This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
[v2][AArch64, 4/6] Reimplement frsqrts intrinsics
- From: Jiong Wang <jiong dot wang at foss dot arm dot com>
- To: James Greenhalgh <james dot greenhalgh at arm dot com>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>
- Date: Mon, 6 Jun 2016 14:40:33 +0100
- Subject: [v2][AArch64, 4/6] Reimplement frsqrts intrinsics
- Authentication-results: sourceware.org; auth=none
- References: <57430251 dot 6060902 at foss dot arm dot com> <57440F88 dot 2060603 at foss dot arm dot com> <20160527130344 dot GF26495 at arm dot com> <57487B41 dot 8020200 at foss dot arm dot com> <6af07de4-8179-c0bf-410c-317ef52876dd at foss dot arm dot com> <7cb1e234-46f9-76b4-aefd-1eacabfb4ca7 at foss dot arm dot com> <49a7c4d8-3fdc-8806-a4df-affa742cc5d7 at foss dot arm dot com>
Similar as [3/6], these intrinsics were implemented before the instruction
pattern "aarch64_rsqrts<mode>" added, that these intrinsics were implemented
through inline assembly.
This mirgrate the implementation to builtin.
gcc/
2016-06-06 Jiong Wang<jiong.wang@arm.com>
* config/aarch64/aarch64-builtins.def (rsqrts): New builtins for modes
VALLF.
* config/aarch64/aarch64-simd.md (aarch64_rsqrts_<mode>3): Rename to
"aarch64_rsqrts<mode>".
* config/aarch64/aarch64.c (get_rsqrts_type): Update gen* name.
* config/aarch64/arm_neon.h (vrsqrtss_f32): Remove inline assembly. Use
builtin.
(vrsqrtsd_f64): Likewise.
(vrsqrts_f32): Likewise.
(vrsqrts_f64): Likewise.
(vrsqrtsq_f32): Likewise.
(vrsqrtsq_f64): Likewise.
diff --git a/gcc/config/aarch64/aarch64-simd-builtins.def b/gcc/config/aarch64/aarch64-simd-builtins.def
index 32bcd06ec6e483c53b01caf1e30305e0b2b3fb21..1955d171d727e8995795d343ea766f130be0985e 100644
--- a/gcc/config/aarch64/aarch64-simd-builtins.def
+++ b/gcc/config/aarch64/aarch64-simd-builtins.def
@@ -462,3 +462,6 @@
/* Implemented by aarch64_rsqrte<mode>. */
BUILTIN_VALLF (UNOP, rsqrte, 0)
+
+ /* Implemented by aarch64_rsqrts<mode>. */
+ BUILTIN_VALLF (BINOP, rsqrts, 0)
diff --git a/gcc/config/aarch64/aarch64-simd.md b/gcc/config/aarch64/aarch64-simd.md
index 568dd20ad3436e4aa4c3e7cf6b6f766b7fc127db..78a87b1fb52b5b5e21ef5cd7dbe090c863369775 100644
--- a/gcc/config/aarch64/aarch64-simd.md
+++ b/gcc/config/aarch64/aarch64-simd.md
@@ -390,7 +390,7 @@
"frsqrte\\t%<v>0<Vmtype>, %<v>1<Vmtype>"
[(set_attr "type" "neon_fp_rsqrte_<Vetype><q>")])
-(define_insn "aarch64_rsqrts_<mode>3"
+(define_insn "aarch64_rsqrts<mode>"
[(set (match_operand:VALLF 0 "register_operand" "=w")
(unspec:VALLF [(match_operand:VALLF 1 "register_operand" "w")
(match_operand:VALLF 2 "register_operand" "w")]
diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
index acfb39dc025d74fe531d439bb87c52d18955ee7c..b60e5c52df6310a87635c523d723eee9768d7aef 100644
--- a/gcc/config/aarch64/aarch64.c
+++ b/gcc/config/aarch64/aarch64.c
@@ -7368,11 +7368,11 @@ get_rsqrts_type (machine_mode mode)
{
switch (mode)
{
- case DFmode: return gen_aarch64_rsqrts_df3;
- case SFmode: return gen_aarch64_rsqrts_sf3;
- case V2DFmode: return gen_aarch64_rsqrts_v2df3;
- case V2SFmode: return gen_aarch64_rsqrts_v2sf3;
- case V4SFmode: return gen_aarch64_rsqrts_v4sf3;
+ case DFmode: return gen_aarch64_rsqrtsdf;
+ case SFmode: return gen_aarch64_rsqrtssf;
+ case V2DFmode: return gen_aarch64_rsqrtsv2df;
+ case V2SFmode: return gen_aarch64_rsqrtsv2sf;
+ case V4SFmode: return gen_aarch64_rsqrtsv4sf;
default: gcc_unreachable ();
}
}
diff --git a/gcc/config/aarch64/arm_neon.h b/gcc/config/aarch64/arm_neon.h
index 49d572ff8e5007ad07672568ed4dccbea4e0e139..2177703180ca50acedd64d613e4e665264371fb2 100644
--- a/gcc/config/aarch64/arm_neon.h
+++ b/gcc/config/aarch64/arm_neon.h
@@ -9185,61 +9185,6 @@ vrsqrteq_u32 (uint32x4_t a)
return result;
}
-__extension__ static __inline float32x2_t __attribute__ ((__always_inline__))
-vrsqrts_f32 (float32x2_t a, float32x2_t b)
-{
- float32x2_t result;
- __asm__ ("frsqrts %0.2s,%1.2s,%2.2s"
- : "=w"(result)
- : "w"(a), "w"(b)
- : /* No clobbers */);
- return result;
-}
-
-__extension__ static __inline float64_t __attribute__ ((__always_inline__))
-vrsqrtsd_f64 (float64_t a, float64_t b)
-{
- float64_t result;
- __asm__ ("frsqrts %d0,%d1,%d2"
- : "=w"(result)
- : "w"(a), "w"(b)
- : /* No clobbers */);
- return result;
-}
-
-__extension__ static __inline float32x4_t __attribute__ ((__always_inline__))
-vrsqrtsq_f32 (float32x4_t a, float32x4_t b)
-{
- float32x4_t result;
- __asm__ ("frsqrts %0.4s,%1.4s,%2.4s"
- : "=w"(result)
- : "w"(a), "w"(b)
- : /* No clobbers */);
- return result;
-}
-
-__extension__ static __inline float64x2_t __attribute__ ((__always_inline__))
-vrsqrtsq_f64 (float64x2_t a, float64x2_t b)
-{
- float64x2_t result;
- __asm__ ("frsqrts %0.2d,%1.2d,%2.2d"
- : "=w"(result)
- : "w"(a), "w"(b)
- : /* No clobbers */);
- return result;
-}
-
-__extension__ static __inline float32_t __attribute__ ((__always_inline__))
-vrsqrtss_f32 (float32_t a, float32_t b)
-{
- float32_t result;
- __asm__ ("frsqrts %s0,%s1,%s2"
- : "=w"(result)
- : "w"(a), "w"(b)
- : /* No clobbers */);
- return result;
-}
-
#define vshrn_high_n_s16(a, b, c) \
__extension__ \
({ \
@@ -21476,6 +21421,45 @@ vrsqrteq_f64 (float64x2_t __a)
return __builtin_aarch64_rsqrtev2df (__a);
}
+/* vrsqrts. */
+
+__extension__ static __inline float32_t __attribute__ ((__always_inline__))
+vrsqrtss_f32 (float32_t __a, float32_t __b)
+{
+ return __builtin_aarch64_rsqrtssf (__a, __b);
+}
+
+__extension__ static __inline float64_t __attribute__ ((__always_inline__))
+vrsqrtsd_f64 (float64_t __a, float64_t __b)
+{
+ return __builtin_aarch64_rsqrtsdf (__a, __b);
+}
+
+__extension__ static __inline float32x2_t __attribute__ ((__always_inline__))
+vrsqrts_f32 (float32x2_t __a, float32x2_t __b)
+{
+ return __builtin_aarch64_rsqrtsv2sf (__a, __b);
+}
+
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vrsqrts_f64 (float64x1_t __a, float64x1_t __b)
+{
+ return (float64x1_t) {vrsqrtsd_f64 (vget_lane_f64 (__a, 0),
+ vget_lane_f64 (__b, 0))};
+}
+
+__extension__ static __inline float32x4_t __attribute__ ((__always_inline__))
+vrsqrtsq_f32 (float32x4_t __a, float32x4_t __b)
+{
+ return __builtin_aarch64_rsqrtsv4sf (__a, __b);
+}
+
+__extension__ static __inline float64x2_t __attribute__ ((__always_inline__))
+vrsqrtsq_f64 (float64x2_t __a, float64x2_t __b)
+{
+ return __builtin_aarch64_rsqrtsv2df (__a, __b);
+}
+
/* vrsra */
__extension__ static __inline int8x8_t __attribute__ ((__always_inline__))