linux-kernel - Re: [PATCH] clk: tests: Add tests for clk lookup by name

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <aOgFSi_h95tfD8x2@redhat.com>
Date: Thu, 9 Oct 2025 14:56:10 -0400
From: Brian Masney <bmasney@...hat.com>
To: Chen-Yu Tsai <wenst@...omium.org>
Cc: Stephen Boyd <sboyd@...nel.org>, linux-clk@...r.kernel.org,
	linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] clk: tests: Add tests for clk lookup by name

On Thu, Oct 09, 2025 at 11:24:22AM +0800, Chen-Yu Tsai wrote:
> On Thu, Oct 9, 2025 at 7:58 AM Brian Masney <bmasney@...hat.com> wrote:
> >
> > On Thu, Oct 02, 2025 at 05:20:35PM +0800, Chen-Yu Tsai wrote:
> > > Clk lookup (by name) recently gained some performance improvements at
> > > the expense of more complexity within the lookup code.
> > >
> > > To make sure that this works as intended and doesn't break, add some
> > > basic tests for this part of the CCF.
> > >
> > > A new "clk_hw_lookup()" function is added purely for running kunit
> > > tests.
> > >
> > > Signed-off-by: Chen-Yu Tsai <wenst@...omium.org>
> > > ---
> > >  drivers/clk/clk.c      | 11 +++++++
> > >  drivers/clk/clk.h      |  4 +++
> > >  drivers/clk/clk_test.c | 66 +++++++++++++++++++++++++++++++++++++++++-
> > >  3 files changed, 80 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> > > index 85d2f2481acf..a17d0070d11f 100644
> > > --- a/drivers/clk/clk.c
> > > +++ b/drivers/clk/clk.c
> > > @@ -778,6 +778,17 @@ struct clk *__clk_lookup(const char *name)
> > >       return !core ? NULL : core->hw->clk;
> > >  }
> > >
> > > +#if IS_ENABLED(CONFIG_CLK_KUNIT_TEST)
> > > +/* This is only provided for kunit tests to test the core lookup functions. */
> > > +struct clk_hw *clk_hw_lookup(const char *name)
> > > +{
> > > +     struct clk_core *core = clk_core_lookup(name);
> > > +
> > > +     return !core ? NULL : core->hw;
> > > +}
> > > +EXPORT_SYMBOL_GPL(clk_hw_lookup);
> > > +#endif
> >
> > Use EXPORT_SYMBOL_IF_KUNIT instead for consistency with the rest of the
> > kernel. In clk_test.c, you'll also need to add:
> >
> > MODULE_IMPORT_NS("EXPORTED_FOR_KUNIT_TESTING");
> 
> Didn't know about this one. Thanks!
> 
> > Since clk_hw_lookup() is only used by kunit, why not just put this new
> > function in clk-test.c, and use EXPORT_SYMBOL_IF_KUNIT on
> > clk_core_lookup?
> 
> Then we end up sort of exposing clk_core_lookup as well?
> 
> I believe Stephen wants to keep things contained as much as possible.

I agree about keeping things contained as much as possible as well.
clk_core_lookup() would only be exposed to the kunit tests if you used
EXPORT_SYMBOL_IF_KUNIT. Definitely go with whatever approach you think
Stephen will prefer.

Brian