tun: windows: don't spin unless we really need it

This commit is contained in:
Jason A. Donenfeld 2019-08-18 11:49:37 +02:00
parent b16dba47a7
commit 01786286c1

View File

@ -19,11 +19,14 @@ import (
) )
const ( const (
packetAlignment uint32 = 4 // Number of bytes packets are aligned to in rings packetAlignment = 4 // Number of bytes packets are aligned to in rings
packetSizeMax = 0xffff // Maximum packet size packetSizeMax = 0xffff // Maximum packet size
packetCapacity = 0x800000 // Ring capacity, 8MiB packetCapacity = 0x800000 // Ring capacity, 8MiB
packetTrailingSize = uint32(unsafe.Sizeof(packetHeader{})) + ((packetSizeMax + (packetAlignment - 1)) &^ (packetAlignment - 1)) - packetAlignment packetTrailingSize = uint32(unsafe.Sizeof(packetHeader{})) + ((packetSizeMax + (packetAlignment - 1)) &^ (packetAlignment - 1)) - packetAlignment
ioctlRegisterRings = (51820 << 16) | (0x970 << 2) | 0 /*METHOD_BUFFERED*/ | (0x3 /*FILE_READ_DATA | FILE_WRITE_DATA*/ << 14) ioctlRegisterRings = (51820 << 16) | (0x970 << 2) | 0 /*METHOD_BUFFERED*/ | (0x3 /*FILE_READ_DATA | FILE_WRITE_DATA*/ << 14)
rateMeasurementGranularity = uint64((time.Second / 2) / time.Nanosecond)
spinloopRateThreshold = 800000000 / 8 // 800mbps
spinloopDuration = uint64(time.Millisecond / 80 / time.Nanosecond) // ~1gbit/s
) )
type packetHeader struct { type packetHeader struct {
@ -50,6 +53,13 @@ type ringDescriptor struct {
} }
} }
type rateJuggler struct {
current uint64
nextByteCount uint64
nextStartTime int64
changing int32
}
type NativeTun struct { type NativeTun struct {
wt *wintun.Wintun wt *wintun.Wintun
handle windows.Handle handle windows.Handle
@ -58,8 +68,15 @@ type NativeTun struct {
events chan Event events chan Event
errors chan error errors chan error
forcedMTU int forcedMTU int
rate rateJuggler
} }
//go:linkname procyield runtime.procyield
func procyield(cycles uint32)
//go:linkname nanotime runtime.nanotime
func nanotime() int64
func packetAlign(size uint32) uint32 { func packetAlign(size uint32) uint32 {
return (size + (packetAlignment - 1)) &^ (packetAlignment - 1) return (size + (packetAlignment - 1)) &^ (packetAlignment - 1)
} }
@ -184,9 +201,6 @@ func (tun *NativeTun) ForceMTU(mtu int) {
tun.forcedMTU = mtu tun.forcedMTU = mtu
} }
//go:linkname procyield runtime.procyield
func procyield(cycles uint32)
// Note: Read() and Write() assume the caller comes only from a single thread; there's no locking. // Note: Read() and Write() assume the caller comes only from a single thread; there's no locking.
func (tun *NativeTun) Read(buff []byte, offset int) (int, error) { func (tun *NativeTun) Read(buff []byte, offset int) (int, error) {
@ -205,7 +219,8 @@ retry:
return 0, os.ErrClosed return 0, os.ErrClosed
} }
start := time.Now() start := nanotime()
shouldSpin := atomic.LoadUint64(&tun.rate.current) >= spinloopRateThreshold && uint64(start-atomic.LoadInt64(&tun.rate.nextStartTime)) <= rateMeasurementGranularity*2
var buffTail uint32 var buffTail uint32
for { for {
buffTail = atomic.LoadUint32(&tun.rings.send.ring.tail) buffTail = atomic.LoadUint32(&tun.rings.send.ring.tail)
@ -215,7 +230,7 @@ retry:
if tun.close { if tun.close {
return 0, os.ErrClosed return 0, os.ErrClosed
} }
if time.Since(start) >= time.Millisecond/80 /* ~1gbit/s */ { if !shouldSpin || uint64(nanotime()-start) >= spinloopDuration {
windows.WaitForSingleObject(tun.rings.send.tailMoved, windows.INFINITE) windows.WaitForSingleObject(tun.rings.send.tailMoved, windows.INFINITE)
goto retry goto retry
} }
@ -243,6 +258,7 @@ retry:
copy(buff[offset:], packet.data[:packet.size]) copy(buff[offset:], packet.data[:packet.size])
buffHead = tun.rings.send.ring.wrap(buffHead + alignedPacketSize) buffHead = tun.rings.send.ring.wrap(buffHead + alignedPacketSize)
atomic.StoreUint32(&tun.rings.send.ring.head, buffHead) atomic.StoreUint32(&tun.rings.send.ring.head, buffHead)
tun.rate.update(uint64(packet.size))
return int(packet.size), nil return int(packet.size), nil
} }
@ -256,6 +272,7 @@ func (tun *NativeTun) Write(buff []byte, offset int) (int, error) {
} }
packetSize := uint32(len(buff) - offset) packetSize := uint32(len(buff) - offset)
tun.rate.update(uint64(packetSize))
alignedPacketSize := packetAlign(uint32(unsafe.Sizeof(packetHeader{})) + packetSize) alignedPacketSize := packetAlign(uint32(unsafe.Sizeof(packetHeader{})) + packetSize)
buffHead := atomic.LoadUint32(&tun.rings.receive.ring.head) buffHead := atomic.LoadUint32(&tun.rings.receive.ring.head)
@ -292,3 +309,18 @@ func (tun *NativeTun) LUID() uint64 {
func (rb *ring) wrap(value uint32) uint32 { func (rb *ring) wrap(value uint32) uint32 {
return value & (packetCapacity - 1) return value & (packetCapacity - 1)
} }
func (rate *rateJuggler) update(packetLen uint64) {
now := nanotime()
total := atomic.AddUint64(&rate.nextByteCount, packetLen)
period := uint64(now - atomic.LoadInt64(&rate.nextStartTime))
if period >= rateMeasurementGranularity {
if !atomic.CompareAndSwapInt32(&rate.changing, 0, 1) {
return
}
atomic.StoreInt64(&rate.nextStartTime, now)
atomic.StoreUint64(&rate.current, total*uint64(time.Second/time.Nanosecond)/period)
atomic.StoreUint64(&rate.nextByteCount, 0)
atomic.StoreInt32(&rate.changing, 0)
}
}