fsck: handle sparse files efficiently, fix xfstests generic/285

If we encounter a 128KB block of zeros, try to skip to the next
data section by calling File.SeekData().

This fixes xfstests generic/285, which creates a 17TB sparse file,
and runs fsck afterwards. Without this optimization, fsck would
take ages.

fsck.go

@@ -1,6 +1,7 @@
 package main
 
 import (
+	"bytes"
 	"fmt"
 	"os"
 	"path/filepath"
@@ -115,6 +116,7 @@ func (ck *fsckObj) file(path string) {
 		return
 	}
 	defer f.Release()
+	allZero := make([]byte, fuse.MAX_KERNEL_WRITE)
 	buf := make([]byte, fuse.MAX_KERNEL_WRITE)
 	var off int64
 	// Read() through the whole file and catch transparently mitigated corruptions
@@ -132,6 +134,15 @@ func (ck *fsckObj) file(path string) {
 			return
 		}
 		off += int64(result.Size())
+		// If we seem to be in the middle of a file hole, try to skip to the next
+		// data section.
+		if bytes.Equal(buf, allZero) {
+			f2 := f.(*fusefrontend.File)
+			nextOff, err := f2.SeekData(off)
+			if err == nil {
+				off = nextOff
+			}
+		}
 	}
 }
 

tests/fsck/fsck_test.go

@@ -4,8 +4,10 @@ import (
 	"encoding/base64"
 	"os"
 	"os/exec"
+	"runtime"
 	"strings"
 	"testing"
+	"time"
 
 	"github.com/pkg/xattr"
 
@@ -84,3 +86,41 @@ func TestExampleFses(t *testing.T) {
 		}
 	}
 }
+
+// TestTerabyteFile verifies that fsck does something intelligent when it hits
+// a 1-terabyte sparse file (trying to read the whole file is not intelligent).
+func TestTerabyteFile(t *testing.T) {
+	if runtime.GOOS != "linux" {
+		t.Skipf("Only linux supports SEEK_DATA")
+	}
+	cDir := test_helpers.InitFS(t)
+	pDir := cDir + ".mnt"
+	test_helpers.MountOrFatal(t, cDir, pDir, "-extpass", "echo test")
+	defer test_helpers.UnmountErr(pDir)
+	exabyteFile := pDir + "/exabyteFile"
+	fd, err := os.Create(exabyteFile)
+	if err != nil {
+		t.Fatal(err)
+	}
+	defer fd.Close()
+	var oneTiB int64 = 1024 * 1024 * 1024 * 1024
+	_, err = fd.WriteAt([]byte("foobar"), oneTiB)
+	if err != nil {
+		t.Fatal(err)
+	}
+	fi, err := fd.Stat()
+	if err != nil {
+		t.Fatal(err)
+	}
+	t.Logf("size=%d, running fsck", fi.Size())
+	cmd := exec.Command(test_helpers.GocryptfsBinary, "-fsck", "-extpass", "echo test", cDir)
+	cmd.Stderr = os.Stderr
+	cmd.Stdout = os.Stdout
+	cmd.Start()
+	timer := time.AfterFunc(10*time.Second, func() {
+		cmd.Process.Kill()
+		t.Fatalf("timeout")
+	})
+	cmd.Wait()
+	timer.Stop()
+}