[PATCH V3] qemu: Enabled SDL when compiling for target architecture

[PATCH V3] qemu: Enabled SDL when compiling for target architecture

[Mihai Prica]

Enables qemu to run images with video output without the need for vncviewer.

Signed-off-by: Mihai Prica <mihai.prica@intel.com>
---
 meta/recipes-devtools/qemu/qemu.inc |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/meta/recipes-devtools/qemu/qemu.inc b/meta/recipes-devtools/qemu/qemu.inc
index 1b5b569..06fa88b 100644
--- a/meta/recipes-devtools/qemu/qemu.inc
+++ b/meta/recipes-devtools/qemu/qemu.inc
@@ -1,7 +1,7 @@
 DESCRIPTION = "open source processor emulator"
 HOMEPAGE = "http://qemu.org"
 LICENSE = "GPLv2 & LGPLv2.1"
-DEPENDS = "glib-2.0 zlib alsa-lib virtual/libx11 pixman dtc"
+DEPENDS = "glib-2.0 zlib alsa-lib virtual/libx11 pixman dtc libsdl"
 DEPENDS_class-native = "zlib-native alsa-lib-native glib-2.0-native pixman-native dtc-native"
 DEPENDS_class-nativesdk = "nativesdk-zlib nativesdk-libsdl nativesdk-glib-2.0 nativesdk-pixman nativesdk-dtc"
 RDEPENDS_${PN}_class-nativesdk = "nativesdk-libsdl"
@@ -13,7 +13,7 @@ BBCLASSEXTEND = "native nativesdk"
 # QEMU_TARGETS is overridable variable
 QEMU_TARGETS ?= "arm i386 mips mipsel mips64 mips64el ppc sh4 x86_64"
 
-SDL ?= "--disable-sdl"
+SDL ?= ""
 SDL_class-native ?= ""
 SDL_class-nativesdk ?= ""
 
-- 
1.7.9.5

Re: [PATCH V3] qemu: Enabled SDL when compiling for target architecture

[Khem Raj]

On Apr 4, 2013, at 5:09 AM, Mihai Prica <mihai.prica@intel.com> wrote:

> Enables qemu to run images with video output without the need for vncviewer.
> 
> Signed-off-by: Mihai Prica <mihai.prica@intel.com>
> ---
> meta/recipes-devtools/qemu/qemu.inc |    4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/meta/recipes-devtools/qemu/qemu.inc b/meta/recipes-devtools/qemu/qemu.inc
> index 1b5b569..06fa88b 100644
> --- a/meta/recipes-devtools/qemu/qemu.inc
> +++ b/meta/recipes-devtools/qemu/qemu.inc
> @@ -1,7 +1,7 @@
> DESCRIPTION = "open source processor emulator"
> HOMEPAGE = "http://qemu.org"
> LICENSE = "GPLv2 & LGPLv2.1"
> -DEPENDS = "glib-2.0 zlib alsa-lib virtual/libx11 pixman dtc"
> +DEPENDS = "glib-2.0 zlib alsa-lib virtual/libx11 pixman dtc libsdl"
> DEPENDS_class-native = "zlib-native alsa-lib-native glib-2.0-native pixman-native dtc-native"
> DEPENDS_class-nativesdk = "nativesdk-zlib nativesdk-libsdl nativesdk-glib-2.0 nativesdk-pixman nativesdk-dtc"
> RDEPENDS_${PN}_class-nativesdk = "nativesdk-libsdl"
> @@ -13,7 +13,7 @@ BBCLASSEXTEND = "native nativesdk"
> # QEMU_TARGETS is overridable variable
> QEMU_TARGETS ?= "arm i386 mips mipsel mips64 mips64el ppc sh4 x86_64"
> 
> -SDL ?= "--disable-sdl"
> +SDL ?= ""

I think this is not acceptable change. Since many systems are headless and its useless for them
secondly, converting this to PACKAGECONFIG and keeping the behavior as it was would be fine.


> SDL_class-native ?= ""
> SDL_class-nativesdk ?= ""
> 
> -- 
> 1.7.9.5
> 
> 
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

Re: [PATCH V3] qemu: Enabled SDL when compiling for target architecture

[Prica, Mihai]

> -----Original Message-----
> From: Khem Raj [mailto:raj.khem@gmail.com]
> Sent: Thursday, April 04, 2013 5:59 PM
> To: Prica, Mihai
> Cc: openembedded-core@lists.openembedded.org
> Subject: Re: [OE-core] [PATCH V3] qemu: Enabled SDL when compiling for target
> architecture
> 
> 
> On Apr 4, 2013, at 5:09 AM, Mihai Prica <mihai.prica@intel.com> wrote:
> 
> > Enables qemu to run images with video output without the need for vncviewer.
> >
> > Signed-off-by: Mihai Prica <mihai.prica@intel.com>
> > ---
> > meta/recipes-devtools/qemu/qemu.inc |    4 ++--
> > 1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/meta/recipes-devtools/qemu/qemu.inc
> > b/meta/recipes-devtools/qemu/qemu.inc
> > index 1b5b569..06fa88b 100644
> > --- a/meta/recipes-devtools/qemu/qemu.inc
> > +++ b/meta/recipes-devtools/qemu/qemu.inc
> > @@ -1,7 +1,7 @@
> > DESCRIPTION = "open source processor emulator"
> > HOMEPAGE = "http://qemu.org"
> > LICENSE = "GPLv2 & LGPLv2.1"
> > -DEPENDS = "glib-2.0 zlib alsa-lib virtual/libx11 pixman dtc"
> > +DEPENDS = "glib-2.0 zlib alsa-lib virtual/libx11 pixman dtc libsdl"
> > DEPENDS_class-native = "zlib-native alsa-lib-native glib-2.0-native pixman-
> native dtc-native"
> > DEPENDS_class-nativesdk = "nativesdk-zlib nativesdk-libsdl nativesdk-glib-2.0
> nativesdk-pixman nativesdk-dtc"
> > RDEPENDS_${PN}_class-nativesdk = "nativesdk-libsdl"
> > @@ -13,7 +13,7 @@ BBCLASSEXTEND = "native nativesdk"
> > # QEMU_TARGETS is overridable variable QEMU_TARGETS ?= "arm i386 mips
> > mipsel mips64 mips64el ppc sh4 x86_64"
> >
> > -SDL ?= "--disable-sdl"
> > +SDL ?= ""
> 
> I think this is not acceptable change. Since many systems are headless and its
> useless for them secondly, converting this to PACKAGECONFIG and keeping the
> behavior as it was would be fine.

I sent a patch to meta-virtualization that just enabled SDL using a bbappend and the
maintainers said I should to integrate this into poky. I talked with Richard and he 
said I should use a PACKAGECONFIG and enable sdl by default for the target. For 
headless images the user can disable it. I sent a V2 with this solution. Saul replied 
and told me there are some problems with this approach and I should stick with the
SDL variable.

> 
> 
> > SDL_class-native ?= ""
> > SDL_class-nativesdk ?= ""
> >
> > --
> > 1.7.9.5
> >
> >
> > _______________________________________________
> > Openembedded-core mailing list
> > Openembedded-core@lists.openembedded.org
> > http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

Re: [PATCH V3] qemu: Enabled SDL when compiling for target architecture

[Khem Raj]

On Apr 4, 2013, at 8:56 AM, "Prica, Mihai" <mihai.prica@intel.com> wrote:

> I sent a patch to meta-virtualization that just enabled SDL using a bbappend and the
> maintainers said I should to integrate this into poky. I talked with Richard and he 
> said I should use a PACKAGECONFIG and enable sdl by default for the target. For 
> headless images the user can disable it. I sent a V2 with this solution. Saul replied 
> and told me there are some problems with this approach and I should stick with the
> SDL variable.

OK so you been diligent. I see this an an image/distro feature. If I don't have X in my features
this should not be enabled either. What was problem with PACKAGECONFIG ?

Re: [PATCH V3] qemu: Enabled SDL when compiling for target architecture

[Richard Purdie]

On Thu, 2013-04-04 at 15:56 +0000, Prica, Mihai wrote:
> > On Apr 4, 2013, at 5:09 AM, Mihai Prica <mihai.prica@intel.com> wrote:
> > > Enables qemu to run images with video output without the need for vncviewer.
> > >
> > > Signed-off-by: Mihai Prica <mihai.prica@intel.com>
> > > ---
> > > meta/recipes-devtools/qemu/qemu.inc |    4 ++--
> > > 1 file changed, 2 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/meta/recipes-devtools/qemu/qemu.inc
> > > b/meta/recipes-devtools/qemu/qemu.inc
> > > index 1b5b569..06fa88b 100644
> > > --- a/meta/recipes-devtools/qemu/qemu.inc
> > > +++ b/meta/recipes-devtools/qemu/qemu.inc
> > > @@ -1,7 +1,7 @@
> > > DESCRIPTION = "open source processor emulator"
> > > HOMEPAGE = "http://qemu.org"
> > > LICENSE = "GPLv2 & LGPLv2.1"
> > > -DEPENDS = "glib-2.0 zlib alsa-lib virtual/libx11 pixman dtc"
> > > +DEPENDS = "glib-2.0 zlib alsa-lib virtual/libx11 pixman dtc libsdl"
> > > DEPENDS_class-native = "zlib-native alsa-lib-native glib-2.0-native pixman-
> > native dtc-native"
> > > DEPENDS_class-nativesdk = "nativesdk-zlib nativesdk-libsdl nativesdk-glib-2.0
> > nativesdk-pixman nativesdk-dtc"
> > > RDEPENDS_${PN}_class-nativesdk = "nativesdk-libsdl"
> > > @@ -13,7 +13,7 @@ BBCLASSEXTEND = "native nativesdk"
> > > # QEMU_TARGETS is overridable variable QEMU_TARGETS ?= "arm i386 mips
> > > mipsel mips64 mips64el ppc sh4 x86_64"
> > >
> > > -SDL ?= "--disable-sdl"
> > > +SDL ?= ""
> > 
> > I think this is not acceptable change. Since many systems are headless and its
> > useless for them secondly, converting this to PACKAGECONFIG and keeping the
> > behavior as it was would be fine.
> 
> I sent a patch to meta-virtualization that just enabled SDL using a bbappend and the
> maintainers said I should to integrate this into poky. I talked with Richard and he 
> said I should use a PACKAGECONFIG and enable sdl by default for the target. For 
> headless images the user can disable it. I sent a V2 with this solution. Saul replied 
> and told me there are some problems with this approach and I should stick with the
> SDL variable.

I'd ultimately like to see PACKAGECONFIG used here but we haven't
managed the right patchset yet so I took this one earlier today as a
short term fix :/.

Cheers,

Richard

Re: [PATCH V3] qemu: Enabled SDL when compiling for target architecture

[Khem Raj]

On Apr 4, 2013, at 9:18 AM, Richard Purdie <richard.purdie@linuxfoundation.org> wrote:

> I'd ultimately like to see PACKAGECONFIG used here but we haven't
> managed the right patchset yet so I took this one earlier today as a
> short term fix :/.

i think its not a good thing for headless systems as I said earlier there are additional packages showing
up on target for no reasons, I could reverse it via a bbappend in my own layer but thats not pleasant

Re: [PATCH V3] qemu: Enabled SDL when compiling for target architecture

[Richard Purdie]

On Thu, 2013-04-04 at 11:07 -0700, Khem Raj wrote:
> On Apr 4, 2013, at 9:18 AM, Richard Purdie <richard.purdie@linuxfoundation.org> wrote:
> 
> > I'd ultimately like to see PACKAGECONFIG used here but we haven't
> > managed the right patchset yet so I took this one earlier today as a
> > short term fix :/.
> 
> i think its not a good thing for headless systems as I said earlier there are additional packages showing
> up on target for no reasons, 

"no reason" isn't true, there are use cases where this is breaking for
some users.

> I could reverse it via a bbappend in my own layer but thats not pleasant

So we need to make this a PACKAGECONFIG option which can be controlled
from the distro, no argument.

We've had various problems getting the patch right for this, that is the
main issue. You'd have to look at the archives to see what those are.

Cheers,

Richard