add xtensa bit width 4 decompression code

tensorflow/lite/micro/micro_context.cc

@@ -26,6 +26,10 @@ limitations under the License.
 #include "tensorflow/lite/micro/micro_utils.h"
 #include "tensorflow/lite/portable_type_to_tflitetype.h"
 
+#ifdef HIFI5
+#include "tensorflow/lite/micro/kernels/xtensa/xtensa.h"
+#endif  // HIFI5
+
 namespace tflite {
 namespace {
 
@@ -57,6 +61,10 @@ struct DecompressionState {
   template <typename T>
   T* DecompressToBuffer(void* buffer);
 
+#ifdef HIFI5
+  void DecompressToBufferWidth4_Xtensa(int8_t* buffer);
+#endif  // HIFI5
+
   void DecompressToBufferWidth4_16(int8_t* buffer);
   void DecompressToBufferWidth3_32(int8_t* buffer);
   void DecompressToBufferWidth2_16(int8_t* buffer);
@@ -476,6 +484,52 @@ void DecompressionState::DecompressToBufferWidthAny(T* buffer) {
   }
 }
 
+#ifdef HIFI5
+void DecompressionState::DecompressToBufferWidth4_Xtensa(int8_t* buffer) {
+  MicroProfiler* profiler =
+      static_cast<MicroProfiler*>(micro_context_->external_context());
+  ScopedMicroProfiler scoped_profiler(__func__, profiler);
+
+  char shuffle_pattern_1[8] = {0x08, 0x19, 0x2A, 0x3B, 0x4C, 0x5D, 0x6E, 0x7F};
+  ae_int8x8 d_shuffle_t = *(ae_int8x8*)&shuffle_pattern_1[0];
+
+  char shuffle_pattern_2[8] = {0xFB, 0x73, 0xEA, 0x62, 0xD9, 0x51, 0xC8, 0x40};
+  ae_int8x8 d_d_shuffle_t2 = *(ae_int8x8*)&shuffle_pattern_2[0];
+
+  ae_int8x8 d_out1, d_out2;
+  ae_int8x8 d_value_0, d_value_1;
+  ae_int8x8 d_index;
+
+  ae_int8x8* pIn_tmp = (ae_int8x8*)compressed_indices_;
+  ae_int8* p_out_tmp = (ae_int8*)buffer;
+
+  const size_t stride = comp_data_.data.lut_data->value_table_channel_stride;
+  const uint8_t* value_table =
+      static_cast<const uint8_t*>(comp_data_.data.lut_data->value_table);
+
+#ifdef notdef
+  MicroPrintf("indices %p buffer %p value_table %p stride %u",
+              compressed_indices_, buffer, value_table, stride);
+#endif
+
+  for (size_t i = 0; i < num_channels_; i++) {
+    ae_int8x8 d_value_0_t = *(ae_int8x8*)&value_table[0];
+    ae_int8x8 d_value_1_t = *(ae_int8x8*)&value_table[8];
+
+    AE_DSEL8X8(d_value_0, d_value_1, d_value_0_t, d_value_1_t, d_shuffle_t);
+
+    for (size_t j = 0; j < elements_per_channel_; j += 16) {
+      AE_L8X8_IP(d_index, pIn_tmp, 8);
+      AE_DSEL8X8(d_out1, d_out2, d_value_0, d_value_1, d_index);
+      AE_DSEL8X8(d_out1, d_out2, d_out1, d_out2, d_d_shuffle_t2);
+      AE_S8X8X2_IP(d_out1, d_out2, (ae_int8x16*)p_out_tmp, 16);
+    }
+
+    value_table += stride;
+  }
+}
+#endif  // HIFI5
+
 template <typename T>
 T* DecompressionState::DecompressToBuffer(void* buffer) {
   TFLITE_DCHECK(compressed_bit_width_ <= LookupTableData::kMaxBitWidth);
@@ -484,7 +538,16 @@ T* DecompressionState::DecompressToBuffer(void* buffer) {
   if (std::is_same<T, int8_t>::value &&
       comp_data_.data.lut_data->compressed_bit_width == 4 &&
       !comp_data_.data.lut_data->use_alternate_axis) {
+#ifdef HIFI5
+    if (!(elements_per_channel_ & 0x0F) &&
+        comp_data_.data.lut_data->value_table_channel_stride == 16) {
+      DecompressToBufferWidth4_Xtensa(static_cast<int8_t*>(buffer));
+    } else {
+      DecompressToBufferWidth4_16(static_cast<int8_t*>(buffer));
+    }
+#else   // HIFI5
     DecompressToBufferWidth4_16(static_cast<int8_t*>(buffer));
+#endif  // HIFI5
   } else if (std::is_same<T, int8_t>::value &&
              comp_data_.data.lut_data->compressed_bit_width == 2 &&
              !comp_data_.data.lut_data->use_alternate_axis) {