mbox series

[GIT,PULL] Please pull RDMA subsystem changes

Message ID ZJzUeFT7lLqEjMJn@nvidia.com (mailing list archive)
State Superseded
Commit 884fe9da1b7ccbea31b118f902fbc78f58366b4a
Headers show
Series [GIT,PULL] Please pull RDMA subsystem changes | expand

Pull-request

git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma.git tags/for-linus

Message

Jason Gunthorpe June 29, 2023, 12:46 a.m. UTC
Hi Linus,

Here are the changes for RDMA for this cycle, there was a small rxe
conflict with v6.4 that I resolved in the usual way.

Thanks,
Jason

The following changes since commit 6995e2de6891c724bfeb2db33d7b87775f913ad1:

  Linux 6.4 (2023-06-25 16:29:58 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma.git tags/for-linus

for you to fetch changes up to 5f004bcaee4cb552cf1b46a505f18f08777db7e5:

  Merge tag 'v6.4' into rdma.git for-next (2023-06-27 14:06:29 -0300)

----------------------------------------------------------------
v6.5 merge window RDMA pull request

This cycle saw a focus on rxe and bnxt_re drivers:

- Code cleanups for irdma, rxe, rtrs, hns, vmw_pvrdma

- rxe uses workqueues instead of tasklets

- rxe has better compliance around access checks for MRs and rereg_mr

- mana supportst he 'v2' FW interface for RX coalescing

- hfi1 bug fix for stale cache entries in its MR cache

- mlx5 buf fix to handle FW failures when destroying QPs

- erdma HW has a new doorbell allocation mechanism for uverbs that is
  secure

- Lots of small cleanups and rework in bnxt_re
   * Use the common mmap functions
   * Support disassociation
   * Improve FW command flow

- bnxt_re support for "low latency push", this allows a packet

----------------------------------------------------------------
Arnd Bergmann (1):
      RDMA/irdma: avoid fortify-string warning in irdma_clr_wqes

Bob Pearson (10):
      RDMA/rxe: Add workqueue support for rxe tasks
      RDMA/rxe: Rename IB_ACCESS_REMOTE
      RDMA//rxe: Optimize send path in rxe_resp.c
      RDMA/rxe: Fix access checks in rxe_check_bind_mw
      RDMA/rxe: Introduce rxe access supported flags
      RDMA/rxe: Let rkey == lkey for local access
      RDMA/rxe: Implement rereg_user_mr
      RDMA/rxe: Send last wqe reached event on qp cleanup
      RDMA/rxe: Fixes mr access supported list
      RDMA/rxe: Simplify cq->notify code

Brendan Cunningham (3):
      IB/hfi1: Fix wrong mmu_node used for user SDMA packet after invalidate
      IB/hfi1: Add mmu_rb_node refcount to hfi1_mmu_rb_template tracepoints
      IB/hfi1: Remove unused struct mmu_rb_ops fields .insert, .invalidate

Bryan Tan (1):
      RDMA/vmw_pvrdma: Remove unnecessary check on wr->opcode

Cheng Xu (4):
      RDMA/erdma: Configure PAGE_SIZE to hardware
      RDMA/erdma: Allocate doorbell resources from hardware
      RDMA/erdma: Associate QPs/CQs with doorbells for authorization
      RDMA/erdma: Refactor the original doorbell allocation mechanism

Chengchang Tang (2):
      RDMA/hns: Fix hns_roce_table_get return value
      RDMA/hns: Add clear_hem return value to log

Colin Ian King (1):
      RDMA/bnxt_re: Fix spelling mistake "priviledged" -> "privileged"

Daisuke Matsuda (1):
      RDMA/rxe: Fix comments about removed tasklets

Dan Carpenter (1):
      RDMA/bnxt_re: Fix an IS_ERR() vs NULL check

Jason Gunthorpe (1):
      Merge tag 'v6.4' into rdma.git for-next

Junxian Huang (1):
      RDMA/hns: Remove unnecessary QP type checks

Kalesh AP (6):
      RDMA/bnxt_re: Fix to remove unnecessary return labels
      RDMA/bnxt_re: Use unique names while registering interrupts
      RDMA/bnxt_re: Remove a redundant check inside bnxt_re_update_gid
      RDMA/bnxt_re: Fix to remove an unnecessary log
      RDMA/bnxt_re: Return directly without goto jumps
      RDMA/bnxt_re: Remove unnecessary checks

Kamal Heib (3):
      RDMA/irdma: Return void from irdma_init_iw_device()
      RDMA/irdma: Return void from irdma_init_rdma_device()
      RDMA/irdma: Move iw device ops initialization

Kashyap Desai (19):
      RDMA/bnxt_re: wraparound mbox producer index
      RDMA/bnxt_re: Avoid calling wake_up threads from spin_lock context
      RDMA/bnxt_re: remove virt_func check while creating RoCE FW channel
      RDMA/bnxt_re: set fixed command queue depth
      RDMA/bnxt_re: Enhance the existing functions that wait for FW responses
      RDMA/bnxt_re: Avoid the command wait if firmware is inactive
      RDMA/bnxt_re: use shadow qd while posting non blocking rcfw command
      RDMA/bnxt_re: Simplify the function that sends the FW commands
      RDMA/bnxt_re: add helper function __poll_for_resp
      RDMA/bnxt_re: handle command completions after driver detect a timedout
      RDMA/bnxt_re: Add firmware stall check detection
      RDMA/bnxt_re: post destroy_ah for delayed completion of AH creation
      RDMA/bnxt_re: consider timeout of destroy ah as success.
      RDMA/bnxt_re: cancel all control path command waiters upon error
      RDMA/bnxt_re: use firmware provided max request timeout
      RDMA/bnxt_re: remove redundant cmdq_bitmap
      RDMA/bnxt_re: optimize the parameters passed to helper functions
      RDMA/bnxt_re: Remove incorrect return check from slow path
      RDMA/bnxt_re: Refactor code around bnxt_qplib_map_rc()

Leon Romanovsky (2):
      RDMA/mlx5: Reduce QP table exposure
      RDMA/bnxt_re: Initialize opcode while sending message

Li Zhijian (1):
      RDMA/rtrs: Remove duplicate cq_num assignment

Long Li (1):
      RDMA/mana_ib: Use v2 version of cfg_rx_steer_req to enable RX coalescing

Nicolas Morey (1):
      RDMA/rxe: Remove dangling declaration of rxe_cq_disable()

Patrisious Haddad (3):
      net/mlx5: Nullify qp->dbg pointer post destruction
      RDMA/mlx5: Handle DCT QP logic separately from low level QP interface
      RDMA/mlx5: Return the firmware result upon destroying QP/RQ

Selvin Xavier (8):
      RDMA/bnxt_re: Disable/kill tasklet only if it is enabled
      RDMA/bnxt_re: Use the common mmap helper functions
      RDMA/bnxt_re: Add disassociate ucontext support
      RDMA/bnxt_re: Optimize the bnxt_re_init_hwrm_hdr usage
      RDMA/bnxt_re: Query function capabilities from firmware
      RDMA/bnxt_re: Move the interface version to chip context structure
      RDMA/bnxt_re: Reorg the bar mapping
      RDMA/bnxt_re: Enable low latency push

Yang Li (2):
      RDMA/cma: Remove NULL check before dev_{put, hold}
      RDMA/bnxt_re: Remove duplicated include in bnxt_re/main.c

 drivers/infiniband/core/cma.c                     |   3 +-
 drivers/infiniband/hw/bnxt_re/bnxt_re.h           |   5 +
 drivers/infiniband/hw/bnxt_re/ib_verbs.c          | 292 +++++++++-
 drivers/infiniband/hw/bnxt_re/ib_verbs.h          |  19 +
 drivers/infiniband/hw/bnxt_re/main.c              | 192 +++++--
 drivers/infiniband/hw/bnxt_re/qplib_fp.c          | 104 ++--
 drivers/infiniband/hw/bnxt_re/qplib_fp.h          |   2 +-
 drivers/infiniband/hw/bnxt_re/qplib_rcfw.c        | 664 +++++++++++++++++-----
 drivers/infiniband/hw/bnxt_re/qplib_rcfw.h        |  52 +-
 drivers/infiniband/hw/bnxt_re/qplib_res.c         | 177 ++++--
 drivers/infiniband/hw/bnxt_re/qplib_res.h         |  34 +-
 drivers/infiniband/hw/bnxt_re/qplib_sp.c          |  19 +-
 drivers/infiniband/hw/bnxt_re/qplib_sp.h          |   5 +-
 drivers/infiniband/hw/erdma/erdma.h               |  16 +-
 drivers/infiniband/hw/erdma/erdma_hw.h            |  64 ++-
 drivers/infiniband/hw/erdma/erdma_main.c          |  53 +-
 drivers/infiniband/hw/erdma/erdma_verbs.c         | 178 +++---
 drivers/infiniband/hw/erdma/erdma_verbs.h         |  13 +-
 drivers/infiniband/hw/hfi1/ipoib_tx.c             |   4 +-
 drivers/infiniband/hw/hfi1/mmu_rb.c               | 108 ++--
 drivers/infiniband/hw/hfi1/mmu_rb.h               |  10 +-
 drivers/infiniband/hw/hfi1/sdma.c                 |  23 +-
 drivers/infiniband/hw/hfi1/sdma.h                 |  47 +-
 drivers/infiniband/hw/hfi1/sdma_txreq.h           |   2 +
 drivers/infiniband/hw/hfi1/trace_mmu.h            |  48 +-
 drivers/infiniband/hw/hfi1/user_sdma.c            | 137 ++---
 drivers/infiniband/hw/hfi1/user_sdma.h            |   1 -
 drivers/infiniband/hw/hfi1/vnic_sdma.c            |   4 +-
 drivers/infiniband/hw/hns/hns_roce_hem.c          |  51 +-
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c        |  24 +-
 drivers/infiniband/hw/irdma/uk.c                  |  10 +-
 drivers/infiniband/hw/irdma/verbs.c               |  41 +-
 drivers/infiniband/hw/mana/qp.c                   |   5 +-
 drivers/infiniband/hw/mlx5/mlx5_ib.h              |   1 +
 drivers/infiniband/hw/mlx5/qp.h                   |  12 +-
 drivers/infiniband/hw/mlx5/qpc.c                  |  93 +--
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c      |   8 -
 drivers/infiniband/sw/rxe/rxe.c                   |   9 +-
 drivers/infiniband/sw/rxe/rxe_comp.c              |   2 +-
 drivers/infiniband/sw/rxe/rxe_cq.c                |   5 +-
 drivers/infiniband/sw/rxe/rxe_loc.h               |   2 -
 drivers/infiniband/sw/rxe/rxe_mr.c                |  21 +-
 drivers/infiniband/sw/rxe/rxe_mw.c                |  22 +-
 drivers/infiniband/sw/rxe/rxe_opcode.h            |   3 +
 drivers/infiniband/sw/rxe/rxe_param.h             |   2 +-
 drivers/infiniband/sw/rxe/rxe_qp.c                |   7 +
 drivers/infiniband/sw/rxe/rxe_req.c               |   2 +-
 drivers/infiniband/sw/rxe/rxe_resp.c              |  25 +-
 drivers/infiniband/sw/rxe/rxe_task.c              | 110 ++--
 drivers/infiniband/sw/rxe/rxe_task.h              |   6 +-
 drivers/infiniband/sw/rxe/rxe_verbs.c             |  45 +-
 drivers/infiniband/sw/rxe/rxe_verbs.h             |  21 +
 drivers/infiniband/ulp/rtrs/rtrs-clt.c            |   1 -
 drivers/net/ethernet/mellanox/mlx5/core/debugfs.c |   6 +-
 drivers/net/ethernet/microsoft/mana/mana_en.c     |   5 +-
 include/linux/mlx5/driver.h                       |  10 -
 include/net/mana/mana.h                           |   4 +-
 include/uapi/rdma/bnxt_re-abi.h                   |  27 +
 58 files changed, 1955 insertions(+), 901 deletions(-)

Comments

Linus Torvalds June 30, 2023, 4:10 a.m. UTC | #1
On Wed, 28 Jun 2023 at 17:46, Jason Gunthorpe <jgg@nvidia.com> wrote:
>
> Here are the changes for RDMA for this cycle, there was a small rxe
> conflict with v6.4 that I resolved in the usual way.

Please just don't. I'd rather know about the conflicts. See

   Documentation/maintainer/rebasing-and-merging.rst

and about gazillion emails on the issue.

> - Lots of small cleanups and rework in bnxt_re
>    * Use the common mmap functions
>    * Support disassociation
>    * Improve FW command flow
>
> - bnxt_re support for "low latency push", this allows a packet

This allows a packet WHAT?

On a positive note, I see that you have a blue checkmark by google
now, and Google says

  "The sender of this email has verified that
    they own nvidia.com and the logo in the
    profile image. Learn more"

which is lovely. I expect great things from nvidia now that you
apparently own it. Congratulations! Champagne all around!

             Linus
pr-tracker-bot@kernel.org June 30, 2023, 4:16 a.m. UTC | #2
The pull request you sent on Wed, 28 Jun 2023 21:46:48 -0300:

> git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma.git tags/for-linus

has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/7ede5f78a0d74b574791c7eb0e2ca6e54b80c93c

Thank you!
Jason Gunthorpe July 3, 2023, 9:19 p.m. UTC | #3
On Thu, Jun 29, 2023 at 09:10:17PM -0700, Linus Torvalds wrote:
> On Wed, 28 Jun 2023 at 17:46, Jason Gunthorpe <jgg@nvidia.com> wrote:
> >
> > Here are the changes for RDMA for this cycle, there was a small rxe
> > conflict with v6.4 that I resolved in the usual way.
> 
> Please just don't. I'd rather know about the conflicts. See
> 
>    Documentation/maintainer/rebasing-and-merging.rst
> 
> and about gazillion emails on the issue.

OK, I thought the release merges were OK, I've sent them this way for
a long time now. Easy to change
 
> > - Lots of small cleanups and rework in bnxt_re
> >    * Use the common mmap functions
> >    * Support disassociation
> >    * Improve FW command flow
> >
> > - bnxt_re support for "low latency push", this allows a packet
> 
> This allows a packet WHAT?

Oh it got lost "a packet to be Tx'd without DMA"

> On a positive note, I see that you have a blue checkmark by google
> now, and Google says
> 
>   "The sender of this email has verified that
>     they own nvidia.com and the logo in the
>     profile image. Learn more"
> 
> which is lovely. I expect great things from nvidia now that you
> apparently own it. Congratulations! Champagne all around!

I thought only Twitter did blue checkmarks!

Thanks,
Jason