re PR target/85323 (SSE/AVX/AVX512 shift by 0 not optimized away)

	PR target/85323
	* config/i386/i386.c: Include tree-vector-builder.h.
	(ix86_vector_shift_count): New function.
	(ix86_fold_builtin): Fold shift builtins by scalar count.
	(ix86_gimple_fold_builtin): Likewise.

	* gcc.target/i386/pr85323-1.c: New test.
	* gcc.target/i386/pr85323-2.c: New test.
	* gcc.target/i386/pr85323-3.c: New test.

From-SVN: r260311

gcc/ChangeLog

 2018-05-17  Jakub Jelinek  <jakub@redhat.com>
 
+	PR target/85323
+	* config/i386/i386.c: Include tree-vector-builder.h.
+	(ix86_vector_shift_count): New function.
+	(ix86_fold_builtin): Fold shift builtins by scalar count.
+	(ix86_gimple_fold_builtin): Likewise.
+
 	* config/i386/avx512fintrin.h (_mm512_set_epi16, _mm512_set_epi8,
 	_mm512_setzero): New intrinsics.
 

gcc/testsuite/ChangeLog

 2018-05-17  Jakub Jelinek  <jakub@redhat.com>
 
+	PR target/85323
+	* gcc.target/i386/pr85323-1.c: New test.
+	* gcc.target/i386/pr85323-2.c: New test.
+	* gcc.target/i386/pr85323-3.c: New test.
+
 	* gcc.target/i386/avx512f-set-v32hi-1.c: New test.
 	* gcc.target/i386/avx512f-set-v32hi-2.c: New test.
 	* gcc.target/i386/avx512f-set-v32hi-3.c: New test.

gcc/testsuite/gcc.target/i386/pr85323-1.c

+/* PR target/85323 */
+/* { dg-do compile } */
+/* { dg-options "-O2 -mavx512vl -mavx512bw -fdump-tree-optimized" } */
+/* { dg-final { scan-tree-dump-times "return x_\[0-9]*.D.;" 3 "optimized" } } */
+
+#include <x86intrin.h>
+
+__m128i
+foo (__m128i x)
+{
+  x = _mm_sll_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm_sll_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm_sll_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm_srl_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm_srl_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm_srl_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm_sra_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm_sra_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm_sra_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm_slli_epi64 (x, 0);
+  x = _mm_slli_epi32 (x, 0);
+  x = _mm_slli_epi16 (x, 0);
+  x = _mm_srli_epi64 (x, 0);
+  x = _mm_srli_epi32 (x, 0);
+  x = _mm_srli_epi16 (x, 0);
+  x = _mm_srai_epi64 (x, 0);
+  x = _mm_srai_epi32 (x, 0);
+  x = _mm_srai_epi16 (x, 0);
+  return x;
+}
+
+__m256i
+bar (__m256i x)
+{
+  x = _mm256_sll_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm256_sll_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm256_sll_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm256_srl_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm256_srl_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm256_srl_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm256_sra_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm256_sra_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm256_sra_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm256_slli_epi64 (x, 0);
+  x = _mm256_slli_epi32 (x, 0);
+  x = _mm256_slli_epi16 (x, 0);
+  x = _mm256_srli_epi64 (x, 0);
+  x = _mm256_srli_epi32 (x, 0);
+  x = _mm256_srli_epi16 (x, 0);
+  x = _mm256_srai_epi64 (x, 0);
+  x = _mm256_srai_epi32 (x, 0);
+  x = _mm256_srai_epi16 (x, 0);
+  return x;
+}
+
+__m512i
+baz (__m512i x)
+{
+  x = _mm512_sll_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm512_sll_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm512_sll_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm512_srl_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm512_srl_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm512_srl_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm512_sra_epi64 (x, _mm_set1_epi32 (0));
+  x = _mm512_sra_epi32 (x, _mm_set1_epi32 (0));
+  x = _mm512_sra_epi16 (x, _mm_set1_epi32 (0));
+  x = _mm512_slli_epi64 (x, 0);
+  x = _mm512_slli_epi32 (x, 0);
+  x = _mm512_slli_epi16 (x, 0);
+  x = _mm512_srli_epi64 (x, 0);
+  x = _mm512_srli_epi32 (x, 0);
+  x = _mm512_srli_epi16 (x, 0);
+  x = _mm512_srai_epi64 (x, 0);
+  x = _mm512_srai_epi32 (x, 0);
+  x = _mm512_srai_epi16 (x, 0);
+  return x;
+}

gcc/testsuite/gcc.target/i386/pr85323-2.c

+/* PR target/85323 */
+/* { dg-do compile } */
+/* { dg-options "-O2 -mavx512vl -mavx512bw -fdump-tree-optimized" } */
+/* { dg-final { scan-tree-dump-times "= \{ 0, 0 \};" 12 "optimized" } } */
+/* { dg-final { scan-tree-dump-times "= \{ 0, 0, 0, 0 \};" 12 "optimized" } } */
+/* { dg-final { scan-tree-dump-times "= \{ 0, 0, 0, 0, 0, 0, 0, 0 \};" 12 "optimized" } } */
+
+#include <x86intrin.h>
+
+void
+foo (__m128i x[12])
+{
+  x[0] = _mm_sll_epi64 (x[0], _mm_set1_epi64x (64));
+  x[1] = _mm_sll_epi32 (x[1], _mm_set1_epi64x (32));
+  x[2] = _mm_sll_epi16 (x[2], _mm_set1_epi64x (16));
+  x[3] = _mm_srl_epi64 (x[3], _mm_set1_epi64x (65));
+  x[4] = _mm_srl_epi32 (x[4], _mm_set1_epi64x (33));
+  x[5] = _mm_srl_epi16 (x[5], _mm_set1_epi64x (17));
+  x[6] = _mm_slli_epi64 (x[6], 66);
+  x[7] = _mm_slli_epi32 (x[7], 34);
+  x[8] = _mm_slli_epi16 (x[8], 18);
+  x[9] = _mm_srli_epi64 (x[9], 67);
+  x[10] = _mm_srli_epi32 (x[10], 35);
+  x[11] = _mm_srli_epi16 (x[11], 19);
+}
+
+void
+bar (__m256i x[12])
+{
+  x[0] = _mm256_sll_epi64 (x[0], _mm_set1_epi64x (64));
+  x[1] = _mm256_sll_epi32 (x[1], _mm_set1_epi64x (32));
+  x[2] = _mm256_sll_epi16 (x[2], _mm_set1_epi64x (16));
+  x[3] = _mm256_srl_epi64 (x[3], _mm_set1_epi64x (65));
+  x[4] = _mm256_srl_epi32 (x[4], _mm_set1_epi64x (33));
+  x[5] = _mm256_srl_epi16 (x[5], _mm_set1_epi64x (17));
+  x[6] = _mm256_slli_epi64 (x[6], 66);
+  x[7] = _mm256_slli_epi32 (x[7], 34);
+  x[8] = _mm256_slli_epi16 (x[8], 18);
+  x[9] = _mm256_srli_epi64 (x[9], 67);
+  x[10] = _mm256_srli_epi32 (x[10], 35);
+  x[11] = _mm256_srli_epi16 (x[11], 19);
+}
+
+void
+baz (__m512i x[12])
+{
+  x[0] = _mm512_sll_epi64 (x[0], _mm_set1_epi64x (64));
+  x[1] = _mm512_sll_epi32 (x[1], _mm_set1_epi64x (32));
+  x[2] = _mm512_sll_epi16 (x[2], _mm_set1_epi64x (16));
+  x[3] = _mm512_srl_epi64 (x[3], _mm_set1_epi64x (65));
+  x[4] = _mm512_srl_epi32 (x[4], _mm_set1_epi64x (33));
+  x[5] = _mm512_srl_epi16 (x[5], _mm_set1_epi64x (17));
+  x[6] = _mm512_slli_epi64 (x[6], 66);
+  x[7] = _mm512_slli_epi32 (x[7], 34);
+  x[8] = _mm512_slli_epi16 (x[8], 18);
+  x[9] = _mm512_srli_epi64 (x[9], 67);
+  x[10] = _mm512_srli_epi32 (x[10], 35);
+  x[11] = _mm512_srli_epi16 (x[11], 19);
+}