Re: [PATCH v11 00/25] DPAA driver fixes and improvements

On Thu, 13 Aug 2026 20:11:40 +0530
Hemant Agrawal [off-list ref] wrote:

v11: address AI review comments; skipping false positive ones.
  - net/dpaa: fix lost close error in rte_dpaa_remove()
  - net/dpaa: fix NULL dereference on rte_dpaa_bpid_info in oldev rx
    queue setup
  - net/dpaa: fix error logging in oldev tx queue setup to include
    return code and return proper -EINVAL

v10: address further AI review comments
  - dma/dpaa: expose compile-time static flags (sg_enable,
    data_validation, pci_read) as probe-time devargs with corrected
    logic and updated documentation
  - bus/dpaa: fix pool-channel FQ shutdown always returning -EINVAL
    (wrong affinity check; dedicated and pool channel ranges never
    overlap)
  - mempool/dpaa: fix double-free of bpid_info on pool release

v9: fix bissect issue and other AI review comments

This series collects a set of correctness fixes, cleanups and feature
additions across the NXP DPAA bus, net, mempool and DMA drivers.

1. Bus/fman infrastructure cleanups (patches 01, 02, 12)
   - bus/dpaa: refine fman naming and fix global scope
   - bus/dpaa: scan max BPID from DTS
   - bus/dpaa: improve log macro and fix bus detection

2. Process-type guards (patch 03)
   - drivers: add process-type guards to prevent segfaults in secondary

3. FQ shutdown hardening (patches 04-08)
   - bus/dpaa: shutdown DPAA FQ by fq descriptor
   - bus/dpaa: improve FQ shutdown with channel validation
   - bus/dpaa: add DPAA cgrid cleanup support

4. net/dpaa improvements (patches 09, 11, 13)
   - net/dpaa: optimize FM deconfig
   - net/dpaa: optimize FMC MAC type parsing

5. Statistics (patch 10)
   - drivers: add BMI Tx statistics

6. mempool/dpaa (patch 14)
   - drivers: release DPAA bpid on driver destructor

7. dma/dpaa (patches 15-16)
   - dma/dpaa: add SG data validation and ERR050757 fix
   - dma/dpaa: expose static flags as probe-time devargs

8. net/dpaa features (patches 17-22)
   - net/dpaa: support Rx/Tx taildrop threshold devarg
   - net/dpaa: add Tx rate limiting API
   - bus/dpaa: orp queue create and burst enqueue
   - net/dpaa: support fmcless rxq number as devargs
   - net/dpaa: support non fmX-macY type of shared Ethernet name
   - bus/dpaa: optimize DPAA multi-entry buffer pool operations

9. Shutdown and cleanup (patches 23-24)
   - drivers: improve shutdown fq with channel
   - drivers: improve crypto fq resource handling
   - net/dpaa: enhance VSP port support
   - drivers: add offline (O/H) port device support
   - drivers: add dpaax enter destructor to gate EAL memory frees

10. Documentation and release notes (patch 25)
    - doc: update release notes with NXP DPAA changes

Gagandeep Singh (3):
  net/dpaa: fix device remove
  dma/dpaa: add SG data validation and ERR050757
  drivers: add offline (O/H) port device support

Hemant Agrawal (6):
  net/dpaa: optimize FM deconfig
  net/dpaa: support Rx/Tx taildrop threshold devarg
  net/dpaa: add Tx rate limiting API
  net/dpaa: support non fmX-macY type of shared Ethernet name
  bus/dpaa: improve log macro and fix bus detection
  doc: update release notes with NXP DPAA changes

Jun Yang (14):
  bus/dpaa: refine fman naming and fix global scope
  bus/dpaa: scan max BPID from DTS
  drivers: shutdown DPAA FQ by fq descriptor
  bus/dpaa: improve FQ shutdown with channel validation
  bus/dpaa: add DPAA cgrid cleanup support
  drivers: add BMI Tx statistics
  net/dpaa: optimize FMC MAC type parsing
  drivers: release DPAA bpid on driver destructor
  bus/dpaa: orp queue create and burst enqueue
  net/dpaa: support fmcless rxq number as devargs
  bus/dpaa: optimize DPAA multi-entry buffer pool operations
  drivers: improve shutdown fq with channel
  drivers: improve  crypto fq resource handling
  drivers: add dpaax enter destructor to gate EAL memory frees

Prashant Gupta (1):
  drivers: add process-type guards for secondary process

Vanshika Shukla (1):
  net/dpaa: enhance VSP port support

 doc/guides/dmadevs/dpaa.rst               |  29 +-
 doc/guides/nics/dpaa.rst                  |  39 +-
 doc/guides/rel_notes/release_26_11.rst    |  29 +
 drivers/bus/dpaa/base/fman/fman.c         |  23 +-
 drivers/bus/dpaa/base/fman/fman_hw.c      | 111 ++--
 drivers/bus/dpaa/base/qbman/bman.c        |  59 +-
 drivers/bus/dpaa/base/qbman/bman_driver.c |  51 +-
 drivers/bus/dpaa/base/qbman/qman.c        | 207 +++++--
 drivers/bus/dpaa/base/qbman/qman.h        |  23 +-
 drivers/bus/dpaa/base/qbman/qman_driver.c |  29 +-
 drivers/bus/dpaa/bus_dpaa_driver.h        |  11 +-
 drivers/bus/dpaa/dpaa_bus.c               | 164 +++--
 drivers/bus/dpaa/dpaa_bus_base_symbols.c  |   4 +
 drivers/bus/dpaa/include/fman.h           |  20 +-
 drivers/bus/dpaa/include/fsl_bman.h       |  49 +-
 drivers/bus/dpaa/include/fsl_qman.h       |  54 +-
 drivers/common/dpaax/compat.h             |  22 +-
 drivers/common/dpaax/dpaax_iova_table.c   |  25 +-
 drivers/crypto/dpaa_sec/dpaa_sec.c        | 109 +++-
 drivers/dma/dpaa/dpaa_qdma.c              | 110 +++-
 drivers/mempool/dpaa/dpaa_mempool.c       |  64 +-
 drivers/mempool/dpaa/dpaa_mempool.h       |   3 +-
 drivers/net/dpaa/dpaa_ethdev.c            | 310 ++++++++--
 drivers/net/dpaa/dpaa_ethdev.h            |  25 +-
 drivers/net/dpaa/dpaa_flow.c              | 221 +++++--
 drivers/net/dpaa/dpaa_flow.h              |   7 +-
 drivers/net/dpaa/dpaa_fmc.c               |  73 ++-
 drivers/net/dpaa/dpaa_oldev.c             | 705 ++++++++++++++++++++++
 drivers/net/dpaa/fmlib/fm_lib.c           |  32 +-
 drivers/net/dpaa/fmlib/fm_port_ext.h      |   4 +-
 drivers/net/dpaa/meson.build              |  11 +-
 drivers/net/dpaa/rte_pmd_dpaa.h           |  23 +-
 drivers/net/dpaa/rte_pmd_dpaa_oldev.h     |  93 +++
 33 files changed, 2282 insertions(+), 457 deletions(-)
 create mode 100644 drivers/net/dpaa/dpaa_oldev.c
 create mode 100644 drivers/net/dpaa/rte_pmd_dpaa_oldev.h

Using Claude Opus 5 and latest AGENTS.md, it still sees some
issues worth investigating.

Errors

Patch 04/25: drivers: add process-type guards for secondary process

  The qman.c hunk is an unrelated, undocumented bug fix.  The old code
  read the FQD dest_wq as:

      channel = dest_wq & 0x7;
      wq      = dest_wq >> 3;

  and the new qm_fqd_get_chan()/qm_fqd_get_wq() helpers reverse that:
  channel is bits 15:3, wq is bits 2:0.  The new split is the correct
  one -- qman_shutdown_fq() compares channel against the pool-channel
  range, which cannot work with a 0-7 value -- but this is a real
  functional fix landing in a patch whose subject and body describe
  only RTE_PROC_PRIMARY guards.  Split it into its own commit with a
  Fixes: tag and Cc: stable@dpdk.org.

  Info: qm_fqd_get_wq(), qm_fqid_set() and QM_FQID_MASK are added but
  never used anywhere in the tree.  Commit message says "rivers".

Patch 08/25: drivers: add BMI Tx statistics

  The Tx BMI counters are never enabled and never reset.  get_all()
  now correctly reads tx_regs, but:

    fman_if_bmi_stats_enable()  -> writes only regs->fmbm_rstc
    fman_if_bmi_stats_disable() -> writes only regs->fmbm_rstc
    fman_if_bmi_stats_reset()   -> zeroes only the eight Rx counters

  FMBM_TSTC has the same counters-enable bit as FMBM_RSTC, so with the
  bit clear the four new tx_* xstats read 0 for the lifetime of the
  port.  Even if the hardware counted unconditionally,
  rte_eth_xstats_reset() would still leave tx_bad_frames_count,
  tx_frame_length_discard, tx_frames_unsupported_format and
  tx_buf_deallocate at their old values -- a stats_reset that silently
  resets only part of the set.

  Apply FMAN_BMI_COUNTERS_EN to tx_regs->fmbm_tstc in enable/disable,
  and zero the four Tx counters in reset.  The register offsets
  themselves check out: fmbm_tstc lands at 0x200, matching fmbm_rstc.

Warnings

Patch 06/25: bus/dpaa: improve FQ shutdown with channel validation

  The commit message describes the opposite of the code.  It claims
  "add validation that the portal's affine channel matches the FQ's
  channel for pool-channel FQs, and only restore SDQCR when it was
  actually changed."  The diff removes the affinity check (the inline
  comment explains, correctly, that it could never pass) and restores
  SDQCR unconditionally.  Rewrite the message to match.

  Info: pr_warn("Pool channel(%04x) configured != default(0x%04x)")
  -- the first value is missing its 0x prefix.

Patch 07/25: bus/dpaa: add DPAA cgrid cleanup support

  qman_find_fq_by_cgrid() is a linear MC-command scan of the FQID
  space, and dpaa_eth_dev_close() calls it once per Rx CGR and once
  per Tx CGR.  Each iteration is a QUERYFQ_NP hardware round trip,
  plus a QUERYFQ for every non-OOS FQ.  It terminates early on
  ERR_FQID, so the bound is the provisioned FQD count rather than
  2^24, but on a 16+16 queue port that is still 32 full scans of every
  FQ on the SoC on each port close.  Consider tracking CGR->FQ
  association in software, or at minimum bounding the scan to the
  FQID range the driver itself allocated.

  Second issue: the function returns on the first match.  If more than
  one FQ is bound to the CGR, only one is shut down and the CGR is
  deleted with the others still attached -- which is the exact
  condition the patch is trying to prevent.  It needs to loop until no
  FQ remains.

Patch 10/25: net/dpaa: optimize FMC MAC type parsing

  Unknown mac_type values now fall through to a bare index compare.
  The old code ended with an explicit
  "Invalid MAC(mac_idx=%d) type(%d)" / -EINVAL for any type outside
  {1g, 2.5g, 10g}.  In the new form those types skip all three
  pport->type filters and match on fif->mac_idx == num alone, so an
  fman_onic or fman_offline_internal interface can bind to a port
  entry of any direction or speed.  Restore the explicit reject.

  dpaa_port_fmc_get_idx_from_name() uses atoi() with no validation --
  a malformed suffix yields index 0, which silently matches mac_idx 0.
  Use strtol() and check the end pointer.

  DPAA_PMD_INFO("MAC index of %s is %d") fires for every port entry on
  every interface during FMC parse.  This belongs at DEBUG.

  Info: the && continuations are indented one tab, level with the
  statement body:

      if (fif->mac_type == fman_mac_1g &&
              pport->type != e_FM_PORT_TYPE_RX)
              return -ENODEV;

  DPDK style wants a double-indent continuation here.

Patch 02/25: bus/dpaa: refine fman naming and fix global scope

  Subject promises a global-scope fix; the patch is a pure rename of
  ccsr_map -> memac_map and bmi_map -> rx_bmi_map.  Drop the second
  clause from the subject.

Patch 11/25: drivers: release DPAA bpid on driver destructor

  dpaa_mbuf_free_pool() now frees rte_dpaa_bpid_info and NULLs it once
  the last mempool is gone.  That pointer is cached in
  rxq->bp_array by both dpaa_ethdev.c and dpaa_oldev.c at queue-setup
  time, and dpaa_rxtx.c reads rte_dpaa_bpid_info on the fast path.
  Freeing an mbuf pool while any port still holds a queue set up
  against it leaves those bp_array pointers dangling.  Either
  reference-count the table or leave it allocated until the
  destructor.

  Info: #define RTE_PRIORITY_104 104 puts a new RTE_-prefixed macro in
  a driver .c file.  dpaa_finish() in net/dpaa passes 103 as a literal;
  match that.

Not yet covered: patches 12-25 (DMA SG validation, taildrop devargs, Tx
rate limiting, ORP enqueue, fmcless/shared-MAC devargs, bpool
optimization, the log-macro/bus-detection patch that carried the v6
dpaa_bus_dev_compare error, VSP, the 705-line offline-port driver,
crypto FQ handling, the dpaax destructor gate, and the release notes).