1
0
mirror of https://github.com/golang/go synced 2024-11-08 01:46:14 -07:00
go/src/runtime/os_netbsd_arm.go
Ben Shi 69261ecad6 runtime: use hardware divider to improve performance
The hardware divider is an optional component of ARMv7. This patch
detects whether it is available in runtime and use it or not.

1. The hardware divider is detected at startup and a flag is set/clear
   according to a perticular bit of runtime.hwcap.
2. Each call of runtime.udiv will check this flag and decide if
   use the hardware division instruction.

A rough test shows the performance improves 40-50% for ARMv7. And
the compatibility of ARMv5/v6 is not broken.

fixes #19118

Change-Id: Ic586bc9659ebc169553ca2004d2bdb721df823ac
Reviewed-on: https://go-review.googlesource.com/37496
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
2017-04-11 12:25:55 +00:00

38 lines
1.2 KiB
Go

// Copyright 2013 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package runtime
import "unsafe"
var hardDiv bool // TODO: set if a hardware divider is available
func lwp_mcontext_init(mc *mcontextt, stk unsafe.Pointer, mp *m, gp *g, fn uintptr) {
// Machine dependent mcontext initialisation for LWP.
mc.__gregs[_REG_R15] = uint32(funcPC(lwp_tramp))
mc.__gregs[_REG_R13] = uint32(uintptr(stk))
mc.__gregs[_REG_R0] = uint32(uintptr(unsafe.Pointer(mp)))
mc.__gregs[_REG_R1] = uint32(uintptr(unsafe.Pointer(gp)))
mc.__gregs[_REG_R2] = uint32(fn)
}
func checkgoarm() {
// TODO(minux): FP checks like in os_linux_arm.go.
// osinit not called yet, so ncpu not set: must use getncpu directly.
if getncpu() > 1 && goarm < 7 {
print("runtime: this system has multiple CPUs and must use\n")
print("atomic synchronization instructions. Recompile using GOARM=7.\n")
exit(1)
}
}
//go:nosplit
func cputicks() int64 {
// Currently cputicks() is used in blocking profiler and to seed runtime·fastrand().
// runtime·nanotime() is a poor approximation of CPU ticks that is enough for the profiler.
// TODO: need more entropy to better seed fastrand.
return nanotime()
}