| Age | Commit message (Collapse) | Author |
|---|
|
It's possible for tasks to stage symlinks that point to non-existent
files; an example is ncurses-native.do_populate_sysroot. There wasn't
any error checking here so this broke the build when "task" was included
in BUILDHISTORY_FEATURES. In any case we shouldn't be following symlinks
and getting the sha256sum of the link target - we need concern ourselves
only with the target path, so check if the file is a link and sha256 the
target path instead if it is. If it's neither a regular file nor a
symlink (perhaps a pipe or a device), just skip it.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
|
|
Save a file per task listing sha256sums for each file staged, i.e.
the output of the task. Some caveats:
1) This only covers sstate tasks since it uses SSTATEPOSTUNPACKFUNCS,
however those are generally the most interesting in terms of output
anyway.
2) The signature is taken before applying any relocations, so any
relocated files will actually have different signatures, but that's
churn that you probably won't want to see here.
3) At the moment if you run the same build twice without sstate you will
very likely see changes in the output for certain tasks due to things
like timestamps being present in the binary output. Fixing that is
a general Linux ecosystem problem - see this page for our efforts to
resolve it on our side:
https://wiki.yoctoproject.org/wiki/Reproducible_Builds
NOTE: you need to set your BUILDHISTORY_FEATURES value to include
"task" to enable collection of these signatures as it is is disabled by
default.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If we want to determine what changed since the last build, one angle
from which to look at it is to check the signatures. However, if we
don't actually have the signatures from the last build we don't have
anywhere to start. Save the signatures on each build in order to give us
the starting point.
NOTE: you need to set your BUILDHISTORY_FEATURES value to include
"task" to enable collection of these signatures as it is is disabled by
default.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Enabling or disabling buildhistory caused a rebuild of images, which
is undesirable. For example, it prevented image reuse from a main
build with buildhistory in a following oe-selftest where buildhistory
must be disabled.
The reason are the additional ROOTFS_POSTUNINSTALL_COMMAND and
ROOTFS_POSTUNINSTALL_COMMAND entries. Those need to be excluded both
via vardepvalueexclude and vardepsexclude.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
It is useful to know which layer provided a given recipe and its
binary packages.
Many projects combine a number of layers and some of them
also provide same recipe names in which case bitbake
can prioritize between them. buildhistory can record
the decision by saving the layer from where the recipe
was taken from.
Also, if a project is split to sub projects which maintain
recipes in different meta layers, then meta layer specific
summaries of e.g. disk usage can be calculated if
source recipes meta layer name is recorded for example in
buildhistory.
If source layer is not in build history, then layer providing
the recipe can be exported from build environment using
'bitbake-layers show-recipes', but it takes a long time to execute
since all recipes are parsed again and requires full source tree
with correct build configuration.
This patch exports the name of layer as configured in BBFILE_COLLECTIONS
append of its layer.conf. 'bitbake-layers show-recipes' exports the
meta layers directory path name. For several open source layers
these are different, e.g. meta-openembedded/meta-perl/conf/layer.conf
is perl-layer, poky/meta/conf/layer.conf is core,
poky/meta-skeleton/conf/layer.conf is skeleton etc.
Signed-off-by: Mikko Rapeli <mikko.rapeli@bmw.de>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If sdk ins't in BUILDHISTORY_FEATURES the get_extra_sdkinfo fails
because no information about sdk is generated in buildhistory repo.
Signed-off-by: Aníbal Limón <anibal.limon@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
All other fields are separated by tabs except KiB and binary package
name. This helps users, *cough managers*, who import this file into
MS Excel to calculate file system usage summaries.
Signed-off-by: Mikko Rapeli <mikko.rapeli@bmw.de>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Now that the datastore works dynamically we don't need the update_data calls
so we can just remove them. They're not actually done anything at all for
a while.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
getVar() now defaults to expanding by default, thus remove the True
option from getVar() calls with a regex search and replace.
Search made with the following regex: getVar ?\(( ?[^,()]*), True\)
Signed-off-by: Joshua Lock <joshua.g.lock@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
buildhistory was writing srcrevs.values() as SRCREV when only one
srcrev/branch exists. This returns a view of the dictionary values in python
3, and used to return a list in python 2, neither of which is an appropriate
value for SRCREV. It was resulting in latest_srcrev files like this:
# SRCREV = "346584bf6e38232be8773c24fd7dedcbd7b3d9ed"
SRCREV = "dict_values(['346584bf6e38232be8773c24fd7dedcbd7b3d9ed'])"
Which in turn would result in invalid output in buildhistory-collect-srcrevs.
Fix by calling `next(iter())` on the `.values()`
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If attempting to patch a git repo without a proper git config setup,
an error will occur saying user.name/user.email are needed by git
am/apply. After some code was removed from kernel-yocto, it was
simple enough to reproduce this error by creating a kernel patch and
using a container to build.
This patch abstracts out functionality that existed in buildhistory
for use in other classes. It also adds a call to this functionality
to the kernel-yocto class.
Fixes [YOCTO #10346]
introduced in OE-core revision
0f698dfd1c8bbc0d53ae7977e26685a7a3df52a3
Signed-off-by: Stephano Cetola <stephano.cetola@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This code was outputting variables by iterating a dictionary. In Py2 this
always results in the same iteration order but with Py3 the order changes every
execution, which resulted in buildhistory having to store diffs where fields
were simply re-ordered.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
I got fed up with seeing items dance around in sstate-package-sizes.txt
in the buildhistory git repo simply because they have the same size.
Let's sort the list first by size and then also by name to ensure items
with the same size are deterministically sorted.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add SDK_INCLUDE_PKGDATA and SDK_INCLUDE_TOOLCHAIN to the variables that
we put into sdk-info.txt
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This patch contains all the other misc pieces of the transition to
python3 which didn't make sense to be broken into individual patches.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Iterators now return views, not lists in python3. Where we need
lists, handle this explicitly.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
python3 standardises its use of iteration operations. Update
the code to match the for python3 requires.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In python3, strings are unicode by default. We need to encode/decode
from command pipelines and other places where we interface with the
real world using the correct locales. This patch updates various
call sites to use the correct encoding/decodings.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This class adds functions to the SDK creation hooks, so ensure that they're
ignored in task stamps.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Fix a minor grammatical error in the comments here.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
For images we produce a number of filtered dependency .dot files for
readability, the first of which is depends-nokernel.dot which filters
out just the kernel itself (not kernel modules). Unfortunately the
filter specifications hadn't been updated for the dash-to-underscore
removal or the 4.x kernel upgrade, thus the filtering wasn't actually
doing anything.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Invoking oe-pkgdata-util in turn for every package in the list was slow
with a large image. Modify oe-pkgdata-util's read-value command to take
an option to read the list of packages from a file, as well as prefix
the value with the package name; we can then use this to get all of the
package sizes at once. This reduces the time to gather this information
from minutes to just a second or two.
Fixes [YOCTO #7339].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
buildhistory_get_imageinfo() assumed that the buildhistory directory
for the image had already been created earlier. That assumption is not
true for special images (like the virtual swupd images from
meta-swupd) where the entire traditional do_rootfs/do_image is
skipped.
Creating files-in-image.txt still makes sense for such images, so
support them by creating the directory also in
buildhistory_get_imageinfo().
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
We've recently added SDK_EXT_TYPE and SDK_RECRDEP_TASKS and these will
have an effect on the size and functionality of the SDK, so record them
in sdk-info.txt.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
buildhistory.bbclass by design is incremental: each build adds or
updates information. Information is never removed.
Sometimes it can be useful to reduce the information only to those
recipes that were build during a specific bitbake invocation, for
example when the invocation does a full world build.
This is now possible by setting BUILDHISTORY_RESET as explained in the
modified class. The comment on the variable also mentions the caveats
associated with using this mode.
In this mode, buildhistory.bbclass first moves all existing
information into a temporary directory called "old" inside the build
history directory. There the information is used for the "version
going backwards" QA check. Then when the build is complete and before
(potentially) committing to git, the temporary directory gets deleted.
Because information that has not changed during the build will be
reconstructed during full world rebuilds, a git log will then only
show real updates, additions and removals.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Variable expressions are no longer expanded in python functions as of
BitBake commit 8bf33a8e92c0e188fa392030025756196c96fcbb, so we've now
got to do this explicitly here.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Its valid to have dependencies like XXX (=2.1) without spaces, as injected
by debian.bbclass. The code was breaking these into separate components
and destroying them so improve the regex to handle them.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Previously, in order to determine the existence of an already
initialized Git repository we checked if a directory named '.git' was
present in the buildhistory dir. However, e.g. in the case of git
submodules '.git' may also be a regular file referencing some other
location which was causing unwanted behavior. This patch changes
buildhistory.bbclass to check for any file named '.git' which fixes
these problems.
[YOCTO #8911]
Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This patch changes the use list_pkgs() instead of list()
from class RpmPkgsList. The change is in two functions,
image_list_installed_packages from rootfs.py and
sdk_list_installed_packages from sdk.py.
With this change the functions calling the functions
listed above, must format the output as they required.
The formatting can be done using format_pkg_list() from
oe.utils.
The classes calling the afected functions are changed too
with this patch, to keep the same functionality using the
new data structure.
[YOCTO #7427]
Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
These form part of the configuration for the extensible SDK, we should
really be recording what goes into them.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
I was using bytes here for the sizes and writing out KiB as the suffix.
Change it to actually write out size values in KiB.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
* Ensure extensible SDK and standard SDK go into their own directories
* Record extra extensible SDK variables
* Write sstate-package-sizes.txt and sstate-task-sizes.txt files so you
can analyse the size of the contents
* Add BUILDHISTORY_SDK_FILES (similar to BUILDHISTORY_IMAGE_FILES) and
default it to pick up config files installed in the extensible SDK
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
After OE-Core revision baa4e43a29e45df17eaa3456acc179b08d571db6 we lost
recording SDK the contents in buildhistory. This was due to the
SDK_POSTPROCESS_COMMAND variable being set with = in
populate_sdk_base.bbclass which overwrote any value set with += in
buildhistory.bbclass; to fix it, use _append in buildhistory.bbclass
instead.
Fixes [YOCTO #8839].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The idea behind the implementation of Yocto #8138 was that an
additional class can write additional files in the recipe directories,
for example by hooking into the functions of buildhistory.bbclass or
by implementing its own SSTATEPOSTINSTFUNCS function.
However, when these additional files get created before
buildhistory_emit_pkghistory(), they get removed again by that
function because it contains code which removes everything it does
not know about. The reason for that is that these unknown items
are probably obsolete.
This logic is the reason why the additional "kconfig" file from
buildhistory-extra.bbclass never showed up in the final build history.
To fix this, the hard-coded list of known files in
buildhistory_emit_pkghistory() must be turned into a variable which
derived classes can extend.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Fix some spelling mistakes in bbclass files
Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
We don't actually use values we read in here that are likely to
contain = characters but we might as well split the value properly in
case we do in future.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
We have the command in the commit message, we might as well have the
build result as well (succeeded/failed and whether or not it was
interrupted by the user). The interrupted part relies upon a change to
BitBake to extend the BuildCompleted event to include an attribute for
that, but will not fail if the attribute is not present.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
We do already commit these into the repository itself, but have them in
the commit message as well as a reference. As part of this, refactor out
running "git commit" into a separate function so we don't have to
duplicate the code in the two places we call it.
Implements [YOCTO #7966].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Avoid an error when attempting to remove previous data if it's not a
subdirectory - we were assuming that anything that wasn't named "latest"
or "latest_srcrev" had to be a directory. This makes it possible to have
a buildhistory_emit_pkghistory_append which writes additional files at
the recipe level.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If we want the correct file permissions to show up here as they would on
the target, we need to run the command under pseudo. Normally we'd set
the fakeroot varflag on the function and that would be enough, but it
turns out that setting fakeroot on a non-task function that you execute
using bb.build.exec_func() isn't working at the moment. Work around this
by simply using FAKEROOTENV and FAKEROOTCMD. Unfortunately that means we
have to duplicate the command for the two cases but I couldn't find a
better means of doing that that actually works.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
* Avoid using ${...} for shell variables (since they could be
expanded as bitbake variables if present)
* Use files-in-package.txt rather than files-in-<packagename>.txt; the
file is already in a subdirectory named with the package name and this
naming is consistent with that of files-in-image.txt.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Shell functions use tabs in this file.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
We don't care about the permissions on the top-level directory in which
the files are contained, just everything under it; this also avoids
lists with just this entry in it for empty packages. Affects file
listings for both images and packages.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If there aren't any changes, we still make a commit to the buildhistory
repo, but this wasn't being pushed if BUILDHISTORY_PUSH_REPO is set.
Move the push to the end to make it unconditional.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
files-in-$pkg.txt
This is just a safe check to make sure the output folder is present, before
creating the files-in-$pkg.txt file.
Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
A file similar (in format) to files-in-image.txt is created for each
recipe's build package. This would allow to easily see what the
recipe's package may install when including it into an image.
These files reside on the specific buildhistory recipe package (i.e.
buildhistory/packages/i586-poky-linux/busybox/busybox-dev/files-in-busybox-dev.txt
for busybox-dev package)
[YOCTO #5870]
Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The git user.email and user.name settings must be set or the commit to
the buildhistory repo (when BUILDHISTORY_COMMIT is set to "1") will
fail. If the user hasn't set these, set some dummy values since the
actual value is unlikely to be particularly useful and it's far worse
for the commit not to be done leaving the repository in a partially set
up state.
Prompted by [YOCTO #6973].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
* Use argparse instead of optparse for standardised help output, options
and a much cleaner code structure
* Look up pkgdata directory automatically so the user doesn't have to
specify it
* Use standard logging
NOTE: this does mean a slight change in syntax - if you do want to
specify the pkgdata directory (usually only necessary if you're calling
it from within the build process) you need to use the parameter -p (or
--pkgdata-dir) and specify this before the command, not after it.
Examples:
oe-pkgdata-util find-path /sbin/mke2fs
oe-pkgdata-util lookup-recipe libelf1
oe-pkgdata-util read-value PKGSIZE libc6
oe-pkgdata-util -p /home/user/oe/build/tmp/sysroots/qemux86-64/pkgdata read-value PKGSIZE libc6
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
|
|
Since the rewrite of the image construction code in python a few
releases ago, we remove a couple of packages from the image as one of
the final steps when constructing the image (notably update-rc.d and
run-postinsts). However, because of the order of operations, these
packages are still listed both in the buildhistory
installed_package*.txt files and in the manifest file created next to
the image, which is wrong.
There were two possible solutions to this: (1) change the order such
that the uninstallation occurs before calling ROOTFS_POSTPROCESS_COMMAND
or (2) add another hook variable in such that we can have the
package list collection code run at the right time. Because it's
currently possible (but very much not recommended) to install additional
packages within ROOTFS_POSTPROCESS_COMMAND, which may have postinstall
scripts and thus require the packages we would otherwise uninstall if we
were to take option 1, option 2 is really the least likely to cause
problems. Therefore, add ROOTFS_POSTUNINSTALL_COMMAND and make the image
and buildhistory classes use it.
Fixes [YOCTO #6479].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
|
|
Writes build information to target filesystem on /etc/build such as enabled
layers, their current status and commit.
squashspaces was moved to oe/utils.py to make it available to different classes
and avoid code duplication.
[YOCTO #6770]
Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
Paul Eggleton
Patrick Ohly
Mikko Rapeli
Aníbal Limón
Richard Purdie
Joshua Lock
Christopher Larson
Stephano Cetola
Ross Burton
Markus Lehtonen
Mariano Lopez
Maxin B. John
Leonardo Sandoval
Alejandro Hernandez