Re: [oe] Fedora14 and glibc-initial
On 09/12/2010 06:21 PM, Holger Freyther wrote: Hi, has anyone else tried to build glibc-initial with Fedora14/rawhide? I am having a build issue. At first make install tries to copy something to /usr/local/include/sys/limits.h and now it is complaining about the Makefile having mixed mode: Makefile:235: *** mixed implicit and normal rules. Stop. the first error vanished (maybe fixed upstream?), I have fixed the Makefile error in glibc-2.9 and busybox-1.13.2 and now... Is anyone else having build issues with procps? ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
[oe] Question regarding the staging layout
Hi all, I am seeing a lot of messages like: /home/ich/build/arm/xscale/tmp/sysroots/i686-linux/usr/armv5te/lib/gcc/arm-angstrom-linux-gnueabi/4.3.3/../../../../arm-angstrom-linux-gnueabi/bin/ld: warning: libc.so, needed by /home/ich/build/arm/xscale/tmp/sysroots/i686-linux/usr/armv5te/lib/gcc/arm-angstrom-linux-gnueabi/4.3.3/../../../../arm-angstrom-linux-gnueabi/lib/libgcc_s.so, not found (try using -rpath or -rpath-link) is that to be expected? ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
[oe] svn://svn.om.vptt.ch issue
Not sure if this is permanent but the following 5 recipes did not fetch for me: /home/frans/oe/openembedded/recipes/e17/bubble-keyboard_svn.bb /home/frans/oe/openembedded/recipes/e17/essential-dialer_svn.bb /home/frans/oe/openembedded/recipes/e17/gridpad_svn.bb /home/frans/oe/openembedded/recipes/e17/intuition_svn.bb /home/frans/oe/openembedded/recipes/efl1/libefso_svn.bb These all fetch from svn.om.vptt.ch Anyone any idea on this? Frans ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
[oe] COMPATIBLE_MACHINE and no buildable providers
Bumped upon the folliwing issue. If a recipe depends on another recipe but that one does have a COMPATIBLE_MACHINE and you are not in the list, you get an error like: ERROR: Required build target xyz' has no buildable providers. Is this desired? Or in other words: should this be fixed by adding the COMPATIBLE_MACHINE clause to the recipe that has the depends? Or should COMPATIBLE_MACHINE be allowed to trickle up so when building a recipe that depends on another recipe that is not for that MACHINE one only gets a warning. This is for instance an issue when doing bitbake world. Frans ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] OEDEM 2010 - Confirmed
Hello. On Sat, 2010-09-18 at 13:33, Leon Woestenberg wrote: On Wed, Sep 15, 2010 at 6:56 PM, Florian Boor florian.b...@kernelconcepts.de wrote: Members and non-members of the OE e.V. are welcome. It is likely that an e.V. general assembly will be held during OEDEM but that is for the e.V. board to decide and announce. We will do this. this = Decide or announce? Triggered by Philip's remark on the six week deadline: The six week deadline has slipped yesterday if the assembly is on the 29th. The six week deadline is today if the assembly is on the 30th. It was announced to the members mailing list in time. And Leon found it there after this mail. Just to make sure other people are not confused about this. It is also noted on the Agenda for OEDEM 2010: http://wiki.openembedded.net/index.php/Oedem/2010#Friday regards Stefan Schmidt ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] svn://svn.om.vptt.ch issue
Hello. On Sat, 2010-09-18 at 13:28, Frans Meulenbroeks wrote: Not sure if this is permanent but the following 5 recipes did not fetch for me: /home/frans/oe/openembedded/recipes/e17/bubble-keyboard_svn.bb /home/frans/oe/openembedded/recipes/e17/essential-dialer_svn.bb /home/frans/oe/openembedded/recipes/e17/gridpad_svn.bb /home/frans/oe/openembedded/recipes/e17/intuition_svn.bb /home/frans/oe/openembedded/recipes/efl1/libefso_svn.bb These all fetch from svn.om.vptt.ch Anyone any idea on this? I just checked back with the guys hosting the svn and it should be available. He promised me to check back with the admin on monday. So please keep the recipes. regards Stefan Schmidt ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] svn://svn.om.vptt.ch issue
2010/9/18 Stefan Schmidt ste...@datenfreihafen.org: Hello. On Sat, 2010-09-18 at 13:28, Frans Meulenbroeks wrote: Not sure if this is permanent but the following 5 recipes did not fetch for me: /home/frans/oe/openembedded/recipes/e17/bubble-keyboard_svn.bb /home/frans/oe/openembedded/recipes/e17/essential-dialer_svn.bb /home/frans/oe/openembedded/recipes/e17/gridpad_svn.bb /home/frans/oe/openembedded/recipes/e17/intuition_svn.bb /home/frans/oe/openembedded/recipes/efl1/libefso_svn.bb These all fetch from svn.om.vptt.ch Anyone any idea on this? I just checked back with the guys hosting the svn and it should be available. He promised me to check back with the admin on monday. So please keep the recipes. Sure. It was not a removal request, just trying to find out what was wrong. Thanks for the info! Frans. ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] [PATCH] clutter: moved COMPATIBLE_MACHINE to clutter-common.inc
On Saturday 18 September 2010 01:47:08 pm Frans Meulenbroeks wrote: moved COMPATIBLE_MACHINE from clutter.inc to clutter-common.inc the other recipes all DEPEND on clutter but do not have a COMPATIBLE_MACHINE defined, but of course do not work for other machines. This patch moves COMPATIBLE_MACHINE to clutter-common.inc which is included by all other inc file and so it applies to all Is there any fallout expected by removing COMPATIBLE_MACHINE? All machines not having an explicit DEPENDS / EXTRA_OECONF get virtual/libgl (mesa) as depends and --with-flavour=glx as defaults which should build for all (?). If there is no damage caused by removing COMPATIBLE_MACHINE there is no need to enter (new) machines here in the future.. Andreas ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] OEDEM 2010 - Confirmed
Hello, On Sat, Sep 18, 2010 at 1:53 PM, Stefan Schmidt ste...@datenfreihafen.org wrote: It was announced to the members mailing list in time. And Leon found it there after this mail. Just to make sure other people are not confused about this. It is also noted on the Agenda for OEDEM 2010: http://wiki.openembedded.net/index.php/Oedem/2010#Friday Indeed, thanks. In that thread I asked for it to be added to the news items of the wiki. For people with -EINFOOVERLOAD :-) Regards, -- Leon ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] [PATCH] xkbd_0.8.15.bb: Fix geometry flag to accept negative offsets.
On 09/17/2010 01:51 PM, Graham Gower wrote: Where is this patch coming from? is it upstream? ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] Fedora14 and glibc-initial
On Sat, Sep 18, 2010 at 2:55 AM, Holger Freyther holger...@freyther.de wrote: On 09/12/2010 06:21 PM, Holger Freyther wrote: Hi, has anyone else tried to build glibc-initial with Fedora14/rawhide? I am having a build issue. At first make install tries to copy something to /usr/local/include/sys/limits.h and now it is complaining about the Makefile having mixed mode: Makefile:235: *** mixed implicit and normal rules. Stop. the first error vanished (maybe fixed upstream?), I have fixed the Makefile error in glibc-2.9 and busybox-1.13.2 and now... Is anyone else having build issues with procps? is F14 using 3.82 make ? may be not many have moved to make 3.82 yet. ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel -- -Khem ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] Fedora14 and glibc-initial
On 09/19/2010 12:19 AM, Khem Raj wrote: is F14 using 3.82 make ? may be not many have moved to make 3.82 yet. yes ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
[oe] python-tlslite
Hi, python-tlslite does not fetch any more. When fetched a html page is returned saying h3Error: Page cannot be displayed. Please contact service provider for more details. (3)/h3 And the url http://trevp.net/tlslite gives a generic page like you get with some of those domain resellers. python-tlslite is defined in python/python-tlslite_0.3.8.bb:HOMEPAGE = http://trevp.net/tlslite/; python/python-tlslite_0.3.8.bb:SRCNAME = tlslite and used in: tasks/task-mamona-python.bb:python-tlslite \ tasks/task-python-everything.bb: python-tlslite \ tasks/task-python-sharprom_20060425.bb: python-tlslite \ suggestion: move the recipe to nonworking and take it out of the tasks. Frans. ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] python-tlslite
On 09/19/2010 02:49 AM, Frans Meulenbroeks wrote: python/python-tlslite_0.3.8.bb:HOMEPAGE = http://trevp.net/tlslite/; python/python-tlslite_0.3.8.bb:SRCNAME = tlslite Did you consider using a Gentoo/Debian mirror for the version 0.3.8. At least Gentoo is still having this recipe. ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] python-tlslite
2010/9/18 Holger Freyther holger...@freyther.de: On 09/19/2010 02:49 AM, Frans Meulenbroeks wrote: python/python-tlslite_0.3.8.bb:HOMEPAGE = http://trevp.net/tlslite/; python/python-tlslite_0.3.8.bb:SRCNAME = tlslite Did you consider using a Gentoo/Debian mirror for the version 0.3.8. At least Gentoo is still having this recipe. Good plan, will do. Frans ___ Openembedded-devel mailing list Openembedded-devel@lists.openembedded.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
Re: [oe] [PATCH (resend)] image.bbclass: implement IMAGE_FEATURES
On 08/03/2010 09:40 PM, Chris Larson wrote:
From: Christopher Larsonclar...@mvista.com
This adds the ability to include groups of packages in an image. It is
similar in concept to the IMAGE_FEATURES used in Poky, but here we use
IMAGE_FEATURE_feature and IMAGE_FEATURE_feature_OPTIONAL, rather than
base_contains calls.
Initially defined groups:
base - the packages requested via IMAGE_INSTALL IMAGE_BOOT
dbg - debug packages
dev - development packages
doc - documentation packages (of limited use at the moment due to task.bbclass)
vmlinux - kernel debugging packages (just kernel-vmlinux at the moment)
To add groups to your images:
IMAGE_FEATURES += dbg dev
To add a new package group:
IMAGE_FEATURE_myfeature = gstreamer python pulseaudio
Signed-off-by: Chris Larsonclar...@mvista.com
Acked-by: Philip Balister phi...@balister.org
Works for me. I'm using it to install and the -dbg packages in an image
so I can oprofile better.
Philip
---
classes/image.bbclass | 101 -
1 files changed, 99 insertions(+), 2 deletions(-)
diff --git a/classes/image.bbclass b/classes/image.bbclass
index 2105195..30137be 100644
--- a/classes/image.bbclass
+++ b/classes/image.bbclass
@@ -29,11 +29,10 @@ ${IMAGE_DEV_MANAGER} \
${IMAGE_INIT_MANAGER} \
${IMAGE_LOGIN_MANAGER}
-RDEPENDS += ${IMAGE_INSTALL} ${IMAGE_BOOT}
+RDEPENDS += ${__PACKAGE_INSTALL}
# export IMAGE_BASENAME not supported at this time
IMAGE_BASENAME[export] = 1
-export PACKAGE_INSTALL ?= ${IMAGE_INSTALL} ${IMAGE_BOOT}
# We need to recursively follow RDEPENDS and RRECOMMENDS for images
do_rootfs[recrdeptask] += do_deploy do_populate_sysroot
@@ -61,6 +60,7 @@ python () {
bb.data.setVarFlag('do_rootfs', 'depends', deps, d)
runtime_mapping_rename(PACKAGE_INSTALL, d)
+runtime_mapping_rename(PACKAGE_INSTALL_ATTEMPTONLY, d)
}
#
@@ -107,6 +107,103 @@ IMAGE_LINGUAS ?= de-de fr-fr en-gb
LINGUAS_INSTALL = ${@ .join(map(lambda s: locale-base-%s % s,
bb.data.getVar('IMAGE_LINGUAS', d, 1).split()))}
+def __pkgmap(d):
+from os import listdir
+from os.path import isdir, join, dirname
+
+os = d.getVar(TARGET_OS, True)
+vendor = d.getVar(TARGET_VENDOR, True)
+basedir = dirname(d.getVar(PKGDATA_DIR, True))
+
+dirs = (%s%s-%s % (arch, vendor, os)
+for arch in d.getVar(PACKAGE_ARCHS, True).split())
+
+for pkgdatadir in (join(basedir, sys) for sys in dirs):
+try:
+files = listdir(pkgdatadir)
+except OSError:
+continue
+
+for pn in filter(lambda f: not isdir(join(pkgdatadir, f)), files):
+try:
+pkgdata = read_pkgdatafile(join(pkgdatadir, pn))
+except OSError:
+continue
+
+for pkg in pkgdata[PACKAGES].split():
+yield pkg, pn
+
+def pkgmap(d):
+pkgmap_data = d.getVar(__pkgmap_data, False)
+if pkgmap_data is None:
+pkgmap_data = dict(__pkgmap(d))
+d.setVar(__pkgmap_data, pkgmap_data)
+
+return pkgmap_data
+
+def recipename(pkg, d):
+return pkgmap(d).get(pkg)
+
+def pkgjoin(l):
+return ' '.join(set(l))
+
+def pkgsplit(var, d):
+return (d.getVar(var, True) or ).split()
+
+def __packages(features, d):
+for feature in features:
+for pkg in pkgsplit('IMAGE_FEATURE_%s' % feature, d):
+yield pkg
+
+def required_packages(features, d):
+for pkg in __packages((f for f in features if not
d.getVar(IMAGE_FEATURE_%s_OPTIONAL % f, True)), d):
+yield pkg
+
+def optional_packages(features, d):
+for pkg in __packages((f for f in features if
d.getVar(IMAGE_FEATURE_%s_OPTIONAL % f, True)), d):
+yield pkg
+
+def active_packages(features, d):
+from itertools import chain
+
+for pkg in chain(required_packages(features, d),
optional_packages(features, d)):
+yield pkg
+
+def active_recipes(features, d):
+for pkg in active_packages(features, d):
+recipe = recipename(pkg, d)
+if recipe:
+yield recipe
+
+def image_features_noextras(d):
+for f in d.getVar(IMAGE_FEATURES, True).split():
+if not f in ('dbg', 'dev', 'doc'):
+yield f
+
+def dbg_packages(d):
+from itertools import chain
+
+features = image_features_noextras(d)
+return pkgjoin(%s-dbg % pkg for pkg in chain(active_packages(features,
d), active_recipes(features, d)))
+
+__PACKAGE_INSTALL = ${@' '.join(required_packages(d.getVar('IMAGE_FEATURES',
True).split(), d))}
+PACKAGE_INSTALL = ${__PACKAGE_INSTALL}
+PACKAGE_INSTALL_ATTEMPTONLY = ${@'
'.join(optional_packages(d.getVar('IMAGE_FEATURES', True).split(), d))}
+
+IMAGE_FEATURES ?=
+IMAGE_FEATURES_prepend = base
+
+# Define our default available image features
+IMAGE_FEATURE_base = ${IMAGE_INSTALL} ${IMAGE_BOOT}
+IMAGE_FEATURE_vmlinux = kernel-vmlinux
+
+IMAGE_FEATURE_dbg = $...@dbg_packages(d)}
+IMAGE_FEATURE_dbg_OPTIONAL = 1
+IMAGE_FEATURE_dev =
