fix ascend vnpu addResource

[DSFans2014]

What type of PR is this?

What this PR does / why we need it:

Which issue(s) this PR fixes:

Fixes #5360

Special notes for your reviewer:

Does this PR introduce a user-facing change?

[gemini-code-assist]

Code Review

This pull request introduces a PodMap to AscendDevice to track device usage per pod, preventing duplicate resource additions or subtractions. It also refactors AddResourceUsage and SubResourceUsage to accept *AscendDevice directly and switches from InRequestDevices to SupportDevices. Feedback on the changes highlights an inefficiency in SubResource where a new DeviceUsage is redundantly allocated and assigned to PodMap immediately before the key is deleted.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Adds per-pod tracking for Ascend device allocations to make resource accounting idempotent and improves unit coverage around resource addition.

Changes:

• Introduces PodMap on AscendDevice to track per-pod device usage and prevent double-counting.
• Updates add/sub resource logic to use PodMap and to skip unknown devices safely.
• Adds a table-driven test suite for AscendDevices.AddResource.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 4 comments.

File	Description
pkg/scheduler/api/devices/ascend/hami/device_info.go	Adds per-pod usage tracking (PodMap), updates add/sub accounting flow, and extends deepcopy/snapshot to include the new map.
pkg/scheduler/api/devices/ascend/hami/device_info_test.go	Adds table-driven tests and helper constructors for AddResource behavior (single/multi/duplicate/invalid annotation, etc.).

Comments suppressed due to low confidence (2)

pkg/scheduler/api/devices/ascend/hami/device_info.go:154

• This changes the public method signatures from (id string, ...) to (*AscendDevice, ...), which is a breaking API change for any external callers. Consider keeping the original id-based exported methods as wrappers (lookup + call an unexported helper), or make the pointer-taking variants unexported to avoid forcing downstream changes.

func (ads *AscendDevices) AddResourceUsage(dev *AscendDevice, cores int32, mem int32) error {
	dev.DeviceUsage.Used++
	dev.DeviceUsage.Usedcores += cores
	dev.DeviceUsage.Usedmem += mem
	return nil
}

func (ads *AscendDevices) SubResourceUsage(dev *AscendDevice, cores int32, mem int32) error {
	dev.DeviceUsage.Used--
	dev.DeviceUsage.Usedcores -= cores
	dev.DeviceUsage.Usedmem -= mem
	return nil
}

pkg/scheduler/api/devices/ascend/hami/device_info.go:171

• The SubResource behavior changed materially (annotation key source + PodMap-gated subtraction). There are new tests for AddResource, but no corresponding unit tests asserting SubResource decrements exactly once and cleans up PodMap (including cases like missing device, repeated SubResource calls for same pod, and PodMap missing entry behavior). Adding targeted tests here would prevent regressions.

func (ads *AscendDevices) SubResource(pod *v1.Pod) {
	if ads == nil {
		return
	}
	ano_key := devices.SupportDevices[ads.Type]
	ano, ok := pod.Annotations[ano_key]
	if !ok {
		return
	}

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[JesseStutler]

Please help /cc @archlitchi

[archlitchi]

@DSFans2014 have you tested on a NPU environment?

[DSFans2014]

@DSFans2014 have you tested on a NPU environment?

Yes, I've already tested this fix locally.

[archlitchi]

/lgtm

[volcano-sh-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: archlitchi

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

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• pkg/scheduler/OWNERS [archlitchi]

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

[JesseStutler]

/cherrypick release-1.15

[volcano-sh-bot]

@JesseStutler: new pull request created: #5377

Details

In response to this:

/cherrypick release-1.15

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.