+11
−2
Loading
When running in interrupt mode, NVMeoF TCP doesn't have a chance to flush PDU again if partial flush happened, so here return correct errno to let NVMeoF TCP doing more retries to flush one PDU completely. Even for a full flush request in interrupt mode, the `sock->queued_reqs` list maybe not empty, we let the caller to do more retries for this case too. Fix issue #3697. Change-Id: I01edf17a2b31b67f41f170c8a479a940294ac4c5 Signed-off-by:Changpeng Liu <changpeliu@tencent.com> Reviewed-on: https://review.spdk.io/c/spdk/spdk/+/26263 Tested-by:
SPDK Automated Test System <spdkbot@gmail.com> Reviewed-by:
Konrad Sztyber <ksztyber@nvidia.com> Community-CI: Mellanox Build Bot Reviewed-by:
Jim Harris <jim.harris@nvidia.com> Reviewed-by:
GangCao <gang.cao@intel.com>