On 6/8/2016 6:41 AM, Leon Romanovsky wrote: > On Wed, Jun 08, 2016 at 02:52:10AM -0700, Nicholas A. Bellinger wrote: >> Hey Moni & HCH, Co. >> >> (Adding Sagi CC') >> >> On Wed, 2016-06-08 at 11:49 +0300, Moni Shoua wrote: >>> On Fri, Jun 3, 2016 at 8:25 AM, Christoph Hellwig wrote: >>>> Do you have a changelog compred to the last version? >>>> >>> >>> - Moved from drivers/staging/rdma/rxe to drivers/infiniband/hw/rxe >>> - Rebase to newer kernel (was 4.4-rc6) >>> - Removed support in phys. MR >>> - Removed memory window codes (buggy) >>> - Mind IS_ENABLED(CONFIG_IPV6) >>> - Dispatch IB event on port UP/DOWN >>> - Validate ICRC in receive >>> - Validate dgid in receive >>> - Scatter IPv4 header to right place for UD QP >>> - Other bug fixes >>> - Copyright dates were updated in file headers >>> >>> >>>> How well do FRWRs work, does this include Sagi's work in that area? >>> FRWR is not supported yet >>>> How well does posting of fenced WRs work? >>> We ran successfully some simple tests with IBV_SEND_FENCE >>> >> >> Great. >> >> A Soft-RoCE environment against v4.7-rc2 for iser and early nvme-fabrics >> code would be extremely useful. :-) >> >> Can we get this code into 0-day build ASAP..? > > Hi Nicholas, > > Per your request, I added Soft-RoCE code to my tree for testing (kbuild > and linux-next tests). Please be aware that official RDMA tree is Doug's > tree and only appearance in his tree will be sign for acceptance. > > https://git.kernel.org/cgit/linux/kernel/git/leon/linux-rdma.git/log/?h=topic/rxe > https://git.kernel.org/cgit/linux/kernel/git/leon/linux-rdma.git/log/?h=testing/rdma-next I have a branch in my tree with this code in it already. This isn't a statement of inclusion, as I still have to finish the review, but it's on the rxe branch and I haven't seen any 0-day failure so far (it won't undergo linux-next until it officially hits my next branch).