mbox series

[for-next,0/9] IB/hfi1: Updates for 5.3

Message ID 20190614162724.44714.22604.stgit@awfm-01.aw.intel.com (mailing list archive)
Headers show
Series IB/hfi1: Updates for 5.3 | expand

Message

Dennis Dalessandro June 14, 2019, 4:27 p.m. UTC
This is the first round of fixes and updates for 5.3. There are some bugs but
nothing that really warranted making its way to RC. So this can all go to
next. Of note there is a patch for a fix Jason requested.

---

Dennis Dalessandro (1):
      IB/hfi1: No need to use try_module_get for debugfs

Kamenee Arumugam (1):
      IB/{hfi1,qib,rdmavt}: Put qp in error state when cq is full

Michael J. Ruhl (4):
      IB/rdmavt: Set QP allowed opcodes after QP allocation
      IB/{rdmavt, hfi1, qib}: Remove AH refcount for UD QPs
      IB/{rdmavt, hfi1, qib}: Add helpers to hide SWQE WR details
      IB/hfi1: Reduce excessive aspm inlines

Mike Marciniszyn (3):
      IB/hfi1: Add missing INVALIDATE opcodes for trace
      IB/rdmavt: Enhance trace information for FRWR debug
      IB/rdmavt: Add trace for map_mr_sg


 drivers/infiniband/hw/hfi1/Makefile       |    1 
 drivers/infiniband/hw/hfi1/aspm.c         |  270 +++++++++++++++++++++++++++++
 drivers/infiniband/hw/hfi1/aspm.h         |  262 +---------------------------
 drivers/infiniband/hw/hfi1/debugfs.c      |    5 -
 drivers/infiniband/hw/hfi1/pcie.c         |    6 -
 drivers/infiniband/hw/hfi1/qp.c           |    4 
 drivers/infiniband/hw/hfi1/rc.c           |    3 
 drivers/infiniband/hw/hfi1/trace_ibhdrs.h |    2 
 drivers/infiniband/hw/hfi1/uc.c           |    3 
 drivers/infiniband/hw/hfi1/ud.c           |   36 ++--
 drivers/infiniband/hw/qib/qib_qp.c        |    4 
 drivers/infiniband/hw/qib/qib_rc.c        |    3 
 drivers/infiniband/hw/qib/qib_uc.c        |    3 
 drivers/infiniband/hw/qib/qib_ud.c        |   28 ++-
 drivers/infiniband/sw/rdmavt/ah.c         |    6 -
 drivers/infiniband/sw/rdmavt/cq.c         |   15 +-
 drivers/infiniband/sw/rdmavt/mr.c         |    3 
 drivers/infiniband/sw/rdmavt/qp.c         |   96 +++++++---
 drivers/infiniband/sw/rdmavt/trace_mr.h   |   56 ++++++
 drivers/infiniband/sw/rdmavt/vt.h         |    9 +
 include/rdma/rdma_vt.h                    |    3 
 include/rdma/rdmavt_cq.h                  |    3 
 include/rdma/rdmavt_qp.h                  |  119 ++++++++++++-
 23 files changed, 584 insertions(+), 356 deletions(-)
 create mode 100644 drivers/infiniband/hw/hfi1/aspm.c

--
-Denny

Comments

Dennis Dalessandro June 14, 2019, 4:55 p.m. UTC | #1
On 6/14/2019 12:27 PM, Dennis Dalessandro wrote:
> This is the first round of fixes and updates for 5.3. There are some bugs but
> nothing that really warranted making its way to RC. So this can all go to
> next. Of note there is a patch for a fix Jason requested.
> 
> ---
> 
> Dennis Dalessandro (1):
>        IB/hfi1: No need to use try_module_get for debugfs
> 
> Kamenee Arumugam (1):
>        IB/{hfi1,qib,rdmavt}: Put qp in error state when cq is full
> 
> Michael J. Ruhl (4):
>        IB/rdmavt: Set QP allowed opcodes after QP allocation
>        IB/{rdmavt, hfi1, qib}: Remove AH refcount for UD QPs
>        IB/{rdmavt, hfi1, qib}: Add helpers to hide SWQE WR details
>        IB/hfi1: Reduce excessive aspm inlines
> 
> Mike Marciniszyn (3):
>        IB/hfi1: Add missing INVALIDATE opcodes for trace
>        IB/rdmavt: Enhance trace information for FRWR debug
>        IB/rdmavt: Add trace for map_mr_sg
> 
> 
>   drivers/infiniband/hw/hfi1/Makefile       |    1
>   drivers/infiniband/hw/hfi1/aspm.c         |  270 +++++++++++++++++++++++++++++
>   drivers/infiniband/hw/hfi1/aspm.h         |  262 +---------------------------
>   drivers/infiniband/hw/hfi1/debugfs.c      |    5 -
>   drivers/infiniband/hw/hfi1/pcie.c         |    6 -
>   drivers/infiniband/hw/hfi1/qp.c           |    4
>   drivers/infiniband/hw/hfi1/rc.c           |    3
>   drivers/infiniband/hw/hfi1/trace_ibhdrs.h |    2
>   drivers/infiniband/hw/hfi1/uc.c           |    3
>   drivers/infiniband/hw/hfi1/ud.c           |   36 ++--
>   drivers/infiniband/hw/qib/qib_qp.c        |    4
>   drivers/infiniband/hw/qib/qib_rc.c        |    3
>   drivers/infiniband/hw/qib/qib_uc.c        |    3
>   drivers/infiniband/hw/qib/qib_ud.c        |   28 ++-
>   drivers/infiniband/sw/rdmavt/ah.c         |    6 -
>   drivers/infiniband/sw/rdmavt/cq.c         |   15 +-
>   drivers/infiniband/sw/rdmavt/mr.c         |    3
>   drivers/infiniband/sw/rdmavt/qp.c         |   96 +++++++---
>   drivers/infiniband/sw/rdmavt/trace_mr.h   |   56 ++++++
>   drivers/infiniband/sw/rdmavt/vt.h         |    9 +
>   include/rdma/rdma_vt.h                    |    3
>   include/rdma/rdmavt_cq.h                  |    3
>   include/rdma/rdmavt_qp.h                  |  119 ++++++++++++-
>   23 files changed, 584 insertions(+), 356 deletions(-)
>   create mode 100644 drivers/infiniband/hw/hfi1/aspm.c

Note this needs the Completion rework patches to avoid conflicts.

https://patchwork.kernel.org/patch/10992251/
https://patchwork.kernel.org/patch/10992247/

-Denny
Jason Gunthorpe June 25, 2019, 5:41 p.m. UTC | #2
On Fri, Jun 14, 2019 at 12:27:56PM -0400, Dennis Dalessandro wrote:
> This is the first round of fixes and updates for 5.3. There are some bugs but
> nothing that really warranted making its way to RC. So this can all go to
> next. Of note there is a patch for a fix Jason requested.
> 
> 
> Dennis Dalessandro (1):
>       IB/hfi1: No need to use try_module_get for debugfs
> 
> Kamenee Arumugam (1):
>       IB/{hfi1,qib,rdmavt}: Put qp in error state when cq is full
> 
> Michael J. Ruhl (4):
>       IB/rdmavt: Set QP allowed opcodes after QP allocation
>       IB/{rdmavt, hfi1, qib}: Remove AH refcount for UD QPs
>       IB/{rdmavt, hfi1, qib}: Add helpers to hide SWQE WR details
>       IB/hfi1: Reduce excessive aspm inlines
> 
> Mike Marciniszyn (3):
>       IB/hfi1: Add missing INVALIDATE opcodes for trace
>       IB/rdmavt: Enhance trace information for FRWR debug
>       IB/rdmavt: Add trace for map_mr_sg

This also has rebasing problems, please sent it on top of the prior
series, which is on top of for-testing.

Thanks,
Jason