mailarchive of the ptxdist mailing list
 help / color / mirror / Atom feed
From: Michael Olbrich <m.olbrich@pengutronix.de>
To: ptxdist@pengutronix.de
Subject: Re: [ptxdist] Mesa DRI drivers path
Date: Mon, 2 Jun 2014 10:27:32 +0200	[thread overview]
Message-ID: <20140602082732.GL26228@pengutronix.de> (raw)
In-Reply-To: <CAEPGbJT=ZEWj-P3pGMj5jQ8akehPV49qT0mrbLps3kennw-uYA@mail.gmail.com>

On Mon, May 19, 2014 at 02:19:02PM +0200, Roger Karis wrote:
> while building an i568-based linux distribution with PTXdist 2014.4 we
> discovered a deployment problem involving mesalib and xorg-server.
> 
> The xorg-server package uses pkg-config during the prepare phase to
> determine the absolute path where DRI drivers are installed. The path
> is stored in the variable 'dridriverdir' in dri.pc.
> 
> In previous versions of mesalib (e.g. 8.0.3), this path variable was
> fully expanded at package preparation time (for example to
> /usr/lib/dri). In the current packaged version (10.1.0), this path is
> not fully expanded but remains relative to the install prefix (which
> is set to /usr at prepare). At install time the prefix is replaced by
> ${pcfiledir}/../.. which makes the dridriverdir relative to the
> sysroot-target folder.
> 
> When starting the xorg-server on the target, this means that the
> drivers will be loaded from an absolute path using
> .../platform-foo/sysroot-target/.... Because the drivers are located
> at /usr/lib/dri, this fails with a AIGLX dlopen() error.
> 
> The workaround we currently employ is to explicitly specify the driver
> directory (--with-dri-driverdir) during mesalib preparation, but of
> course it would be much better if there was a proper solution to this
> problem.

I think setting --with-dri-driverdir explicitly is actually the only sane
way to handle this. The problem ist that pkg-config cannot really handle
sysroot/destdir. Take look at dri.pc there is 'includedir'. That will
probably be uses as '-I$(includedir)' so it needs to bee
'$(SYSROOT)/something'. On the other hand 'dridriverdir' probably used
while installing: "... $(DESTDIR)$(dridriverdir) ..." so it must not
contain '$(SYSROOT)' as you noticed. I've come across cases where such a
variable is used both ways. This is usually where really ugly patches start
to appear. Setting "--with-dri-driverdir" with an appropriate comment is
a rather clean and simple fix in comparison.

Michael

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

-- 
ptxdist mailing list
ptxdist@pengutronix.de

      reply	other threads:[~2014-06-02  8:27 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-05-19 12:19 Roger Karis
2014-06-02  8:27 ` Michael Olbrich [this message]

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=20140602082732.GL26228@pengutronix.de \
    --to=m.olbrich@pengutronix.de \
    --cc=ptxdist@pengutronix.de \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox