build failure with xf86-video-omap

build failure with xf86-video-omap

[Nicolas Dechesne]

[-- Attachment #1: Type: text/plain, Size: 2083 bytes --]

hi ,

since xf86-video-omap was recently merged in oe-core and poky, i am trying
to remove the recipes we had staged in meta-ti trees for some weeks (
http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf).
so i was trying to build  poky/master with meta-ti/master, for pandaboard,
and I get build failures in xf86-video-omap [1] and [2].

so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i
can send a patch in oe-core for that.

[2] is a bit more cryptic to me.. so I quickly tried to build with danny (
i just cherry picked the commit that introduced xf86-video-omap from
master) and this time it built fine.

has anyone actually tested this recipe? if so I must be doing something
wrong, any hint would be helpful.

i also tried a build without poky, and just oe-core, and got the same
failure.

cheers

nicolas

[1] : from log_do.configure for xf86-video-omap
checking for XORG... no
configure: error: Package requirements (xorg-server >= 1.3 xproto
fontsproto libdrm >= 2.4.36 libdrm_omap xf86driproto  randrproto
renderproto videoproto xextproto) were not met:



No package 'xf86driproto' found

Consider adjusting the PKG_CONFIG_PATH environment variable if you
installed software in a non-standard prefix.


[2] : from log_do.configure for xf86-video-omap
checking dependency style of arm-poky-linux-gnueabi-gcc  -march=armv7-a
-mthumb-\
interwork -mfloat-abi=softfp -mfpu=neon -mtune=cortex-a9
--sysroot=/data/psi_omap_bu\
ilds_admins/ndechesne/yocto/build-poky2/tmp/sysroots/pandaboard... (cached)
none
checking for sys/ioctl.h... (cached) yes
checking if RANDR is defined... yes
checking if RENDER is defined... yes
checking if XV is defined... yes
checking if DPMSExtension is defined... yes
checking for XORG... yes
checking for ANSI C header files... (cached) yes
checking whether to include DRI support... checking for
/usr/include/xorg/dri.h...
configure: error: cannot check for file existence when cross compiling
... and configure just

[-- Attachment #2: Type: text/html, Size: 2896 bytes --]

Re: build failure with xf86-video-omap

[Martin Jansa]

[-- Attachment #1: Type: text/plain, Size: 2613 bytes --]

On Wed, Nov 21, 2012 at 11:52:30PM +0100, Nicolas Dechesne wrote:
> hi ,
> 
> since xf86-video-omap was recently merged in oe-core and poky, i am trying
> to remove the recipes we had staged in meta-ti trees for some weeks (
> http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf).
> so i was trying to build  poky/master with meta-ti/master, for pandaboard,
> and I get build failures in xf86-video-omap [1] and [2].
> 
> so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i
> can send a patch in oe-core for that.
> 
> [2] is a bit more cryptic to me.. so I quickly tried to build with danny (
> i just cherry picked the commit that introduced xf86-video-omap from
> master) and this time it built fine.

Check this tread:
http://lists.linuxtogo.org/pipermail/openembedded-core/2012-November/032012.html
 
> has anyone actually tested this recipe? 

obviously no

Cheers,

> if so I must be doing something > wrong, any hint would be helpful.
> 
> i also tried a build without poky, and just oe-core, and got the same
> failure.
> 
> cheers
> 
> nicolas
> 
> [1] : from log_do.configure for xf86-video-omap
> checking for XORG... no
> configure: error: Package requirements (xorg-server >= 1.3 xproto
> fontsproto libdrm >= 2.4.36 libdrm_omap xf86driproto  randrproto
> renderproto videoproto xextproto) were not met:
> 
> 
> 
> No package 'xf86driproto' found
> 
> Consider adjusting the PKG_CONFIG_PATH environment variable if you
> installed software in a non-standard prefix.
> 
> 
> [2] : from log_do.configure for xf86-video-omap
> checking dependency style of arm-poky-linux-gnueabi-gcc  -march=armv7-a
> -mthumb-\
> interwork -mfloat-abi=softfp -mfpu=neon -mtune=cortex-a9
> --sysroot=/data/psi_omap_bu\
> ilds_admins/ndechesne/yocto/build-poky2/tmp/sysroots/pandaboard... (cached)
> none
> checking for sys/ioctl.h... (cached) yes
> checking if RANDR is defined... yes
> checking if RENDER is defined... yes
> checking if XV is defined... yes
> checking if DPMSExtension is defined... yes
> checking for XORG... yes
> checking for ANSI C header files... (cached) yes
> checking whether to include DRI support... checking for
> /usr/include/xorg/dri.h...
> configure: error: cannot check for file existence when cross compiling
> ... and configure just

> _______________________________________________
> yocto mailing list
> yocto@yoctoproject.org
> https://lists.yoctoproject.org/listinfo/yocto


-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 205 bytes --]

Re: [yocto] build failure with xf86-video-omap

[Martin Jansa]

[-- Attachment #1: Type: text/plain, Size: 2613 bytes --]

On Wed, Nov 21, 2012 at 11:52:30PM +0100, Nicolas Dechesne wrote:
> hi ,
> 
> since xf86-video-omap was recently merged in oe-core and poky, i am trying
> to remove the recipes we had staged in meta-ti trees for some weeks (
> http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf).
> so i was trying to build  poky/master with meta-ti/master, for pandaboard,
> and I get build failures in xf86-video-omap [1] and [2].
> 
> so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i
> can send a patch in oe-core for that.
> 
> [2] is a bit more cryptic to me.. so I quickly tried to build with danny (
> i just cherry picked the commit that introduced xf86-video-omap from
> master) and this time it built fine.

Check this tread:
http://lists.linuxtogo.org/pipermail/openembedded-core/2012-November/032012.html
 
> has anyone actually tested this recipe? 

obviously no

Cheers,

> if so I must be doing something > wrong, any hint would be helpful.
> 
> i also tried a build without poky, and just oe-core, and got the same
> failure.
> 
> cheers
> 
> nicolas
> 
> [1] : from log_do.configure for xf86-video-omap
> checking for XORG... no
> configure: error: Package requirements (xorg-server >= 1.3 xproto
> fontsproto libdrm >= 2.4.36 libdrm_omap xf86driproto  randrproto
> renderproto videoproto xextproto) were not met:
> 
> 
> 
> No package 'xf86driproto' found
> 
> Consider adjusting the PKG_CONFIG_PATH environment variable if you
> installed software in a non-standard prefix.
> 
> 
> [2] : from log_do.configure for xf86-video-omap
> checking dependency style of arm-poky-linux-gnueabi-gcc  -march=armv7-a
> -mthumb-\
> interwork -mfloat-abi=softfp -mfpu=neon -mtune=cortex-a9
> --sysroot=/data/psi_omap_bu\
> ilds_admins/ndechesne/yocto/build-poky2/tmp/sysroots/pandaboard... (cached)
> none
> checking for sys/ioctl.h... (cached) yes
> checking if RANDR is defined... yes
> checking if RENDER is defined... yes
> checking if XV is defined... yes
> checking if DPMSExtension is defined... yes
> checking for XORG... yes
> checking for ANSI C header files... (cached) yes
> checking whether to include DRI support... checking for
> /usr/include/xorg/dri.h...
> configure: error: cannot check for file existence when cross compiling
> ... and configure just

> _______________________________________________
> yocto mailing list
> yocto@yoctoproject.org
> https://lists.yoctoproject.org/listinfo/yocto


-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 205 bytes --]

Re: build failure with xf86-video-omap

[Burton, Ross]

On 21 November 2012 22:52, Nicolas Dechesne <ndec13@gmail.com> wrote:
> since xf86-video-omap was recently merged in oe-core and poky, i am trying
> to remove the recipes we had staged in meta-ti trees for some weeks
> (http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf).
> so i was trying to build  poky/master with meta-ti/master, for pandaboard,
> and I get build failures in xf86-video-omap [1] and [2].
>
> so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i can
> send a patch in oe-core for that.

Please do.

> [2] is a bit more cryptic to me.. so I quickly tried to build with danny ( i
> just cherry picked the commit that introduced xf86-video-omap from master)
> and this time it built fine.
>
> has anyone actually tested this recipe? if so I must be doing something
> wrong, any hint would be helpful.

This was discussed a few hours ago.  It's caused by a) the
configure.ac being broken and b) the workaround for this being removed
by me in e828faf027540bc5def8b7371a959884caad9557 (which is in master
only).  Presumably this code was test built before that patch landed.

As this is a genuine upstream bug (ARM-specific driver that doesn't
cross-compile?!) I've just filed fdo #57386.

Ross

Re: [yocto] build failure with xf86-video-omap

[Burton, Ross]

On 21 November 2012 22:52, Nicolas Dechesne <ndec13@gmail.com> wrote:
> since xf86-video-omap was recently merged in oe-core and poky, i am trying
> to remove the recipes we had staged in meta-ti trees for some weeks
> (http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf).
> so i was trying to build  poky/master with meta-ti/master, for pandaboard,
> and I get build failures in xf86-video-omap [1] and [2].
>
> so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i can
> send a patch in oe-core for that.

Please do.

> [2] is a bit more cryptic to me.. so I quickly tried to build with danny ( i
> just cherry picked the commit that introduced xf86-video-omap from master)
> and this time it built fine.
>
> has anyone actually tested this recipe? if so I must be doing something
> wrong, any hint would be helpful.

This was discussed a few hours ago.  It's caused by a) the
configure.ac being broken and b) the workaround for this being removed
by me in e828faf027540bc5def8b7371a959884caad9557 (which is in master
only).  Presumably this code was test built before that patch landed.

As this is a genuine upstream bug (ARM-specific driver that doesn't
cross-compile?!) I've just filed fdo #57386.

Ross

Re: build failure with xf86-video-omap

[Nicolas Dechesne]

[-- Attachment #1: Type: text/plain, Size: 1948 bytes --]

looping rob clark (upstream author)


On Thu, Nov 22, 2012 at 12:07 AM, Burton, Ross <ross.burton@intel.com>wrote:

> On 21 November 2012 22:52, Nicolas Dechesne <ndec13@gmail.com> wrote:
> > since xf86-video-omap was recently merged in oe-core and poky, i am
> trying
> > to remove the recipes we had staged in meta-ti trees for some weeks
> > (
> http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf
> ).
> > so i was trying to build  poky/master with meta-ti/master, for
> pandaboard,
> > and I get build failures in xf86-video-omap [1] and [2].
> >
> > so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i
> can
> > send a patch in oe-core for that.
>
> Please do.
>

sure, will do.

>
> > [2] is a bit more cryptic to me.. so I quickly tried to build with danny
> ( i
> > just cherry picked the commit that introduced xf86-video-omap from
> master)
> > and this time it built fine.
> >
> > has anyone actually tested this recipe? if so I must be doing something
> > wrong, any hint would be helpful.
>
> This was discussed a few hours ago.  It's caused by a) the
> configure.ac being broken and b) the workaround for this being removed
> by me in e828faf027540bc5def8b7371a959884caad9557 (which is in master
> only).  Presumably this code was test built before that patch landed.
>
> As this is a genuine upstream bug (ARM-specific driver that doesn't
> cross-compile?!) I've just filed fdo #57386.
>

i suspect you opened a OE bug, since I can't find it in yocto BZ, however
bugs.openembedded.org doesn't seem to answer at the moment..

so, I checked our configure.ac , and clearly, that needs some love...

Rob: any chance you can look into that? it's the xf86-video-omap
configure.ac file that's wrong, we should mimic what's in the
xf-video-intel one i guess..  otherwise i can try later this week.




> Ross
>

[-- Attachment #2: Type: text/html, Size: 3140 bytes --]

Re: [yocto] build failure with xf86-video-omap

[Nicolas Dechesne]

[-- Attachment #1: Type: text/plain, Size: 1948 bytes --]

looping rob clark (upstream author)


On Thu, Nov 22, 2012 at 12:07 AM, Burton, Ross <ross.burton@intel.com>wrote:

> On 21 November 2012 22:52, Nicolas Dechesne <ndec13@gmail.com> wrote:
> > since xf86-video-omap was recently merged in oe-core and poky, i am
> trying
> > to remove the recipes we had staged in meta-ti trees for some weeks
> > (
> http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf
> ).
> > so i was trying to build  poky/master with meta-ti/master, for
> pandaboard,
> > and I get build failures in xf86-video-omap [1] and [2].
> >
> > so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i
> can
> > send a patch in oe-core for that.
>
> Please do.
>

sure, will do.

>
> > [2] is a bit more cryptic to me.. so I quickly tried to build with danny
> ( i
> > just cherry picked the commit that introduced xf86-video-omap from
> master)
> > and this time it built fine.
> >
> > has anyone actually tested this recipe? if so I must be doing something
> > wrong, any hint would be helpful.
>
> This was discussed a few hours ago.  It's caused by a) the
> configure.ac being broken and b) the workaround for this being removed
> by me in e828faf027540bc5def8b7371a959884caad9557 (which is in master
> only).  Presumably this code was test built before that patch landed.
>
> As this is a genuine upstream bug (ARM-specific driver that doesn't
> cross-compile?!) I've just filed fdo #57386.
>

i suspect you opened a OE bug, since I can't find it in yocto BZ, however
bugs.openembedded.org doesn't seem to answer at the moment..

so, I checked our configure.ac , and clearly, that needs some love...

Rob: any chance you can look into that? it's the xf86-video-omap
configure.ac file that's wrong, we should mimic what's in the
xf-video-intel one i guess..  otherwise i can try later this week.




> Ross
>

[-- Attachment #2: Type: text/html, Size: 3140 bytes --]

Re: build failure with xf86-video-omap

[Burton, Ross]

On 21 November 2012 23:26, Nicolas Dechesne <ndec13@gmail.com> wrote:
>> As this is a genuine upstream bug (ARM-specific driver that doesn't
>> cross-compile?!) I've just filed fdo #57386.
>
> i suspect you opened a OE bug, since I can't find it in yocto BZ, however
> bugs.openembedded.org doesn't seem to answer at the moment..
>
> so, I checked our configure.ac , and clearly, that needs some love...
>
> Rob: any chance you can look into that? it's the xf86-video-omap
> configure.ac file that's wrong, we should mimic what's in the xf-video-intel
> one i guess..  otherwise i can try later this week.

fdo - https://bugs.freedesktop.org/show_bug.cgi?id=57386

I started replacing the AC_CHECK_FILE with PKG_CHECK_EXISTS, but now
I'm wondering why DRI_CFLAGS DRM_CFLAGS are being used without being
apparently defined.  Is this configure.ac needing a bit of a clean or
am I misunderstanding something?

Ross

Re: [yocto] build failure with xf86-video-omap

[Burton, Ross]

On 21 November 2012 23:26, Nicolas Dechesne <ndec13@gmail.com> wrote:
>> As this is a genuine upstream bug (ARM-specific driver that doesn't
>> cross-compile?!) I've just filed fdo #57386.
>
> i suspect you opened a OE bug, since I can't find it in yocto BZ, however
> bugs.openembedded.org doesn't seem to answer at the moment..
>
> so, I checked our configure.ac , and clearly, that needs some love...
>
> Rob: any chance you can look into that? it's the xf86-video-omap
> configure.ac file that's wrong, we should mimic what's in the xf-video-intel
> one i guess..  otherwise i can try later this week.

fdo - https://bugs.freedesktop.org/show_bug.cgi?id=57386

I started replacing the AC_CHECK_FILE with PKG_CHECK_EXISTS, but now
I'm wondering why DRI_CFLAGS DRM_CFLAGS are being used without being
apparently defined.  Is this configure.ac needing a bit of a clean or
am I misunderstanding something?

Ross

Re: build failure with xf86-video-omap

[Burton, Ross]

On 21 November 2012 23:33, Burton, Ross <ross.burton@intel.com> wrote:
> fdo - https://bugs.freedesktop.org/show_bug.cgi?id=57386
>
> I started replacing the AC_CHECK_FILE with PKG_CHECK_EXISTS, but now
> I'm wondering why DRI_CFLAGS DRM_CFLAGS are being used without being
> apparently defined.  Is this configure.ac needing a bit of a clean or
> am I misunderstanding something?

Looked a bit more, looks like 99% of the DRI checking code is dead, so
I've attached a patch to FDO #57386 which mostly deletes stuff.

(caveat: untested)

Ross

Re: [yocto] build failure with xf86-video-omap

[Burton, Ross]

On 21 November 2012 23:33, Burton, Ross <ross.burton@intel.com> wrote:
> fdo - https://bugs.freedesktop.org/show_bug.cgi?id=57386
>
> I started replacing the AC_CHECK_FILE with PKG_CHECK_EXISTS, but now
> I'm wondering why DRI_CFLAGS DRM_CFLAGS are being used without being
> apparently defined.  Is this configure.ac needing a bit of a clean or
> am I misunderstanding something?

Looked a bit more, looks like 99% of the DRI checking code is dead, so
I've attached a patch to FDO #57386 which mostly deletes stuff.

(caveat: untested)

Ross

Re: [meta-ti] build failure with xf86-video-omap

[Denys Dmytriyenko]

On Thu, Nov 22, 2012 at 12:26:12AM +0100, Nicolas Dechesne wrote:
> looping rob clark (upstream author)
> 
> 
> On Thu, Nov 22, 2012 at 12:07 AM, Burton, Ross <ross.burton@intel.com>wrote:
> 
> > On 21 November 2012 22:52, Nicolas Dechesne <ndec13@gmail.com> wrote:
> > > since xf86-video-omap was recently merged in oe-core and poky, i am
> > trying
> > > to remove the recipes we had staged in meta-ti trees for some weeks
> > > (
> > http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf
> > ).
> > > so i was trying to build  poky/master with meta-ti/master, for
> > pandaboard,
> > > and I get build failures in xf86-video-omap [1] and [2].
> > >
> > > so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i
> > can
> > > send a patch in oe-core for that.
> >
> > Please do.
> >
> 
> sure, will do.
> 
> >
> > > [2] is a bit more cryptic to me.. so I quickly tried to build with danny
> > ( i
> > > just cherry picked the commit that introduced xf86-video-omap from
> > master)
> > > and this time it built fine.
> > >
> > > has anyone actually tested this recipe? if so I must be doing something
> > > wrong, any hint would be helpful.
> >
> > This was discussed a few hours ago.  It's caused by a) the
> > configure.ac being broken and b) the workaround for this being removed
> > by me in e828faf027540bc5def8b7371a959884caad9557 (which is in master
> > only).  Presumably this code was test built before that patch landed.
> >
> > As this is a genuine upstream bug (ARM-specific driver that doesn't
> > cross-compile?!) I've just filed fdo #57386.
> >
> 
> i suspect you opened a OE bug, since I can't find it in yocto BZ, however
> bugs.openembedded.org doesn't seem to answer at the moment..

AFAIK, OE bugzilla was deprecated some time ago. We were considering taking it 
down, as a matter of fact. OE-Core issues are logged into Yocto bugzilla. Not 
sure where Ross filed this one though.


> so, I checked our configure.ac , and clearly, that needs some love...
> 
> Rob: any chance you can look into that? it's the xf86-video-omap
> configure.ac file that's wrong, we should mimic what's in the
> xf-video-intel one i guess..  otherwise i can try later this week.

-- 
Denys

Re: [yocto] build failure with xf86-video-omap

[Denys Dmytriyenko]

On Thu, Nov 22, 2012 at 12:26:12AM +0100, Nicolas Dechesne wrote:
> looping rob clark (upstream author)
> 
> 
> On Thu, Nov 22, 2012 at 12:07 AM, Burton, Ross <ross.burton@intel.com>wrote:
> 
> > On 21 November 2012 22:52, Nicolas Dechesne <ndec13@gmail.com> wrote:
> > > since xf86-video-omap was recently merged in oe-core and poky, i am
> > trying
> > > to remove the recipes we had staged in meta-ti trees for some weeks
> > > (
> > http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/commit/?id=410dc026f2ee24a2346e7563a83f0181c79809cf
> > ).
> > > so i was trying to build  poky/master with meta-ti/master, for
> > pandaboard,
> > > and I get build failures in xf86-video-omap [1] and [2].
> > >
> > > so [1] is actually trivial, as we are missing a DEPENDS in the recipe, i
> > can
> > > send a patch in oe-core for that.
> >
> > Please do.
> >
> 
> sure, will do.
> 
> >
> > > [2] is a bit more cryptic to me.. so I quickly tried to build with danny
> > ( i
> > > just cherry picked the commit that introduced xf86-video-omap from
> > master)
> > > and this time it built fine.
> > >
> > > has anyone actually tested this recipe? if so I must be doing something
> > > wrong, any hint would be helpful.
> >
> > This was discussed a few hours ago.  It's caused by a) the
> > configure.ac being broken and b) the workaround for this being removed
> > by me in e828faf027540bc5def8b7371a959884caad9557 (which is in master
> > only).  Presumably this code was test built before that patch landed.
> >
> > As this is a genuine upstream bug (ARM-specific driver that doesn't
> > cross-compile?!) I've just filed fdo #57386.
> >
> 
> i suspect you opened a OE bug, since I can't find it in yocto BZ, however
> bugs.openembedded.org doesn't seem to answer at the moment..

AFAIK, OE bugzilla was deprecated some time ago. We were considering taking it 
down, as a matter of fact. OE-Core issues are logged into Yocto bugzilla. Not 
sure where Ross filed this one though.


> so, I checked our configure.ac , and clearly, that needs some love...
> 
> Rob: any chance you can look into that? it's the xf86-video-omap
> configure.ac file that's wrong, we should mimic what's in the
> xf-video-intel one i guess..  otherwise i can try later this week.

-- 
Denys