Kernel configurations in the CIP project


Lukas Bulwahn <Lukas.Bulwahn@...>
 

Hi Robert,

If I recall our conversation and follow the discussion on cip-dev correctly, you are still maintaining the test builds and infrastructure from the CIP project, right?

Could you provide to Markus Kreidl, Nicholas McGuire (see CC) and me the kernel configurations you use in the test builds for the different CIP members?

Background:

Markus Kreidl has written a tool that determines which patches, which hunks of those patches are relevant to a given configuration. We would be interested in providing an overview of which patches are especially relevant for the given configs of a new CIP kernel patchlevel release, and hence, the CIP maintainers (hopefully, it is not just Ben) are informed about those and can have a good look at those specific patches when they provide a new kernel release version.

If it useful, we would follow-up how to make that fit nicely into the existing review and reporting process. If it not useful, we learned a lesson about our tool.

Best regards,

Lukas



Daniel Sangorrin <daniel.sangorrin@...>
 

Hello Lukas,

-----Original Message-----
From: cip-dev-bounces@...
Hi Robert,

If I recall our conversation and follow the discussion on cip-dev correctly, you are still
maintaining the test builds and infrastructure from the CIP project, right?

Could you provide to Markus Kreidl, Nicholas McGuire (see CC) and me the kernel
configurations you use in the test builds for the different CIP members?
Not sure about the configs that Robert uses for testing, but the configurations from CIP members can be found here.
https://gitlab.com/cip-project/cip-kernel/cip-kernel-config

Background:

Markus Kreidl has written a tool that determines which patches, which hunks of
those patches are relevant to a given configuration. We would be interested in
providing an overview of which patches are especially relevant for the given configs
of a new CIP kernel patchlevel release, and hence, the CIP maintainers (hopefully, it
is not just Ben) are informed about those and can have a good look at those specific
patches when they provide a new kernel release version.

If it useful, we would follow-up how to make that fit nicely into the existing review
and reporting process. If it not useful, we learned a lesson about our tool.
That sounds very interesting indeed.
It would be nice if you could also map the patches to the LTP tests that need to be run.

Thanks,
Daniel


Robert Marshall <robert.marshall@...>
 

Lukas

Thanks for your email, comments below.

Lukas Bulwahn <Lukas.Bulwahn@...> writes:

Hi Robert,

If I recall our conversation and follow the discussion on cip-dev
correctly, you are still maintaining the test builds and
infrastructure from the CIP project, right?
That's correct

Could you provide to Markus Kreidl, Nicholas McGuire (see CC) and me
the kernel configurations you use in the test builds for the different
CIP members?
I'm using the kernel versions with the latest CIP tag, I use exactly
the procedure described on this page

https://wiki.linuxfoundation.org/civilinfrastructureplatform/cipsystembuildhowto#building-the-cip-kernel-with-kernel-ci

building for the beaglebone black and the Renesas iwg20m board.


Robert

Background:

Markus Kreidl has written a tool that determines which patches, which
hunks of those patches are relevant to a given configuration. We would
be interested in providing an overview of which patches are especially
relevant for the given configs of a new CIP kernel patchlevel release,
and hence, the CIP maintainers (hopefully, it is not just Ben) are
informed about those and can have a good look at those specific
patches when they provide a new kernel release version.

If it useful, we would follow-up how to make that fit nicely into the
existing review and reporting process. If it not useful, we learned a
lesson about our tool.

Best regards,

Lukas

--
Robert Marshall, Software Developer Codethink Ltd
Telephone: +44 7762 840 414 3rd Floor, Dale House, 35 Dale Street
https://www.codethink.co.uk/ MANCHESTER, M1 2HF. United Kingdom


Ben Hutchings <ben.hutchings@...>
 

On Wed, 2018-10-03 at 08:33 +0900, Daniel Sangorrin wrote:
Hello Lukas,

-----Original Message-----
From: cip-dev-bounces@...
Hi Robert,

If I recall our conversation and follow the discussion on cip-dev correctly, you are still
maintaining the test builds and infrastructure from the CIP project, right?

Could you provide to Markus Kreidl, Nicholas McGuire (see CC) and me the kernel
configurations you use in the test builds for the different CIP members?
Not sure about the configs that Robert uses for testing, but the configurations from CIP members can be found here.
https://gitlab.com/cip-project/cip-kernel/cip-kernel-config
[...]

I think these are the configurations that Lukas needs - they are the
configurations that should be used for determining whether commits are
relevant to CIP.

Ben.

--
Ben Hutchings, Software Developer   Codethink Ltd
https://www.codethink.co.uk/ Dale House, 35 Dale Street
Manchester, M1 2HF, United Kingdom


Lukas Bulwahn <Lukas.Bulwahn@...>
 

On 2018-10-04, 15:52, "Ben Hutchings" <ben.hutchings@...> wrote:

On Wed, 2018-10-03 at 08:33 +0900, Daniel Sangorrin wrote:
> Hello Lukas,
>
> > -----Original Message-----
> > From: cip-dev-bounces@...
> > Hi Robert,
> >
> > If I recall our conversation and follow the discussion on cip-dev correctly, you are still
> > maintaining the test builds and infrastructure from the CIP project, right?
> >
> > Could you provide to Markus Kreidl, Nicholas McGuire (see CC) and me the kernel
> > configurations you use in the test builds for the different CIP members?
>
> Not sure about the configs that Robert uses for testing, but the configurations from CIP members can be found here.
> https://gitlab.com/cip-project/cip-kernel/cip-kernel-config
[...]

I think these are the configurations that Lukas needs - they are the
configurations that should be used for determining whether commits are
relevant to CIP.

Thanks for all the responses. I believe Markus Kreidl is already looking how he can produce useful information for the cip releases and those configs.


Lukas


Nicholas Mc Guire <der.herr@...>
 

On Thu, Oct 04, 2018 at 02:51:46PM +0100, Ben Hutchings wrote:
On Wed, 2018-10-03 at 08:33 +0900, Daniel Sangorrin wrote:
Hello Lukas,

-----Original Message-----
From: cip-dev-bounces@...
Hi Robert,

If I recall our conversation and follow the discussion on cip-dev correctly, you are still
maintaining the test builds and infrastructure from the CIP project, right?

Could you provide to Markus Kreidl, Nicholas McGuire (see CC) and me the kernel
configurations you use in the test builds for the different CIP members?
Not sure about the configs that Robert uses for testing, but the configurations from CIP members can be found here.
https://gitlab.com/cip-project/cip-kernel/cip-kernel-config
[...]

I think these are the configurations that Lukas needs - they are the
configurations that should be used for determining whether commits are
relevant to CIP.
ok - we started playing with them the siemens iot config seems very large to me is that
a designed config or did that "happen" ?

thx!
hofrat


Jan Kiszka
 

On 05.10.18 06:10, Nicholas Mc Guire wrote:
On Thu, Oct 04, 2018 at 02:51:46PM +0100, Ben Hutchings wrote:
On Wed, 2018-10-03 at 08:33 +0900, Daniel Sangorrin wrote:
Hello Lukas,

-----Original Message-----
From: cip-dev-bounces@...
Hi Robert,

If I recall our conversation and follow the discussion on cip-dev correctly, you are still
maintaining the test builds and infrastructure from the CIP project, right?

Could you provide to Markus Kreidl, Nicholas McGuire (see CC) and me the kernel
configurations you use in the test builds for the different CIP members?
Not sure about the configs that Robert uses for testing, but the configurations from CIP members can be found here.
https://gitlab.com/cip-project/cip-kernel/cip-kernel-config
[...]

I think these are the configurations that Lukas needs - they are the
configurations that should be used for determining whether commits are
relevant to CIP.
ok - we started playing with them the siemens iot config seems very large to me is that
a designed config or did that "happen" ?
The IOT2000 is an open platform, and that naturally brings in the need to have more features preconfigured than one would selected in a dedicated devices, e.g. drivers for pluggable devices.

Jan

--
Siemens AG, Corporate Technology, CT RDA IOT SES-DE
Corporate Competence Center Embedded Linux


Nicholas Mc Guire <der.herr@...>
 

On Mon, Oct 08, 2018 at 02:33:32PM +0200, Jan Kiszka wrote:
On 05.10.18 06:10, Nicholas Mc Guire wrote:
On Thu, Oct 04, 2018 at 02:51:46PM +0100, Ben Hutchings wrote:
On Wed, 2018-10-03 at 08:33 +0900, Daniel Sangorrin wrote:
Hello Lukas,

-----Original Message-----
From: cip-dev-bounces@...
Hi Robert,

If I recall our conversation and follow the discussion on cip-dev correctly, you are still
maintaining the test builds and infrastructure from the CIP project, right?

Could you provide to Markus Kreidl, Nicholas McGuire (see CC) and me the kernel
configurations you use in the test builds for the different CIP members?
Not sure about the configs that Robert uses for testing, but the configurations from CIP members can be found here.
https://gitlab.com/cip-project/cip-kernel/cip-kernel-config
[...]

I think these are the configurations that Lukas needs - they are the
configurations that should be used for determining whether commits are
relevant to CIP.
ok - we started playing with them the siemens iot config seems very large to me is that
a designed config or did that "happen" ?
The IOT2000 is an open platform, and that naturally brings in the need to
have more features preconfigured than one would selected in a dedicated
devices, e.g. drivers for pluggable devices.
ok - but that makes the evaluation of actual impact quite hard
so it would be very helpful to have a realistic reference example for
a particular use-case if that is possible. The current configuration
will give a very pesemistic view of the potential impact. Do you know
if there is some Use-Case that we could use as a comparison ?

thx!
hofrat