[Patch AArch64] Add some more missing intrinsics

[James Greenhalgh <james dot greenhalgh at arm dot com>]

Hi,

Inspired by Jiong's recent work, here are some more missing intrinsics,
and a smoke test for each of them.

This patch covers:

  vcvt_n_f64_s64
  vcvt_n_f64_u64
  vcvt_n_s64_f64
  vcvt_n_u64_f64
  vcvt_f64_s64
  vrecpe_f64
  vcvt_f64_u64
  vrecps_f64

Tested on aarch64-none-elf, and on an internal testsuite for Neon
intrinsics.

Note that the new tests will ICE without the fixups in
https://gcc.gnu.org/ml/gcc-patches/2016-06/msg00805.html

OK?

Thanks,
James

---
gcc/ChangeLog

2016-06-10  James Greenhalgh  <james.greenhalgh@arm.com>

	* config/aarch64/arm_neon.h (vcvt_n_f64_s64): New.
	(vcvt_n_f64_u64): Likewise.
	(vcvt_n_s64_f64): Likewise.
	(vcvt_n_u64_f64): Likewise.
	(vcvt_f64_s64): Likewise.
	(vrecpe_f64): Likewise.
	(vcvt_f64_u64): Likewise.
	(vrecps_f64): Likewise.

gcc/testsuite/ChangeLog

2016-06-10  James Greenhalgh  <james.greenhalgh@arm.com>

	* gcc.target/aarch64/vcvt_f64_1.c: New.
	* gcc.target/aarch64/vcvt_n_f64_1.c: New.
	* gcc.target/aarch64/vrecp_f64_1.c: New.

diff --git a/gcc/config/aarch64/arm_neon.h b/gcc/config/aarch64/arm_neon.h
index f70b6d3..2f90938 100644
--- a/gcc/config/aarch64/arm_neon.h
+++ b/gcc/config/aarch64/arm_neon.h
@@ -12447,6 +12447,20 @@ vcvt_n_f32_u32 (uint32x2_t __a, const int __b)
   return __builtin_aarch64_ucvtfv2si_sus (__a, __b);
 }
 
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vcvt_n_f64_s64 (int64x1_t __a, const int __b)
+{
+  return (float64x1_t)
+    { __builtin_aarch64_scvtfdi (vget_lane_s64 (__a, 0), __b) };
+}
+
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vcvt_n_f64_u64 (uint64x1_t __a, const int __b)
+{
+  return (float64x1_t)
+    { __builtin_aarch64_ucvtfdi_sus (vget_lane_u64 (__a, 0), __b) };
+}
+
 __extension__ static __inline float32x4_t __attribute__ ((__always_inline__))
 vcvtq_n_f32_s32 (int32x4_t __a, const int __b)
 {
@@ -12509,6 +12523,20 @@ vcvt_n_u32_f32 (float32x2_t __a, const int __b)
   return __builtin_aarch64_fcvtzuv2sf_uss (__a, __b);
 }
 
+__extension__ static __inline int64x1_t __attribute__ ((__always_inline__))
+vcvt_n_s64_f64 (float64x1_t __a, const int __b)
+{
+  return (int64x1_t)
+    { __builtin_aarch64_fcvtzsdf (vget_lane_f64 (__a, 0), __b) };
+}
+
+__extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
+vcvt_n_u64_f64 (float64x1_t __a, const int __b)
+{
+  return (uint64x1_t)
+    { __builtin_aarch64_fcvtzudf_uss (vget_lane_f64 (__a, 0), __b) };
+}
+
 __extension__ static __inline int32x4_t __attribute__ ((__always_inline__))
 vcvtq_n_s32_f32 (float32x4_t __a, const int __b)
 {
@@ -12571,6 +12599,18 @@ vcvt_f32_u32 (uint32x2_t __a)
   return __builtin_aarch64_floatunsv2siv2sf ((int32x2_t) __a);
 }
 
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vcvt_f64_s64 (int64x1_t __a)
+{
+  return (float64x1_t) { vget_lane_s64 (__a, 0) };
+}
+
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vcvt_f64_u64 (uint64x1_t __a)
+{
+  return (float64x1_t) { vget_lane_u64 (__a, 0) };
+}
+
 __extension__ static __inline float32x4_t __attribute__ ((__always_inline__))
 vcvtq_f32_s32 (int32x4_t __a)
 {
@@ -20659,6 +20699,12 @@ vrecpe_f32 (float32x2_t __a)
   return __builtin_aarch64_frecpev2sf (__a);
 }
 
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vrecpe_f64 (float64x1_t __a)
+{
+  return (float64x1_t) { vrecped_f64 (vget_lane_f64 (__a, 0)) };
+}
+
 __extension__ static __inline float32x4_t __attribute__ ((__always_inline__))
 vrecpeq_f32 (float32x4_t __a)
 {
@@ -20691,6 +20737,13 @@ vrecps_f32 (float32x2_t __a, float32x2_t __b)
   return __builtin_aarch64_frecpsv2sf (__a, __b);
 }
 
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vrecps_f64 (float64x1_t __a, float64x1_t __b)
+{
+  return (float64x1_t) { vrecpsd_f64  (vget_lane_f64 (__a, 0),
+				       vget_lane_f64 (__b, 0)) };
+}
+
 __extension__ static __inline float32x4_t __attribute__ ((__always_inline__))
 vrecpsq_f32 (float32x4_t __a, float32x4_t __b)
 {
diff --git a/gcc/testsuite/gcc.target/aarch64/vcvt_f64_1.c b/gcc/testsuite/gcc.target/aarch64/vcvt_f64_1.c
new file mode 100644
index 0000000..b7ee7af
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/vcvt_f64_1.c
@@ -0,0 +1,48 @@
+/* { dg-do compile } */
+/* { dg-options "-O2" } */
+
+#include "arm_neon.h"
+
+/* For each of these intrinsics, we're mapping to a simple C cast.
+   While the compiler has some freedom in terms of choice of instruction,
+   we'd hope that for this simple case it would always pick the single
+   instruction form given in these tests.  Anything else is likely a
+   regression, so check for an exact instruction pattern and
+   register allocation decision.  */
+
+/* Test that if we have a value already in Advanced-SIMD registers, we use
+   the scalar register forms.  */
+
+float64x1_t
+test_vcvt_f64_s64_fpr (int64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "scvtf\td0, d0" 1 } } */
+  return vcvt_f64_s64 (a);
+}
+
+float64x1_t
+test_vcvt_f64_u64_fpr (uint64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "ucvtf\td0, d0" 1 } } */
+  return vcvt_f64_u64 (a);
+}
+
+/* Test that if we have an integer constructor, we use the general-purpose
+   register forms for scvtf and ucvtf.  */
+
+float64x1_t
+test_vcvt_f64_s64_gpr (int64_t a)
+{
+  /* { dg-final { scan-assembler-times "scvtf\td0, x0" 1 } } */
+  int64x1_t b = (int64x1_t) { a };
+  return vcvt_f64_s64 (b);
+}
+
+float64x1_t
+test_vcvt_f64_u64_gpr (uint64_t a)
+{
+  /* { dg-final { scan-assembler-times "ucvtf\td0, x0" 1 } } */
+  uint64x1_t b = (uint64x1_t) { a };
+  return vcvt_f64_u64 (b);
+}
+
diff --git a/gcc/testsuite/gcc.target/aarch64/vcvt_n_f64_1.c b/gcc/testsuite/gcc.target/aarch64/vcvt_n_f64_1.c
new file mode 100644
index 0000000..6fe16de
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/vcvt_n_f64_1.c
@@ -0,0 +1,80 @@
+/* { dg-do compile } */
+/* { dg-options "-O2" } */
+
+#include "arm_neon.h"
+
+/* For each of these intrinsics, we map directly to an unspec in RTL.
+   We're just using the argument directly and returning the result, so we
+   can precisely specify the exact instruction pattern and register
+   allocations we expect.  */
+
+/* Test that if we have a value already in Advanced-SIMD registers, we use
+   the scalar register forms.  */
+
+float64x1_t
+test_vcvt_n_f64_s64_fpr (int64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "scvtf\td0, d0, #3" 1 } } */
+  return vcvt_n_f64_s64 (a, 3);
+}
+
+float64x1_t
+test_vcvt_n_f64_u64_fpr (uint64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "ucvtf\td0, d0, #3" 1 } } */
+  return vcvt_n_f64_u64 (a, 3);
+}
+
+/* Test that if we have an integer constructor, we use the general-purpose
+   register forms for scvtf and ucvtf.  */
+
+float64x1_t
+test_vcvt_n_f64_s64_gpr (int64_t a)
+{
+  /* { dg-final { scan-assembler-times "scvtf\td0, x0, #3" 1 } } */
+  int64x1_t b = (int64x1_t) { a };
+  return vcvt_n_f64_s64 (b, 3);
+}
+
+float64x1_t
+test_vcvt_n_f64_u64_gpr (uint64_t a)
+{
+  /* { dg-final { scan-assembler-times "ucvtf\td0, x0, #3" 1 } } */
+  uint64x1_t b = (uint64x1_t) { a };
+  return vcvt_n_f64_u64 (b, 3);
+}
+
+/* Test that a normal return through the Advanced-SIMD registers uses
+   the scalar register form.  */
+
+int64x1_t
+test_vcvt_n_s64_f64_fpr (float64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "fcvtzs\td0, d0, #3" 1 } } */
+  return vcvt_n_s64_f64 (a, 3);
+}
+
+uint64x1_t
+test_vcvt_n_u64_f64_fpr (float64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "fcvtzu\td0, d0, #3" 1 } } */
+  return vcvt_n_u64_f64 (a, 3);
+}
+
+/* Test that a lane extracted return as a plain [u]int64_t uses
+   the general-register forms of fcvtzs and fcvtzu.  */
+
+int64_t
+test_vcvt_n_s64_f64_gpr (float64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "fcvtzs\tx0, d0, #3" 1 } } */
+  return vget_lane_s64 (vcvt_n_s64_f64 (a, 3), 0);
+}
+
+uint64_t
+test_vcvt_n_u64_f64_gpr (float64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "fcvtzu\tx0, d0, #3" 1 } } */
+  return vget_lane_u64 (vcvt_n_u64_f64 (a, 3), 0);
+}
+
diff --git a/gcc/testsuite/gcc.target/aarch64/vrecp_f64_1.c b/gcc/testsuite/gcc.target/aarch64/vrecp_f64_1.c
new file mode 100644
index 0000000..c61b2f1
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/vrecp_f64_1.c
@@ -0,0 +1,23 @@
+/* { dg-do compile } */
+/* { dg-options "-O2" } */
+
+#include "arm_neon.h"
+
+/* For each of these intrinsics, we're mapping to an unspec in RTL.
+   We therefore know the expected instruction choice and register pattern,
+   so we can look for it exactly.  */
+
+float64x1_t
+test_vrecpe_f64 (float64x1_t a)
+{
+  /* { dg-final { scan-assembler-times "frecpe\td0, d0" 1 } } */
+  return vrecpe_f64 (a);
+}
+
+float64x1_t
+test_vrecps_f64 (float64x1_t a, float64x1_t b)
+{
+  /* { dg-final { scan-assembler-times "frecps\td0, d0, d1" 1 } } */
+  return vrecps_f64 (a, b);
+}
+