ShuriZma
/
xenia
mirror of https://github.com/xenia-project/xenia.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Use SSE intrinsics for uint16_t byte swapping 

Part of https://github.com/benvanik/xenia/issues/308
This commit is contained in:
raven02 2015-07-02 23:04:48 +08:00
parent 4ec0655751
commit fea115a6f4
1 changed files with 8 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
src/xenia/base/memory_generic.cc
View File

@ -43,7 +43,14 @@ void copy_and_swap_16_aligned(uint16_t* dest, const uint16_t* src,
void copy_and_swap_16_unaligned(uint16_t* dest, const uint16_t* src, void copy_and_swap_16_unaligned(uint16_t* dest, const uint16_t* src,
size_t count) { size_t count) {
for (size_t i = 0; i < count; ++i) { size_t i;
for (i = 0; i + 8 <= count; i += 8) {
__m128i s = _mm_loadu_si128((__m128i*)&src[i]);
__m128i d = _mm_or_si128(_mm_slli_epi16(s, 8), _mm_srli_epi16(s, 8));
_mm_storeu_si128((__m128i*)&dest[i], d);
}
for (; i < count; ++i) { // handle residual elements
dest[i] = byte_swap(src[i]); dest[i] = byte_swap(src[i]);
} }
} }