openmp: Fix up vectorization simd call badness computation [PR99100]

As mentioned in the PR, ix86_simd_clone_usable didn't make it more desirable
to use 'e' mangled AVX512F entrypoints over 'd' mangled ones (AVX2) with the
same simdlen.  This patch fixes that.  I have tweaked the generic code too
to make more room for these target specific badness factors.

2021-02-16  Jakub Jelinek  <jakub@redhat.com>

	PR target/99100
	* tree-vect-stmts.c (vectorizable_simd_clone_call): For num_calls != 1
	multiply by 4096 and for inbranch by 8192.
	* config/i386/i386.c (ix86_simd_clone_usable): For TARGET_AVX512F,
	return 3, 2 or 1 for mangle letters 'b', 'c' or 'd'.

	* gcc.target/i386/pr99100.c: New test.

diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c
index 48f9aa0..2fe182f 100644 (file)
--- a/gcc/config/i386/i386.c
+++ b/gcc/config/i386/i386.c
@@ -22657,15 +22657,15 @@ ix86_simd_clone_usable (struct cgraph_node *node)
        return -1;
       if (!TARGET_AVX)
        return 0;
-      return TARGET_AVX2 ? 2 : 1;
+      return TARGET_AVX512F ? 3 : TARGET_AVX2 ? 2 : 1;
     case 'c':
       if (!TARGET_AVX)
        return -1;
-      return TARGET_AVX2 ? 1 : 0;
+      return TARGET_AVX512F ? 2 : TARGET_AVX2 ? 1 : 0;
     case 'd':
       if (!TARGET_AVX2)
        return -1;
-      return 0;
+      return TARGET_AVX512F ? 1 : 0;
     case 'e':
       if (!TARGET_AVX512F)
        return -1;

diff --git a/gcc/testsuite/gcc.target/i386/pr99100.c b/gcc/testsuite/gcc.target/i386/pr99100.c
new file mode 100644 (file)
index 0000000..a17fe28
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/pr99100.c
@@ -0,0 +1,22 @@
+/* PR target/99100 */
+/* { dg-do compile } */
+/* { dg-options "-Ofast -mavx512f -fopenmp-simd -mprefer-vector-width=512" } */
+/* { dg-final { scan-assembler "_ZGVeN8v_myfunc" } } */
+/* { dg-final { scan-assembler "_ZGVeN8v_sin" } } */
+
+#pragma omp declare simd notinbranch
+double sin (double x);
+#pragma omp declare simd simdlen(8) notinbranch
+__attribute__((const)) double myfunc (double x);
+
+#define N 1024
+__attribute__((__aligned__ (256))) double a[N], b[N], c[N];
+
+void
+foo ()
+{
+  for (int i = 0; i < N; i++)
+    a[i] = myfunc (b[i]);
+  for (int i = 0; i < N; i++)
+    c[i] = sin (b[i]);
+}

diff --git a/gcc/tree-vect-stmts.c b/gcc/tree-vect-stmts.c
index 083f38b..85d3161 100644 (file)
--- a/gcc/tree-vect-stmts.c
+++ b/gcc/tree-vect-stmts.c
@@ -3914,9 +3914,9 @@ vectorizable_simd_clone_call (vec_info *vinfo, stmt_vec_info stmt_info,
            || n->simdclone->nargs != nargs)
          continue;
        if (num_calls != 1)
-         this_badness += exact_log2 (num_calls) * 1024;
+         this_badness += exact_log2 (num_calls) * 4096;
        if (n->simdclone->inbranch)
-         this_badness += 2048;
+         this_badness += 8192;
        int target_badness = targetm.simd_clone.usable (n);
        if (target_badness < 0)
          continue;