On 8/22/23 10:07 AM, Dennis Dalessandro wrote: > Much scaled down version of Brendan's previous patches. This doesn't touch uAPI > just refactors some code. > > Small fixup from Doug. > > Would have sent a couple weeks ago but had been dealing with the isert > regression. Reverting that give us a clean bill of health. If too late > for 6.6 can wait another cycle even. > > --- > > Brendan Cunningham (1): > RDMA/hfi1: Move user SDMA system memory pinning code to its own file > > Douglas Miller (1): > IB/hfi1: Reduce printing of errors during driver shut down > > > drivers/infiniband/hw/hfi1/Makefile | 1 + > drivers/infiniband/hw/hfi1/chip.c | 8 +- > drivers/infiniband/hw/hfi1/hfi.h | 4 +- > drivers/infiniband/hw/hfi1/pin_system.c | 474 ++++++++++++++++++++++++ > drivers/infiniband/hw/hfi1/pinning.h | 20 + > drivers/infiniband/hw/hfi1/user_sdma.c | 441 +--------------------- > drivers/infiniband/hw/hfi1/user_sdma.h | 17 +- > 7 files changed, 510 insertions(+), 455 deletions(-) > create mode 100644 drivers/infiniband/hw/hfi1/pin_system.c > create mode 100644 drivers/infiniband/hw/hfi1/pinning.h > > -- > -Denny > My bad on the subject line here. Should be obvious what it's for though, small fix and a refactor patch for hfi1. -Denny
