[PATCH][GCC][Aarch64] Add vectorize patten for copysign.
Tamar Christina
Tamar.Christina@arm.com
Thu Jan 19 09:43:00 GMT 2017
Hi All,
This is a slight modification of the earlier patch (Using a different constant in the mask creation.)
< + HOST_WIDE_INT_M1 << bits));
---
> + HOST_WIDE_INT_M1U << bits));
Kind Regards,
Tamar
________________________________________
From: gcc-patches-owner@gcc.gnu.org <gcc-patches-owner@gcc.gnu.org> on behalf of Tamar Christina <Tamar.Christina@arm.com>
Sent: Tuesday, January 17, 2017 2:50:19 PM
To: GCC Patches; James Greenhalgh; Marcus Shawcroft; Richard Earnshaw
Cc: nd
Subject: [PATCH][GCC][Aarch64] Add vectorize patten for copysign.
Hi All,
This patch vectorizes the copysign builtin for AArch64
similar to how it is done for Arm.
AArch64 now generates:
...
.L4:
ldr q1, [x6, x3]
add w4, w4, 1
ldr q0, [x5, x3]
cmp w4, w7
bif v1.16b, v2.16b, v3.16b
fmul v0.2d, v0.2d, v1.2d
str q0, [x5, x3]
for the input:
x * copysign(1.0, y)
On 481.wrf in Spec2006 on AArch64 this gives us a speedup of 9.1%.
Regtested on aarch64-none-linux-gnu and no regressions.
Ok for trunk?
gcc/
2017-01-17 Tamar Christina <tamar.christina@arm.com>
* config/aarch64/aarch64-builtins.c
(aarch64_builtin_vectorized_function): Added CASE_CFN_COPYSIGN.
* config/aarch64/aarch64.c (aarch64_simd_gen_const_vector_dup):
Changed int to HOST_WIDE_INT.
* config/aarch64/aarch64-protos.h
(aarch64_simd_gen_const_vector_dup): Likewise.
* config/aarch64/aarch64-simd-builtins.def: Added copysign BINOP.
* config/aarch64/aarch64-simd.md: Added copysign<mode>3.
gcc/testsuite/
2017-01-17 Tamar Christina <tamar.christina@arm.com>
* gcc.target/arm/vect-copysignf.c: Move to...
* gcc.dg/vect/vect-copysignf.c: ... Here.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: vect-copysign-2.patch
Type: text/x-patch
Size: 4732 bytes
Desc: vect-copysign-2.patch
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20170119/3b56c476/attachment.bin>
More information about the Gcc-patches
mailing list