2
2
mirror of https://github.com/octoleo/restic.git synced 2024-12-13 14:43:12 +00:00
restic/internal/fuse/file.go
Michael Eischer 51173c5003 fuse: forget fs.Node instances on request by the kernel
Forget fs.Node instances once the kernel frees the corresponding nodeId.
This ensures that restic does not run out of memory on large snapshots.
2024-09-14 18:11:44 +02:00

182 lines
4.5 KiB
Go

//go:build darwin || freebsd || linux
// +build darwin freebsd linux
package fuse
import (
"context"
"sort"
"github.com/restic/restic/internal/debug"
"github.com/restic/restic/internal/errors"
"github.com/restic/restic/internal/restic"
"github.com/anacrolix/fuse"
"github.com/anacrolix/fuse/fs"
)
// The default block size to report in stat
const blockSize = 512
// Statically ensure that *file and *openFile implement the given interfaces
var _ = fs.HandleReader(&openFile{})
var _ = fs.NodeForgetter(&file{})
var _ = fs.NodeGetxattrer(&file{})
var _ = fs.NodeListxattrer(&file{})
var _ = fs.NodeOpener(&file{})
type file struct {
root *Root
forget forgetFn
node *restic.Node
inode uint64
}
type openFile struct {
file
// cumsize[i] holds the cumulative size of blobs[:i].
cumsize []uint64
}
func newFile(root *Root, forget forgetFn, inode uint64, node *restic.Node) (fusefile *file, err error) {
debug.Log("create new file for %v with %d blobs", node.Name, len(node.Content))
return &file{
inode: inode,
forget: forget,
root: root,
node: node,
}, nil
}
func (f *file) Attr(_ context.Context, a *fuse.Attr) error {
debug.Log("Attr(%v)", f.node.Name)
a.Inode = f.inode
a.Mode = f.node.Mode
a.Size = f.node.Size
a.Blocks = (f.node.Size + blockSize - 1) / blockSize
a.BlockSize = blockSize
a.Nlink = uint32(f.node.Links)
if !f.root.cfg.OwnerIsRoot {
a.Uid = f.node.UID
a.Gid = f.node.GID
}
a.Atime = f.node.AccessTime
a.Ctime = f.node.ChangeTime
a.Mtime = f.node.ModTime
return nil
}
func (f *file) Open(ctx context.Context, _ *fuse.OpenRequest, _ *fuse.OpenResponse) (fs.Handle, error) {
debug.Log("open file %v with %d blobs", f.node.Name, len(f.node.Content))
var bytes uint64
cumsize := make([]uint64, 1+len(f.node.Content))
for i, id := range f.node.Content {
if ctx.Err() != nil {
return nil, ctx.Err()
}
size, found := f.root.repo.LookupBlobSize(restic.DataBlob, id)
if !found {
return nil, errors.Errorf("id %v not found in repository", id)
}
bytes += uint64(size)
cumsize[i+1] = bytes
}
var of = openFile{file: *f}
if bytes != f.node.Size {
debug.Log("sizes do not match: node.Size %v != size %v, using real size", f.node.Size, bytes)
// Make a copy of the node with correct size
nodenew := *f.node
nodenew.Size = bytes
of.file.node = &nodenew
}
of.cumsize = cumsize
return &of, nil
}
func (f *openFile) getBlobAt(ctx context.Context, i int) (blob []byte, err error) {
blob, err = f.root.blobCache.GetOrCompute(f.node.Content[i], func() ([]byte, error) {
return f.root.repo.LoadBlob(ctx, restic.DataBlob, f.node.Content[i], nil)
})
if err != nil {
debug.Log("LoadBlob(%v, %v) failed: %v", f.node.Name, f.node.Content[i], err)
return nil, unwrapCtxCanceled(err)
}
return blob, nil
}
func (f *openFile) Read(ctx context.Context, req *fuse.ReadRequest, resp *fuse.ReadResponse) error {
debug.Log("Read(%v, %v, %v), file size %v", f.node.Name, req.Size, req.Offset, f.node.Size)
offset := uint64(req.Offset)
// as stated in https://godoc.org/bazil.org/fuse/fs#HandleReader there
// is no need to check if offset > size
// handle special case: file is empty
if f.node.Size == 0 {
resp.Data = resp.Data[:0]
return nil
}
// Skip blobs before the offset
startContent := -1 + sort.Search(len(f.cumsize), func(i int) bool {
return f.cumsize[i] > offset
})
offset -= f.cumsize[startContent]
dst := resp.Data[0:req.Size]
readBytes := 0
remainingBytes := req.Size
// The documentation of bazil/fuse actually says that synchronization is
// required (see https://godoc.org/bazil.org/fuse#hdr-Service_Methods):
//
// Multiple goroutines may call service methods simultaneously;
// the methods being called are responsible for appropriate synchronization.
//
// However, no lock needed here as getBlobAt can be called concurrently
// (blobCache has its own locking)
for i := startContent; remainingBytes > 0 && i < len(f.cumsize)-1; i++ {
blob, err := f.getBlobAt(ctx, i)
if err != nil {
return err
}
if offset > 0 {
blob = blob[offset:]
offset = 0
}
copied := copy(dst, blob)
remainingBytes -= copied
readBytes += copied
dst = dst[copied:]
}
resp.Data = resp.Data[:readBytes]
return nil
}
func (f *file) Listxattr(_ context.Context, req *fuse.ListxattrRequest, resp *fuse.ListxattrResponse) error {
nodeToXattrList(f.node, req, resp)
return nil
}
func (f *file) Getxattr(_ context.Context, req *fuse.GetxattrRequest, resp *fuse.GetxattrResponse) error {
return nodeGetXattr(f.node, req, resp)
}
func (f *file) Forget() {
f.forget()
}