disk: use xattr to check if the disk is managed by us

[huww98]

What type of PR is this?

/kind bug

What this PR does / why we need it:

to fix the issue that when DISK_AD_CONTROLLER is enabled and disk is detaching, MaxVolumesPerNode may be smaller than expected

Which issue(s) this PR fixes:

Fixes #

Special notes for your reviewer:

Does this PR introduce a user-facing change?

fixed that when DISK_AD_CONTROLLER is enabled and disk is detaching, MaxVolumesPerNode may be smaller than expected

Additional documentation e.g., KEPs (Kubernetes Enhancement Proposals), usage docs, etc.:

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: huww98
Once this PR has been reviewed and has the lgtm label, please assign mowangdk for approval. For more information see the Kubernetes Code Review Process.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[huww98]

/auto-cc

[huww98]

torvalds/linux@b09e0fa

This tmpfs xattr support is added in Linux Kernel 3.0

[huww98]

I've made some changes to the design:

• The xattr is now added at NodeUnstageVolume, because:
  • this also works for volumes attached before CSI upgraded
  • we may switch drivers for the disk for some use-case, and the xattr may be lost.
• I now iterate all the /dev to find xattr, because some disk may lack of serial number, and cannot be found by link.
• Detaching disks are now excluded from attachedDisks, because we can't know whether it's device in /dev is present. Even if it is really an unmanaged disk, it is going to be detached, so it does not matter.
• To avoid race condition, the informations are now retrieved in the order of: xattr -> OpenAPI -> Node status. This is the only valid order from my analyses.