| Age | Commit message (Collapse) | Author | Files |
|---|
|
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
When files overlap in the sysroot, something bad usually happened. We've had
two independent cases recently where a couple of months after one of these
warnings was shown, builds failed due to corruption.
This change moves the warning to become a fatal error. The complaint I've had
about this is that we need to tell the user what happened and more importantly
how to recover from it. If we could recover from it, great but the trouble is
we simply don't know what happened.
As a compromise, we can document several of the possible scenarios in the error
message. We don't normally go to this level of detail however in this case, I'm
lacking other viable alternatives.
I do believe it is important to stop as corruption occurs rather than letting the
build contunue into territory that is not deterministic amongst other things.
The complex message is followed by a simpler one in case the long message is too
much for the user.
(From OE-Core rev: 179ac7de03977b6e440409eddb2166819e07286a)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Taking run-postinsts and building for two machines which have different
OVERRIDES leads to two different sets of stamps for an allarch package.
We don't need to depend on OVERRIDES in these classes, the end resulting
variables are good enough. We can therefore exclude the dependency
and allow a single package to be generated for run-postinsts.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
These patches are the start of a new idea, a way of allowing a single set of
cross/native sstate to work over mutliple distros, even old ones.
The assumption is that our own C library is basically up to date. We build
and share a small tarball (~2MB) of a prebuilt copy of this along with a
patchelf binary (which sadly is C++ based so libstdc++ is in there). This
tarball can be generated from our usual SDK generation process through
the supplied recipe, uninative-tarball.
At the start of the build, if its not been extracted into the sysroot, this
tarball is extracted there and configured for the specified path.
When we install binaries from a "uninative" sstate feed, we change the
dynamic loader to point at this dynamic loader and C librbary. This works
exactly the same way as our relocatable SDK does. The only real difference
is a switch to use patchelf, so even if the interpreter section is too small,
it can still adjust the binary.
Right now this implements a working proof of concept. If you build the tarball
and place it at the head of the tree (in COREBASE), you can run a build from
sstate and successfully build packages and construct images.
There is some improvement needed, its hardcoded for x86_64 right now, its trivial
to add 32 bit support too. The tarball isn't fetched right now, there is just a
harcoded path assumption and there is no error handling. I haven't figured
out the best delivery mechanism for that yet. BuildStarted is probably not
the right event to hook on either.
I've merged this to illustrate how with a small change, we might make the
native/cross sstate much more reusable and hence improve the accessibility
of lower overhead builds. With this change, its possible the Yocto Project may
be able to support a configured sstate mirror out the box. This also has
positive implications for our developer workflow/SDK improvements.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>
|
|
Currently we have the horrible situation where meta-environment packages
the toolchain environment files and they get included in the SDK but are
broken, then, the SDK code overwrites them with good versions. This is
suboptimal.
This change fixes the code in meta-environment to create working
files and adds in the multilib support from populate_sdk_base, then
we remove the code in that base bbclass and rely on the packages
being installed if/as/when needed. This removes the duplication
and the broken versions of the files, hopefully making all well.
[YOCTO #6608]
[YOCTO #6613]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
UHD and GNU radio use the cmake build system. The toolchain file made
from cmake.bbclass does not set the variable needs by cmake projects
that use .S files. UHD added some .S files and these changes are required
to build recent UHD.
Signed-off-by: Philip Balister <philip@balister.org>
|
|
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
|
|
I've been giving things some thought, specifically why sstate doesn't
get used more and why we have people requesting external toolchains. I'm
guessing the issue is that people don't like how often sstate can change
and the lack of an easy way to lock it down.
Locking it down is actually quite easy so patch implements some basics
of how you can do this (for example to a specific toolchain). With an
addition like this to local.conf (or wherever):
SIGGEN_LOCKEDSIGS = "\
gcc-cross:do_populate_sysroot:a8d91b35b98e1494957a2ddaf4598956 \
eglibc:do_populate_sysroot:13e8c68553dc61f9d67564f13b9b2d67 \
eglibc:do_packagedata:bfca0db1782c719d373f8636282596ee \
gcc-cross:do_packagedata:4b601ff4f67601395ee49c46701122f6 \
"
the code at the end of the email will force the hashes to those values
for the recipes mentioned. The system would then find and use those
specific objects from the sstate cache instead of trying to build
anything.
Obviously this is a little simplistic, you might need to put an override
against this to only apply those revisions for a specific architecture
for example. You'd also probably want to put code in the sstate hash
validation code to ensure it really did install these from sstate since
if it didn't you'd want to abort the build.
This patch also implements support to add to bitbake -S which dumps the
locked sstate checksums for each task into a ready prepared include file
locked-sigs.inc (currently placed into cwd). There is a function,
bb.parse.siggen.dump_lockedsigs() which can be called to trigger the
same functionality from task space.
A warning is added to sstate.bbclass through a call back into the siggen
class to warn if objects are not used from the locked cache. The
SIGGEN_ENFORCE_LOCKEDSIGS variable controls whether this is just a warning
or a fatal error.
A script is provided to generate sstate directory from a locked-sigs file.
(From OE-Core rev: 7e14784f2493a19c6bfe3ec3f05a5cf9797a2f22)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The "grep -e (x|y)" doesn't work, for example:
$ echo xy | grep -e '(x|y)'
No output
We can use "grep -E" (extended regexp) or "grep -e x -e y" to fix it.
It only affected the cross recipes.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Currently copies of the license files are made which wastes disk space
and adversely affects performance. We can link these instead in most
cases for small performance gains.
(From OE-Core rev: 0b0f3631fd22f731b6aeedb73965e367b695028b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The use of [ and && here means $? is reset and the exit 1 error
interception wasn't working, leading to "file changed as we read it"
errors from sstate_create_package when heavily using hardlinks.
Fix this by placing $? into a variable.
(From OE-Core rev: 6e51f900b76b06c09a3d6927f8db7398e2c035ed)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add an SDK_POST_INSTALL_COMMAND variable which allows additional
commands to be added towards the end of the SDK install script, for e.g.
additional processing that needs to be done as part of installing the
SDK.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
When the PACKAGE_ARCH is dependant of a override, the expanding is
done too late triggering the false-positive error of wrong inclusion
order.
To fix this we use immediate expansion operator to evaluate the
PACKAGE_ARCH value in order to choose to load 'allarch' class or not.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
|
|
A few firmware files use this filename extension, like korg/k1212.dsp .
Create respective packages.
Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
SSTATEPOSTINSTFUNCS is now set with = in sstate.bbclass, and because
the line here in buildhistory.bbclass ends up being parsed before that
we now need to use _append or the value we are adding will be wiped out.
This fixes buildhistory no longer recording package information since
OE-Core revision 9d659c6f20fa4a141b491c62a3ef0dfb1f896d9c.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
* in some cases (e.g. with external toolchain which doesn't respect our
reasonably old version set in OLDEST_KERNEL) it's possible to have libc
which requires newer kernel than what we have on builders, qemu supports
-r param to use different uname than what's returned by host system.
* change qemu_run_binary to pass -r ${OLDEST_KERNEL} and add the same to
QEMU_OPTIONS which are used by qemuwrapper-cross
* maybe we should eventually convert all qemu_run_binary usages always include
qemuwrapper-cross dependency and always call qemu through that (it
seems very strange that qemu_target_binary is called from qemuwrapper
and for allarch recipes it can return qemu-allarch as qemu binary).
qemu_run_binary is used by:
meta/classes/gtk-immodules-cache.bbclass: ${@qemu_run_binary(d, '$D', '${bindir}/gtk-query-immodules-$maj_ver.0')} \
meta/classes/qemu.bbclass:def qemu_run_binary(data, rootfs_path, binary):
meta/recipes-core/systemd/systemd_213.bb: ${@qemu_run_binary(d, '$D', '${base_bindir}/udevadm')} hwdb --update \
meta/recipes-graphics/pango/pango.inc: ${@qemu_run_binary(d, '$D','${bindir}/${MLPREFIX}pango-querymodules')} \
and qemuwrapper directly by:
scripts/postinst-intercepts/update_font_cache:PSEUDO_UNLOAD=1 qemuwrapper -L $D -E LD_LIBRARY_PATH=$D/${libdir}:$D/${base_libdir}\
scripts/postinst-intercepts/update_pixbuf_cache:PSEUDO_UNLOAD=1 qemuwrapper -L $D -E LD_LIBRARY_PATH=$D/${libdir}:$D/${base_libdir}\
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In some cases we do either need to add extra sstate manipulation
functions, or change the existing modification functions. This patch
parametrizes it to SSTATEPOSTCREATEFUNCS after sstate_create_package
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The ipk or deb can't depend on file such as "/bin/bash" or
"/usr/bin/perl", so it knows nothing about the pkg depends bash or perl,
thus there would be dependencies problems when we run "apt-get
install/remove <pkg>" on the target, this check can help us find the
issues and then fix them manually.
* Benefits:
- Help to fix file rdepends issues for ipk and deb
- Help to fix abnormal rdepends.
- Help to check whether the rdepends is OK or not after build each
recipe (don't have to install to the image), for example, a recipe may
generate 10 binary packages, only a part of them will be installed to
the image by default, we can know whether the rdepends are OK or
not for the ones which are installed, but can't know the ones which
are not installed, this patch can help check all the 10 binary
packages' rdepends.
* Basic designs:
- Get all the RDEPENDS on the chain.
- Get the pkg's FILERPROVIDES from oe.packagedata.read_subpkgdata()
and save to set filerdepends.
- Get each RPDEPENDS' FILERPROVIDES, RPROVIDES and FILERPROVIDESFLIST,
and save to set rdep_rprovides.
- Do the set "filerdepends -= rdep_rprovides" and QA issue if
filerdepends is not null.
[YOCTO #1662]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
When using the useradd-staticids.bbclass under meta/classes,
this error occurs:
"<username> - <username>: Username does not have a static uid defined."
There was a problem with the regular expression for parsing parameters,
it was sometimes returning an empty string.
I have fixed this by skipping empty strings.
Signed-off-by: Fabrice Coulon <fabrice@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
During spec generation, ideally directories should not be auto
packaged under the %file section of rpm packages but take ownership of
specific directories.
* packages only empty directories or explict directory.
See:
- http://www.rpm.org/max-rpm/s1-rpm-inside-files-list-directives.html
- "The %dir Directive"
* This will prevent the overlapping of security permission.
For example, in Tizen the directory /etc have smack label 'System::Shared'
So Only one package should own and set the label of /etc to prevent
the overwriting of the smack label.
Existing behaviour is maintained if DIRFILES is not set. If it is set,
the modified behaviour is used. If can be set to an empty value by
core recipes to trigger the modified behaviour.
[RP: Modified to allow optional usage of DIRFILES]
Signed-off-by: Ronan Le Martret <ronan@fridu.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The 'and' logic for the check is wrong. To make things clear, please
see the example below.
Say that we have a recipe A, which has a sysv-init style init script named
initA and no corresponding service file. The recipe inherits update-rc.d,
but it doesn't inherit systemd.bbclass. The DISTRO_FEATURES has 'systemd'
inside it, but it doesn't have 'sysvinit'. Now if we build an image, with
the 'and' logic in the check, the symlinks for initA would not be installed
into /etc/rc?.d directory.
This is incorrect. Because there's no corresponding service. The symlinks
in /etc/rc?.d/ should be created so that the service would be correctly
started at boot time.
The logic should really be 'or' in the check. This is actually what the code
was when it was originally written.
Several different situations are listed below to prove the correctness of the
'or' logic.
If 'sysvinit' is in DISTRO_FEATURES, the initA script would always be installed
with corresponding preisnt/postinst generated and added.
If 'sysvinit' is not in DISTRO_FEATURES, we have three situations.
1) A has initA and A.service.
In such situation, systemd.bbclass would set INHIBIT_UPDATERCD_BBCLASS,
so no preinst/postinst about update-rc.d would be added and the symlinks
for initA would not be created.
2) A has initA, and the functionality of initA is not implemented internally
in systemd.
In such situation, symlinks for initA would be installed.
3) A has initA, and the functionality of initA is implemented internally in
systemd or in some other recipe.
Examples for such situation are alsa-state and keymaps in OE.
In such situation, we need to set INHIBIT_UPDATERCD_BBCLASS in the recipe
so that there would be no preinst/postinst scripts about update-rc.d added.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
A distribution compression policy should be established, and used by all
packages. It compressed man pages in ${mandir} and info pages in ${infodir}
1. The doc will be compressed to gz format by default, which is configured
in ${DOC_COMPRESS}
2. It will automatically correct the compressed doc which is not
in ${DOC_COMPRESS} but in ${DOC_COMPRESS_LIST} to the format
of ${DOC_COMPRESS} policy
3. It is easy to add a new type compression by editing
local.conf, such as:
...
DOC_COMPRESS_LIST_append = ' abc'
DOC_COMPRESS = 'abc'
DOC_COMPRESS_CMD[abc] = 'abc compress cmd ***'
DOC_DECOMPRESS_CMD[abc] = 'abc decompress cmd ***'
...
[YOCTO #1238]
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This allows its usage in other RPM macros so files in ${S} can be found.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add hooks to allow customisation of the rpm spec files. Since python functions
aren't visible in the data store, one variable is used to trigger the call to
a separately named function. A dummy function is not provided since this then
triggers various class ordering complexities which are best avoided.
Ultimately this will be replaced by a refactor of the code to generate the
spec file using a python class.
This allows the tizen layer to add hooks for the security manifests for
example.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
The native recipe doesn't need package_qa, and it has a race issue with
do_unpack which requires the clean dir of ${S}, then causes error:
NOTE: recipe opkg-utils-native-0.1.8+gitAUTOINC+eae0d8fa44-r0: task do_package_qa: Failed
ERROR: Task 283 (virtual:native:/path/to/opkg-utils_git.bb, do_package_qa) failed with exit code '1'
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
--is-ancestor is a relatively new git option [commit 5907cda1, Aug 30 2012].
To support build machines with older versions of git installed, we can use
the basic porcelain commands to acheive the same check.
merge-base: "--is-ancestor A B" can be replaced with:
if test "$(git rev-parse --verify A)" = "$(git merge-base A B)"
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Use the bbinfo, bberror, bbfatal equivalents to the existing echo statements
within the kernel-yocto processing. This makes us consistent with the other
messages from the build system.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
When custom respositories are built (like a pure kernel.org
repo), the machine_meta SRCREV format is not applicable. As
such, we shouldn't check for the meta branch and we shouldn't
only check SRCREV_machine based revisions.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Since the git fetcher ensures that branches exist, we no longer need to
validate the branch and have a conditional checkout of the source.
We can remove some checks and ensure that whenever we exit the
do_kernel_checkout routine that a branch is always checked out.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
It's better to check a branches existence via show-ref versus the end
user branch commands. So we make the switch.
Also as part of this change, we move the conversion of remote branches
to local branches above the meta branch checking. This is required to
ensure that the branch is local for the show-ref check.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
The checking of machine and meta branch SRCREVs was inconsistent and
didn't allow a mixed AUTOREV machine/meta branch combination. By
simplifying the checks and changing the logic, we can now allow this
combination.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
KBRANCH_DEFAULT was introduced as a way to trigger the enforced build
of a particular branch of the tree. With the fetcher now enforcing
SRCREVs existing on a branch, we can simply validate that the SRCREV
is reachable from the final branch and no longer care about enforcing
a given branch.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Now that the fetcher will enforce branch existence, we no longer need to
confirm that a branch exists, and that it was the branch requested to
be built.
We know the branch exists and we'll confirm that the specified SRCREV
is going to be built after we've patched the tree.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
We no longer need to check if the KBRANCH matches the branch specified
in the SRC_URI. This is taken care of by the fetcher at the beginning
and SRCREV ancestor validation after patching.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
The bitbake fetcher now enforces that a commit is contained by a branch,
so this code can be dropped from do_validate_branches.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Rather than attempting to condition the entire tree to machine SRCREV (since
we don't know what branch will be built), we can instead wait until patching
has completed and then confirm that we are indeed building a decendant of the
specified SRCREV. The result is a much simpler check, and no mangling of the
tree.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Parsing the output of git show is error prone, since it changes based on
the type of issue with bad comit IDs. Since the output is no longer used
in the case of a valid ref, we can switch to git-cat-file and simply
check the return code.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
do_validate_branches checks to ensure that a valid machine SRCREV was
set. A test against an empty SRCREV is done in two separate locations,
we only need one, since the first check immediately returns and the
second check never hits.
At the same time, we can stop referring to the same commit hash by
3 different names. Instead we assign to a local variable at the
top of the routine, and refer to it at all times.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
A recipe will sometimes be rebuilt unnecessarily if it adds users via
useradd and the pseudo-native task has not been populated.
This patch adds the correct dependency.
Signed-off-by: Jate Sujjavanich <jatedev@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This quite coincidentally invalidates the sstate for do_package which
is needed due to the correction of oe.utils.multiprocess_exec().
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Some packages have additional configuration options that are currently
not accessible through cpan_build.bbclass. Similar to cpan.bbclass, add
EXTRA_CPAN_BUILD_FLAGS to pass the optional parameters to perl Build.PL.
Signed-off-by: Tim Orling <TicoTimo@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Currently this variable isn't passed through the magic rename
mangling. This means that if you try adding "nativesdk-eglibc"
to an image directly, you can't since the package is renamed
by debian.bbclass and nothing sees the renaming.
This is annoying since I wanted to exactly that. This code
change passes it through the standard renaming function, the
tricky part is that we have to set PKGDATA_DIR to point to
the correct sysroot during the call. We create a copy of the
datastore for the purposes of the call to do this.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Fixed:
- There might be a space between "#!" and "/pat/to/usr/bin/perl", e.g.:
"#! /usr/bin/perl", now also matches it.
- There might be the lines like the following in the body:
eval 'exec /path/to/usr/bin/perl-native/perl -S $0 ${1+"$@"}'
Now we only check "#! */path/to/usr/bin/perl" to make sure it is a
perl script, but match and fix the lines in both header and body.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In some cases we do either need to add extra sstate manipulation
functions, or change the existing modification functions. This patch
parametrises them to SSTATECREATEFUNCS and SSTATEPOSTUNPACKFUNCS and
abstracts the "hardcoded path" functions into separate functions using
these new variables.
We may use this new functionality to improve binary relocating using
patchelf for example, this at least lets us have the hooks to be able to
experiment.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In order to use hasFeature and hasPackage methods in sdk test modules, we need specific fields to be set in the TestContext object.
Adding pkgmanifest, imagefeatures and distrofeatures to the TestContext.
Signed-off-by: Corneliu Stoicescu <corneliux.stoicescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Similar to the way BSP images have rootfs a manifest, the toolchain now also has a manifest file created alongside the sdk image.
Signed-off-by: Corneliu Stoicescu <corneliux.stoicescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
When archiving original source, insert the first 10
characters of srcrev, after stripping off any
AUTOINC+, into the filename for tarfiles created
from directories.
Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
