| Age | Commit message (Collapse) | Author | Files |
|---|
|
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Rather than pause for 5 seconds, test the return code of the command and
require user input before exiting on failure. This avoids pausing after
successful command execution as well as possibly not waiting long enough
if the user happens to be doing something else for 5 seconds.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
|
|
perf has been coupled to the kernel packages via kernel.bbclass.
While maintaining the build of perf out of the kernel source tree
is desired the package coupling has proved to be awkward in
several situations such as:
- when a kernel recipe doesn't want to build/provide perf
- when licensing of dependencies would prohibit perf and hence
the kernel from being built.
To solve some of these problems, this recipe is the extraction of
the linux-tools.inc provided perf compilation into a standalone
perf recipe that builds out of the kernel source, but is otherwise
independent.
No new functionality is provided above what the linux-tools.inc
variant provided, but the separate recipe provides baseline for
adding new functionality.
Signed-off-by: Liang Li <liang.li@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The kernel source tree in the sysroot has all unecessary source
code removed. The existing use case is to support module building
out of the sysroot, but as more toolsa are moved into the kernel
tree itself there are new use cases for the kernel sysroot source.
To avoid putting dependencies on the kernel, and to be able to
individually build and package these tools out of the source tree,
we can save $kerndir/tools and $kernddir/lib from being removed.
This enables tools like perf to be built our of the kernel source
in the sysroot, without significantly increasing the amount of
source in the sysroot.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Fixes:
* Remove the "echo '# Remove manifest padding....' > remove.manifest,
The remove.manifest would be used via "rpm -e `remove.manifest`",
there would be error since there is no pkg called: Remove, manifest or
padding
* The incremental.manifest can't be null when used by rpm, so check it
before use.
* The rpm needs:
--root "${target_rootfs}/install"
when use:
-D "_dbpath ${target_rootfs}/install"
Otherwise it would use the ${target_rootfs} as the root, and use the
${target_rootfs}/var/lib/rpm as the dbpath, this is OK in a fresh
installation, but there would be errors when increment rpm generation.
[YOCTO #2617]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Ensure the following results in the kernel being rebuilt, repackaged and
re-deployed in the final step:
bitbake virtual/kernel
bitbake -c menuconfig virtual/kernel
[ make changes to the kernel configuration and save ]
bitbake virtual/kernel
If there are no changes to the configuration saved, the rebuild will not
be triggered.
Note that this relies on a function recently added to BitBake and
requires full hashing (i.e. BB_SIGNATURE_HANDLER must be set to a
signature handler that inherits from BasicHash) - if this is not the
case or the function is not available in the version of BitBake being
used this change will do nothing.
Fixes [YOCTO #2256].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
|
|
* without this, kernel upgrades where KERNEL_VERSION is changed
e.g. 3.4.2 -> 3.4.3 generate .dep for running 3.4.2 and after reboot user ends
up without any module loaded to make it worse after reboot nothing is upgraded
to trigger another kernel(-module) postinst to generate .dep for now running 3.4.3
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
Since we now have PN as the end of the package list, we can almost get rid of
lib_package, each recipe can just add PACKAGES =+ PN-bin instead of the inherit
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Since we have re-order the full PACKAGES list, we do not need to reset
the -dev list, we still need this since the -bin package needs to be inserted.
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
This is part of the the package reorder, by having binconfig add -config
will the files first due to the order being greedy.
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Using the host objdump can lead to errors like:
objdump: library.so.1.0.0: File format is ambiguous
objdump: Matching formats: elf32-littlearm elf32-littlearm-symbian elf32-littlearm-vxworks
with certain configurations of binutils.
Signed-off-by: James Limbouris <james@digitalmatter.com.au>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Remove some of the older fields that are not in the new list as we
are able to better automagically generate this directly from the
recipe files the extra files will go away.
To use this, one will have to include the appropirate files, such
as maintainers.inc, upstream_status.inc
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If BBPATH references the working directory, the user is warned and asked
to fix the problem.
[Yocto #1465]
Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Currently, the task just exits if something goes wrong. This adds the
ncurses-native dependency. It also adds a small delay before closing the
window so any messages displayed there can be seen.
Trying to get the kernel build system to correctly find and link with
our copy of ncurses is some kind of nightmare. I ended up having to add
it to HOST_LOADLIBES globally for this task which is rather nasty but I
couldn't find any other way.
[YOCTO #2513]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
An empty manifest will cause a failure condition. We ensure that any install
actions occur only with a manifest larger then size 0. Also ensure that padding
is added to the end of the manifest, instead of the beginning to enable this
size check. (Padding is required for very small manifest files..)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Move the creation of the database configuration files to just before we
perform the install actions. This ensures they will exist even in SDK or
other non-target filesystem images.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add functionality to RPM to directly query the packageorigin (path) from
the resolver database, instead of having to do this via an indirect method.
This results in a minor performance improvement.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
RPM 5.4.8 requires db 5.3.x, so both are upgraded together.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Upstream Gnome projects are starting to migrate to the .xz compress format,
so we need to add this to allow recipes to override the default of .bz2 as
the upstreams make the transition.
[YOCTO #2241]
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In some cases we need to specify linker flags and right
now we do not have a way to communicate that to cmake
based systems. cmake defines CMAKE_C_LINK_FLAGS and CMAKE_CXX_LINK_FLAGS
for these needs. This patch therefore defines two local variables
namely OECMAKE_C_LINK_FLAGS and OECMAKE_CXX_LINK_FLAGS which
can be altered by recipes to tweak linker flags
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
The rpmlib was removed when images that add
"remove_packaging_data_files" to ROOTFS_POSTPROCESS_COMMAND, which would
make the increment rpm image generation doesn't work in the second
build, since list_installed_packages would get incorrect value in the
second build, move the rpmlib to ${T} rather than remove it, and move it
back when INC_RPM_IMAGE_GEN =1.
[YOCTO #2440]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The content to modify this bbclass is as follow:
- Use the existing functions to get license as a directory instead of
rewriting it for avoiding code duplication.
- Use SPDXLICENSEMAP to map licenses
[YOCTO #2473]
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Remove the long obtuse sysroot path from the ldd output.
Make the error message significantly easier to read and understand.
Old Style:
WARNING: QA Issue: keyutils: /home/mhatle/git/oss/oe-core/build-i386/tmp-eglibc/work/i586-oe-linux/keyutils-1.5.5-r1/packages-split/keyutils/sbin/request-key links to something under exec_prefix
WARNING: QA Issue: ldd reports: libkeyutils.so.1 => /home/mhatle/git/oss/oe-core/build-i386/tmp-eglibc/sysroots/qemux86/usr/lib/libkeyutils.so.1 (0xdead1000)
libc.so.6 => /home/mhatle/git/oss/oe-core/build-i386/tmp-eglibc/sysroots/qemux86/lib/libc.so.6 (0xdead2000)
/lib/ld-linux.so.2 => /home/mhatle/git/oss/oe-core/build-i386/tmp-eglibc/sysroots/qemux86/lib/ld-linux.so.2 (0xdead3000)
New style:
WARNING: QA Issue: keyutils: /sbin/request-key, installed in the base_prefix, requires a shared library under exec_prefix (/usr): libkeyutils.so.1 => /usr/lib/libkeyutils.so.1 (0xdead1000)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This fixes an ugly way I was trying to find pkg runtime
data for package and license manifest creation.
rootfs generation times for core-image-minimal:
Prior to patch
real 0m41.570s
user 1m40.466s
sys 0m6.768s
With patch
real 0m27.527s
user 0m9.833s
sys 0m3.496s
Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Conflicts:
meta/classes/license.bbclass
|
|
Per request, adding the package version to the package manifest
file.
Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
|
|
The usability of the archiver classes can be improved, beyond the
simple addition of default values for the variables. A user could
well inherit just archiver rather than the individual useful classes,
and not realize it will do nothing.
[YOCTO #2472]
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
|
|
If there is a patch to Makefile.PL, a Makefile.PL but no Makefile
will be placed in ${B}/.pc/xxx.patch/ after do_patch.
And no Makefile will be generated for *this* Makefile.PL.
While do_configure, the original code tries to sed Makefiles
matching with each Makefile.PL in {B}, so this would fail.
Signed-off-by: Xin Ouyang <Xin.Ouyang@windriver.com>
|
|
After the class changes, absolute symlinks are not being handled correctly
by the class file. This adds handling for absolute symlinks to account for the
pkgdest directory, removing dangling symlink messages from recipes like bzip2.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
'//' in a FILES variable causes hard to track down issues with
packaging. This adds a warning and attempts to auto-correct the issue to
try and make the problem more user friendly.
[YOCTO #2448]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This reverts commit e83d8e58a6b107eea87df0ec233a1bc932b2c6e as the conversion
is not correct. Its replacing readlines() calls which generate an array with
what are effectively strings. There are split("\n") calls missing in many
cases so this needs to be reverted until it gets fixed.
|
|
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Currently, ccache is used if it is present. When building from scratch it gives
no performance improvement and creates a ton of empty directories even when its
not in use.
This change moves ccache support to a bbclass file which the user can choose to
enable. This should make builds more determinstic and make it easier/clearer
to the end user when its being used and when it is not.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Currently we copy the aclocal directory to the build so that autotools
doesn't see .m4 files disappear when its processing them. This can happen
if for example, package X is being rebuilt at the same time as Y and it
gets uninstalled from sstate (assuming there are no dependencies between
X and Y). This code making the copy was added to avoid races but introduces
a race of its own, namely that the files can disappear during the copy.
This patch adds a cp-noerror script which silently ignores such errors
and gives the behaviour we need in this case. It hence fixes issues which
crop up for users and the autobuilder occasionally.
[YOCTO #2485]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In order to show a friendlier error message within Hob that does not
bury the actual sanity error in our typical preamble about disabling
sanity checks, use a separate event to indicate that sanity checks
failed.
This change is intended to work together with the related change to
BitBake, however it has a check to ensure that it does not fail with
older versions that do not include that change.
Fixes [YOCTO #2336].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The variable name has been typo'd as TARGE_ARCH since it was introduced
some time ago, so the check has never worked. Fixing the typo shows that
the test is not quite functional, so let's just remove it:
http://lists.linuxtogo.org/pipermail/openembedded-core/2011-October/010613.html
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Replace os.popen with subprocess.Popen since the older function would
fail (more or less) silently if the executed program cannot be found
There are both bb.process.run() and bb.process.Popen() which wraps the
subprocess module, use it for simplifying the code.
Note: We don't need the "2>/dev/null" or "2>&1" since bb.process.run()
can handle it, it will raise exception when error occurs, we should
handle the exception ourselves if we want to ignore the error.
More info:
http://docs.python.org/library/subprocess.html#subprocess-replacements
[YOCTO #2454]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Replace os.system with subprocess.call since the older function would
fail (more or less) silently if the executed program cannot be found
More info:
http://docs.python.org/library/subprocess.html#subprocess-replacements
[YOCTO #2454]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In order to better support multilib processing, switch from PN
to BPN.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
Refactor in order to:
* Deprecate the old interfaces, but keep them for compatibility
* Provide a new, interface -- capable of working with split packages
* Each update-alternative will now set proper "per-file" provides
Note: this adds a warning message when the older deprecated behavior is
used. The older behavior has been fully tested using oe-core.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
Some image classes such as bootimg save files into ${S} as part of rootfs
generation. For correctness we should therefore clean this at the start of
image generation to ensure reproducibility.
I found this issue when some files I thought should disappear from my rootfs
would not disappear.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
When trying to understand why a QA wanring such as:
ERROR: QA Issue: foo rdepends on bar-dev
it is very difficult to figure out where the bar-dev dependency
comes from, since many of them are added dynamically.
This adds a debug statement that says which dependency adds an
rdepends to the system.
Also, while doing this work, it was noted that the same dependencies
were being scanned for over and over. Instead we shorten the list
by only added to the dep list if the dependency was not already there.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
The following functions in meta/classes/utils.bbclass were not
used by anyone, and they didn't work when I tried to use them:
def oe_popen_env(d):
def oe_run(d, cmd, **kwargs):
def oe_popen(d, cmd, **kwargs):
def oe_system(d, cmd, **kwargs):
There error was:
AttributeError: type object 'str' has no attribute 'getVar'
We have bitbake/lib/bb/process.py to instead, so remove them.
[YOCTO #2489]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
|
|
This enables a user to use bitbake -e even when the sanity checks are
failing.
Signed-off-by: Joshua Lock <josh@linux.intel.com>
|
|
The call to bb.mkdirhier() in check_create_long_filename() can fail with an
OSError, explicitly catch this and report something useful to the user.
Signed-off-by: Joshua Lock <josh@linux.intel.com>
|
|
If the user does not have write permissions to SSTATE_CACHE, detected by
the check_create_long_filename() test failing with a "Permission denied"
value in strerror, then suggest they might want to use the location as
an entry in SSTATE_MIRRORS.
Signed-off-by: Joshua Lock <josh@linux.intel.com>
|
|
Each failure in the sanity message should be reported on a new line.
Signed-off-by: Joshua Lock <josh@linux.intel.com>
|
|
Replace calls to data.getVar(VARIABLE, data_object, expand) to
direct calls to the getVar method the the data_object.
Signed-off-by: Joshua Lock <josh@linux.intel.com>
|
|
At the ConfigParsed event the datastore has yet to be finalised and thus
appends and overrides have not been set.
To ensure the sanity check is being run against the configuration values
the user has set call finalize() on a copy of the datastore and pass that
for all sanity checks.
Signed-off-by: Joshua Lock <josh@linux.intel.com>
|
|
This has caused problems for several users, including the Yocto Project
autobuilder. Since the message was added in order to be more user friendly
revert the change.
This reverts commit 0c0c4efbf92bcf0f8942f17c18525a4b4ed1798c.
|
|
* Deleting .pyo files causes them to get compiled on the target.
* First boot gets *really* slow for python based projects.
* No space gets saved on the target.
* The package manager doesn't know about the files and
therefore fails to uninstall them, occupying space and causing
uninstalled python scripts to remain executable.
* It's inconsistent, because python itself and autotools based
projects already ship .pyo files.
* Probably .pyo files were deleted because .pyc files were
available earlier, but this has changed and OE-Core's python
now only generates optimized .pyo files. Deletion of .pyo was
introduced in 2008, python/04-default-is-optimized.patch
was introduced in 2009.
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
|
