| Message ID | cover.1626857976.git.leonro@nvidia.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | Separate user/kernel QP creation logic | expand |
On Wed, Jul 21, 2021 at 12:07:03PM +0300, Leon Romanovsky wrote: > From: Leon Romanovsky <leonro@nvidia.com> > > Changelog: > iv1: > * Fixed typo: incline -> inline/ > * Dropped ib_create_qp_uverbs() wrapper in favour of direct call. > * Moved kernel-doc to the actual ib_create_qp() function that users will use. > v0: https://lore.kernel.org/lkml/cover.1626846795.git.leonro@nvidia.com > > --------------------------------------------------------------------------- > Hi, > > The "QP allocation" series shows clearly how convoluted the create QP > flow and especially XRC_TGT flow, where it calls to kernel verb just > to pass some parameters as NULL to the user create QP verb. > > This series is a small step to make clean XRC_TGT flow by providing > more clean user/kernel create QP verb separation. > > It is based on the "QP allocation" series. > > Thanks > > Leon Romanovsky (7): > RDMA/mlx5: Delete not-available udata check > RDMA/core: Delete duplicated and unreachable code > RDMA/core: Remove protection from wrong in-kernel API usage > RDMA/core: Reorganize create QP low-level functions > RDMA/core: Configure selinux QP during creation > RDMA/core: Properly increment and decrement QP usecnts > RDMA/core: Create clean QP creations interface for uverbs > > drivers/infiniband/core/core_priv.h | 59 +---- > drivers/infiniband/core/uverbs_cmd.c | 31 +-- > drivers/infiniband/core/uverbs_std_types_qp.c | 29 +-- > drivers/infiniband/core/verbs.c | 208 +++++++++++------- > drivers/infiniband/hw/mlx5/qp.c | 3 - > include/rdma/ib_verbs.h | 16 +- > 6 files changed, 157 insertions(+), 189 deletions(-) Jason, Can we progress with this series too? Thanks > > -- > 2.31.1 >
On Tue, Aug 03, 2021 at 09:10:56PM +0300, Leon Romanovsky wrote: > On Wed, Jul 21, 2021 at 12:07:03PM +0300, Leon Romanovsky wrote: > > From: Leon Romanovsky <leonro@nvidia.com> > > > > Changelog: > > iv1: > > * Fixed typo: incline -> inline/ > > * Dropped ib_create_qp_uverbs() wrapper in favour of direct call. > > * Moved kernel-doc to the actual ib_create_qp() function that users will use. > > v0: https://lore.kernel.org/lkml/cover.1626846795.git.leonro@nvidia.com > > > > Hi, > > > > The "QP allocation" series shows clearly how convoluted the create QP > > flow and especially XRC_TGT flow, where it calls to kernel verb just > > to pass some parameters as NULL to the user create QP verb. > > > > This series is a small step to make clean XRC_TGT flow by providing > > more clean user/kernel create QP verb separation. > > > > It is based on the "QP allocation" series. > > > > Thanks > > > > Leon Romanovsky (7): > > RDMA/mlx5: Delete not-available udata check > > RDMA/core: Delete duplicated and unreachable code > > RDMA/core: Remove protection from wrong in-kernel API usage > > RDMA/core: Reorganize create QP low-level functions > > RDMA/core: Configure selinux QP during creation > > RDMA/core: Properly increment and decrement QP usecnts > > RDMA/core: Create clean QP creations interface for uverbs > > > > drivers/infiniband/core/core_priv.h | 59 +---- > > drivers/infiniband/core/uverbs_cmd.c | 31 +-- > > drivers/infiniband/core/uverbs_std_types_qp.c | 29 +-- > > drivers/infiniband/core/verbs.c | 208 +++++++++++------- > > drivers/infiniband/hw/mlx5/qp.c | 3 - > > include/rdma/ib_verbs.h | 16 +- > > 6 files changed, 157 insertions(+), 189 deletions(-) > > Jason, > > Can we progress with this series too? It doesn't apply, can you resend it quickly? Jason
On Tue, Aug 03, 2021 at 03:13:12PM -0300, Jason Gunthorpe wrote: > On Tue, Aug 03, 2021 at 09:10:56PM +0300, Leon Romanovsky wrote: > > On Wed, Jul 21, 2021 at 12:07:03PM +0300, Leon Romanovsky wrote: > > > From: Leon Romanovsky <leonro@nvidia.com> > > > > > > Changelog: > > > iv1: > > > * Fixed typo: incline -> inline/ > > > * Dropped ib_create_qp_uverbs() wrapper in favour of direct call. > > > * Moved kernel-doc to the actual ib_create_qp() function that users will use. > > > v0: https://lore.kernel.org/lkml/cover.1626846795.git.leonro@nvidia.com > > > > > > Hi, > > > > > > The "QP allocation" series shows clearly how convoluted the create QP > > > flow and especially XRC_TGT flow, where it calls to kernel verb just > > > to pass some parameters as NULL to the user create QP verb. > > > > > > This series is a small step to make clean XRC_TGT flow by providing > > > more clean user/kernel create QP verb separation. > > > > > > It is based on the "QP allocation" series. > > > > > > Thanks > > > > > > Leon Romanovsky (7): > > > RDMA/mlx5: Delete not-available udata check > > > RDMA/core: Delete duplicated and unreachable code > > > RDMA/core: Remove protection from wrong in-kernel API usage > > > RDMA/core: Reorganize create QP low-level functions > > > RDMA/core: Configure selinux QP during creation > > > RDMA/core: Properly increment and decrement QP usecnts > > > RDMA/core: Create clean QP creations interface for uverbs > > > > > > drivers/infiniband/core/core_priv.h | 59 +---- > > > drivers/infiniband/core/uverbs_cmd.c | 31 +-- > > > drivers/infiniband/core/uverbs_std_types_qp.c | 29 +-- > > > drivers/infiniband/core/verbs.c | 208 +++++++++++------- > > > drivers/infiniband/hw/mlx5/qp.c | 3 - > > > include/rdma/ib_verbs.h | 16 +- > > > 6 files changed, 157 insertions(+), 189 deletions(-) > > > > Jason, > > > > Can we progress with this series too? > > It doesn't apply, can you resend it quickly? Why? It is in my tree and it was on top of QP allocation patches. Thanks > > Jason
On Tue, Aug 03, 2021 at 09:17:42PM +0300, Leon Romanovsky wrote: > On Tue, Aug 03, 2021 at 03:13:12PM -0300, Jason Gunthorpe wrote: > > On Tue, Aug 03, 2021 at 09:10:56PM +0300, Leon Romanovsky wrote: > > > On Wed, Jul 21, 2021 at 12:07:03PM +0300, Leon Romanovsky wrote: > > > > From: Leon Romanovsky <leonro@nvidia.com> > > > > > > > > Changelog: > > > > iv1: > > > > * Fixed typo: incline -> inline/ > > > > * Dropped ib_create_qp_uverbs() wrapper in favour of direct call. > > > > * Moved kernel-doc to the actual ib_create_qp() function that users will use. > > > > v0: https://lore.kernel.org/lkml/cover.1626846795.git.leonro@nvidia.com > > > > > > > > Hi, > > > > > > > > The "QP allocation" series shows clearly how convoluted the create QP > > > > flow and especially XRC_TGT flow, where it calls to kernel verb just > > > > to pass some parameters as NULL to the user create QP verb. > > > > > > > > This series is a small step to make clean XRC_TGT flow by providing > > > > more clean user/kernel create QP verb separation. > > > > > > > > It is based on the "QP allocation" series. > > > > > > > > Thanks > > > > > > > > Leon Romanovsky (7): > > > > RDMA/mlx5: Delete not-available udata check > > > > RDMA/core: Delete duplicated and unreachable code > > > > RDMA/core: Remove protection from wrong in-kernel API usage > > > > RDMA/core: Reorganize create QP low-level functions > > > > RDMA/core: Configure selinux QP during creation > > > > RDMA/core: Properly increment and decrement QP usecnts > > > > RDMA/core: Create clean QP creations interface for uverbs > > > > > > > > drivers/infiniband/core/core_priv.h | 59 +---- > > > > drivers/infiniband/core/uverbs_cmd.c | 31 +-- > > > > drivers/infiniband/core/uverbs_std_types_qp.c | 29 +-- > > > > drivers/infiniband/core/verbs.c | 208 +++++++++++------- > > > > drivers/infiniband/hw/mlx5/qp.c | 3 - > > > > include/rdma/ib_verbs.h | 16 +- > > > > 6 files changed, 157 insertions(+), 189 deletions(-) > > > > > > Jason, > > > > > > Can we progress with this series too? > > > > It doesn't apply, can you resend it quickly? > > Why? > > It is in my tree and it was on top of QP allocation patches. I resent the series. Thanks > > Thanks > > > > > > Jason
On Tue, Aug 03, 2021 at 09:21:18PM +0300, Leon Romanovsky wrote: > > > > Can we progress with this series too? > > > > > > It doesn't apply, can you resend it quickly? > > > > Why? > > > > It is in my tree and it was on top of QP allocation patches. > > I resent the series. That one applied Jason
On Wed, Jul 21, 2021 at 12:07:03PM +0300, Leon Romanovsky wrote: > From: Leon Romanovsky <leonro@nvidia.com> > > Changelog: > iv1: > * Fixed typo: incline -> inline/ > * Dropped ib_create_qp_uverbs() wrapper in favour of direct call. > * Moved kernel-doc to the actual ib_create_qp() function that users will use. > v0: https://lore.kernel.org/lkml/cover.1626846795.git.leonro@nvidia.com > > Hi, > > The "QP allocation" series shows clearly how convoluted the create QP > flow and especially XRC_TGT flow, where it calls to kernel verb just > to pass some parameters as NULL to the user create QP verb. > > This series is a small step to make clean XRC_TGT flow by providing > more clean user/kernel create QP verb separation. > > It is based on the "QP allocation" series. > > Thanks > > Leon Romanovsky (7): > RDMA/mlx5: Delete not-available udata check > RDMA/core: Delete duplicated and unreachable code > RDMA/core: Remove protection from wrong in-kernel API usage > RDMA/core: Reorganize create QP low-level functions > RDMA/core: Configure selinux QP during creation > RDMA/core: Properly increment and decrement QP usecnts > RDMA/core: Create clean QP creations interface for uverbs Applied to for-next, thanks Jason
From: Leon Romanovsky <leonro@nvidia.com> Changelog: iv1: * Fixed typo: incline -> inline/ * Dropped ib_create_qp_uverbs() wrapper in favour of direct call. * Moved kernel-doc to the actual ib_create_qp() function that users will use. v0: https://lore.kernel.org/lkml/cover.1626846795.git.leonro@nvidia.com --------------------------------------------------------------------------- Hi, The "QP allocation" series shows clearly how convoluted the create QP flow and especially XRC_TGT flow, where it calls to kernel verb just to pass some parameters as NULL to the user create QP verb. This series is a small step to make clean XRC_TGT flow by providing more clean user/kernel create QP verb separation. It is based on the "QP allocation" series. Thanks Leon Romanovsky (7): RDMA/mlx5: Delete not-available udata check RDMA/core: Delete duplicated and unreachable code RDMA/core: Remove protection from wrong in-kernel API usage RDMA/core: Reorganize create QP low-level functions RDMA/core: Configure selinux QP during creation RDMA/core: Properly increment and decrement QP usecnts RDMA/core: Create clean QP creations interface for uverbs drivers/infiniband/core/core_priv.h | 59 +---- drivers/infiniband/core/uverbs_cmd.c | 31 +-- drivers/infiniband/core/uverbs_std_types_qp.c | 29 +-- drivers/infiniband/core/verbs.c | 208 +++++++++++------- drivers/infiniband/hw/mlx5/qp.c | 3 - include/rdma/ib_verbs.h | 16 +- 6 files changed, 157 insertions(+), 189 deletions(-)