*/
    data class Script(val source: ProgramSource) : Program()

    /**
     * A [Stage1] program followed by a stage 2 [Script] program.
     */
    data class Staged(val stage1: Stage1, val stage2: Script) : Program()

    /**
     * Any stage 1 program, one of [Buildscript], [Plugins] or [a sequence of the two][Stage1Sequence].
     */
    abstract class Stage1 : Program()

	ucfg.goversion = set(ucfg.GoVersion)
	ucfg.program = make(map[string]bool, len(ucfg.Programs))
	ucfg.pgversion = make(map[pgkey]bool, len(ucfg.Programs))
	ucfg.pgcounter = make(map[pgkey]bool, len(ucfg.Programs))
	ucfg.pgcounterprefix = make(map[pgkey]bool, len(ucfg.Programs))
	ucfg.pgstack = make(map[pgkey]bool, len(ucfg.Programs))
	ucfg.rate = make(map[pgkey]float64)
	for _, p := range ucfg.Programs {
		ucfg.program[p.Name] = true

delivered to the Notify channel, and the write will fail with an EPIPE
error.

This means that, by default, command line programs will behave like
typical Unix command line programs, while other programs will not
crash with SIGPIPE when writing to a closed network connection.

# Go programs that use cgo or SWIG

In a Go program that includes non-Go code, typically C/C++ code
accessed using cgo or SWIG, Go's startup code normally runs first. It

with "go get") should work with Go programs, but we cannot guarantee that.
What we test is that the "go tool pprof" shipped with each Go release works
with programs from that release.

2. Pprof is used inside Google for C++, Java, and Go programs.
Because it was developed for that broader context, it is overgeneralized
when used here for the specific use case of profiling standard Go programs.

! stdout foo

# Wait should set the output to the concatenated outputs of the background
# programs, in the order in which they were started.
wait
stdout 'foo\nbar'

# The end of the test should interrupt or kill any remaining background
# programs, but that should not cause the test to fail if it does not
# care about the exit status of those programs.
[exec:sleep] ? exec sleep 86400 &

// license that can be found in the LICENSE file.

// Pprof interprets and displays profiles of Go programs.
//
// Basic usage:
//
//	go tool pprof binary profile
//
// For detailed usage information:
//
//	go tool pprof -h
//
// For an example, see https://blog.golang.org/profiling-go-programs.

// It guarantees to make a copy of s into a new allocation,
// which can be important when retaining only a small substring
// of a much larger string. Using Clone can help such programs
// use less memory. Of course, since using Clone makes a copy,
// overuse of Clone can make programs use more memory.
// Clone should typically be used only rarely, and only when
// profiling indicates that it is needed.

overhead of the added check. A new check that finds only a handful of problems
across all existing programs, even if the problem is significant, is not worth
adding to the suite everyone runs daily.

Precision:

Most of vet's checks are heuristic and can generate both false positives (flagging
correct programs) and false negatives (not flagging incorrect ones). The rate of

variable.
Test names are defined as the name of the `.go` file that generates the
trace, but with the `.go` extension removed.

## Trace test programs

The trace test programs in the `testprog` directory generate traces to
stdout.
Otherwise they're just normal programs.

## Trace debug commands

The `cmd` directory contains helpful tools for debugging traces.

* `gotraceraw` parses traces without validation.

## Notable Packages

- The [x/telemetry/counter](https://pkg.go.dev/golang.org/x/telemetry/counter)
  package provides a library for instrumenting programs with counters and stack
  reports.
- The [x/telemetry/upload](https://pkg.go.dev/golang.org/x/telemetry/upload)
  package provides a hook for Go toolchain programs to upload telemetry data,
  if the user has opted in to telemetry uploading.