Add plumbing for importing fds in VFS2, along with non-socket, non-TTY impl.

In VFS2, imported file descriptors are stored in a kernfs-based filesystem.
Upon calling ImportFD, the host fd can be accessed in two ways:
1. a FileDescription that can be added to the FDTable, and
2. a Dentry in the host.filesystem mount, which we will want to access through
magic symlinks in /proc/[pid]/fd/.

An implementation of the kernfs.Inode interface stores a unique host fd. This
inode can be inserted into file descriptions as well as dentries.

This change also plumbs in three FileDescriptionImpls corresponding to fds for
sockets, TTYs, and other files (only the latter is implemented here).
These implementations will mostly make corresponding syscalls to the host.
Where possible, the logic is ported over from pkg/sentry/fs/host.

Updates #1672

PiperOrigin-RevId: 299417263

3 files changed, 37 insertions, 18 deletions

diff --git a/pkg/sentry/fsimpl/proc/subtasks.go b/pkg/sentry/fsimpl/proc/subtasks.go
index f3f4e49b4..611645f3f 100644
--- a/pkg/sentry/fsimpl/proc/subtasks.go
+++ b/pkg/sentry/fsimpl/proc/subtasks.go
@@ -121,8 +121,13 @@ func (i *subtasksInode) Open(rp *vfs.ResolvingPath, vfsd *vfs.Dentry, opts vfs.O
 }
 
 // Stat implements kernfs.Inode.
-func (i *subtasksInode) Stat(vsfs *vfs.Filesystem) linux.Statx {
-	stat := i.InodeAttrs.Stat(vsfs)
-	stat.Nlink += uint32(i.task.ThreadGroup().Count())
-	return stat
+func (i *subtasksInode) Stat(vsfs *vfs.Filesystem, opts vfs.StatOptions) (linux.Statx, error) {
+	stat, err := i.InodeAttrs.Stat(vsfs, opts)
+	if err != nil {
+		return linux.Statx{}, err
+	}
+	if opts.Mask&linux.STATX_NLINK != 0 {
+		stat.Nlink += uint32(i.task.ThreadGroup().Count())
+	}
+	return stat, nil
 }
diff --git a/pkg/sentry/fsimpl/proc/task.go b/pkg/sentry/fsimpl/proc/task.go
index 18e5cd6f6..c0d643f51 100644
--- a/pkg/sentry/fsimpl/proc/task.go
+++ b/pkg/sentry/fsimpl/proc/task.go
@@ -154,12 +154,21 @@ func newTaskOwnedDir(task *kernel.Task, ino uint64, perm linux.FileMode, childre
 }
 
 // Stat implements kernfs.Inode.
-func (i *taskOwnedInode) Stat(fs *vfs.Filesystem) linux.Statx {
-	stat := i.Inode.Stat(fs)
-	uid, gid := i.getOwner(linux.FileMode(stat.Mode))
-	stat.UID = uint32(uid)
-	stat.GID = uint32(gid)
-	return stat
+func (i *taskOwnedInode) Stat(fs *vfs.Filesystem, opts vfs.StatOptions) (linux.Statx, error) {
+	stat, err := i.Inode.Stat(fs, opts)
+	if err != nil {
+		return linux.Statx{}, err
+	}
+	if opts.Mask&(linux.STATX_UID|linux.STATX_GID) != 0 {
+		uid, gid := i.getOwner(linux.FileMode(stat.Mode))
+		if opts.Mask&linux.STATX_UID != 0 {
+			stat.UID = uint32(uid)
+		}
+		if opts.Mask&linux.STATX_GID != 0 {
+			stat.GID = uint32(gid)
+		}
+	}
+	return stat, nil
 }
 
 // CheckPermissions implements kernfs.Inode.
@@ -236,7 +245,7 @@ func newNamespaceSymlink(task *kernel.Task, ino uint64, ns string) *kernfs.Dentr
 // member, there is one entry containing three colon-separated fields:
 //   hierarchy-ID:controller-list:cgroup-path"
 func newCgroupData(controllers map[string]string) dynamicInode {
-	buf := bytes.Buffer{}
+	var buf bytes.Buffer
 
 	// The hierarchy ids must be positive integers (for cgroup v1), but the
 	// exact number does not matter, so long as they are unique. We can
diff --git a/pkg/sentry/fsimpl/proc/tasks.go b/pkg/sentry/fsimpl/proc/tasks.go
index 10c08fa90..b1e39c82f 100644
--- a/pkg/sentry/fsimpl/proc/tasks.go
+++ b/pkg/sentry/fsimpl/proc/tasks.go
@@ -211,17 +211,22 @@ func (i *tasksInode) Open(rp *vfs.ResolvingPath, vfsd *vfs.Dentry, opts vfs.Open
 	return fd.VFSFileDescription(), nil
 }
 
-func (i *tasksInode) Stat(vsfs *vfs.Filesystem) linux.Statx {
-	stat := i.InodeAttrs.Stat(vsfs)
+func (i *tasksInode) Stat(vsfs *vfs.Filesystem, opts vfs.StatOptions) (linux.Statx, error) {
+	stat, err := i.InodeAttrs.Stat(vsfs, opts)
+	if err != nil {
+		return linux.Statx{}, err
+	}
 
-	// Add dynamic children to link count.
-	for _, tg := range i.pidns.ThreadGroups() {
-		if leader := tg.Leader(); leader != nil {
-			stat.Nlink++
+	if opts.Mask&linux.STATX_NLINK != 0 {
+		// Add dynamic children to link count.
+		for _, tg := range i.pidns.ThreadGroups() {
+			if leader := tg.Leader(); leader != nil {
+				stat.Nlink++
+			}
 		}
 	}
 
-	return stat
+	return stat, nil
 }
 
 func cpuInfoData(k *kernel.Kernel) string {