move #if to avoid unused warning

Signed-off-by: Loic Dachary <loic@dachary.org>
(cherry picked from commit 29899ad4439aca07ccd2efe2a8b7086efdec72b0)

6 files changed, 104 insertions, 108 deletions

diff --git a/src/gf_w128.c b/src/gf_w128.c
index abf0d81..b7cacf5 100644
--- a/src/gf_w128.c
+++ b/src/gf_w128.c
@@ -81,6 +81,7 @@ int xor)
     }
 }
 
+#if defined(INTEL_SSE4_PCLMUL)
 static
 void
 gf_w128_clm_multiply_region_from_single(gf_t *gf, void *src, void *dest, gf_val_128_t val, int bytes,
@@ -91,7 +92,6 @@ int xor)
     gf_val_128_t d128;
     uint64_t c128[2];
     gf_region_data rd;
-#if defined(INTEL_SSE4_PCLMUL)
     __m128i     a,b;
     __m128i     result0,result1;
     __m128i     prim_poly;
@@ -184,8 +184,8 @@ int xor)
         d128[i+1] = (uint64_t)_mm_extract_epi64(result1,0);
       }
     }
-#endif
 }
+#endif
 
 /*
  * Some w128 notes:
@@ -599,11 +599,11 @@ gf_w128_split_4_128_multiply_region(gf_t *gf, void *src, void *dest, gf_val_128_
   }
 }
 
+#ifdef INTEL_SSSE3
 static
 void
 gf_w128_split_4_128_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_128_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSSE3
   gf_internal_t *h;
   int i, m, j, k, tindex;
   uint64_t pp, v[2], s, *s64, *d64, *top;
@@ -695,14 +695,14 @@ gf_w128_split_4_128_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_
   /* Doing this instead of gf_do_final_region_alignment() because that doesn't hold 128-bit vals */
 
   gf_w128_multiply_region_from_single(gf, rd.s_top, rd.d_top, val, ((char*)src+bytes)-(char*)rd.s_top, xor);
-#endif
 }
+#endif
 
+#ifdef INTEL_SSSE3
 static
 void
 gf_w128_split_4_128_sse_altmap_multiply_region(gf_t *gf, void *src, void *dest, gf_val_128_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSSE3
   gf_internal_t *h;
   int i, m, j, k, tindex;
   uint64_t pp, v[2], s, *s64, *d64, *top;
@@ -805,8 +805,8 @@ gf_w128_split_4_128_sse_altmap_multiply_region(gf_t *gf, void *src, void *dest,
   /* Doing this instead of gf_do_final_region_alignment() because that doesn't hold 128-bit vals */
 
   gf_w128_multiply_region_from_single(gf, rd.s_top, rd.d_top, val, ((char*)src+bytes)-(char*)rd.s_top, xor);
-#endif
 }
+#endif
 
 static
 void
@@ -1495,10 +1495,10 @@ void gf_w128_group_r_init(gf_t *gf)
   return;
 }
 
+#if defined(INTEL_SSE4)
   static
 void gf_w128_group_r_sse_init(gf_t *gf)
 {
-#if defined(INTEL_SSE4)
   int i, j;
   int g_r;
   uint64_t pp;
@@ -1520,8 +1520,8 @@ void gf_w128_group_r_sse_init(gf_t *gf)
     }
   }
   return;
-#endif
 }
+#endif
 
   static 
 int gf_w128_split_init(gf_t *gf)
diff --git a/src/gf_w16.c b/src/gf_w16.c
index 75cad50..babddd7 100644
--- a/src/gf_w16.c
+++ b/src/gf_w16.c
@@ -125,6 +125,7 @@ gf_w16_multiply_region_from_single(gf_t *gf, void *src, void *dest, gf_val_32_t
   gf_do_final_region_alignment(&rd);
 }
 
+#if defined(INTEL_SSE4_PCLMUL)
 static
 void
 gf_w16_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
@@ -132,8 +133,6 @@ gf_w16_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, gf_val
   gf_region_data rd;
   uint16_t *s16;
   uint16_t *d16;
-
-#if defined(INTEL_SSE4_PCLMUL)
   __m128i         a, b;
   __m128i         result;
   __m128i         prim_poly;
@@ -186,9 +185,10 @@ gf_w16_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, gf_val
     } 
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
+#if defined(INTEL_SSE4_PCLMUL)
 static
 void
 gf_w16_clm_multiply_region_from_single_3(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
@@ -197,8 +197,6 @@ gf_w16_clm_multiply_region_from_single_3(gf_t *gf, void *src, void *dest, gf_val
   uint16_t *s16;
   uint16_t *d16;
 
-#if defined(INTEL_SSE4_PCLMUL)
-
   __m128i         a, b;
   __m128i         result;
   __m128i         prim_poly;
@@ -255,9 +253,10 @@ gf_w16_clm_multiply_region_from_single_3(gf_t *gf, void *src, void *dest, gf_val
     } 
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
+#if defined(INTEL_SSE4_PCLMUL)
 static
 void
 gf_w16_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
@@ -266,8 +265,6 @@ gf_w16_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, gf_val
   uint16_t *s16;
   uint16_t *d16;
 
-#if defined(INTEL_SSE4_PCLMUL)
-
   __m128i         a, b;
   __m128i         result;
   __m128i         prim_poly;
@@ -328,8 +325,8 @@ gf_w16_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, gf_val
     } 
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 static
 inline
@@ -605,13 +602,13 @@ int gf_w16_shift_init(gf_t *gf)
 static 
 int gf_w16_cfm_init(gf_t *gf)
 {
+#if defined(INTEL_SSE4_PCLMUL)
   gf_internal_t *h;
 
   h = (gf_internal_t *) gf->scratch;
   
   /*Ben: Determining how many reductions to do */
   
-#if defined(INTEL_SSE4_PCLMUL)
   if ((0xfe00 & h->prim_poly) == 0) {
     gf->multiply.w32 = gf_w16_clm_multiply_2;
     gf->multiply_region.w32 = gf_w16_clm_multiply_region_from_single_2;
@@ -1548,11 +1545,11 @@ gf_w16_bytwo_p_nosse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_
       prod = _mm_xor_si128(prod, t1); \
       v = _mm_srli_epi64(v, 1); }
 
+#ifdef INTEL_SSE2
 static
 void 
 gf_w16_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *s8, *d8;
   uint32_t vrev;
@@ -1609,14 +1606,14 @@ gf_w16_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t
     s8 += 16;
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w16_bytwo_b_sse_region_2_noxor(gf_region_data *rd, struct gf_w16_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1635,14 +1632,14 @@ gf_w16_bytwo_b_sse_region_2_noxor(gf_region_data *rd, struct gf_w16_bytwo_data *
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w16_bytwo_b_sse_region_2_xor(gf_region_data *rd, struct gf_w16_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1663,15 +1660,15 @@ gf_w16_bytwo_b_sse_region_2_xor(gf_region_data *rd, struct gf_w16_bytwo_data *bt
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
 
+#ifdef INTEL_SSE2
 static
 void 
 gf_w16_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   int itb;
   uint8_t *d8, *s8;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1719,8 +1716,8 @@ gf_w16_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t
   }
 
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 static
 void 
diff --git a/src/gf_w32.c b/src/gf_w32.c
index f472e02..d8234b3 100644
--- a/src/gf_w32.c
+++ b/src/gf_w32.c
@@ -120,13 +120,13 @@ xor)
   }
 }
 
+#if defined(INTEL_SSE4_PCLMUL)
+
 static 
 void
 gf_w32_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, uint32_t val, int bytes, int xor)
 {
 
-#if defined(INTEL_SSE4_PCLMUL)
-
   int i;
   uint32_t *s32;
   uint32_t *d32;
@@ -167,16 +167,16 @@ gf_w32_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, uint32
       d32[i] = ((gf_val_32_t)_mm_extract_epi32(result, 0));
     } 
   }
-#endif
 }
+#endif
+
+#if defined(INTEL_SSE4_PCLMUL) 
 
 static 
 void
 gf_w32_clm_multiply_region_from_single_3(gf_t *gf, void *src, void *dest, uint32_t val, int bytes, int xor)
 {
 
-#if defined(INTEL_SSE4_PCLMUL) 
-
   int i;
   uint32_t *s32;
   uint32_t *d32;
@@ -222,14 +222,14 @@ gf_w32_clm_multiply_region_from_single_3(gf_t *gf, void *src, void *dest, uint32
       d32[i] = ((gf_val_32_t)_mm_extract_epi32(result, 0));
     } 
   }
-#endif
 }
+#endif
 
+#if defined(INTEL_SSE4_PCLMUL)
 static 
 void
 gf_w32_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, uint32_t val, int bytes, int xor)
 {
-#if defined(INTEL_SSE4_PCLMUL)
   int i;
   uint32_t *s32;
   uint32_t *d32;
@@ -279,8 +279,8 @@ gf_w32_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, uint32
       d32[i] = ((gf_val_32_t)_mm_extract_epi32(result, 0));
     } 
   }
-#endif
 }
+#endif
 
 static
 inline
@@ -983,11 +983,11 @@ gf_w32_bytwo_p_nosse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_
       prod = _mm_xor_si128(prod, t1); \
       v = _mm_srli_epi64(v, 1); }
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w32_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *s8, *d8;
   uint32_t vrev;
@@ -1036,8 +1036,8 @@ gf_w32_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t
     s8 += 16;
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 static
 void
@@ -1177,11 +1177,11 @@ gf_w32_bytwo_b_nosse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_
   gf_do_final_region_alignment(&rd);
 }
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w32_bytwo_b_sse_region_2_noxor(gf_region_data *rd, struct gf_w32_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1200,14 +1200,14 @@ gf_w32_bytwo_b_sse_region_2_noxor(gf_region_data *rd, struct gf_w32_bytwo_data *
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w32_bytwo_b_sse_region_2_xor(gf_region_data *rd, struct gf_w32_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1228,15 +1228,15 @@ gf_w32_bytwo_b_sse_region_2_xor(gf_region_data *rd, struct gf_w32_bytwo_data *bt
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
 
+#ifdef INTEL_SSE2
 static
 void 
 gf_w32_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   uint32_t itb;
   uint8_t *d8, *s8;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1284,8 +1284,8 @@ gf_w32_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t
   }
 
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 static
 int gf_w32_bytwo_init(gf_t *gf)
@@ -1552,11 +1552,11 @@ gf_w32_split_2_32_lazy_multiply_region(gf_t *gf, void *src, void *dest, uint32_t
   gf_do_final_region_alignment(&rd);
 }
 
+#ifdef INTEL_SSSE3
 static
 void
 gf_w32_split_2_32_lazy_sse_multiply_region(gf_t *gf, void *src, void *dest, uint32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSSE3
   gf_internal_t *h;
   int i, m, j, tindex;
   uint32_t pp, v, v2, s, *s32, *d32, *top;
@@ -1631,8 +1631,8 @@ gf_w32_split_2_32_lazy_sse_multiply_region(gf_t *gf, void *src, void *dest, uint
 
   gf_do_final_region_alignment(&rd);
 
-#endif
 }
+#endif
 
 static
 void
diff --git a/src/gf_w4.c b/src/gf_w4.c
index 9a0dfb5..cba23b7 100644
--- a/src/gf_w4.c
+++ b/src/gf_w4.c
@@ -414,11 +414,11 @@ gf_w4_single_table_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t
 
 #define MM_PRINT(s, r) { uint8_t blah[16]; printf("%-12s", s); _mm_storeu_si128((__m128i *)blah, r); for (i = 0; i < 16; i++) printf(" %02x", blah[i]); printf("\n"); }
 
+#ifdef INTEL_SSSE3
 static
 void 
 gf_w4_single_table_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSSE3
   gf_region_data rd;
   uint8_t *base, *sptr, *dptr, *top;
   __m128i  tl, loset, h4, r, va, th;
@@ -460,8 +460,8 @@ gf_w4_single_table_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_3
   }
   gf_do_final_region_alignment(&rd);
 
-#endif
 }
+#endif
 
 static 
 int gf_w4_single_table_init(gf_t *gf)
@@ -916,11 +916,11 @@ gf_w4_bytwo_p_nosse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t
       prod = _mm_xor_si128(prod, t1); \
       v = _mm_srli_epi64(v, 1); }
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w4_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *s8, *d8;
   uint8_t vrev;
@@ -965,8 +965,8 @@ gf_w4_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t v
     s8 += 16;
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 /*
 static
@@ -1034,11 +1034,11 @@ gf_w4_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t v
 }
 */
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_2_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1057,14 +1057,14 @@ gf_w4_bytwo_b_sse_region_2_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_2_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1085,14 +1085,14 @@ gf_w4_bytwo_b_sse_region_2_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_4_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1112,14 +1112,14 @@ gf_w4_bytwo_b_sse_region_4_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_4_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1141,15 +1141,15 @@ gf_w4_bytwo_b_sse_region_4_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_3_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1170,14 +1170,14 @@ gf_w4_bytwo_b_sse_region_3_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_3_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1198,14 +1198,14 @@ gf_w4_bytwo_b_sse_region_3_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_5_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1227,14 +1227,14 @@ gf_w4_bytwo_b_sse_region_5_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_5_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1256,14 +1256,14 @@ gf_w4_bytwo_b_sse_region_5_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_7_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1286,14 +1286,14 @@ gf_w4_bytwo_b_sse_region_7_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_7_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1316,14 +1316,14 @@ gf_w4_bytwo_b_sse_region_7_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_6_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1345,14 +1345,14 @@ gf_w4_bytwo_b_sse_region_6_noxor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static 
 void
 gf_w4_bytwo_b_sse_region_6_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1374,14 +1374,14 @@ gf_w4_bytwo_b_sse_region_6_xor(gf_region_data *rd, struct gf_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static
 void 
 gf_w4_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
   struct gf_bytwo_data *btd;
@@ -1489,8 +1489,8 @@ gf_w4_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t v
     }
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 static
 void 
diff --git a/src/gf_w64.c b/src/gf_w64.c
index 332bd5d..37bac5e 100644
--- a/src/gf_w64.c
+++ b/src/gf_w64.c
@@ -87,6 +87,7 @@ xor)
   }
 }
 
+#if defined(INTEL_SSE4_PCLMUL) 
 static
 void
 gf_w64_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, gf_val_64_t val, int bytes, int
@@ -96,7 +97,6 @@ xor)
   gf_val_64_t *s64, *d64, *top;
   gf_region_data rd;
 
-#if defined(INTEL_SSE4_PCLMUL) 
   __m128i         a, b;
   __m128i         result, r1;
   __m128i         prim_poly;
@@ -175,9 +175,10 @@ xor)
     }
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
+#if defined(INTEL_SSE4_PCLMUL)
 static
 void
 gf_w64_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, gf_val_64_t val, int bytes, int
@@ -187,7 +188,6 @@ xor)
   gf_val_64_t *s64, *d64, *top;
   gf_region_data rd;
 
-#if defined(INTEL_SSE4_PCLMUL)
   __m128i         a, b;
   __m128i         result, r1;
   __m128i         prim_poly;
@@ -263,8 +263,8 @@ xor)
     }
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 static
   inline
@@ -1369,11 +1369,11 @@ void gf_w64_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_
 #endif
 }
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w64_bytwo_b_sse_region_2_xor(gf_region_data *rd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint64_t one64, amask;
   uint8_t *d8, *s8, tb;
@@ -1400,14 +1400,14 @@ gf_w64_bytwo_b_sse_region_2_xor(gf_region_data *rd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w64_bytwo_b_sse_region_2_noxor(gf_region_data *rd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint64_t one64, amask;
   uint8_t *d8, *s8, tb;
@@ -1432,14 +1432,14 @@ gf_w64_bytwo_b_sse_region_2_noxor(gf_region_data *rd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static
 void
 gf_w64_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_64_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   uint64_t itb, amask, one64;
   uint8_t *d8, *s8;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1490,8 +1490,8 @@ gf_w64_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_64_t
   }
 
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 
 static
@@ -1712,11 +1712,11 @@ int gf_w64_composite_init(gf_t *gf)
   return 1;
 }
 
+#ifdef INTEL_SSSE3
 static
   void
 gf_w64_split_4_64_lazy_sse_altmap_multiply_region(gf_t *gf, void *src, void *dest, uint64_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSSE3
   gf_internal_t *h;
   int i, m, j, k, tindex;
   uint64_t pp, v, s, *s64, *d64, *top;
@@ -1793,14 +1793,14 @@ gf_w64_split_4_64_lazy_sse_altmap_multiply_region(gf_t *gf, void *src, void *des
     }
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE4
 static
   void
 gf_w64_split_4_64_lazy_sse_multiply_region(gf_t *gf, void *src, void *dest, uint64_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE4
   gf_internal_t *h;
   int i, m, j, k, tindex;
   uint64_t pp, v, s, *s64, *d64, *top;
@@ -1997,8 +1997,8 @@ gf_w64_split_4_64_lazy_sse_multiply_region(gf_t *gf, void *src, void *dest, uint
   }
 
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 #define GF_MULTBY_TWO(p) (((p) & GF_FIRST_BIT) ? (((p) << 1) ^ h->prim_poly) : (p) << 1);
 
diff --git a/src/gf_w8.c b/src/gf_w8.c
index f47c40c..b171639 100644
--- a/src/gf_w8.c
+++ b/src/gf_w8.c
@@ -364,6 +364,7 @@ gf_w8_multiply_region_from_single(gf_t *gf, void *src, void *dest, gf_val_32_t v
   gf_do_final_region_alignment(&rd);
 }
 
+#if defined(INTEL_SSE4_PCLMUL)
 static
 void
 gf_w8_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int
@@ -373,8 +374,6 @@ gf_w8_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, gf_val_
   uint8_t *s8;
   uint8_t *d8;
 
-#if defined(INTEL_SSE4_PCLMUL)
-
   __m128i         a, b;
   __m128i         result;
   __m128i         prim_poly;
@@ -420,9 +419,10 @@ gf_w8_clm_multiply_region_from_single_2(gf_t *gf, void *src, void *dest, gf_val_
     }
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
+#if defined(INTEL_SSE4_PCLMUL)
 static
 void
 gf_w8_clm_multiply_region_from_single_3(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int
@@ -432,8 +432,6 @@ gf_w8_clm_multiply_region_from_single_3(gf_t *gf, void *src, void *dest, gf_val_
   uint8_t *s8;
   uint8_t *d8;
 
-#if defined(INTEL_SSE4_PCLMUL)
-
   __m128i         a, b;
   __m128i         result;
   __m128i         prim_poly;
@@ -483,9 +481,10 @@ gf_w8_clm_multiply_region_from_single_3(gf_t *gf, void *src, void *dest, gf_val_
     }
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
+#if defined(INTEL_SSE4_PCLMUL)
 static
 void
 gf_w8_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int
@@ -495,8 +494,6 @@ gf_w8_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, gf_val_
   uint8_t *s8;
   uint8_t *d8;
 
-#if defined(INTEL_SSE4_PCLMUL)
-
   __m128i         a, b;
   __m128i         result;
   __m128i         prim_poly;
@@ -550,8 +547,8 @@ gf_w8_clm_multiply_region_from_single_4(gf_t *gf, void *src, void *dest, gf_val_
     }
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 /* ------------------------------------------------------------
 IMPLEMENTATION: SHIFT:
@@ -588,11 +585,11 @@ gf_w8_shift_multiply (gf_t *gf, uint32_t a8, uint32_t b8)
 static 
 int gf_w8_cfm_init(gf_t *gf)
 { 
+#if defined(INTEL_SSE4_PCLMUL)
   gf_internal_t *h;
 
   h = (gf_internal_t *) gf->scratch;
 
-#if defined(INTEL_SSE4_PCLMUL)
     if ((0xe0 & h->prim_poly) == 0){
       gf->multiply.w32 = gf_w8_clm_multiply_2;
       gf->multiply_region.w32 = gf_w8_clm_multiply_region_from_single_2;
@@ -941,6 +938,7 @@ gf_w8_default_multiply(gf_t *gf, gf_val_32_t a, gf_val_32_t b)
   return (ftd->multtable[a][b]);
 }
 
+#ifdef INTEL_SSSE3
 static
   gf_val_32_t
 gf_w8_default_divide(gf_t *gf, gf_val_32_t a, gf_val_32_t b)
@@ -950,6 +948,7 @@ gf_w8_default_divide(gf_t *gf, gf_val_32_t a, gf_val_32_t b)
   ftd = (struct gf_w8_default_data *) ((gf_internal_t *) gf->scratch)->private;
   return (ftd->divtable[a][b]);
 }
+#endif
 
 static
   gf_val_32_t
@@ -1054,11 +1053,11 @@ gf_w8_table_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, in
   }
 }
 
+#ifdef INTEL_SSSE3
 static
   void
 gf_w8_split_multiply_region_sse(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSSE3
   uint8_t *s8, *d8, *bh, *bl, *sptr, *dptr, *top;
   __m128i  tbl, loset, t1, r, va, mth, mtl;
   uint64_t altable[4];
@@ -1114,8 +1113,8 @@ gf_w8_split_multiply_region_sse(gf_t *gf, void *src, void *dest, gf_val_32_t val
   }
 
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 
 /* ------------------------------------------------------------
@@ -1669,11 +1668,11 @@ gf_w8_bytwo_p_nosse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t
   prod = _mm_xor_si128(prod, t1); \
   v = _mm_srli_epi64(v, 1); }
 
+#ifdef INTEL_SSE2
 static
   void 
 gf_w8_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *s8, *d8;
   uint8_t vrev;
@@ -1722,14 +1721,14 @@ gf_w8_bytwo_p_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t v
     s8 += 16;
   }
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static
   void
 gf_w8_bytwo_b_sse_region_2_noxor(gf_region_data *rd, struct gf_w8_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1748,14 +1747,14 @@ gf_w8_bytwo_b_sse_region_2_noxor(gf_region_data *rd, struct gf_w8_bytwo_data *bt
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
+#ifdef INTEL_SSE2
 static
   void
 gf_w8_bytwo_b_sse_region_2_xor(gf_region_data *rd, struct gf_w8_bytwo_data *btd)
 {
-#ifdef INTEL_SSE2
   int i;
   uint8_t *d8, *s8, tb;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1776,15 +1775,15 @@ gf_w8_bytwo_b_sse_region_2_xor(gf_region_data *rd, struct gf_w8_bytwo_data *btd)
     d8 += 16;
     s8 += 16;
   }
-#endif
 }
+#endif
 
 
+#ifdef INTEL_SSE2
 static
   void 
 gf_w8_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t val, int bytes, int xor)
 {
-#ifdef INTEL_SSE2
   int itb;
   uint8_t *d8, *s8;
   __m128i pp, m1, m2, t1, t2, va, vb;
@@ -1832,8 +1831,8 @@ gf_w8_bytwo_b_sse_multiply_region(gf_t *gf, void *src, void *dest, gf_val_32_t v
   }
 
   gf_do_final_region_alignment(&rd);
-#endif
 }
+#endif
 
 static
   void