compiler: support file-level //go:linkname directives

Modern golang.org/x/sys/unix (v0.36+) declares its linknames
detached from function declarations, e.g.:

    func syscall_syscall(...)
    //go:linkname syscall_syscall syscall.syscall

TinyGo's pragma parser only inspected function doc comments and
therefore missed these, producing link errors like:

    undefined symbol: _golang.org/x/sys/unix.syscall_syscall

Extend parsePragmas to also walk the enclosing *ast.File's
free-standing comments for //go:linkname directives matching the
function's name. Function-attached directives still take
precedence. The existing 'unsafe' import gate is preserved.

Fixes #4395, #5365

Author: neomantra

compiler/compiler.go

@@ -90,7 +90,8 @@ type compilerContext struct {
 	astComments      map[string]*ast.CommentGroup
 	embedGlobals     map[string][]*loader.EmbedFile
 	pkg              *types.Package
-	packageDir       string // directory for this package
+	loaderPkg        *loader.Package // current package being compiled (for AST access)
+	packageDir       string          // directory for this package
 	runtimePkg       *types.Package
 }
 
@@ -294,6 +295,7 @@ func CompilePackage(moduleName string, pkg *loader.Package, ssaPkg *ssa.Package,
 	c.packageDir = pkg.OriginalDir()
 	c.embedGlobals = pkg.EmbedGlobals
 	c.pkg = pkg.Pkg
+	c.loaderPkg = pkg
 	c.runtimePkg = ssaPkg.Prog.ImportedPackage("runtime").Pkg
 	c.program = ssaPkg.Prog
 

compiler/symbol.go

@@ -346,6 +346,51 @@ func (c *compilerContext) parsePragmas(info *functionInfo, f *ssa.Function) {
 		}
 	}
 
+	// Also scan file-level //go:linkname directives. These appear as
+	// free-standing comments in *ast.File.Comments (not attached to any
+	// declaration), and are used by modern golang.org/x/sys/unix and others.
+	// Function-attached directives (above) take precedence — we only add
+	// file-level ones if no doc-comment linkname was found for this function.
+	//
+	// TODO: the hasUnsafeImport gate enforced downstream (see the
+	// //go:linkname case below) is package-level. gc enforces it per
+	// file, on the file containing the directive. For file-level
+	// linknames this is more important than for function-attached ones,
+	// because the directive can live in a file separate from the
+	// function. A stricter implementation would check whether the file
+	// returned by fileForFunc imports "unsafe", not whether any file in
+	// the package does.
+	hasFunctionLinkname := false
+	for _, comment := range pragmas {
+		if strings.HasPrefix(comment.Text, "//go:linkname ") {
+			parts := strings.Fields(comment.Text)
+			if len(parts) == 3 && parts[1] == f.Name() {
+				hasFunctionLinkname = true
+				break
+			}
+		}
+	}
+	if !hasFunctionLinkname {
+		if file := c.fileForFunc(f); file != nil {
+			for _, group := range file.Comments {
+				// Skip the function's own doc comment — already handled above.
+				if decl, ok := syntax.(*ast.FuncDecl); ok && group == decl.Doc {
+					continue
+				}
+				for _, comment := range group.List {
+					if !strings.HasPrefix(comment.Text, "//go:linkname ") {
+						continue
+					}
+					parts := strings.Fields(comment.Text)
+					if len(parts) != 3 || parts[1] != f.Name() {
+						continue
+					}
+					pragmas = append(pragmas, comment)
+				}
+			}
+		}
+	}
+
 	// Parse each pragma.
 	for _, comment := range pragmas {
 		parts := strings.Fields(comment.Text)
@@ -637,6 +682,34 @@ type globalInfo struct {
 	section  string // go:section
 }
 
+// fileForFunc returns the *ast.File that contains the declaration of f, or
+// nil if it cannot be determined. File-level pragmas are only consulted for
+// functions in the package currently being compiled — functions imported from
+// other packages have their file-level pragmas processed when those packages
+// are compiled.
+func (c *compilerContext) fileForFunc(f *ssa.Function) *ast.File {
+	if c.loaderPkg == nil || f.Pkg == nil || f.Pkg.Pkg != c.loaderPkg.Pkg {
+		return nil
+	}
+	syntax := f.Syntax()
+	if f.Origin() != nil {
+		syntax = f.Origin().Syntax()
+	}
+	if syntax == nil {
+		return nil
+	}
+	pos := syntax.Pos()
+	if !pos.IsValid() {
+		return nil
+	}
+	for _, file := range c.loaderPkg.Files {
+		if file.FileStart <= pos && pos < file.FileEnd {
+			return file
+		}
+	}
+	return nil
+}
+
 // loadASTComments loads comments on globals from the AST, for use later in the
 // program. In particular, they are required for //go:extern pragmas on globals.
 func (c *compilerContext) loadASTComments(pkg *loader.Package) {

compiler/testdata/pragma.go

@@ -115,3 +115,31 @@ func doesNotEscapeParam(a *int, b []int, c chan int, d *[0]byte)
 //go:noescape
 func stillEscapes(a *int, b []int, c chan int, d *[0]byte) {
 }
+
+// Define a function in a different package using a file-level go:linkname.
+// (Same as withLinkageName1, but with the //go:linkname directive detached
+// from the function declaration — see https://github.com/tinygo-org/tinygo/issues/4395)
+func withFileLevelLinkageName1() {
+}
+
+// Import a function from a different package using a file-level go:linkname.
+// (Same as withLinkageName2, but with the //go:linkname directive detached
+// from the function declaration.)
+func withFileLevelLinkageName2()
+
+//go:linkname withFileLevelLinkageName1 somepkg.someFileLevelFunction1
+//go:linkname withFileLevelLinkageName2 somepkg.someFileLevelFunction2
+
+// File-level linkname directives can also appear between two function
+// declarations, in which case Go's AST attaches them as the doc comment
+// of the following function — even when the directive's localname refers
+// to a different function. Exercise that case: the directive below names
+// withAdjacentLinkageName, but Go will attach it to
+// sentinelAfterAdjacentLinkname's Doc. The file-level scan must find it
+// by walking comment groups regardless of which decl they're attached to.
+func withAdjacentLinkageName() {
+}
+
+//go:linkname withAdjacentLinkageName somepkg.someAdjacentFunction
+func sentinelAfterAdjacentLinkname() {
+}

compiler/testdata/pragma.ll

@@ -93,6 +93,26 @@ entry:
   ret void
 }
 
+; Function Attrs: nounwind
+define hidden void @somepkg.someFileLevelFunction1(ptr %context) unnamed_addr #2 {
+entry:
+  ret void
+}
+
+declare void @somepkg.someFileLevelFunction2(ptr) #1
+
+; Function Attrs: nounwind
+define hidden void @somepkg.someAdjacentFunction(ptr %context) unnamed_addr #2 {
+entry:
+  ret void
+}
+
+; Function Attrs: nounwind
+define hidden void @main.sentinelAfterAdjacentLinkname(ptr %context) unnamed_addr #2 {
+entry:
+  ret void
+}
+
 attributes #0 = { allockind("alloc,zeroed") allocsize(0) "alloc-family"="runtime.alloc" "target-features"="+bulk-memory,+bulk-memory-opt,+call-indirect-overlong,+mutable-globals,+nontrapping-fptoint,+sign-ext,-multivalue,-reference-types" }
 attributes #1 = { "target-features"="+bulk-memory,+bulk-memory-opt,+call-indirect-overlong,+mutable-globals,+nontrapping-fptoint,+sign-ext,-multivalue,-reference-types" }
 attributes #2 = { nounwind "target-features"="+bulk-memory,+bulk-memory-opt,+call-indirect-overlong,+mutable-globals,+nontrapping-fptoint,+sign-ext,-multivalue,-reference-types" }