2008-10-15 17:11:51 -07:00
|
|
|
// Copyright 2009 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.
|
|
|
|
|
|
2011-04-20 09:57:05 +10:00
|
|
|
// Package reflect implements run-time reflection, allowing a program to
|
|
|
|
|
// manipulate objects with arbitrary types. The typical use is to take a value
|
|
|
|
|
// with static type interface{} and extract its dynamic type information by
|
2011-04-25 13:39:16 -04:00
|
|
|
// calling TypeOf, which returns a Type.
|
2009-11-08 15:57:25 -08:00
|
|
|
//
|
2011-04-25 13:39:16 -04:00
|
|
|
// A call to ValueOf returns a Value representing the run-time data.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// Zero takes a Type and returns a Value representing a zero value
|
|
|
|
|
// for that type.
|
2011-09-10 09:35:25 +10:00
|
|
|
//
|
|
|
|
|
// See "The Laws of Reflection" for an introduction to reflection in Go:
|
2015-07-10 17:17:11 -06:00
|
|
|
// https://golang.org/doc/articles/laws_of_reflection.html
|
2008-10-15 17:11:51 -07:00
|
|
|
package reflect
|
|
|
|
|
|
2009-01-16 14:58:14 -08:00
|
|
|
import (
|
all: merge NaCl branch (part 1)
See golang.org/s/go13nacl for design overview.
This CL is the mostly mechanical changes from rsc's Go 1.2 based NaCl branch, specifically 39cb35750369 to 500771b477cf from https://code.google.com/r/rsc-go13nacl. This CL does not include working NaCl support, there are probably two or three more large merges to come.
CL 15750044 is not included as it involves more invasive changes to the linker which will need to be merged separately.
The exact change lists included are
15050047: syscall: support for Native Client
15360044: syscall: unzip implementation for Native Client
15370044: syscall: Native Client SRPC implementation
15400047: cmd/dist, cmd/go, go/build, test: support for Native Client
15410048: runtime: support for Native Client
15410049: syscall: file descriptor table for Native Client
15410050: syscall: in-memory file system for Native Client
15440048: all: update +build lines for Native Client port
15540045: cmd/6g, cmd/8g, cmd/gc: support for Native Client
15570045: os: support for Native Client
15680044: crypto/..., hash/crc32, reflect, sync/atomic: support for amd64p32
15690044: net: support for Native Client
15690048: runtime: support for fake time like on Go Playground
15690051: build: disable various tests on Native Client
LGTM=rsc
R=rsc
CC=golang-codereviews
https://golang.org/cl/68150047
2014-02-25 09:47:42 -05:00
|
|
|
"runtime"
|
2009-12-15 15:40:16 -08:00
|
|
|
"strconv"
|
2011-03-03 13:20:17 -05:00
|
|
|
"sync"
|
2009-12-15 15:40:16 -08:00
|
|
|
"unsafe"
|
2008-10-15 17:11:51 -07:00
|
|
|
)
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// Type is the representation of a Go type.
|
|
|
|
|
//
|
|
|
|
|
// Not all methods apply to all kinds of types. Restrictions,
|
|
|
|
|
// if any, are noted in the documentation for each method.
|
|
|
|
|
// Use the Kind method to find out the kind of type before
|
|
|
|
|
// calling kind-specific methods. Calling a method
|
|
|
|
|
// inappropriate to the kind of type causes a run-time panic.
|
|
|
|
|
type Type interface {
|
|
|
|
|
// Methods applicable to all types.
|
|
|
|
|
|
|
|
|
|
// Align returns the alignment in bytes of a value of
|
|
|
|
|
// this type when allocated in memory.
|
|
|
|
|
Align() int
|
|
|
|
|
|
|
|
|
|
// FieldAlign returns the alignment in bytes of a value of
|
|
|
|
|
// this type when used as a field in a struct.
|
|
|
|
|
FieldAlign() int
|
|
|
|
|
|
|
|
|
|
// Method returns the i'th method in the type's method set.
|
|
|
|
|
// It panics if i is not in the range [0, NumMethod()).
|
|
|
|
|
//
|
|
|
|
|
// For a non-interface type T or *T, the returned Method's Type and Func
|
|
|
|
|
// fields describe a function whose first argument is the receiver.
|
|
|
|
|
//
|
|
|
|
|
// For an interface type, the returned Method's Type field gives the
|
|
|
|
|
// method signature, without a receiver, and the Func field is nil.
|
|
|
|
|
Method(int) Method
|
|
|
|
|
|
2011-06-29 13:11:49 +10:00
|
|
|
// MethodByName returns the method with that name in the type's
|
|
|
|
|
// method set and a boolean indicating if the method was found.
|
|
|
|
|
//
|
|
|
|
|
// For a non-interface type T or *T, the returned Method's Type and Func
|
|
|
|
|
// fields describe a function whose first argument is the receiver.
|
|
|
|
|
//
|
|
|
|
|
// For an interface type, the returned Method's Type field gives the
|
|
|
|
|
// method signature, without a receiver, and the Func field is nil.
|
|
|
|
|
MethodByName(string) (Method, bool)
|
|
|
|
|
|
reflect: more efficient; cannot Set result of NewValue anymore
* Reduces malloc counts during gob encoder/decoder test from 6/6 to 3/5.
The current reflect uses Set to mean two subtly different things.
(1) If you have a reflect.Value v, it might just represent
itself (as in v = reflect.NewValue(42)), in which case calling
v.Set only changed v, not any other data in the program.
(2) If you have a reflect Value v derived from a pointer
or a slice (as in x := []int{42}; v = reflect.NewValue(x).Index(0)),
v represents the value held there. Changing x[0] affects the
value returned by v.Int(), and calling v.Set affects x[0].
This was not really by design; it just happened that way.
The motivation for the new reflect implementation was
to remove mallocs. The use case (1) has an implicit malloc
inside it. If you can do:
v := reflect.NewValue(0)
v.Set(42)
i := v.Int() // i = 42
then that implies that v is referring to some underlying
chunk of memory in order to remember the 42; that is,
NewValue must have allocated some memory.
Almost all the time you are using reflect the goal is to
inspect or to change other data, not to manipulate data
stored solely inside a reflect.Value.
This CL removes use case (1), so that an assignable
reflect.Value must always refer to some other piece of data
in the program. Put another way, removing this case would
make
v := reflect.NewValue(0)
v.Set(42)
as illegal as
0 = 42.
It would also make this illegal:
x := 0
v := reflect.NewValue(x)
v.Set(42)
for the same reason. (Note that right now, v.Set(42) "succeeds"
but does not change the value of x.)
If you really wanted to make v refer to x, you'd start with &x
and dereference it:
x := 0
v := reflect.NewValue(&x).Elem() // v = *&x
v.Set(42)
It's pretty rare, except in tests, to want to use NewValue and then
call Set to change the Value itself instead of some other piece of
data in the program. I haven't seen it happen once yet while
making the tree build with this change.
For the same reasons, reflect.Zero (formerly reflect.MakeZero)
would also return an unassignable, unaddressable value.
This invalidates the (awkward) idiom:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.PointTo(v)
which, when the API changed, turned into:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.Set(v.Addr())
In both, it is far from clear what the code is trying to do. Now that
it is possible, this CL adds reflect.New(Type) Value that does the
obvious thing (same as Go's new), so this code would be replaced by:
pv := ... some Ptr Value we have ...
pv.Set(reflect.New(pv.Type().Elem()))
The changes just described can be confusing to think about,
but I believe it is because the old API was confusing - it was
conflating two different kinds of Values - and that the new API
by itself is pretty simple: you can only Set (or call Addr on)
a Value if it actually addresses some real piece of data; that is,
only if it is the result of dereferencing a Ptr or indexing a Slice.
If you really want the old behavior, you'd get it by translating:
v := reflect.NewValue(x)
into
v := reflect.New(reflect.Typeof(x)).Elem()
v.Set(reflect.NewValue(x))
Gofix will not be able to help with this, because whether
and how to change the code depends on whether the original
code meant use (1) or use (2), so the developer has to read
and think about the code.
You can see the effect on packages in the tree in
https://golang.org/cl/4423043/.
R=r
CC=golang-dev
https://golang.org/cl/4435042
2011-04-18 14:35:33 -04:00
|
|
|
// NumMethod returns the number of methods in the type's method set.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
NumMethod() int
|
|
|
|
|
|
|
|
|
|
// Name returns the type's name within its package.
|
|
|
|
|
// It returns an empty string for unnamed types.
|
|
|
|
|
Name() string
|
|
|
|
|
|
2012-03-15 17:15:57 -04:00
|
|
|
// PkgPath returns a named type's package path, that is, the import path
|
|
|
|
|
// that uniquely identifies the package, such as "encoding/base64".
|
|
|
|
|
// If the type was predeclared (string, error) or unnamed (*T, struct{}, []int),
|
|
|
|
|
// the package path will be the empty string.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
PkgPath() string
|
|
|
|
|
|
|
|
|
|
// Size returns the number of bytes needed to store
|
|
|
|
|
// a value of the given type; it is analogous to unsafe.Sizeof.
|
|
|
|
|
Size() uintptr
|
|
|
|
|
|
|
|
|
|
// String returns a string representation of the type.
|
|
|
|
|
// The string representation may use shortened package names
|
2011-08-22 13:22:42 +10:00
|
|
|
// (e.g., base64 instead of "encoding/base64") and is not
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// guaranteed to be unique among types. To test for equality,
|
|
|
|
|
// compare the Types directly.
|
|
|
|
|
String() string
|
|
|
|
|
|
|
|
|
|
// Kind returns the specific kind of this type.
|
|
|
|
|
Kind() Kind
|
|
|
|
|
|
2015-02-17 15:44:42 -08:00
|
|
|
// Implements reports whether the type implements the interface type u.
|
2011-04-20 16:24:45 -04:00
|
|
|
Implements(u Type) bool
|
|
|
|
|
|
2015-02-17 15:44:42 -08:00
|
|
|
// AssignableTo reports whether a value of the type is assignable to type u.
|
2011-04-20 16:24:45 -04:00
|
|
|
AssignableTo(u Type) bool
|
|
|
|
|
|
2015-02-17 15:44:42 -08:00
|
|
|
// ConvertibleTo reports whether a value of the type is convertible to type u.
|
2012-09-22 08:52:27 -04:00
|
|
|
ConvertibleTo(u Type) bool
|
|
|
|
|
|
2015-02-17 15:44:42 -08:00
|
|
|
// Comparable reports whether values of this type are comparable.
|
2014-09-16 17:40:10 -04:00
|
|
|
Comparable() bool
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// Methods applicable only to some types, depending on Kind.
|
|
|
|
|
// The methods allowed for each kind are:
|
|
|
|
|
//
|
|
|
|
|
// Int*, Uint*, Float*, Complex*: Bits
|
|
|
|
|
// Array: Elem, Len
|
|
|
|
|
// Chan: ChanDir, Elem
|
|
|
|
|
// Func: In, NumIn, Out, NumOut, IsVariadic.
|
|
|
|
|
// Map: Key, Elem
|
|
|
|
|
// Ptr: Elem
|
|
|
|
|
// Slice: Elem
|
|
|
|
|
// Struct: Field, FieldByIndex, FieldByName, FieldByNameFunc, NumField
|
|
|
|
|
|
|
|
|
|
// Bits returns the size of the type in bits.
|
|
|
|
|
// It panics if the type's Kind is not one of the
|
|
|
|
|
// sized or unsized Int, Uint, Float, or Complex kinds.
|
|
|
|
|
Bits() int
|
|
|
|
|
|
|
|
|
|
// ChanDir returns a channel type's direction.
|
|
|
|
|
// It panics if the type's Kind is not Chan.
|
|
|
|
|
ChanDir() ChanDir
|
|
|
|
|
|
2015-02-17 15:44:42 -08:00
|
|
|
// IsVariadic reports whether a function type's final input parameter
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// is a "..." parameter. If so, t.In(t.NumIn() - 1) returns the parameter's
|
|
|
|
|
// implicit actual type []T.
|
|
|
|
|
//
|
2011-07-27 13:29:44 +10:00
|
|
|
// For concreteness, if t represents func(x int, y ... float64), then
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
//
|
|
|
|
|
// t.NumIn() == 2
|
|
|
|
|
// t.In(0) is the reflect.Type for "int"
|
2011-07-27 13:29:44 +10:00
|
|
|
// t.In(1) is the reflect.Type for "[]float64"
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// t.IsVariadic() == true
|
|
|
|
|
//
|
|
|
|
|
// IsVariadic panics if the type's Kind is not Func.
|
|
|
|
|
IsVariadic() bool
|
|
|
|
|
|
|
|
|
|
// Elem returns a type's element type.
|
|
|
|
|
// It panics if the type's Kind is not Array, Chan, Map, Ptr, or Slice.
|
|
|
|
|
Elem() Type
|
|
|
|
|
|
|
|
|
|
// Field returns a struct type's i'th field.
|
|
|
|
|
// It panics if the type's Kind is not Struct.
|
|
|
|
|
// It panics if i is not in the range [0, NumField()).
|
|
|
|
|
Field(i int) StructField
|
|
|
|
|
|
|
|
|
|
// FieldByIndex returns the nested field corresponding
|
|
|
|
|
// to the index sequence. It is equivalent to calling Field
|
|
|
|
|
// successively for each index i.
|
|
|
|
|
// It panics if the type's Kind is not Struct.
|
|
|
|
|
FieldByIndex(index []int) StructField
|
|
|
|
|
|
|
|
|
|
// FieldByName returns the struct field with the given name
|
|
|
|
|
// and a boolean indicating if the field was found.
|
|
|
|
|
FieldByName(name string) (StructField, bool)
|
|
|
|
|
|
|
|
|
|
// FieldByNameFunc returns the first struct field with a name
|
|
|
|
|
// that satisfies the match function and a boolean indicating if
|
|
|
|
|
// the field was found.
|
|
|
|
|
FieldByNameFunc(match func(string) bool) (StructField, bool)
|
|
|
|
|
|
|
|
|
|
// In returns the type of a function type's i'th input parameter.
|
|
|
|
|
// It panics if the type's Kind is not Func.
|
|
|
|
|
// It panics if i is not in the range [0, NumIn()).
|
|
|
|
|
In(i int) Type
|
|
|
|
|
|
|
|
|
|
// Key returns a map type's key type.
|
|
|
|
|
// It panics if the type's Kind is not Map.
|
|
|
|
|
Key() Type
|
|
|
|
|
|
|
|
|
|
// Len returns an array type's length.
|
|
|
|
|
// It panics if the type's Kind is not Array.
|
|
|
|
|
Len() int
|
|
|
|
|
|
|
|
|
|
// NumField returns a struct type's field count.
|
|
|
|
|
// It panics if the type's Kind is not Struct.
|
|
|
|
|
NumField() int
|
|
|
|
|
|
|
|
|
|
// NumIn returns a function type's input parameter count.
|
|
|
|
|
// It panics if the type's Kind is not Func.
|
|
|
|
|
NumIn() int
|
|
|
|
|
|
|
|
|
|
// NumOut returns a function type's output parameter count.
|
|
|
|
|
// It panics if the type's Kind is not Func.
|
|
|
|
|
NumOut() int
|
|
|
|
|
|
|
|
|
|
// Out returns the type of a function type's i'th output parameter.
|
|
|
|
|
// It panics if the type's Kind is not Func.
|
|
|
|
|
// It panics if i is not in the range [0, NumOut()).
|
|
|
|
|
Out(i int) Type
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
common() *rtype
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
uncommon() *uncommonType
|
|
|
|
|
}
|
|
|
|
|
|
2013-09-13 13:56:39 -04:00
|
|
|
// BUG(rsc): FieldByName and related functions consider struct field names to be equal
|
|
|
|
|
// if the names are equal, even if they are unexported names originating
|
|
|
|
|
// in different packages. The practical effect of this is that the result of
|
|
|
|
|
// t.FieldByName("x") is not well defined if the struct type t contains
|
|
|
|
|
// multiple fields named x (embedded from different packages).
|
|
|
|
|
// FieldByName may return one of the fields named x or may report that there are none.
|
|
|
|
|
// See golang.org/issue/4876 for more details.
|
|
|
|
|
|
2012-09-18 14:22:41 -04:00
|
|
|
/*
|
2015-03-11 12:58:47 -07:00
|
|
|
* These data structures are known to the compiler (../../cmd/internal/gc/reflect.go).
|
2012-09-18 14:22:41 -04:00
|
|
|
* A few are known to ../runtime/type.go to convey to debuggers.
|
2015-03-11 12:58:47 -07:00
|
|
|
* They are also known to ../runtime/type.go.
|
2012-09-18 14:22:41 -04:00
|
|
|
*/
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// A Kind represents the specific kind of type that a Type represents.
|
|
|
|
|
// The zero Kind is not a valid kind.
|
2011-11-16 19:18:25 -05:00
|
|
|
type Kind uint
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
|
|
|
|
|
const (
|
|
|
|
|
Invalid Kind = iota
|
|
|
|
|
Bool
|
|
|
|
|
Int
|
|
|
|
|
Int8
|
|
|
|
|
Int16
|
|
|
|
|
Int32
|
|
|
|
|
Int64
|
|
|
|
|
Uint
|
|
|
|
|
Uint8
|
|
|
|
|
Uint16
|
|
|
|
|
Uint32
|
|
|
|
|
Uint64
|
|
|
|
|
Uintptr
|
|
|
|
|
Float32
|
|
|
|
|
Float64
|
|
|
|
|
Complex64
|
|
|
|
|
Complex128
|
|
|
|
|
Array
|
|
|
|
|
Chan
|
|
|
|
|
Func
|
|
|
|
|
Interface
|
|
|
|
|
Map
|
|
|
|
|
Ptr
|
|
|
|
|
Slice
|
|
|
|
|
String
|
|
|
|
|
Struct
|
|
|
|
|
UnsafePointer
|
|
|
|
|
)
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// rtype is the common implementation of most values.
|
2010-06-21 13:19:29 -07:00
|
|
|
// It is embedded in other, public struct types, but always
|
2011-07-27 13:29:44 +10:00
|
|
|
// with a unique tag like `reflect:"array"` or `reflect:"ptr"`
|
|
|
|
|
// so that code cannot convert from, say, *arrayType to *ptrType.
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
type rtype struct {
|
2014-09-12 07:29:19 -04:00
|
|
|
size uintptr
|
2015-05-04 21:43:30 -04:00
|
|
|
ptrdata uintptr
|
2015-08-21 15:23:57 +12:00
|
|
|
hash uint32 // hash of type; avoids computation in hash tables
|
|
|
|
|
_ uint8 // unused/padding
|
|
|
|
|
align uint8 // alignment of variable with this type
|
|
|
|
|
fieldAlign uint8 // alignment of struct field with this type
|
|
|
|
|
kind uint8 // enumeration for C
|
|
|
|
|
alg *typeAlg // algorithm table
|
|
|
|
|
gcdata *byte // garbage collection data
|
2016-02-16 22:23:14 -05:00
|
|
|
string string // string form; unnecessary but undeniably useful
|
2015-08-21 15:23:57 +12:00
|
|
|
*uncommonType // (relatively) uncommon fields
|
2012-02-12 23:26:20 -05:00
|
|
|
}
|
|
|
|
|
|
2014-12-27 20:32:11 -08:00
|
|
|
// a copy of runtime.typeAlg
|
2014-09-16 17:40:10 -04:00
|
|
|
type typeAlg struct {
|
|
|
|
|
// function for hashing objects of this type
|
2015-01-27 10:04:11 +01:00
|
|
|
// (ptr to object, seed) -> hash
|
|
|
|
|
hash func(unsafe.Pointer, uintptr) uintptr
|
2014-09-16 17:40:10 -04:00
|
|
|
// function for comparing objects of this type
|
2015-01-27 10:04:11 +01:00
|
|
|
// (ptr to object A, ptr to object B) -> ==?
|
|
|
|
|
equal func(unsafe.Pointer, unsafe.Pointer) bool
|
2014-09-16 17:40:10 -04:00
|
|
|
}
|
|
|
|
|
|
2012-02-12 23:26:20 -05:00
|
|
|
// Method on non-interface type
|
2009-07-07 11:02:44 -07:00
|
|
|
type method struct {
|
2012-02-12 23:26:20 -05:00
|
|
|
name *string // name of method
|
|
|
|
|
pkgPath *string // nil for exported Names; otherwise import path
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
mtyp *rtype // method type (without receiver)
|
|
|
|
|
typ *rtype // .(*FuncType) underneath (with receiver)
|
2012-02-12 23:26:20 -05:00
|
|
|
ifn unsafe.Pointer // fn used in interface call (one-word receiver)
|
|
|
|
|
tfn unsafe.Pointer // fn used for normal method call
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// uncommonType is present only for types with names or methods
|
|
|
|
|
// (if T is a named type, the uncommonTypes for T and *T have methods).
|
|
|
|
|
// Using a pointer to this struct reduces the overall size required
|
|
|
|
|
// to describe an unnamed type with no methods.
|
2009-07-07 11:02:44 -07:00
|
|
|
type uncommonType struct {
|
2012-02-12 23:26:20 -05:00
|
|
|
name *string // name of type
|
|
|
|
|
pkgPath *string // import path; nil for built-in types like int, string
|
|
|
|
|
methods []method // methods associated with type
|
2008-10-23 12:41:06 -07:00
|
|
|
}
|
|
|
|
|
|
2009-07-07 11:02:44 -07:00
|
|
|
// ChanDir represents a channel type's direction.
|
|
|
|
|
type ChanDir int
|
2009-10-07 11:55:06 -07:00
|
|
|
|
2009-07-07 11:02:44 -07:00
|
|
|
const (
|
2012-02-12 23:26:20 -05:00
|
|
|
RecvDir ChanDir = 1 << iota // <-chan
|
|
|
|
|
SendDir // chan<-
|
|
|
|
|
BothDir = RecvDir | SendDir // chan
|
2009-07-07 11:02:44 -07:00
|
|
|
)
|
2008-10-15 17:11:51 -07:00
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// arrayType represents a fixed array type.
|
|
|
|
|
type arrayType struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
rtype `reflect:"array"`
|
|
|
|
|
elem *rtype // array element type
|
|
|
|
|
slice *rtype // slice type
|
|
|
|
|
len uintptr
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// chanType represents a channel type.
|
|
|
|
|
type chanType struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
rtype `reflect:"chan"`
|
|
|
|
|
elem *rtype // channel element type
|
|
|
|
|
dir uintptr // channel direction (ChanDir)
|
2008-10-15 17:11:51 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// funcType represents a function type.
|
|
|
|
|
type funcType struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
rtype `reflect:"func"`
|
|
|
|
|
dotdotdot bool // last input parameter is ...
|
|
|
|
|
in []*rtype // input parameter types
|
|
|
|
|
out []*rtype // output parameter types
|
2008-10-17 18:06:29 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// imethod represents a method on an interface type
|
2009-07-07 11:02:44 -07:00
|
|
|
type imethod struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
name *string // name of method
|
|
|
|
|
pkgPath *string // nil for exported Names; otherwise import path
|
|
|
|
|
typ *rtype // .(*FuncType) underneath
|
2009-03-30 23:19:31 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// interfaceType represents an interface type.
|
|
|
|
|
type interfaceType struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
rtype `reflect:"interface"`
|
|
|
|
|
methods []imethod // sorted by hash
|
2008-10-15 17:11:51 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// mapType represents a map type.
|
|
|
|
|
type mapType struct {
|
2014-08-01 14:38:56 -07:00
|
|
|
rtype `reflect:"map"`
|
|
|
|
|
key *rtype // map key type
|
|
|
|
|
elem *rtype // map element (value) type
|
|
|
|
|
bucket *rtype // internal bucket structure
|
2016-02-23 07:46:01 +00:00
|
|
|
hmap *rtype // internal map header
|
2014-08-01 14:38:56 -07:00
|
|
|
keysize uint8 // size of key slot
|
|
|
|
|
indirectkey uint8 // store ptr to key instead of key itself
|
|
|
|
|
valuesize uint8 // size of value slot
|
|
|
|
|
indirectvalue uint8 // store ptr to value instead of value itself
|
|
|
|
|
bucketsize uint16 // size of bucket
|
2014-12-12 13:45:19 -08:00
|
|
|
reflexivekey bool // true if k==k for all keys
|
2015-06-08 08:42:28 -07:00
|
|
|
needkeyupdate bool // true if we need to update key on an overwrite
|
2008-10-15 17:11:51 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// ptrType represents a pointer type.
|
|
|
|
|
type ptrType struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
rtype `reflect:"ptr"`
|
|
|
|
|
elem *rtype // pointer element (pointed at) type
|
2008-10-15 17:11:51 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// sliceType represents a slice type.
|
|
|
|
|
type sliceType struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
rtype `reflect:"slice"`
|
|
|
|
|
elem *rtype // slice element type
|
2008-10-15 17:11:51 -07:00
|
|
|
}
|
|
|
|
|
|
2009-07-07 11:02:44 -07:00
|
|
|
// Struct field
|
|
|
|
|
type structField struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
name *string // nil for embedded fields
|
|
|
|
|
pkgPath *string // nil for exported Names; otherwise import path
|
|
|
|
|
typ *rtype // type of field
|
|
|
|
|
tag *string // nil if no tag
|
|
|
|
|
offset uintptr // byte offset of field within struct
|
2008-10-17 18:06:29 -07:00
|
|
|
}
|
2008-10-15 17:11:51 -07:00
|
|
|
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
// structType represents a struct type.
|
|
|
|
|
type structType struct {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
rtype `reflect:"struct"`
|
|
|
|
|
fields []structField // sorted by offset
|
2009-03-30 23:19:31 -07:00
|
|
|
}
|
|
|
|
|
|
2009-07-07 11:02:44 -07:00
|
|
|
/*
|
|
|
|
|
* The compiler knows the exact layout of all the data structures above.
|
|
|
|
|
* The compiler does not know about the data structures and methods below.
|
|
|
|
|
*/
|
2008-10-15 17:11:51 -07:00
|
|
|
|
2009-07-07 11:02:44 -07:00
|
|
|
// Method represents a single method.
|
|
|
|
|
type Method struct {
|
2012-03-15 17:15:57 -04:00
|
|
|
// Name is the method name.
|
|
|
|
|
// PkgPath is the package path that qualifies a lower case (unexported)
|
|
|
|
|
// method name. It is empty for upper case (exported) method names.
|
|
|
|
|
// The combination of PkgPath and Name uniquely identifies a method
|
2012-10-30 13:38:01 -07:00
|
|
|
// in a method set.
|
2015-07-10 17:17:11 -06:00
|
|
|
// See https://golang.org/ref/spec#Uniqueness_of_identifiers
|
2009-12-15 15:40:16 -08:00
|
|
|
Name string
|
2012-03-15 17:15:57 -04:00
|
|
|
PkgPath string
|
|
|
|
|
|
|
|
|
|
Type Type // method type
|
|
|
|
|
Func Value // func with receiver as first argument
|
|
|
|
|
Index int // index for Type.Method
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
|
|
|
|
|
2013-03-26 11:50:29 -07:00
|
|
|
const (
|
2014-08-18 21:13:11 -04:00
|
|
|
kindDirectIface = 1 << 5
|
|
|
|
|
kindGCProg = 1 << 6 // Type.gc points to GC program
|
|
|
|
|
kindNoPointers = 1 << 7
|
|
|
|
|
kindMask = (1 << 5) - 1
|
2013-03-26 11:50:29 -07:00
|
|
|
)
|
2010-06-20 12:16:25 -07:00
|
|
|
|
|
|
|
|
func (k Kind) String() string {
|
|
|
|
|
if int(k) < len(kindNames) {
|
|
|
|
|
return kindNames[k]
|
|
|
|
|
}
|
|
|
|
|
return "kind" + strconv.Itoa(int(k))
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
var kindNames = []string{
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
Invalid: "invalid",
|
2010-06-20 12:16:25 -07:00
|
|
|
Bool: "bool",
|
|
|
|
|
Int: "int",
|
|
|
|
|
Int8: "int8",
|
|
|
|
|
Int16: "int16",
|
|
|
|
|
Int32: "int32",
|
|
|
|
|
Int64: "int64",
|
|
|
|
|
Uint: "uint",
|
|
|
|
|
Uint8: "uint8",
|
|
|
|
|
Uint16: "uint16",
|
|
|
|
|
Uint32: "uint32",
|
|
|
|
|
Uint64: "uint64",
|
|
|
|
|
Uintptr: "uintptr",
|
|
|
|
|
Float32: "float32",
|
|
|
|
|
Float64: "float64",
|
2011-01-19 23:09:00 -05:00
|
|
|
Complex64: "complex64",
|
|
|
|
|
Complex128: "complex128",
|
2010-06-20 12:16:25 -07:00
|
|
|
Array: "array",
|
|
|
|
|
Chan: "chan",
|
|
|
|
|
Func: "func",
|
|
|
|
|
Interface: "interface",
|
|
|
|
|
Map: "map",
|
|
|
|
|
Ptr: "ptr",
|
|
|
|
|
Slice: "slice",
|
|
|
|
|
String: "string",
|
|
|
|
|
Struct: "struct",
|
|
|
|
|
UnsafePointer: "unsafe.Pointer",
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-09 17:27:49 -07:00
|
|
|
func (t *uncommonType) uncommon() *uncommonType {
|
2009-11-09 12:07:39 -08:00
|
|
|
return t
|
2009-07-09 17:27:49 -07:00
|
|
|
}
|
|
|
|
|
|
2009-07-17 14:20:33 -07:00
|
|
|
func (t *uncommonType) PkgPath() string {
|
|
|
|
|
if t == nil || t.pkgPath == nil {
|
2009-11-09 12:07:39 -08:00
|
|
|
return ""
|
2008-10-26 08:28:33 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
return *t.pkgPath
|
2009-07-17 14:20:33 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
func (t *uncommonType) Name() string {
|
|
|
|
|
if t == nil || t.name == nil {
|
2009-11-09 12:07:39 -08:00
|
|
|
return ""
|
2009-03-30 23:19:31 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
return *t.name
|
2008-10-17 18:06:29 -07:00
|
|
|
}
|
|
|
|
|
|
2016-02-16 22:23:14 -05:00
|
|
|
func (t *rtype) String() string { return t.string }
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Size() uintptr { return t.size }
|
2008-10-15 17:11:51 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Bits() int {
|
reflect: more efficient; cannot Set result of NewValue anymore
* Reduces malloc counts during gob encoder/decoder test from 6/6 to 3/5.
The current reflect uses Set to mean two subtly different things.
(1) If you have a reflect.Value v, it might just represent
itself (as in v = reflect.NewValue(42)), in which case calling
v.Set only changed v, not any other data in the program.
(2) If you have a reflect Value v derived from a pointer
or a slice (as in x := []int{42}; v = reflect.NewValue(x).Index(0)),
v represents the value held there. Changing x[0] affects the
value returned by v.Int(), and calling v.Set affects x[0].
This was not really by design; it just happened that way.
The motivation for the new reflect implementation was
to remove mallocs. The use case (1) has an implicit malloc
inside it. If you can do:
v := reflect.NewValue(0)
v.Set(42)
i := v.Int() // i = 42
then that implies that v is referring to some underlying
chunk of memory in order to remember the 42; that is,
NewValue must have allocated some memory.
Almost all the time you are using reflect the goal is to
inspect or to change other data, not to manipulate data
stored solely inside a reflect.Value.
This CL removes use case (1), so that an assignable
reflect.Value must always refer to some other piece of data
in the program. Put another way, removing this case would
make
v := reflect.NewValue(0)
v.Set(42)
as illegal as
0 = 42.
It would also make this illegal:
x := 0
v := reflect.NewValue(x)
v.Set(42)
for the same reason. (Note that right now, v.Set(42) "succeeds"
but does not change the value of x.)
If you really wanted to make v refer to x, you'd start with &x
and dereference it:
x := 0
v := reflect.NewValue(&x).Elem() // v = *&x
v.Set(42)
It's pretty rare, except in tests, to want to use NewValue and then
call Set to change the Value itself instead of some other piece of
data in the program. I haven't seen it happen once yet while
making the tree build with this change.
For the same reasons, reflect.Zero (formerly reflect.MakeZero)
would also return an unassignable, unaddressable value.
This invalidates the (awkward) idiom:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.PointTo(v)
which, when the API changed, turned into:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.Set(v.Addr())
In both, it is far from clear what the code is trying to do. Now that
it is possible, this CL adds reflect.New(Type) Value that does the
obvious thing (same as Go's new), so this code would be replaced by:
pv := ... some Ptr Value we have ...
pv.Set(reflect.New(pv.Type().Elem()))
The changes just described can be confusing to think about,
but I believe it is because the old API was confusing - it was
conflating two different kinds of Values - and that the new API
by itself is pretty simple: you can only Set (or call Addr on)
a Value if it actually addresses some real piece of data; that is,
only if it is the result of dereferencing a Ptr or indexing a Slice.
If you really want the old behavior, you'd get it by translating:
v := reflect.NewValue(x)
into
v := reflect.New(reflect.Typeof(x)).Elem()
v.Set(reflect.NewValue(x))
Gofix will not be able to help with this, because whether
and how to change the code depends on whether the original
code meant use (1) or use (2), so the developer has to read
and think about the code.
You can see the effect on packages in the tree in
https://golang.org/cl/4423043/.
R=r
CC=golang-dev
https://golang.org/cl/4435042
2011-04-18 14:35:33 -04:00
|
|
|
if t == nil {
|
|
|
|
|
panic("reflect: Bits of nil Type")
|
|
|
|
|
}
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
k := t.Kind()
|
|
|
|
|
if k < Int || k > Complex128 {
|
reflect: more efficient; cannot Set result of NewValue anymore
* Reduces malloc counts during gob encoder/decoder test from 6/6 to 3/5.
The current reflect uses Set to mean two subtly different things.
(1) If you have a reflect.Value v, it might just represent
itself (as in v = reflect.NewValue(42)), in which case calling
v.Set only changed v, not any other data in the program.
(2) If you have a reflect Value v derived from a pointer
or a slice (as in x := []int{42}; v = reflect.NewValue(x).Index(0)),
v represents the value held there. Changing x[0] affects the
value returned by v.Int(), and calling v.Set affects x[0].
This was not really by design; it just happened that way.
The motivation for the new reflect implementation was
to remove mallocs. The use case (1) has an implicit malloc
inside it. If you can do:
v := reflect.NewValue(0)
v.Set(42)
i := v.Int() // i = 42
then that implies that v is referring to some underlying
chunk of memory in order to remember the 42; that is,
NewValue must have allocated some memory.
Almost all the time you are using reflect the goal is to
inspect or to change other data, not to manipulate data
stored solely inside a reflect.Value.
This CL removes use case (1), so that an assignable
reflect.Value must always refer to some other piece of data
in the program. Put another way, removing this case would
make
v := reflect.NewValue(0)
v.Set(42)
as illegal as
0 = 42.
It would also make this illegal:
x := 0
v := reflect.NewValue(x)
v.Set(42)
for the same reason. (Note that right now, v.Set(42) "succeeds"
but does not change the value of x.)
If you really wanted to make v refer to x, you'd start with &x
and dereference it:
x := 0
v := reflect.NewValue(&x).Elem() // v = *&x
v.Set(42)
It's pretty rare, except in tests, to want to use NewValue and then
call Set to change the Value itself instead of some other piece of
data in the program. I haven't seen it happen once yet while
making the tree build with this change.
For the same reasons, reflect.Zero (formerly reflect.MakeZero)
would also return an unassignable, unaddressable value.
This invalidates the (awkward) idiom:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.PointTo(v)
which, when the API changed, turned into:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.Set(v.Addr())
In both, it is far from clear what the code is trying to do. Now that
it is possible, this CL adds reflect.New(Type) Value that does the
obvious thing (same as Go's new), so this code would be replaced by:
pv := ... some Ptr Value we have ...
pv.Set(reflect.New(pv.Type().Elem()))
The changes just described can be confusing to think about,
but I believe it is because the old API was confusing - it was
conflating two different kinds of Values - and that the new API
by itself is pretty simple: you can only Set (or call Addr on)
a Value if it actually addresses some real piece of data; that is,
only if it is the result of dereferencing a Ptr or indexing a Slice.
If you really want the old behavior, you'd get it by translating:
v := reflect.NewValue(x)
into
v := reflect.New(reflect.Typeof(x)).Elem()
v.Set(reflect.NewValue(x))
Gofix will not be able to help with this, because whether
and how to change the code depends on whether the original
code meant use (1) or use (2), so the developer has to read
and think about the code.
You can see the effect on packages in the tree in
https://golang.org/cl/4423043/.
R=r
CC=golang-dev
https://golang.org/cl/4435042
2011-04-18 14:35:33 -04:00
|
|
|
panic("reflect: Bits of non-arithmetic Type " + t.String())
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
}
|
|
|
|
|
return int(t.size) * 8
|
|
|
|
|
}
|
2010-06-21 13:19:29 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Align() int { return int(t.align) }
|
2008-10-16 18:09:38 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) FieldAlign() int { return int(t.fieldAlign) }
|
2008-10-16 18:09:38 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Kind() Kind { return Kind(t.kind & kindMask) }
|
2010-06-20 12:16:25 -07:00
|
|
|
|
2013-12-19 15:15:24 -08:00
|
|
|
func (t *rtype) pointers() bool { return t.kind&kindNoPointers == 0 }
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) common() *rtype { return t }
|
2011-03-03 13:20:17 -05:00
|
|
|
|
2009-07-07 11:02:44 -07:00
|
|
|
func (t *uncommonType) Method(i int) (m Method) {
|
|
|
|
|
if t == nil || i < 0 || i >= len(t.methods) {
|
2011-07-18 11:34:13 +10:00
|
|
|
panic("reflect: Method index out of range")
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
p := &t.methods[i]
|
2009-07-07 11:02:44 -07:00
|
|
|
if p.name != nil {
|
2009-11-09 12:07:39 -08:00
|
|
|
m.Name = *p.name
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
2014-10-17 12:54:31 -04:00
|
|
|
fl := flag(Func)
|
2009-07-07 11:02:44 -07:00
|
|
|
if p.pkgPath != nil {
|
2009-11-09 12:07:39 -08:00
|
|
|
m.PkgPath = *p.pkgPath
|
2015-08-28 09:33:51 +02:00
|
|
|
fl |= flagStickyRO
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
mt := p.typ
|
2011-11-16 19:18:25 -05:00
|
|
|
m.Type = mt
|
2013-02-21 17:01:13 -05:00
|
|
|
fn := unsafe.Pointer(&p.tfn)
|
2014-10-15 14:24:18 -04:00
|
|
|
m.Func = Value{mt, fn, fl}
|
2011-06-29 13:11:49 +10:00
|
|
|
m.Index = i
|
2009-12-15 15:40:16 -08:00
|
|
|
return
|
2008-10-23 12:41:06 -07:00
|
|
|
}
|
|
|
|
|
|
2009-07-07 11:02:44 -07:00
|
|
|
func (t *uncommonType) NumMethod() int {
|
|
|
|
|
if t == nil {
|
2009-11-09 12:07:39 -08:00
|
|
|
return 0
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
return len(t.methods)
|
2009-03-30 23:19:31 -07:00
|
|
|
}
|
|
|
|
|
|
2011-06-29 13:11:49 +10:00
|
|
|
func (t *uncommonType) MethodByName(name string) (m Method, ok bool) {
|
|
|
|
|
if t == nil {
|
|
|
|
|
return
|
|
|
|
|
}
|
|
|
|
|
var p *method
|
|
|
|
|
for i := range t.methods {
|
|
|
|
|
p = &t.methods[i]
|
|
|
|
|
if p.name != nil && *p.name == name {
|
|
|
|
|
return t.Method(i), true
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return
|
|
|
|
|
}
|
|
|
|
|
|
2015-06-24 09:50:12 +10:00
|
|
|
// TODO(rsc): gc supplies these, but they are not
|
2009-07-07 11:02:44 -07:00
|
|
|
// as efficient as they could be: they have commonType
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// as the receiver instead of *rtype.
|
|
|
|
|
func (t *rtype) NumMethod() int {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() == Interface {
|
|
|
|
|
tt := (*interfaceType)(unsafe.Pointer(t))
|
|
|
|
|
return tt.NumMethod()
|
|
|
|
|
}
|
|
|
|
|
return t.uncommonType.NumMethod()
|
|
|
|
|
}
|
2008-10-16 18:09:38 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Method(i int) (m Method) {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() == Interface {
|
|
|
|
|
tt := (*interfaceType)(unsafe.Pointer(t))
|
|
|
|
|
return tt.Method(i)
|
|
|
|
|
}
|
|
|
|
|
return t.uncommonType.Method(i)
|
|
|
|
|
}
|
2008-10-16 18:09:38 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) MethodByName(name string) (m Method, ok bool) {
|
2011-06-29 13:11:49 +10:00
|
|
|
if t.Kind() == Interface {
|
|
|
|
|
tt := (*interfaceType)(unsafe.Pointer(t))
|
|
|
|
|
return tt.MethodByName(name)
|
|
|
|
|
}
|
|
|
|
|
return t.uncommonType.MethodByName(name)
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) PkgPath() string {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
return t.uncommonType.PkgPath()
|
|
|
|
|
}
|
2009-07-17 14:20:33 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Name() string {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
return t.uncommonType.Name()
|
|
|
|
|
}
|
2008-10-15 17:11:51 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) ChanDir() ChanDir {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Chan {
|
|
|
|
|
panic("reflect: ChanDir of non-chan type")
|
|
|
|
|
}
|
|
|
|
|
tt := (*chanType)(unsafe.Pointer(t))
|
|
|
|
|
return ChanDir(tt.dir)
|
|
|
|
|
}
|
2008-10-15 17:11:51 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) IsVariadic() bool {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Func {
|
|
|
|
|
panic("reflect: IsVariadic of non-func type")
|
|
|
|
|
}
|
|
|
|
|
tt := (*funcType)(unsafe.Pointer(t))
|
|
|
|
|
return tt.dotdotdot
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Elem() Type {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
switch t.Kind() {
|
|
|
|
|
case Array:
|
|
|
|
|
tt := (*arrayType)(unsafe.Pointer(t))
|
|
|
|
|
return toType(tt.elem)
|
|
|
|
|
case Chan:
|
|
|
|
|
tt := (*chanType)(unsafe.Pointer(t))
|
|
|
|
|
return toType(tt.elem)
|
|
|
|
|
case Map:
|
|
|
|
|
tt := (*mapType)(unsafe.Pointer(t))
|
|
|
|
|
return toType(tt.elem)
|
|
|
|
|
case Ptr:
|
|
|
|
|
tt := (*ptrType)(unsafe.Pointer(t))
|
|
|
|
|
return toType(tt.elem)
|
|
|
|
|
case Slice:
|
|
|
|
|
tt := (*sliceType)(unsafe.Pointer(t))
|
|
|
|
|
return toType(tt.elem)
|
|
|
|
|
}
|
2012-02-15 13:51:49 +11:00
|
|
|
panic("reflect: Elem of invalid type")
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
}
|
2008-10-22 13:02:43 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Field(i int) StructField {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Struct {
|
|
|
|
|
panic("reflect: Field of non-struct type")
|
|
|
|
|
}
|
|
|
|
|
tt := (*structType)(unsafe.Pointer(t))
|
|
|
|
|
return tt.Field(i)
|
|
|
|
|
}
|
2011-02-22 09:21:50 -08:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) FieldByIndex(index []int) StructField {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Struct {
|
|
|
|
|
panic("reflect: FieldByIndex of non-struct type")
|
|
|
|
|
}
|
|
|
|
|
tt := (*structType)(unsafe.Pointer(t))
|
|
|
|
|
return tt.FieldByIndex(index)
|
|
|
|
|
}
|
2008-10-17 18:06:29 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) FieldByName(name string) (StructField, bool) {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Struct {
|
|
|
|
|
panic("reflect: FieldByName of non-struct type")
|
|
|
|
|
}
|
|
|
|
|
tt := (*structType)(unsafe.Pointer(t))
|
|
|
|
|
return tt.FieldByName(name)
|
|
|
|
|
}
|
2008-10-15 17:11:51 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) FieldByNameFunc(match func(string) bool) (StructField, bool) {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Struct {
|
|
|
|
|
panic("reflect: FieldByNameFunc of non-struct type")
|
2008-10-16 18:09:38 -07:00
|
|
|
}
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
tt := (*structType)(unsafe.Pointer(t))
|
|
|
|
|
return tt.FieldByNameFunc(match)
|
2008-10-15 17:11:51 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) In(i int) Type {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Func {
|
|
|
|
|
panic("reflect: In of non-func type")
|
2008-10-16 16:38:33 -07:00
|
|
|
}
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
tt := (*funcType)(unsafe.Pointer(t))
|
|
|
|
|
return toType(tt.in[i])
|
2008-10-16 16:38:33 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Key() Type {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Map {
|
|
|
|
|
panic("reflect: Key of non-map type")
|
|
|
|
|
}
|
|
|
|
|
tt := (*mapType)(unsafe.Pointer(t))
|
|
|
|
|
return toType(tt.key)
|
|
|
|
|
}
|
2010-02-01 00:25:59 -08:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Len() int {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Array {
|
|
|
|
|
panic("reflect: Len of non-array type")
|
|
|
|
|
}
|
|
|
|
|
tt := (*arrayType)(unsafe.Pointer(t))
|
|
|
|
|
return int(tt.len)
|
|
|
|
|
}
|
2008-10-31 15:26:14 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) NumField() int {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Struct {
|
|
|
|
|
panic("reflect: NumField of non-struct type")
|
2008-10-30 17:29:53 -07:00
|
|
|
}
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
tt := (*structType)(unsafe.Pointer(t))
|
|
|
|
|
return len(tt.fields)
|
2008-10-16 16:38:33 -07:00
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) NumIn() int {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Func {
|
2012-02-15 13:51:49 +11:00
|
|
|
panic("reflect: NumIn of non-func type")
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
}
|
|
|
|
|
tt := (*funcType)(unsafe.Pointer(t))
|
|
|
|
|
return len(tt.in)
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) NumOut() int {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Func {
|
2012-02-15 13:51:49 +11:00
|
|
|
panic("reflect: NumOut of non-func type")
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
}
|
|
|
|
|
tt := (*funcType)(unsafe.Pointer(t))
|
|
|
|
|
return len(tt.out)
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Out(i int) Type {
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() != Func {
|
|
|
|
|
panic("reflect: Out of non-func type")
|
|
|
|
|
}
|
|
|
|
|
tt := (*funcType)(unsafe.Pointer(t))
|
|
|
|
|
return toType(tt.out[i])
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
func (d ChanDir) String() string {
|
|
|
|
|
switch d {
|
|
|
|
|
case SendDir:
|
|
|
|
|
return "chan<-"
|
|
|
|
|
case RecvDir:
|
|
|
|
|
return "<-chan"
|
|
|
|
|
case BothDir:
|
|
|
|
|
return "chan"
|
|
|
|
|
}
|
|
|
|
|
return "ChanDir" + strconv.Itoa(int(d))
|
|
|
|
|
}
|
2008-11-05 13:01:33 -08:00
|
|
|
|
2011-02-16 11:01:21 -08:00
|
|
|
// Method returns the i'th method in the type's method set.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
func (t *interfaceType) Method(i int) (m Method) {
|
2009-07-07 11:02:44 -07:00
|
|
|
if i < 0 || i >= len(t.methods) {
|
2009-11-09 12:07:39 -08:00
|
|
|
return
|
2008-10-16 16:38:33 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
p := &t.methods[i]
|
|
|
|
|
m.Name = *p.name
|
2009-07-07 11:02:44 -07:00
|
|
|
if p.pkgPath != nil {
|
2009-11-09 12:07:39 -08:00
|
|
|
m.PkgPath = *p.pkgPath
|
2008-10-16 16:38:33 -07:00
|
|
|
}
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
m.Type = toType(p.typ)
|
2011-06-29 13:11:49 +10:00
|
|
|
m.Index = i
|
2009-12-15 15:40:16 -08:00
|
|
|
return
|
2008-10-16 16:38:33 -07:00
|
|
|
}
|
|
|
|
|
|
2011-02-16 11:01:21 -08:00
|
|
|
// NumMethod returns the number of interface methods in the type's method set.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
func (t *interfaceType) NumMethod() int { return len(t.methods) }
|
2008-11-05 08:17:01 -08:00
|
|
|
|
2011-06-29 13:11:49 +10:00
|
|
|
// MethodByName method with the given name in the type's method set.
|
|
|
|
|
func (t *interfaceType) MethodByName(name string) (m Method, ok bool) {
|
|
|
|
|
if t == nil {
|
|
|
|
|
return
|
|
|
|
|
}
|
|
|
|
|
var p *imethod
|
|
|
|
|
for i := range t.methods {
|
|
|
|
|
p = &t.methods[i]
|
|
|
|
|
if *p.name == name {
|
|
|
|
|
return t.Method(i), true
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return
|
|
|
|
|
}
|
|
|
|
|
|
2012-03-15 17:15:57 -04:00
|
|
|
// A StructField describes a single field in a struct.
|
2009-07-07 11:02:44 -07:00
|
|
|
type StructField struct {
|
2012-03-15 17:15:57 -04:00
|
|
|
// Name is the field name.
|
2015-10-27 16:25:08 +01:00
|
|
|
Name string
|
2012-03-15 17:15:57 -04:00
|
|
|
// PkgPath is the package path that qualifies a lower case (unexported)
|
|
|
|
|
// field name. It is empty for upper case (exported) field names.
|
2015-07-10 17:17:11 -06:00
|
|
|
// See https://golang.org/ref/spec#Uniqueness_of_identifiers
|
2012-03-15 17:15:57 -04:00
|
|
|
PkgPath string
|
|
|
|
|
|
|
|
|
|
Type Type // field type
|
|
|
|
|
Tag StructTag // field tag string
|
|
|
|
|
Offset uintptr // offset within struct, in bytes
|
|
|
|
|
Index []int // index sequence for Type.FieldByIndex
|
2013-02-01 21:01:46 -05:00
|
|
|
Anonymous bool // is an embedded field
|
2008-10-16 18:09:38 -07:00
|
|
|
}
|
|
|
|
|
|
2011-06-29 09:52:34 -04:00
|
|
|
// A StructTag is the tag string in a struct field.
|
|
|
|
|
//
|
|
|
|
|
// By convention, tag strings are a concatenation of
|
|
|
|
|
// optionally space-separated key:"value" pairs.
|
|
|
|
|
// Each key is a non-empty string consisting of non-control
|
|
|
|
|
// characters other than space (U+0020 ' '), quote (U+0022 '"'),
|
|
|
|
|
// and colon (U+003A ':'). Each value is quoted using U+0022 '"'
|
|
|
|
|
// characters and Go string literal syntax.
|
|
|
|
|
type StructTag string
|
|
|
|
|
|
|
|
|
|
// Get returns the value associated with key in the tag string.
|
|
|
|
|
// If there is no such key in the tag, Get returns the empty string.
|
|
|
|
|
// If the tag does not have the conventional format, the value
|
2011-09-16 15:07:13 -07:00
|
|
|
// returned by Get is unspecified.
|
2011-06-29 09:52:34 -04:00
|
|
|
func (tag StructTag) Get(key string) string {
|
2015-02-05 15:23:52 +11:00
|
|
|
// When modifying this code, also update the validateStructTag code
|
|
|
|
|
// in golang.org/x/tools/cmd/vet/structtag.go.
|
|
|
|
|
|
2011-06-29 09:52:34 -04:00
|
|
|
for tag != "" {
|
2015-02-05 15:23:52 +11:00
|
|
|
// Skip leading space.
|
2011-06-29 09:52:34 -04:00
|
|
|
i := 0
|
|
|
|
|
for i < len(tag) && tag[i] == ' ' {
|
|
|
|
|
i++
|
|
|
|
|
}
|
|
|
|
|
tag = tag[i:]
|
|
|
|
|
if tag == "" {
|
|
|
|
|
break
|
|
|
|
|
}
|
|
|
|
|
|
2015-02-05 15:23:52 +11:00
|
|
|
// Scan to colon. A space, a quote or a control character is a syntax error.
|
|
|
|
|
// Strictly speaking, control chars include the range [0x7f, 0x9f], not just
|
|
|
|
|
// [0x00, 0x1f], but in practice, we ignore the multi-byte control characters
|
|
|
|
|
// as it is simpler to inspect the tag's bytes than the tag's runes.
|
2011-06-29 09:52:34 -04:00
|
|
|
i = 0
|
2015-02-05 15:23:52 +11:00
|
|
|
for i < len(tag) && tag[i] > ' ' && tag[i] != ':' && tag[i] != '"' && tag[i] != 0x7f {
|
2011-06-29 09:52:34 -04:00
|
|
|
i++
|
|
|
|
|
}
|
2015-02-05 15:23:52 +11:00
|
|
|
if i == 0 || i+1 >= len(tag) || tag[i] != ':' || tag[i+1] != '"' {
|
2011-06-29 09:52:34 -04:00
|
|
|
break
|
|
|
|
|
}
|
|
|
|
|
name := string(tag[:i])
|
|
|
|
|
tag = tag[i+1:]
|
|
|
|
|
|
2015-02-05 15:23:52 +11:00
|
|
|
// Scan quoted string to find value.
|
2011-06-29 09:52:34 -04:00
|
|
|
i = 1
|
|
|
|
|
for i < len(tag) && tag[i] != '"' {
|
|
|
|
|
if tag[i] == '\\' {
|
|
|
|
|
i++
|
|
|
|
|
}
|
|
|
|
|
i++
|
|
|
|
|
}
|
|
|
|
|
if i >= len(tag) {
|
|
|
|
|
break
|
|
|
|
|
}
|
|
|
|
|
qvalue := string(tag[:i+1])
|
|
|
|
|
tag = tag[i+1:]
|
|
|
|
|
|
|
|
|
|
if key == name {
|
2015-02-05 15:23:52 +11:00
|
|
|
value, err := strconv.Unquote(qvalue)
|
|
|
|
|
if err != nil {
|
|
|
|
|
break
|
|
|
|
|
}
|
2011-06-29 09:52:34 -04:00
|
|
|
return value
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return ""
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-07 11:02:44 -07:00
|
|
|
// Field returns the i'th struct field.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
func (t *structType) Field(i int) (f StructField) {
|
2009-07-07 11:02:44 -07:00
|
|
|
if i < 0 || i >= len(t.fields) {
|
2009-11-09 12:07:39 -08:00
|
|
|
return
|
2008-10-16 18:09:38 -07:00
|
|
|
}
|
2011-11-16 19:18:25 -05:00
|
|
|
p := &t.fields[i]
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
f.Type = toType(p.typ)
|
2009-07-07 11:02:44 -07:00
|
|
|
if p.name != nil {
|
2009-11-09 12:07:39 -08:00
|
|
|
f.Name = *p.name
|
2008-11-05 08:17:01 -08:00
|
|
|
} else {
|
2009-12-15 15:40:16 -08:00
|
|
|
t := f.Type
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if t.Kind() == Ptr {
|
|
|
|
|
t = t.Elem()
|
2009-07-28 17:01:46 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
f.Name = t.Name()
|
|
|
|
|
f.Anonymous = true
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
|
|
|
|
if p.pkgPath != nil {
|
2009-11-09 12:07:39 -08:00
|
|
|
f.PkgPath = *p.pkgPath
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
|
|
|
|
if p.tag != nil {
|
2011-06-29 09:52:34 -04:00
|
|
|
f.Tag = StructTag(*p.tag)
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
f.Offset = p.offset
|
2012-01-27 16:11:17 -05:00
|
|
|
|
|
|
|
|
// NOTE(rsc): This is the only allocation in the interface
|
|
|
|
|
// presented by a reflect.Type. It would be nice to avoid,
|
|
|
|
|
// at least in the common cases, but we need to make sure
|
|
|
|
|
// that misbehaving clients of reflect cannot affect other
|
|
|
|
|
// uses of reflect. One possibility is CL 5371098, but we
|
|
|
|
|
// postponed that ugliness until there is a demonstrated
|
|
|
|
|
// need for the performance. This is issue 2320.
|
2009-12-15 15:40:16 -08:00
|
|
|
f.Index = []int{i}
|
|
|
|
|
return
|
2009-07-07 11:02:44 -07:00
|
|
|
}
|
|
|
|
|
|
2009-08-05 15:56:44 -07:00
|
|
|
// TODO(gri): Should there be an error/bool indicator if the index
|
|
|
|
|
// is wrong for FieldByIndex?
|
|
|
|
|
|
|
|
|
|
// FieldByIndex returns the nested field corresponding to index.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
func (t *structType) FieldByIndex(index []int) (f StructField) {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
f.Type = toType(&t.rtype)
|
2009-08-05 15:56:44 -07:00
|
|
|
for i, x := range index {
|
|
|
|
|
if i > 0 {
|
2009-12-15 15:40:16 -08:00
|
|
|
ft := f.Type
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
if ft.Kind() == Ptr && ft.Elem().Kind() == Struct {
|
|
|
|
|
ft = ft.Elem()
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
f.Type = ft
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
f = f.Type.Field(x)
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
return
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
|
|
|
|
|
2012-09-05 09:35:53 -04:00
|
|
|
// A fieldScan represents an item on the fieldByNameFunc scan work list.
|
|
|
|
|
type fieldScan struct {
|
|
|
|
|
typ *structType
|
|
|
|
|
index []int
|
|
|
|
|
}
|
2009-08-05 15:56:44 -07:00
|
|
|
|
2012-09-05 09:35:53 -04:00
|
|
|
// FieldByNameFunc returns the struct field with a name that satisfies the
|
|
|
|
|
// match function and a boolean to indicate if the field was found.
|
|
|
|
|
func (t *structType) FieldByNameFunc(match func(string) bool) (result StructField, ok bool) {
|
|
|
|
|
// This uses the same condition that the Go language does: there must be a unique instance
|
|
|
|
|
// of the match at a given depth level. If there are multiple instances of a match at the
|
|
|
|
|
// same depth, they annihilate each other and inhibit any possible match at a lower level.
|
|
|
|
|
// The algorithm is breadth first search, one depth level at a time.
|
|
|
|
|
|
|
|
|
|
// The current and next slices are work queues:
|
|
|
|
|
// current lists the fields to visit on this depth level,
|
|
|
|
|
// and next lists the fields on the next lower level.
|
|
|
|
|
current := []fieldScan{}
|
|
|
|
|
next := []fieldScan{{typ: t}}
|
|
|
|
|
|
|
|
|
|
// nextCount records the number of times an embedded type has been
|
|
|
|
|
// encountered and considered for queueing in the 'next' slice.
|
|
|
|
|
// We only queue the first one, but we increment the count on each.
|
|
|
|
|
// If a struct type T can be reached more than once at a given depth level,
|
|
|
|
|
// then it annihilates itself and need not be considered at all when we
|
|
|
|
|
// process that next depth level.
|
|
|
|
|
var nextCount map[*structType]int
|
|
|
|
|
|
|
|
|
|
// visited records the structs that have been considered already.
|
|
|
|
|
// Embedded pointer fields can create cycles in the graph of
|
|
|
|
|
// reachable embedded types; visited avoids following those cycles.
|
|
|
|
|
// It also avoids duplicated effort: if we didn't find the field in an
|
|
|
|
|
// embedded type T at level 2, we won't find it in one at level 4 either.
|
|
|
|
|
visited := map[*structType]bool{}
|
|
|
|
|
|
|
|
|
|
for len(next) > 0 {
|
|
|
|
|
current, next = next, current[:0]
|
|
|
|
|
count := nextCount
|
|
|
|
|
nextCount = nil
|
|
|
|
|
|
|
|
|
|
// Process all the fields at this depth, now listed in 'current'.
|
|
|
|
|
// The loop queues embedded fields found in 'next', for processing during the next
|
|
|
|
|
// iteration. The multiplicity of the 'current' field counts is recorded
|
|
|
|
|
// in 'count'; the multiplicity of the 'next' field counts is recorded in 'nextCount'.
|
|
|
|
|
for _, scan := range current {
|
|
|
|
|
t := scan.typ
|
|
|
|
|
if visited[t] {
|
|
|
|
|
// We've looked through this type before, at a higher level.
|
|
|
|
|
// That higher level would shadow the lower level we're now at,
|
|
|
|
|
// so this one can't be useful to us. Ignore it.
|
|
|
|
|
continue
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
2012-09-05 09:35:53 -04:00
|
|
|
visited[t] = true
|
|
|
|
|
for i := range t.fields {
|
|
|
|
|
f := &t.fields[i]
|
|
|
|
|
// Find name and type for field f.
|
|
|
|
|
var fname string
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
var ntyp *rtype
|
2012-09-05 09:35:53 -04:00
|
|
|
if f.name != nil {
|
|
|
|
|
fname = *f.name
|
|
|
|
|
} else {
|
|
|
|
|
// Anonymous field of type T or *T.
|
|
|
|
|
// Name taken from type.
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
ntyp = f.typ
|
2012-09-05 09:35:53 -04:00
|
|
|
if ntyp.Kind() == Ptr {
|
|
|
|
|
ntyp = ntyp.Elem().common()
|
|
|
|
|
}
|
|
|
|
|
fname = ntyp.Name()
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
|
|
|
|
|
2012-09-05 09:35:53 -04:00
|
|
|
// Does it match?
|
|
|
|
|
if match(fname) {
|
|
|
|
|
// Potential match
|
|
|
|
|
if count[t] > 1 || ok {
|
|
|
|
|
// Name appeared multiple times at this level: annihilate.
|
|
|
|
|
return StructField{}, false
|
|
|
|
|
}
|
|
|
|
|
result = t.Field(i)
|
|
|
|
|
result.Index = nil
|
|
|
|
|
result.Index = append(result.Index, scan.index...)
|
|
|
|
|
result.Index = append(result.Index, i)
|
|
|
|
|
ok = true
|
|
|
|
|
continue
|
|
|
|
|
}
|
2009-08-05 15:56:44 -07:00
|
|
|
|
2012-09-05 09:35:53 -04:00
|
|
|
// Queue embedded struct fields for processing with next level,
|
|
|
|
|
// but only if we haven't seen a match yet at this level and only
|
2012-11-13 10:45:30 -08:00
|
|
|
// if the embedded types haven't already been queued.
|
2012-09-05 09:35:53 -04:00
|
|
|
if ok || ntyp == nil || ntyp.Kind() != Struct {
|
|
|
|
|
continue
|
|
|
|
|
}
|
|
|
|
|
styp := (*structType)(unsafe.Pointer(ntyp))
|
|
|
|
|
if nextCount[styp] > 0 {
|
2012-11-13 10:45:30 -08:00
|
|
|
nextCount[styp] = 2 // exact multiple doesn't matter
|
2012-09-05 09:35:53 -04:00
|
|
|
continue
|
|
|
|
|
}
|
|
|
|
|
if nextCount == nil {
|
|
|
|
|
nextCount = map[*structType]int{}
|
|
|
|
|
}
|
|
|
|
|
nextCount[styp] = 1
|
2012-11-13 10:45:30 -08:00
|
|
|
if count[t] > 1 {
|
|
|
|
|
nextCount[styp] = 2 // exact multiple doesn't matter
|
|
|
|
|
}
|
2012-09-05 09:35:53 -04:00
|
|
|
var index []int
|
|
|
|
|
index = append(index, scan.index...)
|
|
|
|
|
index = append(index, i)
|
|
|
|
|
next = append(next, fieldScan{styp, index})
|
|
|
|
|
}
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
2012-09-05 09:35:53 -04:00
|
|
|
if ok {
|
|
|
|
|
break
|
2011-11-16 19:18:25 -05:00
|
|
|
}
|
2009-07-16 18:21:14 -07:00
|
|
|
}
|
2009-12-15 15:40:16 -08:00
|
|
|
return
|
2009-07-16 18:21:14 -07:00
|
|
|
}
|
|
|
|
|
|
2009-08-05 15:56:44 -07:00
|
|
|
// FieldByName returns the struct field with the given name
|
|
|
|
|
// and a boolean to indicate if the field was found.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
func (t *structType) FieldByName(name string) (f StructField, present bool) {
|
2012-09-05 09:35:53 -04:00
|
|
|
// Quick check for top-level name, or struct without anonymous fields.
|
|
|
|
|
hasAnon := false
|
|
|
|
|
if name != "" {
|
|
|
|
|
for i := range t.fields {
|
|
|
|
|
tf := &t.fields[i]
|
|
|
|
|
if tf.name == nil {
|
|
|
|
|
hasAnon = true
|
|
|
|
|
continue
|
|
|
|
|
}
|
|
|
|
|
if *tf.name == name {
|
|
|
|
|
return t.Field(i), true
|
|
|
|
|
}
|
|
|
|
|
}
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
2012-09-05 09:35:53 -04:00
|
|
|
if !hasAnon {
|
|
|
|
|
return
|
|
|
|
|
}
|
|
|
|
|
return t.FieldByNameFunc(func(s string) bool { return s == name })
|
2009-08-05 15:56:44 -07:00
|
|
|
}
|
|
|
|
|
|
2015-04-14 09:56:05 -07:00
|
|
|
// TypeOf returns the reflection Type that represents the dynamic type of i.
|
|
|
|
|
// If i is a nil interface value, TypeOf returns nil.
|
2011-04-25 13:39:16 -04:00
|
|
|
func TypeOf(i interface{}) Type {
|
reflect: more efficient; cannot Set result of NewValue anymore
* Reduces malloc counts during gob encoder/decoder test from 6/6 to 3/5.
The current reflect uses Set to mean two subtly different things.
(1) If you have a reflect.Value v, it might just represent
itself (as in v = reflect.NewValue(42)), in which case calling
v.Set only changed v, not any other data in the program.
(2) If you have a reflect Value v derived from a pointer
or a slice (as in x := []int{42}; v = reflect.NewValue(x).Index(0)),
v represents the value held there. Changing x[0] affects the
value returned by v.Int(), and calling v.Set affects x[0].
This was not really by design; it just happened that way.
The motivation for the new reflect implementation was
to remove mallocs. The use case (1) has an implicit malloc
inside it. If you can do:
v := reflect.NewValue(0)
v.Set(42)
i := v.Int() // i = 42
then that implies that v is referring to some underlying
chunk of memory in order to remember the 42; that is,
NewValue must have allocated some memory.
Almost all the time you are using reflect the goal is to
inspect or to change other data, not to manipulate data
stored solely inside a reflect.Value.
This CL removes use case (1), so that an assignable
reflect.Value must always refer to some other piece of data
in the program. Put another way, removing this case would
make
v := reflect.NewValue(0)
v.Set(42)
as illegal as
0 = 42.
It would also make this illegal:
x := 0
v := reflect.NewValue(x)
v.Set(42)
for the same reason. (Note that right now, v.Set(42) "succeeds"
but does not change the value of x.)
If you really wanted to make v refer to x, you'd start with &x
and dereference it:
x := 0
v := reflect.NewValue(&x).Elem() // v = *&x
v.Set(42)
It's pretty rare, except in tests, to want to use NewValue and then
call Set to change the Value itself instead of some other piece of
data in the program. I haven't seen it happen once yet while
making the tree build with this change.
For the same reasons, reflect.Zero (formerly reflect.MakeZero)
would also return an unassignable, unaddressable value.
This invalidates the (awkward) idiom:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.PointTo(v)
which, when the API changed, turned into:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.Set(v.Addr())
In both, it is far from clear what the code is trying to do. Now that
it is possible, this CL adds reflect.New(Type) Value that does the
obvious thing (same as Go's new), so this code would be replaced by:
pv := ... some Ptr Value we have ...
pv.Set(reflect.New(pv.Type().Elem()))
The changes just described can be confusing to think about,
but I believe it is because the old API was confusing - it was
conflating two different kinds of Values - and that the new API
by itself is pretty simple: you can only Set (or call Addr on)
a Value if it actually addresses some real piece of data; that is,
only if it is the result of dereferencing a Ptr or indexing a Slice.
If you really want the old behavior, you'd get it by translating:
v := reflect.NewValue(x)
into
v := reflect.New(reflect.Typeof(x)).Elem()
v.Set(reflect.NewValue(x))
Gofix will not be able to help with this, because whether
and how to change the code depends on whether the original
code meant use (1) or use (2), so the developer has to read
and think about the code.
You can see the effect on packages in the tree in
https://golang.org/cl/4423043/.
R=r
CC=golang-dev
https://golang.org/cl/4435042
2011-04-18 14:35:33 -04:00
|
|
|
eface := *(*emptyInterface)(unsafe.Pointer(&i))
|
|
|
|
|
return toType(eface.typ)
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
}
|
2011-03-03 13:20:17 -05:00
|
|
|
|
|
|
|
|
// ptrMap is the cache for PtrTo.
|
|
|
|
|
var ptrMap struct {
|
|
|
|
|
sync.RWMutex
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
m map[*rtype]*ptrType
|
reflect: more efficient; cannot Set result of NewValue anymore
* Reduces malloc counts during gob encoder/decoder test from 6/6 to 3/5.
The current reflect uses Set to mean two subtly different things.
(1) If you have a reflect.Value v, it might just represent
itself (as in v = reflect.NewValue(42)), in which case calling
v.Set only changed v, not any other data in the program.
(2) If you have a reflect Value v derived from a pointer
or a slice (as in x := []int{42}; v = reflect.NewValue(x).Index(0)),
v represents the value held there. Changing x[0] affects the
value returned by v.Int(), and calling v.Set affects x[0].
This was not really by design; it just happened that way.
The motivation for the new reflect implementation was
to remove mallocs. The use case (1) has an implicit malloc
inside it. If you can do:
v := reflect.NewValue(0)
v.Set(42)
i := v.Int() // i = 42
then that implies that v is referring to some underlying
chunk of memory in order to remember the 42; that is,
NewValue must have allocated some memory.
Almost all the time you are using reflect the goal is to
inspect or to change other data, not to manipulate data
stored solely inside a reflect.Value.
This CL removes use case (1), so that an assignable
reflect.Value must always refer to some other piece of data
in the program. Put another way, removing this case would
make
v := reflect.NewValue(0)
v.Set(42)
as illegal as
0 = 42.
It would also make this illegal:
x := 0
v := reflect.NewValue(x)
v.Set(42)
for the same reason. (Note that right now, v.Set(42) "succeeds"
but does not change the value of x.)
If you really wanted to make v refer to x, you'd start with &x
and dereference it:
x := 0
v := reflect.NewValue(&x).Elem() // v = *&x
v.Set(42)
It's pretty rare, except in tests, to want to use NewValue and then
call Set to change the Value itself instead of some other piece of
data in the program. I haven't seen it happen once yet while
making the tree build with this change.
For the same reasons, reflect.Zero (formerly reflect.MakeZero)
would also return an unassignable, unaddressable value.
This invalidates the (awkward) idiom:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.PointTo(v)
which, when the API changed, turned into:
pv := ... some Ptr Value we have ...
v := reflect.Zero(pv.Type().Elem())
pv.Set(v.Addr())
In both, it is far from clear what the code is trying to do. Now that
it is possible, this CL adds reflect.New(Type) Value that does the
obvious thing (same as Go's new), so this code would be replaced by:
pv := ... some Ptr Value we have ...
pv.Set(reflect.New(pv.Type().Elem()))
The changes just described can be confusing to think about,
but I believe it is because the old API was confusing - it was
conflating two different kinds of Values - and that the new API
by itself is pretty simple: you can only Set (or call Addr on)
a Value if it actually addresses some real piece of data; that is,
only if it is the result of dereferencing a Ptr or indexing a Slice.
If you really want the old behavior, you'd get it by translating:
v := reflect.NewValue(x)
into
v := reflect.New(reflect.Typeof(x)).Elem()
v.Set(reflect.NewValue(x))
Gofix will not be able to help with this, because whether
and how to change the code depends on whether the original
code meant use (1) or use (2), so the developer has to read
and think about the code.
You can see the effect on packages in the tree in
https://golang.org/cl/4423043/.
R=r
CC=golang-dev
https://golang.org/cl/4435042
2011-04-18 14:35:33 -04:00
|
|
|
}
|
|
|
|
|
|
2011-03-03 13:20:17 -05:00
|
|
|
// PtrTo returns the pointer type with element t.
|
|
|
|
|
// For example, if t represents type Foo, PtrTo(t) represents *Foo.
|
reflect: new Type and Value definitions
Type is now an interface that implements all the possible type methods.
Instead of a type switch on a reflect.Type t, switch on t.Kind().
If a method is invoked on the wrong kind of type (for example,
calling t.Field(0) when t.Kind() != Struct), the call panics.
There is one method renaming: t.(*ChanType).Dir() is now t.ChanDir().
Value is now a struct value that implements all the possible value methods.
Instead of a type switch on a reflect.Value v, switch on v.Kind().
If a method is invoked on the wrong kind of value (for example,
calling t.Recv() when t.Kind() != Chan), the call panics.
Since Value is now a struct, not an interface, its zero value
cannot be compared to nil. Instead of v != nil, use v.IsValid().
Instead of other uses of nil as a Value, use Value{}, the zero value.
Many methods have been renamed, most due to signature conflicts:
OLD NEW
v.(*ArrayValue).Elem v.Index
v.(*BoolValue).Get v.Bool
v.(*BoolValue).Set v.SetBool
v.(*ChanType).Dir v.ChanDir
v.(*ChanValue).Get v.Pointer
v.(*ComplexValue).Get v.Complex
v.(*ComplexValue).Overflow v.OverflowComplex
v.(*ComplexValue).Set v.SetComplex
v.(*FloatValue).Get v.Float
v.(*FloatValue).Overflow v.OverflowFloat
v.(*FloatValue).Set v.SetFloat
v.(*FuncValue).Get v.Pointer
v.(*InterfaceValue).Get v.InterfaceData
v.(*IntValue).Get v.Int
v.(*IntValue).Overflow v.OverflowInt
v.(*IntValue).Set v.SetInt
v.(*MapValue).Elem v.MapIndex
v.(*MapValue).Get v.Pointer
v.(*MapValue).Keys v.MapKeys
v.(*MapValue).SetElem v.SetMapIndex
v.(*PtrValue).Get v.Pointer
v.(*SliceValue).Elem v.Index
v.(*SliceValue).Get v.Pointer
v.(*StringValue).Get v.String
v.(*StringValue).Set v.SetString
v.(*UintValue).Get v.Uint
v.(*UintValue).Overflow v.OverflowUint
v.(*UintValue).Set v.SetUint
v.(*UnsafePointerValue).Get v.Pointer
v.(*UnsafePointerValue).Set v.SetPointer
Part of the motivation for this change is to enable a more
efficient implementation of Value, one that does not allocate
memory during most operations. To reduce the size of the CL,
this CL's implementation is a wrapper around the old API.
Later CLs will make the implementation more efficient without
changing the API.
Other CLs to be submitted at the same time as this one
add support for this change to gofix (4343047) and update
the Go source tree (4353043).
R=gri, iant, niemeyer, r, rog, gustavo, r2
CC=golang-dev
https://golang.org/cl/4281055
2011-04-08 12:26:51 -04:00
|
|
|
func PtrTo(t Type) Type {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
return t.(*rtype).ptrTo()
|
2011-11-16 19:18:25 -05:00
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) ptrTo() *rtype {
|
2016-02-18 06:31:57 -05:00
|
|
|
// Check the cache.
|
2011-03-03 13:20:17 -05:00
|
|
|
ptrMap.RLock()
|
|
|
|
|
if m := ptrMap.m; m != nil {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
if p := m[t]; p != nil {
|
2011-03-03 13:20:17 -05:00
|
|
|
ptrMap.RUnlock()
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
return &p.rtype
|
2011-03-03 13:20:17 -05:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
ptrMap.RUnlock()
|
2016-02-18 06:31:57 -05:00
|
|
|
|
2011-03-03 13:20:17 -05:00
|
|
|
ptrMap.Lock()
|
|
|
|
|
if ptrMap.m == nil {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
ptrMap.m = make(map[*rtype]*ptrType)
|
2011-03-03 13:20:17 -05:00
|
|
|
}
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
p := ptrMap.m[t]
|
2011-03-03 13:20:17 -05:00
|
|
|
if p != nil {
|
|
|
|
|
// some other goroutine won the race and created it
|
|
|
|
|
ptrMap.Unlock()
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
return &p.rtype
|
2011-03-03 13:20:17 -05:00
|
|
|
}
|
|
|
|
|
|
2015-03-29 23:11:47 +00:00
|
|
|
// Look in known types.
|
2016-02-16 22:23:14 -05:00
|
|
|
s := "*" + t.string
|
2015-03-29 23:11:47 +00:00
|
|
|
for _, tt := range typesByString(s) {
|
|
|
|
|
p = (*ptrType)(unsafe.Pointer(tt))
|
|
|
|
|
if p.elem == t {
|
|
|
|
|
ptrMap.m[t] = p
|
|
|
|
|
ptrMap.Unlock()
|
|
|
|
|
return &p.rtype
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// Create a new ptrType starting with the description
|
|
|
|
|
// of an *unsafe.Pointer.
|
|
|
|
|
p = new(ptrType)
|
|
|
|
|
var iptr interface{} = (*unsafe.Pointer)(nil)
|
|
|
|
|
prototype := *(**ptrType)(unsafe.Pointer(&iptr))
|
|
|
|
|
*p = *prototype
|
2011-03-03 13:20:17 -05:00
|
|
|
|
2016-02-16 22:23:14 -05:00
|
|
|
p.string = s
|
2011-03-03 13:20:17 -05:00
|
|
|
|
|
|
|
|
// For the type structures linked into the binary, the
|
|
|
|
|
// compiler provides a good hash of the string.
|
|
|
|
|
// Create a good hash for the new string by using
|
|
|
|
|
// the FNV-1 hash's mixing function to combine the
|
|
|
|
|
// old hash and the new "*".
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
p.hash = fnv1(t.hash, '*')
|
2011-03-03 13:20:17 -05:00
|
|
|
|
|
|
|
|
p.uncommonType = nil
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
p.elem = t
|
2011-03-03 13:20:17 -05:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
ptrMap.m[t] = p
|
2011-03-03 13:20:17 -05:00
|
|
|
ptrMap.Unlock()
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
return &p.rtype
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// fnv1 incorporates the list of bytes into the hash x using the FNV-1 hash function.
|
|
|
|
|
func fnv1(x uint32, list ...byte) uint32 {
|
|
|
|
|
for _, b := range list {
|
|
|
|
|
x = x*16777619 ^ uint32(b)
|
|
|
|
|
}
|
|
|
|
|
return x
|
2011-03-03 13:20:17 -05:00
|
|
|
}
|
2011-04-20 16:24:45 -04:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) Implements(u Type) bool {
|
2011-04-20 16:24:45 -04:00
|
|
|
if u == nil {
|
|
|
|
|
panic("reflect: nil type passed to Type.Implements")
|
|
|
|
|
}
|
|
|
|
|
if u.Kind() != Interface {
|
|
|
|
|
panic("reflect: non-interface type passed to Type.Implements")
|
|
|
|
|
}
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
return implements(u.(*rtype), t)
|
2011-04-20 16:24:45 -04:00
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) AssignableTo(u Type) bool {
|
2011-04-20 16:24:45 -04:00
|
|
|
if u == nil {
|
|
|
|
|
panic("reflect: nil type passed to Type.AssignableTo")
|
|
|
|
|
}
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
uu := u.(*rtype)
|
2011-04-20 16:24:45 -04:00
|
|
|
return directlyAssignable(uu, t) || implements(uu, t)
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func (t *rtype) ConvertibleTo(u Type) bool {
|
2012-09-22 08:52:27 -04:00
|
|
|
if u == nil {
|
2012-09-24 11:47:27 -04:00
|
|
|
panic("reflect: nil type passed to Type.ConvertibleTo")
|
2012-09-22 08:52:27 -04:00
|
|
|
}
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
uu := u.(*rtype)
|
2012-09-22 08:52:27 -04:00
|
|
|
return convertOp(uu, t) != nil
|
|
|
|
|
}
|
|
|
|
|
|
2014-09-16 17:40:10 -04:00
|
|
|
func (t *rtype) Comparable() bool {
|
|
|
|
|
return t.alg != nil && t.alg.equal != nil
|
|
|
|
|
}
|
|
|
|
|
|
2015-02-17 15:44:42 -08:00
|
|
|
// implements reports whether the type V implements the interface type T.
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func implements(T, V *rtype) bool {
|
2011-04-20 16:24:45 -04:00
|
|
|
if T.Kind() != Interface {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
t := (*interfaceType)(unsafe.Pointer(T))
|
|
|
|
|
if len(t.methods) == 0 {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// The same algorithm applies in both cases, but the
|
|
|
|
|
// method tables for an interface type and a concrete type
|
|
|
|
|
// are different, so the code is duplicated.
|
|
|
|
|
// In both cases the algorithm is a linear scan over the two
|
|
|
|
|
// lists - T's methods and V's methods - simultaneously.
|
|
|
|
|
// Since method tables are stored in a unique sorted order
|
|
|
|
|
// (alphabetical, with no duplicate method names), the scan
|
|
|
|
|
// through V's methods must hit a match for each of T's
|
|
|
|
|
// methods along the way, or else V does not implement T.
|
|
|
|
|
// This lets us run the scan in overall linear time instead of
|
|
|
|
|
// the quadratic time a naive search would require.
|
2015-03-11 12:58:47 -07:00
|
|
|
// See also ../runtime/iface.go.
|
2011-04-20 16:24:45 -04:00
|
|
|
if V.Kind() == Interface {
|
|
|
|
|
v := (*interfaceType)(unsafe.Pointer(V))
|
|
|
|
|
i := 0
|
|
|
|
|
for j := 0; j < len(v.methods); j++ {
|
|
|
|
|
tm := &t.methods[i]
|
|
|
|
|
vm := &v.methods[j]
|
2015-03-30 02:03:03 +00:00
|
|
|
if *vm.name == *tm.name && vm.pkgPath == tm.pkgPath && vm.typ == tm.typ {
|
2011-04-20 16:24:45 -04:00
|
|
|
if i++; i >= len(t.methods) {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
v := V.uncommon()
|
|
|
|
|
if v == nil {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
i := 0
|
|
|
|
|
for j := 0; j < len(v.methods); j++ {
|
|
|
|
|
tm := &t.methods[i]
|
|
|
|
|
vm := &v.methods[j]
|
2015-03-30 02:03:03 +00:00
|
|
|
if *vm.name == *tm.name && vm.pkgPath == tm.pkgPath && vm.mtyp == tm.typ {
|
2011-04-20 16:24:45 -04:00
|
|
|
if i++; i >= len(t.methods) {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
|
2015-02-17 15:44:42 -08:00
|
|
|
// directlyAssignable reports whether a value x of type V can be directly
|
2011-04-20 16:24:45 -04:00
|
|
|
// assigned (using memmove) to a value of type T.
|
2015-07-10 17:17:11 -06:00
|
|
|
// https://golang.org/doc/go_spec.html#Assignability
|
2011-04-20 16:24:45 -04:00
|
|
|
// Ignoring the interface rules (implemented elsewhere)
|
|
|
|
|
// and the ideal constant rules (no ideal constants at run time).
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func directlyAssignable(T, V *rtype) bool {
|
2011-04-20 16:24:45 -04:00
|
|
|
// x's type V is identical to T?
|
|
|
|
|
if T == V {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Otherwise at least one of T and V must be unnamed
|
|
|
|
|
// and they must have the same kind.
|
|
|
|
|
if T.Name() != "" && V.Name() != "" || T.Kind() != V.Kind() {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
|
2012-09-22 08:52:27 -04:00
|
|
|
// x's type T and V must have identical underlying types.
|
|
|
|
|
return haveIdenticalUnderlyingType(T, V)
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func haveIdenticalUnderlyingType(T, V *rtype) bool {
|
2012-09-22 08:52:27 -04:00
|
|
|
if T == V {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
kind := T.Kind()
|
|
|
|
|
if kind != V.Kind() {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Non-composite types of equal kind have same underlying type
|
|
|
|
|
// (the predefined instance of the type).
|
|
|
|
|
if Bool <= kind && kind <= Complex128 || kind == String || kind == UnsafePointer {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Composite types.
|
|
|
|
|
switch kind {
|
2011-04-20 16:24:45 -04:00
|
|
|
case Array:
|
|
|
|
|
return T.Elem() == V.Elem() && T.Len() == V.Len()
|
|
|
|
|
|
|
|
|
|
case Chan:
|
|
|
|
|
// Special case:
|
|
|
|
|
// x is a bidirectional channel value, T is a channel type,
|
|
|
|
|
// and x's type V and T have identical element types.
|
|
|
|
|
if V.ChanDir() == BothDir && T.Elem() == V.Elem() {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Otherwise continue test for identical underlying type.
|
|
|
|
|
return V.ChanDir() == T.ChanDir() && T.Elem() == V.Elem()
|
|
|
|
|
|
|
|
|
|
case Func:
|
|
|
|
|
t := (*funcType)(unsafe.Pointer(T))
|
|
|
|
|
v := (*funcType)(unsafe.Pointer(V))
|
|
|
|
|
if t.dotdotdot != v.dotdotdot || len(t.in) != len(v.in) || len(t.out) != len(v.out) {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
for i, typ := range t.in {
|
|
|
|
|
if typ != v.in[i] {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
for i, typ := range t.out {
|
|
|
|
|
if typ != v.out[i] {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return true
|
|
|
|
|
|
|
|
|
|
case Interface:
|
|
|
|
|
t := (*interfaceType)(unsafe.Pointer(T))
|
|
|
|
|
v := (*interfaceType)(unsafe.Pointer(V))
|
|
|
|
|
if len(t.methods) == 0 && len(v.methods) == 0 {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
// Might have the same methods but still
|
|
|
|
|
// need a run time conversion.
|
|
|
|
|
return false
|
|
|
|
|
|
|
|
|
|
case Map:
|
|
|
|
|
return T.Key() == V.Key() && T.Elem() == V.Elem()
|
|
|
|
|
|
|
|
|
|
case Ptr, Slice:
|
|
|
|
|
return T.Elem() == V.Elem()
|
|
|
|
|
|
|
|
|
|
case Struct:
|
|
|
|
|
t := (*structType)(unsafe.Pointer(T))
|
|
|
|
|
v := (*structType)(unsafe.Pointer(V))
|
|
|
|
|
if len(t.fields) != len(v.fields) {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
for i := range t.fields {
|
|
|
|
|
tf := &t.fields[i]
|
|
|
|
|
vf := &v.fields[i]
|
2012-11-07 11:55:35 -08:00
|
|
|
if tf.name != vf.name && (tf.name == nil || vf.name == nil || *tf.name != *vf.name) {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
if tf.pkgPath != vf.pkgPath && (tf.pkgPath == nil || vf.pkgPath == nil || *tf.pkgPath != *vf.pkgPath) {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
if tf.typ != vf.typ {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
if tf.tag != vf.tag && (tf.tag == nil || vf.tag == nil || *tf.tag != *vf.tag) {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
if tf.offset != vf.offset {
|
2011-04-20 16:24:45 -04:00
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return false
|
|
|
|
|
}
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
|
|
|
|
|
// typelinks is implemented in package runtime.
|
2013-03-20 16:32:37 -07:00
|
|
|
// It returns a slice of all the 'typelink' information in the binary,
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// which is to say a slice of known types, sorted by string.
|
|
|
|
|
// Note that strings are not unique identifiers for types:
|
|
|
|
|
// there can be more than one with a given string.
|
|
|
|
|
// Only types we might want to look up are included:
|
2016-02-18 06:31:57 -05:00
|
|
|
// pointers, channels, maps, slices, and arrays.
|
2015-03-29 21:59:00 +00:00
|
|
|
func typelinks() [][]*rtype
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
|
|
|
|
|
// typesByString returns the subslice of typelinks() whose elements have
|
|
|
|
|
// the given string representation.
|
|
|
|
|
// It may be empty (no known types with that string) or may have
|
|
|
|
|
// multiple elements (multiple types with that string).
|
|
|
|
|
func typesByString(s string) []*rtype {
|
2015-03-29 21:59:00 +00:00
|
|
|
typs := typelinks()
|
|
|
|
|
var ret []*rtype
|
|
|
|
|
|
|
|
|
|
for _, typ := range typs {
|
|
|
|
|
// We are looking for the first index i where the string becomes >= s.
|
|
|
|
|
// This is a copy of sort.Search, with f(h) replaced by (*typ[h].string >= s).
|
|
|
|
|
i, j := 0, len(typ)
|
|
|
|
|
for i < j {
|
|
|
|
|
h := i + (j-i)/2 // avoid overflow when computing h
|
|
|
|
|
// i ≤ h < j
|
2016-02-16 22:23:14 -05:00
|
|
|
if !(typ[h].string >= s) {
|
2015-03-29 21:59:00 +00:00
|
|
|
i = h + 1 // preserves f(i-1) == false
|
|
|
|
|
} else {
|
|
|
|
|
j = h // preserves f(j) == true
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
// i == j, f(i-1) == false, and f(j) (= f(i)) == true => answer is i.
|
|
|
|
|
|
|
|
|
|
// Having found the first, linear scan forward to find the last.
|
|
|
|
|
// We could do a second binary search, but the caller is going
|
|
|
|
|
// to do a linear scan anyway.
|
|
|
|
|
j = i
|
2016-02-16 22:23:14 -05:00
|
|
|
for j < len(typ) && typ[j].string == s {
|
2015-03-29 21:59:00 +00:00
|
|
|
j++
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
}
|
|
|
|
|
|
2015-03-29 21:59:00 +00:00
|
|
|
if j > i {
|
|
|
|
|
if ret == nil {
|
|
|
|
|
ret = typ[i:j:j]
|
|
|
|
|
} else {
|
|
|
|
|
ret = append(ret, typ[i:j]...)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return ret
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// The lookupCache caches ChanOf, MapOf, and SliceOf lookups.
|
|
|
|
|
var lookupCache struct {
|
|
|
|
|
sync.RWMutex
|
|
|
|
|
m map[cacheKey]*rtype
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// A cacheKey is the key for use in the lookupCache.
|
|
|
|
|
// Four values describe any of the types we are looking for:
|
|
|
|
|
// type kind, one or two subtypes, and an extra integer.
|
|
|
|
|
type cacheKey struct {
|
|
|
|
|
kind Kind
|
|
|
|
|
t1 *rtype
|
|
|
|
|
t2 *rtype
|
|
|
|
|
extra uintptr
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// cacheGet looks for a type under the key k in the lookupCache.
|
|
|
|
|
// If it finds one, it returns that type.
|
|
|
|
|
// If not, it returns nil with the cache locked.
|
|
|
|
|
// The caller is expected to use cachePut to unlock the cache.
|
|
|
|
|
func cacheGet(k cacheKey) Type {
|
|
|
|
|
lookupCache.RLock()
|
|
|
|
|
t := lookupCache.m[k]
|
|
|
|
|
lookupCache.RUnlock()
|
|
|
|
|
if t != nil {
|
|
|
|
|
return t
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
lookupCache.Lock()
|
|
|
|
|
t = lookupCache.m[k]
|
|
|
|
|
if t != nil {
|
|
|
|
|
lookupCache.Unlock()
|
|
|
|
|
return t
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if lookupCache.m == nil {
|
|
|
|
|
lookupCache.m = make(map[cacheKey]*rtype)
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// cachePut stores the given type in the cache, unlocks the cache,
|
|
|
|
|
// and returns the type. It is expected that the cache is locked
|
|
|
|
|
// because cacheGet returned nil.
|
|
|
|
|
func cachePut(k cacheKey, t *rtype) Type {
|
|
|
|
|
lookupCache.m[k] = t
|
|
|
|
|
lookupCache.Unlock()
|
|
|
|
|
return t
|
|
|
|
|
}
|
|
|
|
|
|
2014-12-23 15:19:30 +11:00
|
|
|
// The funcLookupCache caches FuncOf lookups.
|
|
|
|
|
// FuncOf does not share the common lookupCache since cacheKey is not
|
|
|
|
|
// sufficient to represent functions unambiguously.
|
|
|
|
|
var funcLookupCache struct {
|
|
|
|
|
sync.RWMutex
|
|
|
|
|
m map[uint32][]*rtype // keyed by hash calculated in FuncOf
|
|
|
|
|
}
|
|
|
|
|
|
2012-11-22 02:58:24 +08:00
|
|
|
// ChanOf returns the channel type with the given direction and element type.
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// For example, if t represents int, ChanOf(RecvDir, t) represents <-chan int.
|
|
|
|
|
//
|
|
|
|
|
// The gc runtime imposes a limit of 64 kB on channel element types.
|
|
|
|
|
// If t's size is equal to or exceeds this limit, ChanOf panics.
|
|
|
|
|
func ChanOf(dir ChanDir, t Type) Type {
|
|
|
|
|
typ := t.(*rtype)
|
|
|
|
|
|
|
|
|
|
// Look in cache.
|
|
|
|
|
ckey := cacheKey{Chan, typ, nil, uintptr(dir)}
|
|
|
|
|
if ch := cacheGet(ckey); ch != nil {
|
|
|
|
|
return ch
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// This restriction is imposed by the gc compiler and the runtime.
|
|
|
|
|
if typ.size >= 1<<16 {
|
|
|
|
|
lookupCache.Unlock()
|
|
|
|
|
panic("reflect.ChanOf: element size too large")
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Look in known types.
|
|
|
|
|
// TODO: Precedence when constructing string.
|
|
|
|
|
var s string
|
|
|
|
|
switch dir {
|
|
|
|
|
default:
|
|
|
|
|
lookupCache.Unlock()
|
|
|
|
|
panic("reflect.ChanOf: invalid dir")
|
|
|
|
|
case SendDir:
|
2016-02-16 22:23:14 -05:00
|
|
|
s = "chan<- " + typ.string
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
case RecvDir:
|
2016-02-16 22:23:14 -05:00
|
|
|
s = "<-chan " + typ.string
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
case BothDir:
|
2016-02-16 22:23:14 -05:00
|
|
|
s = "chan " + typ.string
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
}
|
|
|
|
|
for _, tt := range typesByString(s) {
|
|
|
|
|
ch := (*chanType)(unsafe.Pointer(tt))
|
|
|
|
|
if ch.elem == typ && ch.dir == uintptr(dir) {
|
|
|
|
|
return cachePut(ckey, tt)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Make a channel type.
|
|
|
|
|
var ichan interface{} = (chan unsafe.Pointer)(nil)
|
|
|
|
|
prototype := *(**chanType)(unsafe.Pointer(&ichan))
|
|
|
|
|
ch := new(chanType)
|
|
|
|
|
*ch = *prototype
|
2015-01-01 21:38:12 -05:00
|
|
|
ch.dir = uintptr(dir)
|
2016-02-16 22:23:14 -05:00
|
|
|
ch.string = s
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
ch.hash = fnv1(typ.hash, 'c', byte(dir))
|
|
|
|
|
ch.elem = typ
|
|
|
|
|
ch.uncommonType = nil
|
|
|
|
|
|
|
|
|
|
return cachePut(ckey, &ch.rtype)
|
|
|
|
|
}
|
|
|
|
|
|
2013-03-26 11:50:29 -07:00
|
|
|
func ismapkey(*rtype) bool // implemented in runtime
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// MapOf returns the map type with the given key and element types.
|
|
|
|
|
// For example, if k represents int and e represents string,
|
|
|
|
|
// MapOf(k, e) represents map[int]string.
|
|
|
|
|
//
|
|
|
|
|
// If the key type is not a valid map key type (that is, if it does
|
2013-03-26 11:50:29 -07:00
|
|
|
// not implement Go's == operator), MapOf panics.
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
func MapOf(key, elem Type) Type {
|
|
|
|
|
ktyp := key.(*rtype)
|
|
|
|
|
etyp := elem.(*rtype)
|
|
|
|
|
|
2013-03-26 11:50:29 -07:00
|
|
|
if !ismapkey(ktyp) {
|
|
|
|
|
panic("reflect.MapOf: invalid key type " + ktyp.String())
|
|
|
|
|
}
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
|
|
|
|
|
// Look in cache.
|
|
|
|
|
ckey := cacheKey{Map, ktyp, etyp, 0}
|
|
|
|
|
if mt := cacheGet(ckey); mt != nil {
|
|
|
|
|
return mt
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Look in known types.
|
2016-02-16 22:23:14 -05:00
|
|
|
s := "map[" + ktyp.string + "]" + etyp.string
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
for _, tt := range typesByString(s) {
|
|
|
|
|
mt := (*mapType)(unsafe.Pointer(tt))
|
|
|
|
|
if mt.key == ktyp && mt.elem == etyp {
|
|
|
|
|
return cachePut(ckey, tt)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Make a map type.
|
|
|
|
|
var imap interface{} = (map[unsafe.Pointer]unsafe.Pointer)(nil)
|
|
|
|
|
mt := new(mapType)
|
2015-01-26 21:04:41 +03:00
|
|
|
*mt = **(**mapType)(unsafe.Pointer(&imap))
|
2016-02-16 22:23:14 -05:00
|
|
|
mt.string = s
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
mt.hash = fnv1(etyp.hash, 'm', byte(ktyp.hash>>24), byte(ktyp.hash>>16), byte(ktyp.hash>>8), byte(ktyp.hash))
|
|
|
|
|
mt.key = ktyp
|
|
|
|
|
mt.elem = etyp
|
2013-08-31 09:09:50 -07:00
|
|
|
mt.bucket = bucketOf(ktyp, etyp)
|
2014-08-01 14:38:56 -07:00
|
|
|
if ktyp.size > maxKeySize {
|
|
|
|
|
mt.keysize = uint8(ptrSize)
|
|
|
|
|
mt.indirectkey = 1
|
|
|
|
|
} else {
|
|
|
|
|
mt.keysize = uint8(ktyp.size)
|
|
|
|
|
mt.indirectkey = 0
|
|
|
|
|
}
|
|
|
|
|
if etyp.size > maxValSize {
|
|
|
|
|
mt.valuesize = uint8(ptrSize)
|
|
|
|
|
mt.indirectvalue = 1
|
|
|
|
|
} else {
|
|
|
|
|
mt.valuesize = uint8(etyp.size)
|
|
|
|
|
mt.indirectvalue = 0
|
|
|
|
|
}
|
|
|
|
|
mt.bucketsize = uint16(mt.bucket.size)
|
2014-12-12 13:45:19 -08:00
|
|
|
mt.reflexivekey = isReflexive(ktyp)
|
2015-06-08 08:42:28 -07:00
|
|
|
mt.needkeyupdate = needKeyUpdate(ktyp)
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
mt.uncommonType = nil
|
2013-03-26 11:50:29 -07:00
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
return cachePut(ckey, &mt.rtype)
|
|
|
|
|
}
|
|
|
|
|
|
2014-12-23 15:19:30 +11:00
|
|
|
// FuncOf returns the function type with the given argument and result types.
|
|
|
|
|
// For example if k represents int and e represents string,
|
|
|
|
|
// FuncOf([]Type{k}, []Type{e}, false) represents func(int) string.
|
|
|
|
|
//
|
|
|
|
|
// The variadic argument controls whether the function is variadic. FuncOf
|
|
|
|
|
// panics if the in[len(in)-1] does not represent a slice and variadic is
|
|
|
|
|
// true.
|
|
|
|
|
func FuncOf(in, out []Type, variadic bool) Type {
|
|
|
|
|
if variadic && (len(in) == 0 || in[len(in)-1].Kind() != Slice) {
|
|
|
|
|
panic("reflect.FuncOf: last arg of variadic func must be slice")
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Make a func type.
|
|
|
|
|
var ifunc interface{} = (func())(nil)
|
|
|
|
|
prototype := *(**funcType)(unsafe.Pointer(&ifunc))
|
|
|
|
|
ft := new(funcType)
|
|
|
|
|
*ft = *prototype
|
|
|
|
|
|
|
|
|
|
// Build a hash and minimally populate ft.
|
|
|
|
|
var hash uint32
|
|
|
|
|
var fin, fout []*rtype
|
|
|
|
|
for _, in := range in {
|
|
|
|
|
t := in.(*rtype)
|
|
|
|
|
fin = append(fin, t)
|
|
|
|
|
hash = fnv1(hash, byte(t.hash>>24), byte(t.hash>>16), byte(t.hash>>8), byte(t.hash))
|
|
|
|
|
}
|
|
|
|
|
if variadic {
|
|
|
|
|
hash = fnv1(hash, 'v')
|
|
|
|
|
}
|
|
|
|
|
hash = fnv1(hash, '.')
|
|
|
|
|
for _, out := range out {
|
|
|
|
|
t := out.(*rtype)
|
|
|
|
|
fout = append(fout, t)
|
|
|
|
|
hash = fnv1(hash, byte(t.hash>>24), byte(t.hash>>16), byte(t.hash>>8), byte(t.hash))
|
|
|
|
|
}
|
|
|
|
|
ft.hash = hash
|
|
|
|
|
ft.in = fin
|
|
|
|
|
ft.out = fout
|
|
|
|
|
ft.dotdotdot = variadic
|
|
|
|
|
|
|
|
|
|
// Look in cache.
|
|
|
|
|
funcLookupCache.RLock()
|
|
|
|
|
for _, t := range funcLookupCache.m[hash] {
|
|
|
|
|
if haveIdenticalUnderlyingType(&ft.rtype, t) {
|
|
|
|
|
funcLookupCache.RUnlock()
|
|
|
|
|
return t
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
funcLookupCache.RUnlock()
|
|
|
|
|
|
|
|
|
|
// Not in cache, lock and retry.
|
|
|
|
|
funcLookupCache.Lock()
|
|
|
|
|
defer funcLookupCache.Unlock()
|
|
|
|
|
if funcLookupCache.m == nil {
|
|
|
|
|
funcLookupCache.m = make(map[uint32][]*rtype)
|
|
|
|
|
}
|
|
|
|
|
for _, t := range funcLookupCache.m[hash] {
|
|
|
|
|
if haveIdenticalUnderlyingType(&ft.rtype, t) {
|
|
|
|
|
return t
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Look in known types for the same string representation.
|
|
|
|
|
str := funcStr(ft)
|
|
|
|
|
for _, tt := range typesByString(str) {
|
|
|
|
|
if haveIdenticalUnderlyingType(&ft.rtype, tt) {
|
|
|
|
|
funcLookupCache.m[hash] = append(funcLookupCache.m[hash], tt)
|
|
|
|
|
return tt
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Populate the remaining fields of ft and store in cache.
|
2016-02-16 22:23:14 -05:00
|
|
|
ft.string = str
|
2014-12-23 15:19:30 +11:00
|
|
|
ft.uncommonType = nil
|
|
|
|
|
funcLookupCache.m[hash] = append(funcLookupCache.m[hash], &ft.rtype)
|
|
|
|
|
|
2015-05-15 14:46:20 -04:00
|
|
|
return &ft.rtype
|
2014-12-23 15:19:30 +11:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// funcStr builds a string representation of a funcType.
|
|
|
|
|
func funcStr(ft *funcType) string {
|
|
|
|
|
repr := make([]byte, 0, 64)
|
|
|
|
|
repr = append(repr, "func("...)
|
|
|
|
|
for i, t := range ft.in {
|
|
|
|
|
if i > 0 {
|
|
|
|
|
repr = append(repr, ", "...)
|
|
|
|
|
}
|
|
|
|
|
if ft.dotdotdot && i == len(ft.in)-1 {
|
|
|
|
|
repr = append(repr, "..."...)
|
2016-02-16 22:23:14 -05:00
|
|
|
repr = append(repr, (*sliceType)(unsafe.Pointer(t)).elem.string...)
|
2014-12-23 15:19:30 +11:00
|
|
|
} else {
|
2016-02-16 22:23:14 -05:00
|
|
|
repr = append(repr, t.string...)
|
2014-12-23 15:19:30 +11:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
repr = append(repr, ')')
|
|
|
|
|
if l := len(ft.out); l == 1 {
|
|
|
|
|
repr = append(repr, ' ')
|
|
|
|
|
} else if l > 1 {
|
|
|
|
|
repr = append(repr, " ("...)
|
|
|
|
|
}
|
|
|
|
|
for i, t := range ft.out {
|
|
|
|
|
if i > 0 {
|
|
|
|
|
repr = append(repr, ", "...)
|
|
|
|
|
}
|
2016-02-16 22:23:14 -05:00
|
|
|
repr = append(repr, t.string...)
|
2014-12-23 15:19:30 +11:00
|
|
|
}
|
|
|
|
|
if len(ft.out) > 1 {
|
|
|
|
|
repr = append(repr, ')')
|
|
|
|
|
}
|
|
|
|
|
return string(repr)
|
|
|
|
|
}
|
|
|
|
|
|
2014-12-12 13:45:19 -08:00
|
|
|
// isReflexive reports whether the == operation on the type is reflexive.
|
|
|
|
|
// That is, x == x for all values x of type t.
|
|
|
|
|
func isReflexive(t *rtype) bool {
|
|
|
|
|
switch t.Kind() {
|
|
|
|
|
case Bool, Int, Int8, Int16, Int32, Int64, Uint, Uint8, Uint16, Uint32, Uint64, Uintptr, Chan, Ptr, String, UnsafePointer:
|
|
|
|
|
return true
|
|
|
|
|
case Float32, Float64, Complex64, Complex128, Interface:
|
|
|
|
|
return false
|
|
|
|
|
case Array:
|
|
|
|
|
tt := (*arrayType)(unsafe.Pointer(t))
|
|
|
|
|
return isReflexive(tt.elem)
|
|
|
|
|
case Struct:
|
|
|
|
|
tt := (*structType)(unsafe.Pointer(t))
|
|
|
|
|
for _, f := range tt.fields {
|
|
|
|
|
if !isReflexive(f.typ) {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return true
|
|
|
|
|
default:
|
|
|
|
|
// Func, Map, Slice, Invalid
|
|
|
|
|
panic("isReflexive called on non-key type " + t.String())
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2015-06-08 08:42:28 -07:00
|
|
|
// needKeyUpdate reports whether map overwrites require the key to be copied.
|
|
|
|
|
func needKeyUpdate(t *rtype) bool {
|
|
|
|
|
switch t.Kind() {
|
|
|
|
|
case Bool, Int, Int8, Int16, Int32, Int64, Uint, Uint8, Uint16, Uint32, Uint64, Uintptr, Chan, Ptr, UnsafePointer:
|
|
|
|
|
return false
|
|
|
|
|
case Float32, Float64, Complex64, Complex128, Interface, String:
|
|
|
|
|
// Float keys can be updated from +0 to -0.
|
|
|
|
|
// String keys can be updated to use a smaller backing store.
|
|
|
|
|
// Interfaces might have floats of strings in them.
|
|
|
|
|
return true
|
|
|
|
|
case Array:
|
|
|
|
|
tt := (*arrayType)(unsafe.Pointer(t))
|
|
|
|
|
return needKeyUpdate(tt.elem)
|
|
|
|
|
case Struct:
|
|
|
|
|
tt := (*structType)(unsafe.Pointer(t))
|
|
|
|
|
for _, f := range tt.fields {
|
|
|
|
|
if needKeyUpdate(f.typ) {
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return false
|
|
|
|
|
default:
|
|
|
|
|
// Func, Map, Slice, Invalid
|
|
|
|
|
panic("needKeyUpdate called on non-key type " + t.String())
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2014-09-08 00:06:45 -04:00
|
|
|
// Make sure these routines stay in sync with ../../runtime/hashmap.go!
|
2013-08-31 09:09:50 -07:00
|
|
|
// These types exist only for GC, so we only fill out GC relevant info.
|
|
|
|
|
// Currently, that's just size and the GC program. We also fill in string
|
|
|
|
|
// for possible debugging use.
|
|
|
|
|
const (
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
bucketSize uintptr = 8
|
|
|
|
|
maxKeySize uintptr = 128
|
|
|
|
|
maxValSize uintptr = 128
|
2013-08-31 09:09:50 -07:00
|
|
|
)
|
|
|
|
|
|
|
|
|
|
func bucketOf(ktyp, etyp *rtype) *rtype {
|
2015-01-26 21:04:41 +03:00
|
|
|
// See comment on hmap.overflow in ../runtime/hashmap.go.
|
|
|
|
|
var kind uint8
|
2015-02-14 16:10:06 +03:00
|
|
|
if ktyp.kind&kindNoPointers != 0 && etyp.kind&kindNoPointers != 0 &&
|
|
|
|
|
ktyp.size <= maxKeySize && etyp.size <= maxValSize {
|
2015-01-26 21:04:41 +03:00
|
|
|
kind = kindNoPointers
|
|
|
|
|
}
|
|
|
|
|
|
2014-07-29 11:01:02 +04:00
|
|
|
if ktyp.size > maxKeySize {
|
2013-08-31 09:09:50 -07:00
|
|
|
ktyp = PtrTo(ktyp).(*rtype)
|
|
|
|
|
}
|
2014-07-29 11:01:02 +04:00
|
|
|
if etyp.size > maxValSize {
|
2013-08-31 09:09:50 -07:00
|
|
|
etyp = PtrTo(etyp).(*rtype)
|
|
|
|
|
}
|
|
|
|
|
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
// Prepare GC data if any.
|
|
|
|
|
// A bucket is at most bucketSize*(1+maxKeySize+maxValSize)+2*ptrSize bytes,
|
|
|
|
|
// or 2072 bytes, or 259 pointer-size words, or 33 bytes of pointer bitmap.
|
|
|
|
|
// Normally the enforced limit on pointer maps is 16 bytes,
|
|
|
|
|
// but larger ones are acceptable, 33 bytes isn't too too big,
|
|
|
|
|
// and it's easier to generate a pointer bitmap than a GC program.
|
|
|
|
|
// Note that since the key and value are known to be <= 128 bytes,
|
|
|
|
|
// they're guaranteed to have bitmaps instead of GC programs.
|
|
|
|
|
var gcdata *byte
|
|
|
|
|
var ptrdata uintptr
|
cmd/compile, runtime: fix placement of map bucket overflow pointer on nacl
On most systems, a pointer is the worst case alignment, so adding
a pointer field at the end of a struct guarantees there will be no
padding added after that field (to satisfy overall struct alignment
due to some more-aligned field also present).
In the runtime, the map implementation needs a quick way to
get to the overflow pointer, which is last in the bucket struct,
so it uses size - sizeof(pointer) as the offset.
NaCl/amd64p32 is the exception, as always.
The worst case alignment is 64 bits but pointers are 32 bits.
There's a long history that is not worth going into, but when
we moved the overflow pointer to the end of the struct,
we didn't get the padding computation right.
The compiler computed the regular struct size and then
on amd64p32 added another 32-bit field.
And the runtime assumed it could step back two 32-bit fields
(one 64-bit register size) to get to the overflow pointer.
But in fact if the struct needed 64-bit alignment, the computation
of the regular struct size would have added a 32-bit pad already,
and then the code unconditionally added a second 32-bit pad.
This placed the overflow pointer three words from the end, not two.
The last two were padding, and since the runtime was consistent
about using the second-to-last word as the overflow pointer,
no harm done in the sense of overwriting useful memory.
But writing the overflow pointer to a non-pointer word of memory
means that the GC can't see the overflow blocks, so it will
collect them prematurely. Then bad things happen.
Correct all this in a few steps:
1. Add an explicit check at the end of the bucket layout in the
compiler that the overflow field is last in the struct, never
followed by padding.
2. When padding is needed on nacl (not always, just when needed),
insert it before the overflow pointer, to preserve the "last in the struct"
property.
3. Let the compiler have the final word on the width of the struct,
by inserting an explicit padding field instead of overwriting the
results of the width computation it does.
4. For the same reason (tell the truth to the compiler), set the type
of the overflow field when we're trying to pretend its not a pointer
(in this case the runtime maintains a list of the overflow blocks
elsewhere).
5. Make the runtime use "last in the struct" as its location algorithm.
This fixes TestTraceStress on nacl/amd64p32.
The 'bad map state' and 'invalid free list' failures no longer occur.
Fixes #11838.
Change-Id: If918887f8f252d988db0a35159944d2b36512f92
Reviewed-on: https://go-review.googlesource.com/12971
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
2015-07-30 22:05:51 -04:00
|
|
|
var overflowPad uintptr
|
|
|
|
|
|
|
|
|
|
// On NaCl, pad if needed to make overflow end at the proper struct alignment.
|
|
|
|
|
// On other systems, align > ptrSize is not possible.
|
|
|
|
|
if runtime.GOARCH == "amd64p32" && (ktyp.align > ptrSize || etyp.align > ptrSize) {
|
|
|
|
|
overflowPad = ptrSize
|
|
|
|
|
}
|
|
|
|
|
size := bucketSize*(1+ktyp.size+etyp.size) + overflowPad + ptrSize
|
|
|
|
|
if size&uintptr(ktyp.align-1) != 0 || size&uintptr(etyp.align-1) != 0 {
|
|
|
|
|
panic("reflect: bad size computation in MapOf")
|
|
|
|
|
}
|
|
|
|
|
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
if kind != kindNoPointers {
|
|
|
|
|
nptr := (bucketSize*(1+ktyp.size+etyp.size) + ptrSize) / ptrSize
|
|
|
|
|
mask := make([]byte, (nptr+7)/8)
|
|
|
|
|
base := bucketSize / ptrSize
|
|
|
|
|
|
|
|
|
|
if ktyp.kind&kindNoPointers == 0 {
|
|
|
|
|
if ktyp.kind&kindGCProg != 0 {
|
|
|
|
|
panic("reflect: unexpected GC program in MapOf")
|
|
|
|
|
}
|
|
|
|
|
kmask := (*[16]byte)(unsafe.Pointer(ktyp.gcdata))
|
|
|
|
|
for i := uintptr(0); i < ktyp.size/ptrSize; i++ {
|
|
|
|
|
if (kmask[i/8]>>(i%8))&1 != 0 {
|
|
|
|
|
for j := uintptr(0); j < bucketSize; j++ {
|
|
|
|
|
word := base + j*ktyp.size/ptrSize + i
|
|
|
|
|
mask[word/8] |= 1 << (word % 8)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
base += bucketSize * ktyp.size / ptrSize
|
|
|
|
|
|
|
|
|
|
if etyp.kind&kindNoPointers == 0 {
|
|
|
|
|
if etyp.kind&kindGCProg != 0 {
|
|
|
|
|
panic("reflect: unexpected GC program in MapOf")
|
|
|
|
|
}
|
|
|
|
|
emask := (*[16]byte)(unsafe.Pointer(etyp.gcdata))
|
|
|
|
|
for i := uintptr(0); i < etyp.size/ptrSize; i++ {
|
|
|
|
|
if (emask[i/8]>>(i%8))&1 != 0 {
|
|
|
|
|
for j := uintptr(0); j < bucketSize; j++ {
|
|
|
|
|
word := base + j*etyp.size/ptrSize + i
|
|
|
|
|
mask[word/8] |= 1 << (word % 8)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
base += bucketSize * etyp.size / ptrSize
|
cmd/compile, runtime: fix placement of map bucket overflow pointer on nacl
On most systems, a pointer is the worst case alignment, so adding
a pointer field at the end of a struct guarantees there will be no
padding added after that field (to satisfy overall struct alignment
due to some more-aligned field also present).
In the runtime, the map implementation needs a quick way to
get to the overflow pointer, which is last in the bucket struct,
so it uses size - sizeof(pointer) as the offset.
NaCl/amd64p32 is the exception, as always.
The worst case alignment is 64 bits but pointers are 32 bits.
There's a long history that is not worth going into, but when
we moved the overflow pointer to the end of the struct,
we didn't get the padding computation right.
The compiler computed the regular struct size and then
on amd64p32 added another 32-bit field.
And the runtime assumed it could step back two 32-bit fields
(one 64-bit register size) to get to the overflow pointer.
But in fact if the struct needed 64-bit alignment, the computation
of the regular struct size would have added a 32-bit pad already,
and then the code unconditionally added a second 32-bit pad.
This placed the overflow pointer three words from the end, not two.
The last two were padding, and since the runtime was consistent
about using the second-to-last word as the overflow pointer,
no harm done in the sense of overwriting useful memory.
But writing the overflow pointer to a non-pointer word of memory
means that the GC can't see the overflow blocks, so it will
collect them prematurely. Then bad things happen.
Correct all this in a few steps:
1. Add an explicit check at the end of the bucket layout in the
compiler that the overflow field is last in the struct, never
followed by padding.
2. When padding is needed on nacl (not always, just when needed),
insert it before the overflow pointer, to preserve the "last in the struct"
property.
3. Let the compiler have the final word on the width of the struct,
by inserting an explicit padding field instead of overwriting the
results of the width computation it does.
4. For the same reason (tell the truth to the compiler), set the type
of the overflow field when we're trying to pretend its not a pointer
(in this case the runtime maintains a list of the overflow blocks
elsewhere).
5. Make the runtime use "last in the struct" as its location algorithm.
This fixes TestTraceStress on nacl/amd64p32.
The 'bad map state' and 'invalid free list' failures no longer occur.
Fixes #11838.
Change-Id: If918887f8f252d988db0a35159944d2b36512f92
Reviewed-on: https://go-review.googlesource.com/12971
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
2015-07-30 22:05:51 -04:00
|
|
|
base += overflowPad / ptrSize
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
|
|
|
|
|
word := base
|
|
|
|
|
mask[word/8] |= 1 << (word % 8)
|
|
|
|
|
gcdata = &mask[0]
|
|
|
|
|
ptrdata = (word + 1) * ptrSize
|
|
|
|
|
|
cmd/compile, runtime: fix placement of map bucket overflow pointer on nacl
On most systems, a pointer is the worst case alignment, so adding
a pointer field at the end of a struct guarantees there will be no
padding added after that field (to satisfy overall struct alignment
due to some more-aligned field also present).
In the runtime, the map implementation needs a quick way to
get to the overflow pointer, which is last in the bucket struct,
so it uses size - sizeof(pointer) as the offset.
NaCl/amd64p32 is the exception, as always.
The worst case alignment is 64 bits but pointers are 32 bits.
There's a long history that is not worth going into, but when
we moved the overflow pointer to the end of the struct,
we didn't get the padding computation right.
The compiler computed the regular struct size and then
on amd64p32 added another 32-bit field.
And the runtime assumed it could step back two 32-bit fields
(one 64-bit register size) to get to the overflow pointer.
But in fact if the struct needed 64-bit alignment, the computation
of the regular struct size would have added a 32-bit pad already,
and then the code unconditionally added a second 32-bit pad.
This placed the overflow pointer three words from the end, not two.
The last two were padding, and since the runtime was consistent
about using the second-to-last word as the overflow pointer,
no harm done in the sense of overwriting useful memory.
But writing the overflow pointer to a non-pointer word of memory
means that the GC can't see the overflow blocks, so it will
collect them prematurely. Then bad things happen.
Correct all this in a few steps:
1. Add an explicit check at the end of the bucket layout in the
compiler that the overflow field is last in the struct, never
followed by padding.
2. When padding is needed on nacl (not always, just when needed),
insert it before the overflow pointer, to preserve the "last in the struct"
property.
3. Let the compiler have the final word on the width of the struct,
by inserting an explicit padding field instead of overwriting the
results of the width computation it does.
4. For the same reason (tell the truth to the compiler), set the type
of the overflow field when we're trying to pretend its not a pointer
(in this case the runtime maintains a list of the overflow blocks
elsewhere).
5. Make the runtime use "last in the struct" as its location algorithm.
This fixes TestTraceStress on nacl/amd64p32.
The 'bad map state' and 'invalid free list' failures no longer occur.
Fixes #11838.
Change-Id: If918887f8f252d988db0a35159944d2b36512f92
Reviewed-on: https://go-review.googlesource.com/12971
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
2015-07-30 22:05:51 -04:00
|
|
|
// overflow word must be last
|
|
|
|
|
if ptrdata != size {
|
|
|
|
|
panic("reflect: bad layout computation in MapOf")
|
|
|
|
|
}
|
2014-12-19 20:44:18 -08:00
|
|
|
}
|
2013-08-31 09:09:50 -07:00
|
|
|
|
|
|
|
|
b := new(rtype)
|
cmd/compile, runtime: fix placement of map bucket overflow pointer on nacl
On most systems, a pointer is the worst case alignment, so adding
a pointer field at the end of a struct guarantees there will be no
padding added after that field (to satisfy overall struct alignment
due to some more-aligned field also present).
In the runtime, the map implementation needs a quick way to
get to the overflow pointer, which is last in the bucket struct,
so it uses size - sizeof(pointer) as the offset.
NaCl/amd64p32 is the exception, as always.
The worst case alignment is 64 bits but pointers are 32 bits.
There's a long history that is not worth going into, but when
we moved the overflow pointer to the end of the struct,
we didn't get the padding computation right.
The compiler computed the regular struct size and then
on amd64p32 added another 32-bit field.
And the runtime assumed it could step back two 32-bit fields
(one 64-bit register size) to get to the overflow pointer.
But in fact if the struct needed 64-bit alignment, the computation
of the regular struct size would have added a 32-bit pad already,
and then the code unconditionally added a second 32-bit pad.
This placed the overflow pointer three words from the end, not two.
The last two were padding, and since the runtime was consistent
about using the second-to-last word as the overflow pointer,
no harm done in the sense of overwriting useful memory.
But writing the overflow pointer to a non-pointer word of memory
means that the GC can't see the overflow blocks, so it will
collect them prematurely. Then bad things happen.
Correct all this in a few steps:
1. Add an explicit check at the end of the bucket layout in the
compiler that the overflow field is last in the struct, never
followed by padding.
2. When padding is needed on nacl (not always, just when needed),
insert it before the overflow pointer, to preserve the "last in the struct"
property.
3. Let the compiler have the final word on the width of the struct,
by inserting an explicit padding field instead of overwriting the
results of the width computation it does.
4. For the same reason (tell the truth to the compiler), set the type
of the overflow field when we're trying to pretend its not a pointer
(in this case the runtime maintains a list of the overflow blocks
elsewhere).
5. Make the runtime use "last in the struct" as its location algorithm.
This fixes TestTraceStress on nacl/amd64p32.
The 'bad map state' and 'invalid free list' failures no longer occur.
Fixes #11838.
Change-Id: If918887f8f252d988db0a35159944d2b36512f92
Reviewed-on: https://go-review.googlesource.com/12971
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
2015-07-30 22:05:51 -04:00
|
|
|
b.align = ptrSize
|
|
|
|
|
if overflowPad > 0 {
|
|
|
|
|
b.align = 8
|
|
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
b.size = size
|
2015-05-04 21:43:30 -04:00
|
|
|
b.ptrdata = ptrdata
|
2015-01-26 21:04:41 +03:00
|
|
|
b.kind = kind
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
b.gcdata = gcdata
|
2016-02-16 22:23:14 -05:00
|
|
|
s := "bucket(" + ktyp.string + "," + etyp.string + ")"
|
|
|
|
|
b.string = s
|
2013-08-31 09:09:50 -07:00
|
|
|
return b
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// SliceOf returns the slice type with element type t.
|
|
|
|
|
// For example, if t represents int, SliceOf(t) represents []int.
|
|
|
|
|
func SliceOf(t Type) Type {
|
|
|
|
|
typ := t.(*rtype)
|
|
|
|
|
|
|
|
|
|
// Look in cache.
|
|
|
|
|
ckey := cacheKey{Slice, typ, nil, 0}
|
|
|
|
|
if slice := cacheGet(ckey); slice != nil {
|
|
|
|
|
return slice
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Look in known types.
|
2016-02-16 22:23:14 -05:00
|
|
|
s := "[]" + typ.string
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
for _, tt := range typesByString(s) {
|
|
|
|
|
slice := (*sliceType)(unsafe.Pointer(tt))
|
|
|
|
|
if slice.elem == typ {
|
|
|
|
|
return cachePut(ckey, tt)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Make a slice type.
|
|
|
|
|
var islice interface{} = ([]unsafe.Pointer)(nil)
|
|
|
|
|
prototype := *(**sliceType)(unsafe.Pointer(&islice))
|
|
|
|
|
slice := new(sliceType)
|
|
|
|
|
*slice = *prototype
|
2016-02-16 22:23:14 -05:00
|
|
|
slice.string = s
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
slice.hash = fnv1(typ.hash, '[')
|
|
|
|
|
slice.elem = typ
|
|
|
|
|
slice.uncommonType = nil
|
|
|
|
|
|
|
|
|
|
return cachePut(ckey, &slice.rtype)
|
|
|
|
|
}
|
|
|
|
|
|
2015-06-07 22:47:59 -04:00
|
|
|
// See cmd/compile/internal/gc/reflect.go for derivation of constant.
|
|
|
|
|
const maxPtrmaskBytes = 2048
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// ArrayOf returns the array type with the given count and element type.
|
|
|
|
|
// For example, if t represents int, ArrayOf(5, t) represents [5]int.
|
|
|
|
|
//
|
|
|
|
|
// If the resulting type would be larger than the available address space,
|
|
|
|
|
// ArrayOf panics.
|
2015-01-27 10:04:11 +01:00
|
|
|
func ArrayOf(count int, elem Type) Type {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
typ := elem.(*rtype)
|
2015-01-27 10:04:11 +01:00
|
|
|
// call SliceOf here as it calls cacheGet/cachePut.
|
|
|
|
|
// ArrayOf also calls cacheGet/cachePut and thus may modify the state of
|
|
|
|
|
// the lookupCache mutex.
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
slice := SliceOf(elem)
|
|
|
|
|
|
|
|
|
|
// Look in cache.
|
|
|
|
|
ckey := cacheKey{Array, typ, nil, uintptr(count)}
|
2015-01-27 10:04:11 +01:00
|
|
|
if array := cacheGet(ckey); array != nil {
|
|
|
|
|
return array
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Look in known types.
|
2016-02-16 22:23:14 -05:00
|
|
|
s := "[" + strconv.Itoa(count) + "]" + typ.string
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
for _, tt := range typesByString(s) {
|
2015-01-27 10:04:11 +01:00
|
|
|
array := (*arrayType)(unsafe.Pointer(tt))
|
|
|
|
|
if array.elem == typ {
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
return cachePut(ckey, tt)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Make an array type.
|
|
|
|
|
var iarray interface{} = [1]unsafe.Pointer{}
|
|
|
|
|
prototype := *(**arrayType)(unsafe.Pointer(&iarray))
|
|
|
|
|
array := new(arrayType)
|
|
|
|
|
*array = *prototype
|
2016-02-16 22:23:14 -05:00
|
|
|
array.string = s
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
array.hash = fnv1(typ.hash, '[')
|
|
|
|
|
for n := uint32(count); n > 0; n >>= 8 {
|
|
|
|
|
array.hash = fnv1(array.hash, byte(n))
|
|
|
|
|
}
|
|
|
|
|
array.hash = fnv1(array.hash, ']')
|
|
|
|
|
array.elem = typ
|
|
|
|
|
max := ^uintptr(0) / typ.size
|
|
|
|
|
if uintptr(count) > max {
|
|
|
|
|
panic("reflect.ArrayOf: array size would exceed virtual address space")
|
|
|
|
|
}
|
|
|
|
|
array.size = typ.size * uintptr(count)
|
2015-05-04 21:43:30 -04:00
|
|
|
if count > 0 && typ.ptrdata != 0 {
|
|
|
|
|
array.ptrdata = typ.size*uintptr(count-1) + typ.ptrdata
|
2015-05-04 14:37:45 -04:00
|
|
|
}
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
array.align = typ.align
|
|
|
|
|
array.fieldAlign = typ.fieldAlign
|
|
|
|
|
array.uncommonType = nil
|
|
|
|
|
array.len = uintptr(count)
|
|
|
|
|
array.slice = slice.(*rtype)
|
|
|
|
|
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
array.kind &^= kindNoPointers
|
|
|
|
|
switch {
|
|
|
|
|
case typ.kind&kindNoPointers != 0 || array.size == 0:
|
|
|
|
|
// No pointers.
|
2015-01-27 10:04:11 +01:00
|
|
|
array.kind |= kindNoPointers
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
array.gcdata = nil
|
|
|
|
|
array.ptrdata = 0
|
|
|
|
|
|
|
|
|
|
case count == 1:
|
|
|
|
|
// In memory, 1-element array looks just like the element.
|
|
|
|
|
array.kind |= typ.kind & kindGCProg
|
|
|
|
|
array.gcdata = typ.gcdata
|
|
|
|
|
array.ptrdata = typ.ptrdata
|
|
|
|
|
|
2015-06-07 22:47:59 -04:00
|
|
|
case typ.kind&kindGCProg == 0 && array.size <= maxPtrmaskBytes*8*ptrSize:
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
// Element is small with pointer mask; array is still small.
|
|
|
|
|
// Create direct pointer mask by turning each 1 bit in elem
|
|
|
|
|
// into count 1 bits in larger mask.
|
|
|
|
|
mask := make([]byte, (array.ptrdata/ptrSize+7)/8)
|
|
|
|
|
elemMask := (*[1 << 30]byte)(unsafe.Pointer(typ.gcdata))[:]
|
|
|
|
|
elemWords := typ.size / ptrSize
|
|
|
|
|
for j := uintptr(0); j < typ.ptrdata/ptrSize; j++ {
|
|
|
|
|
if (elemMask[j/8]>>(j%8))&1 != 0 {
|
|
|
|
|
for i := uintptr(0); i < array.len; i++ {
|
|
|
|
|
k := i*elemWords + j
|
|
|
|
|
mask[k/8] |= 1 << (k % 8)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
array.gcdata = &mask[0]
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
// Create program that emits one element
|
|
|
|
|
// and then repeats to make the array.
|
|
|
|
|
prog := []byte{0, 0, 0, 0} // will be length of prog
|
|
|
|
|
elemGC := (*[1 << 30]byte)(unsafe.Pointer(typ.gcdata))[:]
|
|
|
|
|
elemPtrs := typ.ptrdata / ptrSize
|
|
|
|
|
if typ.kind&kindGCProg == 0 {
|
|
|
|
|
// Element is small with pointer mask; use as literal bits.
|
|
|
|
|
mask := elemGC
|
|
|
|
|
// Emit 120-bit chunks of full bytes (max is 127 but we avoid using partial bytes).
|
|
|
|
|
var n uintptr
|
|
|
|
|
for n = elemPtrs; n > 120; n -= 120 {
|
|
|
|
|
prog = append(prog, 120)
|
|
|
|
|
prog = append(prog, mask[:15]...)
|
|
|
|
|
mask = mask[15:]
|
|
|
|
|
}
|
|
|
|
|
prog = append(prog, byte(n))
|
|
|
|
|
prog = append(prog, mask[:(n+7)/8]...)
|
|
|
|
|
} else {
|
|
|
|
|
// Element has GC program; emit one element.
|
|
|
|
|
elemProg := elemGC[4 : 4+*(*uint32)(unsafe.Pointer(&elemGC[0]))-1]
|
|
|
|
|
prog = append(prog, elemProg...)
|
|
|
|
|
}
|
|
|
|
|
// Pad from ptrdata to size.
|
|
|
|
|
elemWords := typ.size / ptrSize
|
|
|
|
|
if elemPtrs < elemWords {
|
|
|
|
|
// Emit literal 0 bit, then repeat as needed.
|
|
|
|
|
prog = append(prog, 0x01, 0x00)
|
|
|
|
|
if elemPtrs+1 < elemWords {
|
|
|
|
|
prog = append(prog, 0x81)
|
|
|
|
|
prog = appendVarint(prog, elemWords-elemPtrs-1)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
// Repeat count-1 times.
|
|
|
|
|
if elemWords < 0x80 {
|
|
|
|
|
prog = append(prog, byte(elemWords|0x80))
|
|
|
|
|
} else {
|
|
|
|
|
prog = append(prog, 0x80)
|
|
|
|
|
prog = appendVarint(prog, elemWords)
|
|
|
|
|
}
|
|
|
|
|
prog = appendVarint(prog, uintptr(count)-1)
|
|
|
|
|
prog = append(prog, 0)
|
|
|
|
|
*(*uint32)(unsafe.Pointer(&prog[0])) = uint32(len(prog) - 4)
|
|
|
|
|
array.kind |= kindGCProg
|
|
|
|
|
array.gcdata = &prog[0]
|
|
|
|
|
array.ptrdata = array.size // overestimate but ok; must match program
|
2015-01-27 10:04:11 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
etyp := typ.common()
|
|
|
|
|
esize := etyp.Size()
|
|
|
|
|
ealg := etyp.alg
|
|
|
|
|
|
|
|
|
|
array.alg = new(typeAlg)
|
|
|
|
|
if ealg.equal != nil {
|
|
|
|
|
eequal := ealg.equal
|
|
|
|
|
array.alg.equal = func(p, q unsafe.Pointer) bool {
|
|
|
|
|
for i := 0; i < count; i++ {
|
|
|
|
|
pi := arrayAt(p, i, esize)
|
|
|
|
|
qi := arrayAt(q, i, esize)
|
|
|
|
|
if !eequal(pi, qi) {
|
|
|
|
|
return false
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
}
|
|
|
|
|
return true
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
if ealg.hash != nil {
|
|
|
|
|
ehash := ealg.hash
|
|
|
|
|
array.alg.hash = func(ptr unsafe.Pointer, seed uintptr) uintptr {
|
|
|
|
|
o := seed
|
|
|
|
|
for i := 0; i < count; i++ {
|
|
|
|
|
o = ehash(arrayAt(ptr, i, esize), o)
|
|
|
|
|
}
|
|
|
|
|
return o
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
switch {
|
|
|
|
|
case count == 1 && !ifaceIndir(typ):
|
|
|
|
|
// array of 1 direct iface type can be direct
|
|
|
|
|
array.kind |= kindDirectIface
|
|
|
|
|
default:
|
|
|
|
|
array.kind &^= kindDirectIface
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
return cachePut(ckey, &array.rtype)
|
|
|
|
|
}
|
|
|
|
|
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
func appendVarint(x []byte, v uintptr) []byte {
|
|
|
|
|
for ; v >= 0x80; v >>= 7 {
|
|
|
|
|
x = append(x, byte(v|0x80))
|
|
|
|
|
}
|
|
|
|
|
x = append(x, byte(v))
|
|
|
|
|
return x
|
|
|
|
|
}
|
|
|
|
|
|
reflect: add ArrayOf, ChanOf, MapOf, SliceOf
In order to add these, we need to be able to find references
to such types that already exist in the binary. To do that, introduce
a new linker section holding a list of the types corresponding to
arrays, chans, maps, and slices.
To offset the storage cost of this list, and to simplify the code,
remove the interface{} header from the representation of a
runtime type. It was used in early versions of the code but was
made obsolete by the kind field: a switch on kind is more efficient
than a type switch.
In the godoc binary, removing the interface{} header cuts two
words from each of about 10,000 types. Adding back the list of pointers
to array, chan, map, and slice types reintroduces one word for
each of about 500 types. On a 64-bit machine, then, this CL *removes*
a net 156 kB of read-only data from the binary.
This CL does not include the needed support for precise garbage
collection. I have created issue 4375 to track that.
This CL also does not set the 'algorithm' - specifically the equality
and copy functions - for a new array correctly, so I have unexported
ArrayOf for now. That is also part of issue 4375.
Fixes #2339.
R=r, remyoudompheng, mirtchovski, iant
CC=golang-dev
https://golang.org/cl/6572043
2012-11-13 13:06:29 -05:00
|
|
|
// toType converts from a *rtype to a Type that can be returned
|
|
|
|
|
// to the client of package reflect. In gc, the only concern is that
|
|
|
|
|
// a nil *rtype must be replaced by a nil Type, but in gccgo this
|
|
|
|
|
// function takes care of ensuring that multiple *rtype for the same
|
|
|
|
|
// type are coalesced into a single Type.
|
|
|
|
|
func toType(t *rtype) Type {
|
|
|
|
|
if t == nil {
|
|
|
|
|
return nil
|
|
|
|
|
}
|
|
|
|
|
return t
|
|
|
|
|
}
|
2014-01-15 13:56:59 -08:00
|
|
|
|
|
|
|
|
type layoutKey struct {
|
|
|
|
|
t *rtype // function signature
|
|
|
|
|
rcvr *rtype // receiver type, or nil if none
|
|
|
|
|
}
|
|
|
|
|
|
2014-03-20 22:22:07 +01:00
|
|
|
type layoutType struct {
|
|
|
|
|
t *rtype
|
|
|
|
|
argSize uintptr // size of arguments
|
|
|
|
|
retOffset uintptr // offset of return values.
|
2014-09-12 07:29:19 -04:00
|
|
|
stack *bitVector
|
2014-12-22 22:31:55 +03:00
|
|
|
framePool *sync.Pool
|
2014-03-20 22:22:07 +01:00
|
|
|
}
|
|
|
|
|
|
2014-01-15 13:56:59 -08:00
|
|
|
var layoutCache struct {
|
|
|
|
|
sync.RWMutex
|
2014-03-20 22:22:07 +01:00
|
|
|
m map[layoutKey]layoutType
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// funcLayout computes a struct type representing the layout of the
|
|
|
|
|
// function arguments and return values for the function type t.
|
|
|
|
|
// If rcvr != nil, rcvr specifies the type of the receiver.
|
|
|
|
|
// The returned type exists only for GC, so we only fill out GC relevant info.
|
|
|
|
|
// Currently, that's just size and the GC program. We also fill in
|
|
|
|
|
// the name for possible debugging use.
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
func funcLayout(t *rtype, rcvr *rtype) (frametype *rtype, argSize, retOffset uintptr, stk *bitVector, framePool *sync.Pool) {
|
2014-01-15 13:56:59 -08:00
|
|
|
if t.Kind() != Func {
|
2014-04-16 11:52:27 -04:00
|
|
|
panic("reflect: funcLayout of non-func type")
|
|
|
|
|
}
|
|
|
|
|
if rcvr != nil && rcvr.Kind() == Interface {
|
|
|
|
|
panic("reflect: funcLayout with interface receiver " + rcvr.String())
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
|
|
|
|
k := layoutKey{t, rcvr}
|
|
|
|
|
layoutCache.RLock()
|
2014-03-20 22:22:07 +01:00
|
|
|
if x := layoutCache.m[k]; x.t != nil {
|
2014-01-15 13:56:59 -08:00
|
|
|
layoutCache.RUnlock()
|
2014-12-22 22:31:55 +03:00
|
|
|
return x.t, x.argSize, x.retOffset, x.stack, x.framePool
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
|
|
|
|
layoutCache.RUnlock()
|
|
|
|
|
layoutCache.Lock()
|
2014-03-20 22:22:07 +01:00
|
|
|
if x := layoutCache.m[k]; x.t != nil {
|
2014-01-15 13:56:59 -08:00
|
|
|
layoutCache.Unlock()
|
2014-12-22 22:31:55 +03:00
|
|
|
return x.t, x.argSize, x.retOffset, x.stack, x.framePool
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
tt := (*funcType)(unsafe.Pointer(t))
|
|
|
|
|
|
2014-09-12 07:29:19 -04:00
|
|
|
// compute gc program & stack bitmap for arguments
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
ptrmap := new(bitVector)
|
2014-09-12 07:29:19 -04:00
|
|
|
var offset uintptr
|
2014-01-15 13:56:59 -08:00
|
|
|
if rcvr != nil {
|
|
|
|
|
// Reflect uses the "interface" calling convention for
|
|
|
|
|
// methods, where receivers take one word of argument
|
|
|
|
|
// space no matter how big they actually are.
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
if ifaceIndir(rcvr) || rcvr.pointers() {
|
|
|
|
|
ptrmap.append(1)
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
2014-09-12 07:29:19 -04:00
|
|
|
offset += ptrSize
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
|
|
|
|
for _, arg := range tt.in {
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
offset += -offset & uintptr(arg.align-1)
|
|
|
|
|
addTypeBits(ptrmap, offset, arg)
|
|
|
|
|
offset += arg.size
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
argN := ptrmap.n
|
|
|
|
|
argSize = offset
|
2014-03-20 22:22:07 +01:00
|
|
|
if runtime.GOARCH == "amd64p32" {
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
offset += -offset & (8 - 1)
|
2014-03-20 22:22:07 +01:00
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
offset += -offset & (ptrSize - 1)
|
|
|
|
|
retOffset = offset
|
2014-01-15 13:56:59 -08:00
|
|
|
for _, res := range tt.out {
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
offset += -offset & uintptr(res.align-1)
|
|
|
|
|
addTypeBits(ptrmap, offset, res)
|
|
|
|
|
offset += res.size
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
offset += -offset & (ptrSize - 1)
|
2014-01-15 13:56:59 -08:00
|
|
|
|
|
|
|
|
// build dummy rtype holding gc program
|
|
|
|
|
x := new(rtype)
|
cmd/compile, runtime: fix placement of map bucket overflow pointer on nacl
On most systems, a pointer is the worst case alignment, so adding
a pointer field at the end of a struct guarantees there will be no
padding added after that field (to satisfy overall struct alignment
due to some more-aligned field also present).
In the runtime, the map implementation needs a quick way to
get to the overflow pointer, which is last in the bucket struct,
so it uses size - sizeof(pointer) as the offset.
NaCl/amd64p32 is the exception, as always.
The worst case alignment is 64 bits but pointers are 32 bits.
There's a long history that is not worth going into, but when
we moved the overflow pointer to the end of the struct,
we didn't get the padding computation right.
The compiler computed the regular struct size and then
on amd64p32 added another 32-bit field.
And the runtime assumed it could step back two 32-bit fields
(one 64-bit register size) to get to the overflow pointer.
But in fact if the struct needed 64-bit alignment, the computation
of the regular struct size would have added a 32-bit pad already,
and then the code unconditionally added a second 32-bit pad.
This placed the overflow pointer three words from the end, not two.
The last two were padding, and since the runtime was consistent
about using the second-to-last word as the overflow pointer,
no harm done in the sense of overwriting useful memory.
But writing the overflow pointer to a non-pointer word of memory
means that the GC can't see the overflow blocks, so it will
collect them prematurely. Then bad things happen.
Correct all this in a few steps:
1. Add an explicit check at the end of the bucket layout in the
compiler that the overflow field is last in the struct, never
followed by padding.
2. When padding is needed on nacl (not always, just when needed),
insert it before the overflow pointer, to preserve the "last in the struct"
property.
3. Let the compiler have the final word on the width of the struct,
by inserting an explicit padding field instead of overwriting the
results of the width computation it does.
4. For the same reason (tell the truth to the compiler), set the type
of the overflow field when we're trying to pretend its not a pointer
(in this case the runtime maintains a list of the overflow blocks
elsewhere).
5. Make the runtime use "last in the struct" as its location algorithm.
This fixes TestTraceStress on nacl/amd64p32.
The 'bad map state' and 'invalid free list' failures no longer occur.
Fixes #11838.
Change-Id: If918887f8f252d988db0a35159944d2b36512f92
Reviewed-on: https://go-review.googlesource.com/12971
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
2015-07-30 22:05:51 -04:00
|
|
|
x.align = ptrSize
|
|
|
|
|
if runtime.GOARCH == "amd64p32" {
|
|
|
|
|
x.align = 8
|
|
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
x.size = offset
|
|
|
|
|
x.ptrdata = uintptr(ptrmap.n) * ptrSize
|
|
|
|
|
if ptrmap.n > 0 {
|
|
|
|
|
x.gcdata = &ptrmap.data[0]
|
|
|
|
|
} else {
|
2014-12-22 19:07:05 -08:00
|
|
|
x.kind |= kindNoPointers
|
|
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
ptrmap.n = argN
|
|
|
|
|
|
2014-01-15 13:56:59 -08:00
|
|
|
var s string
|
|
|
|
|
if rcvr != nil {
|
2016-02-16 22:23:14 -05:00
|
|
|
s = "methodargs(" + rcvr.string + ")(" + t.string + ")"
|
2014-01-15 13:56:59 -08:00
|
|
|
} else {
|
2016-02-16 22:23:14 -05:00
|
|
|
s = "funcargs(" + t.string + ")"
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
2016-02-16 22:23:14 -05:00
|
|
|
x.string = s
|
2014-01-15 13:56:59 -08:00
|
|
|
|
|
|
|
|
// cache result for future callers
|
|
|
|
|
if layoutCache.m == nil {
|
2014-03-20 22:22:07 +01:00
|
|
|
layoutCache.m = make(map[layoutKey]layoutType)
|
|
|
|
|
}
|
2014-12-22 22:31:55 +03:00
|
|
|
framePool = &sync.Pool{New: func() interface{} {
|
|
|
|
|
return unsafe_New(x)
|
|
|
|
|
}}
|
2014-03-20 22:22:07 +01:00
|
|
|
layoutCache.m[k] = layoutType{
|
|
|
|
|
t: x,
|
|
|
|
|
argSize: argSize,
|
|
|
|
|
retOffset: retOffset,
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
stack: ptrmap,
|
2014-12-22 22:31:55 +03:00
|
|
|
framePool: framePool,
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
|
|
|
|
layoutCache.Unlock()
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
return x, argSize, retOffset, ptrmap, framePool
|
2014-01-15 13:56:59 -08:00
|
|
|
}
|
2014-08-18 21:13:11 -04:00
|
|
|
|
2014-10-15 14:24:18 -04:00
|
|
|
// ifaceIndir reports whether t is stored indirectly in an interface value.
|
|
|
|
|
func ifaceIndir(t *rtype) bool {
|
|
|
|
|
return t.kind&kindDirectIface == 0
|
2014-08-18 21:13:11 -04:00
|
|
|
}
|
2014-09-12 07:29:19 -04:00
|
|
|
|
|
|
|
|
// Layout matches runtime.BitVector (well enough).
|
|
|
|
|
type bitVector struct {
|
|
|
|
|
n uint32 // number of bits
|
|
|
|
|
data []byte
|
|
|
|
|
}
|
|
|
|
|
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
// append a bit to the bitmap.
|
|
|
|
|
func (bv *bitVector) append(bit uint8) {
|
2014-09-12 07:29:19 -04:00
|
|
|
if bv.n%8 == 0 {
|
|
|
|
|
bv.data = append(bv.data, 0)
|
|
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
bv.data[bv.n/8] |= bit << (bv.n % 8)
|
|
|
|
|
bv.n++
|
2014-09-12 07:29:19 -04:00
|
|
|
}
|
|
|
|
|
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
func addTypeBits(bv *bitVector, offset uintptr, t *rtype) {
|
|
|
|
|
if t.kind&kindNoPointers != 0 {
|
2014-09-12 07:29:19 -04:00
|
|
|
return
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
switch Kind(t.kind & kindMask) {
|
|
|
|
|
case Chan, Func, Map, Ptr, Slice, String, UnsafePointer:
|
|
|
|
|
// 1 pointer at start of representation
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
for bv.n < uint32(offset/uintptr(ptrSize)) {
|
|
|
|
|
bv.append(0)
|
2014-09-12 07:29:19 -04:00
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
bv.append(1)
|
2014-09-12 07:29:19 -04:00
|
|
|
|
|
|
|
|
case Interface:
|
|
|
|
|
// 2 pointers
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
for bv.n < uint32(offset/uintptr(ptrSize)) {
|
|
|
|
|
bv.append(0)
|
2014-09-12 07:29:19 -04:00
|
|
|
}
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
bv.append(1)
|
|
|
|
|
bv.append(1)
|
2014-09-12 07:29:19 -04:00
|
|
|
|
|
|
|
|
case Array:
|
|
|
|
|
// repeat inner type
|
|
|
|
|
tt := (*arrayType)(unsafe.Pointer(t))
|
|
|
|
|
for i := 0; i < int(tt.len); i++ {
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
addTypeBits(bv, offset+uintptr(i)*tt.elem.size, tt.elem)
|
2014-09-12 07:29:19 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
case Struct:
|
|
|
|
|
// apply fields
|
|
|
|
|
tt := (*structType)(unsafe.Pointer(t))
|
|
|
|
|
for i := range tt.fields {
|
|
|
|
|
f := &tt.fields[i]
|
runtime: replace GC programs with simpler encoding, faster decoder
Small types record the location of pointers in their memory layout
by using a simple bitmap. In Go 1.4 the bitmap held 4-bit entries,
and in Go 1.5 the bitmap holds 1-bit entries, but in both cases using
a bitmap for a large type containing arrays does not make sense:
if someone refers to the type [1<<28]*byte in a program in such
a way that the type information makes it into the binary, it would be
a waste of space to write a 128 MB (for 4-bit entries) or even 32 MB
(for 1-bit entries) bitmap full of 1s into the binary or even to keep
one in memory during the execution of the program.
For large types containing arrays, it is much more compact to describe
the locations of pointers using a notation that can express repetition
than to lay out a bitmap of pointers. Go 1.4 included such a notation,
called ``GC programs'' but it was complex, required recursion during
decoding, and was generally slow. Dmitriy measured the execution of
these programs writing directly to the heap bitmap as being 7x slower
than copying from a preunrolled 4-bit mask (and frankly that code was
not terribly fast either). For some tests, unrollgcprog1 was seen costing
as much as 3x more than the rest of malloc combined.
This CL introduces a different form for the GC programs. They use a
simple Lempel-Ziv-style encoding of the 1-bit pointer information,
in which the only operations are (1) emit the following n bits
and (2) repeat the last n bits c more times. This encoding can be
generated directly from the Go type information (using repetition
only for arrays or large runs of non-pointer data) and it can be decoded
very efficiently. In particular the decoding requires little state and
no recursion, so that the entire decoding can run without any memory
accesses other than the reads of the encoding and the writes of the
decoded form to the heap bitmap. For recursive types like arrays of
arrays of arrays, the inner instructions are only executed once, not
n times, so that large repetitions run at full speed. (In contrast, large
repetitions in the old programs repeated the individual bit-level layout
of the inner data over and over.) The result is as much as 25x faster
decoding compared to the old form.
Because the old decoder was so slow, Go 1.4 had three (or so) cases
for how to set the heap bitmap bits for an allocation of a given type:
(1) If the type had an even number of words up to 32 words, then
the 4-bit pointer mask for the type fit in no more than 16 bytes;
store the 4-bit pointer mask directly in the binary and copy from it.
(1b) If the type had an odd number of words up to 15 words, then
the 4-bit pointer mask for the type, doubled to end on a byte boundary,
fit in no more than 16 bytes; store that doubled mask directly in the
binary and copy from it.
(2) If the type had an even number of words up to 128 words,
or an odd number of words up to 63 words (again due to doubling),
then the 4-bit pointer mask would fit in a 64-byte unrolled mask.
Store a GC program in the binary, but leave space in the BSS for
the unrolled mask. Execute the GC program to construct the mask the
first time it is needed, and thereafter copy from the mask.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
(This is the case that was 7x slower than the other two.)
Because the new pointer masks store 1-bit entries instead of 4-bit
entries and because using the decoder no longer carries a significant
overhead, after this CL (that is, for Go 1.5) there are only two cases:
(1) If the type is 128 words or less (no condition about odd or even),
store the 1-bit pointer mask directly in the binary and use it to
initialize the heap bitmap during malloc. (Implemented in CL 9702.)
(2) There is no case 2 anymore.
(3) Otherwise, store a GC program and execute it to write directly to
the heap bitmap each time an object of that type is allocated.
Executing the GC program directly into the heap bitmap (case (3) above)
was disabled for the Go 1.5 dev cycle, both to avoid needing to use
GC programs for typedmemmove and to avoid updating that code as
the heap bitmap format changed. Typedmemmove no longer uses this
type information; as of CL 9886 it uses the heap bitmap directly.
Now that the heap bitmap format is stable, we reintroduce GC programs
and their space savings.
Benchmarks for heapBitsSetType, before this CL vs this CL:
name old mean new mean delta
SetTypePtr 7.59ns × (0.99,1.02) 5.16ns × (1.00,1.00) -32.05% (p=0.000)
SetTypePtr8 21.0ns × (0.98,1.05) 21.4ns × (1.00,1.00) ~ (p=0.179)
SetTypePtr16 24.1ns × (0.99,1.01) 24.6ns × (1.00,1.00) +2.41% (p=0.001)
SetTypePtr32 31.2ns × (0.99,1.01) 32.4ns × (0.99,1.02) +3.72% (p=0.001)
SetTypePtr64 45.2ns × (1.00,1.00) 47.2ns × (1.00,1.00) +4.42% (p=0.000)
SetTypePtr126 75.8ns × (0.99,1.01) 79.1ns × (1.00,1.00) +4.25% (p=0.000)
SetTypePtr128 74.3ns × (0.99,1.01) 77.6ns × (1.00,1.01) +4.55% (p=0.000)
SetTypePtrSlice 726ns × (1.00,1.01) 712ns × (1.00,1.00) -1.95% (p=0.001)
SetTypeNode1 20.0ns × (0.99,1.01) 20.7ns × (1.00,1.00) +3.71% (p=0.000)
SetTypeNode1Slice 112ns × (1.00,1.00) 113ns × (0.99,1.00) ~ (p=0.070)
SetTypeNode8 23.9ns × (1.00,1.00) 24.7ns × (1.00,1.01) +3.18% (p=0.000)
SetTypeNode8Slice 294ns × (0.99,1.02) 287ns × (0.99,1.01) -2.38% (p=0.015)
SetTypeNode64 52.8ns × (0.99,1.03) 51.8ns × (0.99,1.01) ~ (p=0.069)
SetTypeNode64Slice 1.13µs × (0.99,1.05) 1.14µs × (0.99,1.00) ~ (p=0.767)
SetTypeNode64Dead 36.0ns × (1.00,1.01) 32.5ns × (0.99,1.00) -9.67% (p=0.000)
SetTypeNode64DeadSlice 1.43µs × (0.99,1.01) 1.40µs × (1.00,1.00) -2.39% (p=0.001)
SetTypeNode124 75.7ns × (1.00,1.01) 79.0ns × (1.00,1.00) +4.44% (p=0.000)
SetTypeNode124Slice 1.94µs × (1.00,1.01) 2.04µs × (0.99,1.01) +4.98% (p=0.000)
SetTypeNode126 75.4ns × (1.00,1.01) 77.7ns × (0.99,1.01) +3.11% (p=0.000)
SetTypeNode126Slice 1.95µs × (0.99,1.01) 2.03µs × (1.00,1.00) +3.74% (p=0.000)
SetTypeNode128 85.4ns × (0.99,1.01) 122.0ns × (1.00,1.00) +42.89% (p=0.000)
SetTypeNode128Slice 2.20µs × (1.00,1.01) 2.36µs × (0.98,1.02) +7.48% (p=0.001)
SetTypeNode130 83.3ns × (1.00,1.00) 123.0ns × (1.00,1.00) +47.61% (p=0.000)
SetTypeNode130Slice 2.30µs × (0.99,1.01) 2.40µs × (0.98,1.01) +4.37% (p=0.000)
SetTypeNode1024 498ns × (1.00,1.00) 537ns × (1.00,1.00) +7.96% (p=0.000)
SetTypeNode1024Slice 15.5µs × (0.99,1.01) 17.8µs × (1.00,1.00) +15.27% (p=0.000)
The above compares always using a cached pointer mask (and the
corresponding waste of memory) against using the programs directly.
Some slowdown is expected, in exchange for having a better general algorithm.
The GC programs kick in for SetTypeNode128, SetTypeNode130, SetTypeNode1024,
along with the slice variants of those.
It is possible that the cutoff of 128 words (bits) should be raised
in a followup CL, but even with this low cutoff the GC programs are
faster than Go 1.4's "fast path" non-GC program case.
Benchmarks for heapBitsSetType, Go 1.4 vs this CL:
name old mean new mean delta
SetTypePtr 6.89ns × (1.00,1.00) 5.17ns × (1.00,1.00) -25.02% (p=0.000)
SetTypePtr8 25.8ns × (0.97,1.05) 21.5ns × (1.00,1.00) -16.70% (p=0.000)
SetTypePtr16 39.8ns × (0.97,1.02) 24.7ns × (0.99,1.01) -37.81% (p=0.000)
SetTypePtr32 68.8ns × (0.98,1.01) 32.2ns × (1.00,1.01) -53.18% (p=0.000)
SetTypePtr64 130ns × (1.00,1.00) 47ns × (1.00,1.00) -63.67% (p=0.000)
SetTypePtr126 241ns × (0.99,1.01) 79ns × (1.00,1.01) -67.25% (p=0.000)
SetTypePtr128 2.07µs × (1.00,1.00) 0.08µs × (1.00,1.00) -96.27% (p=0.000)
SetTypePtrSlice 1.05µs × (0.99,1.01) 0.72µs × (0.99,1.02) -31.70% (p=0.000)
SetTypeNode1 16.0ns × (0.99,1.01) 20.8ns × (0.99,1.03) +29.91% (p=0.000)
SetTypeNode1Slice 184ns × (0.99,1.01) 112ns × (0.99,1.01) -39.26% (p=0.000)
SetTypeNode8 29.5ns × (0.97,1.02) 24.6ns × (1.00,1.00) -16.50% (p=0.000)
SetTypeNode8Slice 624ns × (0.98,1.02) 285ns × (1.00,1.00) -54.31% (p=0.000)
SetTypeNode64 135ns × (0.96,1.08) 52ns × (0.99,1.02) -61.32% (p=0.000)
SetTypeNode64Slice 3.83µs × (1.00,1.00) 1.14µs × (0.99,1.01) -70.16% (p=0.000)
SetTypeNode64Dead 134ns × (0.99,1.01) 32ns × (1.00,1.01) -75.74% (p=0.000)
SetTypeNode64DeadSlice 3.83µs × (0.99,1.00) 1.40µs × (1.00,1.01) -63.42% (p=0.000)
SetTypeNode124 240ns × (0.99,1.01) 79ns × (1.00,1.01) -67.05% (p=0.000)
SetTypeNode124Slice 7.27µs × (1.00,1.00) 2.04µs × (1.00,1.00) -71.95% (p=0.000)
SetTypeNode126 2.06µs × (0.99,1.01) 0.08µs × (0.99,1.01) -96.23% (p=0.000)
SetTypeNode126Slice 64.4µs × (1.00,1.00) 2.0µs × (1.00,1.00) -96.85% (p=0.000)
SetTypeNode128 2.09µs × (1.00,1.01) 0.12µs × (1.00,1.00) -94.15% (p=0.000)
SetTypeNode128Slice 65.4µs × (1.00,1.00) 2.4µs × (0.99,1.03) -96.39% (p=0.000)
SetTypeNode130 2.11µs × (1.00,1.00) 0.12µs × (1.00,1.00) -94.18% (p=0.000)
SetTypeNode130Slice 66.3µs × (1.00,1.00) 2.4µs × (0.97,1.08) -96.34% (p=0.000)
SetTypeNode1024 16.0µs × (1.00,1.01) 0.5µs × (1.00,1.00) -96.65% (p=0.000)
SetTypeNode1024Slice 512µs × (1.00,1.00) 18µs × (0.98,1.04) -96.45% (p=0.000)
SetTypeNode124 uses a 124 data + 2 ptr = 126-word allocation.
Both Go 1.4 and this CL are using pointer bitmaps for this case,
so that's an overall 3x speedup for using pointer bitmaps.
SetTypeNode128 uses a 128 data + 2 ptr = 130-word allocation.
Both Go 1.4 and this CL are running the GC program for this case,
so that's an overall 17x speedup when using GC programs (and
I've seen >20x on other systems).
Comparing Go 1.4's SetTypeNode124 (pointer bitmap) against
this CL's SetTypeNode128 (GC program), the slow path in the
code in this CL is 2x faster than the fast path in Go 1.4.
The Go 1 benchmarks are basically unaffected compared to just before this CL.
Go 1 benchmarks, before this CL vs this CL:
name old mean new mean delta
BinaryTree17 5.87s × (0.97,1.04) 5.91s × (0.96,1.04) ~ (p=0.306)
Fannkuch11 4.38s × (1.00,1.00) 4.37s × (1.00,1.01) -0.22% (p=0.006)
FmtFprintfEmpty 90.7ns × (0.97,1.10) 89.3ns × (0.96,1.09) ~ (p=0.280)
FmtFprintfString 282ns × (0.98,1.04) 287ns × (0.98,1.07) +1.72% (p=0.039)
FmtFprintfInt 269ns × (0.99,1.03) 282ns × (0.97,1.04) +4.87% (p=0.000)
FmtFprintfIntInt 478ns × (0.99,1.02) 481ns × (0.99,1.02) +0.61% (p=0.048)
FmtFprintfPrefixedInt 399ns × (0.98,1.03) 400ns × (0.98,1.05) ~ (p=0.533)
FmtFprintfFloat 563ns × (0.99,1.01) 570ns × (1.00,1.01) +1.37% (p=0.000)
FmtManyArgs 1.89µs × (0.99,1.01) 1.92µs × (0.99,1.02) +1.88% (p=0.000)
GobDecode 15.2ms × (0.99,1.01) 15.2ms × (0.98,1.05) ~ (p=0.609)
GobEncode 11.6ms × (0.98,1.03) 11.9ms × (0.98,1.04) +2.17% (p=0.000)
Gzip 648ms × (0.99,1.01) 648ms × (1.00,1.01) ~ (p=0.835)
Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) ~ (p=0.169)
HTTPClientServer 90.5µs × (0.98,1.03) 91.5µs × (0.98,1.04) +1.04% (p=0.045)
JSONEncode 31.5ms × (0.98,1.03) 31.4ms × (0.98,1.03) ~ (p=0.549)
JSONDecode 111ms × (0.99,1.01) 107ms × (0.99,1.01) -3.21% (p=0.000)
Mandelbrot200 6.01ms × (1.00,1.00) 6.01ms × (1.00,1.00) ~ (p=0.878)
GoParse 6.54ms × (0.99,1.02) 6.61ms × (0.99,1.03) +1.08% (p=0.004)
RegexpMatchEasy0_32 160ns × (1.00,1.01) 161ns × (1.00,1.00) +0.40% (p=0.000)
RegexpMatchEasy0_1K 560ns × (0.99,1.01) 559ns × (0.99,1.01) ~ (p=0.088)
RegexpMatchEasy1_32 138ns × (0.99,1.01) 138ns × (1.00,1.00) ~ (p=0.380)
RegexpMatchEasy1_1K 877ns × (1.00,1.00) 878ns × (1.00,1.00) ~ (p=0.157)
RegexpMatchMedium_32 251ns × (0.99,1.00) 251ns × (1.00,1.01) +0.28% (p=0.021)
RegexpMatchMedium_1K 72.6µs × (1.00,1.00) 72.6µs × (1.00,1.00) ~ (p=0.539)
RegexpMatchHard_32 3.84µs × (1.00,1.00) 3.84µs × (1.00,1.00) ~ (p=0.378)
RegexpMatchHard_1K 117µs × (1.00,1.00) 117µs × (1.00,1.00) ~ (p=0.067)
Revcomp 904ms × (0.99,1.02) 904ms × (0.99,1.01) ~ (p=0.943)
Template 125ms × (0.99,1.02) 127ms × (0.99,1.01) +1.79% (p=0.000)
TimeParse 627ns × (0.99,1.01) 622ns × (0.99,1.01) -0.88% (p=0.000)
TimeFormat 655ns × (0.99,1.02) 655ns × (0.99,1.02) ~ (p=0.976)
For the record, Go 1 benchmarks, Go 1.4 vs this CL:
name old mean new mean delta
BinaryTree17 4.61s × (0.97,1.05) 5.91s × (0.98,1.03) +28.35% (p=0.000)
Fannkuch11 4.40s × (0.99,1.03) 4.41s × (0.99,1.01) ~ (p=0.212)
FmtFprintfEmpty 102ns × (0.99,1.01) 84ns × (0.99,1.02) -18.38% (p=0.000)
FmtFprintfString 302ns × (0.98,1.01) 303ns × (0.99,1.02) ~ (p=0.203)
FmtFprintfInt 313ns × (0.97,1.05) 270ns × (0.99,1.01) -13.69% (p=0.000)
FmtFprintfIntInt 524ns × (0.98,1.02) 477ns × (0.99,1.00) -8.87% (p=0.000)
FmtFprintfPrefixedInt 424ns × (0.98,1.02) 386ns × (0.99,1.01) -8.96% (p=0.000)
FmtFprintfFloat 652ns × (0.98,1.02) 594ns × (0.97,1.05) -8.97% (p=0.000)
FmtManyArgs 2.13µs × (0.99,1.02) 1.94µs × (0.99,1.01) -8.92% (p=0.000)
GobDecode 17.1ms × (0.99,1.02) 14.9ms × (0.98,1.03) -13.07% (p=0.000)
GobEncode 13.5ms × (0.98,1.03) 11.5ms × (0.98,1.03) -15.25% (p=0.000)
Gzip 656ms × (0.99,1.02) 647ms × (0.99,1.01) -1.29% (p=0.000)
Gunzip 143ms × (0.99,1.02) 144ms × (0.99,1.01) ~ (p=0.204)
HTTPClientServer 88.2µs × (0.98,1.02) 90.8µs × (0.98,1.01) +2.93% (p=0.000)
JSONEncode 32.2ms × (0.98,1.02) 30.9ms × (0.97,1.04) -4.06% (p=0.001)
JSONDecode 121ms × (0.98,1.02) 110ms × (0.98,1.05) -8.95% (p=0.000)
Mandelbrot200 6.06ms × (0.99,1.01) 6.11ms × (0.98,1.04) ~ (p=0.184)
GoParse 6.76ms × (0.97,1.04) 6.58ms × (0.98,1.05) -2.63% (p=0.003)
RegexpMatchEasy0_32 195ns × (1.00,1.01) 155ns × (0.99,1.01) -20.43% (p=0.000)
RegexpMatchEasy0_1K 479ns × (0.98,1.03) 535ns × (0.99,1.02) +11.59% (p=0.000)
RegexpMatchEasy1_32 169ns × (0.99,1.02) 131ns × (0.99,1.03) -22.44% (p=0.000)
RegexpMatchEasy1_1K 1.53µs × (0.99,1.01) 0.87µs × (0.99,1.02) -43.07% (p=0.000)
RegexpMatchMedium_32 334ns × (0.99,1.01) 242ns × (0.99,1.01) -27.53% (p=0.000)
RegexpMatchMedium_1K 125µs × (1.00,1.01) 72µs × (0.99,1.03) -42.53% (p=0.000)
RegexpMatchHard_32 6.03µs × (0.99,1.01) 3.79µs × (0.99,1.01) -37.12% (p=0.000)
RegexpMatchHard_1K 189µs × (0.99,1.02) 115µs × (0.99,1.01) -39.20% (p=0.000)
Revcomp 935ms × (0.96,1.03) 926ms × (0.98,1.02) ~ (p=0.083)
Template 146ms × (0.97,1.05) 119ms × (0.99,1.01) -18.37% (p=0.000)
TimeParse 660ns × (0.99,1.01) 624ns × (0.99,1.02) -5.43% (p=0.000)
TimeFormat 670ns × (0.98,1.02) 710ns × (1.00,1.01) +5.97% (p=0.000)
This CL is a bit larger than I would like, but the compiler, linker, runtime,
and package reflect all need to be in sync about the format of these programs,
so there is no easy way to split this into independent changes (at least
while keeping the build working at each change).
Fixes #9625.
Fixes #10524.
Change-Id: I9e3e20d6097099d0f8532d1cb5b1af528804989a
Reviewed-on: https://go-review.googlesource.com/9888
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
2015-05-08 01:43:18 -04:00
|
|
|
addTypeBits(bv, offset+f.offset, f.typ)
|
2014-09-12 07:29:19 -04:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|