python 2.7.2 make test under bitbake fails

Hi,

I'm building python 2.7.2 and everything goes well except make test.
There are several failed UTs, which however succeed ran via shell.
Shallow investigation showed that it's because the way bitbake runs
tasks (child processes with input output customization).

So the questions are:
1) Does anybody else faced this issue?
2) Any solution except disabling these UTs (which I did already and this
is how this is solved in gentoo ebuilds)?
3) Any advice? (e.g. secret flag like task['work_fine'] = "1":))

Just to be clear, I'm using BB outside OE, just created simple OE-like
setup.
And my BB is 1.12.0 from http://developer.berlios.de/projects/bitbake/

Problems building classpath in stable/2009

Hi,
(Oeps forgotten the attachements)

I have a build a system on the CM-270 with the OE stable/2009  branch. Now a am trying to build the cacao jvm
(0.99.3) along with the classpath (ver. 0.98). During the compilation task of the "classpath" i have got
the message:

	 java -classpath  antlr.Tool -o ../tools/generated/gnu/classpath/tools/gjdoc/expr/ \
		./gnu/classpath/tools/gjdoc/expr/java-expression.g
	 Unrecognised command line option: -o

	 where options include:
	   -client	   compatibility (ignored)
	   -server	   compatibility (ignored)
	 
	   -cp		   <jar/zip files and directories separated by :>
	   -classpath	   <jar/zip files and directories separated by :>
	 		   locations where to find application classes
	   -D<name>=<value> set a system property
	   -verbose[:class|gc|jni]
	 		   :class print out information about class loading, etc.
	 		   :gc print out results of garbage collection
	 		   :jni print out native method dynamic resolution
	   -version	   print out version number and copyright information
	   -showversion     show version number and copyright and continue
	   -fullversion     show jpackage-compatible version number and exit
	   -? -help	   print out this message
	   -X		   show help on non-standard options

The complete error listing is attached to this email, it seems to me it uses cacao or jamvm to compile?

Problems building classpath in stable/2009

Hi,

I have a build a system on the CM-270 with the OE stable/2009  branch. Now a am trying to build the cacao jvm
(0.99.3) along with the classpath (ver. 0.98). During the compilation task of the "classpath" i have got
the message:

	 java -classpath  antlr.Tool -o ../tools/generated/gnu/classpath/tools/gjdoc/expr/ \
		./gnu/classpath/tools/gjdoc/expr/java-expression.g
	 Unrecognised command line option: -o

	 where options include:
	   -client	   compatibility (ignored)
	   -server	   compatibility (ignored)
	 
	   -cp		   <jar/zip files and directories separated by :>
	   -classpath	   <jar/zip files and directories separated by :>
	 		   locations where to find application classes
	   -D<name>=<value> set a system property
	   -verbose[:class|gc|jni]
	 		   :class print out information about class loading, etc.
	 		   :gc print out results of garbage collection
	 		   :jni print out native method dynamic resolution
	   -version	   print out version number and copyright information
	   -showversion     show version number and copyright and continue
	   -fullversion     show jpackage-compatible version number and exit
	   -? -help	   print out this message
	   -X		   show help on non-standard options

The complete error listing is attached to this email, it seems to me it uses cacao or jamvm to compile?
I tried also with setting the PREFERRED_PROVIDER_virtual/java-native to jamvm-native

Stable branch in git now?

Hello,

Just to make it unambiguous - does stable branch live in git now,
monotone is no longer used for anything?

--

-- 
Best regards,
 Paul                          mailto:pmiscml <at> gmail.com

[PATCH] linux-rp-2.6.24: fix tosa kernel

I'd like to ask for the merge of the following patch into the stable
branch. It's just a rebase of the latest fixes that went into .dev on
top of the stable branch. And Anthony Callegaro verified that this patch
does really fix all remaining tosa problems.

With this patch applied tosa should boot w/o problems with 2.6.24 OE
kernels.

From 996f640aa84139765af6b791c9796e547c043a67 Mon Sep 17 00:00:00 2001
From: Dmitry Baryshkov <dbaryshkov <at> gmail.com>
Date: Tue, 18 Nov 2008 04:18:54 +0300
Subject: [PATCH] linux-rp-2.6.24: fix tosa kernel

add two patches: one to correctly initialise the AC97 interface on PXA,
other to fixup the tosa-bl driver (most important thing is pxa_gpio_mode() setting).

Signed-off-by: Dmitry Baryshkov <dbaryshkov <at> gmail.com>
---
 ...97-switch-AC-unit-to-correct-state-before.patch |   56 +++++++++++
 .../linux/linux-rp-2.6.24/tosa/tosa-bl-fixup.diff  |  103 ++++++++++++++++++++
 packages/linux/linux-rp_2.6.24.bb                  |    4 +-
 3 files changed, 162 insertions(+), 1 deletions(-)
 create mode 100644 packages/linux/linux-rp-2.6.24/tosa/0001-pxa2xx-ac97-switch-AC-unit-to-correct-state-before.patch
 create mode 100644 packages/linux/linux-rp-2.6.24/tosa/tosa-bl-fixup.diff

diff --git
a/packages/linux/linux-rp-2.6.24/tosa/0001-pxa2xx-ac97-switch-AC-unit-to-correct-state-before.patch b/packages/linux/linux-rp-2.6.24/tosa/0001-pxa2xx-ac97-switch-AC-unit-to-correct-state-before.patch
new file mode 100644
index 0000000..09f0cb9
--- /dev/null

[Review] Opie-mediaplayer1 backport & portabase related fixes

Hi there,

I'd like to apply the following to stable:

1) opie-mediaplayer1-oe.patch
opie-mediaplayer1: Add 1.2.3 versions of mediaplayer1 and plugins
Applied to .dev in: 03740c225501e66bf2a7218536868643b53812ce

2) beecrypt-merge.patch
beecrypt: merge duplicate recipes
 * Merge do_stage from portabase/beecrypt into beecrypt
 * Delete portabase/beecrypt (patches no longer seem to be necessary)
Applied to .dev in: 88676b03209e4a8147b6f9ab4d6f5c66d0132128

3) metakit_fix_backport.patch
metakit: fix fetch path & library install/packaging
Applied to .dev in: f2ebdb1826372581987a96a662d30d97b544c438 & 
c9cfaab10b1cc5547009c788d31c5e670d194bc9

Cheers,
Paul

_______________________________________________
Openembedded-stablebranch mailing list
Openembedded-stablebranch <at> openembedded.org
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-stablebranch

[Review] Add qpe-nmap to stable

Hi all,

qpe-nmap is a more usable and featureful nmap frontend for QPE/Opie. I'd like 
to add it to the stable branch (it was previously added to the .dev branch in 
revision 47c8cc979dde8a4e69611fa492570b57a97bcd1e). Please refer to the 
attached patch.

Thanks,
Paul

_______________________________________________
Openembedded-stablebranch mailing list
Openembedded-stablebranch <at> openembedded.org
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-stablebranch

[FYI] More Opie fixes

Hi all,

I would like to apply the following fixes from .dev in .stable:

1) opie_c7x0_disable_oe.patch
libopie2: disable W100 driver for c7x0 (addresses oebug #4191)
 * Add a patch to disable the w100 driver for c7x0 which is currently 
   disabled in qte. Fixes Opie not starting on c7x0 until the w100 
   driver is fixed and reenabled.
Applied to .dev in: b4fa1d211a61db498fb41875679b1cc49e399545

2) qcop_path_fix_oe.patch
Fix qcop paths in postinst/postrm for multiple packages
Applied to .dev in: 231120beb086196600461c605e8af972d11ff350,  
f55b0eb745c295a90ea8e6e188f87c0493386d01

Cheers,
Paul

_______________________________________________
Openembedded-stablebranch mailing list
Openembedded-stablebranch <at> openembedded.org
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-stablebranch

[BUG]: building uclibc fails when USE_NLS is set to "yes"

Hi,

uClibc does not build if UCLIBC_HAS_LOCALE is enabled.

I am building for arm/uclibc/Angstrom 2007.12, but the same happens with
the stable branch and Angstrom 2008.1

Set  USE_NLS="yes", I had to work around a problem so that UCLIBC_HAS_LOCALE
got set in .config (see bug #4592 which is also present in .stable), however 
when the above option is set the build fails.

I attached the build log.

Any ideas what is happening? It works fine if UCLIBC_HAS_LOCALE is not set...

Kind regards,
Jin

| NOTE: make OPTIMIZATION=
CPU_CFLAGS=-isystem/opt/OE/build/angstrom/staging/arm-angstrom-linux-uclibcgnueabi/include
-fexpensive-optimizations -fomit-frame-pointer -frename-registers -Os STRIPTOOL=true
LD=arm-angstrom-linux-uclibcgnueabi-ld LOCALE_DATA_FILENAME=uClibc-locale-030818.arm.tgz
CC=ccache arm-angstrom-linux-uclibcgnueabi-gcc
| make[1]: `conf' is up to date.
| make locale_headers
| gcc extra/locale/gen_wc8bit.c  -o extra/locale/gen_wc8bit
-L/opt/OE/build/angstrom/staging/i686-linux/lib
-Wl,-rpath-link,/opt/OE/build/angstrom/staging/i686-linux/lib

[FYI] conf/distro now reflects maintenance status in the stable branch

Hi,

I keep getting mails from people using $unmaintaineddistro in .stable  
complaining about how everything is broken, so I cleaned up conf/distro.

If your distro config got moved to umaintained and it in fact is  
maintained for org.oe.stable please request it to get moved back. But  
judging from the mailarchives, irc logs and commit log this list is  
accurate. But mistakes happen, so please double check :)

regards,

Koen

_______________________________________________
Openembedded-stablebranch mailing list
Openembedded-stablebranch <at> openembedded.org
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-stablebranch

feed update in stable branch

The added packages (mostly user requested) are documented here [1]. More 
opie applications will be added during the week.

next planned feed update is 10/6/2008.

Regards
Gordon

[1] http://linuxtogo.org/gowiki/AngstromStableChangelog