From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============8467276220702466591==" MIME-Version: 1.0 From: Harris, James R Subject: Re: [SPDK] A problem with SPDK 19.01 NVMeoF/RDMA target Date: Wed, 06 Feb 2019 00:35:38 +0000 Message-ID: <48D118FF-A7D9-4EDD-B2A6-2A78D1A5B1E7@intel.com> In-Reply-To: op.zwqlezwvns8zkg@sam8 List-ID: To: spdk@lists.01.org --===============8467276220702466591== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi Valeriy, Can you provide more details on this test host module? Thanks, Jim > On Feb 5, 2019, at 4:41 AM, Valeriy Glushkov wrote: > = > Hi All! > = > It seems there is an issue in the 19.01 code that leads to a request stay= s unanswered and our test NVMeoF/RDMA host module does not receive the comp= letion. > = > From the debug log of the target: > =3D=3D=3D=3D=3D=3D=3D > rdma.c:1508:spdk_nvmf_rdma_request_process: *DEBUG*: Request 0x1402a20 en= tering state 1 > rdma.c:1508:spdk_nvmf_rdma_request_process: *DEBUG*: Request 0x1402a20 en= tering state 2 > rdma.c:1414:spdk_nvmf_rdma_request_parse_sgl: *DEBUG*: Request 0x1402a20 = took 1 buffer/s from central pool > rdma.c:1508:spdk_nvmf_rdma_request_process: *DEBUG*: Request 0x1402a20 en= tering state 5 > request.c: 121:nvmf_trace_command: *DEBUG*: Admin cmd: opc 0x02 fuse 0 ci= d 2751 nsid 4294967295 cdw10 0x001e0002 > request.c: 127:nvmf_trace_command: *DEBUG*: psdt 0 > request.c: 136:nvmf_trace_command: *DEBUG*: SGL: Keyed (Inv): addr 0xffff= b28f67223dd0 key 0x8294c len 0x200 > ctrlr.c:1285:spdk_nvmf_ctrlr_get_log_page: *DEBUG*: Get log page: LID=3D0= x02 offset=3D0x0 len=3D0x7c > request.c: 92:spdk_nvmf_request_complete: *DEBUG*: cpl: cid=3D2751 cdw0= =3D0x00000000 rsvd1=3D0 status=3D0x0000 > rdma.c:1508:spdk_nvmf_rdma_request_process: *DEBUG*: Request 0x1402a20 en= tering state 7 > rdma.c:1508:spdk_nvmf_rdma_request_process: *DEBUG*: Request 0x1402a20 en= tering state 8 > rdma.c:1508:spdk_nvmf_rdma_request_process: *DEBUG*: Request 0x1402a20 en= tering state 8 > =3D=3D=3D=3D=3D=3D=3D=3D > = > The same host module works well with the SPDK 18.10.1 NVMf target. > = > Is this a bug in the recent SPDK code? > = > -- > Best regards, > Valeriy Glushkov > www.starwind.com > valeriy.glushkov(a)starwind.com > = > = > = > _______________________________________________ > SPDK mailing list > SPDK(a)lists.01.org > https://lists.01.org/mailman/listinfo/spdk --===============8467276220702466591==--