Remove filesystem structure from vfs.Dentry.

This change:

- Drastically simplifies the synchronization model: filesystem structure is
  both implementation-defined and implementation-synchronized.

- Allows implementations of vfs.DentryImpl to use implementation-specific
  dentry types, reducing casts during path traversal.

- Doesn't require dentries representing non-directory files to waste space on a
  map of children.

- Allows dentry revalidation and mount lookup to be correctly ordered (fixed
  FIXME in fsimpl/gofer/filesystem.go).

- Removes the need to have two separate maps in gofer.dentry
  (dentry.vfsd.children and dentry.negativeChildren) for positive and negative
  lookups respectively.

//pkg/sentry/fsimpl/tmpfs/benchmark_test.go:
name                        old time/op  new time/op  delta
VFS2TmpfsStat/1-112          172ns ± 4%   165ns ± 3%   -4.08%  (p=0.002 n=9+9)
VFS2TmpfsStat/2-112          199ns ± 3%   195ns ±10%     ~     (p=0.132 n=8+9)
VFS2TmpfsStat/3-112          230ns ± 2%   216ns ± 2%   -6.15%  (p=0.000 n=8+8)
VFS2TmpfsStat/8-112          390ns ± 2%   358ns ± 4%   -8.33%  (p=0.000 n=9+8)
VFS2TmpfsStat/64-112        2.20µs ± 3%  2.01µs ± 3%   -8.48%  (p=0.000 n=10+8)
VFS2TmpfsStat/100-112       3.42µs ± 9%  3.08µs ± 2%   -9.82%  (p=0.000 n=9+8)
VFS2TmpfsMountStat/1-112     278ns ± 1%   286ns ±15%     ~     (p=0.712 n=8+10)
VFS2TmpfsMountStat/2-112     311ns ± 4%   298ns ± 2%   -4.27%  (p=0.000 n=9+8)
VFS2TmpfsMountStat/3-112     339ns ± 3%   330ns ± 9%     ~     (p=0.070 n=8+9)
VFS2TmpfsMountStat/8-112     503ns ± 3%   466ns ± 3%   -7.38%  (p=0.000 n=8+8)
VFS2TmpfsMountStat/64-112   2.53µs ±16%  2.17µs ± 7%  -14.19%  (p=0.000 n=10+9)
VFS2TmpfsMountStat/100-112  3.60µs ± 4%  3.30µs ± 8%   -8.33%  (p=0.001 n=8+9)

Updates #1035

PiperOrigin-RevId: 307655892

9 files changed, 185 insertions, 304 deletions

diff --git a/pkg/sentry/vfs/dentry.go b/pkg/sentry/vfs/dentry.go
index 35b208721..8624dbd5d 100644
--- a/pkg/sentry/vfs/dentry.go
+++ b/pkg/sentry/vfs/dentry.go
@@ -15,34 +15,17 @@
 package vfs
 
 import (
-	"fmt"
 	"sync/atomic"
 
 	"gvisor.dev/gvisor/pkg/sync"
 	"gvisor.dev/gvisor/pkg/syserror"
 )
 
-// Dentry represents a node in a Filesystem tree which may represent a file.
+// Dentry represents a node in a Filesystem tree at which a file exists.
 //
 // Dentries are reference-counted. Unless otherwise specified, all Dentry
 // methods require that a reference is held.
 //
-// A Dentry transitions through up to 3 different states through its lifetime:
-//
-// - Dentries are initially "independent". Independent Dentries have no parent,
-// and consequently no name.
-//
-// - Dentry.InsertChild() causes an independent Dentry to become a "child" of
-// another Dentry. A child node has a parent node, and a name in that parent,
-// both of which are mutable by DentryMoveChild(). Each child Dentry's name is
-// unique within its parent.
-//
-// - Dentry.RemoveChild() causes a child Dentry to become "disowned". A
-// disowned Dentry can still refer to its former parent and its former name in
-// said parent, but the disowned Dentry is no longer reachable from its parent,
-// and a new Dentry with the same name may become a child of the parent. (This
-// is analogous to a struct dentry being "unhashed" in Linux.)
-//
 // Dentry is loosely analogous to Linux's struct dentry, but:
 //
 // - VFS does not associate Dentries with inodes. gVisor interacts primarily
@@ -57,9 +40,6 @@ import (
 // and/or FileDescriptionImpl methods in gVisor's VFS. Filesystems that do
 // support inodes may store appropriate state in implementations of DentryImpl.
 //
-// - VFS does not provide synchronization for mutable Dentry fields, other than
-// mount-related ones.
-//
 // - VFS does not require that Dentries are instantiated for all paths accessed
 // through VFS, only those that are tracked beyond the scope of a single
 // Filesystem operation. This includes file descriptions, mount points, mount
@@ -67,6 +47,10 @@ import (
 // of Dentries for operations on mutable remote filesystems that can't actually
 // cache any state in the Dentry.
 //
+// - VFS does not track filesystem structure (i.e. relationships between
+// Dentries), since both the relevant state and synchronization are
+// filesystem-specific.
+//
 // - For the reasons above, VFS is not directly responsible for managing Dentry
 // lifetime. Dentry reference counts only indicate the extent to which VFS
 // requires Dentries to exist; Filesystems may elect to cache or discard
@@ -74,36 +58,23 @@ import (
 //
 // +stateify savable
 type Dentry struct {
-	// parent is this Dentry's parent in this Filesystem. If this Dentry is
-	// independent, parent is nil.
-	parent *Dentry
-
-	// name is this Dentry's name in parent.
-	name string
+	// mu synchronizes deletion/invalidation and mounting over this Dentry.
+	mu sync.Mutex `state:"nosave"`
 
-	flags uint32
+	// dead is true if the file represented by this Dentry has been deleted (by
+	// CommitDeleteDentry or CommitRenameReplaceDentry) or invalidated (by
+	// InvalidateDentry). dead is protected by mu.
+	dead bool
 
 	// mounts is the number of Mounts for which this Dentry is Mount.point.
 	// mounts is accessed using atomic memory operations.
 	mounts uint32
 
-	// children are child Dentries.
-	children map[string]*Dentry
-
-	// mu synchronizes disowning and mounting over this Dentry.
-	mu sync.Mutex `state:"nosave"`
-
 	// impl is the DentryImpl associated with this Dentry. impl is immutable.
 	// This should be the last field in Dentry.
 	impl DentryImpl
 }
 
-const (
-	// dflagsDisownedMask is set in Dentry.flags if the Dentry has been
-	// disowned.
-	dflagsDisownedMask = 1 << iota
-)
-
 // Init must be called before first use of d.
 func (d *Dentry) Init(impl DentryImpl) {
 	d.impl = impl
@@ -134,20 +105,6 @@ type DentryImpl interface {
 	DecRef()
 }
 
-// IsDisowned returns true if d is disowned.
-func (d *Dentry) IsDisowned() bool {
-	return atomic.LoadUint32(&d.flags)&dflagsDisownedMask != 0
-}
-
-// Preconditions: !d.IsDisowned().
-func (d *Dentry) setDisowned() {
-	atomic.AddUint32(&d.flags, dflagsDisownedMask)
-}
-
-func (d *Dentry) isMounted() bool {
-	return atomic.LoadUint32(&d.mounts) != 0
-}
-
 // IncRef increments d's reference count.
 func (d *Dentry) IncRef() {
 	d.impl.IncRef()
@@ -164,104 +121,26 @@ func (d *Dentry) DecRef() {
 	d.impl.DecRef()
 }
 
-// These functions are exported so that filesystem implementations can use
-// them. The vfs package, and users of VFS, should not call these functions.
-// Unless otherwise specified, these methods require that there are no
-// concurrent mutators of d.
-
-// Name returns d's name in its parent in its owning Filesystem. If d is
-// independent, Name returns an empty string.
-func (d *Dentry) Name() string {
-	return d.name
-}
-
-// Parent returns d's parent in its owning Filesystem. It does not take a
-// reference on the returned Dentry. If d is independent, Parent returns nil.
-func (d *Dentry) Parent() *Dentry {
-	return d.parent
-}
-
-// ParentOrSelf is equivalent to Parent, but returns d if d is independent.
-func (d *Dentry) ParentOrSelf() *Dentry {
-	if d.parent == nil {
-		return d
-	}
-	return d.parent
-}
-
-// Child returns d's child with the given name in its owning Filesystem. It
-// does not take a reference on the returned Dentry. If no such child exists,
-// Child returns nil.
-func (d *Dentry) Child(name string) *Dentry {
-	return d.children[name]
-}
-
-// HasChildren returns true if d has any children.
-func (d *Dentry) HasChildren() bool {
-	return len(d.children) != 0
-}
-
-// Children returns a map containing all of d's children.
-func (d *Dentry) Children() map[string]*Dentry {
-	if !d.HasChildren() {
-		return nil
-	}
-	m := make(map[string]*Dentry)
-	for name, child := range d.children {
-		m[name] = child
-	}
-	return m
+// IsDead returns true if d has been deleted or invalidated by its owning
+// filesystem.
+func (d *Dentry) IsDead() bool {
+	d.mu.Lock()
+	defer d.mu.Unlock()
+	return d.dead
 }
 
-// InsertChild makes child a child of d with the given name.
-//
-// InsertChild is a mutator of d and child.
-//
-// Preconditions: child must be an independent Dentry. d and child must be from
-// the same Filesystem. d must not already have a child with the given name.
-func (d *Dentry) InsertChild(child *Dentry, name string) {
-	if checkInvariants {
-		if _, ok := d.children[name]; ok {
-			panic(fmt.Sprintf("parent already contains a child named %q", name))
-		}
-		if child.parent != nil || child.name != "" {
-			panic(fmt.Sprintf("child is not independent: parent = %v, name = %q", child.parent, child.name))
-		}
-	}
-	if d.children == nil {
-		d.children = make(map[string]*Dentry)
-	}
-	d.children[name] = child
-	child.parent = d
-	child.name = name
+func (d *Dentry) isMounted() bool {
+	return atomic.LoadUint32(&d.mounts) != 0
 }
 
-// IsAncestorOf returns true if d is an ancestor of d2; that is, d is either
-// d2's parent or an ancestor of d2's parent.
-func (d *Dentry) IsAncestorOf(d2 *Dentry) bool {
-	for d2.parent != nil {
-		if d2.parent == d {
-			return true
-		}
-		d2 = d2.parent
-	}
-	return false
-}
+// The following functions are exported so that filesystem implementations can
+// use them. The vfs package, and users of VFS, should not call these
+// functions.
 
 // PrepareDeleteDentry must be called before attempting to delete the file
 // represented by d. If PrepareDeleteDentry succeeds, the caller must call
 // AbortDeleteDentry or CommitDeleteDentry depending on the deletion's outcome.
-//
-// Preconditions: d is a child Dentry.
 func (vfs *VirtualFilesystem) PrepareDeleteDentry(mntns *MountNamespace, d *Dentry) error {
-	if checkInvariants {
-		if d.parent == nil {
-			panic("d is independent")
-		}
-		if d.IsDisowned() {
-			panic("d is already disowned")
-		}
-	}
 	vfs.mountMu.Lock()
 	if mntns.mountpoints[d] != 0 {
 		vfs.mountMu.Unlock()
@@ -280,42 +159,27 @@ func (vfs *VirtualFilesystem) AbortDeleteDentry(d *Dentry) {
 	d.mu.Unlock()
 }
 
-// CommitDeleteDentry must be called after the file represented by d is
-// deleted, and causes d to become disowned.
-//
-// CommitDeleteDentry is a mutator of d and d.Parent().
-//
-// Preconditions: PrepareDeleteDentry was previously called on d.
+// CommitDeleteDentry must be called after PrepareDeleteDentry if the deletion
+// succeeds.
 func (vfs *VirtualFilesystem) CommitDeleteDentry(d *Dentry) {
-	if d.parent != nil {
-		delete(d.parent.children, d.name)
-	}
-	d.setDisowned()
+	d.dead = true
 	d.mu.Unlock()
 	if d.isMounted() {
-		vfs.forgetDisownedMountpoint(d)
+		vfs.forgetDeadMountpoint(d)
 	}
 }
 
-// ForceDeleteDentry causes d to become disowned. It should only be used in
-// cases where VFS has no ability to stop the deletion (e.g. d represents the
-// local state of a file on a remote filesystem on which the file has already
-// been deleted).
-//
-// ForceDeleteDentry is a mutator of d and d.Parent().
-//
-// Preconditions: d is a child Dentry.
-func (vfs *VirtualFilesystem) ForceDeleteDentry(d *Dentry) {
-	if checkInvariants {
-		if d.parent == nil {
-			panic("d is independent")
-		}
-		if d.IsDisowned() {
-			panic("d is already disowned")
-		}
-	}
+// InvalidateDentry is called when d ceases to represent the file it formerly
+// did for reasons outside of VFS' control (e.g. d represents the local state
+// of a file on a remote filesystem on which the file has already been
+// deleted).
+func (vfs *VirtualFilesystem) InvalidateDentry(d *Dentry) {
 	d.mu.Lock()
-	vfs.CommitDeleteDentry(d)
+	d.dead = true
+	d.mu.Unlock()
+	if d.isMounted() {
+		vfs.forgetDeadMountpoint(d)
+	}
 }
 
 // PrepareRenameDentry must be called before attempting to rename the file
@@ -324,25 +188,9 @@ func (vfs *VirtualFilesystem) ForceDeleteDentry(d *Dentry) {
 // caller must call AbortRenameDentry, CommitRenameReplaceDentry, or
 // CommitRenameExchangeDentry depending on the rename's outcome.
 //
-// Preconditions: from is a child Dentry. If to is not nil, it must be a child
-// Dentry from the same Filesystem. from != to.
+// Preconditions: If to is not nil, it must be a child Dentry from the same
+// Filesystem. from != to.
 func (vfs *VirtualFilesystem) PrepareRenameDentry(mntns *MountNamespace, from, to *Dentry) error {
-	if checkInvariants {
-		if from.parent == nil {
-			panic("from is independent")
-		}
-		if from.IsDisowned() {
-			panic("from is already disowned")
-		}
-		if to != nil {
-			if to.parent == nil {
-				panic("to is independent")
-			}
-			if to.IsDisowned() {
-				panic("to is already disowned")
-			}
-		}
-	}
 	vfs.mountMu.Lock()
 	if mntns.mountpoints[from] != 0 {
 		vfs.mountMu.Unlock()
@@ -376,24 +224,14 @@ func (vfs *VirtualFilesystem) AbortRenameDentry(from, to *Dentry) {
 // is renamed without RENAME_EXCHANGE. If to is not nil, it represents the file
 // that was replaced by from.
 //
-// CommitRenameReplaceDentry is a mutator of from, to, from.Parent(), and
-// to.Parent().
-//
 // Preconditions: PrepareRenameDentry was previously called on from and to.
-// newParent.Child(newName) == to.
-func (vfs *VirtualFilesystem) CommitRenameReplaceDentry(from, newParent *Dentry, newName string, to *Dentry) {
-	if newParent.children == nil {
-		newParent.children = make(map[string]*Dentry)
-	}
-	newParent.children[newName] = from
-	from.parent = newParent
-	from.name = newName
+func (vfs *VirtualFilesystem) CommitRenameReplaceDentry(from, to *Dentry) {
 	from.mu.Unlock()
 	if to != nil {
-		to.setDisowned()
+		to.dead = true
 		to.mu.Unlock()
 		if to.isMounted() {
-			vfs.forgetDisownedMountpoint(to)
+			vfs.forgetDeadMountpoint(to)
 		}
 	}
 }
@@ -401,25 +239,18 @@ func (vfs *VirtualFilesystem) CommitRenameReplaceDentry(from, newParent *Dentry,
 // CommitRenameExchangeDentry must be called after the files represented by
 // from and to are exchanged by rename(RENAME_EXCHANGE).
 //
-// CommitRenameExchangeDentry is a mutator of from, to, from.Parent(), and
-// to.Parent().
-//
 // Preconditions: PrepareRenameDentry was previously called on from and to.
 func (vfs *VirtualFilesystem) CommitRenameExchangeDentry(from, to *Dentry) {
-	from.parent, to.parent = to.parent, from.parent
-	from.name, to.name = to.name, from.name
-	from.parent.children[from.name] = from
-	to.parent.children[to.name] = to
 	from.mu.Unlock()
 	to.mu.Unlock()
 }
 
-// forgetDisownedMountpoint is called when a mount point is deleted to umount
-// all mounts using it in all other mount namespaces.
+// forgetDeadMountpoint is called when a mount point is deleted or invalidated
+// to umount all mounts using it in all other mount namespaces.
 //
-// forgetDisownedMountpoint is analogous to Linux's
+// forgetDeadMountpoint is analogous to Linux's
 // fs/namespace.c:__detach_mounts().
-func (vfs *VirtualFilesystem) forgetDisownedMountpoint(d *Dentry) {
+func (vfs *VirtualFilesystem) forgetDeadMountpoint(d *Dentry) {
 	var (
 		vdsToDecRef    []VirtualDentry
 		mountsToDecRef []*Mount
diff --git a/pkg/sentry/vfs/file_description.go b/pkg/sentry/vfs/file_description.go
index 5976b5ccd..15cc091e2 100644
--- a/pkg/sentry/vfs/file_description.go
+++ b/pkg/sentry/vfs/file_description.go
@@ -127,7 +127,8 @@ func (fd *FileDescription) Init(impl FileDescriptionImpl, statusFlags uint32, mn
 		mount:  mnt,
 		dentry: d,
 	}
-	fd.vd.IncRef()
+	mnt.IncRef()
+	d.IncRef()
 	fd.opts = *opts
 	fd.readable = MayReadFileWithOpenFlags(statusFlags)
 	fd.writable = writable
diff --git a/pkg/sentry/vfs/filesystem.go b/pkg/sentry/vfs/filesystem.go
index a537a29d1..74577bc2f 100644
--- a/pkg/sentry/vfs/filesystem.go
+++ b/pkg/sentry/vfs/filesystem.go
@@ -346,7 +346,10 @@ type FilesystemImpl interface {
 	// ENOTEMPTY.
 	//
 	// Preconditions: !rp.Done(). For the final path component in rp,
-	// !rp.ShouldFollowSymlink(). oldName is not "." or "..".
+	// !rp.ShouldFollowSymlink(). oldParentVD.Dentry() was obtained from a
+	// previous call to
+	// oldParentVD.Mount().Filesystem().Impl().GetParentDentryAt(). oldName is
+	// not "." or "..".
 	//
 	// Postconditions: If RenameAt returns an error returned by
 	// ResolvingPath.Resolve*(), then !rp.Done().
diff --git a/pkg/sentry/vfs/filesystem_impl_util.go b/pkg/sentry/vfs/filesystem_impl_util.go
index 7315a588e..465e610e0 100644
--- a/pkg/sentry/vfs/filesystem_impl_util.go
+++ b/pkg/sentry/vfs/filesystem_impl_util.go
@@ -16,8 +16,6 @@ package vfs
 
 import (
 	"strings"
-
-	"gvisor.dev/gvisor/pkg/fspath"
 )
 
 // GenericParseMountOptions parses a comma-separated list of options of the
@@ -43,27 +41,3 @@ func GenericParseMountOptions(str string) map[string]string {
 	}
 	return m
 }
-
-// GenericPrependPath may be used by implementations of
-// FilesystemImpl.PrependPath() for which a single statically-determined lock
-// or set of locks is sufficient to ensure its preconditions (as opposed to
-// e.g. per-Dentry locks).
-//
-// Preconditions: Dentry.Name() and Dentry.Parent() must be held constant for
-// vd.Dentry() and all of its ancestors.
-func GenericPrependPath(vfsroot, vd VirtualDentry, b *fspath.Builder) error {
-	mnt, d := vd.mount, vd.dentry
-	for {
-		if mnt == vfsroot.mount && d == vfsroot.dentry {
-			return PrependPathAtVFSRootError{}
-		}
-		if d == mnt.root {
-			return nil
-		}
-		if d.parent == nil {
-			return PrependPathAtNonMountRootError{}
-		}
-		b.PrependComponent(d.name)
-		d = d.parent
-	}
-}
diff --git a/pkg/sentry/vfs/genericfstree/BUILD b/pkg/sentry/vfs/genericfstree/BUILD
new file mode 100644
index 000000000..d8fd92677
--- /dev/null
+++ b/pkg/sentry/vfs/genericfstree/BUILD
@@ -0,0 +1,16 @@
+load("//tools/go_generics:defs.bzl", "go_template")
+
+package(
+    default_visibility = ["//:sandbox"],
+    licenses = ["notice"],
+)
+
+go_template(
+    name = "generic_fstree",
+    srcs = [
+        "genericfstree.go",
+    ],
+    types = [
+        "Dentry",
+    ],
+)
diff --git a/pkg/sentry/vfs/genericfstree/genericfstree.go b/pkg/sentry/vfs/genericfstree/genericfstree.go
new file mode 100644
index 000000000..286510195
--- /dev/null
+++ b/pkg/sentry/vfs/genericfstree/genericfstree.go
@@ -0,0 +1,80 @@
+// Copyright 2020 The gVisor Authors.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+// Package genericfstree provides tools for implementing vfs.FilesystemImpls
+// where a single statically-determined lock or set of locks is sufficient to
+// ensure that a Dentry's name and parent are contextually immutable.
+//
+// Clients using this package must use the go_template_instance rule in
+// tools/go_generics/defs.bzl to create an instantiation of this template
+// package, providing types to use in place of Dentry.
+package genericfstree
+
+import (
+	"gvisor.dev/gvisor/pkg/fspath"
+	"gvisor.dev/gvisor/pkg/sentry/vfs"
+)
+
+// Dentry is a required type parameter that is a struct with the given fields.
+type Dentry struct {
+	// vfsd is the embedded vfs.Dentry corresponding to this vfs.DentryImpl.
+	vfsd vfs.Dentry
+
+	// parent is the parent of this Dentry in the filesystem's tree. If this
+	// Dentry is a filesystem root, parent is nil.
+	parent *Dentry
+
+	// name is the name of this Dentry in its parent. If this Dentry is a
+	// filesystem root, name is unspecified.
+	name string
+}
+
+// IsAncestorDentry returns true if d is an ancestor of d2; that is, d is
+// either d2's parent or an ancestor of d2's parent.
+func IsAncestorDentry(d, d2 *Dentry) bool {
+	for {
+		if d2.parent == d {
+			return true
+		}
+		if d2.parent == d2 {
+			return false
+		}
+		d2 = d2.parent
+	}
+}
+
+// ParentOrSelf returns d.parent. If d.parent is nil, ParentOrSelf returns d.
+func ParentOrSelf(d *Dentry) *Dentry {
+	if d.parent != nil {
+		return d.parent
+	}
+	return d
+}
+
+// PrependPath is a generic implementation of FilesystemImpl.PrependPath().
+func PrependPath(vfsroot vfs.VirtualDentry, mnt *vfs.Mount, d *Dentry, b *fspath.Builder) error {
+	for {
+		if mnt == vfsroot.Mount() && &d.vfsd == vfsroot.Dentry() {
+			return vfs.PrependPathAtVFSRootError{}
+		}
+		if &d.vfsd == mnt.Root() {
+			return nil
+		}
+		if d.parent == nil {
+			return vfs.PrependPathAtNonMountRootError{}
+		}
+		b.PrependComponent(d.name)
+		d = d.parent
+	}
+}
diff --git a/pkg/sentry/vfs/mount.go b/pkg/sentry/vfs/mount.go
index f06946103..02850b65c 100644
--- a/pkg/sentry/vfs/mount.go
+++ b/pkg/sentry/vfs/mount.go
@@ -188,6 +188,7 @@ func (vfs *VirtualFilesystem) MountAt(ctx context.Context, creds *auth.Credentia
 	if err != nil {
 		return err
 	}
+
 	// We can't hold vfs.mountMu while calling FilesystemImpl methods due to
 	// lock ordering.
 	vd, err := vfs.GetDentryAt(ctx, creds, target, &GetDentryOptions{})
@@ -199,7 +200,7 @@ func (vfs *VirtualFilesystem) MountAt(ctx context.Context, creds *auth.Credentia
 	vfs.mountMu.Lock()
 	vd.dentry.mu.Lock()
 	for {
-		if vd.dentry.IsDisowned() {
+		if vd.dentry.dead {
 			vd.dentry.mu.Unlock()
 			vfs.mountMu.Unlock()
 			vd.DecRef()
@@ -665,6 +666,12 @@ func (mnt *Mount) submountsLocked() []*Mount {
 	return mounts
 }
 
+// Root returns the mount's root. It does not take a reference on the returned
+// Dentry.
+func (mnt *Mount) Root() *Dentry {
+	return mnt.root
+}
+
 // Root returns mntns' root. A reference is taken on the returned
 // VirtualDentry.
 func (mntns *MountNamespace) Root() VirtualDentry {
diff --git a/pkg/sentry/vfs/pathname.go b/pkg/sentry/vfs/pathname.go
index f21a88034..cd78d66bc 100644
--- a/pkg/sentry/vfs/pathname.go
+++ b/pkg/sentry/vfs/pathname.go
@@ -58,7 +58,7 @@ loop:
 		switch err.(type) {
 		case nil:
 			if vd.mount == vfsroot.mount && vd.mount.root == vfsroot.dentry {
-				// GenericPrependPath() will have returned
+				// genericfstree.PrependPath() will have returned
 				// PrependPathAtVFSRootError in this case since it checks
 				// against vfsroot before mnt.root, but other implementations
 				// of FilesystemImpl.PrependPath() may return nil instead.
@@ -84,7 +84,7 @@ loop:
 		}
 	}
 	b.PrependByte('/')
-	if origD.IsDisowned() {
+	if origD.IsDead() {
 		b.AppendString(" (deleted)")
 	}
 	return b.String(), nil
@@ -136,7 +136,7 @@ loop:
 // PathnameForGetcwd returns an absolute pathname to vd, consistent with
 // Linux's sys_getcwd().
 func (vfs *VirtualFilesystem) PathnameForGetcwd(ctx context.Context, vfsroot, vd VirtualDentry) (string, error) {
-	if vd.dentry.IsDisowned() {
+	if vd.dentry.IsDead() {
 		return "", syserror.ENOENT
 	}
 
diff --git a/pkg/sentry/vfs/resolving_path.go b/pkg/sentry/vfs/resolving_path.go
index 8f31495da..9d047ff88 100644
--- a/pkg/sentry/vfs/resolving_path.go
+++ b/pkg/sentry/vfs/resolving_path.go
@@ -29,7 +29,9 @@ import (
 //
 // From the perspective of FilesystemImpl methods, a ResolvingPath represents a
 // starting Dentry on the associated Filesystem (on which a reference is
-// already held) and a stream of path components relative to that Dentry.
+// already held), a stream of path components relative to that Dentry, and
+// elements of the invoking Context that are commonly required by
+// FilesystemImpl methods.
 //
 // ResolvingPath is loosely analogous to Linux's struct nameidata.
 type ResolvingPath struct {
@@ -251,18 +253,17 @@ func (rp *ResolvingPath) relpathCommit() {
 	rp.origParts[rp.curPart] = rp.pit
 }
 
-// ResolveParent returns the VFS parent of d. It does not take a reference on
-// the returned Dentry.
-//
-// Preconditions: There are no concurrent mutators of d.
-//
-// Postconditions: If the returned error is nil, then the returned Dentry is
-// not nil.
-func (rp *ResolvingPath) ResolveParent(d *Dentry) (*Dentry, error) {
-	var parent *Dentry
+// CheckRoot is called before resolving the parent of the Dentry d. If the
+// Dentry is contextually a VFS root, such that path resolution should treat
+// d's parent as itself, CheckRoot returns (true, nil). If the Dentry is the
+// root of a non-root mount, such that path resolution should switch to another
+// Mount, CheckRoot returns (unspecified, non-nil error). Otherwise, path
+// resolution should resolve d's parent normally, and CheckRoot returns (false,
+// nil).
+func (rp *ResolvingPath) CheckRoot(d *Dentry) (bool, error) {
 	if d == rp.root.dentry && rp.mount == rp.root.mount {
-		// At contextual VFS root.
-		parent = d
+		// At contextual VFS root (due to e.g. chroot(2)).
+		return true, nil
 	} else if d == rp.mount.root {
 		// At mount root ...
 		vd := rp.vfs.getMountpointAt(rp.mount, rp.root)
@@ -270,59 +271,27 @@ func (rp *ResolvingPath) ResolveParent(d *Dentry) (*Dentry, error) {
 			// ... of non-root mount.
 			rp.nextMount = vd.mount
 			rp.nextStart = vd.dentry
-			return nil, resolveMountRootOrJumpError{}
+			return false, resolveMountRootOrJumpError{}
 		}
 		// ... of root mount.
-		parent = d
-	} else if d.parent == nil {
-		// At filesystem root.
-		parent = d
-	} else {
-		parent = d.parent
+		return true, nil
 	}
-	if parent.isMounted() {
-		if mnt := rp.vfs.getMountAt(rp.mount, parent); mnt != nil {
-			rp.nextMount = mnt
-			return nil, resolveMountPointError{}
-		}
-	}
-	return parent, nil
+	return false, nil
 }
 
-// ResolveChild returns the VFS child of d with the given name. It does not
-// take a reference on the returned Dentry. If no such child exists,
-// ResolveChild returns (nil, nil).
-//
-// Preconditions: There are no concurrent mutators of d.
-func (rp *ResolvingPath) ResolveChild(d *Dentry, name string) (*Dentry, error) {
-	child := d.children[name]
-	if child == nil {
-		return nil, nil
+// CheckMount is called after resolving the parent or child of another Dentry
+// to d. If d is a mount point, such that path resolution should switch to
+// another Mount, CheckMount returns a non-nil error. Otherwise, CheckMount
+// returns nil.
+func (rp *ResolvingPath) CheckMount(d *Dentry) error {
+	if !d.isMounted() {
+		return nil
 	}
-	if child.isMounted() {
-		if mnt := rp.vfs.getMountAt(rp.mount, child); mnt != nil {
-			rp.nextMount = mnt
-			return nil, resolveMountPointError{}
-		}
-	}
-	return child, nil
-}
-
-// ResolveComponent returns the Dentry reached by starting at d and resolving
-// the current path component in the stream represented by rp. It does not
-// advance the stream. It does not take a reference on the returned Dentry. If
-// no such Dentry exists, ResolveComponent returns (nil, nil).
-//
-// Preconditions: !rp.Done(). There are no concurrent mutators of d.
-func (rp *ResolvingPath) ResolveComponent(d *Dentry) (*Dentry, error) {
-	switch pc := rp.Component(); pc {
-	case ".":
-		return d, nil
-	case "..":
-		return rp.ResolveParent(d)
-	default:
-		return rp.ResolveChild(d, pc)
+	if mnt := rp.vfs.getMountAt(rp.mount, d); mnt != nil {
+		rp.nextMount = mnt
+		return resolveMountPointError{}
 	}
+	return nil
 }
 
 // ShouldFollowSymlink returns true if, supposing that the current path