04 Go 1.4 Release Notes

Go 1.4 Release Notes

Introduction to Go 1.4 Changes to the language For-range loops Method calls on **T Changes to the supported operating systems and architectures Android NaCl on ARM Plan9 on AMD64 Changes to the compatibility guidelines Changes to the implementations and tools Changes to the runtime Assembly Status of gccgo Internal packages Canonical import paths

Import paths for the subrepositories The go generate subcommand Change to file name handling Other changes to the go command Changes to package source layout SWIG Miscellany Performance Changes to the standard library New packages Major changes to the library Minor changes to the library

Introduction to Go 1.4

The latest Go release, version 1.4, arrives as scheduled six months after 1.3.

It contains only one tiny language change, in the form of a backwards-compatible simple variant of for-rangeloop, and a possibly breaking change to the compiler involving methods on pointers-to-pointers.

The release focuses primarily on implementation work, improving the garbage collector and preparing the ground for a fully concurrent collector to be rolled out in the next few releases. Stacks are now contiguous, reallocated when necessary rather than linking on new "segments"; this release therefore eliminates the notorious "hot stack split" problem. There are some new tools available including support in the go command for build-time source code generation. The release also adds support for ARM processors on Android and Native Client (NaCl) and for AMD64 on Plan 9.

As always, Go 1.4 keeps the promise of compatibility, and almost everything will continue to compile and run without change when moved to 1.4.

Changes to the language

For-range loops

Up until Go 1.3, for-range loop had two forms

for i, v := range x {
	...
}

and

for i := range x {
	...
}

If one was not interested in the loop values, only the iteration itself, it was still necessary to mention a variable (probably the blank identifier, as in for _ = range x), because the form

for range x {
	...
}

was not syntactically permitted.

This situation seemed awkward, so as of Go 1.4 the variable-free form is now legal. The pattern arises rarely but the code can be cleaner when it does.

Updating: The change is strictly backwards compatible to existing Go programs, but tools that analyze Go parse trees may need to be modified to accept this new form as the Key field of RangeStmt may now be nil.

Method calls on **T

Given these declarations,

type T int
func (T) M() {}
var x **T

both gc and gccgo accepted the method call

x.M()

which is a double dereference of the pointer-to-pointer x. The Go specification allows a single dereference to be inserted automatically, but not two, so this call is erroneous according to the language definition. It has therefore been disallowed in Go 1.4, which is a breaking change, although very few programs will be affected.

Updating: Code that depends on the old, erroneous behavior will no longer compile but is easy to fix by adding an explicit dereference.

Changes to the supported operating systems and architectures

Android

Go 1.4 can build binaries for ARM processors running the Android operating system. It can also build a .solibrary that can be loaded by an Android application using the supporting packages in the mobile subrepository. A brief description of the plans for this experimental port are available here.

NaCl on ARM

The previous release introduced Native Client (NaCl) support for the 32-bit x86 (GOARCH=386) and 64-bit x86 using 32-bit pointers (GOARCH=amd64p32). The 1.4 release adds NaCl support for ARM (GOARCH=arm).

Plan9 on AMD64

This release adds support for the Plan 9 operating system on AMD64 processors, provided the kernel supports the nsec system call and uses 4K pages.

Changes to the compatibility guidelines

The unsafe package allows one to defeat Go's type system by exploiting internal details of the implementation or machine representation of data. It was never explicitly specified what use of unsafe meant with respect to compatibility as specified in the Go compatibility guidelines. The answer, of course, is that we can make no promise of compatibility for code that does unsafe things.

We have clarified this situation in the documentation included in the release. The Go compatibility guidelines and the docs for the unsafe package are now explicit that unsafe code is not guaranteed to remain compatible.

Updating: Nothing technical has changed; this is just a clarification of the documentation.

Changes to the implementations and tools

Changes to the runtime

Prior to Go 1.4, the runtime (garbage collector, concurrency support, interface management, maps, slices, strings, ...) was mostly written in C, with some assembler support. In 1.4, much of the code has been translated to Go so that the garbage collector can scan the stacks of programs in the runtime and get accurate information about what variables are active. This change was large but should have no semantic effect on programs.

This rewrite allows the garbage collector in 1.4 to be fully precise, meaning that it is aware of the location of all active pointers in the program. This means the heap will be smaller as there will be no false positives keeping non-pointers alive. Other related changes also reduce the heap size, which is smaller by 10%-30% overall relative to the previous release.

A consequence is that stacks are no longer segmented, eliminating the "hot split" problem. When a stack limit is reached, a new, larger stack is allocated, all active frames for the goroutine are copied there, and any pointers into the stack are updated. Performance can be noticeably better in some cases and is always more predictable. Details are available in the design document.

The use of contiguous stacks means that stacks can start smaller without triggering performance issues, so the default starting size for a goroutine's stack in 1.4 has been reduced from 8192 bytes to 2048 bytes.

As preparation for the concurrent garbage collector scheduled for the 1.5 release, writes to pointer values in the heap are now done by a function call, called a write barrier, rather than directly from the function updating the value. In this next release, this will permit the garbage collector to mediate writes to the heap while it is running. This change has no semantic effect on programs in 1.4, but was included in the release to test the compiler and the resulting performance.

The implementation of interface values has been modified. In earlier releases, the interface contained a word that was either a pointer or a one-word scalar value, depending on the type of the concrete object stored. This implementation was problematical for the garbage collector, so as of 1.4 interface values always hold a pointer. In running programs, most interface values were pointers anyway, so the effect is minimal, but programs that store integers (for example) in interfaces will see more allocations.

As of Go 1.3, the runtime crashes if it finds a memory word that should contain a valid pointer but instead contains an obviously invalid pointer (for example, the value 3). Programs that store integers in pointer values may run afoul of this check and crash. In Go 1.4, setting the GODEBUG variable invalidptr=0 disables the crash as a workaround, but we cannot guarantee that future releases will be able to avoid the crash; the correct fix is to rewrite code not to alias integers and pointers.

Assembly

The language accepted by the assemblers cmd/5a, cmd/6a and cmd/8a has had several changes, mostly to make it easier to deliver type information to the runtime.

First, the textflag.h file that defines flags for TEXT directives has been copied from the linker source directory to a standard location so it can be included with the simple directive

#include "textflag.h"

The more important changes are in how assembler source can define the necessary type information. For most programs it will suffice to move data definitions (DATA and GLOBL directives) out of assembly into Go files and to write a Go declaration for each assembly function. The assembly document describes what to do.

Updating: Assembly files that include textflag.h from its old location will still work, but should be updated. For the type information, most assembly routines will need no change, but all should be examined. Assembly source files that define data, functions with non-empty stack frames, or functions that return pointers need particular attention. A description of the necessary (but simple) changes is in the assembly document.

More information about these changes is in the assembly document.

Status of gccgo

The release schedules for the GCC and Go projects do not coincide. GCC release 4.9 contains the Go 1.2 version of gccgo. The next release, GCC 5, will likely have the Go 1.4 version of gccgo.

Internal packages

Go's package system makes it easy to structure programs into components with clean boundaries, but there are only two forms of access: local (unexported) and global (exported). Sometimes one wishes to have components that are not exported, for instance to avoid acquiring clients of interfaces to code that is part of a public repository but not intended for use outside the program to which it belongs.

The Go language does not have the power to enforce this distinction, but as of Go 1.4 the go command introduces a mechanism to define "internal" packages that may not be imported by packages outside the source subtree in which they reside.

To create such a package, place it in a directory named internal or in a subdirectory of a directory named internal. When the go command sees an import of a package with internal in its path, it verifies that the package doing the import is within the tree rooted at the parent of the internal directory. For example, a package .../a/b/c/internal/d/e/f can be imported only by code in the directory tree rooted at .../a/b/c. It cannot be imported by code in .../a/b/g or in any other repository.

For Go 1.4, the internal package mechanism is enforced for the main Go repository; from 1.5 and onward it will be enforced for any repository.

Full details of the mechanism are in the design document.

Canonical import paths

Code often lives in repositories hosted by public services such as github.com, meaning that the import paths for packages begin with the name of the hosting service, github.com/rsc/pdf for example. One can use an existing mechanism to provide a "custom" or "vanity" import path such as rsc.io/pdf, but that creates two valid import paths for the package. That is a problem: one may inadvertently import the package through the two distinct paths in a single program, which is wasteful; miss an update to a package because the path being used is not recognized to be out of date; or break clients using the old path by moving the package to a different hosting service.

Go 1.4 introduces an annotation for package clauses in Go source that identify a canonical import path for the package. If an import is attempted using a path that is not canonical, the go command will refuse to compile the importing package.

The syntax is simple: put an identifying comment on the package line. For our example, the package clause would read:

package pdf // import "rsc.io/pdf"

With this in place, the go command will refuse to compile a package that imports github.com/rsc/pdf, ensuring that the code can be moved without breaking users.

The check is at build time, not download time, so if go get fails because of this check, the mis-imported package has been copied to the local machine and should be removed manually.

To complement this new feature, a check has been added at update time to verify that the local package's remote repository matches that of its custom import. The go get -u command will fail to update a package if its remote repository has changed since it was first downloaded. The new -f flag overrides this check.

Further information is in the design document.

Import paths for the subrepositories

The Go project subrepositories (code.google.com/p/go.tools and so on) are now available under custom import paths replacing code.google.com/p/go. with golang.org/x/, as in golang.org/x/tools. We will add canonical import comments to the code around June 1, 2015, at which point Go 1.4 and later will stop accepting the old code.google.com paths.

Updating: All code that imports from subrepositories should change to use the new golang.org paths. Go 1.0 and later can resolve and import the new paths, so updating will not break compatibility with older releases. Code that has not updated will stop compiling with Go 1.4 around June 1, 2015.

The go generate subcommand

The go command has a new subcommand, go generate, to automate the running of tools to generate source code before compilation. For example, it can be used to run the yacc compiler-compiler on a .y file to produce the Go source file implementing the grammar, or to automate the generation of String methods for typed constants using the new stringer tool in the golang.org/x/tools subrepository.

For more information, see the design document.

Change to file name handling

Build constraints, also known as build tags, control compilation by including or excluding files (see the documentation /go/build). Compilation can also be controlled by the name of the file itself by "tagging" the file with a suffix (before the .go or .s extension) with an underscore and the name of the architecture or operating system. For instance, the file gopher_arm.go will only be compiled if the target processor is an ARM.

Before Go 1.4, a file called just arm.go was similarly tagged, but this behavior can break sources when new architectures are added, causing files to suddenly become tagged. In 1.4, therefore, a file will be tagged in this manner only if the tag (architecture or operating system name) is preceded by an underscore.

Updating: Packages that depend on the old behavior will no longer compile correctly. Files with names like windows.go or amd64.go should either have explicit build tags added to the source or be renamed to something like os_windows.go or support_amd64.go.

Other changes to the go command

There were a number of minor changes to the cmd/go command worth noting.

• Unless cgo is being used to build the package, the go command now refuses to compile C source files, since the relevant C compilers (6c etc.) are intended to be removed from the installation in some future release. (They are used today only to build part of the runtime.) It is difficult to use them correctly in any case, so any extant uses are likely incorrect, so we have disabled them.
• The go test subcommand has a new flag, -o, to set the name of the resulting binary, corresponding to the same flag in other subcommands. The non-functional -file flag has been removed.
• The go test subcommand will compile and link all *_test.go files in the package, even when there are no Testfunctions in them. It previously ignored such files.
• The behavior of the go build subcommand's -a flag has been changed for non-development installations. For installations running a released distribution, the -a flag will no longer rebuild the standard library and commands, to avoid overwriting the installation's files.

Changes to package source layout

In the main Go source repository, the source code for the packages was kept in the directory src/pkg, which made sense but differed from other repositories, including the Go subrepositories. In Go 1.4, the pkg level of the source tree is now gone, so for example the fmt package's source, once kept in directory src/pkg/fmt, now lives one level higher in src/fmt.

Updating: Tools like godoc that discover source code need to know about the new location. All tools and services maintained by the Go team have been updated.

SWIG

Due to runtime changes in this release, Go 1.4 requires SWIG 3.0.3.

Miscellany

The standard repository's top-level misc directory used to contain Go support for editors and IDEs: plugins, initialization scripts and so on. Maintaining these was becoming time-consuming and needed external help because many of the editors listed were not used by members of the core team. It also required us to make decisions about which plugin was best for a given editor, even for editors we do not use.

The Go community at large is much better suited to managing this information. In Go 1.4, therefore, this support has been removed from the repository. Instead, there is a curated, informative list of what's available on a wiki page.

Performance

Most programs will run about the same speed or slightly faster in 1.4 than in 1.3; some will be slightly slower. There are many changes, making it hard to be precise about what to expect.

As mentioned above, much of the runtime was translated to Go from C, which led to some reduction in heap sizes. It also improved performance slightly because the Go compiler is better at optimization, due to things like inlining, than the C compiler used to build the runtime.

The garbage collector was sped up, leading to measurable improvements for garbage-heavy programs. On the other hand, the new write barriers slow things down again, typically by about the same amount but, depending on their behavior, some programs may be somewhat slower or faster.

Library changes that affect performance are documented below.

Changes to the standard library

New packages

There are no new packages in this release.

Major changes to the library

bufio.Scanner

The Scanner type in the bufio package has had a bug fixed that may require changes to custom split functions. The bug made it impossible to generate an empty token at EOF; the fix changes the end conditions seen by the split function. Previously, scanning stopped at EOF if there was no more data. As of 1.4, the split function will be called once at EOF after input is exhausted, so the split function can generate a final empty token as the documentation already promised.

Updating: Custom split functions may need to be modified to handle empty tokens at EOF as desired.

syscall

The syscall package is now frozen except for changes needed to maintain the core repository. In particular, it will no longer be extended to support new or different system calls that are not used by the core. The reasons are described at length in a separate document.

A new subrepository, golang.org/x/sys, has been created to serve as the location for new developments to support system calls on all kernels. It has a nicer structure, with three packages that each hold the implementation of system calls for one of Unix, Windows and Plan 9. These packages will be curated more generously, accepting all reasonable changes that reflect kernel interfaces in those operating systems. See the documentation and the article mentioned above for more information.

Updating: Existing programs are not affected as the syscall package is largely unchanged from the 1.3 release. Future development that requires system calls not in the syscall package should build on golang.org/x/sysinstead.

Minor changes to the library

The following list summarizes a number of minor changes to the library, mostly additions. See the relevant package documentation for more information about each change.

• The archive/zip package's Writer now supports a Flush method.
• The compress/flate, compress/gzip, and compress/zlib packages now support a Reset method for the decompressors, allowing them to reuse buffers and improve performance. The compress/gzip package also has a Multistream method to control support for multistream files.
• The crypto package now has a Signer interface, implemented by the PrivateKey types in crypto/ecdsa andcrypto/rsa.
• The crypto/tls package now supports ALPN as defined in RFC 7301.
• The crypto/tls package now supports programmatic selection of server certificates through the new CertificateForName function of the Config struct.
• Also in the crypto/tls package, the server now supports TLS_FALLBACK_SCSV to help clients detect fallback attacks. (The Go client does not support fallback at all, so it is not vulnerable to those attacks.)
• The database/sql package can now list all registered Drivers.
• The debug/dwarf package now supports UnspecifiedTypes.
• In the encoding/asn1 package, optional elements with a default value will now only be omitted if they have that value.
• The encoding/csv package no longer quotes empty strings but does quote the end-of-data marker \.(backslash dot). This is permitted by the definition of CSV and allows it to work better with Postgres.
• The encoding/gob package has been rewritten to eliminate the use of unsafe operations, allowing it to be used in environments that do not permit use of the unsafe package. For typical uses it will be 10-30% slower, but the delta is dependent on the type of the data and in some cases, especially involving arrays, it can be faster. There is no functional change.
• The encoding/xml package's Decoder can now report its input offset.
• In the fmt package, formatting of pointers to maps has changed to be consistent with that of pointers to structs, arrays, and so on. For instance, &map[string]int{"one": 1} now prints by default as &map[one: 1] rather than as a hexadecimal pointer value.
• The image package's Image implementations like RGBA and Gray have specialized RGBAAt and GrayAt methods alongside the general At method.
• The image/png package now has an Encoder type to control the compression level used for encoding.
• The math package now has a Nextafter32 function.
• The net/http package's Request type has a new BasicAuth method that returns the username and password from authenticated requests using the HTTP Basic Authentication Scheme.
• The net/http package's Transport type has a new DialTLS hook that allows customizing the behavior of outbound TLS connections.
• The net/http/httputil package's ReverseProxy type has a new field, ErrorLog, that provides user control of logging.
• The os package now implements symbolic links on the Windows operating system through the Symlink function. Other operating systems already have this functionality. There is also a new Unsetenv function.
• The reflect package's Type interface has a new method, Comparable, that reports whether the type implements general comparisons.
• Also in the reflect package, the Value interface is now three instead of four words because of changes to the implementation of interfaces in the runtime. This saves memory but has no semantic effect.
• The runtime package now implements monotonic clocks on Windows, as it already did for the other systems.
• The runtime package's Mallocs counter now counts very small allocations that were missed in Go 1.3. This may break tests using ReadMemStats or AllocsPerRun due to the more accurate answer.
• In the runtime package, an array PauseEnd has been added to the MemStats and GCStats structs. This array is a circular buffer of times when garbage collection pauses ended. The corresponding pause durations are already recorded in PauseNs
• The runtime/race package now supports FreeBSD, which means the go command's -race flag now works on FreeBSD.
• The sync/atomic package has a new type, Value. Value provides an efficient mechanism for atomic loads and stores of values of arbitrary type.
• In the syscall package's implementation on Linux, the Setuid and Setgid have been disabled because those system calls operate on the calling thread, not the whole process, which is different from other platforms and not the expected result.
• The testing package has a new facility to provide more control over running a set of tests. If the test code contains a function

  func TestMain(m *testing.M) 
  

  that function will be called instead of running the tests directly. The M struct contains methods to access and run the tests.
• Also in the testing package, a new Coverage function reports the current test coverage fraction, enabling individual tests to report how much they are contributing to the overall coverage.
• The text/scanner package's Scanner type has a new function, IsIdentRune, allowing one to control the definition of an identifier when scanning.
• The text/template package's boolean functions eq, lt, and so on have been generalized to allow comparison of signed and unsigned integers, simplifying their use in practice. (Previously one could only compare values of the same signedness.) All negative values compare less than all unsigned values.
• The time package now uses the standard symbol for the micro prefix, the micro symbol (U+00B5 'µ'), to print microsecond durations. ParseDuration still accepts us but the package no longer prints microseconds as us. 
  Updating: Code that depends on the output format of durations but does not use ParseDuration will need to be updated.