On Fri, Nov 08, 2024 at 09:56:15AM +0100, Geert Uytterhoeven wrote:
> Hi Stephen,
>
> On Sat, Oct 19, 2024 at 12:21 AM Stephen Boyd <sboyd@xxxxxxxxxx> wrote:
> > Quoting Rob Herring (2024-10-17 13:38:10)
> > > On Wed, Oct 16, 2024 at 02:20:15PM -0700, Stephen Boyd wrote:
> > > > Some configurations want to enable CONFIG_KUNIT without enabling
> > > > CONFIG_OF_OVERLAY. The kunit overlay code already skips if
> > > > CONFIG_OF_OVERLAY isn't enabled, so this select here isn't really doing
> > > > anything besides making it easier to run the tests without them
> > > > skipping. Remove the select and move the config setting to the
> > > > drivers/of/.kunitconfig file so that the overlay tests can be run with
> > > > or without CONFIG_OF_OVERLAY set to test either behavior.
>
> Thanks for your patch, which is now commit 027af8c26fdc0642 ("of: Allow
> overlay kunit tests to run CONFIG_OF_OVERLAY=n") in dt-rh/for-next.
>
> > > > Fixes: 5c9dd72d8385 ("of: Add a KUnit test for overlays and test managed APIs")
> > >
> > > Doesn't really seem like a fix.
> >
> > Ok. Feel free to drop the tag.
> >
> > > Does this need to go into 6.12?
> >
> > It's only important for 6.12 if kernel configurators want to build the
> > kernel with OF_OVERLAY_KUNIT_TEST enabled and not be forced to enable
> > CONFIG_OF_OVERLAY. I don't mind if it waits a while.
>
> I'd say it's a fix, so please keep at last the Fixes tag. Merely
> enabling kunit tests (which can be modular) should not increase the
> possible attack vector on a product by enabling extra unneeded code.
Not sure I buy that that is an actual problem. However, not worth
arguing over. I only really care because if there's a Fixes, then this
really should go to Linus for 6.12 rather than eventually get
auto-selected from 6.13 to go to stable. So I moved it to send to Linus,
but I found that CONFIG_OF_OVERLAY is still selected with this patch.
That's because the clock kunit tests also select CONFIG_OF_OVERLAY.
That's fixed in next, but it's not queued up for 6.12.
Rob
