[noble] Missing fix for continuations with associated metering.

[Frode Nordahl]

[ Impact ]
There is a bug in Open vSwitch which is required for successful build and test run for OVN 24.03.2.
(Bug 2066908: [SRU] ovn 24.03.2 point release)

[ Test plan ]
- Build and test the OVN 24.03.2 package.
- Check openvswitch build log for new test (must pass/'ok')
  `ofproto-dpif - continuation with meters`

[ Regression Potential ]
Minimal, the fix has been available for some time upstream already and has been deemed safe to backport all the way to Open vSwitch 2.17 by the upstream maintainers.

Regressions would likely manifest in the areas modified by the patch:
- invoking a controller action by way of a sample action with a specified meter;
sample action is generated during xlate processing and/or (now) continuation.

[ Other Info ]
The bug has been fixed upstream, and there is a patch available for branch-3.3 [0].

0: https://github.com/openvswitch/ovs/commit/c560f6ca3257dd3b64e01a4a244b02b90216858b

[Mauricio Faria de Oliveira (mfo)]

Clarified/added '(Bug 2066908: [SRU] ovn 24.03.2 point release)' in the Impact section.

[Mauricio Faria de Oliveira (mfo)]

This is arguably a reasonable combination (although with slight differences)
of 'FTBFS' and 'Autopkgtest failures' (i.e., 'for successful build/test run')
acceptable per SRU policy (included below), thus should not need staging.

And even though this SRU (openvswitch) is directed at fixing/preventing a
build-time error in another, upcoming SRU (ovs), it does address run-time
issues in openvswitch (see patch).

This run-time issue/fix has build-time tests added in this SRU, so tests
are covered (added that to 'Test Plan' section); e.g., the patch context
lines show 'ofproto-dpif - continuation with patch port' which is tested
in the build log of current version in noble-proposed:

`1251: ofproto-dpif - continuation with patch port ok`

For reference, the patch adds the build-time test:
`ofproto-dpif - continuation with meters`

...

"""
Other safe cases [0]

FTBFS(Fails To Build From Source) can also be considered. [...]
Usually those bugs should only be SRUed in conjunction with another bug fix.

Autopkgtest failures should also normally be SRUed only in conjunction with other high-priority fixes affecting users at runtime, optionally by staging them.
As an exception, when an SRU of one package will introduce a regression in the autopkgtests of another package, it is appropriate to do an autopkgtest-only SRU of the other package.

Staging low priority uploads [1]

SRUs for bugs which do not affect users at runtime are inappropriate to force users to apply.
There is a cost to our users (and our mirror network) for downloading updates of packages, which should be balanced against the utility of the update to the user downloading it.

However, if such an update otherwise complies with SRU policy, it can be staged to be bundled with a future SRU or security update.
"""

[0] https://wiki.ubuntu.com/StableReleaseUpdates#Other_safe_cases
[1] https://wiki.ubuntu.com/StableReleaseUpdates#Staging_low_priority_uploads
[2] https://launchpadlibrarian.net/727042740/buildlog_ubuntu-noble-amd64.openvswitch_3.3.0-1ubuntu2_BUILDING.txt.gz

[Mauricio Faria de Oliveira (mfo)]

Marking devel/oracular as Fix Released (patch included):

$ pull-lp-source openvswitch
Found openvswitch 3.3.0-5 in oracular
...
$ cd openvswitch-3.3.0

$ grep 'ofproto-dpif - continuation with meters' tests/ofproto-dpif.at
AT_SETUP([ofproto-dpif - continuation with meters])

$ grep -A6 '^put_controller_user_action' ofproto/ofproto-dpif-xlate.c
put_controller_user_action(struct xlate_ctx *ctx,
                           bool dont_send, bool continuation,
                           uint32_t recirc_id, int len,
                           enum ofp_packet_in_reason reason,
                           uint32_t provider_meter_id,
                           uint16_t controller_id)
{

[Mauricio Faria de Oliveira (mfo)]

Added to the 'Regression Potential' an statement along the lines of the intended meaning 'Where problems could occur' [3], which relates to where/what the patch changes.

Please amend/correct in case I missed something/misunderstood the patch; thanks!

[3] https://wiki.ubuntu.com/StableReleaseUpdates#SRU_Bug_Template

[Mauricio Faria de Oliveira (mfo)]

Review feedback:

- Versioning:

It's standard practice that version string for stable releases are
incremented by `ubuntu0.1`, and development release, by `ubuntu1`
(please see [1]).

This SRU/upload incremented by `ubuntu1`, but this was accepted
previously (noble: ubuntu1, noble-proposed: ubuntu2), and isn't
used in the publishing history [2].

So, this isn't an issue this time (but it could, were the version
used in a newer release), so just please note it in future uploads. :)

 openvswitch | 3.3.0-1ubuntu1 | noble | source
 openvswitch | 3.3.0-1ubuntu2 | noble-proposed | source
 openvswitch | 3.3.0-5 | oracular | source

- Patch file

DEP-3 headers [3] Bug-Ubuntu: and Origin: would be nice in the future
(not strictly required, but if you start with git patch(es), like this,
it's the only 2 lines to add, and it may help with future maintenance.)

Thanks for the work, bug report, and upload.
Hope this helps!
Mauricio

[1] https://wiki.ubuntu.com/SecurityTeam/UpdatePreparation#Update_the_packaging
[2] https://launchpad.net/ubuntu/+source/openvswitch/+publishinghistory
[3] https://dep-team.pages.debian.net/deps/dep3/

[Mauricio Faria de Oliveira (mfo)]

Hello Frode, or anyone else affected,

Accepted openvswitch into noble-proposed. The package will build now and be available at https://launchpad.net/ubuntu/+source/openvswitch/3.3.0-1ubuntu3 in a few hours, and then in the -proposed repository.

Please help us by testing this new package. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how to enable and use -proposed. Your feedback will aid us getting this update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug, mentioning the version of the package you tested, what testing has been performed on the package and change the tag from verification-needed-noble to verification-done-noble. If it does not fix the bug for you, please add a comment stating that, and change the tag to verification-failed-noble. In either case, without details of your testing we will not be able to proceed.

Further information regarding the verification process can be found at https://wiki.ubuntu.com/QATeam/PerformingSRUVerification . Thank you in advance for helping!

N.B. The updated package will be released to -updates after the bug(s) fixed by this package have been verified and the package has been in -proposed for a minimum of 7 days.

[Ubuntu SRU Bot (ubuntu-sru-bot)]

All autopkgtests for the newly accepted openvswitch (3.3.0-1ubuntu3) for noble have finished running.
The following regressions have been reported in tests triggered by the package:

netplan.io/unknown (armhf)
openvswitch/3.3.0-1ubuntu3 (ppc64el)

Please visit the excuses page listed below and investigate the failures, proceeding afterwards as per the StableReleaseUpdates policy regarding autopkgtest regressions [1].

https://people.canonical.com/~ubuntu-archive/proposed-migration/noble/update_excuses.html#openvswitch

[1] https://wiki.ubuntu.com/StableReleaseUpdates#Autopkgtest_Regressions

Thank you!

[Frode Nordahl (fnordahl)]

Hello, Mauricio,

Thank you for your thorough review.

The version was chosen knowing that there is a higher order version already present in Oracular and that a point release update will eventually be SRU'ed using the full `0.24.04.1` suffix, I could still have used the micro version as you propose and I will remember that for the next update, thank you for pointing it out.

The patch file could indeed deserve a DEP-3 header, I definitively use those whenever we need to adapt the patch, this one applied straight from the upstream stable branch, so not adding it was an oversight. Will make sure to add those for future uploads.

[Frode Nordahl (fnordahl)]

FTR; the autopkgtest results have been reviewed/triaged and as a result re-runs have been requested.

[Frode Nordahl (fnordahl)]

Build logs do indeed contain a passing `ofproto-dpif - continuation with meters` test.

OVN 24.03.2 successfully built: https://launchpad.net/ubuntu/+source/ovn/24.03.2-0ubuntu0.24.04.1

And has passing autopkgtests:
https://autopkgtest.ubuntu.com/results/autopkgtest-noble/noble/amd64/o/ovn/20240607_090928_25bd5@/log.gz
https://autopkgtest.ubuntu.com/results/autopkgtest-noble/noble/arm64/o/ovn/20240607_120742_4ddc6@/log.gz
https://autopkgtest.ubuntu.com/results/autopkgtest-noble/noble/ppc64el/o/ovn/20240607_090715_e1bc7@/log.gz
https://autopkgtest.ubuntu.com/results/autopkgtest-noble/noble/s390x/o/ovn/20240607_084019_1d3e3@/log.gz

[Launchpad Janitor (janitor)]

This bug was fixed in the package openvswitch - 3.3.0-1ubuntu3

---------------
openvswitch (3.3.0-1ubuntu3) noble; urgency=medium

  * d/p/ofproto-dpif-xlate-Fix-continuations-with-associated-
    metering.patch: Cherry-pick fix required for OVN 24.03.2
    (LP: #2066906).

openvswitch (3.3.0-1ubuntu2) noble; urgency=medium

  * d/tests: Allow populating TEST_LIST from environment.
  * d/tests: Bump number of 1G hugepages to 3 for arm64 (LP: #2063112).
  * d/tests: Skip flaky tests.
  * d/tests/testlist.py: Handle range of single test correctly.
  * d/tests: Skip tests not fitting memory constraints in arm64 CI
    (LP: #2063152).

 -- Frode Nordahl <email address hidden> Thu, 23 May 2024 13:14:27 +0000

[Łukasz Zemczak (sil2100)]

The verification of the Stable Release Update for openvswitch has completed successfully and the package is now being released to -updates. Subsequently, the Ubuntu Stable Release Updates Team is being unsubscribed and will not receive messages about this bug report. In the event that you encounter a regression using the package from -updates please report a new bug using ubuntu-bug and tag the bug report regression-update so we can easily find any regressions.