2020-01-14 16:13:25

by Alan Maguire

Subject: [PATCH] software node: introduce CONFIG_KUNIT_DRIVER_PE_TEST

currently the property entry kunit tests are built if CONFIG_KUNIT=y.
This will cause warnings when merged with the kunit tree that now
supports tristate CONFIG_KUNIT. While the tests appear to compile
as a module, we get a warning about missing module license.

It's better to have a per-test suite CONFIG variable so that
we can do selective building of kunit-based suites, and can
also avoid merge issues like this.

Reported-by: Stephen Rothwell <[email protected]>
Fixes: c032ace71c29 ("software node: add basic tests for property entries")
Signed-off-by: Alan Maguire <[email protected]>
---
drivers/base/test/Kconfig | 3 +++
drivers/base/test/Makefile | 2 +-
2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/base/test/Kconfig b/drivers/base/test/Kconfig
index 86e85da..d29ae95 100644
--- a/drivers/base/test/Kconfig
+++ b/drivers/base/test/Kconfig
@@ -8,3 +8,6 @@ config TEST_ASYNC_DRIVER_PROBE
The module name will be test_async_driver_probe.ko

If unsure say N.
+config KUNIT_DRIVER_PE_TEST
+ bool "KUnit Tests for property entry API"
+ depends on KUNIT
diff --git a/drivers/base/test/Makefile b/drivers/base/test/Makefile
index 2214310..3ca5636 100644
--- a/drivers/base/test/Makefile
+++ b/drivers/base/test/Makefile
@@ -1,4 +1,4 @@
# SPDX-License-Identifier: GPL-2.0
obj-$(CONFIG_TEST_ASYNC_DRIVER_PROBE) += test_async_driver_probe.o

-obj-$(CONFIG_KUNIT) += property-entry-test.o
+obj-$(CONFIG_KUNIT_DRIVER_PE_TEST) += property-entry-test.o
--
1.8.3.1

2020-01-14 16:32:09

by Randy Dunlap

Subject: Re: [PATCH] software node: introduce CONFIG_KUNIT_DRIVER_PE_TEST

Hi Alan,

On 1/14/20 8:09 AM, Alan Maguire wrote:
> currently the property entry kunit tests are built if CONFIG_KUNIT=y.
> This will cause warnings when merged with the kunit tree that now
> supports tristate CONFIG_KUNIT. While the tests appear to compile
> as a module, we get a warning about missing module license.
>
> It's better to have a per-test suite CONFIG variable so that
> we can do selective building of kunit-based suites, and can
> also avoid merge issues like this.
>
> Reported-by: Stephen Rothwell <[email protected]>

Reported-by: Randy Dunlap <[email protected]>

> Fixes: c032ace71c29 ("software node: add basic tests for property entries")
> Signed-off-by: Alan Maguire <[email protected]>
> ---
> drivers/base/test/Kconfig | 3 +++
> drivers/base/test/Makefile | 2 +-
> 2 files changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/base/test/Kconfig b/drivers/base/test/Kconfig
> index 86e85da..d29ae95 100644
> --- a/drivers/base/test/Kconfig
> +++ b/drivers/base/test/Kconfig
> @@ -8,3 +8,6 @@ config TEST_ASYNC_DRIVER_PROBE
> The module name will be test_async_driver_probe.ko
>
> If unsure say N.
> +config KUNIT_DRIVER_PE_TEST
> + bool "KUnit Tests for property entry API"
> + depends on KUNIT

Why is this bool instead of tristate?

> diff --git a/drivers/base/test/Makefile b/drivers/base/test/Makefile
> index 2214310..3ca5636 100644
> --- a/drivers/base/test/Makefile
> +++ b/drivers/base/test/Makefile
> @@ -1,4 +1,4 @@
> # SPDX-License-Identifier: GPL-2.0
> obj-$(CONFIG_TEST_ASYNC_DRIVER_PROBE) += test_async_driver_probe.o
>
> -obj-$(CONFIG_KUNIT) += property-entry-test.o
> +obj-$(CONFIG_KUNIT_DRIVER_PE_TEST) += property-entry-test.o
>

thanks.
--
~Randy

2020-01-14 16:45:06

by Alan Maguire

Subject: Re: [PATCH] software node: introduce CONFIG_KUNIT_DRIVER_PE_TEST

On Tue, 14 Jan 2020, Randy Dunlap wrote:

> Hi Alan,
>
> On 1/14/20 8:09 AM, Alan Maguire wrote:
> > currently the property entry kunit tests are built if CONFIG_KUNIT=y.
> > This will cause warnings when merged with the kunit tree that now
> > supports tristate CONFIG_KUNIT. While the tests appear to compile
> > as a module, we get a warning about missing module license.
> >
> > It's better to have a per-test suite CONFIG variable so that
> > we can do selective building of kunit-based suites, and can
> > also avoid merge issues like this.
> >
> > Reported-by: Stephen Rothwell <[email protected]>
>
> Reported-by: Randy Dunlap <[email protected]>
>

Apologies for missing you out here.

> > Fixes: c032ace71c29 ("software node: add basic tests for property entries")
> > Signed-off-by: Alan Maguire <[email protected]>
> > ---
> > drivers/base/test/Kconfig | 3 +++
> > drivers/base/test/Makefile | 2 +-
> > 2 files changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/base/test/Kconfig b/drivers/base/test/Kconfig
> > index 86e85da..d29ae95 100644
> > --- a/drivers/base/test/Kconfig
> > +++ b/drivers/base/test/Kconfig
> > @@ -8,3 +8,6 @@ config TEST_ASYNC_DRIVER_PROBE
> > The module name will be test_async_driver_probe.ko
> >
> > If unsure say N.
> > +config KUNIT_DRIVER_PE_TEST
> > + bool "KUnit Tests for property entry API"
> > + depends on KUNIT
>
> Why is this bool instead of tristate?
>

The support for building kunit and kunit tests as modules has not merged
into linux-next yet, so if we set the option to tristate the build would
fail for allmodconfig builds. Once it's merged we can revisit though; I
should have mentioned this, thanks for reminding me!

Alan

> > diff --git a/drivers/base/test/Makefile b/drivers/base/test/Makefile
> > index 2214310..3ca5636 100644
> > --- a/drivers/base/test/Makefile
> > +++ b/drivers/base/test/Makefile
> > @@ -1,4 +1,4 @@
> > # SPDX-License-Identifier: GPL-2.0
> > obj-$(CONFIG_TEST_ASYNC_DRIVER_PROBE) += test_async_driver_probe.o
> >
> > -obj-$(CONFIG_KUNIT) += property-entry-test.o
> > +obj-$(CONFIG_KUNIT_DRIVER_PE_TEST) += property-entry-test.o
> >
>
> thanks.
> --
> ~Randy
>
>

2020-01-14 16:47:13

by Randy Dunlap

Subject: Re: [PATCH] software node: introduce CONFIG_KUNIT_DRIVER_PE_TEST

On 1/14/20 8:42 AM, Alan Maguire wrote:
> On Tue, 14 Jan 2020, Randy Dunlap wrote:
>
>> Hi Alan,
>>
>> On 1/14/20 8:09 AM, Alan Maguire wrote:
>>> currently the property entry kunit tests are built if CONFIG_KUNIT=y.
>>> This will cause warnings when merged with the kunit tree that now
>>> supports tristate CONFIG_KUNIT. While the tests appear to compile
>>> as a module, we get a warning about missing module license.
>>>
>>> It's better to have a per-test suite CONFIG variable so that
>>> we can do selective building of kunit-based suites, and can
>>> also avoid merge issues like this.
>>>
>>> Reported-by: Stephen Rothwell <[email protected]>
>>
>> Reported-by: Randy Dunlap <[email protected]>
>>
>
> Apologies for missing you out here.
>
>>> Fixes: c032ace71c29 ("software node: add basic tests for property entries")
>>> Signed-off-by: Alan Maguire <[email protected]>
>>> ---
>>> drivers/base/test/Kconfig | 3 +++
>>> drivers/base/test/Makefile | 2 +-
>>> 2 files changed, 4 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/base/test/Kconfig b/drivers/base/test/Kconfig
>>> index 86e85da..d29ae95 100644
>>> --- a/drivers/base/test/Kconfig
>>> +++ b/drivers/base/test/Kconfig
>>> @@ -8,3 +8,6 @@ config TEST_ASYNC_DRIVER_PROBE
>>> The module name will be test_async_driver_probe.ko
>>>
>>> If unsure say N.
>>> +config KUNIT_DRIVER_PE_TEST
>>> + bool "KUnit Tests for property entry API"
>>> + depends on KUNIT
>>
>> Why is this bool instead of tristate?
>>
>
> The support for building kunit and kunit tests as modules has not merged
> into linux-next yet, so if we set the option to tristate the build would
> fail for allmodconfig builds. Once it's merged we can revisit though; I
> should have mentioned this, thanks for reminding me!

Oh. I see. Thanks.

--
~Randy

2020-01-14 22:46:13

by Rafael J. Wysocki

Subject: Re: [PATCH] software node: introduce CONFIG_KUNIT_DRIVER_PE_TEST

On Tuesday, January 14, 2020 5:45:56 PM CET Randy Dunlap wrote:
> On 1/14/20 8:42 AM, Alan Maguire wrote:
> > On Tue, 14 Jan 2020, Randy Dunlap wrote:
> >
> >> Hi Alan,
> >>
> >> On 1/14/20 8:09 AM, Alan Maguire wrote:
> >>> currently the property entry kunit tests are built if CONFIG_KUNIT=y.
> >>> This will cause warnings when merged with the kunit tree that now
> >>> supports tristate CONFIG_KUNIT. While the tests appear to compile
> >>> as a module, we get a warning about missing module license.
> >>>
> >>> It's better to have a per-test suite CONFIG variable so that
> >>> we can do selective building of kunit-based suites, and can
> >>> also avoid merge issues like this.
> >>>
> >>> Reported-by: Stephen Rothwell <[email protected]>
> >>
> >> Reported-by: Randy Dunlap <[email protected]>
> >>
> >
> > Apologies for missing you out here.
> >
> >>> Fixes: c032ace71c29 ("software node: add basic tests for property entries")
> >>> Signed-off-by: Alan Maguire <[email protected]>
> >>> ---
> >>> drivers/base/test/Kconfig | 3 +++
> >>> drivers/base/test/Makefile | 2 +-
> >>> 2 files changed, 4 insertions(+), 1 deletion(-)
> >>>
> >>> diff --git a/drivers/base/test/Kconfig b/drivers/base/test/Kconfig
> >>> index 86e85da..d29ae95 100644
> >>> --- a/drivers/base/test/Kconfig
> >>> +++ b/drivers/base/test/Kconfig
> >>> @@ -8,3 +8,6 @@ config TEST_ASYNC_DRIVER_PROBE
> >>> The module name will be test_async_driver_probe.ko
> >>>
> >>> If unsure say N.
> >>> +config KUNIT_DRIVER_PE_TEST
> >>> + bool "KUnit Tests for property entry API"
> >>> + depends on KUNIT
> >>
> >> Why is this bool instead of tristate?
> >>
> >
> > The support for building kunit and kunit tests as modules has not merged
> > into linux-next yet, so if we set the option to tristate the build would
> > fail for allmodconfig builds. Once it's merged we can revisit though; I
> > should have mentioned this, thanks for reminding me!
>
> Oh. I see. Thanks.

Patch applied, thanks!

2020-01-14 23:55:10

by Brendan Higgins

Subject: Re: [PATCH] software node: introduce CONFIG_KUNIT_DRIVER_PE_TEST

On Tue, Jan 14, 2020 at 2:43 PM Rafael J. Wysocki <[email protected]> wrote:
>
> On Tuesday, January 14, 2020 5:45:56 PM CET Randy Dunlap wrote:
> > On 1/14/20 8:42 AM, Alan Maguire wrote:
> > > On Tue, 14 Jan 2020, Randy Dunlap wrote:
> > >
> > >> Hi Alan,
> > >>
> > >> On 1/14/20 8:09 AM, Alan Maguire wrote:
> > >>> currently the property entry kunit tests are built if CONFIG_KUNIT=y.
> > >>> This will cause warnings when merged with the kunit tree that now
> > >>> supports tristate CONFIG_KUNIT. While the tests appear to compile
> > >>> as a module, we get a warning about missing module license.
> > >>>
> > >>> It's better to have a per-test suite CONFIG variable so that
> > >>> we can do selective building of kunit-based suites, and can
> > >>> also avoid merge issues like this.
> > >>>
> > >>> Reported-by: Stephen Rothwell <[email protected]>
> > >>
> > >> Reported-by: Randy Dunlap <[email protected]>
> > >>
> > >
> > > Apologies for missing you out here.
> > >
> > >>> Fixes: c032ace71c29 ("software node: add basic tests for property entries")
> > >>> Signed-off-by: Alan Maguire <[email protected]>
> > >>> ---
> > >>> drivers/base/test/Kconfig | 3 +++
> > >>> drivers/base/test/Makefile | 2 +-
> > >>> 2 files changed, 4 insertions(+), 1 deletion(-)
> > >>>
> > >>> diff --git a/drivers/base/test/Kconfig b/drivers/base/test/Kconfig
> > >>> index 86e85da..d29ae95 100644
> > >>> --- a/drivers/base/test/Kconfig
> > >>> +++ b/drivers/base/test/Kconfig
> > >>> @@ -8,3 +8,6 @@ config TEST_ASYNC_DRIVER_PROBE
> > >>> The module name will be test_async_driver_probe.ko
> > >>>
> > >>> If unsure say N.
> > >>> +config KUNIT_DRIVER_PE_TEST
> > >>> + bool "KUnit Tests for property entry API"
> > >>> + depends on KUNIT
> > >>
> > >> Why is this bool instead of tristate?
> > >>
> > >
> > > The support for building kunit and kunit tests as modules has not merged
> > > into linux-next yet, so if we set the option to tristate the build would
> > > fail for allmodconfig builds. Once it's merged we can revisit though; I
> > > should have mentioned this, thanks for reminding me!
> >
> > Oh. I see. Thanks.
>
> Patch applied, thanks!

Cool, looks good. Thanks!