Avoid is_constant calls in vectorizable_bswap

The "new" VEC_PERM_EXPR handling makes it easy to support bswap
for variable-length vectors.

2018-08-24  Richard Sandiford  <richard.sandiford@arm.com>

gcc/
	* tree-vect-stmts.c (vectorizable_bswap): Handle variable-length
	vectors.

gcc/testsuite/
	* gcc.target/aarch64/sve/bswap_1.c: New test.
	* gcc.target/aarch64/sve/bswap_2.c: Likewise.
	* gcc.target/aarch64/sve/bswap_3.c: Likewise.

From-SVN: r263833

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index ac459b6..aa8f392 100644 (file)
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,5 +1,10 @@
 2018-08-24  Richard Sandiford  <richard.sandiford@arm.com>
 
+       * tree-vect-stmts.c (vectorizable_bswap): Handle variable-length
+       vectors.
+
+2018-08-24  Richard Sandiford  <richard.sandiford@arm.com>
+
        * tree-vect-slp.c (vect_transform_slp_perm_load): Separate out
        the case in which the permute needs only a single element and
        repeats for every vector of the result.  Extend that case to

diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog
index db05964..6520f46 100644 (file)
--- a/gcc/testsuite/ChangeLog
+++ b/gcc/testsuite/ChangeLog
@@ -1,5 +1,11 @@
 2018-08-24  Richard Sandiford  <richard.sandiford@arm.com>
 
+       * gcc.target/aarch64/sve/bswap_1.c: New test.
+       * gcc.target/aarch64/sve/bswap_2.c: Likewise.
+       * gcc.target/aarch64/sve/bswap_3.c: Likewise.
+
+2018-08-24  Richard Sandiford  <richard.sandiford@arm.com>
+
        * gcc.target/aarch64/sve/slp_perm_1.c: New test.
        * gcc.target/aarch64/sve/slp_perm_2.c: Likewise.
        * gcc.target/aarch64/sve/slp_perm_3.c: Likewise.

diff --git a/gcc/testsuite/gcc.target/aarch64/sve/bswap_1.c b/gcc/testsuite/gcc.target/aarch64/sve/bswap_1.c
new file mode 100644 (file)
index 0000000..aa4c033
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/sve/bswap_1.c
@@ -0,0 +1,13 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -ftree-vectorize" } */
+
+#include <stdint.h>
+
+void
+f (uint16_t *a, uint16_t *b)
+{
+  for (int i = 0; i < 100; ++i)
+    a[i] = __builtin_bswap16 (b[i]);
+}
+
+/* { dg-final { scan-assembler-times {\trevb\tz[0-9]+\.h, p[0-7]/m, z[0-9]+\.h\n} 1 { xfail aarch64_big_endian } } } */

diff --git a/gcc/testsuite/gcc.target/aarch64/sve/bswap_2.c b/gcc/testsuite/gcc.target/aarch64/sve/bswap_2.c
new file mode 100644 (file)
index 0000000..442c65e
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/sve/bswap_2.c
@@ -0,0 +1,13 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -ftree-vectorize" } */
+
+#include <stdint.h>
+
+void
+f (uint32_t *a, uint32_t *b)
+{
+  for (int i = 0; i < 100; ++i)
+    a[i] = __builtin_bswap32 (b[i]);
+}
+
+/* { dg-final { scan-assembler-times {\trevb\tz[0-9]+\.s, p[0-7]/m, z[0-9]+\.s\n} 1 { xfail aarch64_big_endian } } } */

diff --git a/gcc/testsuite/gcc.target/aarch64/sve/bswap_3.c b/gcc/testsuite/gcc.target/aarch64/sve/bswap_3.c
new file mode 100644 (file)
index 0000000..75acc64
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/sve/bswap_3.c
@@ -0,0 +1,13 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -ftree-vectorize" } */
+
+#include <stdint.h>
+
+void
+f (uint64_t *a, uint64_t *b)
+{
+  for (int i = 0; i < 100; ++i)
+    a[i] = __builtin_bswap64 (b[i]);
+}
+
+/* { dg-final { scan-assembler-times {\trevb\tz[0-9]+\.d, p[0-7]/m, z[0-9]+\.d\n} 1 { xfail aarch64_big_endian } } } */

diff --git a/gcc/tree-vect-stmts.c b/gcc/tree-vect-stmts.c
index 63fb1fb..8d94fca 100644 (file)
--- a/gcc/tree-vect-stmts.c
+++ b/gcc/tree-vect-stmts.c
@@ -2961,13 +2961,10 @@ vectorizable_bswap (stmt_vec_info stmt_info, gimple_stmt_iterator *gsi,
   vec_info *vinfo = stmt_info->vinfo;
   loop_vec_info loop_vinfo = STMT_VINFO_LOOP_VINFO (stmt_info);
   unsigned ncopies;
-  unsigned HOST_WIDE_INT nunits, num_bytes;
 
   op = gimple_call_arg (stmt, 0);
   vectype = STMT_VINFO_VECTYPE (stmt_info);
-
-  if (!TYPE_VECTOR_SUBPARTS (vectype).is_constant (&nunits))
-    return false;
+  poly_uint64 nunits = TYPE_VECTOR_SUBPARTS (vectype);
 
   /* Multiple types in SLP are handled by creating the appropriate number of
      vectorized stmts for each SLP node.  Hence, NCOPIES is always 1 in
@@ -2983,11 +2980,11 @@ vectorizable_bswap (stmt_vec_info stmt_info, gimple_stmt_iterator *gsi,
   if (! char_vectype)
     return false;
 
-  if (!TYPE_VECTOR_SUBPARTS (char_vectype).is_constant (&num_bytes))
+  poly_uint64 num_bytes = TYPE_VECTOR_SUBPARTS (char_vectype);
+  unsigned word_bytes;
+  if (!constant_multiple_p (num_bytes, nunits, &word_bytes))
     return false;
 
-  unsigned word_bytes = num_bytes / nunits;
-
   /* The encoding uses one stepped pattern for each byte in the word.  */
   vec_perm_builder elts (num_bytes, word_bytes, 3);
   for (unsigned i = 0; i < 3; ++i)