[24.0 backport] daemon: daemon.prepareMountPoints(): fix panic if mount is not a volume

[thaJeztah]

• backport of daemon: daemon.prepareMountPoints(): fix panic if mount is not a volume #45902
• fixes Failed to restart daemon after 24.0.2 -> 24.0.3 upgrade #45898
• fixes dockerd segfaults when patching 24.0.2 to 24.0.3 with live-restore enabled #45907
• fixes docker.service fails to start after upgrade from 24.0.2 to 24.0.3 docker/docker-ce-packaging#919
• relates to Restore active mount counts on live-restore #45754

The daemon.lazyInitializeVolume() function only handles restoring Volumes if a Driver is specified. The Container's MountPoints field may also contain other kind of mounts (e.g., bind-mounts). Those were ignored, and don't return an error; https://github.com/moby/moby/blob/1d9c8619cded4657af1529779c5771127e8ad0e7/daemon/volumes.go#L243-L252C2

However, the prepareMountPoints() assumed each MountPoint was a volume, and logged an informational message about the volume being restored;

moby/daemon/mounts.go

Lines 18 to 25 in 1d9c861

	if err := daemon.lazyInitializeVolume(container.ID, config); err != nil {
	return err
	}
	if alive {
	log.G(context.TODO()).WithFields(logrus.Fields{
	"container": container.ID,
	"volume": config.Volume.Name(),
	}).Debug("Live-restoring volume for alive container")

This would panic if the MountPoint was not a volume;

github.com/docker/docker/daemon.(*Daemon).prepareMountPoints(0xc00054b7b8?, 0xc0007c2500)
        /root/rpmbuild/BUILD/src/engine/.gopath/src/github.com/docker/docker/daemon/mounts.go:24 +0x1c0
github.com/docker/docker/daemon.(*Daemon).restore.func5(0xc0007c2500, 0x0?)
        /root/rpmbuild/BUILD/src/engine/.gopath/src/github.com/docker/docker/daemon/daemon.go:552 +0x271
created by github.com/docker/docker/daemon.(*Daemon).restore
        /root/rpmbuild/BUILD/src/engine/.gopath/src/github.com/docker/docker/daemon/daemon.go:530 +0x8d8
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x30 pc=0x564e9be4c7c0]

This issue was introduced in 647c2a6

(cherry picked from commit a490248)

- What I did

- How I did it

- How to verify it

- Description for the changelog

- A picture of a cute animal (not mandatory but encouraged)

[thaJeztah]

@cpuguy83 @neersighted PTAL

[r4co0n]

I think it might be useful to add some test using bind-mounts that might catch if this or anything similar gets wrong again. But I am not familiar with your tests, and how/if this would be feasible, I just want to suggest it, in case this slipped through while you worked on resolving this issue.

[cpuguy83]

Thanks @r4co0n, the change does add a test for the issue.

[r4co0n]

@cpuguy83, I concur, I really don't know why I did not notice the relevant test, thank you for clarifying.