[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CABVgOSnde5VmF8TLfF=B4Nbvo7C5GUJ91OY9Rct-ksyPaTtpcw@mail.gmail.com>
Date: Fri, 30 Jul 2021 13:33:46 +0800
From: David Gow <davidgow@...gle.com>
To: Herbert Xu <herbert@...dor.apana.org.au>
Cc: Daniel Latypov <dlatypov@...gle.com>, davem@...emloft.net,
linux-crypto@...r.kernel.org,
Brendan Higgins <brendanhiggins@...gle.com>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
KUnit Development <kunit-dev@...glegroups.com>
Subject: Re: [RFC v1 1/2] crypto: tcrypt: minimal conversion to run under KUnit
On Fri, Jul 30, 2021 at 10:55 AM Herbert Xu <herbert@...dor.apana.org.au> wrote:
>
> On Fri, Jul 23, 2021 at 12:31:28PM -0700, Daniel Latypov wrote:
> >
> > Thanks, that makes a lot of sense.
> > In that case, how useful would `kunit.py run` be? I.e. Do people
> > mostly want to see numbers on bare metal?
>
> I think it's a mix of both. As in performance on bare metal and
> under virtualisation may be of interest. I don't think you're going
> to be going through kunit for the speed tests though, because you
> need to supply module parameters for tcrypt to do that.
FYI, there is a patch for kunit_tool which will allow kernel
parameters to be passed through:
https://patchwork.kernel.org/project/linux-kselftest/patch/20210715160819.1107685-1-dlatypov@google.com/
That being said, no-one's ever used any of the KUnit tooling for
performance testing before, as far as I know, so whether or not it
turns out to be useful or not remains to be seen. With this patch,
it'd at least be an option if you wanted to try it.
-- David
Powered by blists - more mailing lists