cmd/cgo: permit passing string values directly between Go and C

Permit the C preamble to use the _GoString_ type. Permit Go code to
pass string values directly to those C types. Add accessors for C
code to retrieve sizes and pointers.

Fixes #6907

Change-Id: I190c88319ec88a3ef0ddb99f342a843ba69fcaa3
Reviewed-on: https://go-review.googlesource.com/70890
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Austin Clements <austin@google.com>

misc/cgo/test/cgo_test.go

@@ -83,5 +83,7 @@ func Test20129(t *testing.T)                 { test20129(t) }
 func Test20910(t *testing.T)                 { test20910(t) }
 func Test21708(t *testing.T)                 { test21708(t) }
 func Test21809(t *testing.T)                 { test21809(t) }
+func Test6907(t *testing.T)                  { test6907(t) }
+func Test6907Go(t *testing.T)                { test6907Go(t) }
 
 func BenchmarkCgoCall(b *testing.B) { benchCgoCall(b) }

misc/cgo/test/issue6907.go

@@ -0,0 +1,33 @@
+// Copyright 2017 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.
+
+package cgotest
+
+/*
+#include <stdlib.h>
+#include <string.h>
+
+char* Issue6907CopyString(_GoString_ s) {
+	size_t n;
+	const char *p;
+	char *r;
+
+	n = _GoStringLen(s);
+	p = _GoStringPtr(s);
+	r = malloc(n + 1);
+	memmove(r, p, n);
+	r[n] = '\0';
+	return r;
+}
+*/
+import "C"
+
+import "testing"
+
+func test6907(t *testing.T) {
+	want := "yarn"
+	if got := C.GoString(C.Issue6907CopyString(want)); got != want {
+		t.Errorf("C.GoString(C.Issue6907CopyString(%q)) == %q, want %q", want, got, want)
+	}
+}

misc/cgo/test/issue6907export.go

@@ -0,0 +1,30 @@
+// Copyright 2017 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.
+
+package cgotest
+
+/*
+extern int CheckIssue6907C(_GoString_);
+*/
+import "C"
+
+import (
+	"testing"
+)
+
+const CString = "C string"
+
+//export CheckIssue6907Go
+func CheckIssue6907Go(s string) C.int {
+	if s == CString {
+		return 1
+	}
+	return 0
+}
+
+func test6907Go(t *testing.T) {
+	if got := C.CheckIssue6907C(CString); got != 1 {
+		t.Errorf("C.CheckIssue6907C() == %d, want %d", got, 1)
+	}
+}

misc/cgo/test/issue6907export_c.c

@@ -0,0 +1,11 @@
+// Copyright 2017 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.
+
+#include <string.h>
+
+#include "_cgo_export.h"
+
+int CheckIssue6907C(_GoString_ s) {
+	return CheckIssue6907Go(s);
+}

src/cmd/cgo/doc.go

@@ -134,6 +134,19 @@ struct_, union_, or enum_, as in C.struct_stat.
 The size of any C type T is available as C.sizeof_T, as in
 C.sizeof_struct_stat.
 
+A C function may be declared in the Go file with a parameter type of
+the special name _GoString_. This function may be called with an
+ordinary Go string value. The string length, and a pointer to the
+string contents, may be accessed by calling the C functions
+
+	size_t _GoStringLen(_GoString_ s);
+	const char *_GoStringPtr(_GoString_ s);
+
+These functions are only available in the preamble, not in other C
+files. The C code must not modify the contents of the pointer returned
+by _GoStringPtr. Note that the string contents may not have a trailing
+NUL byte.
+
 As Go doesn't have support for C's union type in the general case,
 C's union types are represented as a Go byte array with the same length.
 
@@ -248,6 +261,12 @@ Not all Go types can be mapped to C types in a useful way.
 Go struct types are not supported; use a C struct type.
 Go array types are not supported; use a C pointer.
 
+Go functions that take arguments of type string may be called with the
+C type _GoString_, described above. The _GoString_ type will be
+automatically defined in the preamble. Note that there is no way for C
+code to create a value of this type; this is only useful for passing
+string values from Go to C and back to Go.
+
 Using //export in a file places a restriction on the preamble:
 since it is copied into two different C output files, it must not
 contain any definitions, only declarations. If a file contains both
@@ -269,6 +288,14 @@ pointer is a Go pointer or a C pointer is a dynamic property
 determined by how the memory was allocated; it has nothing to do with
 the type of the pointer.
 
+Note that values of some Go types, other than the type's zero value,
+always include Go pointers. This is true of string, slice, interface,
+channel, map, and function types. A pointer type may hold a Go pointer
+or a C pointer. Array and struct types may or may not include Go
+pointers, depending on the element types. All the discussion below
+about Go pointers applies not just to pointer types, but also to other
+types that include Go pointers.
+
 Go code may pass a Go pointer to C provided the Go memory to which it
 points does not contain any Go pointers. The C code must preserve
 this property: it must not store any Go pointers in Go memory, even
@@ -279,14 +306,17 @@ the Go memory in question is the entire array or the entire backing
 array of the slice.
 
 C code may not keep a copy of a Go pointer after the call returns.
+This includes the _GoString_ type, which, as noted above, includes a
+Go pointer; _GoString_ values may not be retained by C code.
 
-A Go function called by C code may not return a Go pointer. A Go
-function called by C code may take C pointers as arguments, and it may
-store non-pointer or C pointer data through those pointers, but it may
-not store a Go pointer in memory pointed to by a C pointer. A Go
-function called by C code may take a Go pointer as an argument, but it
-must preserve the property that the Go memory to which it points does
-not contain any Go pointers.
+A Go function called by C code may not return a Go pointer (which
+implies that it may not return a string, slice, channel, and so
+forth). A Go function called by C code may take C pointers as
+arguments, and it may store non-pointer or C pointer data through
+those pointers, but it may not store a Go pointer in memory pointed to
+by a C pointer. A Go function called by C code may take a Go pointer
+as an argument, but it must preserve the property that the Go memory
+to which it points does not contain any Go pointers.
 
 Go code may not store a Go pointer in C memory. C code may store Go
 pointers in C memory, subject to the rule above: it must stop storing

src/cmd/cgo/gcc.go

@@ -192,8 +192,8 @@ func (p *Package) Translate(f *File) {
 // in the file f and saves relevant renamings in f.Name[name].Define.
 func (p *Package) loadDefines(f *File) {
 	var b bytes.Buffer
-	b.WriteString(f.Preamble)
 	b.WriteString(builtinProlog)
+	b.WriteString(f.Preamble)
 	stdout := p.gccDefines(b.Bytes())
 
 	for _, line := range strings.Split(stdout, "\n") {
@@ -312,8 +312,8 @@ func (p *Package) guessKinds(f *File) []*Name {
 	// whether name denotes a type or an expression.
 
 	var b bytes.Buffer
-	b.WriteString(f.Preamble)
 	b.WriteString(builtinProlog)
+	b.WriteString(f.Preamble)
 
 	for i, n := range names {
 		fmt.Fprintf(&b, "#line %d \"not-declared\"\n"+
@@ -461,8 +461,8 @@ func (p *Package) loadDWARF(f *File, names []*Name) {
 	// for each entry in names and then dereference the type we
 	// learn for __cgo__i.
 	var b bytes.Buffer
-	b.WriteString(f.Preamble)
 	b.WriteString(builtinProlog)
+	b.WriteString(f.Preamble)
 	b.WriteString("#line 1 \"cgo-dwarf-inference\"\n")
 	for i, n := range names {
 		fmt.Fprintf(&b, "__typeof__(%s) *__cgo__%d;\n", n.C, i)

src/cmd/cgo/out.go

@@ -539,6 +539,7 @@ func (p *Package) writeOutput(f *File, srcfile string) {
 
 	// While we process the vars and funcs, also write gcc output.
 	// Gcc output starts with the preamble.
+	fmt.Fprintf(fgcc, "%s\n", builtinProlog)
 	fmt.Fprintf(fgcc, "%s\n", f.Preamble)
 	fmt.Fprintf(fgcc, "%s\n", gccProlog)
 	fmt.Fprintf(fgcc, "%s\n", tsanProlog)
@@ -1145,6 +1146,7 @@ func (p *Package) writeExportHeader(fgcch io.Writer) {
 		pkg = p.PackagePath
 	}
 	fmt.Fprintf(fgcch, "/* package %s */\n\n", pkg)
+	fmt.Fprintf(fgcch, "%s\n", builtinExportProlog)
 
 	fmt.Fprintf(fgcch, "/* Start of preamble from import \"C\" comments.  */\n\n")
 	fmt.Fprintf(fgcch, "%s\n", p.Preamble)
@@ -1389,7 +1391,7 @@ const builtinProlog = `
 /* Define intgo when compiling with GCC.  */
 typedef ptrdiff_t intgo;
 
-typedef struct { char *p; intgo n; } _GoString_;
+typedef struct { const char *p; intgo n; } _GoString_;
 typedef struct { char *p; intgo n; intgo c; } _GoBytes_;
 _GoString_ GoString(char *p);
 _GoString_ GoStringN(char *p, int l);
@@ -1397,6 +1399,12 @@ _GoBytes_ GoBytes(void *p, int n);
 char *CString(_GoString_);
 void *CBytes(_GoBytes_);
 void *_CMalloc(size_t);
+
+__attribute__ ((unused))
+static size_t _GoStringLen(_GoString_ s) { return s.n; }
+
+__attribute__ ((unused))
+static const char *_GoStringPtr(_GoString_ s) { return s.p; }
 `
 
 const goProlog = `
@@ -1628,6 +1636,27 @@ void localCgoCheckResult(Eface val) {
 }
 `
 
+// builtinExportProlog is a shorter version of builtinProlog,
+// to be put into the _cgo_export.h file.
+// For historical reasons we can't use builtinProlog in _cgo_export.h,
+// because _cgo_export.h defines GoString as a struct while builtinProlog
+// defines it as a function. We don't change this to avoid unnecessarily
+// breaking existing code.
+const builtinExportProlog = `
+#line 1 "cgo-builtin-prolog"
+
+#include <stddef.h> /* for ptrdiff_t below */
+
+#ifndef GO_CGO_EXPORT_PROLOGUE_H
+#define GO_CGO_EXPORT_PROLOGUE_H
+
+typedef ptrdiff_t intgo;
+
+typedef struct { const char *p; intgo n; } _GoString_;
+
+#endif
+`
+
 func (p *Package) gccExportHeaderProlog() string {
 	return strings.Replace(gccExportHeaderProlog, "GOINTBITS", fmt.Sprint(8*p.IntSize), -1)
 }
@@ -1661,7 +1690,7 @@ typedef double _Complex GoComplex128;
 */
 typedef char _check_for_GOINTBITS_bit_pointer_matching_GoInt[sizeof(void*)==GOINTBITS/8 ? 1:-1];
 
-typedef struct { const char *p; GoInt n; } GoString;
+typedef _GoString_ GoString;
 typedef void *GoMap;
 typedef void *GoChan;
 typedef struct { void *t; void *v; } GoInterface;