From: Denys Dmytriyenko <denys@ti.com>
To: "Robert P. J. Day" <rpjday@crashcourse.ca>
Cc: meta-ti mailing list <meta-ti@yoctoproject.org>
Subject: Re: curious about libdrm-2.4.41 recipe from meta-ti layer
Date: Fri, 18 Jul 2014 11:37:47 -0400 [thread overview]
Message-ID: <20140718153747.GF15788@edge> (raw)
In-Reply-To: <alpine.LFD.2.11.1407180737270.9890@localhost>
On Fri, Jul 18, 2014 at 07:46:36AM -0400, Robert P. J. Day wrote:
>
> i was perusing a number of yocto layers looking for cool examples to
> use in an upcoming class and ran across this oddity in the meta-ti
> layer:
>
> recipes-graphics/drm/libdrm_2.4.41.bb:
> FILESEXTRAPATHS_append := ":${COREBASE}/meta/recipes-graphics/drm/libdrm"
>
> in all of the layers i've looked at, that is the only example of
> *appending* to FILESEXTRAPATHS i've ever seen, so i thought i'd look
> closer to see what was going on.
>
> even though oe-core defines a libdrm recipe file, the meta-ti layer
> doesn't use a bbappend for its version. rather, it explicitly adds the
> libdrm/ directory from oe-core using a reference to ${COREBASE}, then
> adds its own files. that's valid, i guess, it just looks weird and is
> the only time i've ever seen that.
>
> i'm also puzzled that the meta-ti recipe sets:
>
> DEFAULT_PREFERENCE = "-1"
>
> what's that for? given that the oe-core layer already defines a
> recipe for libdrm that has a higher version number, why would the
> above line be necessary?
>
> just trying to understand the rationale behind this unusual
> construction.
You missed line #11 with custom SRC_URI, which points to a forked and
modified sources of a very specific version of libdrm for one platform.
That's why it sets COMPATIBLE_MACHINE and DEFAULT_PREFERENCE - it needs to be
requested specifically and not selected automatically by accident. By default
we want to use the latest upstream version from oe-core.
And it cannot be made as a bbappend - it provides a different version than
oe-core. But it reuses GNU_SOURCE_definition.patch from oe-core, hence
FILESEXTRAPATHS mangling.
Don't ask why it's done with forked git sources instead of patches though -
not my decision :)
--
Denys
prev parent reply other threads:[~2014-07-18 15:38 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-07-18 11:46 curious about libdrm-2.4.41 recipe from meta-ti layer Robert P. J. Day
2014-07-18 15:30 ` Maupin, Chase
2014-07-18 15:44 ` Denys Dmytriyenko
2014-07-18 15:37 ` Denys Dmytriyenko [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=20140718153747.GF15788@edge \
--to=denys@ti.com \
--cc=meta-ti@yoctoproject.org \
--cc=rpjday@crashcourse.ca \
/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.