+71
−20
Loading
Current implementation performs poorly on big number of threads, because it needs to iterate through all of them to find the ones with channels opened to current bdev. This implementation stores such threads in RB-tree, dramaticaly increasing performance. This is important, because spdk_for_each_channel requires g_devlist_mutex locked. Change-Id: If73cc279c87a70cc3b7ae915d0863e6c81a1f83d Signed-off-by:Vladislav Fedyaev <a37206@gmail.com> Reviewed-on: https://review.spdk.io/c/spdk/spdk/+/26121 Tested-by:
SPDK Automated Test System <spdkbot@gmail.com> Reviewed-by:
Shuhei Matsumoto <smatsumoto@nvidia.com> Reviewed-by:
Jim Harris <jim.harris@nvidia.com> Community-CI: Mellanox Build Bot Reviewed-by:
Vasilii Ivanov <iwanovvvasilij@gmail.com>