From 4c1f201b6fc2cc30625d3c706b1f45cc68ef0223 Mon Sep 17 00:00:00 2001
From: Aleksandr Nogikh <nogikh@google.com>
Date: Wed, 11 Jan 2023 19:37:37 +0100
Subject: pkg/subsystem/linux: add the basic subsystem extraction code

---
 pkg/subsystem/linux/subsystems.go | 124 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 124 insertions(+)
 create mode 100644 pkg/subsystem/linux/subsystems.go

(limited to 'pkg/subsystem/linux/subsystems.go')

diff --git a/pkg/subsystem/linux/subsystems.go b/pkg/subsystem/linux/subsystems.go
new file mode 100644
index 000000000..b5ffa924c
--- /dev/null
+++ b/pkg/subsystem/linux/subsystems.go
@@ -0,0 +1,124 @@
+// Copyright 2023 syzkaller project authors. All rights reserved.
+// Use of this source code is governed by Apache 2 LICENSE that can be found in the LICENSE file.
+
+package linux
+
+import (
+	"fmt"
+	"io/fs"
+	"os"
+	"sort"
+
+	"github.com/google/syzkaller/pkg/subsystem/entity"
+)
+
+func ListFromRepo(repo string) ([]*entity.Subsystem, error) {
+	return listFromRepoInner(os.DirFS(repo), linuxSubsystemRules)
+}
+
+// listFromRepoInner allows for better testing.
+func listFromRepoInner(root fs.FS, rules *customRules) ([]*entity.Subsystem, error) {
+	records, err := getMaintainers(root)
+	if err != nil {
+		return nil, err
+	}
+	ctx := &linuxCtx{
+		root:       root,
+		rawRecords: records,
+		extraRules: rules,
+	}
+	ctx.groupByList()
+	list, err := ctx.getSubsystems()
+	if err != nil {
+		return nil, err
+	}
+	// Sort subsystems by name to keep output consistent.
+	sort.Slice(list, func(i, j int) bool { return list[i].Name < list[j].Name })
+	// Sort path rules to keep output consistent.
+	for _, entity := range list {
+		sort.Slice(entity.PathRules, func(i, j int) bool {
+			a, b := entity.PathRules[i], entity.PathRules[j]
+			if a.IncludeRegexp != b.IncludeRegexp {
+				return a.IncludeRegexp < b.IncludeRegexp
+			}
+			return a.ExcludeRegexp < b.ExcludeRegexp
+		})
+	}
+	return list, nil
+}
+
+type linuxCtx struct {
+	root       fs.FS
+	rawRecords []*maintainersRecord
+	extraRules *customRules
+}
+
+type subsystemCandidate struct {
+	records     []*maintainersRecord
+	commonEmail string
+}
+
+func (ctx *linuxCtx) groupByList() []*subsystemCandidate {
+	perList := make(map[string][]*maintainersRecord)
+	for _, record := range ctx.rawRecords {
+		for _, list := range record.lists {
+			perList[list] = append(perList[list], record)
+		}
+	}
+	ret := []*subsystemCandidate{}
+	for email, list := range perList {
+		ret = append(ret, &subsystemCandidate{
+			commonEmail: email,
+			records:     list,
+		})
+	}
+	return ret
+}
+
+func (ctx *linuxCtx) getSubsystems() ([]*entity.Subsystem, error) {
+	ret := []*entity.Subsystem{}
+	for _, raw := range ctx.groupByList() {
+		s := &entity.Subsystem{}
+		mergeRawRecords(s, raw.records)
+		ret = append(ret, s)
+	}
+	return ret, nil
+}
+
+func mergeRawRecords(subsystem *entity.Subsystem, records []*maintainersRecord) {
+	unique := func(list []string) []string {
+		m := make(map[string]struct{})
+		for _, s := range list {
+			m[s] = struct{}{}
+		}
+		ret := []string{}
+		for s := range m {
+			ret = append(ret, s)
+		}
+		return ret
+	}
+	var lists, maintainers []string
+	for _, record := range records {
+		rule := record.ToPathRule()
+		if !rule.IsEmpty() {
+			subsystem.PathRules = append(subsystem.PathRules, rule)
+		}
+		lists = append(lists, record.lists...)
+		maintainers = append(maintainers, record.maintainers...)
+	}
+	subsystem.Lists = unique(lists)
+	// But there's a risk that we collect too many unrelated maintainers, so
+	// let's only merge them if there are no lists.
+	if len(records) <= 1 {
+		subsystem.Maintainers = unique(maintainers)
+	}
+}
+
+func getMaintainers(root fs.FS) ([]*maintainersRecord, error) {
+	f, err := root.Open("MAINTAINERS")
+	if err != nil {
+		return nil, fmt.Errorf("failed to open the MAINTAINERS file: %w", err)
+	}
+	defer f.Close()
+	return parseLinuxMaintainers(f)
+}
-- 
cgit mrf-deployment