| Message ID | 20210209131107.698833-1-leon@kernel.org (mailing list archive) |
|---|---|
| Headers | show |
| Series | Real time/free running timestamp support | expand |
On Tue, Feb 09, 2021 at 03:11:05PM +0200, Leon Romanovsky wrote: > From: Leon Romanovsky <leonro@nvidia.com> > > Add an extra timestamp format for mlx5_ib device. > > Thanks > > Aharon Landau (2): > net/mlx5: Add new timestamp mode bits > RDMA/mlx5: Fail QP creation if the device can not support the CQE TS > > drivers/infiniband/hw/mlx5/qp.c | 104 +++++++++++++++++++++++++++++--- > include/linux/mlx5/mlx5_ifc.h | 54 +++++++++++++++-- > 2 files changed, 145 insertions(+), 13 deletions(-) Since this is a rdma series, and we are at the end of the cycle, I took the IFC file directly to the rdma tree instead of through the shared branch. Applied to for-next, thanks Jason
On Fri, 2021-02-12 at 14:10 -0400, Jason Gunthorpe wrote: > On Tue, Feb 09, 2021 at 03:11:05PM +0200, Leon Romanovsky wrote: > > From: Leon Romanovsky <leonro@nvidia.com> > > > > Add an extra timestamp format for mlx5_ib device. > > > > Thanks > > > > Aharon Landau (2): > > net/mlx5: Add new timestamp mode bits > > RDMA/mlx5: Fail QP creation if the device can not support the CQE > > TS > > > > drivers/infiniband/hw/mlx5/qp.c | 104 > > +++++++++++++++++++++++++++++--- > > include/linux/mlx5/mlx5_ifc.h | 54 +++++++++++++++-- > > 2 files changed, 145 insertions(+), 13 deletions(-) > > Since this is a rdma series, and we are at the end of the cycle, I > took the IFC file directly to the rdma tree instead of through the > shared branch. > > Applied to for-next, thanks > mmm, i was planing to resubmit this patch with the netdev real time support series, since the uplink representor is getting delayed, I thought I could submit the real time stuff today. can you wait on the ifc patch, i will re-send it today if you will, but it must go through the shared branch Thanks.
On Fri, Feb 12, 2021 at 01:09:20PM -0800, Saeed Mahameed wrote: > On Fri, 2021-02-12 at 14:10 -0400, Jason Gunthorpe wrote: > > On Tue, Feb 09, 2021 at 03:11:05PM +0200, Leon Romanovsky wrote: > > > From: Leon Romanovsky <leonro@nvidia.com> > > > > > > Add an extra timestamp format for mlx5_ib device. > > > > > > Thanks > > > > > > Aharon Landau (2): > > > net/mlx5: Add new timestamp mode bits > > > RDMA/mlx5: Fail QP creation if the device can not support the CQE > > > TS > > > > > > drivers/infiniband/hw/mlx5/qp.c | 104 > > > +++++++++++++++++++++++++++++--- > > > include/linux/mlx5/mlx5_ifc.h | 54 +++++++++++++++-- > > > 2 files changed, 145 insertions(+), 13 deletions(-) > > > > Since this is a rdma series, and we are at the end of the cycle, I > > took the IFC file directly to the rdma tree instead of through the > > shared branch. > > > > Applied to for-next, thanks > > > > mmm, i was planing to resubmit this patch with the netdev real time > support series, since the uplink representor is getting delayed, I > thought I could submit the real time stuff today. can you wait on the > ifc patch, i will re-send it today if you will, but it must go through > the shared branch Friday of rc7 is a bit late to be sending new patches for the first time, isn't it?? But sure, if you update the shared branch right now I'll fix up rdma.git Jason
On Fri, 2021-02-12 at 17:14 -0400, Jason Gunthorpe wrote: > On Fri, Feb 12, 2021 at 01:09:20PM -0800, Saeed Mahameed wrote: > > On Fri, 2021-02-12 at 14:10 -0400, Jason Gunthorpe wrote: > > > On Tue, Feb 09, 2021 at 03:11:05PM +0200, Leon Romanovsky wrote: > > > > From: Leon Romanovsky <leonro@nvidia.com> > > > > > > > > Add an extra timestamp format for mlx5_ib device. > > > > > > > > Thanks > > > > > > > > Aharon Landau (2): > > > > net/mlx5: Add new timestamp mode bits > > > > RDMA/mlx5: Fail QP creation if the device can not support the > > > > CQE > > > > TS > > > > > > > > drivers/infiniband/hw/mlx5/qp.c | 104 > > > > +++++++++++++++++++++++++++++--- > > > > include/linux/mlx5/mlx5_ifc.h | 54 +++++++++++++++-- > > > > 2 files changed, 145 insertions(+), 13 deletions(-) > > > > > > Since this is a rdma series, and we are at the end of the cycle, > > > I > > > took the IFC file directly to the rdma tree instead of through > > > the > > > shared branch. > > > > > > Applied to for-next, thanks > > > > > > > mmm, i was planing to resubmit this patch with the netdev real time > > support series, since the uplink representor is getting delayed, I > > thought I could submit the real time stuff today. can you wait on > > the > > ifc patch, i will re-send it today if you will, but it must go > > through > > the shared branch > > Friday of rc7 is a bit late to be sending new patches for the first > time, isn't it?? I know, uplink representor last minute mess ! > > But sure, if you update the shared branch right now I'll fix up > rdma.git > I can't put it in the shared brach without review, i will post it to the netdev/rdma lists for two days at least for review and feedback. Not critical though, this can wait for next rc1.. if you want to proceed with what you already did. Thanks
On Fri, Feb 12, 2021 at 01:19:09PM -0800, Saeed Mahameed wrote: > On Fri, 2021-02-12 at 17:14 -0400, Jason Gunthorpe wrote: > > On Fri, Feb 12, 2021 at 01:09:20PM -0800, Saeed Mahameed wrote: > > > On Fri, 2021-02-12 at 14:10 -0400, Jason Gunthorpe wrote: > > > > On Tue, Feb 09, 2021 at 03:11:05PM +0200, Leon Romanovsky wrote: > > > > > From: Leon Romanovsky <leonro@nvidia.com> > > > > > > > > > > Add an extra timestamp format for mlx5_ib device. > > > > > > > > > > Thanks > > > > > > > > > > Aharon Landau (2): > > > > > net/mlx5: Add new timestamp mode bits > > > > > RDMA/mlx5: Fail QP creation if the device can not support the > > > > > CQE > > > > > TS > > > > > > > > > > drivers/infiniband/hw/mlx5/qp.c | 104 > > > > > +++++++++++++++++++++++++++++--- > > > > > include/linux/mlx5/mlx5_ifc.h | 54 +++++++++++++++-- > > > > > 2 files changed, 145 insertions(+), 13 deletions(-) > > > > > > > > Since this is a rdma series, and we are at the end of the cycle, > > > > I > > > > took the IFC file directly to the rdma tree instead of through > > > > the > > > > shared branch. > > > > > > > > Applied to for-next, thanks > > > > > > > > > > mmm, i was planing to resubmit this patch with the netdev real time > > > support series, since the uplink representor is getting delayed, I > > > thought I could submit the real time stuff today. can you wait on > > > the > > > ifc patch, i will re-send it today if you will, but it must go > > > through > > > the shared branch > > > > Friday of rc7 is a bit late to be sending new patches for the first > > time, isn't it?? > > I know, uplink representor last minute mess ! > > > > > But sure, if you update the shared branch right now I'll fix up > > rdma.git > > > > I can't put it in the shared brach without review, i will post it to > the netdev/rdma lists for two days at least for review and feedback. Well, I'm not going to take any different patches beyond right now unless Linus does a rc8?? Just move this one IFC patch to the shared branch, it is obviously OK Jason
On Fri, Feb 12, 2021 at 05:21:53PM -0400, Jason Gunthorpe wrote: > On Fri, Feb 12, 2021 at 01:19:09PM -0800, Saeed Mahameed wrote: > > On Fri, 2021-02-12 at 17:14 -0400, Jason Gunthorpe wrote: > > > On Fri, Feb 12, 2021 at 01:09:20PM -0800, Saeed Mahameed wrote: > > > > On Fri, 2021-02-12 at 14:10 -0400, Jason Gunthorpe wrote: > > > > > On Tue, Feb 09, 2021 at 03:11:05PM +0200, Leon Romanovsky wrote: > > > > > > From: Leon Romanovsky <leonro@nvidia.com> > > > > > > > > > > > > Add an extra timestamp format for mlx5_ib device. > > > > > > > > > > > > Thanks > > > > > > > > > > > > Aharon Landau (2): > > > > > > net/mlx5: Add new timestamp mode bits > > > > > > RDMA/mlx5: Fail QP creation if the device can not support the > > > > > > CQE > > > > > > TS > > > > > > > > > > > > drivers/infiniband/hw/mlx5/qp.c | 104 > > > > > > +++++++++++++++++++++++++++++--- > > > > > > include/linux/mlx5/mlx5_ifc.h | 54 +++++++++++++++-- > > > > > > 2 files changed, 145 insertions(+), 13 deletions(-) > > > > > > > > > > Since this is a rdma series, and we are at the end of the cycle, > > > > > I > > > > > took the IFC file directly to the rdma tree instead of through > > > > > the > > > > > shared branch. > > > > > > > > > > Applied to for-next, thanks > > > > > > > > > > > > > mmm, i was planing to resubmit this patch with the netdev real time > > > > support series, since the uplink representor is getting delayed, I > > > > thought I could submit the real time stuff today. can you wait on > > > > the > > > > ifc patch, i will re-send it today if you will, but it must go > > > > through > > > > the shared branch > > > > > > Friday of rc7 is a bit late to be sending new patches for the first > > > time, isn't it?? > > > > I know, uplink representor last minute mess ! > > > > > > > > But sure, if you update the shared branch right now I'll fix up > > > rdma.git > > > > > > > I can't put it in the shared brach without review, i will post it to > > the netdev/rdma lists for two days at least for review and feedback. > > Well, I'm not going to take any different patches beyond right now > unless Linus does a rc8?? > > Just move this one IFC patch to the shared branch, it is obviously OK OK, I'm curious to see the end result of all this last minute adventure. Thanks > > Jason
On Sun, Feb 14, 2021 at 08:41:03AM +0200, Leon Romanovsky wrote: > On Fri, Feb 12, 2021 at 05:21:53PM -0400, Jason Gunthorpe wrote: > > On Fri, Feb 12, 2021 at 01:19:09PM -0800, Saeed Mahameed wrote: > > > On Fri, 2021-02-12 at 17:14 -0400, Jason Gunthorpe wrote: > > > > On Fri, Feb 12, 2021 at 01:09:20PM -0800, Saeed Mahameed wrote: > > > > > On Fri, 2021-02-12 at 14:10 -0400, Jason Gunthorpe wrote: > > > > > > On Tue, Feb 09, 2021 at 03:11:05PM +0200, Leon Romanovsky wrote: > > > > > > > From: Leon Romanovsky <leonro@nvidia.com> > > > > > > > > > > > > > > Add an extra timestamp format for mlx5_ib device. > > > > > > > > > > > > > > Thanks > > > > > > > > > > > > > > Aharon Landau (2): > > > > > > > net/mlx5: Add new timestamp mode bits > > > > > > > RDMA/mlx5: Fail QP creation if the device can not support the > > > > > > > CQE > > > > > > > TS > > > > > > > > > > > > > > drivers/infiniband/hw/mlx5/qp.c | 104 > > > > > > > +++++++++++++++++++++++++++++--- > > > > > > > include/linux/mlx5/mlx5_ifc.h | 54 +++++++++++++++-- > > > > > > > 2 files changed, 145 insertions(+), 13 deletions(-) > > > > > > > > > > > > Since this is a rdma series, and we are at the end of the cycle, > > > > > > I > > > > > > took the IFC file directly to the rdma tree instead of through > > > > > > the > > > > > > shared branch. > > > > > > > > > > > > Applied to for-next, thanks > > > > > > > > > > > > > > > > mmm, i was planing to resubmit this patch with the netdev real time > > > > > support series, since the uplink representor is getting delayed, I > > > > > thought I could submit the real time stuff today. can you wait on > > > > > the > > > > > ifc patch, i will re-send it today if you will, but it must go > > > > > through > > > > > the shared branch > > > > > > > > Friday of rc7 is a bit late to be sending new patches for the first > > > > time, isn't it?? > > > > > > I know, uplink representor last minute mess ! > > > > > > > > > > > But sure, if you update the shared branch right now I'll fix up > > > > rdma.git > > > > > > > > > > I can't put it in the shared brach without review, i will post it to > > > the netdev/rdma lists for two days at least for review and feedback. > > > > Well, I'm not going to take any different patches beyond right now > > unless Linus does a rc8?? > > > > Just move this one IFC patch to the shared branch, it is obviously OK > > OK, I'm curious to see the end result of all this last minute adventure. Jason, I took first patch to the shared branch. a6a217dddcd5 net/mlx5: Add new timestamp mode bits Thanks > > Thanks > > > > > Jason
From: Leon Romanovsky <leonro@nvidia.com> Add an extra timestamp format for mlx5_ib device. Thanks Aharon Landau (2): net/mlx5: Add new timestamp mode bits RDMA/mlx5: Fail QP creation if the device can not support the CQE TS drivers/infiniband/hw/mlx5/qp.c | 104 +++++++++++++++++++++++++++++--- include/linux/mlx5/mlx5_ifc.h | 54 +++++++++++++++-- 2 files changed, 145 insertions(+), 13 deletions(-) -- 2.29.2