csi-rbdplugin crashes when decoding volume ID(CSI identifier) failed

[crazytaxii]

Describe the bug

csi-rbdplugin will crash when decoding volume ID failed due to nil pointer

Environment details

• Image/version of Ceph CSI driver : v3.6.2
• Helm chart version :
• Kernel version : 4.18.0
• Mounter used for mounting PVC (for cephFS its fuse or kernel. for rbd its
  krbd or rbd-nbd) : krbd
• Kubernetes cluster version : v1.27.2
• Ceph cluster version : 16.2.13

problematic code:

	rbdVol, err := GenVolFromVolID(ctx, volumeID, cr, req.GetSecrets())
	defer rbdVol.Destroy()
	if err != nil {
		return cs.checkErrAndUndoReserve(ctx, err, volumeID, rbdVol, cr)
	}

defer should be placed under error processing

Tasks

Preview Give feedback

No tasks being tracked yet.

[Madhu-1]

@crazytaxii Do you have any test cases or the steps to produce this panic?

[crazytaxii]

the code itself is incorrect:

	rbdVol, err := GenVolFromVolID(ctx, volumeID, cr, req.GetSecrets())
	defer rbdVol.Destroy()
	if err != nil {
		return cs.checkErrAndUndoReserve(ctx, err, volumeID, rbdVol, cr)
	}

it should be:

	rbdVol, err := GenVolFromVolID(ctx, volumeID, cr, req.GetSecrets())
	if err != nil {
		return cs.checkErrAndUndoReserve(ctx, err, volumeID, rbdVol, cr)
	}
	defer rbdVol.Destroy()