From 1a1d1684776e9c206f9bed54e7d1742627e76c2b Mon Sep 17 00:00:00 2001
From: Dmitry Ivanov <ethercrow@gmail.com>
Date: Fri, 23 Oct 2020 18:48:52 +0200
Subject: [PATCH] Implement intersperse using SSE2

---
 bench/BenchAll.hs |  4 ++++
 cbits/fpstring.c  | 20 +++++++++++++++++++-
 2 files changed, 23 insertions(+), 1 deletion(-)

diff --git a/bench/BenchAll.hs b/bench/BenchAll.hs
index 40d76007f..79eb4d566 100644
--- a/bench/BenchAll.hs
+++ b/bench/BenchAll.hs
@@ -361,6 +361,10 @@ main = do
       , benchFE "floatHexFixed"    $ fromIntegral >$< P.floatHexFixed
       , benchFE "doubleHexFixed"   $ fromIntegral >$< P.doubleHexFixed
       ]
+    , bgroup "intersperse"
+      [ bench "intersperse" $ whnf (S.intersperse 32) byteStringData
+      , bench "intersperse (unaligned)" $ whnf (S.intersperse 32) (S.drop 1 byteStringData)
+      ]
     , bgroup "partition"
       [
         bgroup "strict"
diff --git a/cbits/fpstring.c b/cbits/fpstring.c
index f82bc00e5..9beb00d99 100644
--- a/cbits/fpstring.c
+++ b/cbits/fpstring.c
@@ -30,6 +30,10 @@
  */
 
 #include "fpstring.h"
+#if defined(__x86_64__)
+#include <emmintrin.h>
+#include <xmmintrin.h>
+#endif
 
 /* copy a string in reverse */
 void fps_reverse(unsigned char *q, unsigned char *p, size_t n) {
@@ -44,7 +48,21 @@ void fps_intersperse(unsigned char *q,
                      unsigned char *p,
                      size_t n,
                      unsigned char c) {
-
+#if defined(__x86_64__)
+  {
+    const __m128i separator = _mm_set1_epi8(c);
+    const unsigned char *const p_begin = p;
+    const unsigned char *const p_end = p_begin + n - 9;
+    while (p < p_end) {
+      const __m128i eight_src_bytes = _mm_loadl_epi64((__m128i *)p);
+      const __m128i sixteen_dst_bytes = _mm_unpacklo_epi8(eight_src_bytes, separator);
+      _mm_storeu_si128((__m128i *)q, sixteen_dst_bytes);
+      p += 8;
+      q += 16;
+    }
+    n -= p - p_begin;
+  }
+#endif
     while (n > 1) {
         *q++ = *p++;
         *q++ = c;