All of lore.kernel.org
 help / color / mirror / Atom feed
From: Trevor Woerner <twoerner@gmail.com>
To: yocto@yoctoproject.org
Subject: overzealous packagedata
Date: Thu, 20 Apr 2017 03:26:09 -0400	[thread overview]
Message-ID: <20170420072609.GB25443@linux-uys3> (raw)

I'm trying to build a certain combination of software for a given SoC. mesa
(meta-gl, specifically) provides a libgbm, but if you don't want to use mesa's
libgbm you can disable it with a PACKAGECONFIG. I.e.

	PACKAGECONFIG_remove_pn-mesa-gl = "gbm"

It works, and looking at the packages-split I can verify that nothing gets
installed for either mesa's libgbm nor libgbm-dev.

This particular SoC's BSP, however, provides an SoC-specific libgbm that
should be used in place of mesa's. This is easy, set the PREFERRED_PROVIDER,
tell mesa you don't want its gbm and everything is good, right?

:-(

The build fails because the *packagedata* from both mesa-gl and this SoC's
specific libgbm collide. The packagedata, I guess, states what *might* be
installed. So since both packages might install /usr/lib/libgbm, the build
freaks out:

	ERROR: mesa-gl-2_17.0.2-r0 do_packagedata_setscene: The recipe mesa-gl is trying to install files into a shared area when those files already exist. Those files and their manifest location are:
	  /z/build-master/bbb/build/tmp-glibc/pkgdata/beaglebone/runtime/libgbm-dev
	    (matched in manifest-beaglebone-libgbm.packagedata)
	  /z/build-master/bbb/build/tmp-glibc/pkgdata/beaglebone/runtime/libgbm
	    (matched in manifest-beaglebone-libgbm.packagedata)
	Please verify which recipe should provide the above files.

This is very annoying because, even if hell froze over and somehow both libgbm
and mesa-gl installed both of their libgbm packages, there would actually only
be one provider since mesa-gl's libgbm package is, in fact, empty.

Is there a way to get this to work? I can't not build mesa-gl, because I need
some of its parts (virtual/mesa, virtual/libgl) and I can't not use the SoC's
libgbm because (I'm guessing) that's what I'm supposed to use.


             reply	other threads:[~2017-04-20  7:26 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-04-20  7:26 Trevor Woerner [this message]
2017-04-20  7:30 ` overzealous packagedata Trevor Woerner
2017-04-20 16:48 ` Khem Raj

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=20170420072609.GB25443@linux-uys3 \
    --to=twoerner@gmail.com \
    --cc=yocto@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.