| Age | Commit message (Collapse) | Author | Files |
|---|
|
I ran across this problem when apt 0.7.14 was not fetchable
from regular debian mirrors and existing snapshot mirror
did not have it either since we did not use proper syntax
so it was not hitting that.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
If EXTRA_OECONF_FPU is left set, certain ARM variables related to hard-float
can get pulled in and trigger rebuilds of the crosssdk code. The best solution
is to simply force the variable to a known correct value for the SDK targets
currently supported in the same way as TARGET_FPU.
There is some slight rearrangement of the gcc code to ensure the variable is
always used to call the fpu function.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
There was a bug showing up where the crosssdk recipes were being installed as machine
specific in the sstate-control directory. This turned out to be due to the architecture
fields used by sstate being set incorrectly. The problem is that the crosssdk inherits
the cross class. It therefore needs to be listed in this if statement block before
the cross check, not after.
This should resolve some package architecture issue of crosssdk sstate files.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Having XZ_COMPRESSION_LEVEL on -e -9 and -T 0 will make xz eat more
than 6Gb memory. Reduce this to -6 to make xz to use about 471Mb
on the tested machine.
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
while compressing
Default this variable to 0. This will make xz use as many threads as there are CPU
cores on the system.
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Having -c modifier makes xz to output the compressed data to stdout. In this
way the needed data will be in the do_rootfs log.
Redirect data to ${IMAGE_NAME}.rootfs.${type}.xz .
Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
We should not just replace CCFLAGS with CFLAGS while compiling, because
this may cause run-time errors with perl's DynaLoader.pm.
Tested on qemux86 with new libnet-libidn-perl bb recipe:
root@qemux86:~# perl -e "use Net::LibIDN"
Not a CODE reference at /usr/lib/perl/5.14.2//DynaLoader.pm line 213.
END failed--call queue aborted at /usr/lib/perl/vendor_perl/5.14.2//Net/LibIDN.pm line 213.
Compilation failed in require at -e line 1.
BEGIN failed--compilation aborted at -e line 1.
Signed-off-by: Xin Ouyang <Xin.Ouyang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
It turns out "apt-cache showpkg" does return some information when a
package does not exist but another package recommends it, which can
occur for empty *-dev packages; so use "apt-cache policy" with a
different line count instead.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
In the usual case, the computed path used for debugedit to fix
up path references for the target filesystem is correct. However,
prebuilt binary components, such as prebuilt toolchains, can
have debug paths that do not reflect paths within the local
build directory. Providing an override lets us continue to use
the standard debugedit logic in package.bbclass.
Signed-off-by: Peter Seebach <peter.seebach@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Package install failures due to issues in post install rules.
package_do_shlibs only looks for libraries in a directory "lib", this should be
modified to the variable baselib.
ldconfig_postinst_fragment failure observed on systems without /sbin/ldconfig
[ -x /sbin/ldconfig ] && /sbin/ldconfig
results in the post install rule returning a failure. Modify to
if [ -x /sbin/ldconfig ]; then /sbin/ldconfig; fi
Signed-off-by: Amy Fong <amy.fong@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
This ensures a rebuild does really build from scratch when ccache is in use.
[YOCTO #2696]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If the kernel version string uses characters or symbols that
need to be santized for the package name, we can end up with a
mismatch between module requirements and what the kernel
provides.
The kernel version is pulled from utsrelease.h, which contains
the exact string that was passed to the kernel build, not
one that is santized, this can result in:
echo "CONFIG_LOCALVERSION="\"MYVER+snapshot_standard\" >> ${B}/.config
<build>
% rpm -qp kernel-module-uvesafb-3.4-r0.qemux86.rpm --requires
update-modules
kernel-3.4.3-MYVER+snapshot_standard
% rpm -qp kernel-3.4.3-myver+snapshot-standard-3.4-r0.qemux86.rpm --provides
kernel-3.4.3-myver+snapshot-standard = 3.4-r0
At rootfs assembly time, we'll have a dependency issue with the kernel
providing the santizied string and the modules requiring the utsrelease.h
string.
To not break existing use cases, we can add a second provides to the
kernel packaging with the unsantized version string, and allowing the
kernel module packaging to be unchanged.
RPROVIDES_kernel-base += "kernel-${KERNEL_VERSION}"
% rpm -qp kernel-3.4.3-myver+snapshot-standard-3.4-r0.qemux86.rpm --provides
kernel-3.4.3-MYVER+snapshot_standard
kernel-3.4.3-myver+snapshot-standard = 3.4-r0
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
specific package
This is useful for the scenario where we want to add 'gcc' to
the root file system for all multilib variants
Signed-off-by: Matthew McClintock <msm@freescale.com>
|
|
Prepend the license manifest creation call to ROOTFS_POSTPROCESS_COMMAND
instead of appending to ROOTFS_POSTINSTALL_COMMAND. The latter is not
implemented for the deb backend (and probably ought to just be removed
completely), and by using _prepend we can still ensure it occurs before
package info is removed (and before buildhistory in case it is needed
there in future).
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
|
|
There will be errors when the length of the tmpdir is longer than 410:
1) Longer than 420:
Can't exec "/bin/sh": Argument list too long at /usr/lib/perl/5.10/IO/File.pm line 66.
This error happens on both Ubuntu 10.04 and 10.10 when the pkg needs run
"autoreconf", this is because it passes many files with absolute path to
aclocal, aclocal passes them to perl, this is a limitation of the perl
on Ubuntu 10.04 and 10.10, and the perl-native is not ready at this very
early stage.
2) Longer than 490:
bitbake/lib/bb/persist_data.py", line 197, in connect(database=...)
> return sqlite3.connect(database, timeout=5, isolation_level=None)
OperationalError: unable to open database file
This error happens on Ubuntu 10.04, 10.10 and Fedora 17. This is because
the length of the database in sqlite3 module (host's) can't be longer
than 490 (or little smaller). The python-native is not ready at this
very early stage.
The 2 errors are host related, I think that limit the length of the
TMPDIR to 410 is OK for most of the build, rarely build sets TMPDIR's
longer than 410.
[YOCTO #2434]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
|
|
Enable the remapping for SDK generation, this is required to be able to
create an SDK that targets an alternative multilib. Note, this work does
not finish SDK/multilib support, but it is one more step toward making
it work properly.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
When constructing a multilib based image, it's possible to end up with
the same package listed in multiple steps of the install. During resolution
we use the --replacepkgs option to avoid errors, as the resulting install
solution will be as expected. (We do not enable the replacepkgs option
for the final install step, only the generation of the install solution.)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
An anonymous python chunk configures some multilib settings for both
populate_sdk_rpm and rootfs_rpm. The two classes should contain identical
versions to ensure that the generated multilib configuration is the same
for the SDK and the rootfs.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
bootimg.bbclass using STAGING_DIR_HOST/kernel instead of
STAGING_KERNEL_DIR, resulting in build failure of live images.
| install: cannot stat `/usr/local/dev/yocto/fishriver-test/build/tmp/sysroots/fishriver/kernel/bzImage': No such file or directory
Replace it with STAGING_KERNEL_DIR.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Tested-by: Tom Zanussi <tom.zanussi@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
|
|
On x86, an ELF image file may be stored as a coreboot payload.
The image file is constructed, using the mkelfimage utility,
from a kernel and an initrd.
Signed-off-by: Raymond Danks <ray.danks@se-eng.com>
Acked-by: Darren Hart <dvhart@linux.intel.com>
|
|
This will allow the reporting of these errors as either WARNINGs (default)
or ERRORs if installed_vs_shipped is added to the ERROR_QA of the policy
file (such as a <distro_name>.conf file.
V2: found the code I had intended to send instead of that other junk,
was just not watching what I pushed on that one, sorry. (this is edit in
no in the actual commit message)
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Since this check effective is not a change becase the package
order that was in this file has not really changed no PR Bumps
are needed.
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
[YOCTO #1614]
Add the kernel headers to the kernel-dev package. This packages what was
already built and kept in sysroots for building modules with bitbake.
Making this available on the target requires removing some additional
host binaries.
Move the location to /usr/src/kernel
Before use on the target, the user will need to:
# cd /usr/src/kernel
# make scripts
This renders the kernel-misc recipe empty, so remove it.
As we use /usr/src/kernel in several places (and I missed one in the
previous version), add a KERNEL_SRC_DIR variable and use that throughout
the class to avoid update errors in the future.
Now that we package the kernel headers, drop the
kernel_package_preprocess function which removed them from PKGD.
All *-sdk image recipes include dev-pkgs, so the kernel-dev package will
be installed by default on all such images.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Bruce Ashfield <bruce.ashfield@windriver.com>
CC: Tom Zanussi <tom.zanussi@intel.com>
CC: Khem Raj <raj.khem@gmail.com>
|
|
changes in bitbake
This also deletes the buildall task since I seen usecases for it.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
IMAGE_FEATURES such as 'ssh-server-dropbear' and 'ssh-server-openssh'
can't be both enabled. User can use the following variables to define
the relationship of image features:
IMAGE_FEATURES_REPLACES_foo = "bar" means including image feature "foo"
would replace the image feature "bar".
IMAGE_FEATURES_CONFLICTS_foo = "bar" means including both image features
"foo" and "bar" would cause an parsing error.
Signed-off-by: Lianhao Lu <lianhao.lu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
exclusion list
This appears to be an oversight in the original implementation. All of the
host package types were being ignored except for the SDK cross-canadian type.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
When building an image recipe, you can now build a companion SDK by
calling the populate_sdk task:
bitbake -c populate_sdk core-image-minimal
Note: there are still issues w/ the SDK not working completely with
multilibs.
A lock is required between rootfs and populate_sdk activities to prevent
configuration file clashes and similar package management problems in ipk
and deb based systems. (RPM already had a lock for a different reason.)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
Most of the time skipping per file dependency generation is a bad idea, but
when building a nativesdk or similar you may be required to pickup host
dependencies. These host dependencies can not always be reconciled within
the scope of other nativesdk components, so if we skip them we can facilitate
this unique situation.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
The libc-common attempts to rewrite the package information in a way similar
to debian.bbclass. When it does this, it should be appending to the
dependency variables (RPROVIDES, RREPLACES, and RCONFLICTS), instead
of simply setting a hard coded value. Otherwise the lib package can not
tailor the dependency variables to suite it's needs.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
When configure fails, it usually says "see config.log" yet nobody ever shares
the config.log file meaning the person trying to help invariably has to ask
for more information.
This patch dumps all the config.log files into the main bitbake log files when
configure fails, meaning all the information is present to help someone debug
such failures. It does make the log rather larger but this is preferable to
not having enough information in most cases.
[YOCTO #2463]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In order to enable basic multilib support for the export of an SDK
image, a number of minor changes had to be made:
The value of MULTIMACH_TARGET_SYS needs to be variable. This way we
can define the value to the appropriate multilib. (Also in some cases
the default PACKAGE_ARCH is set to MACHINE_ARCH which is incorrect for
the SDK.) Add a companion REAL_MULTIMACH_TARGET_SYS, based on code
from meta-environment, to allow for this.
We have to convert the do_populate_sdk into a python call, and then
break up the previous items into three parts.
* Image construction
* Setup of environment files
* Generation of the tarball
Then we can iterate over the multilibs to populate the environment files.
Finally, matching changes were needed in the toolchain-scripts file. And
what I presume is an optimization of immediate evaluation for
TOOLCHAIN_CONFIGSITE_NOCACHE and ..._SYSROOTCACHE needed to be done at
runtime, otherwise the wrong values may be used.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
If we combine the do_populate_sdk with the image generation, we want
to avoid the dependency processing unless do_populate_sdk is run.
This requires the bitbake change to implement task based rdepends.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
We want to be able to supply attempt only packages in the SDK in order to
support more advanced SDK images that more closely match specific image
recipes.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
populate_sdk was renamed to populate_sdk_base in order to allow for changes
that may break existing SDK recipes. Any such changes need an analog
in populate_sdk (new version) to restore previous desired behavior.
In addition to the rename, one minor change was made. The _base version
only had the do_populate_sdk as an added task, but no before or after defined.
For compatibility, populate_sdk has do_populate_sdk defined as "after"
do_install and before do_build, this is identical to the original behavior.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
When a recent change, the path to log files may be contained within an
arbitrary directory. To generate the manual log files in the correct path
we should be using the ${BB_LOGFILE}'s path instead of always assuming the
logs go into ${WORKDIR}/temp.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
|
|
PACKAGEVARS
These variables all take package name extensions and should be considered in sstate
dependency processing.
This may also fix some multilib alternatives issues (unconfirmed).
[YOCTO #2056]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If we don't do this it will match on expressions like "NOTE:" which
are not fatal errors.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
these are not strictly dependencies of the do_rootfs task. This patch
moves them to the do_build task allowing bitbake a little more flexibility
about when things need to get built by.
I noticed this when qemu-native failed to build and a rootfs was not generated
which is not the intended behaviour.
Also update the syntax to use appendVarFlag instead of get/setVarFlag
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The autogen-native built error on FC17:
ccache: failed to create /dev/null/.ccache
This is because the default gcc command of FC17 is a symlink to ccache,
so the ccache will always be used regardless to the setting of CCACHE,
ccache uses $HOME/.ccache as the CACHE_DIR by default, but autogen set
HOME=/dev/null, so the error happens.
Disable ccache explicitly if it is not enabled would fix the problem,
otherwise it would always use ccache regardless to the setting of CCACHE
on Fedora 17.
The ccache 3.1.7 has a bug, it would always create $CCCHE_DIR/.ccache
even CCACHE_DISABLE=1.
Unset CCACHE_DISABLE in ccache.bbclass, since ccache only checks whether
there is a CCACHE_DISABLE in the environment or not, it doesn't care about
its value, so we need unset it explicitly when enable ccache.
[YOCTO #2554]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
|
|
This function can miss packages whose license is in
"COPYLEFT_LICENSE_EXCLUDE" and tarball packages with license in
"COPYLEFT_LICENSE_INCLUDE".
[YOCTO #2473]
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
|
|
Previously this was installing schemas in the sysroot, which is wrong for native
packages as nothing should touch the sysroot directly, and even more wrong for
non-native packages as the sysroot is irrelevant.
So, export the environment variable that stops the registration happening at
install time. The postinst script will handle the non-native case, and for the
sysroot I've opened #2648. This isn't a massive problem as nothing to my
knowledge actually installs schemas to the sysroot.
[YOCTO #2245]
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This was needed since it was after the sysroots and thus in some cases
the native sysroot chown was being found and used instead of the script
provided by native-intercept/chown. This was noticed by the non-gplv3
build since it's coreutils depends on coreutils-native.
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Updating the kern-tools SRCREV to pickup the following functionality:
- buildall: provides the ability to build all kernel branches
without a build system, only a cross compiler and configme
are required.
- robustness/cleanups: obselete/unused code removal and general robustness
fixes from Paul Gortmaker and Bruce Ashfield
The following kern-tools commits are part of this series:
b8dfd3d buildall: add whitelist/blacklist support
0ef039c configme: catch errors found during fragment sanitization
5b6498c buildall: remove all instances of it using/reading scc files
2e57550 buildall: support semi seamless restarts
4b5dd4d kconf_check: simplify cmdline args, dont store data per branch
58fbb6e configme: relieve it of all knowledge of scc files
a03e291 configme: strip out alternative meta series logic.
96d2bcf kgit-init: check for valid branchpoint
5598db6 buildall: allow a max cap on the number of builds done
b46abec buildall: add support for randomizing build order
68a04e9 buildall: dont copy failed build logs into main build dir
5575d85 buildall: script to independently build all board kernels
86d6200 configme: delete unused variable
8d4e29d configme: delete unused KPROFILE setting
7e15436 configme: ensure we have a valid machine type set
152b9cb scc: remove depreciated/unused commands
bb4e96a scc: allow includes within conditional statements
7da7951 configme: derive path to tools from $0
152dc45 configme: test for BUILD_DIR != ""
129f7b0 kgit-scc: add warnings about bad input args.
e977662 kgit-scc: add text for no arg and invalid arg case.
[YOCTO #843]
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Allow empty passwords login so that the default root user can login in
through openssh.
Signed-off-by: Lianhao Lu <lianhao.lu@intel.com>
|
|
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>
|
