[PATCHv14 00/11] nvme: In-band authentication support
Chaitanya Kulkarni
chaitanyak at nvidia.com
Fri Jun 10 02:28:55 PDT 2022
On 6/10/22 02:12, Hannes Reinecke wrote:
> On 6/10/22 10:55, Chaitanya Kulkarni wrote:
>>
>>>>
>>>> blktests (master) # ./check nvme/039
>>>> nvme/039 (Create authenticated connections) [failed]
>>>> runtime 1.400s ... 1.707s
>>>> --- tests/nvme/039.out 2022-06-08 18:09:06.239931529 -0700
>>>> +++ /mnt/data/blktests/results/nodev/nvme/039.out.bad
>>>> 2022-06-08
>>>> 18:09:40.596663692 -0700
>>>> @@ -1,6 +1,7 @@
>>>> Running nvme/039
>>>> +tests/nvme/rc: line 269: printf: write error: Invalid argument
>>>> Test unauthenticated connection
>>>> -no controller found
>>>> +no controller found: failed to write to nvme-fabrics device
>>>> NQN:blktests-subsystem-1 disconnected 0 controller(s)
>>>> Test authenticated connection
>>>> ...
>>>> (Run 'diff -u tests/nvme/039.out >
>>>> /mnt/data/blktests/results/nodev/nvme/039.out.bad' to see the entire
>>> diff)
>>>
>>> Hmm. Not sure what has happened here, but all blktests worked on my
>>> testbed. I'll be rechecking with the latest nvme-cli build.
>>>
>>> Which nvme-cli version did you use?
>>>
>>
>> ~ # nvme --version
>> nvme version 2.0 (git 2.0-96-g0b286b1)
>> libnvme version 1.0 (git 1.0-62-gaacaeea)
>> ~ #
>>
>> I'm I using wrong version ?
>>
>> Can we please add check in the blktests to skip the test
>> if we have wrong wrong nvme-cli version ? it will help
>> others to not misdiagnosed the problem...
>>
> Hmm. No, that version looks okay.
>
> What's the kernel message log? Does it say something about 'dev_loss_tmo' ?
>
> If so, can you try with 'export nvme_trtype=tcp' and then running the
> tests? Previous versions of nvme-cli had issues with nvme-loop, and were
> trying to set the 'dev_loss_tmo' parameter, which isn't valid for loop.
>
> Cheers,
blktests (master) # nvme_trtype=tcp ./check nvme/039
nvme/039 (Create authenticated connections) [failed]
runtime 1.398s ... 1.397s
--- tests/nvme/039.out 2022-06-08 18:09:06.239931529 -0700
+++ /mnt/data/blktests/results/nodev/nvme/039.out.bad 2022-06-10
02:28:28.077262601 -0700
@@ -1,7 +1,10 @@
Running nvme/039
+losetup: /tmp/blktest-ns1.img: failed to set up loop device:
Device or resource busy
Test unauthenticated connection
-no controller found
+no controller found: failed to write to nvme-fabrics device
NQN:blktests-subsystem-1 disconnected 0 controller(s)
Test authenticated connection
...
(Run 'diff -u tests/nvme/039.out
/mnt/data/blktests/results/nodev/nvme/039.out.bad' to see the entire diff)
blktests (master) # dmesg -c
[ 2294.783157] run blktests nvme/039 at 2022-06-10 02:28:26
[ 2294.829279] nvmet: adding nsid 1 to subsystem blktests-subsystem-1
[ 2294.829568] nvmet: nvmet_ns_enable_store 535
[ 2294.829573] nvmet: nvmet_ns_enable 555
[ 2294.829574] nvmet: nvmet_ns_enable 559
[ 2294.829575] nvmet: nvmet_ns_enable 565
[ 2294.829576] nvmet: nvmet_ns_enable 573
[ 2294.832318] nvmet: nvmet_ns_enable 580
[ 2294.832325] nvmet: nvmet_ns_enable 596
[ 2294.832327] nvmet: nvmet_ns_enable 601
[ 2294.832328] nvmet: nvmet_ns_enable 605
[ 2294.832329] nvmet: nvmet_ns_enable 607
[ 2294.832330] nvmet: nvmet_ns_enable 610
[ 2294.833335] nvmet_tcp: enabling port 0 (127.0.0.1:4420)
[ 2294.840597] nvmet: creating nvm controller 1 for subsystem
blktests-subsystem-1 for NQN
nqn.2014-08.org.nvmexpress:uuid:e4f6b1f3-b215-498e-88c0-cfd20c7dee98
with DH-HMAC-CHAP.
[ 2294.840641] nvme nvme1: qid 0: no key
[ 2294.840645] nvme nvme1: qid 0: authentication setup failed
[ 2294.840668] nvme nvme1: failed to connect queue: 0 ret=401
[ 2294.857350] nvmet: creating nvm controller 2 for subsystem
blktests-subsystem-1 for NQN
nqn.2014-08.org.nvmexpress:uuid:e4f6b1f3-b215-498e-88c0-cfd20c7dee98
with DH-HMAC-CHAP.
[ 2294.857405] __nvme_auth_work 701
[ 2294.862741] nvme nvme1: qid 0: authenticated with hash hmac(sha256)
dhgroup null
[ 2294.862771] nvme nvme1: qid 0: authenticated
[ 2294.864236] nvme nvme1: creating 48 I/O queues.
[ 2294.866802] nvme nvme1: mapped 48/0/0 default/read/poll queues.
[ 2294.867942] __nvme_auth_work 701
[ 2294.873099] __nvme_auth_work 701
[ 2294.878303] __nvme_auth_work 701
[ 2294.883186] __nvme_auth_work 701
[ 2294.888048] __nvme_auth_work 701
[ 2294.893031] __nvme_auth_work 701
[ 2294.897962] __nvme_auth_work 701
[ 2294.902783] __nvme_auth_work 701
[ 2294.907737] __nvme_auth_work 701
[ 2294.912721] __nvme_auth_work 701
[ 2294.917574] __nvme_auth_work 701
[ 2294.922468] __nvme_auth_work 701
[ 2294.927642] __nvme_auth_work 701
[ 2294.932643] __nvme_auth_work 701
[ 2294.937493] __nvme_auth_work 701
[ 2294.942385] __nvme_auth_work 701
[ 2294.947261] __nvme_auth_work 701
[ 2294.952312] __nvme_auth_work 701
[ 2294.957299] __nvme_auth_work 701
[ 2294.962158] __nvme_auth_work 701
[ 2294.967057] __nvme_auth_work 701
[ 2294.972046] __nvme_auth_work 701
[ 2294.977111] __nvme_auth_work 701
[ 2294.982095] __nvme_auth_work 701
[ 2294.987101] __nvme_auth_work 701
[ 2294.992158] __nvme_auth_work 701
[ 2294.997097] __nvme_auth_work 701
[ 2295.001999] __nvme_auth_work 701
[ 2295.006776] __nvme_auth_work 701
[ 2295.011645] __nvme_auth_work 701
[ 2295.016524] __nvme_auth_work 701
[ 2295.021228] __nvme_auth_work 701
[ 2295.026451] __nvme_auth_work 701
[ 2295.031347] __nvme_auth_work 701
[ 2295.036301] __nvme_auth_work 701
[ 2295.041172] __nvme_auth_work 701
[ 2295.046149] __nvme_auth_work 701
[ 2295.051108] __nvme_auth_work 701
[ 2295.056102] __nvme_auth_work 701
[ 2295.061153] __nvme_auth_work 701
[ 2295.066004] __nvme_auth_work 701
[ 2295.070828] __nvme_auth_work 701
[ 2295.075662] __nvme_auth_work 701
[ 2295.080518] __nvme_auth_work 701
[ 2295.085397] __nvme_auth_work 701
[ 2295.090413] __nvme_auth_work 701
[ 2295.095427] __nvme_auth_work 701
[ 2295.100301] __nvme_auth_work 701
[ 2295.105378] nvme nvme1: new ctrl: NQN "blktests-subsystem-1", addr
127.0.0.1:4420
[ 2296.126821] nvme nvme1: Removing ctrl: NQN "blktests-subsystem-1"
[ 2296.159534] nvmet: nvmet_ns_enable_store 535
blktests (master) #
>
> Hannes
More information about the Linux-nvme
mailing list