commit a2b8556e3cdc59bbb4a89c2779d2094db975eaa5
parent f034efca545cd215e878f525310910b27db24f36
Author: Sebastiano Tronto <sebastiano@tronto.net>
Date: Thu, 2 Nov 2023 18:44:52 +0100
Added transformations for avx2
Diffstat:
5 files changed, 1035 insertions(+), 16 deletions(-)
diff --git a/README.md b/README.md
@@ -15,7 +15,7 @@ $ make test
### Make AVX2 work
-* generate moves and transformations with scripts in utils/
+* inline moves for avx2
* fix base get_ and set_ macros (constant arguments?)
* optimize things that use get_ and set_
diff --git a/src/_trans_avx2.c b/src/_trans_avx2.c
@@ -0,0 +1,999 @@
+static inline cube_t
+flipallcorners(cube_t c)
+{
+ cube_t shleft, shright, summed, newco, cleanco, ret;
+
+ shleft = _mm256_slli_si256(c, 1);
+ shright = _mm256_srli_si256(c, 1);
+ summed = _mm256_or_si256(shleft, shright);
+ newco = _mm256_and_si256(summed, _co_avx2);
+ cleanco = _mm256_andnot_si256_(c, _co_avx2);
+ ret = _mm256_and_si256(cleanco, newco);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_UFr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 7, 6, 5, 4, 3, 2, 1, 0
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 7, 6, 5, 4, 3, 2, 1, 0
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_ULr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 24, 27, 26, 25, 3, 2, 1, 0, 6, 7, 4, 5,
+ 0, 0, 0, 0, 0, 0, 0, 0, 3, 2, 0, 1, 6, 7, 5, 4
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 26, 25, 24, 27, 2, 3, 0, 1, 7, 6, 5, 4,
+ 0, 0, 0, 0, 0, 0, 0, 0, 2, 3, 1, 0, 7, 6, 4, 5
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_UBr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 9, 8, 11, 10, 6, 7, 4, 5, 2, 3, 0, 1,
+ 0, 0, 0, 0, 0, 0, 0, 0, 6, 7, 4, 5, 2, 3, 0, 1
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 9, 8, 11, 10, 6, 7, 4, 5, 2, 3, 0, 1,
+ 0, 0, 0, 0, 0, 0, 0, 0, 6, 7, 4, 5, 2, 3, 0, 1
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_URr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 26, 25, 24, 27, 2, 3, 0, 1, 7, 6, 5, 4,
+ 0, 0, 0, 0, 0, 0, 0, 0, 2, 3, 1, 0, 7, 6, 4, 5
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 24, 27, 26, 25, 3, 2, 1, 0, 6, 7, 4, 5,
+ 0, 0, 0, 0, 0, 0, 0, 0, 3, 2, 0, 1, 6, 7, 5, 4
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_DFr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 10, 11, 8, 9, 5, 4, 7, 6, 0, 1, 2, 3,
+ 0, 0, 0, 0, 0, 0, 0, 0, 5, 4, 7, 6, 1, 0, 3, 2
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 10, 11, 8, 9, 5, 4, 7, 6, 0, 1, 2, 3,
+ 0, 0, 0, 0, 0, 0, 0, 0, 5, 4, 7, 6, 1, 0, 3, 2
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_DLr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 27, 24, 25, 26, 1, 0, 3, 2, 5, 4, 7, 6,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 3, 2, 5, 4, 6, 7
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 27, 24, 25, 26, 1, 0, 3, 2, 5, 4, 7, 6,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 3, 2, 5, 4, 6, 7
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_DBr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 8, 9, 10, 11, 4, 5, 6, 7, 1, 0, 3, 2,
+ 0, 0, 0, 0, 0, 0, 0, 0, 4, 5, 6, 7, 0, 1, 2, 3
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 8, 9, 10, 11, 4, 5, 6, 7, 1, 0, 3, 2,
+ 0, 0, 0, 0, 0, 0, 0, 0, 4, 5, 6, 7, 0, 1, 2, 3
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_DRr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 25, 26, 27, 24, 0, 1, 2, 3, 4, 5, 6, 7,
+ 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 2, 3, 4, 5, 7, 6
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 25, 26, 27, 24, 0, 1, 2, 3, 4, 5, 6, 7,
+ 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 2, 3, 4, 5, 7, 6
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_RUr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 3, 2, 1, 0, 25, 26, 27, 24, 21, 22, 23, 20,
+ 0, 0, 0, 0, 0, 0, 0, 0, 39, 36, 38, 37, 66, 65, 67, 64
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 21, 22, 23, 20, 17, 18, 19, 16, 11, 10, 9, 8,
+ 0, 0, 0, 0, 0, 0, 0, 0, 71, 69, 68, 70, 33, 35, 34, 32
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_RFr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 18, 17, 16, 19, 22, 21, 20, 23, 25, 26, 27, 24,
+ 0, 0, 0, 0, 0, 0, 0, 0, 65, 66, 67, 64, 39, 36, 37, 38
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 17, 18, 19, 16, 20, 23, 22, 21, 24, 27, 26, 25,
+ 0, 0, 0, 0, 0, 0, 0, 0, 67, 64, 65, 66, 37, 38, 39, 36
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_RDr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 1, 0, 3, 2, 26, 25, 24, 27, 22, 21, 20, 23,
+ 0, 0, 0, 0, 0, 0, 0, 0, 36, 39, 37, 38, 65, 66, 64, 67
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 20, 23, 22, 21, 16, 19, 18, 17, 9, 8, 11, 10,
+ 0, 0, 0, 0, 0, 0, 0, 0, 70, 68, 69, 71, 32, 34, 35, 33
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_RBr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 16, 19, 18, 17, 21, 22, 23, 20, 26, 25, 24, 27,
+ 0, 0, 0, 0, 0, 0, 0, 0, 66, 65, 64, 67, 36, 39, 38, 37
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 16, 19, 18, 17, 21, 22, 23, 20, 26, 25, 24, 27,
+ 0, 0, 0, 0, 0, 0, 0, 0, 66, 65, 64, 67, 36, 39, 38, 37
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_LUr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 2, 3, 0, 1, 27, 24, 25, 26, 20, 23, 22, 21,
+ 0, 0, 0, 0, 0, 0, 0, 0, 38, 37, 39, 36, 67, 64, 66, 65
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 23, 20, 21, 22, 18, 17, 16, 19, 10, 11, 8, 9,
+ 0, 0, 0, 0, 0, 0, 0, 0, 69, 71, 70, 68, 35, 33, 32, 34
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_LFr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 17, 18, 19, 16, 20, 23, 22, 21, 24, 27, 26, 25,
+ 0, 0, 0, 0, 0, 0, 0, 0, 67, 64, 65, 66, 37, 38, 39, 36
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 18, 17, 16, 19, 22, 21, 20, 23, 25, 26, 27, 24,
+ 0, 0, 0, 0, 0, 0, 0, 0, 65, 66, 67, 64, 39, 36, 37, 38
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_LDr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 0, 1, 2, 3, 24, 27, 26, 25, 23, 20, 21, 22,
+ 0, 0, 0, 0, 0, 0, 0, 0, 37, 38, 36, 39, 64, 67, 65, 66
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 22, 21, 20, 23, 19, 16, 17, 18, 8, 9, 10, 11,
+ 0, 0, 0, 0, 0, 0, 0, 0, 68, 70, 71, 69, 34, 32, 33, 35
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_LBr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 19, 16, 17, 18, 23, 20, 21, 22, 27, 24, 25, 26,
+ 0, 0, 0, 0, 0, 0, 0, 0, 64, 67, 66, 65, 38, 37, 36, 39
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 19, 16, 17, 18, 23, 20, 21, 22, 27, 24, 25, 26,
+ 0, 0, 0, 0, 0, 0, 0, 0, 64, 67, 66, 65, 38, 37, 36, 39
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_FUr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 6, 7, 4, 5, 10, 11, 8, 9, 17, 18, 19, 16,
+ 0, 0, 0, 0, 0, 0, 0, 0, 35, 33, 34, 32, 71, 69, 70, 68
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 6, 7, 4, 5, 10, 11, 8, 9, 17, 18, 19, 16,
+ 0, 0, 0, 0, 0, 0, 0, 0, 35, 33, 34, 32, 71, 69, 70, 68
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_FRr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 21, 22, 23, 20, 17, 18, 19, 16, 11, 10, 9, 8,
+ 0, 0, 0, 0, 0, 0, 0, 0, 71, 69, 68, 70, 33, 35, 34, 32
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 3, 2, 1, 0, 25, 26, 27, 24, 21, 22, 23, 20,
+ 0, 0, 0, 0, 0, 0, 0, 0, 39, 36, 38, 37, 66, 65, 67, 64
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_FDr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 4, 5, 6, 7, 11, 10, 9, 8, 18, 17, 16, 19,
+ 0, 0, 0, 0, 0, 0, 0, 0, 33, 35, 32, 34, 69, 71, 68, 70
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 7, 6, 5, 4, 8, 9, 10, 11, 16, 19, 18, 17,
+ 0, 0, 0, 0, 0, 0, 0, 0, 34, 32, 35, 33, 70, 68, 71, 69
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_FLr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 23, 20, 21, 22, 18, 17, 16, 19, 10, 11, 8, 9,
+ 0, 0, 0, 0, 0, 0, 0, 0, 69, 71, 70, 68, 35, 33, 32, 34
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 2, 3, 0, 1, 27, 24, 25, 26, 20, 23, 22, 21,
+ 0, 0, 0, 0, 0, 0, 0, 0, 38, 37, 39, 36, 67, 64, 66, 65
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_BUr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 7, 6, 5, 4, 8, 9, 10, 11, 16, 19, 18, 17,
+ 0, 0, 0, 0, 0, 0, 0, 0, 34, 32, 35, 33, 70, 68, 71, 69
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 4, 5, 6, 7, 11, 10, 9, 8, 18, 17, 16, 19,
+ 0, 0, 0, 0, 0, 0, 0, 0, 33, 35, 32, 34, 69, 71, 68, 70
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_BRr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 22, 21, 20, 23, 19, 16, 17, 18, 8, 9, 10, 11,
+ 0, 0, 0, 0, 0, 0, 0, 0, 68, 70, 71, 69, 34, 32, 33, 35
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 0, 1, 2, 3, 24, 27, 26, 25, 23, 20, 21, 22,
+ 0, 0, 0, 0, 0, 0, 0, 0, 37, 38, 36, 39, 64, 67, 65, 66
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_BDr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 5, 4, 7, 6, 9, 8, 11, 10, 19, 16, 17, 18,
+ 0, 0, 0, 0, 0, 0, 0, 0, 32, 34, 33, 35, 68, 70, 69, 71
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 5, 4, 7, 6, 9, 8, 11, 10, 19, 16, 17, 18,
+ 0, 0, 0, 0, 0, 0, 0, 0, 32, 34, 33, 35, 68, 70, 69, 71
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_BLr(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 20, 23, 22, 21, 16, 19, 18, 17, 9, 8, 11, 10,
+ 0, 0, 0, 0, 0, 0, 0, 0, 70, 68, 69, 71, 32, 34, 35, 33
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 1, 0, 3, 2, 26, 25, 24, 27, 22, 21, 20, 23,
+ 0, 0, 0, 0, 0, 0, 0, 0, 36, 39, 37, 38, 65, 66, 64, 67
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_UFm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 10, 11, 8, 9, 6, 7, 4, 5, 3, 2, 1, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 3, 2, 1, 0, 7, 6, 5, 4
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 10, 11, 8, 9, 6, 7, 4, 5, 3, 2, 1, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 3, 2, 1, 0, 7, 6, 5, 4
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_ULm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 25, 26, 27, 24, 3, 2, 1, 0, 7, 6, 5, 4,
+ 0, 0, 0, 0, 0, 0, 0, 0, 7, 6, 4, 5, 2, 3, 1, 0
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 25, 26, 27, 24, 3, 2, 1, 0, 7, 6, 5, 4,
+ 0, 0, 0, 0, 0, 0, 0, 0, 7, 6, 4, 5, 2, 3, 1, 0
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_UBm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 8, 9, 10, 11, 7, 6, 5, 4, 2, 3, 0, 1,
+ 0, 0, 0, 0, 0, 0, 0, 0, 2, 3, 0, 1, 6, 7, 4, 5
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 8, 9, 10, 11, 7, 6, 5, 4, 2, 3, 0, 1,
+ 0, 0, 0, 0, 0, 0, 0, 0, 2, 3, 0, 1, 6, 7, 4, 5
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_URm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 27, 24, 25, 26, 2, 3, 0, 1, 6, 7, 4, 5,
+ 0, 0, 0, 0, 0, 0, 0, 0, 6, 7, 5, 4, 3, 2, 0, 1
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 27, 24, 25, 26, 2, 3, 0, 1, 6, 7, 4, 5,
+ 0, 0, 0, 0, 0, 0, 0, 0, 6, 7, 5, 4, 3, 2, 0, 1
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_DFm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 11, 10, 9, 8, 4, 5, 6, 7, 0, 1, 2, 3,
+ 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 3, 2, 5, 4, 7, 6
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 11, 10, 9, 8, 4, 5, 6, 7, 0, 1, 2, 3,
+ 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 3, 2, 5, 4, 7, 6
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_DLm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 26, 25, 24, 27, 1, 0, 3, 2, 4, 5, 6, 7,
+ 0, 0, 0, 0, 0, 0, 0, 0, 4, 5, 7, 6, 1, 0, 2, 3
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 24, 27, 26, 25, 0, 1, 2, 3, 5, 4, 7, 6,
+ 0, 0, 0, 0, 0, 0, 0, 0, 5, 4, 6, 7, 0, 1, 3, 2
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_DBm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 9, 8, 11, 10, 5, 4, 7, 6, 1, 0, 3, 2,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 7
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 9, 8, 11, 10, 5, 4, 7, 6, 1, 0, 3, 2,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 7
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_DRm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 24, 27, 26, 25, 0, 1, 2, 3, 5, 4, 7, 6,
+ 0, 0, 0, 0, 0, 0, 0, 0, 5, 4, 6, 7, 0, 1, 3, 2
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 26, 25, 24, 27, 1, 0, 3, 2, 4, 5, 6, 7,
+ 0, 0, 0, 0, 0, 0, 0, 0, 4, 5, 7, 6, 1, 0, 2, 3
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_RUm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 3, 2, 1, 0, 24, 27, 26, 25, 20, 23, 22, 21,
+ 0, 0, 0, 0, 0, 0, 0, 0, 35, 32, 34, 33, 70, 69, 71, 68
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 22, 21, 20, 23, 18, 17, 16, 19, 11, 10, 9, 8,
+ 0, 0, 0, 0, 0, 0, 0, 0, 33, 35, 34, 32, 71, 69, 68, 70
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_RFm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 18, 17, 16, 19, 23, 20, 21, 22, 24, 27, 26, 25,
+ 0, 0, 0, 0, 0, 0, 0, 0, 69, 70, 71, 68, 35, 32, 33, 34
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 18, 17, 16, 19, 23, 20, 21, 22, 24, 27, 26, 25,
+ 0, 0, 0, 0, 0, 0, 0, 0, 37, 38, 39, 36, 67, 64, 65, 66
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_RDm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 1, 0, 3, 2, 27, 24, 25, 26, 23, 20, 21, 22,
+ 0, 0, 0, 0, 0, 0, 0, 0, 32, 35, 33, 34, 69, 70, 68, 71
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 23, 20, 21, 22, 19, 16, 17, 18, 9, 8, 11, 10,
+ 0, 0, 0, 0, 0, 0, 0, 0, 32, 34, 35, 33, 70, 68, 69, 71
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_RBm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 16, 19, 18, 17, 20, 23, 22, 21, 27, 24, 25, 26,
+ 0, 0, 0, 0, 0, 0, 0, 0, 70, 69, 68, 71, 32, 35, 34, 33
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 19, 16, 17, 18, 22, 21, 20, 23, 26, 25, 24, 27,
+ 0, 0, 0, 0, 0, 0, 0, 0, 36, 39, 38, 37, 66, 65, 64, 67
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_LUm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 2, 3, 0, 1, 26, 25, 24, 27, 21, 22, 23, 20,
+ 0, 0, 0, 0, 0, 0, 0, 0, 34, 33, 35, 32, 71, 68, 70, 69
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 20, 23, 22, 21, 17, 18, 19, 16, 10, 11, 8, 9,
+ 0, 0, 0, 0, 0, 0, 0, 0, 35, 33, 32, 34, 69, 71, 70, 68
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_LFm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 17, 18, 19, 16, 21, 22, 23, 20, 25, 26, 27, 24,
+ 0, 0, 0, 0, 0, 0, 0, 0, 71, 68, 69, 70, 33, 34, 35, 32
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 17, 18, 19, 16, 21, 22, 23, 20, 25, 26, 27, 24,
+ 0, 0, 0, 0, 0, 0, 0, 0, 39, 36, 37, 38, 65, 66, 67, 64
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_LDm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 0, 1, 2, 3, 25, 26, 27, 24, 22, 21, 20, 23,
+ 0, 0, 0, 0, 0, 0, 0, 0, 33, 34, 32, 35, 68, 71, 69, 70
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 21, 22, 23, 20, 16, 19, 18, 17, 8, 9, 10, 11,
+ 0, 0, 0, 0, 0, 0, 0, 0, 34, 32, 33, 35, 68, 70, 71, 69
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_LBm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 19, 16, 17, 18, 22, 21, 20, 23, 26, 25, 24, 27,
+ 0, 0, 0, 0, 0, 0, 0, 0, 68, 71, 70, 69, 34, 33, 32, 35
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 16, 19, 18, 17, 20, 23, 22, 21, 27, 24, 25, 26,
+ 0, 0, 0, 0, 0, 0, 0, 0, 38, 37, 36, 39, 64, 67, 66, 65
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_FUm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 7, 6, 5, 4, 11, 10, 9, 8, 17, 18, 19, 16,
+ 0, 0, 0, 0, 0, 0, 0, 0, 39, 37, 38, 36, 67, 65, 66, 64
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 7, 6, 5, 4, 11, 10, 9, 8, 17, 18, 19, 16,
+ 0, 0, 0, 0, 0, 0, 0, 0, 71, 69, 70, 68, 35, 33, 34, 32
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_FRm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 20, 23, 22, 21, 17, 18, 19, 16, 10, 11, 8, 9,
+ 0, 0, 0, 0, 0, 0, 0, 0, 67, 65, 64, 66, 37, 39, 38, 36
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 2, 3, 0, 1, 26, 25, 24, 27, 21, 22, 23, 20,
+ 0, 0, 0, 0, 0, 0, 0, 0, 66, 65, 67, 64, 39, 36, 38, 37
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_FDm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 5, 4, 7, 6, 10, 11, 8, 9, 18, 17, 16, 19,
+ 0, 0, 0, 0, 0, 0, 0, 0, 37, 39, 36, 38, 65, 67, 64, 66
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 6, 7, 4, 5, 9, 8, 11, 10, 16, 19, 18, 17,
+ 0, 0, 0, 0, 0, 0, 0, 0, 70, 68, 71, 69, 34, 32, 35, 33
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_FLm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 22, 21, 20, 23, 18, 17, 16, 19, 11, 10, 9, 8,
+ 0, 0, 0, 0, 0, 0, 0, 0, 65, 67, 66, 64, 39, 37, 36, 38
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 3, 2, 1, 0, 24, 27, 26, 25, 20, 23, 22, 21,
+ 0, 0, 0, 0, 0, 0, 0, 0, 67, 64, 66, 65, 38, 37, 39, 36
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_BUm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 6, 7, 4, 5, 9, 8, 11, 10, 16, 19, 18, 17,
+ 0, 0, 0, 0, 0, 0, 0, 0, 38, 36, 39, 37, 66, 64, 67, 65
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 5, 4, 7, 6, 10, 11, 8, 9, 18, 17, 16, 19,
+ 0, 0, 0, 0, 0, 0, 0, 0, 69, 71, 68, 70, 33, 35, 32, 34
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_BRm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 23, 20, 21, 22, 19, 16, 17, 18, 9, 8, 11, 10,
+ 0, 0, 0, 0, 0, 0, 0, 0, 64, 66, 67, 65, 38, 36, 37, 39
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 1, 0, 3, 2, 27, 24, 25, 26, 23, 20, 21, 22,
+ 0, 0, 0, 0, 0, 0, 0, 0, 64, 67, 65, 66, 37, 38, 36, 39
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_BDm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 4, 5, 6, 7, 8, 9, 10, 11, 19, 16, 17, 18,
+ 0, 0, 0, 0, 0, 0, 0, 0, 36, 38, 37, 39, 64, 66, 65, 67
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 4, 5, 6, 7, 8, 9, 10, 11, 19, 16, 17, 18,
+ 0, 0, 0, 0, 0, 0, 0, 0, 68, 70, 69, 71, 32, 34, 33, 35
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
+static inline cube_t
+inline_trans_BLm(cube_t c)
+{
+ cube_t ret;
+
+ cube_t tn = _mm256_set_epi8(
+ 0, 0, 0, 0, 21, 22, 23, 20, 16, 19, 18, 17, 8, 9, 10, 11,
+ 0, 0, 0, 0, 0, 0, 0, 0, 66, 64, 65, 67, 36, 38, 39, 37
+ );
+ cube_t ti = _mm256_set_epi8(
+ 0, 0, 0, 0, 0, 1, 2, 3, 25, 26, 27, 24, 22, 21, 20, 23,
+ 0, 0, 0, 0, 0, 0, 0, 0, 65, 66, 64, 67, 36, 39, 37, 38
+ );
+
+ ret = compose(tn, c);
+ ret = compose(ret, ti);
+ ret = flipallcorners(ret);
+
+ return ret;
+}
+
diff --git a/src/cube.c b/src/cube.c
@@ -126,10 +126,12 @@ cube_arr_t zerocube_arr = { .e = {0}, .c = {0} };
#ifdef CUBE_AVX2
-#define setsolved(cube) cube = _mm256_loadu_si256((__m256i_u *)&solvedcube_arr)
-#define setzero(cube) cube = _mm256_setzero_si256()
+#define _co_avx2 _mm256_set_epi64x(0, 0xF0F0F0F0F0F0F0F0, 0, 0)
+#define _eo_avx2 _mm256_set_epi64x(0, 0, 0xF0F0F0F0, 0xF0F0F0F0F0F0F0F0)
#define _c _mm256_set_epi64x(0, 0, 0, 0xFF)
#define _e _mm256_set_epi64x(0, 0xFF, 0, 0)
+#define setsolved(cube) cube = _mm256_loadu_si256((__m256i_u *)&solvedcube_arr)
+#define setzero(cube) cube = _mm256_setzero_si256()
/* TODO: in the next 4 macros, all 0 should be i, but must be constant! */
#define get_edge(cube, i) _mm256_extract_epi8(cube, 0+16)
#define get_corner(cube, i) _mm256_extract_epi8(cube, 0)
diff --git a/utils/gentranscode.sh b/utils/gentranscode.sh
@@ -1,6 +1,8 @@
#!/bin/sh
-gcc -DDEBUG h48_to_src.c ../src/cube.c -o h48_to_src
+type="${1:-src}"
+
+gcc -DDEBUG h48_to_"$type".c ../src/cube.c -o h48_to_"$type"
gcc -DDEBUG invert.c ../src/cube.c -o invert
# Old version
@@ -9,9 +11,9 @@ genarray() {
trans="$(echo $f | sed 's/.*_// ; s/\.txt//')"
printf '[%s] = ' "$trans"
if [ "$1" = "-i" ]; then
- ./invert <"$f" | ./h48_to_src
+ ./invert <"$f" | ./h48_to_"$type"
else
- ./h48_to_src <"$f"
+ ./h48_to_"$type" <"$f"
fi
printf ',\n'
done
@@ -23,15 +25,12 @@ genfuncs() {
printf 'static inline cube_t\ninline_trans_%s' "$trans"
[ "$1" = "-i" ] && printf '_inverse'
printf '(cube_t c)\n{\n'
- printf '\tcube_t ret;\n'
- printf '\tcube_t tn = {\n'
- ./h48_to_src <"$f" | sed 's/^/\t/' | \
- tail -n 3 | head -n 2
- printf '\t};\n\tcube_t ti = {\n'
- ./invert <"$f" | ./h48_to_src | sed 's/^/\t/' | \
- tail -n 3 | head -n 2
- printf '\t};\n\n'
- printf '\tret = compose(tn, c);\n'
+ printf '\tcube_t ret;\n\n'
+ printf '\tcube_t tn = '
+ ./h48_to_"$type" <"$f" | sed '2,4s/^/\t/'
+ printf ';\n\tcube_t ti = '
+ ./invert <"$f" | ./h48_to_"$type" | sed '2,4 s/^/\t/'
+ printf ';\n\n\tret = compose(tn, c);\n'
printf '\tret = compose(ret, ti);\n'
if [ -n "$(echo "$trans" | grep "m")" ]; then
printf '\tret = flipallcorners(ret);\n'
@@ -41,4 +40,4 @@ genfuncs() {
}
genfuncs
-rm -f h48_to_src invert
+rm -f h48_to_"$type" invert
diff --git a/utils/h48_to_avx.c b/utils/h48_to_avx.c
@@ -0,0 +1,19 @@
+#include <stdio.h>
+#include <inttypes.h>
+#include <stdbool.h>
+
+#include "../src/cube.h"
+
+#define STRLENMAX 1000
+
+int main() {
+ char str[STRLENMAX];
+ cube_t cube;
+
+ fgets(str, STRLENMAX, stdin);
+ cube = readcube(H48, str);
+ writecube(AVX, cube, str);
+ fputs(str, stdout);
+
+ return 0;
+}
