Skip to content

Use unsigned byte comparisons with SSE optimizations #48

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Use unsigned byte comparisons with SSE optimizations #48

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 19 additions & 2 deletions src/art.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,23 @@
#endif
#endif

#if defined(__i386__) || defined(__amd64__)
/* From http://www.alfredklomp.com/programming/sse-intrinsics/ */
static inline __m128i
_mm_cmpgt_epu8(__m128i x, __m128i y)
{
// Returns 0xFF where x > y:
return _mm_andnot_si128(_mm_cmpeq_epi8(x, y),
_mm_cmpeq_epi8(_mm_max_epu8(x, y), x));
}
static inline __m128i
_mm_cmplt_epu8(__m128i x, __m128i y)
{
// Returns 0xFF where x < y:
return _mm_cmpgt_epu8(y, x);
}
#endif

/**
* Macros to manipulate pointer tags
*/
Expand Down Expand Up @@ -416,7 +433,7 @@ static void add_child16(art_node16 *n, art_node **ref, unsigned char c, void *ch
__m128i cmp;

// Compare the key to all 16 stored keys
cmp = _mm_cmplt_epi8(_mm_set1_epi8(c),
cmp = _mm_cmplt_epu8(_mm_set1_epi8(c),
_mm_loadu_si128((__m128i*)n->keys));

// Use a mask to ignore children that don't exist
Expand All @@ -426,7 +443,7 @@ static void add_child16(art_node16 *n, art_node **ref, unsigned char c, void *ch
__m128i cmp;

// Compare the key to all 16 stored keys
cmp = _mm_cmplt_epi8(_mm_set1_epi8(c),
cmp = _mm_cmplt_epu8(_mm_set1_epi8(c),
_mm_loadu_si128((__m128i*)n->keys));

// Use a mask to ignore children that don't exist
Expand Down