public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
* [gcc r12-5161] aarch64: Use type-qualified builtins for unsigned MLA/MLS intrinsics
@ 2021-11-11 15:35 Jonathan Wright
0 siblings, 0 replies; only message in thread
From: Jonathan Wright @ 2021-11-11 15:35 UTC (permalink / raw)
To: gcc-cvs
https://gcc.gnu.org/g:515ef83098325e9e3b9f7e74b11911fb9ef6daf2
commit r12-5161-g515ef83098325e9e3b9f7e74b11911fb9ef6daf2
Author: Jonathan Wright <jonathan.wright@arm.com>
Date: Mon Nov 8 10:42:48 2021 +0000
aarch64: Use type-qualified builtins for unsigned MLA/MLS intrinsics
Declare type-qualified builtins and use them for MLA/MLS Neon
intrinsics that operate on unsigned types. This eliminates lots of
casts in arm_neon.h.
gcc/ChangeLog:
2021-11-08 Jonathan Wright <jonathan.wright@arm.com>
* config/aarch64/aarch64-simd-builtins.def: Declare type-
qualified builtin generators for unsigned MLA/MLS intrinsics.
* config/aarch64/arm_neon.h (vmla_n_u16): Use type-qualified
builtin.
(vmla_n_u32): Likewise.
(vmla_u8): Likewise.
(vmla_u16): Likewise.
(vmla_u32): Likewise.
(vmlaq_n_u16): Likewise.
(vmlaq_n_u32): Likewise.
(vmlaq_u8): Likewise.
(vmlaq_u16): Likewise.
(vmlaq_u32): Likewise.
(vmls_n_u16): Likewise.
(vmls_n_u32): Likewise.
(vmls_u8): Likewise.
(vmls_u16): Likewise.
(vmls_u32): Likewise.
(vmlsq_n_u16): Likewise.
(vmlsq_n_u32): Likewise.
(vmlsq_u8): Likewise.
(vmlsq_u16): Likewise.
(vmlsq_u32): Likewise.
Diff:
---
gcc/config/aarch64/aarch64-simd-builtins.def | 4 ++
gcc/config/aarch64/arm_neon.h | 80 +++++++---------------------
2 files changed, 24 insertions(+), 60 deletions(-)
diff --git a/gcc/config/aarch64/aarch64-simd-builtins.def b/gcc/config/aarch64/aarch64-simd-builtins.def
index 9b0a6eceafe..c39f172513f 100644
--- a/gcc/config/aarch64/aarch64-simd-builtins.def
+++ b/gcc/config/aarch64/aarch64-simd-builtins.def
@@ -238,13 +238,17 @@
/* Implemented by aarch64_mla<mode>. */
BUILTIN_VDQ_BHSI (TERNOP, mla, 0, NONE)
+ BUILTIN_VDQ_BHSI (TERNOPU, mla, 0, NONE)
/* Implemented by aarch64_mla_n<mode>. */
BUILTIN_VDQHS (TERNOP, mla_n, 0, NONE)
+ BUILTIN_VDQHS (TERNOPU, mla_n, 0, NONE)
/* Implemented by aarch64_mls<mode>. */
BUILTIN_VDQ_BHSI (TERNOP, mls, 0, NONE)
+ BUILTIN_VDQ_BHSI (TERNOPU, mls, 0, NONE)
/* Implemented by aarch64_mls_n<mode>. */
BUILTIN_VDQHS (TERNOP, mls_n, 0, NONE)
+ BUILTIN_VDQHS (TERNOPU, mls_n, 0, NONE)
/* Implemented by aarch64_shrn<mode>". */
BUILTIN_VQN (SHIFTIMM, shrn, 0, NONE)
diff --git a/gcc/config/aarch64/arm_neon.h b/gcc/config/aarch64/arm_neon.h
index 2e64f079833..538fc3cddff 100644
--- a/gcc/config/aarch64/arm_neon.h
+++ b/gcc/config/aarch64/arm_neon.h
@@ -6608,18 +6608,14 @@ __extension__ extern __inline uint16x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmla_n_u16 (uint16x4_t __a, uint16x4_t __b, uint16_t __c)
{
- return (uint16x4_t) __builtin_aarch64_mla_nv4hi ((int16x4_t) __a,
- (int16x4_t) __b,
- (int16_t) __c);
+ return __builtin_aarch64_mla_nv4hi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint32x2_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmla_n_u32 (uint32x2_t __a, uint32x2_t __b, uint32_t __c)
{
- return (uint32x2_t) __builtin_aarch64_mla_nv2si ((int32x2_t) __a,
- (int32x2_t) __b,
- (int32_t) __c);
+ return __builtin_aarch64_mla_nv2si_uuuu (__a, __b, __c);
}
__extension__ extern __inline int8x8_t
@@ -6647,27 +6643,21 @@ __extension__ extern __inline uint8x8_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmla_u8 (uint8x8_t __a, uint8x8_t __b, uint8x8_t __c)
{
- return (uint8x8_t) __builtin_aarch64_mlav8qi ((int8x8_t) __a,
- (int8x8_t) __b,
- (int8x8_t) __c);
+ return __builtin_aarch64_mlav8qi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint16x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmla_u16 (uint16x4_t __a, uint16x4_t __b, uint16x4_t __c)
{
- return (uint16x4_t) __builtin_aarch64_mlav4hi ((int16x4_t) __a,
- (int16x4_t) __b,
- (int16x4_t) __c);
+ return __builtin_aarch64_mlav4hi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint32x2_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmla_u32 (uint32x2_t __a, uint32x2_t __b, uint32x2_t __c)
{
- return (uint32x2_t) __builtin_aarch64_mlav2si ((int32x2_t) __a,
- (int32x2_t) __b,
- (int32x2_t) __c);
+ return __builtin_aarch64_mlav2si_uuuu (__a, __b, __c);
}
__extension__ extern __inline int32x4_t
@@ -6955,18 +6945,14 @@ __extension__ extern __inline uint16x8_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlaq_n_u16 (uint16x8_t __a, uint16x8_t __b, uint16_t __c)
{
- return (uint16x8_t) __builtin_aarch64_mla_nv8hi ((int16x8_t) __a,
- (int16x8_t) __b,
- (int16_t) __c);
+ return __builtin_aarch64_mla_nv8hi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint32x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlaq_n_u32 (uint32x4_t __a, uint32x4_t __b, uint32_t __c)
{
- return (uint32x4_t) __builtin_aarch64_mla_nv4si ((int32x4_t) __a,
- (int32x4_t) __b,
- (int32_t) __c);
+ return __builtin_aarch64_mla_nv4si_uuuu (__a, __b, __c);
}
__extension__ extern __inline int8x16_t
@@ -6994,27 +6980,21 @@ __extension__ extern __inline uint8x16_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlaq_u8 (uint8x16_t __a, uint8x16_t __b, uint8x16_t __c)
{
- return (uint8x16_t) __builtin_aarch64_mlav16qi ((int8x16_t) __a,
- (int8x16_t) __b,
- (int8x16_t) __c);
+ return __builtin_aarch64_mlav16qi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint16x8_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlaq_u16 (uint16x8_t __a, uint16x8_t __b, uint16x8_t __c)
{
- return (uint16x8_t) __builtin_aarch64_mlav8hi ((int16x8_t) __a,
- (int16x8_t) __b,
- (int16x8_t) __c);
+ return __builtin_aarch64_mlav8hi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint32x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlaq_u32 (uint32x4_t __a, uint32x4_t __b, uint32x4_t __c)
{
- return (uint32x4_t) __builtin_aarch64_mlav4si ((int32x4_t) __a,
- (int32x4_t) __b,
- (int32x4_t) __c);
+ return __builtin_aarch64_mlav4si_uuuu (__a, __b, __c);
}
__extension__ extern __inline float32x2_t
@@ -7042,18 +7022,14 @@ __extension__ extern __inline uint16x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmls_n_u16 (uint16x4_t __a, uint16x4_t __b, uint16_t __c)
{
- return (uint16x4_t) __builtin_aarch64_mls_nv4hi ((int16x4_t) __a,
- (int16x4_t) __b,
- (int16_t) __c);
+ return __builtin_aarch64_mls_nv4hi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint32x2_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmls_n_u32 (uint32x2_t __a, uint32x2_t __b, uint32_t __c)
{
- return (uint32x2_t) __builtin_aarch64_mls_nv2si ((int32x2_t) __a,
- (int32x2_t) __b,
- (int32_t) __c);
+ return __builtin_aarch64_mls_nv2si_uuuu (__a, __b, __c);
}
__extension__ extern __inline int8x8_t
@@ -7081,27 +7057,21 @@ __extension__ extern __inline uint8x8_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmls_u8 (uint8x8_t __a, uint8x8_t __b, uint8x8_t __c)
{
- return (uint8x8_t) __builtin_aarch64_mlsv8qi ((int8x8_t) __a,
- (int8x8_t) __b,
- (int8x8_t) __c);
+ return __builtin_aarch64_mlsv8qi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint16x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmls_u16 (uint16x4_t __a, uint16x4_t __b, uint16x4_t __c)
{
- return (uint16x4_t) __builtin_aarch64_mlsv4hi ((int16x4_t) __a,
- (int16x4_t) __b,
- (int16x4_t) __c);
+ return __builtin_aarch64_mlsv4hi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint32x2_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmls_u32 (uint32x2_t __a, uint32x2_t __b, uint32x2_t __c)
{
- return (uint32x2_t) __builtin_aarch64_mlsv2si ((int32x2_t) __a,
- (int32x2_t) __b,
- (int32x2_t) __c);
+ return __builtin_aarch64_mlsv2si_uuuu (__a, __b, __c);
}
__extension__ extern __inline int32x4_t
@@ -7393,18 +7363,14 @@ __extension__ extern __inline uint16x8_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlsq_n_u16 (uint16x8_t __a, uint16x8_t __b, uint16_t __c)
{
- return (uint16x8_t) __builtin_aarch64_mls_nv8hi ((int16x8_t) __a,
- (int16x8_t) __b,
- (int16_t) __c);
+ return __builtin_aarch64_mls_nv8hi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint32x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlsq_n_u32 (uint32x4_t __a, uint32x4_t __b, uint32_t __c)
{
- return (uint32x4_t) __builtin_aarch64_mls_nv4si ((int32x4_t) __a,
- (int32x4_t) __b,
- (int32_t) __c);
+ return __builtin_aarch64_mls_nv4si_uuuu (__a, __b, __c);
}
__extension__ extern __inline int8x16_t
@@ -7432,27 +7398,21 @@ __extension__ extern __inline uint8x16_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlsq_u8 (uint8x16_t __a, uint8x16_t __b, uint8x16_t __c)
{
- return (uint8x16_t) __builtin_aarch64_mlsv16qi ((int8x16_t) __a,
- (int8x16_t) __b,
- (int8x16_t) __c);
+ return __builtin_aarch64_mlsv16qi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint16x8_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlsq_u16 (uint16x8_t __a, uint16x8_t __b, uint16x8_t __c)
{
- return (uint16x8_t) __builtin_aarch64_mlsv8hi ((int16x8_t) __a,
- (int16x8_t) __b,
- (int16x8_t) __c);
+ return __builtin_aarch64_mlsv8hi_uuuu (__a, __b, __c);
}
__extension__ extern __inline uint32x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
vmlsq_u32 (uint32x4_t __a, uint32x4_t __b, uint32x4_t __c)
{
- return (uint32x4_t) __builtin_aarch64_mlsv4si ((int32x4_t) __a,
- (int32x4_t) __b,
- (int32x4_t) __c);
+ return __builtin_aarch64_mlsv4si_uuuu (__a, __b, __c);
}
__extension__ extern __inline int16x8_t
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2021-11-11 15:35 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-11-11 15:35 [gcc r12-5161] aarch64: Use type-qualified builtins for unsigned MLA/MLS intrinsics Jonathan Wright
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).