Age | Commit message (Collapse) | Author | Files |
|
|
|
None of the distros in OE appear to want/need gpg or curl support in opkg;
so have opkg.inc remove them by default. This diff makes things more consistent
across all distros/tasks/images.
The slugos recipe is kept intact and further divided from the more modern
recipes. The opkg-native recipe is the sane-srcrev version even when building
slugos, which should allow for changes to opkg related bbclasses without
breaking the slugos build.
Build tested for qemumipsel/minimal-image and nslu2/slugos-image.
minimal-image builds 10% quicker in my setup and is 1mb smaller.
Signed-off-by: Graham Gower <graham.gower@gmail.com>
Acked-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
This patch removes all occurrences of
do_stage() {
autotools_stage_all
}
including all kind of variants w.r.t whitespace
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
Acked-by: Koen Kooi <koen@openembedded.org>
Acked-by: Michael 'Mickey' Lauer <mlauer@vanille-media.de>
|
|
virtual/opkg-dev"
This reverts commit aa45596db92a18851de6f2f241e578047f488307.
Per request from Koen Kooi.
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
Acked-by: Graeme Gregory <dp@xora.org.uk>
Acked-by: Philip Balister <philip@balister.org>
|
|
This will be used in task-sdk-base (instead of ipkg-dev). Also bump all
PRs to force rebuild.
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
|
|
RDEPENDS/RRECOMMENDS/RSUGGESTS/RPROVIDES/RCONFLICTS/RREPLACES cleanup
* only 769 from 1102 recipes were covered by automatic PR bump, list
of remaining 333 recipes not using PR/INC_PR is available for manual
PR add or INC_PR conversion - volunteers?
http://build.shr-project.org/tests/jama/files.not.PR.bumped
* just echo 'PR = "r1"' to recipes without any PR/INC_PR is not so good
idea without manual check, because sometimes they rely on PR set in
their .inc file and "r1" could make it go backwards
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
* there is probably very few cases when it's needed without _${PN},
if you find one in attached patch, please enlighten me
* usually we want those only for ${PN} not for all ${PACKAGES} ie ${PN}-dev, ${PN)-doc etc
* there is a check for this in recipe-sanity.bclass
* there is maybe few possible fixes also in bbclasses
cpan-base.bbclass:RDEPENDS += "${@["perl", ""][(bb.data.inherits_class('native', d))]}"
distutils-base.bbclass:RDEPENDS += "python-core"
gtk-icon-cache.bbclass:RDEPENDS += "hicolor-icon-theme"
image.bbclass:RDEPENDS += "${IMAGE_INSTALL} ${IMAGE_BOOT}"
module.bbclass:RDEPENDS += "kernel (${KERNEL_VERSION}) update-modules"
?opie_i18n.bbclass: rdepends = bb.data.getVar("RDEPENDS", d, 1)
?opie_i18n.bbclass: bb.data.setVar('RDEPENDS_%s' % x[0].lower(), pn, d)
?opie_i18n.bbclass: bb.data.setVar('RDEPENDS', rdepends, d)
rubyextension.bbclass:RDEPENDS += "ruby"
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Acked-by: Koen Kooi <k-kooi@ti.com>
|
|
For .diff/.patch you need to apply manually, you can specify apply=no.
Signed-off-by: Chris Larson <chris_larson@mentor.com>
Acked-by: Denys Dmytriyenko <denis@denix.org>
Acked-by: Khem Raj <raj.khem@gmail.com>
|
|
I think this makes the behavior rather more clear.
Signed-off-by: Chris Larson <chris_larson@mentor.com>
Acked-by: Denys Dmytriyenko <denis@denix.org>
Acked-by: Khem Raj <raj.khem@gmail.com>
|
|
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
Fixes the problem, introduced by 8c1731d818c89a291f80940c1912587ff9ce46b7,
where openssl is added with DEPENDS_append override, preventing -nogpg and
-nocurl variants from clearing the DEPENDS, when they include opkg_svn.bb
file, thus forcing them to depend on openssl too.
Signed-off-by: Denys Dmytriyenko <denis@denix.org>
Acked-by: Chris Larson <clarson@kergoth.com>
|
|
|
|
* if there is link /usr/lib/ipkg -> /usr/lib/opkg, then merger will
create backups of "both" alternatives directories, but they will be
both the same /usr/lib/opkg/alternatives, then it will try to merge
with output that every file is the same, sofar no real harm.
* Problem is after successfull merge, when it removes
/usr/lib/ipkg/alternatives which is in this case the right
alternatives in opkg :/
* If you've already this problem on your target then
* # Restore old alternatives
* rm -rf /usr/lib/opkg/alternatives
* cp -ra /usr/lib/opkg/alternatives-backup /usr/lib/opkg/alternatives
*
* # Regenerate links with right alternatives available
* opkg install -force-reinstall busybox
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
Won't build without it, so no PR bump.
|
|
instead of /usr/lib/ipkg/alternatives
* It's usefull to make it compatible with u-a script in opkg package
* Old entries are merged to new directory in quite verbose postinst
script
* If entry exists only in old it's moved
* If entry exists in both the one with more lines is used
* The one with less lines is not used and warning is shown
* If they have the same number of lines diff is checked
* If they are the same, old one is ignored
* If they are different, old one is ignored and warning is shown
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
provider in sane-feed-ipk
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
This is a temporary measure made necessary by a major set of changes to
the entire set of opkg recipes that not only rendered older versions of
opkg sources unbuildable, it also removed a patch that has not yet been
accepted by opkg upstream. Until these issues can be resolved and
adequate testing on small-memory devices performed, this copy of the
original recipe is a necessary evil that just has to be tolerated.
|
|
|
|
* nylon uses a gcc3-based toolchain which does not support this option, thus added an override
|
|
* opkg_svn can only be built if openssl is staged, this revision adds the missing dependency
|
|
* -native is left alone because packaged-staging needs more testing
* tested install and upgrade on angstrom/beagleboard
* tweaked opkg-nogpg-nocurl recipe to keep it building
|
|
* Just formal change
* No need to bump PE or PR
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
Also fixes a minor memory leak while we're at it, as the code was
strdup'ing the malloc'd string read from stdin without ever
freeing the original.
Signed-off-by: Chris Larson <clarson@mvista.com>
|
|
|
|
Otherwise it does not find patches stored under opkg directory
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
Signed-off-by: Chris Larson <clarson@mvista.com>
|
|
Signed-off-by: Chris Larson <clarson@mvista.com>
|
|
update-alternatives
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
automatically for opkg-nogpg).
|
|
recipes/opkg/opkg*.bb : bump PR for affected recipes
|
|
|
|
this abstracts psplash to be generic. now as long as something provides a
splashfuncs file that sysvinit (and other scripts) can source, and all the
approproate init hooks to start the splash etc. you can use psplash,
exquisitie, usplash or anything that tickles your fancy. this moves splash
toa ${SPLASH} variable to include in your image (or override). the default in
angstrom is psplash - unless you override it. opkg also runs a configure
script that cna take forever - and so this speically sends off some splash
commands (if there) to let you know the systme is alive and working (but just
busy).
this is one commit as if this breaks things you either want to fix the minor
break or totally revert the whole patch. i hope it didn't break anything.
|
|
See links below for more details:
http://thread.gmane.org/gmane.comp.handhelds.openembedded/21326
http://thread.gmane.org/gmane.comp.handhelds.openembedded/21816
Signed-off-by: Denys Dmytriyenko <denis@denix.org>
Acked-by: Mike Westerhof <mwester@dls.net>
Acked-by: Philip Balister <philip@balister.org>
Acked-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Marcin Juszkiewicz <hrw@openembedded.org>
Acked-by: Koen Kooi <koen@openembedded.org>
Acked-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|