What happens at link time depends on whether the final binary is linked
using the internal or external mode. If other packages are compiled in
"external only" mode, then the final link will be an external one.
Otherwise the link will be an internal one.

The linking directives are used according to the kind of final link
used.

In internal mode, cmd/link itself processes all the host object files, in

<p class="rule">
The completion of a single call of <code>f()</code> from <code>once.Do(f)</code>
is synchronized before the return of any call of <code>once.Do(f)</code>.
</p>

<p>
In this program:
</p>

<pre>
var a string
var once sync.Once

func setup() {
	a = "hello, world"
}

func doprint() {
	once.Do(setup)
	print(a)
}

func twoprint() {
	go doprint()

func (s sparseEntry) endOffset() int64 { return s.Offset + s.Length }

// A sparse file can be represented as either a sparseDatas or a sparseHoles.
// As long as the total size is known, they are equivalent and one can be
// converted to the other form and back. The various tar formats with sparse
// file support represent sparse files in the sparseDatas form. That is, they

because it documents the full list of known GODEBUG settings,
which are tied to a specific release.
-->

## Introduction {#intro}

Go's emphasis on backwards compatibility is one of its key strengths.
There are, however, times when we cannot maintain complete compatibility.
If code depends on buggy (including insecure) behavior,
then fixing the bug will break that code.

		switch {
		case len(features) == 0 || (len(required) > 0 && required[0] < features[0]):
			feature := take(&required)
			if exceptionSet[feature] {
				// An "unfortunate" case: the feature was once
				// included in the API (e.g. go1.txt), but was
				// subsequently removed. These are already
				// acknowledged by being in the file
				// "api/except.txt". No need to print them out
				// here.

// arguments of [cmp.Ordered] types. There must be at least one argument.
// If T is a floating-point type and any of the arguments are NaNs,
// max will return NaN.
func max[T cmp.Ordered](x T, y ...T) T

// The min built-in function returns the smallest value of a fixed number of
// arguments of [cmp.Ordered] types. There must be at least one argument.
// If T is a floating-point type and any of the arguments are NaNs,

	// files. Internally, the tar format often uses fake "files" to add meta
	// data that describes the next file. These meta data "files" should not
	// normally be visible to the outside. As such, this loop iterates through
	// one or more "header files" until it finds a "normal file".
	format := FormatUSTAR | FormatPAX | FormatGNU
	for {
		// Discard the remainder of the file and any padding.

	}

	return w.writeDataDescriptor()
}

func (w *fileWriter) writeDataDescriptor() error {
	if !w.hasDataDescriptor() {
		return nil
	}
	// Write data descriptor. This is more complicated than one would
	// think, see e.g. comments in zipfile.c:putextended() and
	// https://bugs.openjdk.org/browse/JDK-7073588.
	// The approach here is to write 8 byte sizes if needed without