On Wed, Mar 06, 2024 at 08:44:48AM +0000, Shinichiro Kawasaki wrote: > > sudo ./check nvme/045 > > > nvme/045 (Test re-authentication) [failed] > > > runtime 8.069s ... 7.639s > > > --- tests/nvme/045.out 2024-03-05 18:09:07.267668493 +0900 > > > +++ /home/shin/Blktests/blktests/results/nodev/nvme/045.out.bad 2024-03-05 18:10:07.735494384 +0900 > > > @@ -9,5 +9,6 @@ > > > Change hash to hmac(sha512) > > > Re-authenticate with changed hash > > > Renew host key on the controller and force reconnect > > > -disconnected 0 controller(s) > > > +controller "nvme1" not deleted within 5 seconds > > > +disconnected 1 controller(s) > > > Test complete > > > > That means the host either successfully reconnected or never > > disconnected. We have another test case just for the disconnect test > > (number of queue changes), so if this test passes, it must be the > > former... Shouldn't really happen, this would mean the auth code has bug. > > The test case nvme/048 passes, so this looks a bug. I'll try to recreate it. > > If you have these patches applied, the test should pass. But we might > > have still some more stuff to unify between the transports. The nvme/045 > > test passes in my setup. Though I have seen runs which were hang for > > some reason. Haven't figured out yet what's happening there. But I > > haven't seen failures. > > Still with the fix of the double-free, I observe the nvme/045 failure for rdma, > tcp and fc transports. I wonder where the difference between your system and > mine comes from. > > FYI, here I share the kernel messages for rdma transport. It shows that > nvme_rdma_reconnect_or_remove() was called repeatedly and it tried to reconnect. > The status argument is -111 or 880, so I think the recon flag is always true > and no effect. I'm interested in the status values in your environment. Do you have these patches applied: https://lore.kernel.org/linux-nvme/20240305080005.3638-1-dwagner@xxxxxxx/ ? > [ 59.117607] run blktests nvme/045 at 2024-03-06 17:05:55 > [ 59.198629] (null): rxe_set_mtu: Set mtu to 1024 > [ 59.211185] PCLMULQDQ-NI instructions are not detected. > [ 59.362952] infiniband ens3_rxe: set active > [ 59.363765] infiniband ens3_rxe: added ens3 > [ 59.540499] nvmet: adding nsid 1 to subsystem blktests-subsystem-1 > [ 59.560541] nvmet_rdma: enabling port 0 (10.0.2.15:4420) > [ 59.688866] nvmet: creating nvm controller 1 for subsystem blktests-subsystem-1 for NQN nqn.2014-08.org.nvmexpress:uuid:0f01fb42-9f7f-4856-b0b3-51e60b8de349 with DH-HMAC-CHAP. > [ 59.701114] nvme nvme1: qid 0: authenticated with hash hmac(sha256) dhgroup ffdhe2048 > [ 59.702195] nvme nvme1: qid 0: controller authenticated > [ 59.703310] nvme nvme1: qid 0: authenticated > [ 59.707478] nvme nvme1: Please enable CONFIG_NVME_MULTIPATH for full support of multi-port devices. > [ 59.709883] nvme nvme1: creating 4 I/O queues. > [ 59.745087] nvme nvme1: mapped 4/0/0 default/read/poll queues. > [ 59.786869] nvme nvme1: new ctrl: NQN "blktests-subsystem-1", addr 10.0.2.15:4420, hostnqn: nqn.2014-08.org.nvmexpress:uuid:0f01fb42-9f7f-4856-b0b3-51e60b8de349 > [ 59.999761] nvme nvme1: re-authenticating controller > [ 60.010902] nvme nvme1: qid 0: authenticated with hash hmac(sha256) dhgroup ffdhe2048 > [ 60.011640] nvme nvme1: qid 0: controller authenticated > [ 60.025652] nvme nvme1: re-authenticating controller > [ 60.035349] nvme nvme1: qid 0: authenticated with hash hmac(sha256) dhgroup ffdhe2048 > [ 60.036375] nvme nvme1: qid 0: controller authenticated > [ 60.050449] nvme nvme1: re-authenticating controller > [ 60.060757] nvme nvme1: qid 0: authenticated with hash hmac(sha256) dhgroup ffdhe2048 > [ 60.061460] nvme nvme1: qid 0: controller authenticated > [ 62.662430] nvme nvme1: re-authenticating controller > [ 62.859510] nvme nvme1: qid 0: authenticated with hash hmac(sha512) dhgroup ffdhe8192 > [ 62.860502] nvme nvme1: qid 0: controller authenticated > [ 63.029182] nvme nvme1: re-authenticating controller > [ 63.192844] nvme nvme1: qid 0: authenticated with hash hmac(sha512) dhgroup ffdhe8192 > [ 63.193900] nvme nvme1: qid 0: controller authenticated > [ 63.608561] nvme nvme1: starting error recovery > [ 63.653699] nvme nvme1: Reconnecting in 1 seconds... > [ 64.712627] nvmet: creating nvm controller 1 for subsystem blktests-subsystem-1 for NQN nqn.2014-08.org.nvmexpress:uuid:0f01fb42-9f7f-4856-b0b3-51e60b8de349 with DH-HMAC-CHAP. > [ 64.868896] nvmet: ctrl 1 qid 0 host response mismatch > [ 64.870065] nvmet: ctrl 1 qid 0 failure1 (1) > [ 64.871152] nvmet: ctrl 1 fatal error occurred! > [ 64.871519] nvme nvme1: qid 0: authentication failed > [ 64.874330] nvme nvme1: failed to connect queue: 0 ret=-111 > [ 64.878612] nvme nvme1: Failed reconnect attempt 1 > [ 64.880472] nvme nvme1: Reconnecting in 1 seconds... This looks like the DNR bit is not considered in the reconnect_or_delete function.
