+1
−0
+3
−1
+4
−0
+11
−4
+3
−1
Loading
RDMA transport uses a pool of WRs to handle multi SGL payload. This pool is common for the entire transport, but its size is quite small. In multithread config the default size might not be enough, this patch gives the user possibility to change the size. Default value of the pool size is changed to 4095 Signed-off-by:Alexey Marchuk <alexeymar@nvidia.com> Signed-off-by:
Allen Zhu <allenzhu@nvidia.com> Change-Id: I0d5e4d20dc2b57f548c53967c61a738b36c72cd7 Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/22833 Reviewed-by:
Jim Harris <jim.harris@samsung.com> Reviewed-by:
Shuhei Matsumoto <smatsumoto@nvidia.com> Community-CI: Mellanox Build Bot Tested-by:
SPDK CI Jenkins <sys_sgci@intel.com>