[PATCH blktests 00/10] nvme: test cases for TLS support

Shinichiro Kawasaki shinichiro.kawasaki at wdc.com
Tue Apr 15 01:00:12 PDT 2025


On Apr 15, 2025 / 09:50, Hannes Reinecke wrote:
> On 4/15/25 09:47, Shinichiro Kawasaki wrote:
> > On Apr 02, 2025 / 16:08, Shin'ichiro Kawasaki wrote:
> > > Hannes created two testcases and shared them as the GitHub PR 158 [1]. Quote:
> > > 
> > >   "This pull request adds two new testcases for nvme TLS support, one for 'plain'
> > >    TLS with TLS PSKs, and the other one for testing 'secure concatenation' where
> > >    TLS is started after DH-HMAC-CHAP authentication."
> > > 
> > > The testcases were missing a few requirement checks. They also modified
> > > systemctl service status after test case runs. To address these problems, I took
> > > the liberty to add some more patches and modified the testcases. Here I post the
> > > modified patch series for wider review.
> > > 
> > > The first five patches are preparation patches to add several helper functions
> > > for requirement checks and systemctl support. The last five patches are
> > > originally created by Hannes to add the new testcases nvme/060 and nvme/061.
> > > 
> > > I ran the two test cases using the kernel v6.14 with the patch series titled
> > > "[PATCHv15 00/10] nvme: implement secure concatenation" [2]. I observed nvme/060
> > > passed, but nvme/061 failed. FYI, I share the failure logs [3]. Actions to fix
> > > the failure will be appreciated.
> > 
> > FYI, I applied this series. Of note is that the two new test cases were
> > renumbered from nvme/060 & 061 to nvme/062 & 063 to avoid the number conflict.
> > 
> > > P.S. Hannes, please check the Copyright year of the second test case nvme/061.
> > >       It is 2022, but I guess you meant 2024 or 2025.
> > 
> > I still suspect that Hannes intended the copyright year 2024 or 2025, but keep
> > it 2022 unless Hannes requests change for it.
> 
> Weelll ... 2022 is actually correct, for that's when I wrote the
> original test for TLS. It really took its time getting upstream ...

Ah, got it. Thanks for the clarification.


More information about the Linux-nvme mailing list