+1
−1
+5
−2
+11
−0
Loading
We already have u.bdev.accel_sequence and internal.accel_sequence, but both of them can be cleared in different cases. The former is cleared when a bdev doesn't support accel or by the module itself to notify bdev layer that it executed the sequence. The latter is cleared before submitting a request to a module supporting accel, so that bdev layer won't touch it again. However, there are cases where it's necessary to know whether an IO was associated with an accel sequence and a module has executed it (e.g. to allow modules support accel without supporting memory domains), so an extra flag is required. Signed-off-by:Konrad Sztyber <konrad.sztyber@intel.com> Change-Id: I58ec789cce84b63151d02e28a0bb3d3bbf85f7bf Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/18771 Community-CI: Mellanox Build Bot Reviewed-by:
Aleksey Marchuk <alexeymar@nvidia.com> Tested-by:
SPDK CI Jenkins <sys_sgci@intel.com> Reviewed-by:
Shuhei Matsumoto <smatsumoto@nvidia.com>