2016-03-01 15:57:46 -07:00
|
|
|
// Copyright 2009 The Go Authors. All rights reserved.
|
2009-09-24 15:27:52 -06:00
|
|
|
// Use of this source code is governed by a BSD-style
|
|
|
|
// license that can be found in the LICENSE file.
|
|
|
|
|
2014-08-12 18:04:45 -06:00
|
|
|
#include "textflag.h"
|
2013-08-12 11:25:18 -06:00
|
|
|
|
2009-09-24 15:27:52 -06:00
|
|
|
// func Sqrt(x float64) float64
|
math: fix sqrt regression on AMD64
1.7 introduced a significant regression compared to 1.6:
SqrtIndirect-4 2.32ns ± 0% 7.86ns ± 0% +238.79% (p=0.000 n=20+18)
This is caused by sqrtsd preserving upper part of destination register.
Which introduces dependency on previous value of X0.
In 1.6 benchmark loop didn't use X0 immediately after call:
callq *%rbx
movsd 0x8(%rsp),%xmm2
movsd 0x20(%rsp),%xmm1
addsd %xmm2,%xmm1
mov 0x18(%rsp),%rax
inc %rax
jmp loop
In 1.7 however xmm0 is used just after call:
callq *%rbx
mov 0x10(%rsp),%rcx
lea 0x1(%rcx),%rax
movsd 0x8(%rsp),%xmm0
movsd 0x18(%rsp),%xmm1
I've verified that this is caused by dependency, by inserting
XORPS X0,X0 in the beginning of math.Sqrt, which puts performance back on 1.6 level.
Splitting SQRTSD mem,reg into:
MOVSD mem,reg
SQRTSD reg,reg
Removes dependency, because MOVSD (load version)
doesn't need to preserve upper part of a register.
And reg,reg operation is solved by renamer in CPU.
As a result of this change regression is gone:
SqrtIndirect-4 7.86ns ± 0% 2.33ns ± 0% -70.36% (p=0.000 n=18+17)
This also removes old Sqrt benchmarks, in favor of benchmarks measuring latency.
Only SqrtIndirect is kept, to show impact of this patch.
Change-Id: Ic7eebe8866445adff5bc38192fa8d64c9a6b8872
Reviewed-on: https://go-review.googlesource.com/28392
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
Reviewed-by: Keith Randall <khr@golang.org>
2016-09-01 12:04:08 -06:00
|
|
|
TEXT ·Sqrt(SB), NOSPLIT, $0
|
2016-09-14 06:23:41 -06:00
|
|
|
XORPS X0, X0 // break dependency
|
|
|
|
SQRTSD x+0(FP), X0
|
|
|
|
MOVSD X0, ret+8(FP)
|
2009-09-24 15:27:52 -06:00
|
|
|
RET
|