Multiple acceptance tests improvements - #1673
Merged
Merged
Conversation
Sakoram
requested review from
DawidWesierski4,
awilczyns,
moleksy and
soopel
as code owners
August 4, 2026 12:40
Tracked files deleted in the working tree were passed to Black and aborted formatting. Filter non-existent paths for every formatter input. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Remove application legs that cannot express the behavior under test. Add BPM to the packing sweep and retain only mechanism-relevant axes. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Exercise pacing implementations across load and shaping, and RSS across baseline and replica stress. Parse the resolved datapath so silent fallback cannot pass. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Consolidate kernel socket, AF_XDP, virtio-user, and PF coverage around representative loads. Verify the backend MTL actually initialized, including AF_XDP zero-copy modes. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Run ST 2022-7 as one four-port session and capture both destinations for compliance. Preserve the original failure when an oracle was never dispatched. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Keep the two runnable mixed-application directions at one baseline stream shape. Give RxTxApp TX and FFmpeg RX distinct ports so both processes can initialize. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Generate a deterministic media corpus and size RGB8 integrity frames correctly. Cover eight RFC 4175 formats plus a reduced packing interaction matrix. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Document behavior-first coverage, matrix depth, load profiles, path assertions, oracles, interoperability, redundancy, and deterministic asset policy. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Require both the MTL timing parser and EBU LIST to report narrow compliance for every expected video stream. Bound capture by media frames and distinguish unavailable EBU analysis. Avoid applying user PTP correction twice to hardware RX timestamps. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
ptp_delta was added to every hardware RX timestamp in mbuf_hw_time_stamp(), but it was never a valid offset for one. With timesync, ptp_timesync_adjust_time() steps the NIC PHC that those timestamps are derived from, so they already carry the correction and re-applying it accumulates without bound. Without timesync -- any port missing MT_IF_FEATURE_TIMESYNC, which is every VF, the kernel/AF_XDP backends, and any PF whose PMD cannot enable it -- that adjust only moves the software no_timesync_delta accumulator, a TSC-domain offset. Adding that to a PTP-domain timestamp shifted every RX timestamp by the whole TSC-to-TAI gap, which made the ST 2110-21 RX timing parser reject every frame and left the public timestamp_first_pkt (documented as TAI) far from TAI. Measured with tests/single/rx_timing on a VF, from MTL's rv_tp_stat: before COMPLIANT NARROW 0 WIDE 0 FAILED 521, after 766 narrow 0 wide 4 failed. Remove the field, its accumulation in ptp_adjust_delta(), and its use in mbuf_hw_time_stamp(). adjust_delta_test asserted the double-accounting as intended behaviour; that expectation now covers no_timesync_delta, the accumulator the correction legitimately belongs in. Where the NIC clock is not itself PTP-synced the timestamp is free-running and no software correction remains to fix it, so a synced PF PHC is now a requirement for VF hardware-timestamp sessions. Signed-off-by: Kasiewicz, Marek <marek.kasiewicz@intel.com>
Sakoram
force-pushed
the
multiple-acceptance-tests-improvments
branch
from
August 27, 2026 14:00
70d15aa to
bf58f6e
Compare
awilczyns
approved these changes
Aug 28, 2026
DawidWesierski4
approved these changes
Aug 28, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.