Added integral transformation u8 SSE

Author: awxkee

src/lib/gaussian/sse/horiz_one_approx.rs

@@ -34,6 +34,14 @@ use std::arch::x86::*;
 #[cfg(target_arch = "x86_64")]
 use std::arch::x86_64::*;
 
+#[inline]
+#[target_feature(enable = "sse4.1")]
+#[cfg(target_arch = "x86")]
+pub unsafe fn _mm_loadu_si16(mem_addr: *const u8) -> __m128i {
+    let item = (mem_addr as *const i16).read_unaligned();
+    _mm_set1_epi16(item)
+}
+
 #[inline]
 #[target_feature(enable = "sse4.1")]
 pub unsafe fn _mm_sum_clamp(v: __m128i) -> i32 {