[PATCH v8 00/28] net/hinic3: add PMD features and fixes

Jiacheng Ye yejiacheng4 at h-partners.com
Fri Nov 20 17:34:25 CET 2026


This series fixes offload capability advertisement, cmdq/compact CQE
configuration, and a handful of driver defects, and adds NIC
type identification, FEC, CIR drop statistics, VF LAG/link state, COS
mask/map, mbox recording, RSS hash configuration, compact CQE Rx/Tx
paths, TSO segment expansion, Tx burst mode reporting, queue statistics,
pseudo-header checksum, per-tile cmdq ops, SP560 NIC support, and device
kvargs tunables.

Changes in v8:
- 0009: fold the v7 "add NIC type identification" patch into the
  pseudo-header checksum fix that follows it, which is the only user of
  the new HINIC3_IS_SP230_NIC() helper, and name the NIC type in that
  commit's subject as well. The merged commit carries the Cc: stable tag.
  The series is 28 patches from here on and the later numbering is one
  lower than in v7. The shared-mbuf rejection covers only the writes the
  software performs: TCP/UDP checksum, TSO and outer UDP checksum; the
  SCTP checksum is computed by the hardware and writes nothing back.
- 0006: drop the rte_atomic_thread_fence(); the cleared index is only
  consumed when the ring is re-armed and the RQ context is sent to the
  device at start, so no explicit barrier is required.
- 0007: drop the dead null check on the secondary process path.
- 0010: reject a queue count that does not fit in the MSI-X vector table,
  on both the intr_conf.rxq and the SP230 RQ-context paths, and document
  the new -EINVAL.
- 0014: stop printing the unused mbox slots in the debug dump.
- 0016: restore admin_down when dev_set_link_up() fails, as
  dev_set_link_down() already does.
- 0019: spell out in the commit message that the TCAM does not support
  VLAN matching, so the vlan_flag mask/value pairs were meaningless and are
  removed, and a rule now hits tagged and untagged packets alike.
- 0021/0022: turn the Runtime Configuration section into a definition
  list, with the compact-CQE switch documented in the compact CQE patch.
- 0022: when an application asks for the separate Rx mode on a NIC that
  reports the SW compact CQE bit, clear that bit as well, so the Rx
  burst, the RQ context, the CEQ attributes and the vport flag agree.
  The HW compact CQE bit is never cleared, and nothing is cleared when
  the compact mode is requested.
- 0022: clear the compact CQE info before each packet, so the LRO, VLAN
  and data-offset fields of a previous 16-byte CQE cannot leak into the
  next packet.
- 0024: reword the copy pool comment: the 256 mbufs per queue cap is
  deliberate, and a burst stops at the packet that cannot be copied.
- 0025: rename the Tx compact-task helpers from *_compact_cqe to
  *_compact_task, report "compact task" in the burst mode string, drop
  the dead !sq_wqe branch, and document that the scalar path no longer
  sets TCPUDP_CS.

Changes in v7:
- 0008: treat a missing FDIR rule as a successful no-op instead of
  returning -EINVAL, and gate the RQ enable command on NIC_F_HTN_FDIR so
  the HTN-only command is not sent to STN devices.
- 0011: when the application does not request Rx interrupts
  (intr_conf.rxq == 0), name the queue vector in the SP230 RQ context and
  leave that vector disabled, so no interrupt is reported while Rx keeps
  polling the descriptor ring. No event fds, intr_vec or MSI-X attribute
  commands are used on this path. Backends without per-vector support
  (igb_uio, uio_pci_generic) keep the previous behaviour and are left
  untouched.
- 0012: move the Fixes: tags back to the trailer block.
- 0016: use the 12-character commit hash in Fixes:.
- 0017: set admin_down before disabling the vport, and clear it again if
  the disable fails, so a link-up event cannot re-enable the vport in
  between. dev_start() also skips enabling the vport while admin_down is
  set.
- 0019: derive the cached per-queue RSS enablement from rss_state and
  rss_type.val, refresh it from hinic3_rss_hash_update(), and access it
  with relaxed atomics, because the control path writes it while the Rx
  data path reads it. Include rte_stdatomic.h explicitly instead of
  relying on another header to pull it in.
- 0022: treat a device argument list that is present but empty as no
  arguments instead of parsing a NULL string.
- 0023: use the existing rx_free_mbuf_bytes counter in the new htn Rx
  path instead of a non-existent field name.
- 0024: record the expected queue depth only after the queue has been set
  up successfully, so a failed setup does not leave a depth behind that
  the next attempt has to match, and check every configured queue at the
  top of dev_start(), before the Rx interrupt setup dereferences them.
- 0026: stop the Tx burst at the first packet that cannot be enqueued and
  keep that mbuf owned by the application, so the returned count stays a
  consumed prefix as rte_eth_tx_burst() requires. The per-burst packet
  counter that was only ever equal to the loop index is gone, so the
  statistics use the returned count directly.

Changes in v6:
- Rebase on current main and address the v5 review comments.
- 0003/0005: correct the Fixes: tags, drop the Cc: stable request for
  fixes to code that has never been in a stable release, and remove the
  dead null check in the vport enable fix.
- 0006: clear the stale hardware consumer index on Rx queue reset.
- 0009: reject a packet whose first segment is shorter than the
  pseudo-header, narrow the shared-mbuf check to the L4/TSO flags,
  compare the tunnel type as a whole enum, and compute the outer and inner
  pseudo-header checksums.
- 0011: advertise OUTER_UDP_CKSUM now that the outer UDP pseudo-header is
  written in the software prepare step. Drop the stale QinQ features.ini
  entry and unused QINQ mbuf flag.
- 0012: select the CIR drop counters by NIC type in calc_num, get_names
  and get.
- 0013: report the needed capability count for a short array, keep the
  GET-side firmware NOT_SET to AUTO mapping, and reject unknown firmware
  values.
- 0014: take the PCI device from the ethdev in priv_dump so a secondary
  process does not dereference a stale pointer.
- 0016: keep a separate admin-down flag so a PF link up event cannot
  override dev_set_link_down().
- 0018: cache RSS enablement per queue and clear rss_type when RSS is
  disabled.
- 0019: reject outer IPv6 with an inner L3, and stop setting the VLAN
  flag mask in the TCAM keys.
- 0021: document that the coalescing tunables only take effect with
  HW compact CQE.
- 0022: report the combined compact-CQE checksum error as an unknown
  result instead of blaming the IP or the L4 layer, and keep the ptype
  table in sync with the advertised ptype set.
- 0023: reject a queue depth that differs from the first queue instead
  of coercing it, reset the expected depth in dev_configure() and
  dev_stop(), validate it in dev_start(), and mark Tx queues stopped in
  dev_stop().
- 0024: reject TSO packets that need more SGEs than the ring can hold,
  reject a tso_segsz that does not fit the 14-bit MSS field, and report
  nb_seg_max/nb_mtu_seg_max.
- 0025: clear queue_info in the scalar burst for compact WQEs like
  hinic3_master, clear the per-packet offload state, guard the prefetch
  against an empty burst, and remove the per-packet data path logs.
  Per-packet unsupported-type messages are lowered to DEBUG.
  Tunnel packets are accepted by type instead of by feature bit;
  VXLAN_GPE TSO and IPIP TSO on SP620 stay rejected.
  SP230 does not get the software pseudo-header writes because its
  hardware derives the inner checksums and TSO headers itself.
- 0027: report the Tx burst mode of the function that is installed.

The series builds with -Dwerror=true.

Jiacheng Ye (28):
  net/hinic3: fix SP230 VF PCI device ID
  net/hinic3: fix double removal in ethertype flow destroy
  net/hinic3: fix TCAM index leak on ethertype flow set
  net/hinic3: fix VXLAN VNI spec and mask write
  net/hinic3: fix vport enable null dereference
  net/hinic3: clear stale Rx hardware CI on queue reset
  net/hinic3: attach ethdev and set ops in secondary process
  net/hinic3: fix per-queue start/stop on non-HTN devices
  net/hinic3: fix pseudo-header checksum and identify NIC type
  net/hinic3: add SP230 Rx queue interrupt handling
  net/hinic3: fix offload and cmdq/compact CQE config
  net/hinic3: add vport statistics and VF LAG support
  net/hinic3: add FEC mode get/set and capability support
  net/hinic3: add mbox message recording and debug dump
  net/hinic3: add COS mask and COS map support
  net/hinic3: add VF link state and promiscuous support
  net/hinic3: convert cmdq ops to static global variables
  net/hinic3: update RSS hash type configuration
  net/hinic3: refactor flow rule parsing
  net/hinic3: add SP560 NIC type identification and handling
  net/hinic3: add device kvargs parameters support
  net/hinic3: add compact CQE Rx path and ptype table
  net/hinic3: ensure queue depth and state consistency
  net/hinic3: expand TSO from 127 to 255 segments
  net/hinic3: split Tx burst function for compact task support
  net/hinic3: add SP230 VF software statistics
  net/hinic3: add Tx burst mode get
  net/hinic3: expose Tx queue copied-mbuf statistics

 .mailmap                                      |    1 +
 doc/guides/nics/features/hinic3.ini           |    5 +-
 doc/guides/nics/hinic3.rst                    |   43 +-
 doc/guides/rel_notes/release_26_11.rst        |   19 +
 drivers/net/hinic3/base/hinic3_cmd.h          |    2 +
 drivers/net/hinic3/base/hinic3_csr.h          |    6 +-
 drivers/net/hinic3/base/hinic3_eqs.c          |   64 +-
 drivers/net/hinic3/base/hinic3_eqs.h          |    9 +-
 drivers/net/hinic3/base/hinic3_hw_cfg.c       |    2 +
 drivers/net/hinic3/base/hinic3_hw_cfg.h       |    8 +-
 drivers/net/hinic3/base/hinic3_hw_comm.c      |    9 +-
 drivers/net/hinic3/base/hinic3_hwdev.h        |    3 +-
 drivers/net/hinic3/base/hinic3_hwif.c         |    4 +-
 drivers/net/hinic3/base/hinic3_mbox.c         |   45 +
 drivers/net/hinic3/base/hinic3_mbox.h         |   19 +
 drivers/net/hinic3/base/hinic3_nic_cfg.c      |  201 +++-
 drivers/net/hinic3/base/hinic3_nic_cfg.h      |  105 +-
 drivers/net/hinic3/base/hinic3_nic_event.c    |   71 +-
 drivers/net/hinic3/hinic3_ethdev.c            | 1003 ++++++++++++++--
 drivers/net/hinic3/hinic3_ethdev.h            |   58 +-
 drivers/net/hinic3/hinic3_fdir.c              |   53 +-
 drivers/net/hinic3/hinic3_fdir.h              |    8 +-
 drivers/net/hinic3/hinic3_flow.c              |  590 ++++++---
 drivers/net/hinic3/hinic3_flow.h              |  196 ---
 drivers/net/hinic3/hinic3_nic_io.c            |  157 +--
 drivers/net/hinic3/hinic3_nic_io.h            |   98 ++
 drivers/net/hinic3/hinic3_rx.c                |  668 +++++++++--
 drivers/net/hinic3/hinic3_rx.h                |  214 ++--
 drivers/net/hinic3/hinic3_tx.c                | 1068 +++++++++++++----
 drivers/net/hinic3/hinic3_tx.h                |   63 +-
 .../net/hinic3/htn_adapt/hinic3_htn_cmdq.c    |   65 +-
 .../net/hinic3/htn_adapt/hinic3_htn_cmdq.h    |    6 +-
 .../net/hinic3/stn_adapt/hinic3_stn_cmdq.c    |   76 +-
 .../net/hinic3/stn_adapt/hinic3_stn_cmdq.h    |    6 +-
 34 files changed, 3752 insertions(+), 1193 deletions(-)
 delete mode 100644 drivers/net/hinic3/hinic3_flow.h

-- 
2.33.0



More information about the dev mailing list