| Age | Commit message (Collapse) | Author | Files |
|---|
|
This fixes builds where master.list doesn't already exist. This
change was meant to be part of the previous sstate commit but ended
up separated.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
We have a potential problem where two sstate packages try and touch the same
file. This adds code which will print a warning whenever this happens.
The implementation does but by maintaining a master file list and comparing
file accesses against this. There are a number of places we have duplicate
accesses which are harmless, mostly in the deploy directory so these
are whitelisted.
For now the code prints warnings, this could be strengthened in future to
become error messages. Whilst working on this code on and off over the past
few months various issues were uncovered, some serious.
[YOCTO #238]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* The "tar-package" is used for saving the "Source" list for rpmbuild,
there is no such a file when "ARCHIVER_MODE[type] ?= srpm", and there
would be errors, it hadn't happen before was becuase that the remove
function didn't work. Let the "rpmbuild --rmsource" to remove the
Sources, and the remove function will just remove the tar-package file.
* Remove several unwanted "try ... exception" sentences, let the error
raise rather than ignore them when the error happens.
* Remove several un-needed code.
[YOCTO #2619]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
A part of the code in copyleft_compliance.bbclass is duplicated to
archiver.bbclass, remove the duplicated lines.
[YOCTO #2619]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
The configuration value from the conf file is lower case, e.g. srpm,
tar, so there is no reason to use the upper case which makes things
complicated.
[YOCTO #2619]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
* Fix the fakeroot issue
The archiver.bbclass is used for archiving sources, patches, and logs,
it uses the "rpmbuild -bs" from the package_rpm.bbclass to generate the
.src.rpm, but it didn't work (it's not easy to explain it clearly):
Reason:
- It directly used the "fakeroot" command, we don't have such a
command in native tools, so it would use the fakeroot from the host,
and it would fail when there is no fakeroot on the host.
- The "rpmbuild -bs" doesn't need to work under root, but it is in the
function do_package_write_rpm which is running under fakeroot, and
"rpmbuild" needs to know the source file's user/group name, the source
file is the tarball which is created by the postfuncs of do_unpack
or do_patch which doesn't use the fakeroot, so the created file's
owner would be the real user, e.g.: robert, but there is no such a
user under our native tools' fakeroot(pseudo), then the rpmbuild would
fail. It worked when use the host's fakeroot in the past was because
that the host's fakeroot knows the users on the host.
Fix:
- Remove the incorrect "fakeroot".
- Change the source file's owner to root.root under fakeroot will fix the
problem.
* Other fixes:
- The typo: "do_remove_taball -> do_remove_tarball" which will cause the
tarball is not removed.
- Add the _sourcedir defination to the rpmbuild command since the the
SOURCES would be added to the specfile when archiver.bbclass is
inherited, otherwise there would be errors when "rpmbuild -bb", though
the build is OK. It only added the defination to "rpmbuild -bs",
didn't add to "rpmbuild -bb".
[YOCTO #2619]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
* Make the comment line under 80 characters.
* Use the triple double quotes for the docstring.
* Add a whitespace behind the comma (,).
* Other minor fixes.
[YOCTO #2619]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Several fixes:
* It uses mixed tab and whitespace as the indent in one function,
Fix them to use "4 spaces" as the indent.
* Remove the unwanted blank, for example, more than one blank lines
appeared together.
* Remove the tail whitespace.
[YOCTO #2619]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
directories from ${datadir}
These cause various file conflicts and aren't useful within the sysroot anyway. We might
as well remove them. Whitelisting the contents of ${datadir} that *is* useful might
be a better option in future.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The commit a9f11849 [linux-yocto: explicitly export KMETA to scripts]
allows the meta branch name to be changed by exporting it to all
phases of the build.
But if a custom kernel without a meta branch is built, we end up
passing an empty string to the creation scripts, which breaks the
build since input is expected.
Inhibiting the export of KMETA to the creation scripts when empty
fixes the problem.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
To support custom repositories that set a SRCREV and that only have
a single master branch, do_validate_branches needs a special case
for 'master'. We can't delete and recreate the branch, since you
cannot delete the current branch, instead we must reset the branch
to the proper SRCREV.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
In the kernel's tools/perf/Makefile CFLAGS was previously hard coded to
contain "-I/usr/include/slang" to work with hosts that have
"/usr/include/slang/slang.h" as well as hosts that have
"/usr/include/slang.h". This path can cause compile warnings like:
cc1: warning: '/usr/include/slang' doesn't exists.
or
cc1: warning: include location "/usr/include/slang" is unsafe for
cross-compilation [-Wpoison-system-directories]
Then in some cases warnings become errors if WERROR is enabled hence
build errors.
In coordination with a kernel fix, we can fix this error for all
kernels by modifying the perf Makefile within the staged kernel
source.
Signed-off-by: Liang Li <liang.li@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
"ERROR: Nothing PROVIDES 'grub-efi-x86_64-native'" is returned when
trying to build meta-intel/meta-cedartrail with DEFAULTTUNE = "core2-64"
Used TRANSLATED_TARGET_ARCH instead of TARGET_ARCH in grub-efi.bbclass
[YOCTO #3013]
Signed-off-by: Mihai Lindner <mihaix.lindner@linux.intel.com>
|
|
The last two distutils changes progressivly broke the builds. Firstly they
moved things from the site_packages directory to being higher up the tree
which introduced package QA warnings as a side effect. Secondly, it interacts
badly with setuptools which passes in --root=${D} itself.
This patch restores the original directory layout, hence fixing the QA
warnings and also passes extra options to setuptools to deal with the
--root option it passes.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This let's the user override install-lib argument again if it needs
to be something else, otherwise things like python-setuptools
won't be able to modify the install-lib dir
This fixes a new issue exposed by my previous distutils patch
that fixed the python modules default install location. Also,
it removes running the install step twice which was inadvertant
Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This function takes quite a number of arguments and can be tricky to use
properly; this is not made easier if it is undocumented, so document all
of the arguments. (No functional changes, comments only.)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The incremental rpm generation usually broke when package_rpm.bbclass
changed, change its implementation to make it more stable:
* It depended on the previous and current saved manifest files in
the past, it would break when the manifest changed. Now query the
previous and current installed pkgs from rootfs/var/lib and
rootfs/install/, this would be more reliable, the manifest's change
would not affect it any more.
* Add explanations before package_install_internal_rpm to explain what
does the function do.
* Remove an unwanted "awk '{print $1}'".
[YOCTO #2906]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In toolchain edge cases it's possible for the RPATH of a library to be
set to something like "/usr/lib/../lib". This should be detected as
"/usr/lib" and generate a warning.
Signed-off-by: Andy Ross <andy.ross@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
It was pointed out we have a number of weird indentations in the python functions.
This patch cleans up 3, 7 and other weird indentations for the core bbclass files.
It also fixes some wierd (odd) shell function indentation which my searches picked up.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
texi2html is not actually required to build world of OE-Core anymore,
so we don't need to mandate it. The only difference without it (verified
with buildhistory) is that groff doesn't produce HTML documentation; the
rest of its docs are still produced and packaged and no other packages
are affected.
Part of the work towards [YOCTO #2423].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* I have 2.6.37 kernel without this directory and do_install now fails
after 813256bf7bb6e26d542d5f769e2802564116ebe5
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The kern-tools scripts can support a meta branch and directory of a name that
isn't "meta", but they need the name passed through the environment variable
KMETA. ensuring that KMETA is exported in the shell environment sets the stage
to support flexible meta branch name.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
The kernel branch is no longer required by the yocto-kern-tools
to locate BSP feature descriptions (it is the MACHINE:KTYPE
descriptor), so we no longer require that the BSP branch be
explicitly set.
If a kernel branch is explicitly set, it is now used to trigger
a checks to ensure that the branch really is being built.
Otherwise the branch that the machine description creates will
be built (just as it always was).
This further simplies the use and configuration of a linux-yocto
based kernel recipe.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
asm/unistd.h includes asm/unistd_64.h on x86_64
and asm/unistd_32.h on i386 but these files are
generated files in 3.4 and when we do 'make clean' they get
deleted and it shows up as an error when building
external modules. May be its a 3.4 kernel bug may be not
but make clean should have left the tree in
a shape to build modules but it does not.
Probably we should run make modules_prepare after having
run make clean.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
PACKAGECONFIG flag code only handles that when it has 3 or 4 items in flag,
it may have a stale data if some flags doesn't need DEPENDS and RDEPENDS.
Signed-off-by: Yao Zhao <yao.zhao@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
There is a race where do_clean tries to clean WORKDIR but there are logfiles
written into ${T} by bitbake and this can lead to exceptions due to open files.
The easiest solution is to redirect T to a different location for the do_clean
task, hence avoiding the errors and also allowing the logfiles to be visible
somewhere. ${LOG_DIR} seems an appropriate place for this.
[YOCTO #2846]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* There would be race issue if we put the lock to the WORKDIR, for
example:
bitbake core-image-sato core-image-sato-sdk
If the lock is in their own WORKDIR, the apt-rootfs.conf and
Packages.gz maybe be written by two tasks at the same time, which
would cause unexpected errors.
* Create ${target_rootfs}/etc since the "tar -C" needs it.
Note:
* The rpm has put the lock to DEPLOY_DIR_RPM
* The ipk doesn't need it since it has locks for each deploy directory
and put the opkg.conf in his own WORKDIR, which doesn't like deb put
the apt-rootfs.conf in ${STAGING_ETCDIR_NATIVE}/apt/.
[YOCTO #2495]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
non-SPDX license name in license.bbclass
Signed-off-by: Florin Sarbu <florin.sarbu@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Without this some modules will be intalled in /usr/lib/python2.6/
instead of /usr/${libdir}/python2.6
Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
During an attemptonly install, if you try to install two packages that
conflict with each other, a message is generated:
error: Failed dependencies:
dropbear conflicts with openssh-6.0p1-r4.ppc603e
dropbear conflicts with openssh-sshd-6.0p1-r4.ppc603e
This is placed onto stderr (vs stdout), which was not being redirected.
This allowed the log-check code to find the magic key of "Failed" causing
the package installation process to fail. Instead the proper behavior is to
simply ignore the error and proceed with the install.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This fixes a case when RPATH embedded in program have one of
its path already relative to ORIGIN. We were losing that path
if such a path existed. This patch appends it to the new edited
rpath being created when we see it.
so RPATH like below
(RPATH) Library rpath:
[$ORIGIN/../lib/amd64/jli:$ORIGIN/../jre/lib/amd64/jli]
would end up being empty
but after this patch its kept intact
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
contain that path
Without this change, a path to "/lib/xxx" or "/usr/lib/xxx" would also
attempt to be remapped to be relative to $ORIGIN which makes no sense.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
When chrpath fails, prefix the error message with the name of the recipe
that is being processed, and include the the output from chrpath, as
well as making the calling task actually fail.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
When building the meta-toolchain, the binaries didn't get relocatable
RPATHs. They were hardcoded to the default path. Hence, if one had
already installed one SDK in the default path and one in another
location, the later toolchain's binaries would search and load libraries
from the first location, ending in a "Segmentation Fault".
[YOCTO #2927]
Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The problem appears if multiple setup environment scripts are found. In
order to find only the script we're interested in, I removed globbing in
matching pattern with ${REAL_MULTIMACH_TARGET_SYS} that will be expanded
to the correct string.
Also, fix a problem when changing the scripts/configs. The grep pattern
matched also files that contained "text" in their name.
Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
It turns out that the result of getVarFlags is not a list, it's
a dict. So "getVarFlags(...) or []" does not reliably produce
something with a .items. This escaped detection because our
local build environment never ends up running builds without
PNBLACKLIST entries.
Signed-off-by: Peter Seebach <peter.seebach@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If chrpath failed here we were just silently ignoring it.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Remove the Pimlico applications (Contacts, Dates, Tasks). oe-core isn't a
reference PIM stack, so we don't need this.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This class registers postinst/postrm hooks to compile the GSetting schemas, and
adds the relevant dependencies. If possible the postinst happens on the host to
avoid running it on the target.
[ YOCTO #2798 ]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Rename rpmresolve's -d option to -t and make -d enable debug output;
add a -o option to specify the output file (so rpm debug output doesn't
go to the output file) and also add a little more detail to some of the
error messages.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
kernel.bbclass is a common class for many different kernel versions,
and as such the default module_autoload and module_conf values that
it contains are not relevant, or out of date for most builds.
Individual kernel recipes or distros can better specify these
values to meet their needs.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
The PNBLACKLIST feature does not currently work with multilibs,
because they have different ${PN}. The obvious thing to do is to
do this at the same point that we do the PREFERRED_PROVIDER
and PREFERRED_VERSION fixups. (Making the PNBLACKLIST check
do the for-each-multilib check requires it to do the multilib
list generation repeatedly.)
Signed-off-by: Peter Seebach <peter.seebach@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
This is useful when you have a rather large set of compatible distros. For
example: Centos 5.4, 5.5, 5.6, etc, RHEL server 5.x, RHEL workstation 5.x.
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
libdir should be specified, or else mklibs won't work for 64bit targets.
It wouldn't be able to find the libs.
Traceback (most recent call last):
File "<build>/bitbake_build/tmp/sysroots/i686-linux/usr/bin/x86_64-wrs-linux/mklibs", line 553, in <module>
header = elf_header(find_lib(libraries.copy().pop()))
File "<build>/bitbake_build/tmp/sysroots/i686-linux/usr/bin/x86_64-wrs-linux/mklibs", line 89, in elf_header
raise Exception("Cannot find lib: " + obj)
Exception: Cannot find lib:
Signed-off-by: Jesse Zhang <sen.zhang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The base_libdir change in 5b8a4798ea2ea7df66bb53c26448251ea7da3dd9
breaks the kernel build for 64 bit machines. Revert this part of the
change.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If ${PN} is not in PACKAGES then don't add it as an RDEPENDS for each
locale package. This fixes the installation of gcc-runtime locale
packages, for example.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Always write the metadata revisions for each layer into a machine-
readable "metadata-revs" file so that you can potentially link changes
in the output back to changes in the metadata. (Unlike the existing
similar build-id file, this is not specific to image changes.)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Save PKG (the actual output package name, which is often different due
to debian renaming), and PKGE/PKGV/PKGR (which may be manipulated in
certain special cases e.g. gitpkgv.bbclass in meta-oe, the
external-sourcery-toolchain recipe, etc.) Note that these are only
written when they are different from the normal package name in the
case of PKG, or PE/PV/PR for the other variables.
Also, use PKGE/PKGV/PKGR instead of PE/PV/PR when comparing package
versions since these actually represent the version that the package
manager sees.
Implements [YOCTO #2787].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Write the value of these package script variables into the packageinfo
so that any changes to them can be tracked (in separate files since they
are multi-line).
Inspired by an earlier patch from Andreas Müller <schnitzeltony@googlemail.com>
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
getlastrecipeversion and readRecipeInfo weren't called anywhere.
Remove them so we don't have to keep them up-to-date with future
additions to the recipe info file.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
