node-feature-discovery

node-feature-discovery copied to clipboard

Avoid trying to create empty "cpu-cpuid." labels for cpuid flags that we don't have a description for.

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: marquiz

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

The pull request process is described here

Needs approval from an approver in each of these files:

• ~~OWNERS~~ [marquiz]

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

✅ Deploy Preview for kubernetes-sigs-nfd ready!

Name	Link
🔨 Latest commit	98fd07b1171e6cf68a629a507422c3a235fb6261
🔍 Latest deploy log	https://app.netlify.com/sites/kubernetes-sigs-nfd/deploys/634d62321e6c520008fa6b83
😎 Deploy Preview	https://deploy-preview-914--kubernetes-sigs-nfd.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site settings.

Not verified /hold

ping @Chandan-Abhyankar @yselkowitz

1. this should be done for all architectures' code with manually curated lists (arm64, ppc64le, and s390x)
2. let's also update the cpu flags for POWER 9 and 10, z14/15/16, and ARM v8.latest.

1. this should be done for all architectures' code with manually curated lists (arm64, ppc64le, and s390x)

   2. let's also update the cpu flags for POWER 9 and 10, z14/15/16, and ARM v8.latest.

I think the defensive fix proposed by @marquiz 'future' proofs the code against changes. I do like the referenced PR from Yaakov. It's a great catch/help. Chandan and I can try this out on a p10.

/retitle cpu: ignore unknown cpuid flags on non-x86

1. this should be done for all architectures' code with manually curated lists (arm64, ppc64le, and s390x)

Ach, good point @yselkowitz. I updated the PR to handle other non-x86 architectures.

PTAL @Chandan-Abhyankar @yselkowitz @prb112

Thank you @marquiz , i don't have time to test it on s390x right now and i will be away next week. However the changes look good to me. Since they are the same for all arches, i'm fine with it if they work on the other arches.

One question unrelated to this PR: When i reviewed the loop, where does the capacity from the string array come from?

	r := make([]string, 0, 20)
	hwcap := uint64(C.gethwcap())
	for i := uint(0); i < 64; i++ {

I mean we have 23 Flags defined for s390x and we check 64 bits. Is there a reason the array only has a cap of 20?

Thanks @jschintag, I'll then be happy with an ack from the arm and ppc guys 😊

One question unrelated to this PR: When i reviewed the loop, where does the capacity from the string array come from?

	r := make([]string, 0, 20)
	hwcap := uint64(C.gethwcap())
	for i := uint(0); i < 64; i++ {

I mean we have 23 Flags defined for s390x and we check 64 bits. Is there a reason the array only has a cap of 20?

Good question, I'm not the original author of this code. I think it's just a (harmless) leftover from some eariler development stage. The slice capacity will get increased if more than 20 flags are present. This could be fixed in a separate PR though, I think (e.g. set cap to len(flagNames_s390x) or something more meaningful).

@marquiz This works on ppc64le. Chandan has tried it out and would resolve https://github.com/kubernetes-sigs/node-feature-discovery/issues/913

Any arm64 confirmation? @zvonkok @ArangoGutierrez ?

Rebased after #919 was merged /unhold

Verified that images build and unit tests pass on arm64 and s390x

/lgtm

@yselkowitz: changing LGTM is restricted to collaborators

In response to this:

/lgtm

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.

/lgtm

ping @fmuyassarov @zvonkok

/lgtm

/lgtm