Updates and Fixes
Richard,
I have reviewed and build/booted these changes.
There are 2 changes from Tom that I need your advice on, please review the gmp and python changes, I do not want to hold these up if they are correct, but I wanted to think if there was a better way to handle the python one.
Thanks
Sau!
The following changes since commit 157db263b50fabf4afcbea3c2a5364c0752ae0e2:
clutter-gst: fix patch filename (2011-11-08 17:33:19 +0000)
are available in the git repository at:
git://git.openembedded.org/openembedded-core-contrib sgw/stage
http://cgit.openembedded.org/cgit.cgi/openembedded-core-contrib/log/?h=sgw/stage
Mark Hatle (1):
pseudo: Uprev pseudo to version 1.2
Matthew McClintock (1):
Add new IMAGE_CLASSES variable for classes for image generation
Otavio Salvador (5):
libcap: fix sstate for native package files/device_table-minimal.txt: add /dev/kmsg
dhcp: drop unused dhcp3.inc file
dhcp: rename dhcp4.inc to dhcp.inc
dhcp: move server configuration to dhcp-server-config
Tom Zanussi (6):
tune-corei7: Add tune file for Intel Core i7 machines
tune-ivb: Add tune file for Intel Ivy Bridge machines machine/include/tune-atom.inc: add atom-specific settings ** gmp_5.0.2: Set CC_FOR_BUILD to BUILD_CC
libzypp: fix mishandling of hyphenated arches
** python: skip setup.py 'import check' when cross-compiling
Xiaofeng Yan (4):
libarchive: update to 2.8.5 distro_tracking_fields.inc: update to libarchive to 2.8.5
qmmp: update to 0.5.2 distro_tracking_fields.inc: Modify the status of qmmp
meta/classes/image.bbclass | 3 +-
.../conf/distro/include/distro_tracking_fields.inc | 19 ++-
meta/conf/local.conf.sample.extended | 6 +
meta/conf/machine/include/tune-atom.inc | 26 +++-
meta/conf/machine/include/tune-corei7.inc | 17 ++
meta/conf/machine/include/tune-ivb.inc | 11 ++
meta/files/device_table-minimal.txt | 1 +
.../dhcp/{dhcp4.inc => dhcp.inc} | 10 +-
meta/recipes-connectivity/dhcp/dhcp3.inc | 49 ------
meta/recipes-connectivity/dhcp/dhcp_4.2.0.bb | 4 +-
.../pseudo/pseudo/realpath_fix.patch | 165 --------------------
.../pseudo/{pseudo_1.1.1.bb => pseudo_1.2.bb} | 7 +-
meta/recipes-devtools/pseudo/pseudo_git.bb | 6 +-
.../python/setup_py_skip_cross_import_check.patch | 27 ++++
meta/recipes-devtools/python/python_2.7.2.bb | 5 +
...rchive_2.8.4.yocto.1.bb => libarchive_2.8.5.bb} | 8 +-
meta/recipes-extended/libzypp/libzypp_git.bb | 8 +-
.../qt-apps/{qmmp_0.5.1.bb => qmmp_0.5.2.bb} | 51 ++++---
meta/recipes-support/gmp/gmp_5.0.2.bb | 3 +-
meta/recipes-support/libcap/libcap.inc | 15 ++-
meta/recipes-support/libcap/libcap_2.22.bb | 2 +-
21 files changed, 174 insertions(+), 269 deletions(-)
create mode 100644 meta/conf/machine/include/tune-corei7.inc
create mode 100644 meta/conf/machine/include/tune-ivb.inc
rename meta/recipes-connectivity/dhcp/{dhcp4.inc => dhcp.inc} (91%)
delete mode 100644 meta/recipes-connectivity/dhcp/dhcp3.inc
delete mode 100644 meta/recipes-devtools/pseudo/pseudo/realpath_fix.patch
rename meta/recipes-devtools/pseudo/{pseudo_1.1.1.bb => pseudo_1.2.bb} (48%)
create mode 100644 meta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch
rename meta/recipes-extended/libarchive/{libarchive_2.8.4.yocto.1.bb => libarchive_2.8.5.bb} (76%)
rename meta/recipes-qt/qt-apps/{qmmp_0.5.1.bb => qmmp_0.5.2.bb} (48%)Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 01/17] pseudo: Uprev pseudo to version 1.2
From: Mark Hatle <[email protected]>
This adds a new feature, PSEUDO_UNLOAD, which can be used to eliminate overhead of LD_PRELOAD when no longer necessary.
Also the, clone(2), support on Linux has been updated to resolve some potential defects in the previous implementation.
---
.../conf/distro/include/distro_tracking_fields.inc | 6 +-
.../pseudo/pseudo/realpath_fix.patch | 165 --------------------
.../pseudo/{pseudo_1.1.1.bb => pseudo_1.2.bb} | 7 +-
meta/recipes-devtools/pseudo/pseudo_git.bb | 6 +-
4 files changed, 9 insertions(+), 175 deletions(-)
delete mode 100644 meta/recipes-devtools/pseudo/pseudo/realpath_fix.patch
rename meta/recipes-devtools/pseudo/{pseudo_1.1.1.bb => pseudo_1.2.bb} (48%)meta/conf/distro/include/distro_tracking_fields.inc +3 -3
@@ -5944,11 +5944,11 @@ RECIPE_LAST_UPDATE_pn-gettext = "May 24, 2011" RECIPE_MANUAL_CHECK_DATE_pn-gettext = "Sep 29, 2011" RECIPE_STATUS_pn-pseudo = "green" -RECIPE_LATEST_VERSION_pn-pseudo = "1.1.1" -RECIPE_LAST_UPDATE_pn-pseudo = "Jun 06, 2011" +RECIPE_LATEST_VERSION_pn-pseudo = "1.2" +RECIPE_LAST_UPDATE_pn-pseudo = "Nov 02, 2011" RECIPE_MAINTAINER_pn-pseudo = "Mark Hatle <[email protected]>" RECIPE_COMMENTS_pn-pseudo = "Yocto Project maintained" -RECIPE_MANUAL_CHECK_DATE_pn-pseudo = "Jun 06, 2011" +RECIPE_MANUAL_CHECK_DATE_pn-pseudo = "Nov 02, 2011" DISTRO_PN_ALIAS_pn-pseudo = "Windriver" DISTRO_PN_ALIAS_pn-rt-tests = "Debian=rt-tests Ubuntu=rt-tests"
meta/recipes-devtools/pseudo/pseudo/realpath_fix.patch deleted +0 -165
165 lines not on this page · raw message
meta/recipes-devtools/pseudo/pseudo_1.2.bb renamed from meta/recipes-devtools/pseudo/pseudo_1.1.1.bb +3 -4
@@ -1,11 +1,10 @@ require pseudo.inc -PR = "r2" +PR = "r3" SRC_URI = "http://www.yoctoproject.org/downloads/${BPN}/${BPN}-${PV}.tar.bz2 \ file://oe-config.patch \ - file://realpath_fix.patch \ file://static_sqlite.patch" -SRC_URI[md5sum] = "dd59766c17e199fe6144fce8a2c67802" -SRC_URI[sha256sum] = "c697f643577d661c3ce826504b9dcd11fa98e78a5d10e3c83931da8942f6bfad" +SRC_URI[md5sum] = "a2819084bab7e991f06626d02cf55048" +SRC_URI[sha256sum] = "4749a22df687f44d24c26e97170d4781a1bd52d5ee092364a40877e4d96ff058"
meta/recipes-devtools/pseudo/pseudo_git.bb +3 -3
@@ -1,8 +1,8 @@ require pseudo.inc -SRCREV = "c2f7c5ad8ef0f9c94a2a8382c109c8c6e16c8b18" -PV = "1.1.1+git${SRCPV}" -PR = "r19" +SRCREV = "17c2233f93692f79684792750001ee6d13e03925" +PV = "1.2+git${SRCPV}" +PR = "r20" DEFAULT_PREFERENCE = "-1"
Signature · 1 line
1.7.6.4
Paul Menzel
Reviewed-byRe: [CONSOLIDATED PULL 01/17] pseudo: Uprev pseudo to version 1.2
Dear Mark,
Am Dienstag, den 08.11.2011, 11:22 -0800 schrieb Saul Wold:
Quoted 1 line
From: Mark Hatle <[email protected]>
I had to read several times to find out the meaning of »Uprev«. Maybe using Update or Upgrade is easier for not native speakers.
Also the second »pseudo« can be dropped in the commit summary.
Quoted 7 lines
This adds a new feature,
PSEUDO_UNLOAD, which can be used to eliminate overhead ofLD_PRELOADwhen no longer necessary.Also the, clone(2), support on Linux has been updated to resolve some potential defects in the previous implementation.
Signed-off-by Mark Hatle <[email protected]>
Quoted 8 lines
---
.../conf/distro/include/distro_tracking_fields.inc | 6 +-
.../pseudo/pseudo/realpath_fix.patch | 165 --------------------
.../pseudo/{pseudo_1.1.1.bb => pseudo_1.2.bb} | 7 +-
meta/recipes-devtools/pseudo/pseudo_git.bb | 6 +-
4 files changed, 9 insertions(+), 175 deletions(-)
delete mode 100644 meta/recipes-devtools/pseudo/pseudo/realpath_fix.patch
rename meta/recipes-devtools/pseudo/{pseudo_1.1.1.bb => pseudo_1.2.bb} (48%)[…]
Thanks,
Paul
[CONSOLIDATED PULL 02/17] libcap: fix sstate for native package
In reply to Saul WoldFrom: Otavio Salvador <[email protected]>
The 'lib' option needs to be given on target and native builds otherwise it installs the binaries at ${libdir}64 when host is 64bit.
--- meta/recipes-support/libcap/libcap.inc | 15 ++++++++++++--- meta/recipes-support/libcap/libcap_2.22.bb | 2 +- 2 files changed, 13 insertions(+), 4 deletions(-)
meta/recipes-support/libcap/libcap.inc +12 -3
@@ -21,9 +21,18 @@ do_configure() { sed -e 's,BUILD_CFLAGS ?=,BUILD_CFLAGS := $(BUILD_CFLAGS),' -i Make.Rules } -EXTRA_OEMAKE = "LIBATTR=yes PAM_CAP=${@base_contains('DISTRO_FEATURES', 'pam', 'yes', 'no', d)} INDENT= SYSTEM_HEADERS=${STAGING_INCDIR} RAISE_SETFCAP=no" -EXTRA_OEMAKE_virtclass-native = "LIBATTR=no PAM_CAP=no INDENT= " -EXTRA_OEMAKE += " lib=${@os.path.basename('${libdir}')}" +EXTRA_OEMAKE = " \ + LIBATTR=yes \ + PAM_CAP=${@base_contains('DISTRO_FEATURES', 'pam', 'yes', 'no', d)} \ + INDENT= SYSTEM_HEADERS=${STAGING_INCDIR} RAISE_SETFCAP=no \ + lib=${@os.path.basename('${libdir}')} \ +" +EXTRA_OEMAKE_virtclass-native = " \ + LIBATTR=no \ + PAM_CAP=no \ + INDENT= \ + lib=${@os.path.basename('${libdir}')} \ +" do_compile() { oe_runmake
meta/recipes-support/libcap/libcap_2.22.bb +1 -1
@@ -1,6 +1,6 @@ require libcap.inc -PR = "r1" +PR = "r2" SRC_URI[md5sum] = "ce64058bdb3f086ddbfca8ce6c919845" SRC_URI[sha256sum] = "73ebbd4877b5f69dd28b72098e510c5b318bc480f8201c4061ac98b78c04050f"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 03/17] files/device_table-minimal.txt: add /dev/kmsg
In reply to Saul WoldFrom: Otavio Salvador <[email protected]>
--- meta/files/device_table-minimal.txt | 1 + 1 files changed, 1 insertions(+), 0 deletions(-)
meta/files/device_table-minimal.txt +1 -0
@@ -15,6 +15,7 @@ /dev/hda b 660 0 6 3 0 - - - /dev/hda b 660 0 6 3 1 1 1 20 /dev/kmem c 640 0 15 1 2 - - - +/dev/kmsg c 600 0 0 1 11 - - - /dev/mem c 640 0 15 1 1 - - - /dev/null c 666 0 0 1 3 - - - /dev/ram b 640 0 0 1 0 0 1 4
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 04/17] dhcp: drop unused dhcp3.inc file
In reply to Saul WoldFrom: Otavio Salvador <[email protected]>
--- meta/recipes-connectivity/dhcp/dhcp3.inc | 49 ------------------------------ 1 files changed, 0 insertions(+), 49 deletions(-) delete mode 100644 meta/recipes-connectivity/dhcp/dhcp3.inc
meta/recipes-connectivity/dhcp/dhcp3.inc deleted +0 -49
@@ -1,49 +0,0 @@ -SECTION = "console/network" -DESCRIPTION = "Internet Software Consortium DHCP package" -HOMEPAGE = "http://www.isc.org/" - -LICENSE = "ISC" -LIC_FILES_CHKSUM = "file://LICENSE;beginline=4;md5=bb6fd41f5895b67088ebea61ad365e74" - -SRC_URI = "ftp://ftp.isc.org/isc/dhcp/dhcp-${PV}.tar.gz \ - file://init-relay file://default-relay \ - file://init-server file://default-server \ - file://dhclient.conf file://dhcpd.conf" - -inherit autotools - -TARGET_CFLAGS += "-D_GNU_SOURCE" - -do_compile() { - make RANLIB=${RANLIB} PREDEFINES='-D_PATH_DHCPD_DB=\"/var/lib/dhcp/dhcpd.leases\" \ - -D_PATH_DHCLIENT_DB=\"/var/lib/dhcp/dhclient.leases\" \ - -D_PATH_DHCLIENT_SCRIPT=\"/sbin/dhclient-script\" \ - -D_PATH_DHCPD_CONF=\"/etc/dhcp/dhcpd.conf\" \ - -D_PATH_DHCLIENT_CONF=\"/etc/dhcp/dhclient.conf\"' -} - -do_install() { - make -e DESTDIR=${D} USRMANDIR=${mandir}/man1 ADMMANDIR=${mandir}/man8 FFMANDIR=${mandir}/man5 LIBMANDIR=${mandir}/man3 LIBDIR=${libdir} INCDIR=${includedir} install - install -d ${D}${sysconfdir}/init.d - install -d ${D}${sysconfdir}/default - install -d ${D}${sysconfdir}/dhcp - install -m 0755 ${WORKDIR}/init-relay ${D}${sysconfdir}/init.d/dhcp-relay - install -m 0644 ${WORKDIR}/default-relay ${D}${sysconfdir}/default/dhcp-relay - install -m 0755 ${WORKDIR}/init-server ${D}${sysconfdir}/init.d/dhcp-server - install -m 0644 ${WORKDIR}/default-server ${D}${sysconfdir}/default/dhcp-server - install -m 0644 ${WORKDIR}/dhclient.conf ${D}${sysconfdir}/dhcp/dhclient.conf - install -m 0644 ${WORKDIR}/dhcpd.conf ${D}${sysconfdir}/dhcp/dhcpd.conf - install -d ${D}${base_sbindir}/ - mv ${D}${sbindir}/dhclient ${D}${base_sbindir}/ - install -m 0755 ${S}/client/scripts/linux ${D}${base_sbindir}/dhclient-script -} - -PACKAGES += "dhcp-server dhcp-client dhcp-relay dhcp-omshell" -FILES_${PN} = "" -FILES_dhcp-server = "${sbindir}/dhcpd ${sysconfdir}/init.d/dhcp-server ${sysconfdir}/default/dhcp-server ${sysconfdir}/dhcp/dhcpd.conf" -FILES_dhcp-relay = "${sbindir}/dhcrelay ${sysconfdir}/init.d/dhcp-relay ${sysconfdir}/default/dhcp-relay" - -FILES_dhcp-client = "${base_sbindir}/dhclient ${base_sbindir}/dhclient-script ${sysconfdir}/dhcp/dhclient.conf" -RDEPENDS_dhcp-client = "bash" - -FILES_dhcp-omshell = "${bindir}/omshell"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 05/17] dhcp: rename dhcp4.inc to dhcp.inc
In reply to Saul WoldFrom: Otavio Salvador <[email protected]>
---
.../dhcp/{dhcp4.inc => dhcp.inc} | 0
meta/recipes-connectivity/dhcp/dhcp_4.2.0.bb | 2 +-
2 files changed, 1 insertions(+), 1 deletions(-)
rename meta/recipes-connectivity/dhcp/{dhcp4.inc => dhcp.inc} (100%)meta/recipes-connectivity/dhcp/dhcp.inc renamed from meta/recipes-connectivity/dhcp/dhcp4.inc +0 -0
meta/recipes-connectivity/dhcp/dhcp_4.2.0.bb +1 -1
@@ -1,4 +1,4 @@ -require dhcp4.inc +require dhcp.inc PR = "r2"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 06/17] dhcp: move server configuration to dhcp-server-config
In reply to Saul WoldFrom: Otavio Salvador <[email protected]>
--- meta/recipes-connectivity/dhcp/dhcp.inc | 10 ++++++++-- meta/recipes-connectivity/dhcp/dhcp_4.2.0.bb | 2 +- 2 files changed, 9 insertions(+), 3 deletions(-)
meta/recipes-connectivity/dhcp/dhcp.inc +8 -2
@@ -53,9 +53,15 @@ do_install_append () { install -m 0755 ${S}/client/scripts/linux ${D}${base_sbindir}/dhclient-script } -PACKAGES += "dhcp-server dhcp-client dhcp-relay dhcp-omshell" +PACKAGES += "dhcp-server dhcp-server-config dhcp-client dhcp-relay dhcp-omshell" + FILES_${PN} = "" -FILES_dhcp-server = "${sbindir}/dhcpd ${sysconfdir}/init.d/dhcp-server ${sysconfdir}/default/dhcp-server ${sysconfdir}/dhcp/dhcpd.conf" + +FILES_dhcp-server = "${sbindir}/dhcpd ${sysconfdir}/init.d/dhcp-server" +RRECOMMENDS_dhcp-server = "dhcp-server-config" + +FILES_dhcp-server-config = "${sysconfdir}/default/dhcp-server ${sysconfdir}/dhcp/dhcpd.conf" + FILES_dhcp-relay = "${sbindir}/dhcrelay ${sysconfdir}/init.d/dhcp-relay ${sysconfdir}/default/dhcp-relay" FILES_dhcp-client = "${base_sbindir}/dhclient ${base_sbindir}/dhclient-script ${sysconfdir}/dhcp/dhclient.conf"
meta/recipes-connectivity/dhcp/dhcp_4.2.0.bb +1 -1
@@ -1,6 +1,6 @@ require dhcp.inc -PR = "r2" +PR = "r3" SRC_URI += "file://fixincludes.patch \ file://dhcp-3.0.3-dhclient-dbus.patch;striplevel=0 \
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 07/17] libarchive: update to 2.8.5
In reply to Saul WoldFrom: Xiaofeng Yan <[email protected]>
Remove patch "0003-Patch-from-upstream-rev-2516.patch" because it has been merged to source codes.
---
...rchive_2.8.4.yocto.1.bb => libarchive_2.8.5.bb} | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)
rename meta/recipes-extended/libarchive/{libarchive_2.8.4.yocto.1.bb => libarchive_2.8.5.bb} (76%)meta/recipes-extended/libarchive/libarchive_2.8.5.bb renamed from meta/recipes-extended/libarchive/libarchive_2.8.4.yocto.1.bb +4 -4
@@ -9,19 +9,19 @@ DEPENDS = "libxml2" # We need to repack the tarball due undistributable content on the upstream one. # More details at http://code.google.com/p/libarchive/issues/detail?id=162 -SRC_URI = "http://autobuilder.yoctoproject.org/sources/libarchive-${PV}.tar.gz \ +SRC_URI = "http://libarchive.googlecode.com/files/libarchive-${PV}.tar.gz \ file://0001-Patch-from-upstream-revision-1990.patch \ file://0002-Patch-from-upstream-revision-1991.patch \ - file://0003-Patch-from-upstream-rev-2516.patch \ file://0004-Patch-from-upstream-rev-2514.patch \ file://0005-Patch-from-upstream-rev-2520.patch \ file://0006-Patch-from-upstream-rev-2521.patch \ file://0007-Ignore-ENOSYS-error-when-setting-up-xattrs.-Closes-5.patch \ " -SRC_URI[md5sum] = "71242da5191f1218f13dd520d95a870e" -SRC_URI[sha256sum] = "8cd55db11b1d6001ff8007e4d22b6f4a4bb215e70750e19ab44b84b99ab76053" +SRC_URI[md5sum] = "9caf51dcf6213e9c9f5a1c27448b9c90" +SRC_URI[sha256sum] = "13993e0ffbd121ccda46ea226b1f8eac218de0fa8da7d8b1f998093d5c32a72d" inherit autotools lib_package BBCLASSEXTEND = "nativesdk" +
Signature · 1 line
1.7.6.4
Re: [CONSOLIDATED PULL 07/17] libarchive: update to 2.8.5
Dear Xiaofeng,
do I need to put people in CC?
Am Dienstag, den 08.11.2011, 11:22 -0800 schrieb Saul Wold:
Quoted 4 lines
From: Xiaofeng Yan <[email protected]>
Remove patch "0003-Patch-from-upstream-rev-2516.patch" because it has been merged to source codes.
Mentioning of the SRC_URI change would be also nice.
Quoted 14 lines
Signed-off-by Xiaofeng Yan <[email protected]>--- ...rchive_2.8.4.yocto.1.bb => libarchive_2.8.5.bb} | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) rename meta/recipes-extended/libarchive/{libarchive_2.8.4.yocto.1.bb => libarchive_2.8.5.bb} (76%)meta/recipes-extended/libarchive/libarchive_2.8.5.bb renamed from meta/recipes-extended/libarchive/libarchive_2.8.4.yocto.1.bb +0
-0
meta/recipes-extended/libarchive/libarchive_2.8.5.bb DEPENDS = "libxml2" · line 12
Quoted 5 lines
meta/recipes-extended/libarchive/libarchive_2.8.5.bb renamed from meta/recipes-extended/libarchive/libarchive_2.8.4.yocto.1.bb +1
-1@@ -9,19 +9,19 @@ DEPENDS = "libxml2" # We need to repack the tarball due undistributable content on the upstream one. # More details at http://code.google.com/p/libarchive/issues/detail?id=162 -SRC_URI = "http://autobuilder.yoctoproject.org/sources/libarchive-${PV}.tar.gz \ +SRC_URI = "http://libarchive.googlecode.com/files/libarchive-${PV}.tar.gz \
Reading the issues mentioned in the comment, I guess
I've just deleted them from libarchive/2.8 branch as well.
means this landed in version 2.8.5? Please remove the comment then.
Quoted 8 lines
meta/recipes-extended/libarchive/libarchive_2.8.5.bb renamed from meta/recipes-extended/libarchive/libarchive_2.8.4.yocto.1.bb +0
-1file://0001-Patch-from-upstream-revision-1990.patch \ file://0002-Patch-from-upstream-revision-1991.patch \ - file://0003-Patch-from-upstream-rev-2516.patch \ file://0004-Patch-from-upstream-rev-2514.patch \ file://0005-Patch-from-upstream-rev-2520.patch \ file://0006-Patch-from-upstream-rev-2521.patch \ file://0007-Ignore-ENOSYS-error-when-setting-up-xattrs.-Closes-5.patch \ "
meta/recipes-extended/libarchive/libarchive_2.8.5.bb DEPENDS = "libxml2" · line 27
Quoted 10 lines
meta/recipes-extended/libarchive/libarchive_2.8.5.bb renamed from meta/recipes-extended/libarchive/libarchive_2.8.4.yocto.1.bb +3
-2-SRC_URI[md5sum] = "71242da5191f1218f13dd520d95a870e" -SRC_URI[sha256sum] = "8cd55db11b1d6001ff8007e4d22b6f4a4bb215e70750e19ab44b84b99ab76053" +SRC_URI[md5sum] = "9caf51dcf6213e9c9f5a1c27448b9c90" +SRC_URI[sha256sum] = "13993e0ffbd121ccda46ea226b1f8eac218de0fa8da7d8b1f998093d5c32a72d" inherit autotools lib_package BBCLASSEXTEND = "nativesdk" +
No new line please.
Thanks,
Paul
PATCH 1/1 libarchive: Remove obsolete comment and empty line at the end
Date: Wed, 9 Nov 2011 00:17:22 +0100
This is a fix up for
commit fb19df5b21e551c5dfdfa340438952560c5fa528
Author: Xiaofeng Yan <[email protected]>
Date: Mon Nov 7 20:03:53 2011 +0800
libarchive: update to 2.8.5
Remove patch "0003-Patch-from-upstream-rev-2516.patch" because it has been merged to source codes.
removing a now obsolete comment because the undistributable content was removed [1] from upstream’s tarball.
Also remove an empty line at the end introduced in the above commit.
[1] http://code.google.com/p/libarchive/issues/detail?id=162
--- .../libarchive/libarchive_2.8.5.bb | 3 --- 1 files changed, 0 insertions(+), 3 deletions(-)
meta/recipes-extended/libarchive/libarchive_2.8.5.bb +0 -3
@@ -7,8 +7,6 @@ PR = "r0" DEPENDS = "libxml2" -# We need to repack the tarball due undistributable content on the upstream one. -# More details at http://code.google.com/p/libarchive/issues/detail?id=162 SRC_URI = "http://libarchive.googlecode.com/files/libarchive-${PV}.tar.gz \ file://0001-Patch-from-upstream-revision-1990.patch \ file://0002-Patch-from-upstream-revision-1991.patch \
@@ -24,4 +22,3 @@ SRC_URI[sha256sum] = "13993e0ffbd121ccda46ea226b1f8eac218de0fa8da7d8b1f998093d5c inherit autotools lib_package BBCLASSEXTEND = "nativesdk" -
Signature · 1 line
1.7.7.2
[CONSOLIDATED PULL 08/17] distro_tracking_fields.inc: update to libarchive to 2.8.5
In reply to Saul WoldFrom: Xiaofeng Yan <[email protected]>
Change the status of libarchive.
--- .../conf/distro/include/distro_tracking_fields.inc | 5 ++++- 1 files changed, 4 insertions(+), 1 deletions(-)
meta/conf/distro/include/distro_tracking_fields.inc +4 -1
@@ -4615,7 +4615,10 @@ RECIPE_MAINTAINER_pn-libx11-diet = "Xiaofeng Yan <[email protected]>" RECIPE_STATUS_pn-directfb = "red" RECIPE_MAINTAINER_pn-directfb = "Xiaofeng Yan <[email protected]>" -RECIPE_STATUS_pn-libarchive = "red" +RECIPE_STATUS_pn-libarchive = "green" +RECIPE_LATEST_VERSION_pn-libarchive = "2.8.5" +RECIPE_LATEST_RELEASE_DATE_pn-libarchive = "Sept 3, 2011" +RECIPE_LAST_UPDATE_pn-libarchive = "Nov 7, 2011" RECIPE_MAINTAINER_pn-libarchive = "Xiaofeng Yan <[email protected]>" RECIPE_STATUS_pn-xserver-kdrive = "red"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 09/17] qmmp: update to 0.5.2
In reply to Saul WoldFrom: Xiaofeng Yan <[email protected]>
qmmp: remove old recipe and add new recipe
Use do_split_package for each grouping and name them correctly besides updating package. The format to name plugins of qmmp:
qmmp-plugin-<group>-<library> with the Description: Qmmp<Group> plugin for %s"
Fix dESCRIPTION -> DESCRIPTION
---
.../qt-apps/{qmmp_0.5.1.bb => qmmp_0.5.2.bb} | 51 +++++++++++--------
1 files changed, 29 insertions(+), 22 deletions(-)
rename meta/recipes-qt/qt-apps/{qmmp_0.5.1.bb => qmmp_0.5.2.bb} (48%)meta/recipes-qt/qt-apps/qmmp_0.5.2.bb renamed from meta/recipes-qt/qt-apps/qmmp_0.5.1.bb +29 -22
@@ -6,18 +6,15 @@ SECTION = "multimedia" PR = "r0" -DEPENDS = "qt4-x11-free taglib libmad libvorbis libogg alsa-lib" -RDEPENDS_${PN} += "taglib alsa-lib curl" +DEPENDS = "taglib libmad libvorbis libogg alsa-lib" SRC_URI = "http://qmmp.ylsoftware.com/files/${BPN}-${PV}.tar.bz2" -SRC_URI[md5sum] = "7934c069256bf2fc4ab762ad2683db2f" -SRC_URI[sha256sum] = "16ab0334f368a5ef14d631a1d43d7cae876ee06b7185643377b5e9cfd6143099" +SRC_URI[md5sum] = "20852f3cce3471bfc5affa9b2e947dc6" +SRC_URI[sha256sum] = "6391dec020d2a381d7f4b7890fae6c49eadf88b3c9aef571fe3c5e96140822ec" -PARALLEL_MAKE = "" - -inherit qmake2 cmake +inherit qmake2 cmake qt4x11 export EXTRA_OECMAKE = "-DQT_QMAKE_EXECUTABLE=${OE_QMAKE_QMAKE} \ -DQT_LRELEASE_EXECUTABLE=${OE_QMAKE_LRELEASE} \
@@ -28,20 +25,30 @@ export EXTRA_OECMAKE = "-DQT_QMAKE_EXECUTABLE=${OE_QMAKE_QMAKE} \ -DQT_HEADERS_DIR=${OE_QMAKE_INCDIR_QT} \ -DQT_QTCORE_INCLUDE_DIR=${OE_QMAKE_INCDIR_QT}/QtCore \ " - -FILES_${PN} = "${bindir}/qmmp ${libdir}/*.so* \ - ${libdir}/qmmp/PlaylistFormats/*.so \ - ${libdir}/qmmp/Output/libalsa.so \ - ${libdir}/qmmp/Transports/libhttp.so \ - ${libdir}/qmmp/Visual/libanalyzer.so \ - ${datadir}/icons/* \ - ${datadir}/qmmp/images/* \ - ${datadir}/applications/qmmp.desktop \ +PACKAGES_DYNAMIC = "qmmp-plugin-* " + + +python populate_packages_prepend () { + import os + qmmp_libdir = bb.data.expand('${libdir}/qmmp', d) + gd = bb.data.expand('${D}/${libdir}/qmmp', d) + plug_dirs = os.listdir(gd) + + for plug_dir in plug_dirs: + g_plug_dir = os.path.join(qmmp_libdir,plug_dir) + do_split_packages(d, g_plug_dir, '^lib(.*)\.so$', 'qmmp-plugin-' + plug_dir.lower() + '-%s', 'Qmmp' + plug_dir + 'plugin for %s') +} + +FILES_${PN} = "\ + ${bindir}/qmmp \ + ${libdir}/lib*${SOLIBS} \ + ${datadir}/icons/* \ + ${datadir}/qmmp/images/* \ + ${datadir}/applications/* \ + " + +FILES_${PN}-dbg += "\ + ${libdir}/qmmp/*/.debug/* \ " -PACKAGES += "${PN}-plugin-input-mad ${PN}-plugin-input-vorbis" - -FILES_${PN}-plugin-input-mad = "${libdir}/qmmp/Input/libmad.so" -RDEPENDS_${PN}-plugin-input-mad = "libmad" -FILES_${PN}-plugin-input-vorbis = "${libdir}/qmmp/Input/libvorbis.so" -RDEPENDS_${PN}-plugin-input-vorbis = "libvorbis libogg" +RDEPENDS_${PN} += "taglib alsa-lib libmad curl"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 10/17] distro_tracking_fields.inc: Modify the status of qmmp
In reply to Saul WoldFrom: Xiaofeng Yan <[email protected]>
update qmmp to 0.5.2
--- .../conf/distro/include/distro_tracking_fields.inc | 8 ++++++-- 1 files changed, 6 insertions(+), 2 deletions(-)
meta/conf/distro/include/distro_tracking_fields.inc +6 -2
@@ -5729,11 +5729,15 @@ RECIPE_LAST_UPDATE_pn-libgdata = "Dec 29, 2010" RECIPE_MAINTAINER_pn-libgdata = "Zhai Edwin <[email protected]>" RECIPE_LATEST_VERSION_pn-man-pages = "3.32" -RECIPE_LATEST_VERSION_pn-qmmp = "0.4.3" -RECIPE_MAINTAINER_pn-qmmp = "Xiaofeng Yan <[email protected]>" RECIPE_LATEST_VERSION_pn-sysprof = "6b5b8432711ef5c747f8375073cd9af88922d3c6" RECIPE_LATEST_VERSION_pn-table = "d42a44938699ee30a998fc42bc149aebf69389db" +RECIPE_STATUS_pn-qmmp = "green" +RECIPE_LATEST_VERSION_pn-qmmp = "0.5.2" +RECIPE_LATEST_RELEASE_DATE_pn-qmmp = "Sep 02, 2011" +RECIPE_LAST_UPDATE_pn-qmmp = "Nov 7, 2011" +RECIPE_MAINTAINER_pn-qmmp = "Xiaofeng Yan <[email protected]>" + RECIPE_LATEST_VERSION_pn-texinfo = "4.13a" RECIPE_LATEST_RELEASE_DATE_pn-texinfo = "Sep 22, 2008" RECIPE_LAST_UPDATE_pn-texinfo = "Sep 16, 2011"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 12/17] tune-corei7: Add tune file for Intel Core i7 machines
In reply to Saul WoldFrom: Tom Zanussi <[email protected]>
Add a tune file for Intel Core i7 machines both with and without AVX support.
--- meta/conf/machine/include/tune-corei7.inc | 17 +++++++++++++++++ 1 files changed, 17 insertions(+), 0 deletions(-) create mode 100644 meta/conf/machine/include/tune-corei7.inc
meta/conf/machine/include/tune-corei7.inc new file +17 -0
@@ -0,0 +1,17 @@ +DEFAULTTUNE ?= "corei7" +TUNE_PKGARCH ?= "${@bb.utils.contains("TUNE_FEATURES", "avx", "corei7-avx", "corei7", d)}" + +require conf/machine/include/tune-x86_64.inc + +TUNEVALID[corei7] = "Enable corei7-specific processor optimizations" +TUNE_CCARGS += "${@bb.utils.contains("TUNE_FEATURES", "corei7", "-march=corei7${COREI7_AVX}", "", d)}" + +COREI7_AVX = "${@bb.utils.contains("TUNE_FEATURES", [ "corei7", "avx" ], "-avx", "", d)}" + +AVAILTUNES += "corei7" +TUNE_FEATURES_tune-corei7 = "${TUNE_FEATURES_tune-x86-64} corei7" +PACKAGE_EXTRA_ARCHS_tune-corei7 = "${PACKAGE_EXTRA_ARCHS_tune-x86-64} corei7" + +AVAILTUNES += "corei7-avx" +TUNE_FEATURES_tune-corei7-avx = "${TUNE_FEATURES_tune-x86-64} corei7 avx" +PACKAGE_EXTRA_ARCHS_tune-corei7-avx = "${PACKAGE_EXTRA_ARCHS_tune-x86-64} corei7-avx"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 11/17] Add new IMAGE_CLASSES variable for classes for image generation
In reply to Saul WoldFrom: Matthew McClintock <[email protected]>
Allows us to import classes only for images and not to the global namespace
--- meta/classes/image.bbclass | 3 ++- meta/conf/local.conf.sample.extended | 6 ++++++ 2 files changed, 8 insertions(+), 1 deletions(-)
meta/classes/image.bbclass +2 -1
@@ -109,7 +109,8 @@ def get_devtable_list(d): str += " %s" % bb.which(bb.data.getVar('BBPATH', d, 1), devtable) return str -inherit image_types +IMAGE_CLASSES ?= "image_types" +inherit ${IMAGE_CLASSES} IMAGE_POSTPROCESS_COMMAND ?= "" MACHINE_POSTPROCESS_COMMAND ?= ""
meta/conf/local.conf.sample.extended +6 -0
@@ -114,3 +114,9 @@ # The network based PR service host and port #PRSERV_HOST = "localhost" #PRSERV_PORT = "8585" + +# Additional image generation features +# +# The following is a list of classes to import to use in the generation of images +# currently an example class is image_types_uboot +# IMAGE_CLASSES = " image_types_uboot"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 14/17] machine/include/tune-atom.inc: add atom-specific settings
In reply to Saul WoldFrom: Tom Zanussi <[email protected]>
Atom tunings are currently the same as core2; this patch changes them to make use of atom-specific settings such as the atom arch and instruction support in gcc.
--- meta/conf/machine/include/tune-atom.inc | 26 ++++++++++++++++++++++++-- 1 files changed, 24 insertions(+), 2 deletions(-)
meta/conf/machine/include/tune-atom.inc +24 -2
@@ -1,2 +1,24 @@ -# Atom tunings are the same as core2 for now... -require conf/machine/include/tune-core2.inc +DEFAULTTUNE ?= "atom" +TUNE_PKGARCH ?= "${@bb.utils.contains("TUNE_FEATURES", "m32", "atom", "atom-64", d)}" + +require conf/machine/include/tune-i586.inc + +# Extra tune features +TUNEVALID[atom] = "Enable atom-specific processor optimizations" +TUNE_CCARGS += "${@bb.utils.contains("TUNE_FEATURES", "atom", "-march=atom -mssse3 -mfpmath=sse", "", d)}" + +# Extra tune selections +AVAILTUNES += "atom" +TUNE_FEATURES_tune-atom ?= "${TUNE_FEATURES_tune-x86} atom" +BASE_LIB_tune-atom ?= "lib" +PACKAGE_EXTRA_ARCHS_tune-atom = "${PACKAGE_EXTRA_ARCHS_tune-x86} i386 i486 i586 i686 atom" + +AVAILTUNES += "atom-64" +TUNE_FEATURES_tune-atom-64 ?= "${TUNE_FEATURES_tune-x86-64} atom" +BASE_LIB_tune-atom-64 ?= "lib64" +PACKAGE_EXTRA_ARCHS_tune-atom-64 = "${PACKAGE_EXTRA_ARCHS_tune-x86-64} atom-64" + +AVAILTUNES += "atom-64-x32" +TUNE_FEATURES_tune-atom-64-x32 ?= "${TUNE_FEATURES_tune-x86-64-x32} atom" +BASE_LIB_tune-atom-64-x32 ?= "libx32" +PACKAGE_EXTRA_ARCHS_tune-atom-64-x32 = "${PACKAGE_EXTRA_ARCHS_tune-x86-64-x32} atom-64-x32"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 13/17] tune-ivb: Add tune file for Intel Ivy Bridge machines
In reply to Saul WoldFrom: Tom Zanussi <[email protected]>
Add a tune file for Intel Ivy Bridge machines.
--- meta/conf/machine/include/tune-ivb.inc | 11 +++++++++++ 1 files changed, 11 insertions(+), 0 deletions(-) create mode 100644 meta/conf/machine/include/tune-ivb.inc
meta/conf/machine/include/tune-ivb.inc new file +11 -0
@@ -0,0 +1,11 @@ +DEFAULTTUNE ?= "ivb" +TUNE_PKGARCH ?= "${@bb.utils.contains("TUNE_FEATURES", "ivb", "ivb", "", d)}" + +require conf/machine/include/tune-x86_64.inc + +TUNEVALID[ivb] = "Enable ivb-specific processor optimizations" +TUNE_CCARGS += "${@bb.utils.contains("TUNE_FEATURES", "ivb", "-march=core-avx-i", "", d)}" + +AVAILTUNES += "ivb" +TUNE_FEATURES_tune-ivb = "${TUNE_FEATURES_tune-x86-64} ivb" +PACKAGE_EXTRA_ARCHS_tune-ivb = "${PACKAGE_EXTRA_ARCHS_tune-x86-64} ivb"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 15/17] gmp_5.0.2: Set CC_FOR_BUILD to BUILD_CC
In reply to Saul WoldFrom: Tom Zanussi <[email protected]>
CC_FOR_BUILD was compiling the test programs using the target's compile options and executing those on the host, causing errors such as:
/bin/sh: line 1: 15032 Illegal instruction ./gen-bases table 64 0 > mpn/mp_bases.c/bin/sh: line 1: 15033 Illegal instruction ./gen-bases header 64 0 > mp_bases.h
Export CC_FOR_BUILD using BUILD_CC to fix the problem.
--- meta/recipes-support/gmp/gmp_5.0.2.bb | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-)
meta/recipes-support/gmp/gmp_5.0.2.bb +2 -1
@@ -2,11 +2,12 @@ require gmp.inc LICENSE="LGPLv3&GPLv3" LIC_FILES_CHKSUM = "file://COPYING;md5=d32239bcb673463ab874e80d47fae504 \ file://version.c;endline=18;md5=d8c56b52b9092346b9f93b4da65ef790" -PR = "r0" +PR = "r1" SRC_URI_append += "file://sh4-asmfix.patch \ file://use-includedir.patch " +export CC_FOR_BUILD = "${BUILD_CC}" SRC_URI[md5sum] = "0bbaedc82fb30315b06b1588b9077cd3" SRC_URI[sha256sum] = "dbc2db76fdd4e99f85d5e35aa378ed62c283e0d586b91bd8703aff75a7804c28"
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 16/17] libzypp: fix mishandling of hyphenated arches
In reply to Saul WoldFrom: Tom Zanussi <[email protected]>
Several hyphen-to-underscore translations were missing, causing compiler errors trying to build arches with hyphens in their names. This adds the missing translations.
--- meta/recipes-extended/libzypp/libzypp_git.bb | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)
meta/recipes-extended/libzypp/libzypp_git.bb +4 -4
@@ -11,7 +11,7 @@ DEPENDS = "rpm boost curl libxml2 zlib sat-solver expat openssl udev" S = "${WORKDIR}/git" SRCREV = "15b6c52260bbc52b3d8e585e271b67e10cc7c433" PV = "0.0-git${SRCPV}" -PR = "r15" +PR = "r16" SRC_URI = "git://github.com/openSUSE/libzypp.git;protocol=git \ file://no-doc.patch \
@@ -98,10 +98,10 @@ do_archgen () { esac if [ "${AVOID_CONSTRUCTOR}" != "true" ]; then echo -n " const Arch Arch_${each_arch} " | tr - _ >> zypp/oe-arch.h - echo "(_${each_arch});" >> zypp/oe-arch.h + echo "(_${each_arch});" | tr - _ >> zypp/oe-arch.h else echo -n " const Arch Arch_${each_arch} " | tr - _ >> zypp/oe-arch.h - echo "( IdString ( \"${each_arch}\" ) );" >> zypp/oe-arch.h + echo "( IdString ( \"${each_arch}\" ) );" | tr - _ >> zypp/oe-arch.h fi done echo "#endif /* OE_PROTO */" >> zypp/oe-arch.h
@@ -142,7 +142,7 @@ do_archgen () { COMPAT_WITH="${CARCH},${COMPAT} $COMPAT_WITH" done for each_compat in ${COMPAT_WITH} ; do - echo " defCompatibleWith( ${each_compat} );" >> zypp/oe-arch.h + echo " defCompatibleWith( ${each_compat} );" | tr - _ >> zypp/oe-arch.h done echo "#endif /* DEF_COMPAT */" >> zypp/oe-arch.h echo "" >> zypp/oe-arch.h
Signature · 1 line
1.7.6.4
[CONSOLIDATED PULL 17/17] python: skip setup.py 'import check' when cross-compiling
In reply to Saul WoldFrom: Tom Zanussi <[email protected]>
build_extension() in setup.py, as part of the build process, does an 'import check' on the built extension. The import check in turn dlopen()'s the shared library associated with the extension, which isn't something that makes sense if that library was cross-compiled for a different architecture.
This was noticed with an x86_64 target that was compiled with avx support, because it caused 'illegal instruction' exceptions:
| /bin/sh: line 1: 14575 Illegal instruction ... -E ./setup.py -q build
For other target architectures, it doesn't necessarily cause illegal instruction exceptions, but still fails. For example, on arm, the failure pathway causes this warning:
*** WARNING: renaming "cmath" since importing it failed: .../cmath.so:
wrong ELF class: ELFCLASS32
This patch to setup.py and the associated recipe changes allow the whole 'import check' logic to be skipped when cross-compiling.
--- .../python/setup_py_skip_cross_import_check.patch | 27 ++++++++++++++++++++ meta/recipes-devtools/python/python_2.7.2.bb | 5 +++ 2 files changed, 32 insertions(+), 0 deletions(-) create mode 100644 meta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch
meta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch new file +27 -0
@@ -0,0 +1,27 @@ +This patch skips over the 'import check' setup.py does when building +extensions. This generally won't work when cross-compiling. + +Upstream-Status: Inappropriate [embedded-specific] + +Signed-off-by: Tom Zanussi <[email protected]> + +Index: Python-2.7.2/setup.py +=================================================================== +--- Python-2.7.2.orig/setup.py 2011-11-04 16:46:34.553796410 -0500 ++++ Python-2.7.2/setup.py 2011-11-04 16:59:49.692802313 -0500 +@@ -287,6 +287,15 @@ + (ext.name, sys.exc_info()[1])) + self.failed.append(ext.name) + return ++ ++ # If we're cross-compiling, we want to skip the import check ++ # i.e. we shouldn't be dynamically loading target shared libs ++ if os.environ.get('CROSS_COMPILE') is not None: ++ self.announce( ++ 'WARNING: skipping import check for cross-compiled "%s"' % ++ ext.name) ++ return ++ + # Workaround for Mac OS X: The Carbon-based modules cannot be + # reliably imported into a command-line Python + if 'Carbon' in ext.extra_link_args:
meta/recipes-devtools/python/python_2.7.2.bb +5 -0
@@ -18,6 +18,7 @@ SRC_URI += "\ file://multilib.patch \ file://cgi_py.patch \ file://remove_sqlite_rpath.patch \ + file://setup_py_skip_cross_import_check.patch \ " S = "${WORKDIR}/Python-${PV}"
@@ -57,6 +58,8 @@ do_compile() { # then call do_install twice we get Makefile.orig == Makefile.sysroot install -m 0644 Makefile Makefile.sysroot + export CROSS_COMPILE="${TARGET_PREFIX}" + oe_runmake HOSTPGEN=${STAGING_BINDIR_NATIVE}/pgen \ HOSTPYTHON=${STAGING_BINDIR_NATIVE}/python \ STAGING_LIBDIR=${STAGING_LIBDIR} \
@@ -78,6 +81,8 @@ do_install() { # make install needs the original Makefile, or otherwise the inclues would # go to ${D}${STAGING...}/... install -m 0644 Makefile.orig Makefile + + export CROSS_COMPILE="${TARGET_PREFIX}" oe_runmake HOSTPGEN=${STAGING_BINDIR_NATIVE}/pgen \ HOSTPYTHON=${STAGING_BINDIR_NATIVE}/python \
Signature · 1 line
1.7.6.4
Re: [CONSOLIDATED PULL 17/17] python: skip setup.py 'import check' when cross-compiling
Missing PR bump!
Op 8 nov. 2011, om 20:22 heeft Saul Wold het volgende geschreven:
Quoted 101 lines
From: Tom Zanussi <[email protected]>
build_extension()insetup.py, as part of the build process, does an 'import check' on the built extension. The import check in turn dlopen()'s the shared library associated with the extension, which isn't something that makes sense if that library was cross-compiled for a different architecture.This was noticed with an
x86_64target that was compiled with avx support, because it caused 'illegal instruction' exceptions:|
/bin/sh: line 1: 14575 Illegal instruction ... -E ./setup.py -q buildFor other target architectures, it doesn't necessarily cause illegal instruction exceptions, but still fails. For example, on arm, the failure pathway causes this warning:
*** WARNING: renaming "cmath" since importing it failed: .../cmath.so:
wrong ELF class: ELFCLASS32This patch to
setup.pyand the associated recipe changes allow the whole 'import check' logic to be skipped when cross-compiling.Signed-off-by Tom Zanussi <[email protected]>---
.../python/setup_py_skip_cross_import_check.patch| 27 ++++++++++++++++++++
meta/recipes-devtools/python/python_2.7.2.bb | 5 +++2 files changed, 32 insertions(+), 0 deletions(-)create mode 100644
meta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patchmeta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch new file +27
-0@@ -0,0 +1,27 @@ +This patch skips over the 'import check' setup.py does when building +extensions. This generally won't work when cross-compiling. + +Upstream-Status: Inappropriate [embedded-specific] + +Signed-off-by: Tom Zanussi <[email protected]> + +Index: Python-2.7.2/setup.py +=================================================================== +--- Python-2.7.2.orig/setup.py 2011-11-04 16:46:34.553796410 -0500 ++++ Python-2.7.2/setup.py 2011-11-04 16:59:49.692802313 -0500 +@@ -287,6 +287,15 @@ + (ext.name, sys.exc_info()[1])) + self.failed.append(ext.name) + return ++ ++ # If we're cross-compiling, we want to skip the import check ++ # i.e. we shouldn't be dynamically loading target shared libs ++ if os.environ.get('CROSS_COMPILE') is not None: ++ self.announce( ++ 'WARNING: skipping import check for cross-compiled "%s"' % ++ ext.name) ++ return ++ + # Workaround for Mac OS X: The Carbon-based modules cannot be + # reliably imported into a command-line Python + if 'Carbon' in ext.extra_link_args:meta/recipes-devtools/python/python_2.7.2.bb +1
-0@@ -18,6 +18,7 @@ SRC_URI += "\ file://multilib.patch \ file://cgi_py.patch \ file://remove_sqlite_rpath.patch \ + file://setup_py_skip_cross_import_check.patch \"
S = "${WORKDIR}/Python-${PV}"
meta/recipes-devtools/python/python_2.7.2.bb +4
-0@@ -57,6 +58,8 @@ do_compile() { # then call do_install twice we get Makefile.orig == Makefile.sysroot install -m 0644 Makefile Makefile.sysroot + export CROSS_COMPILE="${TARGET_PREFIX}" + oe_runmake HOSTPGEN=${STAGING_BINDIR_NATIVE}/pgen \ HOSTPYTHON=${STAGING_BINDIR_NATIVE}/python \ STAGING_LIBDIR=${STAGING_LIBDIR} \@@ -78,6 +81,8 @@ do_install() { # make install needs the original Makefile, or otherwise the inclues would # go to ${D}${STAGING...}/... install -m 0644 Makefile.orig Makefile + + export CROSS_COMPILE="${TARGET_PREFIX}" oe_runmake HOSTPGEN=${STAGING_BINDIR_NATIVE}/pgen \ HOSTPYTHON=${STAGING_BINDIR_NATIVE}/python \Signature · 7 lines
1.7.6.4 _______________________________________________ Openembedded-core mailing list [email protected] http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
Re: [CONSOLIDATED PULL 17/17] python: skip setup.py 'import check' when cross-compiling
In reply to Saul WoldHow I wish I could notice the patch this morning so I could save 1 day!
Anyway, thanks TomZ for the patch! :-)
Thanks,
-- Dexuan
Quoted 124 lines
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Saul Wold
Sent: Wednesday, November 09, 2011 3:23 AM
To: [email protected]
Subject: [OE-core] [CONSOLIDATED PULL 17/17] python: skipsetup.py'import check' when cross-compilingFrom: Tom Zanussi <[email protected]>
build_extension()insetup.py, as part of the build process, does an 'import check' on the built extension. The import check in turn dlopen()'s the shared library associated with the extension, which isn't something that makes sense if that library was cross-compiled for a different architecture.This was noticed with an
x86_64target that was compiled with avx support, because it caused 'illegal instruction' exceptions:|
/bin/sh: line 1: 14575 Illegal instruction ... -E ./setup.py -q buildFor other target architectures, it doesn't necessarily cause illegal instruction exceptions, but still fails. For example, on arm, the failure pathway causes this warning:
*** WARNING: renaming "cmath" since importing it failed: .../cmath.so:
wrong ELF class: ELFCLASS32This patch to
setup.pyand the associated recipe changes allow the whole 'import check' logic to be skipped when cross-compiling.Signed-off-by Tom Zanussi <[email protected]>--- .../python/setup_py_skip_cross_import_check.patch | 27 ++++++++++++++++++++ meta/recipes-devtools/python/python_2.7.2.bb | 5 +++ 2 files changed, 32 insertions(+), 0 deletions(-) create mode 100644meta/recipes-devtools/python/python/setup_py_skip_cross_import_check. patch
diff --git a/meta/recipes-devtools/python/python/setup_py_skip_cross_import_chec k.patch b/meta/recipes-devtools/python/python/setup_py_skip_cross_import_chec k.patch
new file mode 100644
index 0000000..6ccdb94
---/dev/null
+++
b/meta/recipes-devtools/python/python/setup_py_skip_cross_import_chec k.patchmeta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch new file +9
-0@@ -0,0 +1,27 @@ +This patch skips over the 'import check' setup.py does when building +extensions. This generally won't work when cross-compiling. + +Upstream-Status: Inappropriate [embedded-specific] + +Signed-off-by: Tom Zanussi <[email protected]> + +Index: Python-2.7.2/setup.py +===================================================================
+---Python-2.7.2.orig/setup.py2011-11-04 16:46:34.553796410 -0500
++++Python-2.7.2/setup.py2011-11-04 16:59:49.692802313 -0500
+@@ -287,6 +287,15 @@meta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch new file +9
-0+ (ext.name, sys.exc_info()[1])) + self.failed.append(ext.name) + return ++ ++ # If we're cross-compiling, we want to skip the import check ++ # i.e. we shouldn't be dynamically loading target shared libs ++ if os.environ.get('CROSS_COMPILE') is not None: ++ self.announce( ++ 'WARNING: skipping import check for cross-compiled"%s"' %
++ ext.name)
++ return
++
+ # Workaround for Mac OS X: The Carbon-based modules cannot
bemeta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch new file +2
-0+ # reliably imported into a command-line Python + if 'Carbon' in ext.extra_link_args:meta/recipes-devtools/python/python_2.7.2.bb +3
-0@@ -18,6 +18,7 @@ SRC_URI += "\ file://multilib.patch \ file://cgi_py.patch \ file://remove_sqlite_rpath.patch \ + file://setup_py_skip_cross_import_check.patch \ " S = "${WORKDIR}/Python-${PV}"@@ -57,6 +58,8 @@ do_compile() { # then call do_install twice we get Makefile.orig == Makefile.sysroot install -m 0644 Makefile Makefile.sysroot + export CROSS_COMPILE="${TARGET_PREFIX}" + oe_runmake HOSTPGEN=${STAGING_BINDIR_NATIVE}/pgen \ HOSTPYTHON=${STAGING_BINDIR_NATIVE}/python \ STAGING_LIBDIR=${STAGING_LIBDIR} \@@ -78,6 +81,8 @@ do_install() { # make install needs the original Makefile, or otherwise the inclueswould
meta/recipes-devtools/python/python_2.7.2.bb +2
-1# go to ${D}${STAGING...}/... install -m 0644 Makefile.orig Makefile + + export CROSS_COMPILE="${TARGET_PREFIX}" oe_runmake HOSTPGEN=${STAGING_BINDIR_NATIVE}/pgen \ HOSTPYTHON=${STAGING_BINDIR_NATIVE}/python \ --1.7.6.4
_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
Re: [CONSOLIDATED PULL 17/17] python: skip setup.py 'import check' when cross-compiling
Quoted 2 lines · Dexuan
How I wish I could notice the patch this morning so I could save 1 day!
??? Can you please explain what you mean?
Quoted 2 lines
Anyway, thanks TomZ for the patch! :-)
You're welcome, I guess!
Tom
Quoted 128 lines
Thanks,
-- DexuanQuoted 124 lines
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Saul Wold
Sent: Wednesday, November 09, 2011 3:23 AM
To: [email protected]
Subject: [OE-core] [CONSOLIDATED PULL 17/17] python: skipsetup.py'import check' when cross-compilingFrom: Tom Zanussi <[email protected]>
build_extension()insetup.py, as part of the build process, does an 'import check' on the built extension. The import check in turn dlopen()'s the shared library associated with the extension, which isn't something that makes sense if that library was cross-compiled for a different architecture.This was noticed with an
x86_64target that was compiled with avx support, because it caused 'illegal instruction' exceptions:|
/bin/sh: line 1: 14575 Illegal instruction ... -E ./setup.py -q buildFor other target architectures, it doesn't necessarily cause illegal instruction exceptions, but still fails. For example, on arm, the failure pathway causes this warning:
*** WARNING: renaming "cmath" since importing it failed: .../cmath.so:
wrong ELF class: ELFCLASS32This patch to
setup.pyand the associated recipe changes allow the whole 'import check' logic to be skipped when cross-compiling.Signed-off-by Tom Zanussi <[email protected]>--- .../python/setup_py_skip_cross_import_check.patch | 27 ++++++++++++++++++++ meta/recipes-devtools/python/python_2.7.2.bb | 5 +++ 2 files changed, 32 insertions(+), 0 deletions(-) create mode 100644meta/recipes-devtools/python/python/setup_py_skip_cross_import_check. patch
diff --git a/meta/recipes-devtools/python/python/setup_py_skip_cross_import_chec k.patch b/meta/recipes-devtools/python/python/setup_py_skip_cross_import_chec k.patch
new file mode 100644
index 0000000..6ccdb94
---/dev/null
+++
b/meta/recipes-devtools/python/python/setup_py_skip_cross_import_chec k.patchmeta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch new file +9
-0@@ -0,0 +1,27 @@ +This patch skips over the 'import check' setup.py does when building +extensions. This generally won't work when cross-compiling. + +Upstream-Status: Inappropriate [embedded-specific] + +Signed-off-by: Tom Zanussi <[email protected]> + +Index: Python-2.7.2/setup.py +===================================================================
+---Python-2.7.2.orig/setup.py2011-11-04 16:46:34.553796410 -0500
++++Python-2.7.2/setup.py2011-11-04 16:59:49.692802313 -0500
+@@ -287,6 +287,15 @@meta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch new file +9
-0+ (ext.name, sys.exc_info()[1])) + self.failed.append(ext.name) + return ++ ++ # If we're cross-compiling, we want to skip the import check ++ # i.e. we shouldn't be dynamically loading target shared libs ++ if os.environ.get('CROSS_COMPILE') is not None: ++ self.announce( ++ 'WARNING: skipping import check for cross-compiled"%s"' %
++ ext.name)
++ return
++
+ # Workaround for Mac OS X: The Carbon-based modules cannot
bemeta/recipes-devtools/python/python/setup_py_skip_cross_import_check.patch new file +2
-0+ # reliably imported into a command-line Python + if 'Carbon' in ext.extra_link_args:meta/recipes-devtools/python/python_2.7.2.bb +3
-0@@ -18,6 +18,7 @@ SRC_URI += "\ file://multilib.patch \ file://cgi_py.patch \ file://remove_sqlite_rpath.patch \ + file://setup_py_skip_cross_import_check.patch \ " S = "${WORKDIR}/Python-${PV}"@@ -57,6 +58,8 @@ do_compile() { # then call do_install twice we get Makefile.orig == Makefile.sysroot install -m 0644 Makefile Makefile.sysroot + export CROSS_COMPILE="${TARGET_PREFIX}" + oe_runmake HOSTPGEN=${STAGING_BINDIR_NATIVE}/pgen \ HOSTPYTHON=${STAGING_BINDIR_NATIVE}/python \ STAGING_LIBDIR=${STAGING_LIBDIR} \@@ -78,6 +81,8 @@ do_install() { # make install needs the original Makefile, or otherwise the inclueswould
meta/recipes-devtools/python/python_2.7.2.bb +2
-1# go to ${D}${STAGING...}/... install -m 0644 Makefile.orig Makefile + + export CROSS_COMPILE="${TARGET_PREFIX}" oe_runmake HOSTPGEN=${STAGING_BINDIR_NATIVE}/pgen \ HOSTPYTHON=${STAGING_BINDIR_NATIVE}/python \ --1.7.6.4
_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
Re: [CONSOLIDATED PULL 17/17] python: skip setup.py 'import check' when cross-compiling
Zanussi, Tom wrote on 2011-11-09:
Quoted 5 lines · Dexuan
Quoted 2 lines
How I wish I could notice the patch this morning so I could save 1 day!
??? Can you please explain what you mean?
I meant I spent 1 day on debugging the same issue and got the cause but wasn't sure how the patch should be made, so I wanted to ask for suggestion in the ML and before that I tried to find out if anybody reported the same issue and I saw your patch that was sent 1.5 days ago... :-)
Quoted 7 lines
Quoted 2 lines
Anyway, thanks TomZ for the patch! :-)
You're welcome, I guess!
Tom
Thanks,
-- Dexuan
Re: [CONSOLIDATED PULL 17/17] python: skip setup.py 'import check' when cross-compiling
Quoted 8 lines · Dexuan
Zanussi, Tom wrote on 2011-11-09:
Quoted 5 lines · Dexuan
Quoted 2 lines
How I wish I could notice the patch this morning so I could save 1 day!
??? Can you please explain what you mean?
I meant I spent 1 day on debugging the same issue and got the cause but wasn't sure how the patch should be made, so I wanted to ask for suggestion in the ML and before that I tried to find out if anybody reported the same issue and I saw your patch that was sent 1.5 days ago... :-)
Yeah, sure - I'm just kind of curious since I'd never seen it before until I added the avx instruction support.
Just out of curiosity, what were your host and target arches?
Tom
Quoted 12 lines
Quoted 7 lines
Quoted 2 lines
Anyway, thanks TomZ for the patch! :-)
You're welcome, I guess!
Tom
Thanks,
-- Dexuan
Re: [CONSOLIDATED PULL 17/17] python: skip setup.py 'import check' when cross-compiling
Zanussi, Tom wrote on 2011-11-10:
Quoted 14 lines · Dexuan
Quoted 5 lines
Zanussi, Tom wrote on 2011-11-09:
Quoted 4 lines · Dexuan
Quoted 2 lines
How I wish I could notice the patch this morning so I could save 1 day!
Quoted 7 lines
Quoted 1 line
??? Can you please explain what you mean?
I meant I spent 1 day on debugging the same issue and got the cause but wasn't sure how the patch should be made, so I wanted to ask for suggestion in the ML and before that I tried to find out if anybody reported the same issue and I saw your patch that was sent 1.5 days ago... :-)
Quoted 2 lines
Yeah, sure - I'm just kind of curious since I'd never seen it before until I added the avx instruction support.
Actually I think the issue is always there(at least from when python was upgrade to 2.7.x).
e.g., in target, run "python" and try "import grp", we'll get an error saying the built-in module grp couldn't be found; in python's log.do_compile, we can see the obvious warnings.
Quoted 2 lines
Just out of curiosity, what were your host and target arches?
I'm trying the task "self hosted image"(https://wiki.yoctoproject.org/wiki/Build_Appliance_Design):
in target, I found bitbake's do_package failed because "import grp" failed, so I had to look into the issue.
Thanks,
-- Dexuan
Quoted 8 lines · Saul Wold
Richard,
I have reviewed and build/booted these changes.
There are 2 changes from Tom that I need your advice on, please review the gmp and python changes, I do not want to hold these up if they are correct, but I wanted to think if there was a better way to handle the python one.
The gmp one looked correct, the python one looked reasonable too. I've taken them both, if a better solution comes up for the python issue I'll take it though.
Quoted 29 lines
The following changes since commit
157db263b50fabf4afcbea3c2a5364c0752ae0e2:clutter-gst: fix patch filename (2011-11-08 17:33:19 +0000)
are available in the git repository at:
git://git.openembedded.org/openembedded-core-contribsgw/stage
http://cgit.openembedded.org/cgit.cgi/openembedded-core-contrib/log/?h=sgw/stageMark Hatle (1):
pseudo: Uprev pseudo to version 1.2Matthew McClintock (1):
Add newIMAGE_CLASSESvariable for classes for image generationOtavio Salvador (5):
libcap: fix sstate for native packagefiles/device_table-minimal.txt: add/dev/kmsg
dhcp: drop unused dhcp3.inc file
dhcp: rename dhcp4.inc to dhcp.inc
dhcp: move server configuration to dhcp-server-configTom Zanussi (6):
tune-corei7: Add tune file for Intel Core i7 machines
tune-ivb: Add tune file for Intel Ivy Bridge machines machine/include/tune-atom.inc: add atom-specific settings ** gmp_5.0.2: SetCC_FOR_BUILDtoBUILD_CC
libzypp: fix mishandling of hyphenated arches
** python: skipsetup.py'import check' when cross-compiling
Quoted 5 lines
Xiaofeng Yan (4):
libarchive: update to 2.8.5 distro_tracking_fields.inc: update to libarchive to 2.8.5
qmmp: update to 0.5.2 distro_tracking_fields.inc: Modify the status of qmmp
I took everything except the tune file changes from Tom which I need to discuss.
Cheers,
Richard