On 8/21/2017 10:29 AM, Kalderon, Michal wrote: >> From: Doug Ledford [mailto:dledford@xxxxxxxxxx] >> Sent: Friday, August 18, 2017 7:33 PM >> >> On Wed, 2017-07-26 at 14:41 +0300, Michal Kalderon wrote: >>> Hi Doug, >>> >>> This patch series adds iWARP support to our QL4xxxx networking >>> adapters. >>> >>> The first 6 patches of this series were previously sent as an RFC >>> (patches 13-18): https://www.spinics.net/lists/linux-rdma/msg51416.ht >>> ml >>> The qed part of the RFC has already been accepted as part of 4.13: >>> https://www.spinics.net/lists/netdev/msg443224.html >>> >>> Changes from RFC: >>> ----------------- >>> * Functions rdma_protocol_iwarp/roce, are used instead of >>> IS_IWARP/IS_ROCE macros when differentiating flows after device >>> registration. >>> Pointed out by Leon >>> * Additional patch is added for user space iWARP support. >>> * Additional maintainer added for qedr for MAINTAINERS. >>> >>> This series was tested and built against: >>> git://github.com/dledford/linux.git, branch: k.o/for-next >>> >>> Michal Kalderon (8): >>> RDMA/qedr: Add additional maintainer to MAINTAINERS file >>> RDMA/qedr: Rename the qedr_cm file as a preparation for iWARP >>> support >>> RDMA/qedr: Add support for registering an iWARP device >>> RDMA/qedr: Add iWARP support in existing verbs >>> RDMA/qedr: Add support for read with invalidate, supported in iWARP >>> RDMA/qedr: Add iWARP connection management qp related callbacks >>> RDMA/qedr: Add iWARP connection management functions >>> RDMA/qedr: Add support for iWARP in user space >>> >>> MAINTAINERS | 1 + >>> drivers/infiniband/hw/qedr/Makefile | 2 +- >>> drivers/infiniband/hw/qedr/main.c | 105 ++- >>> drivers/infiniband/hw/qedr/qedr.h | 31 +- >>> drivers/infiniband/hw/qedr/qedr_hsi_rdma.h | 6 +- >>> drivers/infiniband/hw/qedr/qedr_iw_cm.c | 746 >>> +++++++++++++++++++++ >>> drivers/infiniband/hw/qedr/qedr_iw_cm.h | 49 ++ >>> .../hw/qedr/{qedr_cm.c => qedr_roce_cm.c} | 2 +- >>> .../hw/qedr/{qedr_cm.h => qedr_roce_cm.h} | 0 >>> drivers/infiniband/hw/qedr/verbs.c | 350 +++++++--- >>> drivers/infiniband/hw/qedr/verbs.h | 2 + >>> 11 files changed, 1203 insertions(+), 91 deletions(-) create mode >>> 100644 drivers/infiniband/hw/qedr/qedr_iw_cm.c >>> create mode 100644 drivers/infiniband/hw/qedr/qedr_iw_cm.h >>> rename drivers/infiniband/hw/qedr/{qedr_cm.c => qedr_roce_cm.c} >>> (99%) >>> rename drivers/infiniband/hw/qedr/{qedr_cm.h => qedr_roce_cm.h} >>> (100%) >> >> I've pulled this into a branch and pushed that branch to github. It's the qedr >> branch. There was some fuzz in taking the patches just due to code float >> since you posted this. Please review my branch and make sure the final >> result is to your satisfaction. > > Thanks Doug! I looked at k.o/for-next-merged branch ( didn't see a qedr branch ) > Qedr Files in this branch look good and as expected. Yeah. That happens. I had them in their own qedr branch, but when I merged it to net-next-based, it was a straight fast-forward so no merge commit was generated. Then when I deleted the qedr branch, it's as though it never existed. > Could you please clarify which kernel version the for-next-merged branch is intended for ? 4.14? It is intended for 4.15. I have it as an early check on the future submission. > I have additional iWARP patches for net-next, would like to know which kernel version to aim them for so that they > Don't make it in before this patch series. Understood. Think of anything based on net-next as a for-next-next branch. When I open for-next in the next submission cycle, I'll ultimately create a branch on -rc2 of the next kernel, and merge this branch in. -- Doug Ledford <dledford@xxxxxxxxxx> GPG Key ID: B826A3330E572FDD Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57 2FDD
Attachment:
signature.asc
Description: OpenPGP digital signature
