Hi Anuj,

On Wed, Jun 10, 2020 at 8:03 PM Mittal, Anuj <[email protected]> wrote:

> Hi Arthur,
>
> On Wed, 2020-06-10 at 18:41 -0700, Arthur She wrote:
> > Hi Anuj,
> > Thanks for your reply.
> >
> > On Tue, Jun 9, 2020 at 10:55 PM Mittal, Anuj <[email protected]>
> > wrote:
> > > Hello,
> > >
> > > On Tue, 2020-06-09 at 22:24 -0700, Arthur She wrote:
> > > > Hi,
> > > > We run igt-gpu-tools with Chamelium for the graphic/DRM test in
> > > > Linaro.
> > > > Something like this.
> > > > The recipe we used is basically copied from meta-intel and did
> > > some
> > > > modifications
> > > > 1. Compile for ARM architecture
> > > > 2. Build git HEAD code instead of a specific version
> > > > 3. Migrate to Meson build system
> > > > 4. Enable Chamelium
> > > >
> > > > We're thinking about upstream the recipe to meta-oe, however,
> > > there
> > > > might be a duplicate recipe problem.
> > > >
> > >
> > > I wonder if this recipe can be in oe-core since it's needed by all
> > > and
> > > helps people validate core functionality and all recipes in
> > > DEPENDS/RDEPENDS are also in core? Would you minding sending it
> > > there
> > > and seek opinion? I will remove it from meta-intel in that case.
> >
> > I've sent it for review and gain some feedback. One of the problems
> > they concerned about is the duplicate recipe.
> > I will base on igt-gpu-tools_1.25.bb to revise the patch and send it
> > to oe-devel and meta-intel for review.
> > Does it work for you?
>
> I just looked through the list archive and noticed that you had sent it
> to oe-devel/meta-oe. I was wondering if it should be sent specifically
> for inclusion in oe-core? The original email sent to oe-core list had
> [meta-oe] tag so I assume that was one of the reasons you were
> requested to send it to oe-devel list ...
>

I submitted the patch to meta-openembedded/meta-oe/recipes-graphics/.
>From the mailing list <https://www.openembedded.org/wiki/Mailing_lists>, I
think it makes sense to send it to oe-devel list.


> >
> > > But your version of the recipe in my opinion needs some
> > > modifications
> > > before it can be accepted there:
> > >
> > > 1, it should be pointing to the release tarball.
> >
> > The core maintainer also mentioned about it. I will use a specific
> > SHA instead of autorev.
> >
> > > 2, the LICENSE field is wrong. The recipe in meta-intel requires
> > > the
> > > xorg-app-common.inc file which sets the LICENSE field. Hence the &
> > > ISC.
> > > But in this version, that information is lost.
> >
> > They also pointed it out.
> > Should it be LICENSE = "MIT"?
>
> I think so.
>
> >
> > > 3, You might have to set SECURITY_LDFLAGS for it to work with
> > > default
> > > poky configuration.
> >
> > Is it particularly used for xorg?
> > We build it with our Weston image and it worked without any problem.
> >
>
> I am not sure if we specifically need to set those when using xorg. I
> never tried Weston. Fedora also does it:
>
>
> https://src.fedoraproject.org/rpms/igt-gpu-tools/blob/master/f/igt-gpu-tools.spec#_95
>
> Do you use meta/conf/distro/include/security_flags.inc that sets
> SECURITY_LDFLAGS and includes them in TARGET_LDFLAGS?
>
> I just checked our recipes, it seems we don't use security_flags.inc.
I rebuilt igt-gpu-tools with SECURITY_LDFLAGS and it still worked with our
image.
I will keep it in the patch.

Thanks,
Arthur

> Thanks,
>
> Anuj
>
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#6613): 
https://lists.yoctoproject.org/g/meta-intel/message/6613
Mute This Topic: https://lists.yoctoproject.org/mt/74790915/21656
Group Owner: [email protected]
Unsubscribe: https://lists.yoctoproject.org/g/meta-intel/unsub  
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to