On Thu, Jan 09, 2020 at 08:20:12PM +0800, Weihang Li wrote: > From: Xi Wang <wangxi11@xxxxxxxxxx> > > When hardware is in resetting stage, we may can't poll back all the > expected work completions as the hardware won't generate cqe anymore. > > This patch allows the driver to compose the expected wc instead of the > hardware during resetting stage. Once the hardware finished resetting, we > can poll cq from hardware again. > > Signed-off-by: Xi Wang <wangxi11@xxxxxxxxxx> > Signed-off-by: Weihang Li <liweihang@xxxxxxxxxx> > --- > Sorry to keep sending new version of this patch. It was because some issues > were found since last version or there are some conflicts to apply it to > the branch. > > Considering that this patch hasn't got comments from any others in the > community, please forget about previous versions and treat this as a new > one. Any suggestions will be appreciated. > > Changes since v3: > - Fix conflicts with recently applied patches. > > Changes since v2: > - Fix cq poll failure for qp1 when device reseting and do a rebase. > > Changes since v1: > - Remove a deplicate cq_clean statement. > > drivers/infiniband/hw/hns/hns_roce_cq.c | 2 + > drivers/infiniband/hw/hns/hns_roce_device.h | 17 +++ > drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 14 ++- > drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 158 +++++++++++++++++++++++----- > drivers/infiniband/hw/hns/hns_roce_main.c | 47 +++++++++ > drivers/infiniband/hw/hns/hns_roce_qp.c | 48 ++++++++- > 6 files changed, 252 insertions(+), 34 deletions(-) Applied to for-next, thanks Jason
