syncthing/lib/fs/filesystem.go

320 lines
8.3 KiB
Go
Raw Normal View History

// Copyright (C) 2016 The Syncthing Authors.
//
// This Source Code Form is subject to the terms of the Mozilla Public
// License, v. 2.0. If a copy of the MPL was not distributed with this file,
// You can obtain one at https://mozilla.org/MPL/2.0/.
package fs
import (
"context"
"errors"
"io"
"os"
"path/filepath"
"strings"
"time"
)
type filesystemWrapperType int32
const (
filesystemWrapperTypeNone filesystemWrapperType = iota
filesystemWrapperTypeMtime
filesystemWrapperTypeCase
filesystemWrapperTypeError
filesystemWrapperTypeWalk
filesystemWrapperTypeLog
)
// The Filesystem interface abstracts access to the file system.
type Filesystem interface {
Chmod(name string, mode FileMode) error
Lchown(name string, uid, gid int) error
Chtimes(name string, atime time.Time, mtime time.Time) error
Create(name string) (File, error)
CreateSymlink(target, name string) error
DirNames(name string) ([]string, error)
Lstat(name string) (FileInfo, error)
Mkdir(name string, perm FileMode) error
MkdirAll(name string, perm FileMode) error
Open(name string) (File, error)
OpenFile(name string, flags int, mode FileMode) (File, error)
ReadSymlink(name string) (string, error)
Remove(name string) error
RemoveAll(name string) error
Rename(oldname, newname string) error
Stat(name string) (FileInfo, error)
SymlinksSupported() bool
Walk(name string, walkFn WalkFunc) error
// If setup fails, returns non-nil error, and if afterwards a fatal (!)
// error occurs, sends that error on the channel. Afterwards this watch
// can be considered stopped.
Watch(path string, ignore Matcher, ctx context.Context, ignorePerms bool) (<-chan Event, <-chan error, error)
Hide(name string) error
Unhide(name string) error
Glob(pattern string) ([]string, error)
Roots() ([]string, error)
Usage(name string) (Usage, error)
Type() FilesystemType
URI() string
Options() []Option
lib/scanner: Fix UTF-8 normalization on ZFS (fixes #4649) It turns out that ZFS doesn't do any normalization when storing files, but does do normalization "as part of any comparison process". In practice, this seems to mean that if you LStat a normalized filename, ZFS will return the FileInfo for the un-normalized version of that filename. This meant that our test to see whether a separate file with a normalized version of the filename already exists was failing, as we were detecting the same file. The fix is to use os.SameFile, to see whether we're getting the same FileInfo from the normalized and un-normalized versions of the same filename. One complication is that ZFS also seems to apply its magic to os.Rename, meaning that we can't use it to rename an un-normalized file to its normalized filename. Instead we have to move via a temporary object. If the move to the temporary object fails, that's OK, we can skip it and move on. If the move from the temporary object fails however, I'm not sure of the best approach: the current one is to leave the temporary file name as-is, and get Syncthing to syncronize it, so at least we don't lose the file. I'm not sure if there are any implications of this however. As part of reworking normalizePath, I spotted that it appeared to be returning the wrong thing: the doc and the surrounding code expecting it to return the normalized filename, but it was returning the un-normalized one. I fixed this, but it seems suspicious that, if the previous behaviour was incorrect, noone ever ran afoul of it. Maybe all filesystems will do some searching and give you a normalized filename if you request an unnormalized one. As part of this, I found that TestNormalization was broken: it was passing, when in fact one of the files it should have verified was present was missing. Maybe this was related to the above issue with normalizePath's return value, I'm not sure. Fixed en route. Kindly tested by @khinsen on the forum, and it appears to work. GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/4646
2018-01-05 18:11:09 +00:00
SameFile(fi1, fi2 FileInfo) bool
// Used for unwrapping things
underlying() (Filesystem, bool)
wrapperType() filesystemWrapperType
}
// The File interface abstracts access to a regular file, being a somewhat
// smaller interface than os.File
type File interface {
io.Closer
io.Reader
io.ReaderAt
io.Seeker
io.Writer
io.WriterAt
Name() string
Truncate(size int64) error
Stat() (FileInfo, error)
Sync() error
}
// The FileInfo interface is almost the same as os.FileInfo, but with the
// Sys method removed (as we don't want to expose whatever is underlying)
// and with a couple of convenience methods added.
type FileInfo interface {
// Standard things present in os.FileInfo
Name() string
Mode() FileMode
Size() int64
ModTime() time.Time
IsDir() bool
// Extensions
IsRegular() bool
IsSymlink() bool
Owner() int
Group() int
}
// FileMode is similar to os.FileMode
type FileMode uint32
func (fm FileMode) String() string {
return os.FileMode(fm).String()
}
// Usage represents filesystem space usage
type Usage struct {
Free uint64
Total uint64
}
type Matcher interface {
ShouldIgnore(name string) bool
SkipIgnoredDirs() bool
}
type MatchResult interface {
IsIgnored() bool
}
type Event struct {
Name string
Type EventType
}
type EventType int
const (
NonRemove EventType = 1 + iota
Remove
Mixed // Should probably not be necessary to be used in filesystem interface implementation
)
// Merge returns Mixed, except if evType and other are the same and not Mixed.
func (evType EventType) Merge(other EventType) EventType {
return evType | other
}
func (evType EventType) String() string {
switch {
case evType == NonRemove:
return "non-remove"
case evType == Remove:
return "remove"
case evType == Mixed:
return "mixed"
default:
panic("bug: Unknown event type")
}
}
var ErrWatchNotSupported = errors.New("watching is not supported")
// Equivalents from os package.
const ModePerm = FileMode(os.ModePerm)
const ModeSetgid = FileMode(os.ModeSetgid)
const ModeSetuid = FileMode(os.ModeSetuid)
const ModeSticky = FileMode(os.ModeSticky)
const ModeSymlink = FileMode(os.ModeSymlink)
const ModeType = FileMode(os.ModeType)
const PathSeparator = os.PathSeparator
const OptAppend = os.O_APPEND
const OptCreate = os.O_CREATE
const OptExclusive = os.O_EXCL
const OptReadOnly = os.O_RDONLY
const OptReadWrite = os.O_RDWR
const OptSync = os.O_SYNC
const OptTruncate = os.O_TRUNC
const OptWriteOnly = os.O_WRONLY
// SkipDir is used as a return value from WalkFuncs to indicate that
// the directory named in the call is to be skipped. It is not returned
// as an error by any function.
var SkipDir = filepath.SkipDir
// IsExist is the equivalent of os.IsExist
var IsExist = os.IsExist
// IsExist is the equivalent of os.ErrExist
var ErrExist = os.ErrExist
// IsNotExist is the equivalent of os.IsNotExist
var IsNotExist = os.IsNotExist
// ErrNotExist is the equivalent of os.ErrNotExist
var ErrNotExist = os.ErrNotExist
// IsPermission is the equivalent of os.IsPermission
var IsPermission = os.IsPermission
// IsPathSeparator is the equivalent of os.IsPathSeparator
var IsPathSeparator = os.IsPathSeparator
// Option modifies a filesystem at creation. An option might be specific
// to a filesystem-type.
//
// String is used to detect options with the same effect, i.e. must be different
// for options with different effects. Meaning if an option has parameters, a
// representation of those must be part of the returned string.
type Option interface {
String() string
apply(Filesystem)
}
func NewFilesystem(fsType FilesystemType, uri string, opts ...Option) Filesystem {
var fs Filesystem
switch fsType {
case FilesystemTypeBasic:
fs = newBasicFilesystem(uri, opts...)
case FilesystemTypeFake:
fs = newFakeFilesystem(uri, opts...)
default:
l.Debugln("Unknown filesystem", fsType, uri)
fs = &errorFilesystem{
fsType: fsType,
uri: uri,
err: errors.New("filesystem with type " + fsType.String() + " does not exist."),
}
}
if l.ShouldDebug("walkfs") {
return NewWalkFilesystem(&logFilesystem{fs})
}
if l.ShouldDebug("fs") {
return &logFilesystem{NewWalkFilesystem(fs)}
}
return NewWalkFilesystem(fs)
}
// IsInternal returns true if the file, as a path relative to the folder
// root, represents an internal file that should always be ignored. The file
// path must be clean (i.e., in canonical shortest form).
func IsInternal(file string) bool {
// fs cannot import config, so we hard code .stfolder here (config.DefaultMarkerName)
internals := []string{".stfolder", ".stignore", ".stversions"}
for _, internal := range internals {
if file == internal {
return true
}
if IsParent(file, internal) {
return true
}
}
return false
}
var (
errPathInvalid = errors.New("path is invalid")
errPathTraversingUpwards = errors.New("relative path traversing upwards (starting with ..)")
)
// Canonicalize checks that the file path is valid and returns it in the "canonical" form:
// - /foo/bar -> foo/bar
// - / -> "."
func Canonicalize(file string) (string, error) {
const pathSep = string(PathSeparator)
if strings.HasPrefix(file, pathSep+pathSep) {
// The relative path may pretend to be an absolute path within
// the root, but the double path separator on Windows implies
// something else and is out of spec.
return "", errPathInvalid
}
// The relative path should be clean from internal dotdots and similar
// funkyness.
file = filepath.Clean(file)
// It is not acceptable to attempt to traverse upwards.
if file == ".." {
return "", errPathTraversingUpwards
}
if strings.HasPrefix(file, ".."+pathSep) {
return "", errPathTraversingUpwards
}
if strings.HasPrefix(file, pathSep) {
if file == pathSep {
return ".", nil
}
return file[1:], nil
}
return file, nil
}
lib/fs: More efficient casefs cache (#6974) This changes the cache to cache less things, yet retain the required efficiency for our walk usecase. This uses less memory. Specifically, instead of keeping result and child caches for each path level, only keep a single cached child. In practice our operations are depth-first, or almost depth-first, and then we retain the same hit ratio for a smaller cache size. I improved the benchmark so that it counts the Lstat and DirNames operations performed, and they do not change significantly. The amount of allocated memory is reduced by 20% and the walk itself is actually slightly faster. This also removes the clear based on number of cached names (as that is not a thing any more) and the timer based clear (which was unused). This means we'll retain the last cache state forever until it's cleared by a write operation, but we did that before too and that state is now a lot smaller... The overhead compared to not using a casefs, for our typical "double walk" (walk the tree then stat everything again) is 2x the dirnames we would otherwise call, and no overhead on the stats (unchanged from old implementation) ``` name old time/op new time/op delta WalkCaseFakeFS100k/rawfs-8 306ms ± 1% 305ms ± 2% ~ (p=0.182 n=9+10) WalkCaseFakeFS100k/casefs-8 579ms ± 5% 557ms ± 1% -3.77% (p=0.000 n=10+10) name old B/entry new B/entry delta WalkCaseFakeFS100k/rawfs-8 590 ± 0% 590 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 1.09k ± 0% 0.87k ± 0% -19.98% (p=0.000 n=10+10) name old DirNames/entry new DirNames/entry delta WalkCaseFakeFS100k/rawfs-8 0.51 ± 0% 0.51 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 1.02 ± 0% 1.02 ± 0% ~ (all equal) name old DirNames/op new DirNames/op delta WalkCaseFakeFS100k/rawfs-8 51.2k ± 0% 51.2k ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 102k ± 0% 102k ± 0% ~ (all equal) name old Lstat/entry new Lstat/entry delta WalkCaseFakeFS100k/rawfs-8 3.02 ± 0% 3.02 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 3.02 ± 0% 3.02 ± 0% ~ (all equal) name old Lstat/op new Lstat/op delta WalkCaseFakeFS100k/rawfs-8 302k ± 0% 302k ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 302k ± 0% 302k ± 0% ~ (all equal) name old allocs/entry new allocs/entry delta WalkCaseFakeFS100k/rawfs-8 15.7 ± 0% 15.7 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 27.5 ± 0% 26.1 ± 0% -5.09% (p=0.000 n=10+10) name old ns/entry new ns/entry delta WalkCaseFakeFS100k/rawfs-8 2.02k ± 1% 2.02k ± 2% ~ (p=0.163 n=9+10) WalkCaseFakeFS100k/casefs-8 3.83k ± 5% 3.68k ± 1% -3.77% (p=0.000 n=10+10) name old alloc/op new alloc/op delta WalkCaseFakeFS100k/rawfs-8 89.2MB ± 0% 89.2MB ± 0% ~ (p=0.364 n=9+10) WalkCaseFakeFS100k/casefs-8 164MB ± 0% 131MB ± 0% -19.97% (p=0.000 n=10+10) name old allocs/op new allocs/op delta WalkCaseFakeFS100k/rawfs-8 2.38M ± 0% 2.38M ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 4.16M ± 0% 3.95M ± 0% -5.05% (p=0.000 n=10+10) ```
2020-09-09 12:38:39 +00:00
// wrapFilesystem should always be used when wrapping a Filesystem.
// It ensures proper wrapping order, which right now means:
// `logFilesystem` needs to be the outermost wrapper for caller lookup.
func wrapFilesystem(fs Filesystem, wrapFn func(Filesystem) Filesystem) Filesystem {
logFs, ok := fs.(*logFilesystem)
if ok {
fs = logFs.Filesystem
}
fs = wrapFn(fs)
if ok {
fs = &logFilesystem{fs}
}
return fs
}
// unwrapFilesystem removes "wrapping" filesystems to expose the filesystem of the requested wrapperType, if it exists.
func unwrapFilesystem(fs Filesystem, wrapperType filesystemWrapperType) (Filesystem, bool) {
var ok bool
lib/fs: More efficient casefs cache (#6974) This changes the cache to cache less things, yet retain the required efficiency for our walk usecase. This uses less memory. Specifically, instead of keeping result and child caches for each path level, only keep a single cached child. In practice our operations are depth-first, or almost depth-first, and then we retain the same hit ratio for a smaller cache size. I improved the benchmark so that it counts the Lstat and DirNames operations performed, and they do not change significantly. The amount of allocated memory is reduced by 20% and the walk itself is actually slightly faster. This also removes the clear based on number of cached names (as that is not a thing any more) and the timer based clear (which was unused). This means we'll retain the last cache state forever until it's cleared by a write operation, but we did that before too and that state is now a lot smaller... The overhead compared to not using a casefs, for our typical "double walk" (walk the tree then stat everything again) is 2x the dirnames we would otherwise call, and no overhead on the stats (unchanged from old implementation) ``` name old time/op new time/op delta WalkCaseFakeFS100k/rawfs-8 306ms ± 1% 305ms ± 2% ~ (p=0.182 n=9+10) WalkCaseFakeFS100k/casefs-8 579ms ± 5% 557ms ± 1% -3.77% (p=0.000 n=10+10) name old B/entry new B/entry delta WalkCaseFakeFS100k/rawfs-8 590 ± 0% 590 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 1.09k ± 0% 0.87k ± 0% -19.98% (p=0.000 n=10+10) name old DirNames/entry new DirNames/entry delta WalkCaseFakeFS100k/rawfs-8 0.51 ± 0% 0.51 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 1.02 ± 0% 1.02 ± 0% ~ (all equal) name old DirNames/op new DirNames/op delta WalkCaseFakeFS100k/rawfs-8 51.2k ± 0% 51.2k ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 102k ± 0% 102k ± 0% ~ (all equal) name old Lstat/entry new Lstat/entry delta WalkCaseFakeFS100k/rawfs-8 3.02 ± 0% 3.02 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 3.02 ± 0% 3.02 ± 0% ~ (all equal) name old Lstat/op new Lstat/op delta WalkCaseFakeFS100k/rawfs-8 302k ± 0% 302k ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 302k ± 0% 302k ± 0% ~ (all equal) name old allocs/entry new allocs/entry delta WalkCaseFakeFS100k/rawfs-8 15.7 ± 0% 15.7 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 27.5 ± 0% 26.1 ± 0% -5.09% (p=0.000 n=10+10) name old ns/entry new ns/entry delta WalkCaseFakeFS100k/rawfs-8 2.02k ± 1% 2.02k ± 2% ~ (p=0.163 n=9+10) WalkCaseFakeFS100k/casefs-8 3.83k ± 5% 3.68k ± 1% -3.77% (p=0.000 n=10+10) name old alloc/op new alloc/op delta WalkCaseFakeFS100k/rawfs-8 89.2MB ± 0% 89.2MB ± 0% ~ (p=0.364 n=9+10) WalkCaseFakeFS100k/casefs-8 164MB ± 0% 131MB ± 0% -19.97% (p=0.000 n=10+10) name old allocs/op new allocs/op delta WalkCaseFakeFS100k/rawfs-8 2.38M ± 0% 2.38M ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 4.16M ± 0% 3.95M ± 0% -5.05% (p=0.000 n=10+10) ```
2020-09-09 12:38:39 +00:00
for {
if fs.wrapperType() == wrapperType {
return fs, true
}
fs, ok = fs.underlying()
if !ok {
return nil, false
lib/fs: More efficient casefs cache (#6974) This changes the cache to cache less things, yet retain the required efficiency for our walk usecase. This uses less memory. Specifically, instead of keeping result and child caches for each path level, only keep a single cached child. In practice our operations are depth-first, or almost depth-first, and then we retain the same hit ratio for a smaller cache size. I improved the benchmark so that it counts the Lstat and DirNames operations performed, and they do not change significantly. The amount of allocated memory is reduced by 20% and the walk itself is actually slightly faster. This also removes the clear based on number of cached names (as that is not a thing any more) and the timer based clear (which was unused). This means we'll retain the last cache state forever until it's cleared by a write operation, but we did that before too and that state is now a lot smaller... The overhead compared to not using a casefs, for our typical "double walk" (walk the tree then stat everything again) is 2x the dirnames we would otherwise call, and no overhead on the stats (unchanged from old implementation) ``` name old time/op new time/op delta WalkCaseFakeFS100k/rawfs-8 306ms ± 1% 305ms ± 2% ~ (p=0.182 n=9+10) WalkCaseFakeFS100k/casefs-8 579ms ± 5% 557ms ± 1% -3.77% (p=0.000 n=10+10) name old B/entry new B/entry delta WalkCaseFakeFS100k/rawfs-8 590 ± 0% 590 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 1.09k ± 0% 0.87k ± 0% -19.98% (p=0.000 n=10+10) name old DirNames/entry new DirNames/entry delta WalkCaseFakeFS100k/rawfs-8 0.51 ± 0% 0.51 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 1.02 ± 0% 1.02 ± 0% ~ (all equal) name old DirNames/op new DirNames/op delta WalkCaseFakeFS100k/rawfs-8 51.2k ± 0% 51.2k ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 102k ± 0% 102k ± 0% ~ (all equal) name old Lstat/entry new Lstat/entry delta WalkCaseFakeFS100k/rawfs-8 3.02 ± 0% 3.02 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 3.02 ± 0% 3.02 ± 0% ~ (all equal) name old Lstat/op new Lstat/op delta WalkCaseFakeFS100k/rawfs-8 302k ± 0% 302k ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 302k ± 0% 302k ± 0% ~ (all equal) name old allocs/entry new allocs/entry delta WalkCaseFakeFS100k/rawfs-8 15.7 ± 0% 15.7 ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 27.5 ± 0% 26.1 ± 0% -5.09% (p=0.000 n=10+10) name old ns/entry new ns/entry delta WalkCaseFakeFS100k/rawfs-8 2.02k ± 1% 2.02k ± 2% ~ (p=0.163 n=9+10) WalkCaseFakeFS100k/casefs-8 3.83k ± 5% 3.68k ± 1% -3.77% (p=0.000 n=10+10) name old alloc/op new alloc/op delta WalkCaseFakeFS100k/rawfs-8 89.2MB ± 0% 89.2MB ± 0% ~ (p=0.364 n=9+10) WalkCaseFakeFS100k/casefs-8 164MB ± 0% 131MB ± 0% -19.97% (p=0.000 n=10+10) name old allocs/op new allocs/op delta WalkCaseFakeFS100k/rawfs-8 2.38M ± 0% 2.38M ± 0% ~ (all equal) WalkCaseFakeFS100k/casefs-8 4.16M ± 0% 3.95M ± 0% -5.05% (p=0.000 n=10+10) ```
2020-09-09 12:38:39 +00:00
}
}
}