[PATCH 8/9] fix: Use INT8_* instead of CHAR_*

For consistency and to prevent bugs, we move from CHAR_* to INT8_*
everywhere in this kernel.

Gbp-Pq: Name 0008-fix-Use-INT8_-instead-of-CHAR_.patch

diff --git a/kernels/volk/volk_32f_s32f_convert_8i.h b/kernels/volk/volk_32f_s32f_convert_8i.h
index ea16ef8953aaab46ac7fd3b0039074d84668fb82..d08ecd5e959ce61488ee7913aeba34af59cdba46 100644 (file)
--- a/kernels/volk/volk_32f_s32f_convert_8i.h
+++ b/kernels/volk/volk_32f_s32f_convert_8i.h
@@ -104,8 +104,8 @@ static inline void volk_32f_s32f_convert_8i_u_avx2(int8_t* outputVector,
     const float* inputVectorPtr = (const float*)inputVector;
     int8_t* outputVectorPtr = outputVector;
 
-    float min_val = CHAR_MIN;
-    float max_val = CHAR_MAX;
+    float min_val = INT8_MIN;
+    float max_val = INT8_MAX;
     float r;
 
     __m256 vScalar = _mm256_set1_ps(scalar);
@@ -176,8 +176,8 @@ static inline void volk_32f_s32f_convert_8i_u_sse2(int8_t* outputVector,
     const float* inputVectorPtr = (const float*)inputVector;
     int8_t* outputVectorPtr = outputVector;
 
-    float min_val = CHAR_MIN;
-    float max_val = CHAR_MAX;
+    float min_val = INT8_MIN;
+    float max_val = INT8_MAX;
     float r;
 
     __m128 vScalar = _mm_set_ps1(scalar);
@@ -245,8 +245,8 @@ static inline void volk_32f_s32f_convert_8i_u_sse(int8_t* outputVector,
     const float* inputVectorPtr = (const float*)inputVector;
     int8_t* outputVectorPtr = outputVector;
 
-    float min_val = CHAR_MIN;
-    float max_val = CHAR_MAX;
+    float min_val = INT8_MIN;
+    float max_val = INT8_MAX;
     float r;
 
     __m128 vScalar = _mm_set_ps1(scalar);
@@ -321,8 +321,8 @@ static inline void volk_32f_s32f_convert_8i_a_avx2(int8_t* outputVector,
     const float* inputVectorPtr = (const float*)inputVector;
     int8_t* outputVectorPtr = outputVector;
 
-    float min_val = CHAR_MIN;
-    float max_val = CHAR_MAX;
+    float min_val = INT8_MIN;
+    float max_val = INT8_MAX;
     float r;
 
     __m256 vScalar = _mm256_set1_ps(scalar);
@@ -393,8 +393,8 @@ static inline void volk_32f_s32f_convert_8i_a_sse2(int8_t* outputVector,
     const float* inputVectorPtr = (const float*)inputVector;
     int8_t* outputVectorPtr = outputVector;
 
-    float min_val = CHAR_MIN;
-    float max_val = CHAR_MAX;
+    float min_val = INT8_MIN;
+    float max_val = INT8_MAX;
     float r;
 
     __m128 vScalar = _mm_set_ps1(scalar);
@@ -460,8 +460,8 @@ static inline void volk_32f_s32f_convert_8i_a_sse(int8_t* outputVector,
 
     const float* inputVectorPtr = (const float*)inputVector;
 
-    float min_val = CHAR_MIN;
-    float max_val = CHAR_MAX;
+    float min_val = INT8_MIN;
+    float max_val = INT8_MAX;
     float r;
 
     int8_t* outputVectorPtr = outputVector;