DLPX-98112 linux-kernel-generic: rebase Delphix patchset onto Ubuntu-hwe-7.0-7.0.0-28.28_24.04.1

[lyriclake]

Rebases the Delphix patchset onto Ubuntu tag Ubuntu-hwe-7.0-7.0.0-28.28_24.04.1, replacing the previous base
Ubuntu-hwe-6.17-6.17.0-40.40_24.04.1. Recreates the rebase that linux-pkg/develop/update-package/linux-kernel-generic failed to complete.

Field	Value
Upstream tag	Ubuntu-hwe-7.0-7.0.0-28.28_24.04.1 (917185778ef2b3cb674c703dfd055eb0c4af67de)
Previous base	Ubuntu-hwe-6.17-6.17.0-40.40_24.04.1 (063d17a50448cbb5ef703221a62f6af6d1fb78d1)
Patchset	9 patches (0 dropped of 9)
Conflicts resolved	1
Jira	DLPX-98112

Note: GitHub will show "This branch has conflicts that must be resolved" and
a diff against the whole Ubuntu bump. Both are expected: this branch does not
descend from develop — the kernel forks are rebased onto each new
Ubuntu tag with git cherry-pick, not merged. Review the patchset commits, not
the aggregate diff. Landing is a force-push, not a GitHub merge.

8f4441107409 — Extract PKG_ABI from mutated string

Conflicting: include/asm-generic/mshyperv.h
Upstream cause: 56f727d2cc7a — UBUNTU: SAUCE: (no-up) hv: Supply vendor ID and package ABI
Verdict: RE_APPLY — Ubuntu SAUCE relocated the #ifndef PKG_ABI block from before hv_do_rep_hypercall() to after it while re-applying the SAUCE stack onto the 7.0 mainline; content unchanged. New upstream still defines PKG_ABI as a bare unquoted integer, so the Delphix ABI-string mutation handling is still needed. Ported the same string-mutation logic to the relocated position verbatim.

Testing Done

ab-pre-push build #14705: FAILURE

• Build: https://selfservice-jenkins.eng-tools-prd.aws.delphixcloud.com/job/appliance-build-orchestrator-pre-push/14705/
• Tested: linux-kernel-generic@e92af378a006 (against develop)
• Suites: build — FAILURE. ipu7-drivers DKMS module failed to compile against the new kernel's V4L2 subdev API (drivers/media/i2c/ov02c10.c: error: implicit declaration of function 'v4l2_subdev_state_get_format' / 'struct v4l2_subdev' has no member named 'entity') — a stale Ubuntu-side DKMS pin, unrelated to this patchset, that our build tooling doesn't exclude. Fix: DLPX-98296 linux-kernel-generic pre-push build fails: ipu7-drivers DKMS module missing do_ipu7=false disable flag linux-pkg#405 (DLPX-98296).

ab-pre-push build #14718: IN PROGRESS

• Build: https://selfservice-jenkins.eng-tools-prd.aws.delphixcloud.com/job/appliance-build-orchestrator-pre-push/14718/
• Tested: linux-kernel-generic@e92af378a006, linux-pkg@065c38acf7f7, zfs@e62b28d4bbc4 (against develop)
• Status: running — re-test with the linux-pkg do_ipu7=false fix (PR #405) and the zfs sockaddr_unsized fix (PR #2305, needed because this hwe-7.0 kernel's headers carry the same API change as aws's) both included as extra repos; re-run recording in completed mode once finished.

ab-pre-push build #14718: FAILURE

• Build: https://selfservice-jenkins.eng-tools-prd.aws.delphixcloud.com/job/appliance-build-orchestrator-pre-push/14718/
• Tested: linux-kernel-generic@e92af378a006, linux-pkg@065c38acf7f7, zfs@e62b28d4bbc4 (against develop)
• Suites: build — SUCCESS (appliance-build » develop » pre-push #7152); both prior fixes (do_ipu7=false, zfs sockaddr_unsized) landed cleanly. Tests — FAILURE, both root-caused as infra flakiness unrelated to this patchset:
  • upgrade-testing #4609 failed inside blackbox-chained #9566 — java.io.EOFException: Unexpected end of ZLIB input stream while expanding drivers.master.v10.tar.gz (corrupted/truncated download).
  • dx-integration-tests #36211 failed in two branches: "Blackbox pre_checkin tests" (blackbox-self-service #226272) hit java.net.UnknownHostException: artifactory.delphix.com (DNS resolution failure); "Unit Tests" (integration-tests » split-precommit-unittest » pre-push #40041) hung on the Jenkins swarm-agent SSH connect for ~2h13m before Broken pipe.
  • Retriggered directly (reusing the same successful build's artifacts — appliance-build #7152's image and the ami-snapshots #9999 AMI — no rebuild needed): upgrade-testing #4613, dx-integration-tests #36269. Results recorded below.

ab-pre-push build #14718: completed

• dx-integration-tests rerun #36269: SUCCESS.
• upgrade-testing rerun #4613 → blackbox-chained #9570: UNSTABLE — 37/46 sub-tests failed, all cascading from a single root cause: the appliance's management stack failed to restart after power-cycling within the 897s timeout during initial_setup_tests::test_setup_swap, which then failed every dependent test (admin profile setup, virtualization SDK install, upgrade request/pre tests). This is an infra/environment-recovery issue, not a regression from the kernel rebase or either fix.
• Reran again, reusing the same build's artifacts: upgrade-testing #4618 (image: appliance-build » pre-push #7152's upgrade artifact). Result recorded below.

upgrade-testing rerun #4618 → blackbox-chained #9575: FAILURE (progress, but same infra pattern)

9/46 sub-tests failed (down from 37/46 last time — the management-stack-restart timeout from the last run didn't recur). All 9 failures trace to a single root cause: SSH/connectivity timeouts to the test appliance starting at test_upload_and_delete_upgrade_image ("management stack did not come up in 300s"), cascading through stop_at_failure dependencies in upgrade_basic_pre/upgrade_pre.

Notably, linux-kernel-aws's rerun (#72) hit the exact same 9 tests, same root cause, in parallel — two unrelated kernel patchsets failing identically points to a shared DCenter/network infra issue at the time, not either kernel rebase.

Reran again reusing the same image: upgrade-testing #4622. Result recorded below.

upgrade-testing rerun #4622 → blackbox-chained #9579: UNSTABLE (effectively a clean pass)

0/97 sub-tests failed. The only two non-passes were skips unrelated to this rebase:

• test_host_cacert_after_upgrade — "No host cert was stored which is required for this test to run" (pre-existing test-environment condition).
• test_validate_mssql_backup_and_validate_vdbs_replication_target — subprocess failure on the replication target, matched known bug DLPX-98231.

Neither implicates the kernel rebase. This clears the infra signature from the last two rounds (management-stack-restart timeout, then SSH/connectivity timeouts to the test appliance during setup) — it did not recur here. Test stage for generic is now green. Compile stage was already confirmed (appliance-build » pre-push #7152: SUCCESS). No further reruns needed on this platform.

linux-kernel-aws's parallel rerun (#72) still hit an SSH connectivity timeout on its next round, but during a different test/stage than the prior rounds (test_upgrade_linux_system, mid-upgrade, not test_upload_and_delete_upgrade_image in setup) — see that PR for the full breakdown and outstanding rerun.

[prakashsurya]

LGTM, pending successful ab-pre-push.

[lyriclake]

Testing status update

ab-pre-push build #14718: completed

• Build: https://selfservice-jenkins.eng-tools-prd.aws.delphixcloud.com/job/appliance-build-orchestrator-pre-push/14718/
• Tested: linux-kernel-generic@e92af378a006, linux-pkg@065c38acf7f7 (linux-pkg#405), zfs@e62b28d4bbc4 (zfs#2305) (against develop)
• Compile stage: SUCCESS — appliance-build » develop » pre-push #7152 (both fixes landed cleanly).
• dx-integration-tests rerun #36269: SUCCESS.
• upgrade-testing rerun #4613 → blackbox-chained #9570: UNSTABLE — 37/46 sub-tests failed, all cascading from a single root cause: the appliance's management stack failed to restart after power-cycling within the 897s timeout during initial_setup_tests::test_setup_swap, which then failed every dependent test (admin profile setup, virtualization SDK install, upgrade request/pre tests). This is an infra/environment-recovery issue, not a regression from the kernel rebase or either fix.
• Reran again, reusing the same build's artifacts: upgrade-testing #4618 (image: appliance-build » pre-push #7152's upgrade artifact). IN PROGRESS.

Will record the final result once it finishes.

[lyriclake]

upgrade-testing rerun #4618 → blackbox-chained #9575: FAILURE (progress, but same infra pattern)

9/46 sub-tests failed (down from 37/46 last time — the management-stack-restart timeout from the last run didn't recur). All 9 failures trace to a single root cause: SSH/connectivity timeouts to the test appliance starting at test_upload_and_delete_upgrade_image ("management stack did not come up in 300s"), cascading through stop_at_failure dependencies in upgrade_basic_pre/upgrade_pre.

Notably, linux-kernel-aws's rerun (#72) hit the exact same 9 tests, same root cause, in parallel — two unrelated kernel patchsets failing identically points to a shared DCenter/network infra issue at the time, not either kernel rebase.

Reran again reusing the same image: upgrade-testing #4622. IN PROGRESS.

[lyriclake]

upgrade-testing rerun #4622 → blackbox-chained #9579: UNSTABLE (effectively a clean pass)

0/97 sub-tests failed. The only two non-passes were skips unrelated to this rebase:

• test_host_cacert_after_upgrade — "No host cert was stored which is required for this test to run" (pre-existing test-environment condition).
• test_validate_mssql_backup_and_validate_vdbs_replication_target — subprocess failure on the replication target, matched known bug DLPX-98231.

Neither implicates the kernel rebase. This clears the infra signature from the last two rounds (management-stack-restart timeout, then SSH/connectivity timeouts to the test appliance during setup) — it did not recur here. Test stage for generic is now green; compile stage was already confirmed (appliance-build » pre-push #7152: SUCCESS). No further reruns needed on this platform.

linux-kernel-aws's parallel rerun (#72) still hit an SSH connectivity timeout, but during a different test/stage than the prior rounds (test_upgrade_linux_system, mid-upgrade, not test_upload_and_delete_upgrade_image in setup) — see that PR for the full breakdown and the open question on whether to rerun once more.

[lyriclake]

Landed on develop as e92af378a006c6b42046924e21370d3ccd7e9b5e via force-push (rebased patchset — the kernel forks cannot be merged through GitHub, see the PR body). Closing manually.

Overridden gates at land time (explicitly confirmed by requester):

• Approval: reviewDecision was REVIEW_REQUIRED, but one APPROVED review exists (prakashsurya); repo is transitioning to a single-review policy.
• Checks: jenkinsci/appliance-build-orchestrator and jenkinsci/build-package showed stale FAILURE from before fix PRs (linux-pkg#405, zfs#2305) landed; Testing Done log confirms a later green build (appliance-build » pre-push #7152: SUCCESS) and clean test rerun with those fixes included.