Re: [Buildroot] [PATCH] package/opengl/libegl: remove BR2_PACKAGE_HAS_LIBEGL_WAYLAND

From: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
To: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>,
	Gary Bisson <bisson.gary@gmail.com>,
	Refik Tuzakli <tuzakli.refik@gmail.com>,
	Buildroot List <buildroot@buildroot.org>,
	Adrian Perez de Castro <aperez@igalia.com>,
	Romain Naour <romain.naour@gmail.com>,
	Fabrice Fontaine <fontaine.fabrice@gmail.com>
Subject: Re: [Buildroot] [PATCH] package/opengl/libegl: remove BR2_PACKAGE_HAS_LIBEGL_WAYLAND
Date: Sat, 24 Sep 2022 23:00:21 +0200	[thread overview]
Message-ID: <20220924230021.23e6a310@windsurf> (raw)
In-Reply-To: <20220924153729.GF1127102@scaer>

Hello Yann,

Thanks for the review!

On Sat, 24 Sep 2022 17:37:29 +0200
"Yann E. MORIN" <yann.morin.1998@free.fr> wrote:

> So, what I understood from upstream commit log, was that NVIdia driver
> does have a libwayland-egl, but our nvidia-driver does not install one.
> 
> So, there seems to be no conflict. Still, this is strange that our
> nvidia-driver did select BR2_PACKAGE_HAS_LIBEGL_WAYLAND without
> installing libwayland-egl.so...

It just works because libwayland-egl.so is now provided by the wayland
package.

> > -	depends on BR2_PACKAGE_HAS_LIBEGL_WAYLAND || \
> > +	depends on BR2_PACKAGE_HAS_LIBEGL || \  
> 
> I am not sure this is equivalent in this case: indeed, it is possivble
> to have EGL without having wayland enabled, and there currently is
> nothing that enforces wayland to be enabled for vte, even conditionally.

In the context of package/vte, the depends on
BR2_PACKAGE_LIBEGL_WAYLAND was useless. Look:

        depends on BR2_PACKAGE_HAS_LIBEGL_WAYLAND || \
                BR2_PACKAGE_HAS_LIBGL
        depends on BR2_PACKAGE_XORG7

See the "depends on BR2_PACKAGE_XORG7" ? So we certainly cannot be in a
wayland situation with the vte package. So you might wonder why there
is a depends on BR2_PACKAGE_HAS_LIBEGL_WAYLAND. Simply because it is a
dependency inherited from libgtk3.

So my patch is simply propagating again the dependency from libgtk3 to
vte.

> > diff --git a/package/weston/weston.mk b/package/weston/weston.mk
> > index 5e71473640..1c7c8a98df 100644
> > --- a/package/weston/weston.mk
> > +++ b/package/weston/weston.mk
> > @@ -71,7 +71,7 @@ else
> >  WESTON_CONF_OPTS += -Ddeprecated-weston-launch=false
> >  endif
> >  
> > -ifeq ($(BR2_PACKAGE_HAS_LIBEGL_WAYLAND)$(BR2_PACKAGE_HAS_LIBGLES),yy)
> > +ifeq ($(BR2_PACKAGE_HAS_LIBGLES),y)  
> 
> Why can we drop the BR2_PACKAGE_HAS_LIBEGL entirely here?
> 
> As far as I can see and remember, the two are orthogonal (even if GLES
> is in practice most often used in conjunction with EGL rather than full
> GL).

That's a very good point, I'll fix that and send a v2. Thanks!

Thomas
-- 
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot