned char c ) { return static_cast<char>( std::tolower(c) ); } );
         if (srcLC == "y" || srcLC == "1" || srcLC == "true" || srcLC == "yes" || srcLC == "on")
             target = true;
         else if (srcLC == "n" || srcLC == "0" || srcLC == "false" || srcLC == "no" || srcLC == "off")
@@ -7816,9 +9732,14 @@ namespace Catch {
                     if( !line.empty() && !startsWith( line, '#' ) ) {
                         if( !startsWith( line, '"' ) )
                             line = '"' + line + '"';
-                        config.testsOrTags.push_back( line + ',' );
+                        config.testsOrTags.push_back( line );
+                        config.testsOrTags.emplace_back( "," );
                     }
                 }
+                //Remove comma in the end
+                if(!config.testsOrTags.empty())
+                    config.testsOrTags.erase( config.testsOrTags.end()-1 );
+
                 return ParserResult::ok( ParseResultType::Matched );
             };
         auto const setTestOrder = [&]( std::string const& order ) {
@@ -7853,14 +9774,16 @@ namespace Catch {
             };
         auto const setWaitForKeypress = [&]( std::string const& keypress ) {
                 auto keypressLc = toLower( keypress );
-                if( keypressLc == "start" )
+                if (keypressLc == "never")
+                    config.waitForKeypress = WaitForKeypress::Never;
+                else if( keypressLc == "start" )
                     config.waitForKeypress = WaitForKeypress::BeforeStart;
                 else if( keypressLc == "exit" )
                     config.waitForKeypress = WaitForKeypress::BeforeExit;
                 else if( keypressLc == "both" )
                     config.waitForKeypress = WaitForKeypress::BeforeStartAndExit;
                 else
-                    return ParserResult::runtimeError( "keypress argument must be one of: start, exit or both. '" + keypress + "' not recognised" );
+                    return ParserResult::runtimeError( "keypress argument must be one of: never, start, exit or both. '" + keypress + "' not recognised" );
             return ParserResult::ok( ParseResultType::Matched );
             };
         auto const setVerbosity = [&]( std::string const& verbosity ) {
@@ -7930,6 +9853,9 @@ namespace Catch {
             | Opt( [&]( bool flag ) { config.showDurations = flag ? ShowDurations::Always : ShowDurations::Never; }, "yes|no" )
                 ["-d"]["--durations"]
                 ( "show test durations" )
+            | Opt( config.minDuration, "seconds" )
+                ["-D"]["--min-duration"]
+                ( "show test durations for tests taking at least the given number of seconds" )
             | Opt( loadTestNamesFromFile, "filename" )
                 ["-f"]["--input-file"]
                 ( "load test names to run from a file" )
@@ -7960,13 +9886,24 @@ namespace Catch {
             | Opt( config.libIdentify )
                 ["--libidentify"]
                 ( "report name and version according to libidentify standard" )
-            | Opt( setWaitForKeypress, "start|exit|both" )
+            | Opt( setWaitForKeypress, "never|start|exit|both" )
                 ["--wait-for-keypress"]
                 ( "waits for a keypress before exiting" )
-            | Opt( config.benchmarkResolutionMultiple, "multiplier" )
-                ["--benchmark-resolution-multiple"]
-                ( "multiple of clock resolution to run benchmarks" )
-
+            | Opt( config.benchmarkSamples, "samples" )
+                ["--benchmark-samples"]
+                ( "number of samples to collect (default: 100)" )
+            | Opt( config.benchmarkResamples, "resamples" )
+                ["--benchmark-resamples"]
+                ( "number of resamples for the bootstrap (default: 100000)" )
+            | Opt( config.benchmarkConfidenceInterval, "confidence interval" )
+                ["--benchmark-confidence-interval"]
+                ( "confidence interval for the bootstrap (between 0 and 1, default: 0.95)" )
+            | Opt( config.benchmarkNoAnalysis )
+                ["--benchmark-no-analysis"]
+                ( "perform only measurements; do not perform any analysis" )
+            | Opt( config.benchmarkWarmupTime, "benchmarkWarmupTime" )
+                ["--benchmark-warmup-time"]
+                ( "amount of time in milliseconds spent on warming up each test (default: 100)" )
             | Arg( config.testsOrTags, "test name|pattern|tags" )
                 ( "which test or tests to use" );
 
@@ -7982,9 +9919,6 @@ namespace Catch {
 
 namespace Catch {
 
-    bool SourceLineInfo::empty() const noexcept {
-        return file[0] == '\0';
-    }
     bool SourceLineInfo::operator == ( SourceLineInfo const& other ) const noexcept {
         return line == other.line && (file == other.file || std::strcmp(file, other.file) == 0);
     }
@@ -8020,14 +9954,23 @@ namespace Catch {
     :   m_data( data ),
         m_stream( openStream() )
     {
-        TestSpecParser parser(ITagAliasRegistry::get());
-        if (data.testsOrTags.empty()) {
-            parser.parse("~[.]"); // All not hidden tests
+        // We need to trim filter specs to avoid trouble with superfluous
+        // whitespace (esp. important for bdd macros, as those are manually
+        // aligned with whitespace).
+
+        for (auto& elem : m_data.testsOrTags) {
+            elem = trim(elem);
         }
-        else {
+        for (auto& elem : m_data.sectionsToRun) {
+            elem = trim(elem);
+        }
+
+        TestSpecParser parser(ITagAliasRegistry::get());
+        if (!m_data.testsOrTags.empty()) {
             m_hasTestFilters = true;
-            for( auto const& testOrTags : data.testsOrTags )
-                parser.parse( testOrTags );
+            for (auto const& testOrTags : m_data.testsOrTags) {
+                parser.parse(testOrTags);
+            }
         }
         m_testSpec = parser.testSpec();
     }
@@ -8060,15 +10003,21 @@ namespace Catch {
     bool Config::warnAboutMissingAssertions() const    { return !!(m_data.warnings & WarnAbout::NoAssertions); }
     bool Config::warnAboutNoTests() const              { return !!(m_data.warnings & WarnAbout::NoTests); }
     ShowDurations::OrNot Config::showDurations() const { return m_data.showDurations; }
+    double Config::minDuration() const                 { return m_data.minDuration; }
     RunTests::InWhatOrder Config::runOrder() const     { return m_data.runOrder; }
     unsigned int Config::rngSeed() const               { return m_data.rngSeed; }
-    int Config::benchmarkResolutionMultiple() const    { return m_data.benchmarkResolutionMultiple; }
     UseColour::YesOrNo Config::useColour() const       { return m_data.useColour; }
     bool Config::shouldDebugBreak() const              { return m_data.shouldDebugBreak; }
     int Config::abortAfter() const                     { return m_data.abortAfter; }
     bool Config::showInvisibles() const                { return m_data.showInvisibles; }
     Verbosity Config::verbosity() const                { return m_data.verbosity; }
 
+    bool Config::benchmarkNoAnalysis() const                      { return m_data.benchmarkNoAnalysis; }
+    int Config::benchmarkSamples() const                          { return m_data.benchmarkSamples; }
+    double Config::benchmarkConfidenceInterval() const            { return m_data.benchmarkConfidenceInterval; }
+    unsigned int Config::benchmarkResamples() const               { return m_data.benchmarkResamples; }
+    std::chrono::milliseconds Config::benchmarkWarmupTime() const { return std::chrono::milliseconds(m_data.benchmarkWarmupTime); }
+
     IStream const* Config::openStream() {
         return Catch::makeStream(m_data.outputFilename);
     }
@@ -8097,6 +10046,36 @@ namespace Catch {
 }
 
 // end catch_errno_guard.h
+// start catch_windows_h_proxy.h
+
+
+#if defined(CATCH_PLATFORM_WINDOWS)
+
+#if !defined(NOMINMAX) && !defined(CATCH_CONFIG_NO_NOMINMAX)
+#  define CATCH_DEFINED_NOMINMAX
+#  define NOMINMAX
+#endif
+#if !defined(WIN32_LEAN_AND_MEAN) && !defined(CATCH_CONFIG_NO_WIN32_LEAN_AND_MEAN)
+#  define CATCH_DEFINED_WIN32_LEAN_AND_MEAN
+#  define WIN32_LEAN_AND_MEAN
+#endif
+
+#ifdef __AFXDLL
+#include <AfxWin.h>
+#else
+#include <windows.h>
+#endif
+
+#ifdef CATCH_DEFINED_NOMINMAX
+#  undef NOMINMAX
+#endif
+#ifdef CATCH_DEFINED_WIN32_LEAN_AND_MEAN
+#  undef WIN32_LEAN_AND_MEAN
+#endif
+
+#endif // defined(CATCH_PLATFORM_WINDOWS)
+
+// end catch_windows_h_proxy.h
 #include <sstream>
 
 namespace Catch {
@@ -8108,7 +10087,7 @@ namespace Catch {
         };
 
         struct NoColourImpl : IColourImpl {
-            void use( Colour::Code ) {}
+            void use( Colour::Code ) override {}
 
             static IColourImpl* instance() {
                 static NoColourImpl s_instance;
@@ -8142,7 +10121,7 @@ namespace {
             originalBackgroundAttributes = csbiInfo.wAttributes & ~( FOREGROUND_GREEN | FOREGROUND_RED | FOREGROUND_BLUE | FOREGROUND_INTENSITY );
         }
 
-        virtual void use( Colour::Code _colourCode ) override {
+        void use( Colour::Code _colourCode ) override {
             switch( _colourCode ) {
                 case Colour::None:      return setTextAttribute( originalForegroundAttributes );
                 case Colour::White:     return setTextAttribute( FOREGROUND_GREEN | FOREGROUND_RED | FOREGROUND_BLUE );
@@ -8205,7 +10184,7 @@ namespace {
     // https://github.com/philsquared/Catch/pull/131
     class PosixColourImpl : public IColourImpl {
     public:
-        virtual void use( Colour::Code _colourCode ) override {
+        void use( Colour::Code _colourCode ) override {
             switch( _colourCode ) {
                 case Colour::None:
                 case Colour::White:     return setColour( "[0m" );
@@ -8240,7 +10219,7 @@ namespace {
 
     bool useColourOnPlatform() {
         return
-#ifdef CATCH_PLATFORM_MAC
+#if defined(CATCH_PLATFORM_MAC) || defined(CATCH_PLATFORM_IPHONE)
             !isDebuggerActive() &&
 #endif
 #if !(defined(__DJGPP__) && defined(__STRICT_ANSI__))
@@ -8281,13 +10260,13 @@ namespace Catch {
 namespace Catch {
 
     Colour::Colour( Code _colourCode ) { use( _colourCode ); }
-    Colour::Colour( Colour&& rhs ) noexcept {
-        m_moved = rhs.m_moved;
-        rhs.m_moved = true;
+    Colour::Colour( Colour&& other ) noexcept {
+        m_moved = other.m_moved;
+        other.m_moved = true;
     }
-    Colour& Colour::operator=( Colour&& rhs ) noexcept {
-        m_moved = rhs.m_moved;
-        rhs.m_moved  = true;
+    Colour& Colour::operator=( Colour&& other ) noexcept {
+        m_moved = other.m_moved;
+        other.m_moved  = true;
         return *this;
     }
 
@@ -8295,7 +10274,13 @@ namespace Catch {
 
     void Colour::use( Code _colourCode ) {
         static IColourImpl* impl = platformColourInstance();
-        impl->use( _colourCode );
+        // Strictly speaking, this cannot possibly happen.
+        // However, under some conditions it does happen (see #1626),
+        // and this change is small enough that we can let practicality
+        // triumph over purity in this case.
+        if (impl != nullptr) {
+            impl->use( _colourCode );
+        }
     }
 
     std::ostream& operator << ( std::ostream& os, Colour const& ) {
@@ -8316,27 +10301,27 @@ namespace Catch {
     class Context : public IMutableContext, NonCopyable {
 
     public: // IContext
-        virtual IResultCapture* getResultCapture() override {
+        IResultCapture* getResultCapture() override {
             return m_resultCapture;
         }
-        virtual IRunner* getRunner() override {
+        IRunner* getRunner() override {
             return m_runner;
         }
 
-        virtual IConfigPtr const& getConfig() const override {
+        IConfigPtr const& getConfig() const override {
             return m_config;
         }
 
-        virtual ~Context() override;
+        ~Context() override;
 
     public: // IMutableContext
-        virtual void setResultCapture( IResultCapture* resultCapture ) override {
+        void setResultCapture( IResultCapture* resultCapture ) override {
             m_resultCapture = resultCapture;
         }
-        virtual void setRunner( IRunner* runner ) override {
+        void setRunner( IRunner* runner ) override {
             m_runner = runner;
         }
-        virtual void setConfig( IConfigPtr const& config ) override {
+        void setConfig( IConfigPtr const& config ) override {
             m_config = config;
         }
 
@@ -8362,6 +10347,12 @@ namespace Catch {
     IContext::~IContext() = default;
     IMutableContext::~IMutableContext() = default;
     Context::~Context() = default;
+
+    SimplePcg32& rng() {
+        static SimplePcg32 s_rng;
+        return s_rng;
+    }
+
 }
 // end catch_context.cpp
 // start catch_debug_console.cpp
@@ -8375,7 +10366,16 @@ namespace Catch {
 }
 
 // end catch_debug_console.h
-#ifdef CATCH_PLATFORM_WINDOWS
+#if defined(CATCH_CONFIG_ANDROID_LOGWRITE)
+#include <android/log.h>
+
+    namespace Catch {
+        void writeToDebugConsole( std::string const& text ) {
+            __android_log_write( ANDROID_LOG_DEBUG, "Catch", text.c_str() );
+        }
+    }
+
+#elif defined(CATCH_PLATFORM_WINDOWS)
 
     namespace Catch {
         void writeToDebugConsole( std::string const& text ) {
@@ -8396,25 +10396,28 @@ namespace Catch {
 // end catch_debug_console.cpp
 // start catch_debugger.cpp
 
-#ifdef CATCH_PLATFORM_MAC
+#if defined(CATCH_PLATFORM_MAC) || defined(CATCH_PLATFORM_IPHONE)
 
-#  include <assert.h>
-#  include <stdbool.h>
+#  include <cassert>
 #  include <sys/types.h>
 #  include <unistd.h>
-#  include <sys/sysctl.h>
 #  include <cstddef>
 #  include <ostream>
 
-namespace Catch {
+#ifdef __apple_build_version__
+    // These headers will only compile with AppleClang (XCode)
+    // For other compilers (Clang, GCC, ... ) we need to exclude them
+#  include <sys/sysctl.h>
+#endif
 
+    namespace Catch {
+        #ifdef __apple_build_version__
         // The following function is taken directly from the following technical note:
-        // http://developer.apple.com/library/mac/#qa/qa2004/qa1361.html
+        // https://developer.apple.com/library/archive/qa/qa1361/_index.html
 
         // Returns true if the current process is being debugged (either
         // running under the debugger or has a debugger attached post facto).
         bool isDebuggerActive(){
-
             int                 mib[4];
             struct kinfo_proc   info;
             std::size_t         size;
@@ -8444,6 +10447,12 @@ namespace Catch {
 
             return ( (info.kp_proc.p_flag & P_TRACED) != 0 );
         }
+        #else
+        bool isDebuggerActive() {
+            // We need to find another way to determine this for non-appleclang compilers on macOS
+            return false;
+        }
+        #endif
     } // namespace Catch
 
 #elif defined(CATCH_PLATFORM_LINUX)
@@ -8514,6 +10523,8 @@ namespace Catch {
 // end catch_decomposer.cpp
 // start catch_enforce.cpp
 
+#include <stdexcept>
+
 namespace Catch {
 #if defined(CATCH_CONFIG_DISABLE_EXCEPTIONS) && !defined(CATCH_CONFIG_DISABLE_EXCEPTIONS_CUSTOM_HANDLER)
     [[noreturn]]
@@ -8523,8 +10534,116 @@ namespace Catch {
         std::terminate();
     }
 #endif
+
+    [[noreturn]]
+    void throw_logic_error(std::string const& msg) {
+        throw_exception(std::logic_error(msg));
+    }
+
+    [[noreturn]]
+    void throw_domain_error(std::string const& msg) {
+        throw_exception(std::domain_error(msg));
+    }
+
+    [[noreturn]]
+    void throw_runtime_error(std::string const& msg) {
+        throw_exception(std::runtime_error(msg));
+    }
+
 } // namespace Catch;
 // end catch_enforce.cpp
+// start catch_enum_values_registry.cpp
+// start catch_enum_values_registry.h
+
+#include <vector>
+#include <memory>
+
+namespace Catch {
+
+    namespace Detail {
+
+        std::unique_ptr<EnumInfo> makeEnumInfo( StringRef enumName, StringRef allValueNames, std::vector<int> const& values );
+
+        class EnumValuesRegistry : public IMutableEnumValuesRegistry {
+
+            std::vector<std::unique_ptr<EnumInfo>> m_enumInfos;
+
+            EnumInfo const& registerEnum( StringRef enumName, StringRef allEnums, std::vector<int> const& values) override;
+        };
+
+        std::vector<StringRef> parseEnums( StringRef enums );
+
+    } // Detail
+
+} // Catch
+
+// end catch_enum_values_registry.h
+
+#include <map>
+#include <cassert>
+
+namespace Catch {
+
+    IMutableEnumValuesRegistry::~IMutableEnumValuesRegistry() {}
+
+    namespace Detail {
+
+        namespace {
+            // Extracts the actual name part of an enum instance
+            // In other words, it returns the Blue part of Bikeshed::Colour::Blue
+            StringRef extractInstanceName(StringRef enumInstance) {
+                // Find last occurrence of ":"
+                size_t name_start = enumInstance.size();
+                while (name_start > 0 && enumInstance[name_start - 1] != ':') {
+                    --name_start;
+                }
+                return enumInstance.substr(name_start, enumInstance.size() - name_start);
+            }
+        }
+
+        std::vector<StringRef> parseEnums( StringRef enums ) {
+            auto enumValues = splitStringRef( enums, ',' );
+            std::vector<StringRef> parsed;
+            parsed.reserve( enumValues.size() );
+            for( auto const& enumValue : enumValues ) {
+                parsed.push_back(trim(extractInstanceName(enumValue)));
+            }
+            return parsed;
+        }
+
+        EnumInfo::~EnumInfo() {}
+
+        StringRef EnumInfo::lookup( int value ) const {
+            for( auto const& valueToName : m_values ) {
+                if( valueToName.first == value )
+                    return valueToName.second;
+            }
+            return "{** unexpected enum value **}"_sr;
+        }
+
+        std::unique_ptr<EnumInfo> makeEnumInfo( StringRef enumName, StringRef allValueNames, std::vector<int> const& values ) {
+            std::unique_ptr<EnumInfo> enumInfo( new EnumInfo );
+            enumInfo->m_name = enumName;
+            enumInfo->m_values.reserve( values.size() );
+
+            const auto valueNames = Catch::Detail::parseEnums( allValueNames );
+            assert( valueNames.size() == values.size() );
+            std::size_t i = 0;
+            for( auto value : values )
+                enumInfo->m_values.emplace_back(value, valueNames[i++]);
+
+            return enumInfo;
+        }
+
+        EnumInfo const& EnumValuesRegistry::registerEnum( StringRef enumName, StringRef allValueNames, std::vector<int> const& values ) {
+            m_enumInfos.push_back(makeEnumInfo(enumName, allValueNames, values));
+            return *m_enumInfos.back();
+        }
+
+    } // Detail
+} // Catch
+
+// end catch_enum_values_registry.cpp
 // start catch_errno_guard.cpp
 
 #include <cerrno>
@@ -8548,7 +10667,7 @@ namespace Catch {
     public:
         ~ExceptionTranslatorRegistry();
         virtual void registerTranslator( const IExceptionTranslator* translator );
-        virtual std::string translateActiveException() const override;
+        std::string translateActiveException() const override;
         std::string tryTranslators() const;
 
     private:
@@ -8635,38 +10754,60 @@ namespace Catch {
 // end catch_exception_translator_registry.cpp
 // start catch_fatal_condition.cpp
 
-#if defined(__GNUC__)
-#    pragma GCC diagnostic push
-#    pragma GCC diagnostic ignored "-Wmissing-field-initializers"
-#endif
+#include <algorithm>
+
+#if !defined( CATCH_CONFIG_WINDOWS_SEH ) && !defined( CATCH_CONFIG_POSIX_SIGNALS )
+
+namespace Catch {
+
+    // If neither SEH nor signal handling is required, the handler impls
+    // do not have to do anything, and can be empty.
+    void FatalConditionHandler::engage_platform() {}
+    void FatalConditionHandler::disengage_platform() {}
+    FatalConditionHandler::FatalConditionHandler() = default;
+    FatalConditionHandler::~FatalConditionHandler() = default;
+
+} // end namespace Catch
+
+#endif // !CATCH_CONFIG_WINDOWS_SEH && !CATCH_CONFIG_POSIX_SIGNALS
+
+#if defined( CATCH_CONFIG_WINDOWS_SEH ) && defined( CATCH_CONFIG_POSIX_SIGNALS )
+#error "Inconsistent configuration: Windows' SEH handling and POSIX signals cannot be enabled at the same time"
+#endif // CATCH_CONFIG_WINDOWS_SEH && CATCH_CONFIG_POSIX_SIGNALS
 
 #if defined( CATCH_CONFIG_WINDOWS_SEH ) || defined( CATCH_CONFIG_POSIX_SIGNALS )
 
 namespace {
-    // Report the error condition
+    //! Signals fatal error message to the run context
     void reportFatal( char const * const message ) {
         Catch::getCurrentContext().getResultCapture()->handleFatalErrorCondition( message );
     }
-}
 
-#endif // signals/SEH handling
+    //! Minimal size Catch2 needs for its own fatal error handling.
+    //! Picked anecdotally, so it might not be sufficient on all
+    //! platforms, and for all configurations.
+    constexpr std::size_t minStackSizeForErrors = 32 * 1024;
+} // end unnamed namespace
+
+#endif // CATCH_CONFIG_WINDOWS_SEH || CATCH_CONFIG_POSIX_SIGNALS
 
 #if defined( CATCH_CONFIG_WINDOWS_SEH )
 
 namespace Catch {
+
     struct SignalDefs { DWORD id; const char* name; };
 
     // There is no 1-1 mapping between signals and windows exceptions.
     // Windows can easily distinguish between SO and SigSegV,
     // but SigInt, SigTerm, etc are handled differently.
     static SignalDefs signalDefs[] = {
-        { EXCEPTION_ILLEGAL_INSTRUCTION,  "SIGILL - Illegal instruction signal" },
-        { EXCEPTION_STACK_OVERFLOW, "SIGSEGV - Stack overflow" },
-        { EXCEPTION_ACCESS_VIOLATION, "SIGSEGV - Segmentation violation signal" },
-        { EXCEPTION_INT_DIVIDE_BY_ZERO, "Divide by zero error" },
+        { static_cast<DWORD>(EXCEPTION_ILLEGAL_INSTRUCTION),  "SIGILL - Illegal instruction signal" },
+        { static_cast<DWORD>(EXCEPTION_STACK_OVERFLOW), "SIGSEGV - Stack overflow" },
+        { static_cast<DWORD>(EXCEPTION_ACCESS_VIOLATION), "SIGSEGV - Segmentation violation signal" },
+        { static_cast<DWORD>(EXCEPTION_INT_DIVIDE_BY_ZERO), "Divide by zero error" },
     };
 
-    LONG CALLBACK FatalConditionHandler::handleVectoredException(PEXCEPTION_POINTERS ExceptionInfo) {
+    static LONG CALLBACK handleVectoredException(PEXCEPTION_POINTERS ExceptionInfo) {
         for (auto const& def : signalDefs) {
             if (ExceptionInfo->ExceptionRecord->ExceptionCode == def.id) {
                 reportFatal(def.name);
@@ -8677,38 +10818,50 @@ namespace Catch {
         return EXCEPTION_CONTINUE_SEARCH;
     }
 
+    // Since we do not support multiple instantiations, we put these
+    // into global variables and rely on cleaning them up in outlined
+    // constructors/destructors
+    static PVOID exceptionHandlerHandle = nullptr;
+
+    // For MSVC, we reserve part of the stack memory for handling
+    // memory overflow structured exception.
     FatalConditionHandler::FatalConditionHandler() {
-        isSet = true;
-        // 32k seems enough for Catch to handle stack overflow,
-        // but the value was found experimentally, so there is no strong guarantee
-        guaranteeSize = 32 * 1024;
-        exceptionHandlerHandle = nullptr;
+        ULONG guaranteeSize = static_cast<ULONG>(minStackSizeForErrors);
+        if (!SetThreadStackGuarantee(&guaranteeSize)) {
+            // We do not want to fully error out, because needing
+            // the stack reserve should be rare enough anyway.
+            Catch::cerr()
+                << "Failed to reserve piece of stack."
+                << " Stack overflows will not be reported successfully.";
+        }
+    }
+
+    // We do not attempt to unset the stack guarantee, because
+    // Windows does not support lowering the stack size guarantee.
+    FatalConditionHandler::~FatalConditionHandler() = default;
+
+    void FatalConditionHandler::engage_platform() {
         // Register as first handler in current chain
         exceptionHandlerHandle = AddVectoredExceptionHandler(1, handleVectoredException);
-        // Pass in guarantee size to be filled
-        SetThreadStackGuarantee(&guaranteeSize);
+        if (!exceptionHandlerHandle) {
+            CATCH_RUNTIME_ERROR("Could not register vectored exception handler");
+        }
     }
 
-    void FatalConditionHandler::reset() {
-        if (isSet) {
-            RemoveVectoredExceptionHandler(exceptionHandlerHandle);
-            SetThreadStackGuarantee(&guaranteeSize);
-            exceptionHandlerHandle = nullptr;
-            isSet = false;
+    void FatalConditionHandler::disengage_platform() {
+        if (!RemoveVectoredExceptionHandler(exceptionHandlerHandle)) {
+            CATCH_RUNTIME_ERROR("Could not unregister vectored exception handler");
         }
+        exceptionHandlerHandle = nullptr;
     }
 
-    FatalConditionHandler::~FatalConditionHandler() {
-        reset();
-    }
+} // end namespace Catch
 
-bool FatalConditionHandler::isSet = false;
-ULONG FatalConditionHandler::guaranteeSize = 0;
-PVOID FatalConditionHandler::exceptionHandlerHandle = nullptr;
+#endif // CATCH_CONFIG_WINDOWS_SEH
 
-} // namespace Catch
+#if defined( CATCH_CONFIG_POSIX_SIGNALS )
 
-#elif defined( CATCH_CONFIG_POSIX_SIGNALS )
+#include <signal.h>
 
 namespace Catch {
 
@@ -8717,10 +10870,6 @@ namespace Catch {
         const char* name;
     };
 
-    // 32kb for the alternate stack seems to be sufficient. However, this value
-    // is experimentally determined, so that's not guaranteed.
-    constexpr static std::size_t sigStackSize = 32768 >= MINSIGSTKSZ ? 32768 : MINSIGSTKSZ;
-
     static SignalDefs signalDefs[] = {
         { SIGINT,  "SIGINT - Terminal interrupt signal" },
         { SIGILL,  "SIGILL - Illegal instruction signal" },
@@ -8730,7 +10879,32 @@ namespace Catch {
         { SIGABRT, "SIGABRT - Abort (abnormal termination) signal" }
     };
 
-    void FatalConditionHandler::handleSignal( int sig ) {
+// Older GCCs trigger -Wmissing-field-initializers for T foo = {}
+// which is zero initialization, but not explicit. We want to avoid
+// that.
+#if defined(__GNUC__)
+#    pragma GCC diagnostic push
+#    pragma GCC diagnostic ignored "-Wmissing-field-initializers"
+#endif
+
+    static char* altStackMem = nullptr;
+    static std::size_t altStackSize = 0;
+    static stack_t oldSigStack{};
+    static struct sigaction oldSigActions[sizeof(signalDefs) / sizeof(SignalDefs)]{};
+
+    static void restorePreviousSignalHandlers() {
+        // We set signal handlers back to the previous ones. Hopefully
+        // nobody overwrote them in the meantime, and doesn't expect
+        // their signal handlers to live past ours given that they
+        // installed them after ours..
+        for (std::size_t i = 0; i < sizeof(signalDefs) / sizeof(SignalDefs); ++i) {
+            sigaction(signalDefs[i].id, &oldSigActions[i], nullptr);
+        }
+        // Return the old stack
+        sigaltstack(&oldSigStack, nullptr);
+    }
+
+    static void handleSignal( int sig ) {
         char const * name = "<unknown signal>";
         for (auto const& def : signalDefs) {
             if (sig == def.id) {
@@ -8738,16 +10912,33 @@ namespace Catch {
                 break;
             }
         }
-        reset();
-        reportFatal(name);
+        // We need to restore previous signal handlers and let them do
+        // their thing, so that the users can have the debugger break
+        // when a signal is raised, and so on.
+        restorePreviousSignalHandlers();
+        reportFatal( name );
         raise( sig );
     }
 
     FatalConditionHandler::FatalConditionHandler() {
-        isSet = true;
+        assert(!altStackMem && "Cannot initialize POSIX signal handler when one already exists");
+        if (altStackSize == 0) {
+            altStackSize = std::max(static_cast<size_t>(SIGSTKSZ), minStackSizeForErrors);
+        }
+        altStackMem = new char[altStackSize]();
+    }
+
+    FatalConditionHandler::~FatalConditionHandler() {
+        delete[] altStackMem;
+        // We signal that another instance can be constructed by zeroing
+        // out the pointer.
+        altStackMem = nullptr;
+    }
+
+    void FatalConditionHandler::engage_platform() {
         stack_t sigStack;
         sigStack.ss_sp = altStackMem;
-        sigStack.ss_size = sigStackSize;
+        sigStack.ss_size = altStackSize;
         sigStack.ss_flags = 0;
         sigaltstack(&sigStack, &oldSigStack);
         struct sigaction sa = { };
@@ -8759,59 +10950,20 @@ namespace Catch {
         }
     }
 
-    FatalConditionHandler::~FatalConditionHandler() {
-        reset();
-    }
-
-    void FatalConditionHandler::reset() {
-        if( isSet ) {
-            // Set signals back to previous values -- hopefully nobody overwrote them in the meantime
-            for( std::size_t i = 0; i < sizeof(signalDefs)/sizeof(SignalDefs); ++i ) {
-                sigaction(signalDefs[i].id, &oldSigActions[i], nullptr);
-            }
-            // Return the old stack
-            sigaltstack(&oldSigStack, nullptr);
-            isSet = false;
-        }
-    }
-
-    bool FatalConditionHandler::isSet = false;
-    struct sigaction FatalConditionHandler::oldSigActions[sizeof(signalDefs)/sizeof(SignalDefs)] = {};
-    stack_t FatalConditionHandler::oldSigStack = {};
-    char FatalConditionHandler::altStackMem[sigStackSize] = {};
-
-} // namespace Catch
-
-#else
-
-namespace Catch {
-    void FatalConditionHandler::reset() {}
-}
-
-#endif // signals/SEH handling
-
 #if defined(__GNUC__)
 #    pragma GCC diagnostic pop
 #endif
-// end catch_fatal_condition.cpp
-// start catch_generators.cpp
-
-// start catch_random_number_generator.h
-
-#include <algorithm>
-#include <random>
-
-namespace Catch {
 
-    struct IConfig;
+    void FatalConditionHandler::disengage_platform() {
+        restorePreviousSignalHandlers();
+    }
 
-    std::mt19937& rng();
-    void seedRng( IConfig const& config );
-    unsigned int rngSeed();
+} // end namespace Catch
 
-}
+#endif // CATCH_CONFIG_POSIX_SIGNALS
+// end catch_fatal_condition.cpp
+// start catch_generators.cpp
 
-// end catch_random_number_generator.h
 #include <limits>
 #include <set>
 
@@ -8827,8 +10979,8 @@ namespace Generators {
 
     GeneratorUntypedBase::~GeneratorUntypedBase() {}
 
-    auto acquireGeneratorTracker( SourceLineInfo const& lineInfo ) -> IGeneratorTracker& {
-        return getResultCapture().acquireGeneratorTracker( lineInfo );
+    auto acquireGeneratorTracker( StringRef generatorName, SourceLineInfo const& lineInfo ) -> IGeneratorTracker& {
+        return getResultCapture().acquireGeneratorTracker( generatorName, lineInfo );
     }
 
 } // namespace Generators
@@ -8884,10 +11036,16 @@ namespace Catch {
 
         void noMatchingTestCases( std::string const& spec ) override;
 
+        void reportInvalidArguments(std::string const&arg) override;
+
         static std::set<Verbosity> getSupportedVerbosities();
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+        void benchmarkPreparing(std::string const& name) override;
         void benchmarkStarting( BenchmarkInfo const& benchmarkInfo ) override;
-        void benchmarkEnded( BenchmarkStats const& benchmarkStats ) override;
+        void benchmarkEnded( BenchmarkStats<> const& benchmarkStats ) override;
+        void benchmarkFailed(std::string const&) override;
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
         void testRunStarting( TestRunInfo const& testRunInfo ) override;
         void testGroupStarting( GroupInfo const& groupInfo ) override;
@@ -9097,7 +11255,7 @@ namespace Catch {
 namespace Catch {
 
     std::size_t listTests( Config const& config ) {
-        TestSpec testSpec = config.testSpec();
+        TestSpec const& testSpec = config.testSpec();
         if( config.hasTestFilters() )
             Catch::cout() << "Matching test cases:\n";
         else {
@@ -9131,7 +11289,7 @@ namespace Catch {
     }
 
     std::size_t listTestsNamesOnly( Config const& config ) {
-        TestSpec testSpec = config.testSpec();
+        TestSpec const& testSpec = config.testSpec();
         std::size_t matchedTests = 0;
         std::vector<TestCase> matchedTestCases = filterTests( getAllTestCasesSorted( config ), testSpec, config );
         for( auto const& testCaseInfo : matchedTestCases ) {
@@ -9153,14 +11311,23 @@ namespace Catch {
     }
 
     std::string TagInfo::all() const {
-        std::string out;
-        for( auto const& spelling : spellings )
-            out += "[" + spelling + "]";
+        size_t size = 0;
+        for (auto const& spelling : spellings) {
+            // Add 2 for the brackes
+            size += spelling.size() + 2;
+        }
+
+        std::string out; out.reserve(size);
+        for (auto const& spelling : spellings) {
+            out += '[';
+            out += spelling;
+            out += ']';
+        }
         return out;
     }
 
     std::size_t listTags( Config const& config ) {
-        TestSpec testSpec = config.testSpec();
+        TestSpec const& testSpec = config.testSpec();
         if( config.hasTestFilters() )
             Catch::cout() << "Tags for matching test cases:\n";
         else {
@@ -9254,6 +11421,29 @@ using Matchers::Impl::MatcherBase;
 
 } // namespace Catch
 // end catch_matchers.cpp
+// start catch_matchers_exception.cpp
+
+namespace Catch {
+namespace Matchers {
+namespace Exception {
+
+bool ExceptionMessageMatcher::match(std::exception const& ex) const {
+    return ex.what() == m_message;
+}
+
+std::string ExceptionMessageMatcher::describe() const {
+    return "exception message matches \"" + m_message + "\"";
+}
+
+}
+Exception::ExceptionMessageMatcher Message(std::string const& message) {
+    return Exception::ExceptionMessageMatcher(message);
+}
+
+// namespace Exception
+} // namespace Matchers
+} // namespace Catch
+// end catch_matchers_exception.cpp
 // start catch_matchers_floating.cpp
 
 // start catch_polyfills.hpp
@@ -9282,74 +11472,101 @@ namespace Catch {
 } // end namespace Catch
 
 // end catch_to_string.hpp
+#include <algorithm>
+#include <cmath>
 #include <cstdlib>
 #include <cstdint>
 #include <cstring>
+#include <sstream>
+#include <type_traits>
+#include <iomanip>
+#include <limits>
 
 namespace Catch {
-namespace Matchers {
-namespace Floating {
-enum class FloatingPointKind : uint8_t {
-    Float,
-    Double
-};
-}
-}
-}
-
 namespace {
 
-template <typename T>
-struct Converter;
-
-template <>
-struct Converter<float> {
-    static_assert(sizeof(float) == sizeof(int32_t), "Important ULP matcher assumption violated");
-    Converter(float f) {
+    int32_t convert(float f) {
+        static_assert(sizeof(float) == sizeof(int32_t), "Important ULP matcher assumption violated");
+        int32_t i;
         std::memcpy(&i, &f, sizeof(f));
+        return i;
     }
-    int32_t i;
-};
 
-template <>
-struct Converter<double> {
-    static_assert(sizeof(double) == sizeof(int64_t), "Important ULP matcher assumption violated");
-    Converter(double d) {
+    int64_t convert(double d) {
+        static_assert(sizeof(double) == sizeof(int64_t), "Important ULP matcher assumption violated");
+        int64_t i;
         std::memcpy(&i, &d, sizeof(d));
+        return i;
     }
-    int64_t i;
-};
 
-template <typename T>
-auto convert(T t) -> Converter<T> {
-    return Converter<T>(t);
-}
+    template <typename FP>
+    bool almostEqualUlps(FP lhs, FP rhs, uint64_t maxUlpDiff) {
+        // Comparison with NaN should always be false.
+        // This way we can rule it out before getting into the ugly details
+        if (Catch::isnan(lhs) || Catch::isnan(rhs)) {
+            return false;
+        }
 
-template <typename FP>
-bool almostEqualUlps(FP lhs, FP rhs, int maxUlpDiff) {
-    // Comparison with NaN should always be false.
-    // This way we can rule it out before getting into the ugly details
-    if (Catch::isnan(lhs) || Catch::isnan(rhs)) {
-        return false;
+        auto lc = convert(lhs);
+        auto rc = convert(rhs);
+
+        if ((lc < 0) != (rc < 0)) {
+            // Potentially we can have +0 and -0
+            return lhs == rhs;
+        }
+
+        // static cast as a workaround for IBM XLC
+        auto ulpDiff = std::abs(static_cast<FP>(lc - rc));
+        return static_cast<uint64_t>(ulpDiff) <= maxUlpDiff;
+    }
+
+#if defined(CATCH_CONFIG_GLOBAL_NEXTAFTER)
+
+    float nextafter(float x, float y) {
+        return ::nextafterf(x, y);
+    }
+
+    double nextafter(double x, double y) {
+        return ::nextafter(x, y);
     }
 
-    auto lc = convert(lhs);
-    auto rc = convert(rhs);
+#endif // ^^^ CATCH_CONFIG_GLOBAL_NEXTAFTER ^^^
 
-    if ((lc.i < 0) != (rc.i < 0)) {
-        // Potentially we can have +0 and -0
-        return lhs == rhs;
+template <typename FP>
+FP step(FP start, FP direction, uint64_t steps) {
+    for (uint64_t i = 0; i < steps; ++i) {
+#if defined(CATCH_CONFIG_GLOBAL_NEXTAFTER)
+        start = Catch::nextafter(start, direction);
+#else
+        start = std::nextafter(start, direction);
+#endif
     }
+    return start;
+}
 
-    auto ulpDiff = std::abs(lc.i - rc.i);
-    return ulpDiff <= maxUlpDiff;
+// Performs equivalent check of std::fabs(lhs - rhs) <= margin
+// But without the subtraction to allow for INFINITY in comparison
+bool marginComparison(double lhs, double rhs, double margin) {
+    return (lhs + margin >= rhs) && (rhs + margin >= lhs);
 }
 
+template <typename FloatingPoint>
+void write(std::ostream& out, FloatingPoint num) {
+    out << std::scientific
+        << std::setprecision(std::numeric_limits<FloatingPoint>::max_digits10 - 1)
+        << num;
 }
 
-namespace Catch {
+} // end anonymous namespace
+
 namespace Matchers {
 namespace Floating {
+
+    enum class FloatingPointKind : uint8_t {
+        Float,
+        Double
+    };
+
     WithinAbsMatcher::WithinAbsMatcher(double target, double margin)
         :m_target{ target }, m_margin{ margin } {
         CATCH_ENFORCE(margin >= 0, "Invalid margin: " << margin << '.'
@@ -9366,10 +11583,11 @@ namespace Floating {
         return "is within " + ::Catch::Detail::stringify(m_margin) + " of " + ::Catch::Detail::stringify(m_target);
     }
 
-    WithinUlpsMatcher::WithinUlpsMatcher(double target, int ulps, FloatingPointKind baseType)
+    WithinUlpsMatcher::WithinUlpsMatcher(double target, uint64_t ulps, FloatingPointKind baseType)
         :m_target{ target }, m_ulps{ ulps }, m_type{ baseType } {
-        CATCH_ENFORCE(ulps >= 0, "Invalid ULP setting: " << ulps << '.'
-            << " ULPs have to be non-negative.");
+        CATCH_ENFORCE(m_type == FloatingPointKind::Double
+                   || m_ulps < (std::numeric_limits<uint32_t>::max)(),
+            "Provided ULP is impossibly large for a float comparison.");
     }
 
 #if defined(__clang__)
@@ -9394,16 +11612,59 @@ namespace Floating {
 #endif
 
     std::string WithinUlpsMatcher::describe() const {
-        return "is within " + Catch::to_string(m_ulps) + " ULPs of " + ::Catch::Detail::stringify(m_target) + ((m_type == FloatingPointKind::Float)? "f" : "");
+        std::stringstream ret;
+
+        ret << "is within " << m_ulps << " ULPs of ";
+
+        if (m_type == FloatingPointKind::Float) {
+            write(ret, static_cast<float>(m_target));
+            ret << 'f';
+        } else {
+            write(ret, m_target);
+        }
+
+        ret << " ([";
+        if (m_type == FloatingPointKind::Double) {
+            write(ret, step(m_target, static_cast<double>(-INFINITY), m_ulps));
+            ret << ", ";
+            write(ret, step(m_target, static_cast<double>( INFINITY), m_ulps));
+        } else {
+            // We have to cast INFINITY to float because of MinGW, see #1782
+            write(ret, step(static_cast<float>(m_target), static_cast<float>(-INFINITY), m_ulps));
+            ret << ", ";
+            write(ret, step(static_cast<float>(m_target), static_cast<float>( INFINITY), m_ulps));
+        }
+        ret << "])";
+
+        return ret.str();
+    }
+
+    WithinRelMatcher::WithinRelMatcher(double target, double epsilon):
+        m_target(target),
+        m_epsilon(epsilon){
+        CATCH_ENFORCE(m_epsilon >= 0., "Relative comparison with epsilon <  0 does not make sense.");
+        CATCH_ENFORCE(m_epsilon  < 1., "Relative comparison with epsilon >= 1 does not make sense.");
+    }
+
+    bool WithinRelMatcher::match(double const& matchee) const {
+        const auto relMargin = m_epsilon * (std::max)(std::fabs(matchee), std::fabs(m_target));
+        return marginComparison(matchee, m_target,
+                                std::isinf(relMargin)? 0 : relMargin);
+    }
+
+    std::string WithinRelMatcher::describe() const {
+        Catch::ReusableStringStream sstr;
+        sstr << "and " << m_target << " are within " << m_epsilon * 100. << "% of each other";
+        return sstr.str();
     }
 
 }// namespace Floating
 
-Floating::WithinUlpsMatcher WithinULP(double target, int maxUlpDiff) {
+Floating::WithinUlpsMatcher WithinULP(double target, uint64_t maxUlpDiff) {
     return Floating::WithinUlpsMatcher(target, maxUlpDiff, Floating::FloatingPointKind::Double);
 }
 
-Floating::WithinUlpsMatcher WithinULP(float target, int maxUlpDiff) {
+Floating::WithinUlpsMatcher WithinULP(float target, uint64_t maxUlpDiff) {
     return Floating::WithinUlpsMatcher(target, maxUlpDiff, Floating::FloatingPointKind::Float);
 }
 
@@ -9411,9 +11672,24 @@ Floating::WithinAbsMatcher WithinAbs(double target, double margin) {
     return Floating::WithinAbsMatcher(target, margin);
 }
 
+Floating::WithinRelMatcher WithinRel(double target, double eps) {
+    return Floating::WithinRelMatcher(target, eps);
+}
+
+Floating::WithinRelMatcher WithinRel(double target) {
+    return Floating::WithinRelMatcher(target, std::numeric_limits<double>::epsilon() * 100);
+}
+
+Floating::WithinRelMatcher WithinRel(float target, float eps) {
+    return Floating::WithinRelMatcher(target, eps);
+}
+
+Floating::WithinRelMatcher WithinRel(float target) {
+    return Floating::WithinRelMatcher(target, std::numeric_limits<float>::epsilon() * 100);
+}
+
 } // namespace Matchers
 } // namespace Catch
-
 // end catch_matchers_floating.cpp
 // start catch_matchers_generic.cpp
 
@@ -9591,14 +11867,23 @@ namespace Catch {
 
     Capturer::Capturer( StringRef macroName, SourceLineInfo const& lineInfo, ResultWas::OfType resultType, StringRef names ) {
         auto trimmed = [&] (size_t start, size_t end) {
-            while (names[start] == ',' || isspace(names[start])) {
+            while (names[start] == ',' || isspace(static_cast<unsigned char>(names[start]))) {
                 ++start;
             }
-            while (names[end] == ',' || isspace(names[end])) {
+            while (names[end] == ',' || isspace(static_cast<unsigned char>(names[end]))) {
                 --end;
             }
             return names.substr(start, end - start + 1);
         };
+        auto skipq = [&] (size_t start, char quote) {
+            for (auto i = start + 1; i < names.size() ; ++i) {
+                if (names[i] == quote)
+                    return i;
+                if (names[i] == '\\')
+                    ++i;
+            }
+            CATCH_INTERNAL_ERROR("CAPTURE parsing encountered unmatched quote");
+        };
 
         size_t start = 0;
         std::stack<char> openings;
@@ -9619,18 +11904,22 @@ namespace Catch {
 //           case '>':
                 openings.pop();
                 break;
+            case '"':
+            case '\'':
+                pos = skipq(pos, c);
+                break;
             case ',':
-                if (start != pos && openings.size() == 0) {
+                if (start != pos && openings.empty()) {
                     m_messages.emplace_back(macroName, lineInfo, resultType);
-                    m_messages.back().message = trimmed(start, pos);
+                    m_messages.back().message = static_cast<std::string>(trimmed(start, pos));
                     m_messages.back().message += " := ";
                     start = pos;
                 }
             }
         }
-        assert(openings.size() == 0 && "Mismatched openings");
+        assert(openings.empty() && "Mismatched openings");
         m_messages.emplace_back(macroName, lineInfo, resultType);
-        m_messages.back().message = trimmed(start, names.size() - 1);
+        m_messages.back().message = static_cast<std::string>(trimmed(start, names.size() - 1));
         m_messages.back().message += " := ";
     }
     Capturer::~Capturer() {
@@ -9816,12 +12105,12 @@ namespace Catch {
         if (tmpnam_s(m_buffer)) {
             CATCH_RUNTIME_ERROR("Could not get a temp filename");
         }
-        if (fopen_s(&m_file, m_buffer, "w")) {
+        if (fopen_s(&m_file, m_buffer, "w+")) {
             char buffer[100];
             if (strerror_s(buffer, errno)) {
                 CATCH_RUNTIME_ERROR("Could not translate errno to a string");
             }
-            CATCH_RUNTIME_ERROR("Coul dnot open the temp file: '" << m_buffer << "' because: " << buffer);
+            CATCH_RUNTIME_ERROR("Could not open the temp file: '" << m_buffer << "' because: " << buffer);
         }
     }
 #else
@@ -9924,20 +12213,61 @@ namespace Catch {
 
 namespace Catch {
 
-    std::mt19937& rng() {
-        static std::mt19937 s_rng;
-        return s_rng;
+namespace {
+
+#if defined(_MSC_VER)
+#pragma warning(push)
+#pragma warning(disable:4146) // we negate uint32 during the rotate
+#endif
+        // Safe rotr implementation thanks to John Regehr
+        uint32_t rotate_right(uint32_t val, uint32_t count) {
+            const uint32_t mask = 31;
+            count &= mask;
+            return (val >> count) | (val << (-count & mask));
+        }
+
+#if defined(_MSC_VER)
+#pragma warning(pop)
+#endif
+
+}
+
+    SimplePcg32::SimplePcg32(result_type seed_) {
+        seed(seed_);
+    }
+
+    void SimplePcg32::seed(result_type seed_) {
+        m_state = 0;
+        (*this)();
+        m_state += seed_;
+        (*this)();
     }
 
-    void seedRng( IConfig const& config ) {
-        if( config.rngSeed() != 0 ) {
-            std::srand( config.rngSeed() );
-            rng().seed( config.rngSeed() );
+    void SimplePcg32::discard(uint64_t skip) {
+        // We could implement this to run in O(log n) steps, but this
+        // should suffice for our use case.
+        for (uint64_t s = 0; s < skip; ++s) {
+            static_cast<void>((*this)());
         }
     }
 
-    unsigned int rngSeed() {
-        return getCurrentContext().getConfig()->rngSeed();
+    SimplePcg32::result_type SimplePcg32::operator()() {
+        // prepare the output value
+        const uint32_t xorshifted = static_cast<uint32_t>(((m_state >> 18u) ^ m_state) >> 27u);
+        const auto output = rotate_right(xorshifted, m_state >> 59u);
+
+        // advance state
+        m_state = m_state * 6364136223846793005ULL + s_inc;
+
+        return output;
+    }
+
+    bool operator==(SimplePcg32 const& lhs, SimplePcg32 const& rhs) {
+        return lhs.m_state == rhs.m_state;
+    }
+
+    bool operator!=(SimplePcg32 const& lhs, SimplePcg32 const& rhs) {
+        return lhs.m_state != rhs.m_state;
     }
 }
 // end catch_random_number_generator.cpp
@@ -9956,6 +12286,8 @@ namespace Catch {
     struct IConfig;
 
     std::vector<TestCase> sortTests( IConfig const& config, std::vector<TestCase> const& unsortedTestCases );
+
+    bool isThrowSafe( TestCase const& testCase, IConfig const& config );
     bool matchTest( TestCase const& testCase, TestSpec const& testSpec, IConfig const& config );
 
     void enforceNoDuplicateTestCases( std::vector<TestCase> const& functions );
@@ -10068,11 +12400,13 @@ namespace Catch {
 namespace Catch {
 
     class StartupExceptionRegistry {
+#if !defined(CATCH_CONFIG_DISABLE_EXCEPTIONS)
     public:
         void add(std::exception_ptr const& exception) noexcept;
         std::vector<std::exception_ptr> const& getExceptions() const noexcept;
     private:
         std::vector<std::exception_ptr> m_exceptions;
+#endif
     };
 
 } // end namespace Catch
@@ -10155,7 +12489,14 @@ namespace Catch {
                 m_tagAliasRegistry.add( alias, tag, lineInfo );
             }
             void registerStartupException() noexcept override {
+#if !defined(CATCH_CONFIG_DISABLE_EXCEPTIONS)
                 m_exceptionRegistry.add(std::current_exception());
+#else
+                CATCH_INTERNAL_ERROR("Attempted to register active exception under CATCH_CONFIG_DISABLE_EXCEPTIONS!");
+#endif
+            }
+            IMutableEnumValuesRegistry& getMutableEnumValuesRegistry() override {
+                return m_enumValuesRegistry;
             }
 
         private:
@@ -10164,6 +12505,7 @@ namespace Catch {
             ExceptionTranslatorRegistry m_exceptionTranslatorRegistry;
             TagAliasRegistry m_tagAliasRegistry;
             StartupExceptionRegistry m_exceptionRegistry;
+            Detail::EnumValuesRegistry m_enumValuesRegistry;
         };
     }
 
@@ -10255,17 +12597,32 @@ namespace Catch {
                 std::shared_ptr<GeneratorTracker> tracker;
 
                 ITracker& currentTracker = ctx.currentTracker();
-                if( TestCaseTracking::ITrackerPtr childTracker = currentTracker.findChild( nameAndLocation ) ) {
+                // Under specific circumstances, the generator we want
+                // to acquire is also the current tracker. If this is
+                // the case, we have to avoid looking through current
+                // tracker's children, and instead return the current
+                // tracker.
+                // A case where this check is important is e.g.
+                //     for (int i = 0; i < 5; ++i) {
+                //         int n = GENERATE(1, 2);
+                //     }
+                //
+                // without it, the code above creates 5 nested generators.
+                if (currentTracker.nameAndLocation() == nameAndLocation) {
+                    auto thisTracker = currentTracker.parent().findChild(nameAndLocation);
+                    assert(thisTracker);
+                    assert(thisTracker->isGeneratorTracker());
+                    tracker = std::static_pointer_cast<GeneratorTracker>(thisTracker);
+                } else if ( TestCaseTracking::ITrackerPtr childTracker = currentTracker.findChild( nameAndLocation ) ) {
                     assert( childTracker );
                     assert( childTracker->isGeneratorTracker() );
                     tracker = std::static_pointer_cast<GeneratorTracker>( childTracker );
-                }
-                else {
+                } else {
                     tracker = std::make_shared<GeneratorTracker>( nameAndLocation, ctx, &currentTracker );
                     currentTracker.addChild( tracker );
                 }
 
-                if( !ctx.completedCycle() && !tracker->isComplete() ) {
+                if( !tracker->isComplete() ) {
                     tracker->open();
                 }
 
@@ -10279,8 +12636,68 @@ namespace Catch {
             }
             void close() override {
                 TrackerBase::close();
-                // Generator interface only finds out if it has another item on atual move
-                if (m_runState == CompletedSuccessfully && m_generator->next()) {
+                // If a generator has a child (it is followed by a section)
+                // and none of its children have started, then we must wait
+                // until later to start consuming its values.
+                // This catches cases where `GENERATE` is placed between two
+                // `SECTION`s.
+                // **The check for m_children.empty cannot be removed**.
+                // doing so would break `GENERATE` _not_ followed by `SECTION`s.
+                const bool should_wait_for_child = [&]() {
+                    // No children -> nobody to wait for
+                    if ( m_children.empty() ) {
+                        return false;
+                    }
+                    // If at least one child started executing, don't wait
+                    if ( std::find_if(
+                             m_children.begin(),
+                             m_children.end(),
+                             []( TestCaseTracking::ITrackerPtr tracker ) {
+                                 return tracker->hasStarted();
+                             } ) != m_children.end() ) {
+                        return false;
+                    }
+
+                    // No children have started. We need to check if they _can_
+                    // start, and thus we should wait for them, or they cannot
+                    // start (due to filters), and we shouldn't wait for them
+                    auto* parent = m_parent;
+                    // This is safe: there is always at least one section
+                    // tracker in a test case tracking tree
+                    while ( !parent->isSectionTracker() ) {
+                        parent = &( parent->parent() );
+                    }
+                    assert( parent &&
+                            "Missing root (test case) level section" );
+
+                    auto const& parentSection =
+                        static_cast<SectionTracker&>( *parent );
+                    auto const& filters = parentSection.getFilters();
+                    // No filters -> no restrictions on running sections
+                    if ( filters.empty() ) {
+                        return true;
+                    }
+
+                    for ( auto const& child : m_children ) {
+                        if ( child->isSectionTracker() &&
+                             std::find( filters.begin(),
+                                        filters.end(),
+                                        static_cast<SectionTracker&>( *child )
+                                            .trimmedName() ) !=
+                                 filters.end() ) {
+                            return true;
+                        }
+                    }
+                    return false;
+                }();
+
+                // This check is a bit tricky, because m_generator->next()
+                // has a side-effect, where it consumes generator's current
+                // value, but we do not want to invoke the side-effect if
+                // this generator is still waiting for any child to start.
+                if ( should_wait_for_child ||
+                     ( m_runState == CompletedSuccessfully &&
+                       m_generator->next() ) ) {
                     m_children.clear();
                     m_runState = Executing;
                 }
@@ -10416,10 +12833,10 @@ namespace Catch {
 
         return true;
     }
-    auto RunContext::acquireGeneratorTracker( SourceLineInfo const& lineInfo ) -> IGeneratorTracker& {
+    auto RunContext::acquireGeneratorTracker( StringRef generatorName, SourceLineInfo const& lineInfo ) -> IGeneratorTracker& {
         using namespace Generators;
-        GeneratorTracker& tracker = GeneratorTracker::acquire( m_trackerContext, TestCaseTracking::NameAndLocation( "generator", lineInfo ) );
-        assert( tracker.isOpen() );
+        GeneratorTracker& tracker = GeneratorTracker::acquire(m_trackerContext,
+                                                              TestCaseTracking::NameAndLocation( static_cast<std::string>(generatorName), lineInfo ) );
         m_lastAssertionInfo.lineInfo = lineInfo;
         return tracker;
     }
@@ -10459,12 +12876,21 @@ namespace Catch {
 
         m_unfinishedSections.push_back(endInfo);
     }
+
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+    void RunContext::benchmarkPreparing(std::string const& name) {
+        m_reporter->benchmarkPreparing(name);
+    }
     void RunContext::benchmarkStarting( BenchmarkInfo const& info ) {
         m_reporter->benchmarkStarting( info );
     }
-    void RunContext::benchmarkEnded( BenchmarkStats const& stats ) {
+    void RunContext::benchmarkEnded( BenchmarkStats<> const& stats ) {
         m_reporter->benchmarkEnded( stats );
     }
+    void RunContext::benchmarkFailed(std::string const & error) {
+        m_reporter->benchmarkFailed(error);
+    }
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
     void RunContext::pushScopedMessage(MessageInfo const & message) {
         m_messages.push_back(message);
@@ -10499,7 +12925,7 @@ namespace Catch {
         // Don't rebuild the result -- the stringification itself can cause more fatal errors
         // Instead, fake a result data.
         AssertionResultData tempResult( ResultWas::FatalErrorCondition, { false } );
-        tempResult.message = message;
+        tempResult.message = static_cast<std::string>(message);
         AssertionResult result(m_lastAssertionInfo, tempResult);
 
         assertionEnded(result);
@@ -10597,9 +13023,8 @@ namespace Catch {
     }
 
     void RunContext::invokeActiveTestCase() {
-        FatalConditionHandler fatalConditionHandler; // Handle signals
+        FatalConditionHandlerGuard _(&m_fatalConditionhandler);
         m_activeTestCase->invoke();
-        fatalConditionHandler.reset();
     }
 
     void RunContext::handleUnfinishedSections() {
@@ -10662,7 +13087,7 @@ namespace Catch {
         m_lastAssertionInfo = info;
 
         AssertionResultData data( resultType, LazyExpression( false ) );
-        data.message = message;
+        data.message = static_cast<std::string>(message);
         AssertionResult assertionResult{ m_lastAssertionInfo, data };
         assertionEnded( assertionResult );
         if( !assertionResult.isOk() )
@@ -10725,6 +13150,18 @@ namespace Catch {
         else
             CATCH_INTERNAL_ERROR("No result capture instance");
     }
+
+    void seedRng(IConfig const& config) {
+        if (config.rngSeed() != 0) {
+            std::srand(config.rngSeed());
+            rng().seed(config.rngSeed());
+        }
+    }
+
+    unsigned int rngSeed() {
+        return getCurrentContext().getConfig()->rngSeed();
+    }
+
 }
 // end catch_run_context.cpp
 // start catch_section.cpp
@@ -10786,7 +13223,7 @@ namespace Catch {
         void libIdentify();
 
         int applyCommandLine( int argc, char const * const * argv );
-    #if defined(CATCH_CONFIG_WCHAR) && defined(WIN32) && defined(UNICODE)
+    #if defined(CATCH_CONFIG_WCHAR) && defined(_WIN32) && defined(UNICODE)
         int applyCommandLine( int argc, wchar_t const * const * argv );
     #endif
 
@@ -10853,6 +13290,8 @@ namespace Catch {
 // end catch_version.h
 #include <cstdlib>
 #include <iomanip>
+#include <set>
+#include <iterator>
 
 namespace Catch {
 
@@ -10886,42 +13325,61 @@ namespace Catch {
             return ret;
         }
 
-        Catch::Totals runTests(std::shared_ptr<Config> const& config) {
-            auto reporter = makeReporter(config);
-
-            RunContext context(config, std::move(reporter));
-
-            Totals totals;
-
-            context.testGroupStarting(config->name(), 1, 1);
-
-            TestSpec testSpec = config->testSpec();
-
-            auto const& allTestCases = getAllTestCasesSorted(*config);
-            for (auto const& testCase : allTestCases) {
-                if (!context.aborting() && matchTest(testCase, testSpec, *config))
-                    totals += context.runTest(testCase);
-                else
-                    context.reporter().skipTest(testCase);
+        class TestGroup {
+        public:
+            explicit TestGroup(std::shared_ptr<Config> const& config)
+            : m_config{config}
+            , m_context{config, makeReporter(config)}
+            {
+                auto const& allTestCases = getAllTestCasesSorted(*m_config);
+                m_matches = m_config->testSpec().matchesByFilter(allTestCases, *m_config);
+                auto const& invalidArgs = m_config->testSpec().getInvalidArgs();
+
+                if (m_matches.empty() && invalidArgs.empty()) {
+                    for (auto const& test : allTestCases)
+                        if (!test.isHidden())
+                            m_tests.emplace(&test);
+                } else {
+                    for (auto const& match : m_matches)
+                        m_tests.insert(match.tests.begin(), match.tests.end());
+                }
             }
 
-            if (config->warnAboutNoTests() && totals.testCases.total() == 0) {
-                ReusableStringStream testConfig;
+            Totals execute() {
+                auto const& invalidArgs = m_config->testSpec().getInvalidArgs();
+                Totals totals;
+                m_context.testGroupStarting(m_config->name(), 1, 1);
+                for (auto const& testCase : m_tests) {
+                    if (!m_context.aborting())
+                        totals += m_context.runTest(*testCase);
+                    else
+                        m_context.reporter().skipTest(*testCase);
+                }
+
+                for (auto const& match : m_matches) {
+                    if (match.tests.empty()) {
+                        m_context.reporter().noMatchingTestCases(match.name);
+                        totals.error = -1;
+                    }
+                }
 
-                bool first = true;
-                for (const auto& input : config->getTestsOrTags()) {
-                    if (!first) { testConfig << ' '; }
-                    first = false;
-                    testConfig << input;
+                if (!invalidArgs.empty()) {
+                    for (auto const& invalidArg: invalidArgs)
+                         m_context.reporter().reportInvalidArguments(invalidArg);
                 }
 
-                context.reporter().noMatchingTestCases(testConfig.str());
-                totals.error = -1;
+                m_context.testGroupEnded(m_config->name(), totals, 1, 1);
+                return totals;
             }
 
-            context.testGroupEnded(config->name(), totals, 1, 1);
-            return totals;
-        }
+        private:
+            using Tests = std::set<TestCase const*>;
+
+            std::shared_ptr<Config> m_config;
+            RunContext m_context;
+            Tests m_tests;
+            TestSpec::Matches m_matches;
+        };
 
         void applyFilenamesAsTags(Catch::IConfig const& config) {
             auto& tests = const_cast<std::vector<TestCase>&>(getAllTestCasesSorted(config));
@@ -10934,6 +13392,10 @@ namespace Catch {
                     filename.erase(0, lastSlash);
                     filename[0] = '#';
                 }
+                else
+                {
+                    filename.insert(0, "#");
+                }
 
                 auto lastDot = filename.find_last_of('.');
                 if (lastDot != std::string::npos) {
@@ -10958,6 +13420,9 @@ namespace Catch {
 #if !defined(CATCH_CONFIG_DISABLE_EXCEPTIONS)
         const auto& exceptions = getRegistryHub().getStartupExceptionRegistry().getExceptions();
         if ( !exceptions.empty() ) {
+            config();
+            getCurrentMutableContext().setConfig(m_config);
+
             m_startupExceptions = true;
             Colour colourGuard( Colour::Red );
             Catch::cerr() << "Errors occurred during startup!" << '\n';
@@ -10987,7 +13452,7 @@ namespace Catch {
     }
     void Session::libIdentify() {
         Catch::cout()
-                << std::left << std::setw(16) << "description: " << "A Catch test executable\n"
+                << std::left << std::setw(16) << "description: " << "A Catch2 test executable\n"
                 << std::left << std::setw(16) << "category: " << "testframework\n"
                 << std::left << std::setw(16) << "framework: " << "Catch Test\n"
                 << std::left << std::setw(16) << "version: " << libraryVersion() << std::endl;
@@ -11018,17 +13483,17 @@ namespace Catch {
         return 0;
     }
 
-#if defined(CATCH_CONFIG_WCHAR) && defined(WIN32) && defined(UNICODE)
+#if defined(CATCH_CONFIG_WCHAR) && defined(_WIN32) && defined(UNICODE)
     int Session::applyCommandLine( int argc, wchar_t const * const * argv ) {
 
         char **utf8Argv = new char *[ argc ];
 
         for ( int i = 0; i < argc; ++i ) {
-            int bufSize = WideCharToMultiByte( CP_UTF8, 0, argv[i], -1, NULL, 0, NULL, NULL );
+            int bufSize = WideCharToMultiByte( CP_UTF8, 0, argv[i], -1, nullptr, 0, nullptr, nullptr );
 
             utf8Argv[ i ] = new char[ bufSize ];
 
-            WideCharToMultiByte( CP_UTF8, 0, argv[i], -1, utf8Argv[i], bufSize, NULL, NULL );
+            WideCharToMultiByte( CP_UTF8, 0, argv[i], -1, utf8Argv[i], bufSize, nullptr, nullptr );
         }
 
         int returnCode = applyCommandLine( argc, utf8Argv );
@@ -11095,7 +13560,12 @@ namespace Catch {
             if( Option<std::size_t> listed = list( m_config ) )
                 return static_cast<int>( *listed );
 
-            auto totals = runTests( m_config );
+            TestGroup tests { m_config };
+            auto const totals = tests.execute();
+
+            if( m_config->warnAboutNoTests() && totals.error == -1 )
+                return 2;
+
             // Note that on unices only the lower 8 bits are usually used, clamping
             // the return value to 255 prevents false negative when some multiple
             // of 256 tests has failed
@@ -11143,6 +13613,7 @@ namespace Catch {
 // end catch_singletons.cpp
 // start catch_startup_exception_registry.cpp
 
+#if !defined(CATCH_CONFIG_DISABLE_EXCEPTIONS)
 namespace Catch {
 void StartupExceptionRegistry::add( std::exception_ptr const& exception ) noexcept {
         CATCH_TRY {
@@ -11158,6 +13629,7 @@ void StartupExceptionRegistry::add( std::exception_ptr const& exception ) noexce
     }
 
 } // end namespace Catch
+#endif
 // end catch_startup_exception_registry.cpp
 // start catch_stream.cpp
 
@@ -11172,7 +13644,7 @@ namespace Catch {
 
     Catch::IStream::~IStream() = default;
 
-    namespace detail { namespace {
+    namespace Detail { namespace {
         template<typename WriterF, std::size_t bufferSize=256>
         class StreamBufImpl : public std::streambuf {
             char data[bufferSize];
@@ -11271,15 +13743,15 @@ namespace Catch {
 
     auto makeStream( StringRef const &filename ) -> IStream const* {
         if( filename.empty() )
-            return new detail::CoutStream();
+            return new Detail::CoutStream();
         else if( filename[0] == '%' ) {
             if( filename == "%debug" )
-                return new detail::DebugOutStream();
+                return new Detail::DebugOutStream();
             else
                 CATCH_ERROR( "Unrecognised stream: '" << filename << "'" );
         }
         else
-            return new detail::FileStream( filename );
+            return new Detail::FileStream( filename );
     }
 
     // This class encapsulates the idea of a pool of ostringstreams that can be reused.
@@ -11336,12 +13808,13 @@ namespace Catch {
 #include <ostream>
 #include <cstring>
 #include <cctype>
+#include <vector>
 
 namespace Catch {
 
     namespace {
         char toLowerCh(char c) {
-            return static_cast<char>( std::tolower( c ) );
+            return static_cast<char>( std::tolower( static_cast<unsigned char>(c) ) );
         }
     }
 
@@ -11376,6 +13849,18 @@ namespace Catch {
         return start != std::string::npos ? str.substr( start, 1+end-start ) : std::string();
     }
 
+    StringRef trim(StringRef ref) {
+        const auto is_ws = [](char c) {
+            return c == ' ' || c == '\t' || c == '\n' || c == '\r';
+        };
+        size_t real_begin = 0;
+        while (real_begin < ref.size() && is_ws(ref[real_begin])) { ++real_begin; }
+        size_t real_end = ref.size();
+        while (real_end > real_begin && is_ws(ref[real_end - 1])) { --real_end; }
+
+        return ref.substr(real_begin, real_end - real_begin);
+    }
+
     bool replaceInPlace( std::string& str, std::string const& replaceThis, std::string const& withThis ) {
         bool replaced = false;
         std::size_t i = str.find( replaceThis );
@@ -11390,6 +13875,21 @@ namespace Catch {
         return replaced;
     }
 
+    std::vector<StringRef> splitStringRef( StringRef str, char delimiter ) {
+        std::vector<StringRef> subStrings;
+        std::size_t start = 0;
+        for(std::size_t pos = 0; pos < str.size(); ++pos ) {
+            if( str[pos] == delimiter ) {
+                if( pos - start > 1 )
+                    subStrings.push_back( str.substr( start, pos-start ) );
+                start = pos+1;
+            }
+        }
+        if( start < str.size() )
+            subStrings.push_back( str.substr( start, str.size()-start ) );
+        return subStrings;
+    }
+
     pluralise::pluralise( std::size_t count, std::string const& label )
     :   m_count( count ),
         m_label( label )
@@ -11406,123 +13906,46 @@ namespace Catch {
 // end catch_string_manip.cpp
 // start catch_stringref.cpp
 
-#if defined(__clang__)
-#    pragma clang diagnostic push
-#    pragma clang diagnostic ignored "-Wexit-time-destructors"
-#endif
-
+#include <algorithm>
 #include <ostream>
 #include <cstring>
 #include <cstdint>
 
-namespace {
-    const uint32_t byte_2_lead = 0xC0;
-    const uint32_t byte_3_lead = 0xE0;
-    const uint32_t byte_4_lead = 0xF0;
-}
-
 namespace Catch {
     StringRef::StringRef( char const* rawChars ) noexcept
     : StringRef( rawChars, static_cast<StringRef::size_type>(std::strlen(rawChars) ) )
     {}
 
-    StringRef::operator std::string() const {
-        return std::string( m_start, m_size );
-    }
-
-    void StringRef::swap( StringRef& other ) noexcept {
-        std::swap( m_start, other.m_start );
-        std::swap( m_size, other.m_size );
-        std::swap( m_data, other.m_data );
-    }
-
     auto StringRef::c_str() const -> char const* {
-        if( isSubstring() )
-           const_cast<StringRef*>( this )->takeOwnership();
+        CATCH_ENFORCE(isNullTerminated(), "Called StringRef::c_str() on a non-null-terminated instance");
         return m_start;
     }
-    auto StringRef::currentData() const noexcept -> char const* {
+    auto StringRef::data() const noexcept -> char const* {
         return m_start;
     }
 
-    auto StringRef::isOwned() const noexcept -> bool {
-        return m_data != nullptr;
-    }
-    auto StringRef::isSubstring() const noexcept -> bool {
-        return m_start[m_size] != '\0';
-    }
-
-    void StringRef::takeOwnership() {
-        if( !isOwned() ) {
-            m_data = new char[m_size+1];
-            memcpy( m_data, m_start, m_size );
-            m_data[m_size] = '\0';
-            m_start = m_data;
-        }
-    }
     auto StringRef::substr( size_type start, size_type size ) const noexcept -> StringRef {
-        if( start < m_size )
-            return StringRef( m_start+start, size );
-        else
+        if (start < m_size) {
+            return StringRef(m_start + start, (std::min)(m_size - start, size));
+        } else {
             return StringRef();
-    }
-    auto StringRef::operator == ( StringRef const& other ) const noexcept -> bool {
-        return
-            size() == other.size() &&
-            (std::strncmp( m_start, other.m_start, size() ) == 0);
-    }
-    auto StringRef::operator != ( StringRef const& other ) const noexcept -> bool {
-        return !operator==( other );
-    }
-
-    auto StringRef::operator[](size_type index) const noexcept -> char {
-        return m_start[index];
-    }
-
-    auto StringRef::numberOfCharacters() const noexcept -> size_type {
-        size_type noChars = m_size;
-        // Make adjustments for uft encodings
-        for( size_type i=0; i < m_size; ++i ) {
-            char c = m_start[i];
-            if( ( c & byte_2_lead ) == byte_2_lead ) {
-                noChars--;
-                if (( c & byte_3_lead ) == byte_3_lead )
-                    noChars--;
-                if( ( c & byte_4_lead ) == byte_4_lead )
-                    noChars--;
-            }
         }
-        return noChars;
     }
-
-    auto operator + ( StringRef const& lhs, StringRef const& rhs ) -> std::string {
-        std::string str;
-        str.reserve( lhs.size() + rhs.size() );
-        str += lhs;
-        str += rhs;
-        return str;
-    }
-    auto operator + ( StringRef const& lhs, const char* rhs ) -> std::string {
-        return std::string( lhs ) + std::string( rhs );
-    }
-    auto operator + ( char const* lhs, StringRef const& rhs ) -> std::string {
-        return std::string( lhs ) + std::string( rhs );
+    auto StringRef::operator == ( StringRef const& other ) const noexcept -> bool {
+        return m_size == other.m_size
+            && (std::memcmp( m_start, other.m_start, m_size ) == 0);
     }
 
     auto operator << ( std::ostream& os, StringRef const& str ) -> std::ostream& {
-        return os.write(str.currentData(), str.size());
+        return os.write(str.data(), str.size());
     }
 
     auto operator+=( std::string& lhs, StringRef const& rhs ) -> std::string& {
-        lhs.append(rhs.currentData(), rhs.size());
+        lhs.append(rhs.data(), rhs.size());
         return lhs;
     }
 
 } // namespace Catch
-
-#if defined(__clang__)
-#    pragma clang diagnostic pop
-#endif
 // end catch_stringref.cpp
 // start catch_tag_alias.cpp
 
@@ -11625,7 +14048,7 @@ namespace Catch {
         void enforceNotReservedTag( std::string const& tag, SourceLineInfo const& _lineInfo ) {
             CATCH_ENFORCE( !isReservedTag(tag),
                           "Tag name: [" << tag << "] is not allowed.\n"
-                          << "Tag names starting with non alpha-numeric characters are reserved\n"
+                          << "Tag names starting with non alphanumeric characters are reserved\n"
                           << _lineInfo );
         }
     }
@@ -11641,8 +14064,7 @@ namespace Catch {
         std::vector<std::string> tags;
         std::string desc, tag;
         bool inTag = false;
-        std::string _descOrTags = nameAndTags.tags;
-        for (char c : _descOrTags) {
+        for (char c : nameAndTags.tags) {
             if( !inTag ) {
                 if( c == '[' )
                     inTag = true;
@@ -11657,6 +14079,12 @@ namespace Catch {
                     else if( prop == TestCaseInfo::None )
                         enforceNotReservedTag( tag, _lineInfo );
 
+                    // Merged hide tags like `[.approvals]` should be added as
+                    // `[.][approvals]`. The `[.]` is added at later point, so
+                    // we only strip the prefix
+                    if (startsWith(tag, '.') && tag.size() > 1) {
+                        tag.erase(0, 1);
+                    }
                     tags.push_back( tag );
                     tag.clear();
                     inTag = false;
@@ -11666,10 +14094,11 @@ namespace Catch {
             }
         }
         if( isHidden ) {
-            tags.push_back( "." );
+            // Add all "hidden" tags to make them behave identically
+            tags.insert( tags.end(), { ".", "!hide" } );
         }
 
-        TestCaseInfo info( nameAndTags.name, _className, desc, tags, _lineInfo );
+        TestCaseInfo info( static_cast<std::string>(nameAndTags.name), _className, desc, tags, _lineInfo );
         return TestCase( _testCase, std::move(info) );
     }
 
@@ -11761,30 +14190,89 @@ namespace Catch {
 // end catch_test_case_info.cpp
 // start catch_test_case_registry_impl.cpp
 
+#include <algorithm>
 #include <sstream>
 
 namespace Catch {
 
-    std::vector<TestCase> sortTests( IConfig const& config, std::vector<TestCase> const& unsortedTestCases ) {
+    namespace {
+        struct TestHasher {
+            using hash_t = uint64_t;
+
+            explicit TestHasher( hash_t hashSuffix ):
+                m_hashSuffix{ hashSuffix } {}
+
+            uint32_t operator()( TestCase const& t ) const {
+                // FNV-1a hash with multiplication fold.
+                const hash_t prime = 1099511628211u;
+                hash_t hash = 14695981039346656037u;
+                for ( const char c : t.name ) {
+                    hash ^= c;
+                    hash *= prime;
+                }
+                hash ^= m_hashSuffix;
+                hash *= prime;
+                const uint32_t low{ static_cast<uint32_t>( hash ) };
+                const uint32_t high{ static_cast<uint32_t>( hash >> 32 ) };
+                return low * high;
+            }
 
-        std::vector<TestCase> sorted = unsortedTestCases;
+        private:
+            hash_t m_hashSuffix;
+        };
+    } // end unnamed namespace
 
+    std::vector<TestCase> sortTests( IConfig const& config, std::vector<TestCase> const& unsortedTestCases ) {
         switch( config.runOrder() ) {
-            case RunTests::InLexicographicalOrder:
-                std::sort( sorted.begin(), sorted.end() );
-                break;
-            case RunTests::InRandomOrder:
-                seedRng( config );
-                std::shuffle( sorted.begin(), sorted.end(), rng() );
-                break;
             case RunTests::InDeclarationOrder:
                 // already in declaration order
                 break;
+
+            case RunTests::InLexicographicalOrder: {
+                std::vector<TestCase> sorted = unsortedTestCases;
+                std::sort( sorted.begin(), sorted.end() );
+                return sorted;
+            }
+
+            case RunTests::InRandomOrder: {
+                seedRng( config );
+                TestHasher h{ config.rngSeed() };
+
+                using hashedTest = std::pair<TestHasher::hash_t, TestCase const*>;
+                std::vector<hashedTest> indexed_tests;
+                indexed_tests.reserve( unsortedTestCases.size() );
+
+                for (auto const& testCase : unsortedTestCases) {
+                    indexed_tests.emplace_back(h(testCase), &testCase);
+                }
+
+                std::sort(indexed_tests.begin(), indexed_tests.end(),
+                          [](hashedTest const& lhs, hashedTest const& rhs) {
+                          if (lhs.first == rhs.first) {
+                              return lhs.second->name < rhs.second->name;
+                          }
+                          return lhs.first < rhs.first;
+                });
+
+                std::vector<TestCase> sorted;
+                sorted.reserve( indexed_tests.size() );
+
+                for (auto const& hashed : indexed_tests) {
+                    sorted.emplace_back(*hashed.second);
+                }
+
+                return sorted;
+            }
         }
-        return sorted;
+        return unsortedTestCases;
     }
+
+    bool isThrowSafe( TestCase const& testCase, IConfig const& config ) {
+        return !testCase.throws() || config.allowThrows();
+    }
+
     bool matchTest( TestCase const& testCase, TestSpec const& testSpec, IConfig const& config ) {
-        return testSpec.matches( testCase ) && ( config.allowThrows() || !testCase.throws() );
+        return testSpec.matches( testCase ) && isThrowSafe( testCase, config );
     }
 
     void enforceNoDuplicateTestCases( std::vector<TestCase> const& functions ) {
@@ -11801,9 +14289,12 @@ namespace Catch {
     std::vector<TestCase> filterTests( std::vector<TestCase> const& testCases, TestSpec const& testSpec, IConfig const& config ) {
         std::vector<TestCase> filtered;
         filtered.reserve( testCases.size() );
-        for( auto const& testCase : testCases )
-            if( matchTest( testCase, testSpec, config ) )
-                filtered.push_back( testCase );
+        for (auto const& testCase : testCases) {
+            if ((!testSpec.hasFilters() && !testCase.isHidden()) ||
+                (testSpec.hasFilters() && matchTest(testCase, testSpec, config))) {
+                filtered.push_back(testCase);
+            }
+        }
         return filtered;
     }
     std::vector<TestCase> const& getAllTestCasesSorted( IConfig const& config ) {
@@ -11842,7 +14333,7 @@ namespace Catch {
     }
 
     std::string extractClassName( StringRef const& classOrQualifiedMethodName ) {
-        std::string className = classOrQualifiedMethodName;
+        std::string className(classOrQualifiedMethodName);
         if( startsWith( className, '&' ) )
         {
             std::size_t lastColons = className.rfind( "::" );
@@ -11879,11 +14370,6 @@ namespace TestCaseTracking {
 
     ITracker::~ITracker() = default;
 
-    TrackerContext& TrackerContext::instance() {
-        static TrackerContext s_instance;
-        return s_instance;
-    }
-
     ITracker& TrackerContext::startRun() {
         m_rootTracker = std::make_shared<SectionTracker>( NameAndLocation( "{root}", CATCH_INTERNAL_LINEINFO ), *this, nullptr );
         m_currentTracker = nullptr;
@@ -11915,15 +14401,12 @@ namespace TestCaseTracking {
         m_currentTracker = tracker;
     }
 
-    TrackerBase::TrackerBase( NameAndLocation const& nameAndLocation, TrackerContext& ctx, ITracker* parent )
-    :   m_nameAndLocation( nameAndLocation ),
+    TrackerBase::TrackerBase( NameAndLocation const& nameAndLocation, TrackerContext& ctx, ITracker* parent ):
+        ITracker(nameAndLocation),
         m_ctx( ctx ),
         m_parent( parent )
     {}
 
-    NameAndLocation const& TrackerBase::nameAndLocation() const {
-        return m_nameAndLocation;
-    }
     bool TrackerBase::isComplete() const {
         return m_runState == CompletedSuccessfully || m_runState == Failed;
     }
@@ -11989,7 +14472,7 @@ namespace TestCaseTracking {
                 m_runState = CompletedSuccessfully;
                 break;
             case ExecutingChildren:
-                if( m_children.empty() || m_children.back()->isComplete() )
+                if( std::all_of(m_children.begin(), m_children.end(), [](ITrackerPtr const& t){ return t->isComplete(); }) )
                     m_runState = CompletedSuccessfully;
                 break;
 
@@ -12024,7 +14507,8 @@ namespace TestCaseTracking {
     }
 
     SectionTracker::SectionTracker( NameAndLocation const& nameAndLocation, TrackerContext& ctx, ITracker* parent )
-    :   TrackerBase( nameAndLocation, ctx, parent )
+    :   TrackerBase( nameAndLocation, ctx, parent ),
+        m_trimmed_name(trim(nameAndLocation.name))
     {
         if( parent ) {
             while( !parent->isSectionTracker() )
@@ -12038,12 +14522,12 @@ namespace TestCaseTracking {
     bool SectionTracker::isComplete() const {
         bool complete = true;
 
-        if ((m_filters.empty() || m_filters[0] == "") ||
-             std::find(m_filters.begin(), m_filters.end(),
-                       m_nameAndLocation.name) != m_filters.end())
+        if (m_filters.empty()
+            || m_filters[0] == ""
+            || std::find(m_filters.begin(), m_filters.end(), m_trimmed_name) != m_filters.end()) {
             complete = TrackerBase::isComplete();
+        }
         return complete;
-
     }
 
     bool SectionTracker::isSectionTracker() const { return true; }
@@ -12067,20 +14551,29 @@ namespace TestCaseTracking {
     }
 
     void SectionTracker::tryOpen() {
-        if( !isComplete() && (m_filters.empty() || m_filters[0].empty() ||  m_filters[0] == m_nameAndLocation.name ) )
+        if( !isComplete() )
             open();
     }
 
     void SectionTracker::addInitialFilters( std::vector<std::string> const& filters ) {
         if( !filters.empty() ) {
-            m_filters.push_back(""); // Root - should never be consulted
-            m_filters.push_back(""); // Test Case - not a section filter
+            m_filters.reserve( m_filters.size() + filters.size() + 2 );
+            m_filters.emplace_back(""); // Root - should never be consulted
+            m_filters.emplace_back(""); // Test Case - not a section filter
             m_filters.insert( m_filters.end(), filters.begin(), filters.end() );
         }
     }
     void SectionTracker::addNextFilters( std::vector<std::string> const& filters ) {
         if( filters.size() > 1 )
-            m_filters.insert( m_filters.end(), ++filters.begin(), filters.end() );
+            m_filters.insert( m_filters.end(), filters.begin()+1, filters.end() );
+    }
+
+    std::vector<std::string> const& SectionTracker::getFilters() const {
+        return m_filters;
+    }
+
+    std::string const& SectionTracker::trimmedName() const {
+        return m_trimmed_name;
     }
 
 } // namespace TestCaseTracking
@@ -12132,47 +14625,81 @@ namespace Catch {
 
 namespace Catch {
 
+    TestSpec::Pattern::Pattern( std::string const& name )
+    : m_name( name )
+    {}
+
     TestSpec::Pattern::~Pattern() = default;
-    TestSpec::NamePattern::~NamePattern() = default;
-    TestSpec::TagPattern::~TagPattern() = default;
-    TestSpec::ExcludedPattern::~ExcludedPattern() = default;
 
-    TestSpec::NamePattern::NamePattern( std::string const& name )
-    : m_wildcardPattern( toLower( name ), CaseSensitive::No )
+    std::string const& TestSpec::Pattern::name() const {
+        return m_name;
+    }
+
+    TestSpec::NamePattern::NamePattern( std::string const& name, std::string const& filterString )
+    : Pattern( filterString )
+    , m_wildcardPattern( toLower( name ), CaseSensitive::No )
     {}
+
     bool TestSpec::NamePattern::matches( TestCaseInfo const& testCase ) const {
-        return m_wildcardPattern.matches( toLower( testCase.name ) );
+        return m_wildcardPattern.matches( testCase.name );
     }
 
-    TestSpec::TagPattern::TagPattern( std::string const& tag ) : m_tag( toLower( tag ) ) {}
+    TestSpec::TagPattern::TagPattern( std::string const& tag, std::string const& filterString )
+    : Pattern( filterString )
+    , m_tag( toLower( tag ) )
+    {}
+
     bool TestSpec::TagPattern::matches( TestCaseInfo const& testCase ) const {
         return std::find(begin(testCase.lcaseTags),
                          end(testCase.lcaseTags),
                          m_tag) != end(testCase.lcaseTags);
     }
 
-    TestSpec::ExcludedPattern::ExcludedPattern( PatternPtr const& underlyingPattern ) : m_underlyingPattern( underlyingPattern ) {}
-    bool TestSpec::ExcludedPattern::matches( TestCaseInfo const& testCase ) const { return !m_underlyingPattern->matches( testCase ); }
+    TestSpec::ExcludedPattern::ExcludedPattern( PatternPtr const& underlyingPattern )
+    : Pattern( underlyingPattern->name() )
+    , m_underlyingPattern( underlyingPattern )
+    {}
+
+    bool TestSpec::ExcludedPattern::matches( TestCaseInfo const& testCase ) const {
+        return !m_underlyingPattern->matches( testCase );
+    }
 
     bool TestSpec::Filter::matches( TestCaseInfo const& testCase ) const {
-        // All patterns in a filter must match for the filter to be a match
-        for( auto const& pattern : m_patterns ) {
-            if( !pattern->matches( testCase ) )
-                return false;
-        }
-        return true;
+        return std::all_of( m_patterns.begin(), m_patterns.end(), [&]( PatternPtr const& p ){ return p->matches( testCase ); } );
+    }
+
+    std::string TestSpec::Filter::name() const {
+        std::string name;
+        for( auto const& p : m_patterns )
+            name += p->name();
+        return name;
     }
 
     bool TestSpec::hasFilters() const {
         return !m_filters.empty();
     }
+
     bool TestSpec::matches( TestCaseInfo const& testCase ) const {
-        // A TestSpec matches if any filter matches
-        for( auto const& filter : m_filters )
-            if( filter.matches( testCase ) )
-                return true;
-        return false;
+        return std::any_of( m_filters.begin(), m_filters.end(), [&]( Filter const& f ){ return f.matches( testCase ); } );
+    }
+
+    TestSpec::Matches TestSpec::matchesByFilter( std::vector<TestCase> const& testCases, IConfig const& config ) const
+    {
+        Matches matches( m_filters.size() );
+        std::transform( m_filters.begin(), m_filters.end(), matches.begin(), [&]( Filter const& filter ){
+            std::vector<TestCase const*> currentMatches;
+            for( auto const& test : testCases )
+                if( isThrowSafe( test, config ) && filter.matches( test ) )
+                    currentMatches.emplace_back( &test );
+            return FilterMatch{ filter.name(), currentMatches };
+        } );
+        return matches;
+    }
+
+    const TestSpec::vectorStrings& TestSpec::getInvalidArgs() const{
+        return  (m_invalidArgs);
     }
+
 }
 // end catch_test_spec.cpp
 // start catch_test_spec_parser.cpp
@@ -12184,64 +14711,136 @@ namespace Catch {
     TestSpecParser& TestSpecParser::parse( std::string const& arg ) {
         m_mode = None;
         m_exclusion = false;
-        m_start = std::string::npos;
         m_arg = m_tagAliases->expandAliases( arg );
         m_escapeChars.clear();
+        m_substring.reserve(m_arg.size());
+        m_patternName.reserve(m_arg.size());
+        m_realPatternPos = 0;
+
         for( m_pos = 0; m_pos < m_arg.size(); ++m_pos )
-            visitChar( m_arg[m_pos] );
-        if( m_mode == Name )
-            addPattern<TestSpec::NamePattern>();
+          //if visitChar fails
+           if( !visitChar( m_arg[m_pos] ) ){
+               m_testSpec.m_invalidArgs.push_back(arg);
+               break;
+           }
+        endMode();
         return *this;
     }
     TestSpec TestSpecParser::testSpec() {
         addFilter();
         return m_testSpec;
     }
+    bool TestSpecParser::visitChar( char c ) {
+        if( (m_mode != EscapedName) && (c == '\\') ) {
+            escape();
+            addCharToPattern(c);
+            return true;
+        }else if((m_mode != EscapedName) && (c == ',') )  {
+            return separate();
+        }
 
-    void TestSpecParser::visitChar( char c ) {
-        if( m_mode == None ) {
-            switch( c ) {
-            case ' ': return;
-            case '~': m_exclusion = true; return;
-            case '[': return startNewMode( Tag, ++m_pos );
-            case '"': return startNewMode( QuotedName, ++m_pos );
-            case '\\': return escape();
-            default: startNewMode( Name, m_pos ); break;
-            }
+        switch( m_mode ) {
+        case None:
+            if( processNoneChar( c ) )
+                return true;
+            break;
+        case Name:
+            processNameChar( c );
+            break;
+        case EscapedName:
+            endMode();
+            addCharToPattern(c);
+            return true;
+        default:
+        case Tag:
+        case QuotedName:
+            if( processOtherChar( c ) )
+                return true;
+            break;
         }
-        if( m_mode == Name ) {
-            if( c == ',' ) {
-                addPattern<TestSpec::NamePattern>();
-                addFilter();
-            }
-            else if( c == '[' ) {
-                if( subString() == "exclude:" )
-                    m_exclusion = true;
-                else
-                    addPattern<TestSpec::NamePattern>();
-                startNewMode( Tag, ++m_pos );
-            }
-            else if( c == '\\' )
-                escape();
+
+        m_substring += c;
+        if( !isControlChar( c ) ) {
+            m_patternName += c;
+            m_realPatternPos++;
+        }
+        return true;
+    }
+    // Two of the processing methods return true to signal the caller to return
+    // without adding the given character to the current pattern strings
+    bool TestSpecParser::processNoneChar( char c ) {
+        switch( c ) {
+        case ' ':
+            return true;
+        case '~':
+            m_exclusion = true;
+            return false;
+        case '[':
+            startNewMode( Tag );
+            return false;
+        case '"':
+            startNewMode( QuotedName );
+            return false;
+        default:
+            startNewMode( Name );
+            return false;
         }
-        else if( m_mode == EscapedName )
-            m_mode = Name;
-        else if( m_mode == QuotedName && c == '"' )
-            addPattern<TestSpec::NamePattern>();
-        else if( m_mode == Tag && c == ']' )
-            addPattern<TestSpec::TagPattern>();
     }
-    void TestSpecParser::startNewMode( Mode mode, std::size_t start ) {
+    void TestSpecParser::processNameChar( char c ) {
+        if( c == '[' ) {
+            if( m_substring == "exclude:" )
+                m_exclusion = true;
+            else
+                endMode();
+            startNewMode( Tag );
+        }
+    }
+    bool TestSpecParser::processOtherChar( char c ) {
+        if( !isControlChar( c ) )
+            return false;
+        m_substring += c;
+        endMode();
+        return true;
+    }
+    void TestSpecParser::startNewMode( Mode mode ) {
         m_mode = mode;
-        m_start = start;
+    }
+    void TestSpecParser::endMode() {
+        switch( m_mode ) {
+        case Name:
+        case QuotedName:
+            return addNamePattern();
+        case Tag:
+            return addTagPattern();
+        case EscapedName:
+            revertBackToLastMode();
+            return;
+        case None:
+        default:
+            return startNewMode( None );
+        }
     }
     void TestSpecParser::escape() {
-        if( m_mode == None )
-            m_start = m_pos;
+        saveLastMode();
         m_mode = EscapedName;
-        m_escapeChars.push_back( m_pos );
+        m_escapeChars.push_back(m_realPatternPos);
+    }
+    bool TestSpecParser::isControlChar( char c ) const {
+        switch( m_mode ) {
+            default:
+                return false;
+            case None:
+                return c == '~';
+            case Name:
+                return c == '[';
+            case EscapedName:
+                return true;
+            case QuotedName:
+                return c == '"';
+            case Tag:
+                return c == '[' || c == ']';
+        }
     }
-    std::string TestSpecParser::subString() const { return m_arg.substr( m_start, m_pos - m_start ); }
 
     void TestSpecParser::addFilter() {
         if( !m_currentFilter.m_patterns.empty() ) {
@@ -12250,6 +14849,86 @@ namespace Catch {
         }
     }
 
+    void TestSpecParser::saveLastMode() {
+      lastMode = m_mode;
+    }
+
+    void TestSpecParser::revertBackToLastMode() {
+      m_mode = lastMode;
+    }
+
+    bool TestSpecParser::separate() {
+      if( (m_mode==QuotedName) || (m_mode==Tag) ){
+         //invalid argument, signal failure to previous scope.
+         m_mode = None;
+         m_pos = m_arg.size();
+         m_substring.clear();
+         m_patternName.clear();
+         m_realPatternPos = 0;
+         return false;
+      }
+      endMode();
+      addFilter();
+      return true; //success
+    }
+
+    std::string TestSpecParser::preprocessPattern() {
+        std::string token = m_patternName;
+        for (std::size_t i = 0; i < m_escapeChars.size(); ++i)
+            token = token.substr(0, m_escapeChars[i] - i) + token.substr(m_escapeChars[i] - i + 1);
+        m_escapeChars.clear();
+        if (startsWith(token, "exclude:")) {
+            m_exclusion = true;
+            token = token.substr(8);
+        }
+
+        m_patternName.clear();
+        m_realPatternPos = 0;
+
+        return token;
+    }
+
+    void TestSpecParser::addNamePattern() {
+        auto token = preprocessPattern();
+
+        if (!token.empty()) {
+            TestSpec::PatternPtr pattern = std::make_shared<TestSpec::NamePattern>(token, m_substring);
+            if (m_exclusion)
+                pattern = std::make_shared<TestSpec::ExcludedPattern>(pattern);
+            m_currentFilter.m_patterns.push_back(pattern);
+        }
+        m_substring.clear();
+        m_exclusion = false;
+        m_mode = None;
+    }
+
+    void TestSpecParser::addTagPattern() {
+        auto token = preprocessPattern();
+
+        if (!token.empty()) {
+            // If the tag pattern is the "hide and tag" shorthand (e.g. [.foo])
+            // we have to create a separate hide tag and shorten the real one
+            if (token.size() > 1 && token[0] == '.') {
+                token.erase(token.begin());
+                TestSpec::PatternPtr pattern = std::make_shared<TestSpec::TagPattern>(".", m_substring);
+                if (m_exclusion) {
+                    pattern = std::make_shared<TestSpec::ExcludedPattern>(pattern);
+                }
+                m_currentFilter.m_patterns.push_back(pattern);
+            }
+
+            TestSpec::PatternPtr pattern = std::make_shared<TestSpec::TagPattern>(token, m_substring);
+
+            if (m_exclusion) {
+                pattern = std::make_shared<TestSpec::ExcludedPattern>(pattern);
+            }
+            m_currentFilter.m_patterns.push_back(pattern);
+        }
+        m_substring.clear();
+        m_exclusion = false;
+        m_mode = None;
+    }
+
     TestSpec parseTestSpec( std::string const& arg ) {
         return TestSpecParser( ITagAliasRegistry::get() ).parse( arg ).testSpec();
     }
@@ -12351,13 +15030,11 @@ namespace Detail {
             enum Arch { Big, Little };
 
             static Arch which() {
-                union _{
-                    int asInt;
-                    char asChar[sizeof (int)];
-                } u;
-
-                u.asInt = 1;
-                return ( u.asChar[sizeof(int)-1] == 1 ) ? Big : Little;
+                int one = 1;
+                // If the lowest byte we read is non-zero, we can assume
+                // that little endian format is used.
+                auto value = *reinterpret_cast<char*>(&one);
+                return value ? Little : Big;
             }
         };
     }
@@ -12481,6 +15158,13 @@ std::string StringMaker<wchar_t *>::convert(wchar_t * str) {
 }
 #endif
 
+#if defined(CATCH_CONFIG_CPP17_BYTE)
+#include <cstddef>
+std::string StringMaker<std::byte>::convert(std::byte value) {
+    return ::Catch::Detail::stringify(std::to_integer<unsigned long long>(value));
+}
+#endif // defined(CATCH_CONFIG_CPP17_BYTE)
+
 std::string StringMaker<int>::convert(int value) {
     return ::Catch::Detail::stringify(static_cast<long long>(value));
 }
@@ -12543,11 +15227,16 @@ std::string StringMaker<std::nullptr_t>::convert(std::nullptr_t) {
     return "nullptr";
 }
 
+int StringMaker<float>::precision = 5;
+
 std::string StringMaker<float>::convert(float value) {
-    return fpToString(value, 5) + 'f';
+    return fpToString(value, precision) + 'f';
 }
+
+int StringMaker<double>::precision = 10;
+
 std::string StringMaker<double>::convert(double value) {
-    return fpToString(value, 10);
+    return fpToString(value, precision);
 }
 
 std::string ratio_string<std::atto>::symbol() { return "a"; }
@@ -12621,11 +15310,48 @@ namespace Catch {
 // end catch_totals.cpp
 // start catch_uncaught_exceptions.cpp
 
+// start catch_config_uncaught_exceptions.hpp
+
+//              Copyright Catch2 Authors
+// Distributed under the Boost Software License, Version 1.0.
+//   (See accompanying file LICENSE_1_0.txt or copy at
+//        https://www.boost.org/LICENSE_1_0.txt)
+
+// SPDX-License-Identifier: BSL-1.0
+
+#ifndef CATCH_CONFIG_UNCAUGHT_EXCEPTIONS_HPP
+#define CATCH_CONFIG_UNCAUGHT_EXCEPTIONS_HPP
+
+#if defined(_MSC_VER)
+#  if _MSC_VER >= 1900 // Visual Studio 2015 or newer
+#    define CATCH_INTERNAL_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS
+#  endif
+#endif
+
+#include <exception>
+
+#if defined(__cpp_lib_uncaught_exceptions) \
+    && !defined(CATCH_INTERNAL_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS)
+
+#  define CATCH_INTERNAL_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS
+#endif // __cpp_lib_uncaught_exceptions
+
+#if defined(CATCH_INTERNAL_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS) \
+    && !defined(CATCH_CONFIG_NO_CPP17_UNCAUGHT_EXCEPTIONS) \
+    && !defined(CATCH_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS)
+
+#  define CATCH_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS
+#endif
+
+#endif // CATCH_CONFIG_UNCAUGHT_EXCEPTIONS_HPP
+// end catch_config_uncaught_exceptions.hpp
 #include <exception>
 
 namespace Catch {
     bool uncaught_exceptions() {
-#if defined(CATCH_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS)
+#if defined(CATCH_CONFIG_DISABLE_EXCEPTIONS)
+        return false;
+#elif defined(CATCH_CONFIG_CPP17_UNCAUGHT_EXCEPTIONS)
         return std::uncaught_exceptions() > 0;
 #else
         return std::uncaught_exception();
@@ -12665,7 +15391,7 @@ namespace Catch {
     }
 
     Version const& libraryVersion() {
-        static Version version( 2, 7, 0, "", 0 );
+        static Version version( 2, 13, 9, "", 0 );
         return version;
     }
 
@@ -12673,14 +15399,12 @@ namespace Catch {
 // end catch_version.cpp
 // start catch_wildcard_pattern.cpp
 
-#include <sstream>
-
 namespace Catch {
 
     WildcardPattern::WildcardPattern( std::string const& pattern,
                                       CaseSensitive::Choice caseSensitivity )
     :   m_caseSensitivity( caseSensitivity ),
-        m_pattern( adjustCase( pattern ) )
+        m_pattern( normaliseString( pattern ) )
     {
         if( startsWith( m_pattern, '*' ) ) {
             m_pattern = m_pattern.substr( 1 );
@@ -12695,28 +15419,27 @@ namespace Catch {
     bool WildcardPattern::matches( std::string const& str ) const {
         switch( m_wildcard ) {
             case NoWildcard:
-                return m_pattern == adjustCase( str );
+                return m_pattern == normaliseString( str );
             case WildcardAtStart:
-                return endsWith( adjustCase( str ), m_pattern );
+                return endsWith( normaliseString( str ), m_pattern );
             case WildcardAtEnd:
-                return startsWith( adjustCase( str ), m_pattern );
+                return startsWith( normaliseString( str ), m_pattern );
             case WildcardAtBothEnds:
-                return contains( adjustCase( str ), m_pattern );
+                return contains( normaliseString( str ), m_pattern );
             default:
                 CATCH_INTERNAL_ERROR( "Unknown enum" );
         }
     }
 
-    std::string WildcardPattern::adjustCase( std::string const& str ) const {
-        return m_caseSensitivity == CaseSensitive::No ? toLower( str ) : str;
+    std::string WildcardPattern::normaliseString( std::string const& str ) const {
+        return trim( m_caseSensitivity == CaseSensitive::No ? toLower( str ) : str );
     }
 }
 // end catch_wildcard_pattern.cpp
 // start catch_xmlwriter.cpp
 
 #include <iomanip>
-
-using uchar = unsigned char;
+#include <type_traits>
 
 namespace Catch {
 
@@ -12756,8 +15479,30 @@ namespace {
         os.flags(f);
     }
 
+    bool shouldNewline(XmlFormatting fmt) {
+        return !!(static_cast<std::underlying_type<XmlFormatting>::type>(fmt & XmlFormatting::Newline));
+    }
+
+    bool shouldIndent(XmlFormatting fmt) {
+        return !!(static_cast<std::underlying_type<XmlFormatting>::type>(fmt & XmlFormatting::Indent));
+    }
+
 } // anonymous namespace
 
+    XmlFormatting operator | (XmlFormatting lhs, XmlFormatting rhs) {
+        return static_cast<XmlFormatting>(
+            static_cast<std::underlying_type<XmlFormatting>::type>(lhs) |
+            static_cast<std::underlying_type<XmlFormatting>::type>(rhs)
+        );
+    }
+
+    XmlFormatting operator & (XmlFormatting lhs, XmlFormatting rhs) {
+        return static_cast<XmlFormatting>(
+            static_cast<std::underlying_type<XmlFormatting>::type>(lhs) &
+            static_cast<std::underlying_type<XmlFormatting>::type>(rhs)
+        );
+    }
+
     XmlEncode::XmlEncode( std::string const& str, ForWhat forWhat )
     :   m_str( str ),
         m_forWhat( forWhat )
@@ -12768,7 +15513,7 @@ namespace {
         // (see: http://www.w3.org/TR/xml/#syntax)
 
         for( std::size_t idx = 0; idx < m_str.size(); ++ idx ) {
-            uchar c = m_str[idx];
+            unsigned char c = m_str[idx];
             switch (c) {
             case '<':   os << "&lt;"; break;
             case '&':   os << "&amp;"; break;
@@ -12828,7 +15573,7 @@ namespace {
                 bool valid = true;
                 uint32_t value = headerValue(c);
                 for (std::size_t n = 1; n < encBytes; ++n) {
-                    uchar nc = m_str[idx + n];
+                    unsigned char nc = m_str[idx + n];
                     valid &= ((nc & 0xC0) == 0x80);
                     value = (value << 6) | (nc & 0x3F);
                 }
@@ -12862,13 +15607,17 @@ namespace {
         return os;
     }
 
-    XmlWriter::ScopedElement::ScopedElement( XmlWriter* writer )
-    :   m_writer( writer )
+    XmlWriter::ScopedElement::ScopedElement( XmlWriter* writer, XmlFormatting fmt )
+    :   m_writer( writer ),
+        m_fmt(fmt)
     {}
 
     XmlWriter::ScopedElement::ScopedElement( ScopedElement&& other ) noexcept
-    :   m_writer( other.m_writer ){
+    :   m_writer( other.m_writer ),
+        m_fmt(other.m_fmt)
+    {
         other.m_writer = nullptr;
+        other.m_fmt = XmlFormatting::None;
     }
     XmlWriter::ScopedElement& XmlWriter::ScopedElement::operator=( ScopedElement&& other ) noexcept {
         if ( m_writer ) {
@@ -12876,16 +15625,19 @@ namespace {
         }
         m_writer = other.m_writer;
         other.m_writer = nullptr;
+        m_fmt = other.m_fmt;
+        other.m_fmt = XmlFormatting::None;
         return *this;
     }
 
     XmlWriter::ScopedElement::~ScopedElement() {
-        if( m_writer )
-            m_writer->endElement();
+        if (m_writer) {
+            m_writer->endElement(m_fmt);
+        }
     }
 
-    XmlWriter::ScopedElement& XmlWriter::ScopedElement::writeText( std::string const& text, bool indent ) {
-        m_writer->writeText( text, indent );
+    XmlWriter::ScopedElement& XmlWriter::ScopedElement::writeText( std::string const& text, XmlFormatting fmt ) {
+        m_writer->writeText( text, fmt );
         return *this;
     }
 
@@ -12895,37 +15647,47 @@ namespace {
     }
 
     XmlWriter::~XmlWriter() {
-        while( !m_tags.empty() )
+        while (!m_tags.empty()) {
             endElement();
+        }
+        newlineIfNecessary();
     }
 
-    XmlWriter& XmlWriter::startElement( std::string const& name ) {
+    XmlWriter& XmlWriter::startElement( std::string const& name, XmlFormatting fmt ) {
         ensureTagClosed();
         newlineIfNecessary();
-        m_os << m_indent << '<' << name;
+        if (shouldIndent(fmt)) {
+            m_os << m_indent;
+            m_indent += "  ";
+        }
+        m_os << '<' << name;
         m_tags.push_back( name );
-        m_indent += "  ";
         m_tagIsOpen = true;
+        applyFormatting(fmt);
         return *this;
     }
 
-    XmlWriter::ScopedElement XmlWriter::scopedElement( std::string const& name ) {
-        ScopedElement scoped( this );
-        startElement( name );
+    XmlWriter::ScopedElement XmlWriter::scopedElement( std::string const& name, XmlFormatting fmt ) {
+        ScopedElement scoped( this, fmt );
+        startElement( name, fmt );
         return scoped;
     }
 
-    XmlWriter& XmlWriter::endElement() {
-        newlineIfNecessary();
-        m_indent = m_indent.substr( 0, m_indent.size()-2 );
+    XmlWriter& XmlWriter::endElement(XmlFormatting fmt) {
+        m_indent = m_indent.substr(0, m_indent.size() - 2);
+
         if( m_tagIsOpen ) {
             m_os << "/>";
             m_tagIsOpen = false;
+        } else {
+            newlineIfNecessary();
+            if (shouldIndent(fmt)) {
+                m_os << m_indent;
+            }
+            m_os << "</" << m_tags.back() << ">";
         }
-        else {
-            m_os << m_indent << "</" << m_tags.back() << ">";
-        }
-        m_os << std::endl;
+        m_os << std::flush;
+        applyFormatting(fmt);
         m_tags.pop_back();
         return *this;
     }
@@ -12941,22 +15703,26 @@ namespace {
         return *this;
     }
 
-    XmlWriter& XmlWriter::writeText( std::string const& text, bool indent ) {
+    XmlWriter& XmlWriter::writeText( std::string const& text, XmlFormatting fmt) {
         if( !text.empty() ){
             bool tagWasOpen = m_tagIsOpen;
             ensureTagClosed();
-            if( tagWasOpen && indent )
+            if (tagWasOpen && shouldIndent(fmt)) {
                 m_os << m_indent;
+            }
             m_os << XmlEncode( text );
-            m_needsNewline = true;
+            applyFormatting(fmt);
         }
         return *this;
     }
 
-    XmlWriter& XmlWriter::writeComment( std::string const& text ) {
+    XmlWriter& XmlWriter::writeComment( std::string const& text, XmlFormatting fmt) {
         ensureTagClosed();
-        m_os << m_indent << "<!--" << text << "-->";
-        m_needsNewline = true;
+        if (shouldIndent(fmt)) {
+            m_os << m_indent;
+        }
+        m_os << "<!--" << text << "-->";
+        applyFormatting(fmt);
         return *this;
     }
 
@@ -12972,11 +15738,16 @@ namespace {
 
     void XmlWriter::ensureTagClosed() {
         if( m_tagIsOpen ) {
-            m_os << ">" << std::endl;
+            m_os << '>' << std::flush;
+            newlineIfNecessary();
             m_tagIsOpen = false;
         }
     }
 
+    void XmlWriter::applyFormatting(XmlFormatting fmt) {
+        m_needsNewline = shouldNewline(fmt);
+    }
+
     void XmlWriter::writeDeclaration() {
         m_os << "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n";
     }
@@ -13022,6 +15793,32 @@ namespace Catch {
         return std::string(buffer);
     }
 
+    bool shouldShowDuration( IConfig const& config, double duration ) {
+        if ( config.showDurations() == ShowDurations::Always ) {
+            return true;
+        }
+        if ( config.showDurations() == ShowDurations::Never ) {
+            return false;
+        }
+        const double min = config.minDuration();
+        return min >= 0 && duration >= min;
+    }
+
+    std::string serializeFilters( std::vector<std::string> const& container ) {
+        ReusableStringStream oss;
+        bool first = true;
+        for (auto&& filter : container)
+        {
+            if (!first)
+                oss << ' ';
+            else
+                first = false;
+
+            oss << filter;
+        }
+        return oss.str();
+    }
+
     TestEventListenerBase::TestEventListenerBase(ReporterConfig const & _config)
         :StreamingReporterBase(_config) {}
 
@@ -13237,24 +16034,25 @@ private:
         if (itMessage == messages.end())
             return;
 
-        // using messages.end() directly yields (or auto) compilation error:
-        std::vector<MessageInfo>::const_iterator itEnd = messages.end();
-        const std::size_t N = static_cast<std::size_t>(std::distance(itMessage, itEnd));
+        const auto itEnd = messages.cend();
+        const auto N = static_cast<std::size_t>(std::distance(itMessage, itEnd));
 
         {
             Colour colourGuard(colour);
             stream << " with " << pluralise(N, "message") << ':';
         }
 
-        for (; itMessage != itEnd; ) {
+        while (itMessage != itEnd) {
             // If this assertion is a warning ignore any INFO messages
             if (printInfoMessages || itMessage->type != ResultWas::Info) {
-                stream << " '" << itMessage->message << '\'';
-                if (++itMessage != itEnd) {
+                printMessage();
+                if (itMessage != itEnd) {
                     Colour colourGuard(dimColour());
                     stream << " and";
                 }
+                continue;
             }
+            ++itMessage;
         }
     }
 
@@ -13272,10 +16070,6 @@ private:
             return "Reports test results on a single line, suitable for IDEs";
         }
 
-        ReporterPreferences CompactReporter::getPreferences() const {
-            return m_reporterPrefs;
-        }
-
         void CompactReporter::noMatchingTestCases( std::string const& spec ) {
             stream << "No test cases matched '" << spec << '\'' << std::endl;
         }
@@ -13302,8 +16096,9 @@ private:
         }
 
         void CompactReporter::sectionEnded(SectionStats const& _sectionStats) {
-            if (m_config->showDurations() == ShowDurations::Always) {
-                stream << getFormattedDuration(_sectionStats.durationInSeconds) << " s: " << _sectionStats.sectionInfo.name << std::endl;
+            double dur = _sectionStats.durationInSeconds;
+            if ( shouldShowDuration( *m_config, dur ) ) {
+                stream << getFormattedDuration( dur ) << " s: " << _sectionStats.sectionInfo.name << std::endl;
             }
         }
 
@@ -13327,8 +16122,13 @@ private:
 #if defined(_MSC_VER)
 #pragma warning(push)
 #pragma warning(disable:4061) // Not all labels are EXPLICITLY handled in switch
- // Note that 4062 (not all labels are handled
- // and default is missing) is enabled
+ // Note that 4062 (not all labels are handled and default is missing) is enabled
+#endif
+
+#if defined(__clang__)
+#  pragma clang diagnostic push
+// For simplicity, benchmarking-only helpers are always enabled
+#  pragma clang diagnostic ignored "-Wunused-function"
 #endif
 
 namespace Catch {
@@ -13510,11 +16310,11 @@ class Duration {
     static const uint64_t s_nanosecondsInASecond = 1000 * s_nanosecondsInAMillisecond;
     static const uint64_t s_nanosecondsInAMinute = 60 * s_nanosecondsInASecond;
 
-    uint64_t m_inNanoseconds;
+    double m_inNanoseconds;
     Unit m_units;
 
 public:
-    explicit Duration(uint64_t inNanoseconds, Unit units = Unit::Auto)
+    explicit Duration(double inNanoseconds, Unit units = Unit::Auto)
         : m_inNanoseconds(inNanoseconds),
         m_units(units) {
         if (m_units == Unit::Auto) {
@@ -13543,7 +16343,7 @@ public:
         case Unit::Minutes:
             return m_inNanoseconds / static_cast<double>(s_nanosecondsInAMinute);
         default:
-            return static_cast<double>(m_inNanoseconds);
+            return m_inNanoseconds;
         }
     }
     auto unitsAsString() const -> std::string {
@@ -13564,7 +16364,7 @@ public:
 
     }
     friend auto operator << (std::ostream& os, Duration const& duration) -> std::ostream& {
-        return os << duration.value() << " " << duration.unitsAsString();
+        return os << duration.value() << ' ' << duration.unitsAsString();
     }
 };
 } // end anon namespace
@@ -13589,10 +16389,16 @@ public:
         if (!m_isOpen) {
             m_isOpen = true;
             *this << RowBreak();
-            for (auto const& info : m_columnInfos)
-                *this << info.name << ColumnBreak();
-            *this << RowBreak();
-            m_os << Catch::getLineOfChars<'-'>() << "\n";
+
+			Columns headerCols;
+			Spacer spacer(2);
+			for (auto const& info : m_columnInfos) {
+				headerCols += Column(info.name).width(static_cast<std::size_t>(info.width - 2));
+				headerCols += spacer;
+			}
+			m_os << headerCols << '\n';
+
+            m_os << Catch::getLineOfChars<'-'>() << '\n';
         }
     }
     void close() {
@@ -13611,30 +16417,29 @@ public:
 
     friend TablePrinter& operator << (TablePrinter& tp, ColumnBreak) {
         auto colStr = tp.m_oss.str();
-        // This takes account of utf8 encodings
-        auto strSize = Catch::StringRef(colStr).numberOfCharacters();
+        const auto strSize = colStr.size();
         tp.m_oss.str("");
         tp.open();
         if (tp.m_currentColumn == static_cast<int>(tp.m_columnInfos.size() - 1)) {
             tp.m_currentColumn = -1;
-            tp.m_os << "\n";
+            tp.m_os << '\n';
         }
         tp.m_currentColumn++;
 
         auto colInfo = tp.m_columnInfos[tp.m_currentColumn];
-        auto padding = (strSize + 2 < static_cast<std::size_t>(colInfo.width))
-            ? std::string(colInfo.width - (strSize + 2), ' ')
+        auto padding = (strSize + 1 < static_cast<std::size_t>(colInfo.width))
+            ? std::string(colInfo.width - (strSize + 1), ' ')
             : std::string();
         if (colInfo.justification == ColumnInfo::Left)
-            tp.m_os << colStr << padding << " ";
+            tp.m_os << colStr << padding << ' ';
         else
-            tp.m_os << padding << colStr << " ";
+            tp.m_os << padding << colStr << ' ';
         return tp;
     }
 
     friend TablePrinter& operator << (TablePrinter& tp, RowBreak) {
         if (tp.m_currentColumn > 0) {
-            tp.m_os << "\n";
+            tp.m_os << '\n';
             tp.m_currentColumn = -1;
         }
         return tp;
@@ -13644,12 +16449,26 @@ public:
 ConsoleReporter::ConsoleReporter(ReporterConfig const& config)
     : StreamingReporterBase(config),
     m_tablePrinter(new TablePrinter(config.stream(),
-    {
-        { "benchmark name", CATCH_CONFIG_CONSOLE_WIDTH - 32, ColumnInfo::Left },
-        { "iters", 8, ColumnInfo::Right },
-        { "elapsed ns", 14, ColumnInfo::Right },
-        { "average", 14, ColumnInfo::Right }
-    })) {}
+        [&config]() -> std::vector<ColumnInfo> {
+        if (config.fullConfig()->benchmarkNoAnalysis())
+        {
+            return{
+                { "benchmark name", CATCH_CONFIG_CONSOLE_WIDTH - 43, ColumnInfo::Left },
+                { "     samples", 14, ColumnInfo::Right },
+                { "  iterations", 14, ColumnInfo::Right },
+                { "        mean", 14, ColumnInfo::Right }
+            };
+        }
+        else
+        {
+            return{
+                { "benchmark name", CATCH_CONFIG_CONSOLE_WIDTH - 43, ColumnInfo::Left },
+                { "samples      mean       std dev", 14, ColumnInfo::Right },
+                { "iterations   low mean   low std dev", 14, ColumnInfo::Right },
+                { "estimated    high mean  high std dev", 14, ColumnInfo::Right }
+            };
+        }
+    }())) {}
 ConsoleReporter::~ConsoleReporter() = default;
 
 std::string ConsoleReporter::getDescription() {
@@ -13660,6 +16479,10 @@ void ConsoleReporter::noMatchingTestCases(std::string const& spec) {
     stream << "No test cases matched '" << spec << '\'' << std::endl;
 }
 
+void ConsoleReporter::reportInvalidArguments(std::string const&arg){
+    stream << "Invalid Filter: " << arg << std::endl;
+}
+
 void ConsoleReporter::assertionStarting(AssertionInfo const&) {}
 
 bool ConsoleReporter::assertionEnded(AssertionStats const& _assertionStats) {
@@ -13680,6 +16503,7 @@ bool ConsoleReporter::assertionEnded(AssertionStats const& _assertionStats) {
 }
 
 void ConsoleReporter::sectionStarting(SectionInfo const& _sectionInfo) {
+    m_tablePrinter->close();
     m_headerPrinted = false;
     StreamingReporterBase::sectionStarting(_sectionInfo);
 }
@@ -13694,8 +16518,9 @@ void ConsoleReporter::sectionEnded(SectionStats const& _sectionStats) {
             stream << "\nNo assertions in test case";
         stream << " '" << _sectionStats.sectionInfo.name << "'\n" << std::endl;
     }
-    if (m_config->showDurations() == ShowDurations::Always) {
-        stream << getFormattedDuration(_sectionStats.durationInSeconds) << " s: " << _sectionStats.sectionInfo.name << std::endl;
+    double dur = _sectionStats.durationInSeconds;
+    if (shouldShowDuration(*m_config, dur)) {
+        stream << getFormattedDuration(dur) << " s: " << _sectionStats.sectionInfo.name << std::endl;
     }
     if (m_headerPrinted) {
         m_headerPrinted = false;
@@ -13703,28 +16528,53 @@ void ConsoleReporter::sectionEnded(SectionStats const& _sectionStats) {
     StreamingReporterBase::sectionEnded(_sectionStats);
 }
 
-void ConsoleReporter::benchmarkStarting(BenchmarkInfo const& info) {
-    lazyPrintWithoutClosingBenchmarkTable();
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+void ConsoleReporter::benchmarkPreparing(std::string const& name) {
+	lazyPrintWithoutClosingBenchmarkTable();
 
-    auto nameCol = Column( info.name ).width( static_cast<std::size_t>( m_tablePrinter->columnInfos()[0].width - 2 ) );
+	auto nameCol = Column(name).width(static_cast<std::size_t>(m_tablePrinter->columnInfos()[0].width - 2));
 
-    bool firstLine = true;
-    for (auto line : nameCol) {
-        if (!firstLine)
-            (*m_tablePrinter) << ColumnBreak() << ColumnBreak() << ColumnBreak();
-        else
-            firstLine = false;
+	bool firstLine = true;
+	for (auto line : nameCol) {
+		if (!firstLine)
+			(*m_tablePrinter) << ColumnBreak() << ColumnBreak() << ColumnBreak();
+		else
+			firstLine = false;
 
-        (*m_tablePrinter) << line << ColumnBreak();
+		(*m_tablePrinter) << line << ColumnBreak();
+	}
+}
+
+void ConsoleReporter::benchmarkStarting(BenchmarkInfo const& info) {
+    (*m_tablePrinter) << info.samples << ColumnBreak()
+        << info.iterations << ColumnBreak();
+    if (!m_config->benchmarkNoAnalysis())
+        (*m_tablePrinter) << Duration(info.estimatedDuration) << ColumnBreak();
+}
+void ConsoleReporter::benchmarkEnded(BenchmarkStats<> const& stats) {
+    if (m_config->benchmarkNoAnalysis())
+    {
+        (*m_tablePrinter) << Duration(stats.mean.point.count()) << ColumnBreak();
+    }
+    else
+    {
+        (*m_tablePrinter) << ColumnBreak()
+            << Duration(stats.mean.point.count()) << ColumnBreak()
+            << Duration(stats.mean.lower_bound.count()) << ColumnBreak()
+            << Duration(stats.mean.upper_bound.count()) << ColumnBreak() << ColumnBreak()
+            << Duration(stats.standardDeviation.point.count()) << ColumnBreak()
+            << Duration(stats.standardDeviation.lower_bound.count()) << ColumnBreak()
+            << Duration(stats.standardDeviation.upper_bound.count()) << ColumnBreak() << ColumnBreak() << ColumnBreak() << ColumnBreak() << ColumnBreak();
     }
 }
-void ConsoleReporter::benchmarkEnded(BenchmarkStats const& stats) {
-    Duration average(stats.elapsedTimeInNanoseconds / stats.iterations);
+
+void ConsoleReporter::benchmarkFailed(std::string const& error) {
+	Colour colour(Colour::Red);
     (*m_tablePrinter)
-        << stats.iterations << ColumnBreak()
-        << stats.elapsedTimeInNanoseconds << ColumnBreak()
-        << average << ColumnBreak();
+        << "Benchmark failed (" << error << ')'
+        << ColumnBreak() << RowBreak();
 }
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
 
 void ConsoleReporter::testCaseEnded(TestCaseStats const& _testCaseStats) {
     m_tablePrinter->close();
@@ -13746,6 +16596,10 @@ void ConsoleReporter::testRunEnded(TestRunStats const& _testRunStats) {
     stream << std::endl;
     StreamingReporterBase::testRunEnded(_testRunStats);
 }
+void ConsoleReporter::testRunStarting(TestRunInfo const& _testInfo) {
+    StreamingReporterBase::testRunStarting(_testInfo);
+    printTestFilters();
+}
 
 void ConsoleReporter::lazyPrint() {
 
@@ -13799,11 +16653,9 @@ void ConsoleReporter::printTestCaseAndSectionHeader() {
 
     SourceLineInfo lineInfo = m_sectionStack.back().lineInfo;
 
-    if (!lineInfo.empty()) {
-        stream << getLineOfChars<'-'>() << '\n';
-        Colour colourGuard(Colour::FileName);
-        stream << lineInfo << '\n';
-    }
+    stream << getLineOfChars<'-'>() << '\n';
+    Colour colourGuard(Colour::FileName);
+    stream << lineInfo << '\n';
     stream << getLineOfChars<'.'>() << '\n' << std::endl;
 }
 
@@ -13927,6 +16779,13 @@ void ConsoleReporter::printSummaryDivider() {
     stream << getLineOfChars<'-'>() << '\n';
 }
 
+void ConsoleReporter::printTestFilters() {
+    if (m_config->testSpec().hasFilters()) {
+        Colour guard(Colour::BrightYellow);
+        stream << "Filters: " << serializeFilters(m_config->getTestsOrTags()) << '\n';
+    }
+}
+
 CATCH_REGISTER_REPORTER("console", ConsoleReporter)
 
 } // end namespace Catch
@@ -13934,6 +16793,10 @@ CATCH_REGISTER_REPORTER("console", ConsoleReporter)
 #if defined(_MSC_VER)
 #pragma warning(pop)
 #endif
+
+#if defined(__clang__)
+#  pragma clang diagnostic pop
+#endif
 // end catch_reporter_console.cpp
 // start catch_reporter_junit.cpp
 
@@ -13941,6 +16804,7 @@ CATCH_REGISTER_REPORTER("console", ConsoleReporter)
 #include <sstream>
 #include <ctime>
 #include <algorithm>
+#include <iomanip>
 
 namespace Catch {
 
@@ -13968,7 +16832,7 @@ namespace Catch {
 #else
             std::strftime(timeStamp, timeStampSize, fmt, timeInfo);
 #endif
-            return std::string(timeStamp);
+            return std::string(timeStamp, timeStampSize-1);
         }
 
         std::string fileNameTag(const std::vector<std::string> &tags) {
@@ -13979,6 +16843,17 @@ namespace Catch {
                 return it->substr(1);
             return std::string();
         }
+
+        // Formats the duration in seconds to 3 decimal places.
+        // This is done because some genius defined Maven Surefire schema
+        // in a way that only accepts 3 decimal places, and tools like
+        // Jenkins use that schema for validation JUnit reporter output.
+        std::string formatDuration( double seconds ) {
+            ReusableStringStream rss;
+            rss << std::fixed << std::setprecision( 3 ) << seconds;
+            return rss.str();
+        }
+
     } // anonymous namespace
 
     JunitReporter::JunitReporter( ReporterConfig const& _config )
@@ -14000,13 +16875,6 @@ namespace Catch {
     void JunitReporter::testRunStarting( TestRunInfo const& runInfo )  {
         CumulativeReporterBase::testRunStarting( runInfo );
         xml.startElement( "testsuites" );
-        if( m_config->rngSeed() != 0 ) {
-            xml.startElement( "properties" );
-            xml.scopedElement( "property" )
-                .writeAttribute( "name", "random-seed" )
-                .writeAttribute( "value", m_config->rngSeed() );
-            xml.endElement();
-        }
     }
 
     void JunitReporter::testGroupStarting( GroupInfo const& groupInfo ) {
@@ -14045,6 +16913,7 @@ namespace Catch {
 
     void JunitReporter::writeGroup( TestGroupNode const& groupNode, double suiteTime ) {
         XmlWriter::ScopedElement e = xml.scopedElement( "testsuite" );
+
         TestGroupStats const& stats = groupNode.value;
         xml.writeAttribute( "name", stats.groupInfo.name );
         xml.writeAttribute( "errors", unexpectedExceptions );
@@ -14054,15 +16923,30 @@ namespace Catch {
         if( m_config->showDurations() == ShowDurations::Never )
             xml.writeAttribute( "time", "" );
         else
-            xml.writeAttribute( "time", suiteTime );
+            xml.writeAttribute( "time", formatDuration( suiteTime ) );
         xml.writeAttribute( "timestamp", getCurrentTimestamp() );
 
+        // Write properties if there are any
+        if (m_config->hasTestFilters() || m_config->rngSeed() != 0) {
+            auto properties = xml.scopedElement("properties");
+            if (m_config->hasTestFilters()) {
+                xml.scopedElement("property")
+                    .writeAttribute("name", "filters")
+                    .writeAttribute("value", serializeFilters(m_config->getTestsOrTags()));
+            }
+            if (m_config->rngSeed() != 0) {
+                xml.scopedElement("property")
+                    .writeAttribute("name", "random-seed")
+                    .writeAttribute("value", m_config->rngSeed());
+            }
+        }
+
         // Write test cases
         for( auto const& child : groupNode.children )
             writeTestCase( *child );
 
-        xml.scopedElement( "system-out" ).writeText( trim( stdOutForSuite ), false );
-        xml.scopedElement( "system-err" ).writeText( trim( stdErrForSuite ), false );
+        xml.scopedElement( "system-out" ).writeText( trim( stdOutForSuite ), XmlFormatting::Newline );
+        xml.scopedElement( "system-err" ).writeText( trim( stdErrForSuite ), XmlFormatting::Newline );
     }
 
     void JunitReporter::writeTestCase( TestCaseNode const& testCaseNode ) {
@@ -14084,12 +16968,13 @@ namespace Catch {
         if ( !m_config->name().empty() )
             className = m_config->name() + "." + className;
 
-        writeSection( className, "", rootSection );
+        writeSection( className, "", rootSection, stats.testInfo.okToFail() );
     }
 
-    void JunitReporter::writeSection(  std::string const& className,
-                        std::string const& rootName,
-                        SectionNode const& sectionNode ) {
+    void JunitReporter::writeSection( std::string const& className,
+                                      std::string const& rootName,
+                                      SectionNode const& sectionNode,
+                                      bool testOkToFail) {
         std::string name = trim( sectionNode.stats.sectionInfo.name );
         if( !rootName.empty() )
             name = rootName + '/' + name;
@@ -14106,20 +16991,30 @@ namespace Catch {
                 xml.writeAttribute( "classname", className );
                 xml.writeAttribute( "name", name );
             }
-            xml.writeAttribute( "time", ::Catch::Detail::stringify( sectionNode.stats.durationInSeconds ) );
+            xml.writeAttribute( "time", formatDuration( sectionNode.stats.durationInSeconds ) );
+            // This is not ideal, but it should be enough to mimic gtest's
+            // junit output.
+            // Ideally the JUnit reporter would also handle `skipTest`
+            // events and write those out appropriately.
+            xml.writeAttribute( "status", "run" );
+
+            if (sectionNode.stats.assertions.failedButOk) {
+                xml.scopedElement("skipped")
+                    .writeAttribute("message", "TEST_CASE tagged with !mayfail");
+            }
 
             writeAssertions( sectionNode );
 
             if( !sectionNode.stdOut.empty() )
-                xml.scopedElement( "system-out" ).writeText( trim( sectionNode.stdOut ), false );
+                xml.scopedElement( "system-out" ).writeText( trim( sectionNode.stdOut ), XmlFormatting::Newline );
             if( !sectionNode.stdErr.empty() )
-                xml.scopedElement( "system-err" ).writeText( trim( sectionNode.stdErr ), false );
+                xml.scopedElement( "system-err" ).writeText( trim( sectionNode.stdErr ), XmlFormatting::Newline );
         }
         for( auto const& childNode : sectionNode.childSections )
             if( className.empty() )
-                writeSection( name, "", *childNode );
+                writeSection( name, "", *childNode, testOkToFail );
             else
-                writeSection( className, name, *childNode );
+                writeSection( className, name, *childNode, testOkToFail );
     }
 
     void JunitReporter::writeAssertions( SectionNode const& sectionNode ) {
@@ -14137,11 +17032,7 @@ namespace Catch {
                     elementName = "error";
                     break;
                 case ResultWas::ExplicitFailure:
-                    elementName = "failure";
-                    break;
                 case ResultWas::ExpressionFailed:
-                    elementName = "failure";
-                    break;
                 case ResultWas::DidntThrowException:
                     elementName = "failure";
                     break;
@@ -14159,10 +17050,25 @@ namespace Catch {
 
             XmlWriter::ScopedElement e = xml.scopedElement( elementName );
 
-            xml.writeAttribute( "message", result.getExpandedExpression() );
+            xml.writeAttribute( "message", result.getExpression() );
             xml.writeAttribute( "type", result.getTestMacroName() );
 
             ReusableStringStream rss;
+            if (stats.totals.assertions.total() > 0) {
+                rss << "FAILED" << ":\n";
+                if (result.hasExpression()) {
+                    rss << "  ";
+                    rss << result.getExpressionInMacro();
+                    rss << '\n';
+                }
+                if (result.hasExpandedExpression()) {
+                    rss << "with expansion:\n";
+                    rss << Column(result.getExpandedExpression()).indent(2) << '\n';
+                }
+            } else {
+                rss << '\n';
+            }
+
             if( !result.getMessage().empty() )
                 rss << result.getMessage() << '\n';
             for( auto const& msg : stats.infoMessages )
@@ -14170,7 +17076,7 @@ namespace Catch {
                     rss << msg.message << '\n';
 
             rss << "at " << result.getSourceInfo();
-            xml.writeText( rss.str(), false );
+            xml.writeText( rss.str(), XmlFormatting::Newline );
         }
     }
 
@@ -14214,19 +17120,41 @@ namespace Catch {
         m_reporter->noMatchingTestCases( spec );
     }
 
+    void ListeningReporter::reportInvalidArguments(std::string const&arg){
+        for ( auto const& listener : m_listeners ) {
+            listener->reportInvalidArguments( arg );
+        }
+        m_reporter->reportInvalidArguments( arg );
+    }
+
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+    void ListeningReporter::benchmarkPreparing( std::string const& name ) {
+		for (auto const& listener : m_listeners) {
+			listener->benchmarkPreparing(name);
+		}
+		m_reporter->benchmarkPreparing(name);
+	}
     void ListeningReporter::benchmarkStarting( BenchmarkInfo const& benchmarkInfo ) {
         for ( auto const& listener : m_listeners ) {
             listener->benchmarkStarting( benchmarkInfo );
         }
         m_reporter->benchmarkStarting( benchmarkInfo );
     }
-    void ListeningReporter::benchmarkEnded( BenchmarkStats const& benchmarkStats ) {
+    void ListeningReporter::benchmarkEnded( BenchmarkStats<> const& benchmarkStats ) {
         for ( auto const& listener : m_listeners ) {
             listener->benchmarkEnded( benchmarkStats );
         }
         m_reporter->benchmarkEnded( benchmarkStats );
     }
 
+	void ListeningReporter::benchmarkFailed( std::string const& error ) {
+		for (auto const& listener : m_listeners) {
+			listener->benchmarkFailed(error);
+		}
+		m_reporter->benchmarkFailed(error);
+	}
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
+
     void ListeningReporter::testRunStarting( TestRunInfo const& testRunInfo ) {
         for ( auto const& listener : m_listeners ) {
             listener->testRunStarting( testRunInfo );
@@ -14357,6 +17285,8 @@ namespace Catch {
         m_xml.startElement( "Catch" );
         if( !m_config->name().empty() )
             m_xml.writeAttribute( "name", m_config->name() );
+        if (m_config->testSpec().hasFilters())
+            m_xml.writeAttribute( "filters", serializeFilters( m_config->getTestsOrTags() ) );
         if( m_config->rngSeed() != 0 )
             m_xml.scopedElement( "Randomness" )
                 .writeAttribute( "seed", m_config->rngSeed() );
@@ -14492,9 +17422,9 @@ namespace Catch {
             e.writeAttribute( "durationInSeconds", m_testCaseTimer.getElapsedSeconds() );
 
         if( !testCaseStats.stdOut.empty() )
-            m_xml.scopedElement( "StdOut" ).writeText( trim( testCaseStats.stdOut ), false );
+            m_xml.scopedElement( "StdOut" ).writeText( trim( testCaseStats.stdOut ), XmlFormatting::Newline );
         if( !testCaseStats.stdErr.empty() )
-            m_xml.scopedElement( "StdErr" ).writeText( trim( testCaseStats.stdErr ), false );
+            m_xml.scopedElement( "StdErr" ).writeText( trim( testCaseStats.stdErr ), XmlFormatting::Newline );
 
         m_xml.endElement();
     }
@@ -14506,6 +17436,10 @@ namespace Catch {
             .writeAttribute( "successes", testGroupStats.totals.assertions.passed )
             .writeAttribute( "failures", testGroupStats.totals.assertions.failed )
             .writeAttribute( "expectedFailures", testGroupStats.totals.assertions.failedButOk );
+        m_xml.scopedElement( "OverallResultsCases")
+            .writeAttribute( "successes", testGroupStats.totals.testCases.passed )
+            .writeAttribute( "failures", testGroupStats.totals.testCases.failed )
+            .writeAttribute( "expectedFailures", testGroupStats.totals.testCases.failedButOk );
         m_xml.endElement();
     }
 
@@ -14515,9 +17449,58 @@ namespace Catch {
             .writeAttribute( "successes", testRunStats.totals.assertions.passed )
             .writeAttribute( "failures", testRunStats.totals.assertions.failed )
             .writeAttribute( "expectedFailures", testRunStats.totals.assertions.failedButOk );
+        m_xml.scopedElement( "OverallResultsCases")
+            .writeAttribute( "successes", testRunStats.totals.testCases.passed )
+            .writeAttribute( "failures", testRunStats.totals.testCases.failed )
+            .writeAttribute( "expectedFailures", testRunStats.totals.testCases.failedButOk );
+        m_xml.endElement();
+    }
+
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+    void XmlReporter::benchmarkPreparing(std::string const& name) {
+        m_xml.startElement("BenchmarkResults")
+            .writeAttribute("name", name);
+    }
+
+    void XmlReporter::benchmarkStarting(BenchmarkInfo const &info) {
+        m_xml.writeAttribute("samples", info.samples)
+            .writeAttribute("resamples", info.resamples)
+            .writeAttribute("iterations", info.iterations)
+            .writeAttribute("clockResolution", info.clockResolution)
+            .writeAttribute("estimatedDuration", info.estimatedDuration)
+            .writeComment("All values in nano seconds");
+    }
+
+    void XmlReporter::benchmarkEnded(BenchmarkStats<> const& benchmarkStats) {
+        m_xml.startElement("mean")
+            .writeAttribute("value", benchmarkStats.mean.point.count())
+            .writeAttribute("lowerBound", benchmarkStats.mean.lower_bound.count())
+            .writeAttribute("upperBound", benchmarkStats.mean.upper_bound.count())
+            .writeAttribute("ci", benchmarkStats.mean.confidence_interval);
+        m_xml.endElement();
+        m_xml.startElement("standardDeviation")
+            .writeAttribute("value", benchmarkStats.standardDeviation.point.count())
+            .writeAttribute("lowerBound", benchmarkStats.standardDeviation.lower_bound.count())
+            .writeAttribute("upperBound", benchmarkStats.standardDeviation.upper_bound.count())
+            .writeAttribute("ci", benchmarkStats.standardDeviation.confidence_interval);
+        m_xml.endElement();
+        m_xml.startElement("outliers")
+            .writeAttribute("variance", benchmarkStats.outlierVariance)
+            .writeAttribute("lowMild", benchmarkStats.outliers.low_mild)
+            .writeAttribute("lowSevere", benchmarkStats.outliers.low_severe)
+            .writeAttribute("highMild", benchmarkStats.outliers.high_mild)
+            .writeAttribute("highSevere", benchmarkStats.outliers.high_severe);
+        m_xml.endElement();
         m_xml.endElement();
     }
 
+    void XmlReporter::benchmarkFailed(std::string const &error) {
+        m_xml.scopedElement("failed").
+            writeAttribute("message", error);
+        m_xml.endElement();
+    }
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
+
     CATCH_REGISTER_REPORTER( "xml", XmlReporter )
 
 } // end namespace Catch
@@ -14543,7 +17526,7 @@ namespace Catch {
 
 #ifndef __OBJC__
 
-#if defined(CATCH_CONFIG_WCHAR) && defined(WIN32) && defined(_UNICODE) && !defined(DO_NOT_USE_WMAIN)
+#if defined(CATCH_CONFIG_WCHAR) && defined(CATCH_PLATFORM_WINDOWS) && defined(_UNICODE) && !defined(DO_NOT_USE_WMAIN)
 // Standard C/C++ Win32 Unicode wmain entry point
 extern "C" int wmain (int argc, wchar_t * argv[], wchar_t * []) {
 #else
@@ -14620,6 +17603,7 @@ int main (int argc, char * const argv[]) {
 #endif // CATCH_CONFIG_DISABLE_MATCHERS
 
 #define CATCH_INFO( msg ) INTERNAL_CATCH_INFO( "CATCH_INFO", msg )
+#define CATCH_UNSCOPED_INFO( msg ) INTERNAL_CATCH_UNSCOPED_INFO( "CATCH_UNSCOPED_INFO", msg )
 #define CATCH_WARN( msg ) INTERNAL_CATCH_MSG( "CATCH_WARN", Catch::ResultWas::Warning, Catch::ResultDisposition::ContinueOnFailure, msg )
 #define CATCH_CAPTURE( ... ) INTERNAL_CATCH_CAPTURE( INTERNAL_CATCH_UNIQUE_NAME(capturer), "CATCH_CAPTURE",__VA_ARGS__ )
 
@@ -14637,14 +17621,22 @@ int main (int argc, char * const argv[]) {
 
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
 #define CATCH_TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE( __VA_ARGS__ )
+#define CATCH_TEMPLATE_TEST_CASE_SIG( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG( __VA_ARGS__ )
 #define CATCH_TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
+#define CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( className, __VA_ARGS__ )
 #define CATCH_TEMPLATE_PRODUCT_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE( __VA_ARGS__ )
+#define CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG( ... ) INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG( __VA_ARGS__ )
 #define CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, __VA_ARGS__ )
+#define CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, __VA_ARGS__ )
 #else
 #define CATCH_TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE( __VA_ARGS__ ) )
+#define CATCH_TEMPLATE_TEST_CASE_SIG( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG( __VA_ARGS__ ) )
 #define CATCH_TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ ) )
+#define CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( className, __VA_ARGS__ ) )
 #define CATCH_TEMPLATE_PRODUCT_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE( __VA_ARGS__ ) )
+#define CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG( __VA_ARGS__ ) )
 #define CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, __VA_ARGS__ ) )
+#define CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, __VA_ARGS__ ) )
 #endif
 
 #if !defined(CATCH_CONFIG_RUNTIME_STATIC_REQUIRE)
@@ -14665,6 +17657,13 @@ int main (int argc, char * const argv[]) {
 #define CATCH_THEN( desc )      INTERNAL_CATCH_DYNAMIC_SECTION( "     Then: " << desc )
 #define CATCH_AND_THEN( desc )  INTERNAL_CATCH_DYNAMIC_SECTION( "      And: " << desc )
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+#define CATCH_BENCHMARK(...) \
+    INTERNAL_CATCH_BENCHMARK(INTERNAL_CATCH_UNIQUE_NAME(C_A_T_C_H_B_E_N_C_H_), INTERNAL_CATCH_GET_1_ARG(__VA_ARGS__,,), INTERNAL_CATCH_GET_2_ARG(__VA_ARGS__,,))
+#define CATCH_BENCHMARK_ADVANCED(name) \
+    INTERNAL_CATCH_BENCHMARK_ADVANCED(INTERNAL_CATCH_UNIQUE_NAME(C_A_T_C_H_B_E_N_C_H_), name)
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
+
 // If CATCH_CONFIG_PREFIX_ALL is not defined then the CATCH_ prefix is not required
 #else
 
@@ -14717,14 +17716,26 @@ int main (int argc, char * const argv[]) {
 
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
 #define TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE( __VA_ARGS__ )
+#define TEMPLATE_TEST_CASE_SIG( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG( __VA_ARGS__ )
 #define TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
+#define TEMPLATE_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( className, __VA_ARGS__ )
 #define TEMPLATE_PRODUCT_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE( __VA_ARGS__ )
+#define TEMPLATE_PRODUCT_TEST_CASE_SIG( ... ) INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG( __VA_ARGS__ )
 #define TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, __VA_ARGS__ )
+#define TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, __VA_ARGS__ )
+#define TEMPLATE_LIST_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE(__VA_ARGS__)
+#define TEMPLATE_LIST_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE_METHOD( className, __VA_ARGS__ )
 #else
 #define TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE( __VA_ARGS__ ) )
+#define TEMPLATE_TEST_CASE_SIG( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG( __VA_ARGS__ ) )
 #define TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ ) )
+#define TEMPLATE_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( className, __VA_ARGS__ ) )
 #define TEMPLATE_PRODUCT_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE( __VA_ARGS__ ) )
+#define TEMPLATE_PRODUCT_TEST_CASE_SIG( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG( __VA_ARGS__ ) )
 #define TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, __VA_ARGS__ ) )
+#define TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, __VA_ARGS__ ) )
+#define TEMPLATE_LIST_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE( __VA_ARGS__ ) )
+#define TEMPLATE_LIST_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_LIST_TEST_CASE_METHOD( className, __VA_ARGS__ ) )
 #endif
 
 #if !defined(CATCH_CONFIG_RUNTIME_STATIC_REQUIRE)
@@ -14750,6 +17761,13 @@ int main (int argc, char * const argv[]) {
 #define THEN( desc )      INTERNAL_CATCH_DYNAMIC_SECTION( "     Then: " << desc )
 #define AND_THEN( desc )  INTERNAL_CATCH_DYNAMIC_SECTION( "      And: " << desc )
 
+#if defined(CATCH_CONFIG_ENABLE_BENCHMARKING)
+#define BENCHMARK(...) \
+    INTERNAL_CATCH_BENCHMARK(INTERNAL_CATCH_UNIQUE_NAME(C_A_T_C_H_B_E_N_C_H_), INTERNAL_CATCH_GET_1_ARG(__VA_ARGS__,,), INTERNAL_CATCH_GET_2_ARG(__VA_ARGS__,,))
+#define BENCHMARK_ADVANCED(name) \
+    INTERNAL_CATCH_BENCHMARK_ADVANCED(INTERNAL_CATCH_UNIQUE_NAME(C_A_T_C_H_B_E_N_C_H_), name)
+#endif // CATCH_CONFIG_ENABLE_BENCHMARKING
+
 using Catch::Detail::Approx;
 
 #else // CATCH_CONFIG_DISABLE
@@ -14789,12 +17807,13 @@ using Catch::Detail::Approx;
 #define CATCH_REQUIRE_THAT( arg, matcher ) (void)(0)
 #endif // CATCH_CONFIG_DISABLE_MATCHERS
 
-#define CATCH_INFO( msg )    (void)(0)
-#define CATCH_WARN( msg )    (void)(0)
-#define CATCH_CAPTURE( msg ) (void)(0)
+#define CATCH_INFO( msg )          (void)(0)
+#define CATCH_UNSCOPED_INFO( msg ) (void)(0)
+#define CATCH_WARN( msg )          (void)(0)
+#define CATCH_CAPTURE( msg )       (void)(0)
 
-#define CATCH_TEST_CASE( ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ))
-#define CATCH_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ))
+#define CATCH_TEST_CASE( ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ))
+#define CATCH_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ))
 #define CATCH_METHOD_AS_TEST_CASE( method, ... )
 #define CATCH_REGISTER_TEST_CASE( Function, ... ) (void)(0)
 #define CATCH_SECTION( ... )
@@ -14803,23 +17822,31 @@ using Catch::Detail::Approx;
 #define CATCH_FAIL_CHECK( ... ) (void)(0)
 #define CATCH_SUCCEED( ... ) (void)(0)
 
-#define CATCH_ANON_TEST_CASE() INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ))
+#define CATCH_ANON_TEST_CASE() INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ))
 
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
-#define CATCH_TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ) )
-#define CATCH_TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), className )
+#define CATCH_TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(__VA_ARGS__)
+#define CATCH_TEMPLATE_TEST_CASE_SIG( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG_NO_REGISTRATION(__VA_ARGS__)
+#define CATCH_TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION(className, __VA_ARGS__)
+#define CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG_NO_REGISTRATION(className, __VA_ARGS__ )
 #define CATCH_TEMPLATE_PRODUCT_TEST_CASE( ... ) CATCH_TEMPLATE_TEST_CASE( __VA_ARGS__ )
+#define CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG( ... ) CATCH_TEMPLATE_TEST_CASE( __VA_ARGS__ )
 #define CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, ... ) CATCH_TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
+#define CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, ... ) CATCH_TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
 #else
-#define CATCH_TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ) ) )
-#define CATCH_TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), className ) )
+#define CATCH_TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(__VA_ARGS__) )
+#define CATCH_TEMPLATE_TEST_CASE_SIG( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG_NO_REGISTRATION(__VA_ARGS__) )
+#define CATCH_TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION(className, __VA_ARGS__ ) )
+#define CATCH_TEMPLATE_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG_NO_REGISTRATION(className, __VA_ARGS__ ) )
 #define CATCH_TEMPLATE_PRODUCT_TEST_CASE( ... ) CATCH_TEMPLATE_TEST_CASE( __VA_ARGS__ )
+#define CATCH_TEMPLATE_PRODUCT_TEST_CASE_SIG( ... ) CATCH_TEMPLATE_TEST_CASE( __VA_ARGS__ )
 #define CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, ... ) CATCH_TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
+#define CATCH_TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, ... ) CATCH_TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
 #endif
 
 // "BDD-style" convenience wrappers
-#define CATCH_SCENARIO( ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ))
-#define CATCH_SCENARIO_METHOD( className, ... ) INTERNAL_CATCH_TESTCASE_METHOD_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ), className )
+#define CATCH_SCENARIO( ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ))
+#define CATCH_SCENARIO_METHOD( className, ... ) INTERNAL_CATCH_TESTCASE_METHOD_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ), className )
 #define CATCH_GIVEN( desc )
 #define CATCH_AND_GIVEN( desc )
 #define CATCH_WHEN( desc )
@@ -14865,11 +17892,12 @@ using Catch::Detail::Approx;
 #endif // CATCH_CONFIG_DISABLE_MATCHERS
 
 #define INFO( msg ) (void)(0)
+#define UNSCOPED_INFO( msg ) (void)(0)
 #define WARN( msg ) (void)(0)
-#define CAPTURE( msg ) (void)(0)
+#define CAPTURE( ... ) (void)(0)
 
-#define TEST_CASE( ... )  INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ))
-#define TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ))
+#define TEST_CASE( ... )  INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ))
+#define TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ))
 #define METHOD_AS_TEST_CASE( method, ... )
 #define REGISTER_TEST_CASE( Function, ... ) (void)(0)
 #define SECTION( ... )
@@ -14877,18 +17905,26 @@ using Catch::Detail::Approx;
 #define FAIL( ... ) (void)(0)
 #define FAIL_CHECK( ... ) (void)(0)
 #define SUCCEED( ... ) (void)(0)
-#define ANON_TEST_CASE() INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ))
+#define ANON_TEST_CASE() INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ))
 
 #ifndef CATCH_CONFIG_TRADITIONAL_MSVC_PREPROCESSOR
-#define TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ) )
-#define TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), className )
+#define TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(__VA_ARGS__)
+#define TEMPLATE_TEST_CASE_SIG( ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG_NO_REGISTRATION(__VA_ARGS__)
+#define TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION(className, __VA_ARGS__)
+#define TEMPLATE_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG_NO_REGISTRATION(className, __VA_ARGS__ )
 #define TEMPLATE_PRODUCT_TEST_CASE( ... ) TEMPLATE_TEST_CASE( __VA_ARGS__ )
+#define TEMPLATE_PRODUCT_TEST_CASE_SIG( ... ) TEMPLATE_TEST_CASE( __VA_ARGS__ )
 #define TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, ... ) TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
+#define TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, ... ) TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
 #else
-#define TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ) ) )
-#define TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_M_P_L_A_T_E____T_E_S_T____ ), className ) )
+#define TEMPLATE_TEST_CASE( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_NO_REGISTRATION(__VA_ARGS__) )
+#define TEMPLATE_TEST_CASE_SIG( ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_SIG_NO_REGISTRATION(__VA_ARGS__) )
+#define TEMPLATE_TEST_CASE_METHOD( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_NO_REGISTRATION(className, __VA_ARGS__ ) )
+#define TEMPLATE_TEST_CASE_METHOD_SIG( className, ... ) INTERNAL_CATCH_EXPAND_VARGS( INTERNAL_CATCH_TEMPLATE_TEST_CASE_METHOD_SIG_NO_REGISTRATION(className, __VA_ARGS__ ) )
 #define TEMPLATE_PRODUCT_TEST_CASE( ... ) TEMPLATE_TEST_CASE( __VA_ARGS__ )
+#define TEMPLATE_PRODUCT_TEST_CASE_SIG( ... ) TEMPLATE_TEST_CASE( __VA_ARGS__ )
 #define TEMPLATE_PRODUCT_TEST_CASE_METHOD( className, ... ) TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
+#define TEMPLATE_PRODUCT_TEST_CASE_METHOD_SIG( className, ... ) TEMPLATE_TEST_CASE_METHOD( className, __VA_ARGS__ )
 #endif
 
 #define STATIC_REQUIRE( ... )       (void)(0)
@@ -14899,8 +17935,8 @@ using Catch::Detail::Approx;
 #define CATCH_TRANSLATE_EXCEPTION( signature ) INTERNAL_CATCH_TRANSLATE_EXCEPTION_NO_REG( INTERNAL_CATCH_UNIQUE_NAME( catch_internal_ExceptionTranslator ), signature )
 
 // "BDD-style" convenience wrappers
-#define SCENARIO( ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ) )
-#define SCENARIO_METHOD( className, ... ) INTERNAL_CATCH_TESTCASE_METHOD_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( ____C_A_T_C_H____T_E_S_T____ ), className )
+#define SCENARIO( ... ) INTERNAL_CATCH_TESTCASE_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ) )
+#define SCENARIO_METHOD( className, ... ) INTERNAL_CATCH_TESTCASE_METHOD_NO_REGISTRATION(INTERNAL_CATCH_UNIQUE_NAME( C_A_T_C_H_T_E_S_T_ ), className )
 
 #define GIVEN( desc )
 #define AND_GIVEN( desc )
diff --git a/test/kernels/add_bias_test.cc b/test/kernels/add_bias_test.cc
index b9e5fd9..50b2313 100644
--- a/test/kernels/add_bias_test.cc
+++ b/test/kernels/add_bias_test.cc
@@ -23,7 +23,11 @@ void kernel_add_bias_test() {
 
   *output.template as<vec_t>() = kernels::add_bias(*input.template as<vec_t>(), bias.begin(), 0);
   for (std::size_t i = 0; i < output.size(); ++i)
+#if !defined( __EXCEPTIONS) && defined(__GNUC__) && (__GNUC__ == 9) && ((__GNUC_MINOR__ == 3) || (__GNUC_MINOR__ == 4))
+    CHECK(output[i] == static_cast<ElemType_>(bias[i] + input[i])); // GCC 9.3 - exceptions breaks std::iota with newer check.
+#else
     CHECK(output[i] == ElemType_(100 + i));
+#endif
 }
 
 template INTGEMM_SSE2 void kernel_add_bias_test<CPUType::SSE2, int8_t>();
diff --git a/test/kernels/unquantize_test.cc b/test/kernels/unquantize_test.cc
index edfafa5..7bb3b36 100644
--- a/test/kernels/unquantize_test.cc
+++ b/test/kernels/unquantize_test.cc
@@ -22,7 +22,11 @@ void kernel_unquantize_test() {
 
   *output.template as<output_vec_t>() = kernels::unquantize(*input.template as<input_vec_t>(), unquant_mult);
   for (std::size_t i = 0; i < output.size(); ++i)
+#if !defined( __EXCEPTIONS) && defined(__GNUC__) && (__GNUC__ == 9) && ((__GNUC_MINOR__ == 3) || (__GNUC_MINOR__ == 4))
+    CHECK(output[i] == input[i] * 0.5f); // GCC 9.3 - exceptions breaks std::iota with newer check.
+#else
     CHECK(output[i] == i * 0.5f);
+#endif
 }
 
 template INTGEMM_SSE2 void kernel_unquantize_test<CPUType::SSE2>();
diff --git a/test/kernels/write_test.cc b/test/kernels/write_test.cc
index a136a86..339b4e9 100644
--- a/test/kernels/write_test.cc
+++ b/test/kernels/write_test.cc
@@ -22,7 +22,11 @@ void kernel_write_test() {
 
   kernels::write(*input.template as<vec_t>(), output.begin(), 0);
   for (std::size_t i = 0; i < VECTOR_LENGTH; ++i)
+#if !defined( __EXCEPTIONS) && defined(__GNUC__) && (__GNUC__ == 9) && ((__GNUC_MINOR__ == 3) || (__GNUC_MINOR__ == 4))
+    CHECK(output[i] == input[i]); // GCC 9.3 - exceptions breaks std::iota with newer check.
+#else
     CHECK(output[i] == ElemType_(i));
+#endif
 }
 
 template INTGEMM_SSE2 void kernel_write_test<CPUType::SSE2, int8_t>();