Commit 14a18cca authored by Changpeng Liu's avatar Changpeng Liu Committed by Jim Harris
Browse files

nvmf: don't submit FLUSH command to backend which can't support it 



As for NVMeoF controller, SPDK always set volatile write cache to
present, so OS may submit FLUSH command to backend block device,
while here, return success for those block devices which can't
support FLUSH command.

Change-Id: Ib764503f1986e7e0a38a1751a91d6c11e1b78865
Signed-off-by: default avatarChangpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/402653


Tested-by: default avatarSPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: default avatarDaniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: default avatarJim Harris <james.r.harris@intel.com>
parent faf87b0c

lib/nvmf/ctrlr_bdev.c

+10 −0
Original line number Diff line number Diff line
@@ -264,6 +264,16 @@ nvmf_bdev_ctrlr_flush_cmd(struct spdk_bdev *bdev, struct spdk_bdev_desc *desc, 
{

	struct spdk_nvme_cpl *response = &req->rsp->nvme_cpl;



	/* As for NVMeoF controller, SPDK always set volatile write

	 * cache bit to 1, return success for those block devices

	 * which can't support FLUSH command.

	 */

	if (!spdk_bdev_io_type_supported(bdev, SPDK_BDEV_IO_TYPE_FLUSH)) {

		response->status.sct = SPDK_NVME_SCT_GENERIC;

		response->status.sc = SPDK_NVME_SC_SUCCESS;

		return SPDK_NVMF_REQUEST_EXEC_STATUS_COMPLETE;

	}



	if (spdk_bdev_flush_blocks(desc, ch, 0, spdk_bdev_get_num_blocks(bdev),

				   nvmf_bdev_ctrlr_complete_cmd, req)) {

		response->status.sc = SPDK_NVME_SC_INTERNAL_DEVICE_ERROR;