vppinfra: add universal array mask_compare and compress funcs

Type: improvement
Change-Id: I6d812339f626ea630ad9354632d2f9506122d379
Signed-off-by: Damjan Marion <damarion@cisco.com>

1 files changed, 183 insertions, 17 deletions

diff --git a/src/vppinfra/vector_funcs.h b/src/vppinfra/vector_funcs.h
index 5c446a5d50d..fed1b5cc3d8 100644
--- a/src/vppinfra/vector_funcs.h
+++ b/src/vppinfra/vector_funcs.h
@@ -5,16 +5,10 @@
 #ifndef included_vector_funcs_h
 #define included_vector_funcs_h
 #include <vppinfra/clib.h>
-
-/** \brief Compare 64 16-bit elemments with provied value and return bitmap
-
-    @param v value to compare elements with
-    @param a array of 64 u16 elements
-    @return u64 bitmap where each bit represents result of comparison
-*/
+#include <vppinfra/memcpy.h>
 
 static_always_inline u64
-clib_compare_u16_x64 (u16 v, u16 *a)
+clib_mask_compare_u16_x64 (u16 v, u16 *a, u32 n_elts)
 {
   u64 mask = 0;
 #if defined(CLIB_HAVE_VEC512)
@@ -42,14 +36,9 @@ clib_compare_u16_x64 (u16 v, u16 *a)
      of that bit represend element index. Finally vaddvq_u16() gives us sum
      of all elements of the vector which will give us u8 bitmap. */
 
-  mask = ((u64) vaddvq_u16 ((av[0] == v8) & m) |
-	  (u64) vaddvq_u16 ((av[1] == v8) & m) << 8 |
-	  (u64) vaddvq_u16 ((av[2] == v8) & m) << 16 |
-	  (u64) vaddvq_u16 ((av[3] == v8) & m) << 24 |
-	  (u64) vaddvq_u16 ((av[4] == v8) & m) << 32 |
-	  (u64) vaddvq_u16 ((av[5] == v8) & m) << 40 |
-	  (u64) vaddvq_u16 ((av[6] == v8) & m) << 48 |
-	  (u64) vaddvq_u16 ((av[7] == v8) & m) << 56);
+  for (int i = 0; i < 8; i++)
+    mask |= (u64) vaddvq_u16 ((av[i] == v8) & m) << (i * 8);
+
 #elif defined(CLIB_HAVE_VEC128) && defined(CLIB_HAVE_VEC128_MSB_MASK)
   u16x8 v8 = u16x8_splat (v);
   u16x8u *av = (u16x8u *) a;
@@ -58,11 +47,188 @@ clib_compare_u16_x64 (u16 v, u16 *a)
 	  (u64) i8x16_msb_mask (i8x16_pack (v8 == av[4], v8 == av[5])) << 32 |
 	  (u64) i8x16_msb_mask (i8x16_pack (v8 == av[6], v8 == av[7])) << 48);
 #else
-  for (int i = 0; i < 64; i++)
+  for (int i = 0; i < n_elts; i++)
     if (a[i] == v)
       mask |= 1ULL << i;
 #endif
   return mask;
 }
 
+/** \brief Compare 16-bit elemments with provied value and return bitmap
+
+    @param v value to compare elements with
+    @param a array of u16 elements
+    @param mask array of u64 where reuslting mask will be stored
+    @param n_elts number of elements in the array
+    @return none
+*/
+
+static_always_inline void
+clib_mask_compare_u16 (u16 v, u16 *a, u64 *mask, u32 n_elts)
+{
+  while (n_elts >= 64)
+    {
+      mask++[0] = clib_mask_compare_u16_x64 (v, a, 64);
+      n_elts -= 64;
+      a += 64;
+    }
+
+  if (PREDICT_TRUE (n_elts == 0))
+    return;
+
+  mask[0] = clib_mask_compare_u16_x64 (v, a, n_elts) & pow2_mask (n_elts);
+}
+
+static_always_inline u64
+clib_mask_compare_u32_x64 (u32 v, u32 *a, u32 n_elts)
+{
+  u64 mask = 0;
+#if defined(CLIB_HAVE_VEC512)
+  u32x16 v16 = u32x16_splat (v);
+  u32x16u *av = (u32x16u *) a;
+  mask = ((u64) u32x16_is_equal_mask (av[0], v16) |
+	  (u64) u32x16_is_equal_mask (av[1], v16) << 16 |
+	  (u64) u32x16_is_equal_mask (av[2], v16) << 32 |
+	  (u64) u32x16_is_equal_mask (av[3], v16) << 48);
+#elif defined(CLIB_HAVE_VEC256)
+  u32x8 v8 = u32x8_splat (v);
+  u32x8u *av = (u32x8u *) a;
+  u32x8 m = { 0, 4, 1, 5, 2, 6, 3, 7 };
+  i8x32 c;
+
+  c = i8x32_pack (i16x16_pack ((i32x8) (v8 == av[0]), (i32x8) (v8 == av[1])),
+		  i16x16_pack ((i32x8) (v8 == av[2]), (i32x8) (v8 == av[3])));
+  mask = i8x32_msb_mask ((i8x32) u32x8_permute ((u32x8) c, m));
+
+  c = i8x32_pack (i16x16_pack ((i32x8) (v8 == av[4]), (i32x8) (v8 == av[5])),
+		  i16x16_pack ((i32x8) (v8 == av[6]), (i32x8) (v8 == av[7])));
+  mask |= (u64) i8x32_msb_mask ((i8x32) u32x8_permute ((u32x8) c, m)) << 32;
+
+#elif defined(CLIB_HAVE_VEC128) && defined(__ARM_NEON)
+  u32x4 v4 = u32x4_splat (v);
+  u32x4 m = { 1, 2, 4, 8 };
+  u32x4u *av = (u32x4u *) a;
+
+  /* compare each u32 elemment with v4, result gives -1 in each element
+     of the resulting vector if comparison result is true.
+     Bitwise AND with m will give us one bit set for true result and offset
+     of that bit represend element index. Finally vaddvq_u32() gives us sum
+     of all elements of the vector which will give us u8 bitmap. */
+
+  for (int i = 0; i < 16; i++)
+    mask |= (u64) vaddvq_u32 ((av[i] == v4) & m) << (i * 4);
+
+#elif defined(CLIB_HAVE_VEC128) && defined(CLIB_HAVE_VEC128_MSB_MASK)
+  u32x4 v4 = u32x4_splat (v);
+  u32x4u *av = (u32x4u *) a;
+
+  for (int i = 0; i < 4; i++)
+    {
+      i16x8 p1 = i16x8_pack (v4 == av[0], v4 == av[1]);
+      i16x8 p2 = i16x8_pack (v4 == av[2], v4 == av[3]);
+      mask |= (u64) i8x16_msb_mask (i8x16_pack (p1, p2)) << (i * 16);
+      av += 4;
+    }
+
+#else
+  for (int i = 0; i < n_elts; i++)
+    if (a[i] == v)
+      mask |= 1ULL << i;
+#endif
+  return mask;
+}
+
+/** \brief Compare 32-bit elemments with provied value and return bitmap
+
+    @param v value to compare elements with
+    @param a array of u32 elements
+    @param mask array of u64 where reuslting mask will be stored
+    @param n_elts number of elements in the array
+    @return none
+*/
+
+static_always_inline void
+clib_mask_compare_u32 (u32 v, u32 *a, u64 *bitmap, u32 n_elts)
+{
+  while (n_elts >= 64)
+    {
+      bitmap++[0] = clib_mask_compare_u32_x64 (v, a, 64);
+      n_elts -= 64;
+      a += 64;
+    }
+
+  if (PREDICT_TRUE (n_elts == 0))
+    return;
+
+  bitmap[0] = clib_mask_compare_u32_x64 (v, a, n_elts) & pow2_mask (n_elts);
+}
+
+static_always_inline u32 *
+clib_compress_u32_x64 (u32 *dst, u32 *src, u64 mask)
+{
+#if defined(CLIB_HAVE_VEC512_COMPRESS)
+  u32x16u *sv = (u32x16u *) src;
+  for (int i = 0; i < 4; i++)
+    {
+      int cnt = _popcnt32 ((u16) mask);
+      u32x16_compress_store (sv[i], mask, dst);
+      dst += cnt;
+      mask >>= 16;
+    }
+
+#elif defined(CLIB_HAVE_VEC256_COMPRESS)
+  u32x8u *sv = (u32x8u *) src;
+  for (int i = 0; i < 8; i++)
+    {
+      int cnt = _popcnt32 ((u8) mask);
+      u32x8_compress_store (sv[i], mask, dst);
+      dst += cnt;
+      mask >>= 8;
+    }
+#else
+  while (mask)
+    {
+      u16 bit = count_trailing_zeros (mask);
+      mask = clear_lowest_set_bit (mask);
+      dst++[0] = src[bit];
+    }
+#endif
+  return dst;
+}
+
+/** \brief Compare array of 32-bit elemments into destination array based on
+ * mask
+
+    @param dst destination array of u32 elements
+    @param src source array of u32 elements
+    @param mask array of u64 values representing compress mask
+    @param n_elts number of elements in the source array
+    @return number of elements stored in destionation array
+*/
+
+static_always_inline u32
+clib_compress_u32 (u32 *dst, u32 *src, u64 *mask, u32 n_elts)
+{
+  u32 *dst0 = dst;
+  while (n_elts >= 64)
+    {
+      if (mask[0] == ~0ULL)
+	{
+	  clib_memcpy_u32 (dst, src, 64);
+	  dst += 64;
+	}
+      else
+	dst = clib_compress_u32_x64 (dst, src, mask[0]);
+
+      mask++;
+      src += 64;
+      n_elts -= 64;
+    }
+
+  if (PREDICT_TRUE (n_elts == 0))
+    return dst - dst0;
+
+  return clib_compress_u32_x64 (dst, src, mask[0] & pow2_mask (n_elts)) - dst0;
+}
+
 #endif