All of lore.kernel.org
 help / color / mirror / Atom feed
From: Alexandre Belloni <alexandre.belloni@free-electrons.com>
To: "meta-freescale@yoctoproject.org"
	<meta-freescale@yoctoproject.org>,
	 poky@yoctoproject.org
Subject: X11 incompatibilities using meta-oe with poky
Date: Fri, 27 Dec 2013 22:22:43 +0100	[thread overview]
Message-ID: <52BDEFA3.2090106@free-electrons.com> (raw)

Hi,

When using core-image-sato (latest dora), I encountered again an issue
that I once solved by hacking around it. It has the following symptoms:
 - touchscreen calibration doesn't start the first time the board is booted
 - matchbox-keyboard doesn't come up when touching the keyboard icon in
the panel
 - matchbox-panel is ugly (compare
http://www.ossystems.com.br/images/blog/boards/boundarydevices-nitrogen6x-hdmi-screenshot-fsl-image-gui.jpg
with http://free-electrons.com/~alexandre/matchbox-panel.jpg)

This time, I decided to investigate that issue a bit more. Here is what
I found:
The cause is the /etc/X11/Xsession script which at some point does:
==

What is really happening is that xserver-nodm-init is provided both by
poky/meta/recipes-graphics/x11-common/xserver-nodm-init.bb and
meta-openembedded/meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init_2.0.bb

layer                 path                                      priority
==========================================================================
meta                  /home/alex/fsl/sources/poky/meta  5
meta-yocto            /home/alex/fsl/sources/poky/meta-yocto  5
meta-oe               /home/alex/fsl/sources/meta-openembedded/meta-oe  6

The meta-oe one is selected ! That one can't actually work with the
current
poky/meta/recipes-sato/matchbox-keyboard/matchbox-keyboard_git.bb recipe
because of the "ls -X" sorting and it doesn't know how to handle shbg
files this will hang trying to use 80matchboxkeyboard.shbg

So my next step was to remove the meta-oe layer (or simply mask
/meta-openembedded/meta-oe/recipes-graphics/). The final result is
better but touch screen calibration (xtscal instead of
xinput-calibrator) fails with:
XCALIBRATE extension missing: Resource temporarily unavailable
XCALIBRATE extension missing: Resource temporarily unavailable
XCALIBRATE extension missing: Resource temporarily unavailable


So, questions:
 - How do we solve that cleanly ?
 - Why doesn't xserver-common conflicts with x11-common despite
RCONFLICTS_${PN} = "xserver-kdrive-common x11-common" ?
 - Shouldn't we stop using xtscal and use xinput-calibrator instead ?
 - Should we update the Xsession script from the poky recipes and stop
using x11-common ?

On my side, what I do is masking
/meta-openembedded/meta-oe/recipes-graphics/ and replacing xtscal by
xinput-calibrator as a dependency of x11-common but I believe we can do
better.

-- 
Alexandre Belloni, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com



             reply	other threads:[~2013-12-27 21:22 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-12-27 21:22 Alexandre Belloni [this message]
2013-12-28  8:32 ` X11 incompatibilities using meta-oe with poky Martin Jansa
2013-12-28  8:32   ` [poky] " Martin Jansa
2013-12-28 10:19   ` Alexandre Belloni
2013-12-28 10:19     ` [poky] " Alexandre Belloni
2013-12-28 11:28     ` Martin Jansa
2013-12-28 11:28       ` [poky] " Martin Jansa
2013-12-28 11:43       ` Alexandre Belloni
2013-12-28 11:43         ` [poky] " Alexandre Belloni
2013-12-28 12:07         ` Martin Jansa
2013-12-28 12:07           ` [poky] " Martin Jansa
2013-12-28 13:00           ` Alexandre Belloni
2013-12-28 13:00             ` [poky] " Alexandre Belloni
2013-12-30  5:09             ` Otavio Salvador
2013-12-31 10:42               ` [meta-freescale] " Alexandre Belloni
2013-12-31 10:42                 ` [poky] " Alexandre Belloni
2013-12-31 11:42                 ` [meta-freescale] " Martin Jansa
2013-12-31 11:42                   ` [poky] " Martin Jansa

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=52BDEFA3.2090106@free-electrons.com \
    --to=alexandre.belloni@free-electrons.com \
    --cc=meta-freescale@yoctoproject.org \
    --cc=poky@yoctoproject.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.