icu will check program install path during do_configure, eg:
checking for a BSD-compatible install... /path/to/install -c
And this path will be writen into pkgdata.inc:
INSTALL_CMD=$(INSTALL-L)
Decided by if install is installed into recipe-sysroot-native during
do_configure stage, the INSTALL_CMD could be
/build/tmp/work/corei7-64-wrs-linux/icu/76-1/recipe-sysroot-native/usr/bin/install
or /build/tmp/hosttools/install if the build is rerun after the sysroot was
extended.
set ac_cv_path_install to install under hosttools to make a deterministic
result of INSTALL_CMD, avoid vary caused by the execute sequence of
another task which DEPENDS on coreutils-native and independent with
do_configure
[RP: Removed paths from ac_cv_path_install to simplify and avoid QA error too]
(From OE-Core rev: 208143e060cda6e22ae1e8c618e033fa9144b323)
Signed-off-by: Changqing Li <changqing.li@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
On big-endian systems the preprocessor define ICU_DATA_DIR
is currently being set to a path ending with the ${PV} of
the recipe.
The PV version string has changed to a '-' separator
since oe-core commit cebe8439cdc656d53355506a31a3782312bf03c5
whereas the build system installs the data files into a
path ending with the dot-separated version of ICU.
This causes the ICU data file to not be detected at runtime,
consequently breaking any dependant applications.
We therefore substitute ${PV} with the dot-separated version
string of ICU, as returned from the icu_install_folder function,
on the ICU_DATA_DIR define on big-endian targets.
(From OE-Core rev: 28cdc0110def011e3d690da1d591076385267ef7)
Signed-off-by: Makarios Christakis <makchrbiz@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update: add JSON parsing library under MIT license.
Drop backported patches which have been merged into the new version.
ICU-22813_rise_buffer_sizes_pkgdata_PR3058.patch - db70adaddc
fix-install-manx.patch
(From OE-Core rev: ef2bd18dd1db7e5f01db76889933bfdbb5415aaf)
Signed-off-by: Jason Schonberg <schonm@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There is a bug libicu that causes libicu to be installed incorrectly when
the build system uses long paths (more than 512 chars).
This condition is not very difficult to trigger on a OE build system
due to the long paths an the deep of the directories that are usually
generated by default.
Also the bug is very subtle and won't be detected by the QA post-install
processes because what this bug causes is that a different version of
libicudata.so.X.Y (one without data) is installed instead of the one
containing the data, but there won't be any file missed on the installation
(just that it installed the wrong one).
See: https://unicode-org.atlassian.net/browse/ICU-22813
This patch backports the fix from upstream/main
(From OE-Core rev: 67d1352873957decacde30ff208fb7bb635b0c5d)
Signed-off-by: Carlos Alberto Lopez Perez <clopez@igalia.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
These were used by external layers and were needed before we had better
unpack instrumenation and the recent unpack directory changes. Drop
them as obsolete.
(From OE-Core rev: 8dae8778ee14caebea92acd0bebf28c5681e39f3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fix is only done for target. Use same code for nativesdk.
(From OE-Core rev: dc6306883cc2c7d4d98d595442e5bf4037a160c5)
Signed-off-by: Oleksiy Obitotskyy <oobitots@cisco.com>
Signed-off-by: Oleksandr Hnatiuk <ohnatiuk@cisco.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The location of files changed during unpackdir transition. Adapt dirs
to fix custom icu data generation.
(From OE-Core rev: bb24c965a55d7e757c1e2a6e76bdb68d2bfb6ce3)
Signed-off-by: Stefan Müller-Klieser <s.mueller-klieser@phytec.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
autoconf-archive is a collection of m4 autoconf macros needed at
build-time, and autoconf-archive-native is a suitable provider as there
is nothing in the recipe that needs to be cross-compiled.
Also if we use DEPENDS=autoconf-archive then the recipe's -dev package
ends up RDEPENDing on autoconf-archive, which isn't correct.
Universally change any DEPENDS on autoconf-archive to the -native form,
and add any missing dependencies that were implicit before.
(From OE-Core rev: 571132415ea7fe2d91c62948f2b6aa553eafa83d)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Upstream have indicated icu-config is deprecated and they don't intend
people to use it any more. Drop our code for it to flush out users and
clean up our recipe a bit. pkg-config is a much better alternative.
(https://github.com/unicode-org/icu/pull/2150)
(From OE-Core rev: 9b121df8d9f2a46bb08992725f97ba03e878a609)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update: copyright years
(From OE-Core rev: 85cbbe2da017b39a985e07bfeefb6aa83ddfea65)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The make_icudata task is set before the configure task then the
dependencies for this task are not populate yet.
Fixed it by adding do_prepare_recipe_sysroot task dependency to the
make_icudata task.
(From OE-Core rev: c134f62062a6359d2d5831ba4cb9694035e4f25a)
Signed-off-by: Kory Maincent <kory.maincent@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is the result of automated script conversion:
scripts/contrib/convert-overrides.py <oe-core directory>
converting the metadata to use ":" as the override character instead of "_".
(From OE-Core rev: 42344347be29f0997cc2f7636d9603b1fe1875ae)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
icu data generation was found to be racy, and causig AB failures;
making it serial and leaving it on is not an option as it regresses
to several minutes.
The specific bug is that rules.mk has:
LD_LIBRARY_PATH=../lib:../stubdata:../tools/ctestfw:$LD_LIBRARY_PATH ../bin/gencnval -s . -d ./out/build/icudt66l mappings/convrtrs.txt
which creates a file and numerous rules like
LD_LIBRARY_PATH=../lib:../stubdata:../tools/ctestfw:$LD_LIBRARY_PATH ../bin/genrb -s ./misc -d ./out/build/icudt67l -i ./out/build/icudt67l -k -q numberingSystems.txt
which quietly read it. There is no prerequisite for the former to complete first.
The race is extra complicated to fix as rules.mk is itself
generated through a custom in-tree python tool.
(From OE-Core rev: df89e8d1136fd406ba35ae573e2cb0cfc88c6aad)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Both parts are rather convoluted, so it is better to have them in a single file.
(From OE-Core rev: e8faa3416bc85dab7e7e638448dcb11ff88354a3)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
- 0001-icu-Added-armeb-support.patch - rebased
- 0001-Fix-big-endian-build.patch - removed, already included in new version
- CVE-2020-10531.patch - removed, already included in new version
- icu-pkgdata-large-cmd.patch - removed, implemented correct size
(From OE-Core rev: 62feb846853bcc8982258a224a3e84090d6559a2)
Signed-off-by: Andrej Valek <andrej.valek@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Musl version is just a wrapper to strtod and not an exposed API
(From OE-Core rev: 660fdaa95858485b1caa779f1137a933c1f5d5f6)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Recent versions of ICU (64+) provides a tool for configuring ICU
locale data file with finer granularity [1]
Default generated size for libicudata.so.66.1 is ~27M, which is quite
large for embedded systems and all of them may not even need all locale
data.
This patch calls the icudata buildtool during configure on the icudata
and utilizes a filter called 'filter.json` ( empty by default) to
create the data, default behavior should remain same but someone can
add a filter.json in own layer to configure this data, e.g.
{
"localeFilter": {
"filterType": "language",
"whitelist": [
"en",
"de",
"it"
]
}
}
would only generate the locale data for english/german/italian
This would reduce the size of libicudata.so.66.1 to 12M
Ensure that icudata is generated using host-tools so it can deal with
endianness correctly, when host and target systems have different
endianness
install the icudtata file back into in/ folder so that main build
can now pickup this data file instead of regenerating it and wiping
out the filter changes that are expected to take effect
Use native compiler tools
Update the big-endian support patch to apply to latest Makefile.in
from icudata source and mark it as backport
defer applying 0001-Fix-big-endian-build.patch after moving new data/ in
[1] https://github.com/unicode-org/icu/blob/master/docs/userguide/icu_data/buildtool.md
(From OE-Core rev: 5e5be67744d7ddf5a9ac433ecba02f697a84a325)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Suggested-by: Wouter Meek <w.meek@metrological.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
-License-Update: Copyright year updated to 2020.
(From OE-Core rev: d8b928f60e1274edc3d97f14ebe60e81aa0057f1)
Signed-off-by: Wang Mingyu <wangmy@cn.fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
New releases of ICU are published on github.
(From OE-Core rev: 1a01c30653c2b0282543aa3cdae8d87af871c24a)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Make icu support arm32 BE.
Upstream-Status: Pending
(From OE-Core rev: 244f2e5fb545dd40d020811799a6022e14468eeb)
Signed-off-by: Lei Maohui <leimaohui@cn.fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-update: copyright years changed.
Drop upstreamed/backported patches.
Add a patch to fix big endian build failure.
(From OE-Core rev: 929d37831624fce84580466c4408217c766410f0)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fix install files conflicts between multlib packages by inherit multilib_script:
| file /usr/bin/cairo-trace conflicts between attempted installs of lib32-libcairo-perf-utils-1.14.12-r0.x86 and libcairo-perf-utils-1.14.12-r0.core2_64
| file /usr/bin/icu-config conflicts between attempted installs of lib32-icu-dev-62.1-r0.x86 and icu-dev-62.1-r0.core2_64
| file /usr/bin/gpgrt-config conflicts between attempted installs of lib32-libgpg-error-dev-1.32-r0.x86 and libgpg-error-dev-1.32-r0.core2_64
(From OE-Core rev: 52f2dd97e9dd20dea0f3cdeb2df490d1a4c646aa)
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
1.Upgrade icu from 61.1 to 62.1.
2.0001-i18n-Drop-include-xlocale.h.patch has been merged.
(From OE-Core rev: 8be6cfb97ea8d5a7938c1ccbf8218a7bfa5e86a2)
Signed-off-by: Hong Liu <hongl.fnst@cn.fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Move the recipe specific over-ride for ASNEEDED into the recipe to
make it more apparent that the over-ride is being applied (and that
it should be re-checked on version updates, etc).
(From OE-Core rev: a4c29153c7ffef024b31e7e3a197a09758a7beb4)
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update: copyright years updated, added terms for Google double-conversion
(From OE-Core rev: b5797e80ccfa080bc1e57c5fb1f2f4a39d0266cf)
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
These functions were moved to meta/lib/oe in 2010 and the base_* functions in
utils.bbclass were intended to be a short-term compatibility layer. They're
still used in a few places, so update the callers to use the new functions.
(From OE-Core rev: c97acbd034532895ce57c6717ed1b3ccc7900b0d)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Make sure build host references do not end up being compiled in the image.
This only affects libicutu and icu-dbg.
(From OE-Core rev: e9dfe7eb7f61b909ae7d034e80cfbebc1fad018b)
Signed-off-by: Juro Bystricky <juro.bystricky@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
ICU library libicudata is created with a wrong endianness if the host
and target have different endianness. (e.g. build ICU for qemuppc on
x86-64 host)
See upstream bug report: http://bugs.icu-project.org/trac/ticket/11758
The discussion in oe-core mailing list:
https://www.mail-archive.com/openembedded-core@lists.openembedded.org/msg65155.html
The root cause is the native pkgdata can not generate correct icu data
for different endianness. By default, this data is built into the shared
library libicudata that would cause it corrupt. But ICU also provides
additional options for loading the data. With option
--with-data-packaging=archive, ICU outputs a single icudt.dat file which
can be loaded by the library.
See document: http://userguide.icu-project.org/icudata
So for big endian system, we can use the separated data file which
re-generates by icupkg and set the correct ICU_DATA environment variable
to make sure the library can load the data without problems.
(From OE-Core rev: 77d1b9cda69d09cd7d3c8174ef14b4ee10e5efb6)
Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
