| Age | Commit message (Collapse) | Author |
|---|
|
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
migrations
Setscene hard dependencies were not being correctly handled during task migration.
For example, do_package of recipe X might become valid due to hashequiv yet we're
still rebuilding pseudo-native, a harddep of do_package. This would mean
it would try to execute that setscene task despite pseudo not being present.
Fix this by ignoring tasks with failed harddependencies. This does mean
stlightly more rebuilds than is optimal but it avoids build crashes. Ultimately
the new runqueue model can likely better handle these cases than the older codebase
could but that is for another more invasive patch.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Bibake is currently too 'chatty' when hash equivalence is enabled. Fix
this by only printing the log output if a rehash happens and it matches
an sstate object.
Also, pass a summary option to the hash checking function. This was
already changed to a mechanism which allows addition of new parameters
so this should be backwards and forwards compatible.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If a hash is reported to the hash server, the stamp written out by the
current task didn't account for any new hash the server may have provided.
Fix this so the correct stamp is written. This means "bitbake X; bitbake X"
no longer rebuilds lots of things when hashequiv is active.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The server is now restarting when running commands which doesn't make
sense. Only restart if its configuration has changed. This should
potentially fix various memory resident bitbake usages too.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The key problem is that currently running setscene tasks are not
accounted for when processing task migrations. This means can allow
two of the same task to execute at the same time with unpredictable
effects.
This change allows us to stop doing that and refactor the code slightly
to make it clearer that these conditions don't arrive even with
deferred tasks.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
bitbake/lib/bb/fetch2/clearcase.py:148: SyntaxWarning: "is" with a literal. Did you mean "=="?
if command is 'mkview':
bitbake/lib/bb/fetch2/clearcase.py:155: SyntaxWarning: "is" with a literal. Did you mean "=="?
elif command is 'rmview':
bitbake/lib/bb/fetch2/clearcase.py:159: SyntaxWarning: "is" with a literal. Did you mean "=="?
elif command is 'setcs':
Python 3.8 is quite correct and we so mean "==" here, fix it to
avoid the warnings.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
md5_file() uses a mmap() window to improve performance when hashing files, so
refactor the code and do the same for SHA1 and SHA256.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Add a basic test for bb.utils.md5_file() etc.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
This should have been removed together with expanded_data in commit
e3694e73 (cooker/command: Drop expanded_data).
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
With PRSERV_HOST = "localhost:0", this message would occasionally pop up
during the initial cache read:
WARNING: /home/matic/ambayocto/poky/bitbake/lib/bb/cache.py:446: ResourceWarning: unclosed <socket.socket fd=10, family=AddressFamily.AF_INET, type=SocketKind.SOCK_STREAM, proto=0, laddr=('127.0.0.1', 45655)>
value = pickled.load()
The file location stated is irrelevant; it just happens to be wherever
CPython decides to run the garbage collector. The issue is that after we
fork off a PRServer, self.socket is also duplicated. The parent side of
it also needs to be closed.
Signed-off-by: Gavin Li <gavin@matician.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
Fix mercurial fetching after breakage from changes to the core fetcher.
Fix username and password usage and setting moddir needed by setup_revisions.
Signed-off-by: Volker Vogelhuber <v.vogelhuber@digitalendoscopy.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
[YOCTO #13586]
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
When bitbake determines if a git clone needs updating, it only checks for the
needed srcrevs, not the revs listed in BB_GIT_SHALLOW_REVS, which will fail if
using shallow and the needed rev was added to the upstream git repo after a
previous fetch. Ensure that we also check for shallow revs.
[YOCTO #13586]
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
|
|
There was a weird error in OE-Core where "devtool modify virtual/kernel"
was showing basehash mismatch errors. This was due to SRCPV sometimes being:
AUTOINC+b867b78b50_47b80ef7bd and sometimes AUTOINC+b867b78b50_255a750d28.
The latter hash comes from KBRANCH and meant sometimes the correct branch
was seen, sometimes it was not. The issue was complicated by the execution
using a remote datastore over tinfoil.
The problem turns out to be a fetcher caching error. If the datastore
changes, the cached url data may not be valid.
We therefore ensure we match cached url data against the datastore that
generated it, which appears to fix this issue.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The bitbake-worker child on the SIGTERM signal handling send the SIGTERM to all
processes in it's process group. In cases when the bitbake-worker child got
SIGTERM after registering own SIGTERM handler and before the os.setsid() call
it can send SIGTERM to unwanted processes.
In the worst case during SIGTERM processing the bitbake-worker child can be in
the group of the process that started BitBake itself. As a result it can kill
processes that not related to BitBake at all.
Signed-off-by: Ivan Efimov <i.efimov@inango-systems.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The syntax to use multiconfig builds changed from multiconfig:foo:target
to mc:foo:target, change the syntax on bitbakes documentation.
Clarify that BBMULTICONFIG defines additional configurations along with
the one coming from local.conf.
Signed-off-by: Alejandro Enedino Hernandez Samaniego <aehs29@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Toaster directly supports the last two stable branches of Yocto
Project. With "Zeus" being released, it is time to replace "Thud".
[YOCTO #13579]
Signed-off-by: David Reyna <David.Reyna@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This allows metadata to depend on SignatureGeneratorUniHashMixIn which was recently added.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Some of the objects that bitbake reports to Toaster as dependencies to packages
are known objects that are not packages, for example library files and kernel
modules. In the Toaster logs, mark these as "Info" instead of "Warning".
[YOCTO #13386]
Signed-off-by: David Reyna <David.Reyna@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
There were three issues in this one bug.
1) The Add Layer button allows empty layers
2) The internal XHR URL was wrong, which caused a hidden AJAX error
and did not correctly complete the action nor disable the button
after an add.
3) There was a race condition between typing in the dependent layer
select text box (which would normally disable the add button), and
the typeahead pull-down selection (which would normally enable the
add button). This forced the user to select the typedahead layer twice.
[YOCTO #13385]
Signed-off-by: David Reyna <David.Reyna@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The hashserve can delete its socket whilst the cleanup us happening leading to
backtraces and test failures.
Add code to avoid this race condition.
[YOCTO #13542]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The full paths make the cache useless in the sdk. They also bloat the
cache size. They're for human debugging benefit only so compromise and
reduce this to the filename.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
A minor performance optmisation to keep lists smaller when running large
builds. We can do this since once a task has been built, we don't need
to worry about it. This improves a major bottleneck that shows up on
performance profile charts in dryruns.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
There are some runqueue code paths where the unihash cache would not be
saved where for example only parsing or an occurred. Save the cache at the
end of runqueue generation to ensure entries are cached.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Use the unihash in the output filename of sigdata files else the contents
of stamp directories is misleading. Write the unihash into the singature to
make it clear what happened.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
We need to set the setscene tasklist before we call into the
taskhash/unihash code else the behaviour is inconsistent.
Avoid reporting hashes for non setscene tasks since we'd never
query that.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Currently runqueue will rerun setscene tasks multiple times as hashes
change. This has caused numerous problems since a setscene task may
become "unavailable" for some future signature combination and the code
then can't easily "unskip" tasks its already passed into the execution
queue.
At least for now, only run setscene once and assume they're equivalent
at that point. In practise that has been much more stable in testing.
Tweak the test to match the change in behaviour.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The setscenetasks list needs to be available in the worker contexts
else the signature behaviour there mismatches what the server does.
Add the data to get/set_taskdata to ensure this happens.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Tasks were not migrating consistently, particularly:
* if a task was rehashed which had already run
* if a task which was valid became invalid due to a rehash
We need to always run the migration code for rehashed tasks and then
reprocess them for hash validity. This means rearranging the code.
It also means several tests are no longer correct and can't be written
correctly to work on all possible workflows so those are removed.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The hash server process is terminated and waited on with join(), so it
should not be a daemon. Daemonizing it cause races with the server
cleanup, especially in the selftest because the process may not have
terminated and cleanup up its socket before the test cleanup runs and
tries to do it.
[YOCTO #13542]
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The HashConnectionError class was moved to the client module and needs
to be updated.
[YOCTO #13537]
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The os module is required to connect to a unix domain socket
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The variable values that result from parsing multiconfig should be
included in the cooker data hash, otherwise changes to these files won't
be detected, which will allow the parsing cache to be loaded with the
old values for the multiconfigs. This can either manifest as the
variable values simply not updating, or getting basehash changed errors
when building.
This bug was previously undetected because all of the multiconfig base
files were a direct file dependency in all parsed recipes. This was
fixed in 34137a00f60 ("bitbake: bitbake: cooker: Rename __depends in all
multiconfigs"), exposing this bug.
[YOCTO #13541]
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Signed-off-by: Jacob Kroon <jacob.kroon@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add a test case to exercise the detection of git-lfs repositories and the
behaviour of the lfs parameter.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Refactor the git-lfs checking: this means both clearer code in the download()
function and allows unit testing to monkeypatch the functionality.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Autobuilder type infrastructure can benefit from deletion of certain files as
background IO due to the way Linux filesystem priority works.
We have problems where build directories as part of oe-selftest being
delete starves the running tasks of IO to the point builds take much
longer to compelte.
Having this option of running the deletion at "idle" helps a lot with
that.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
According to sstate_checkhashes which is defined in sstate.bbclass, the
currentcoun should be a number (0, not None).
Fixed:
$ bitbake base-files -Sprintdiff
> bb.plain("Sstate summary: Wanted %d Found %d Missed %d Current %d (%d%% match, %d%% complete)" % (total, len(found), len(missed), currentcount, match, complete))
TypeError: %d format: a number is required, not NoneType
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Add a new 'lfs' option to the git fetcher so that we can optionally not fetch
git-lfs content, for repositories that contain LFS data that we don't
actually need for building.
By default lfs is set to 1, so if the repository has LFS content then git-lfs is
required. Setting lfs to 0 will mean that git-lfs won't be required to fetch,
and some files will be missing.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Currently, show-recipes will append "(skipped)" marker to recipes which
were skipped due these recipes does not satisfied the configurations.
Example: $ bitbake-layers show-recipes -r
ace
backport-iwlwifi
core-image-rt (skipped)
core-image-rt-sdk (skipped)
core-image-tiny
Add -b/--bare to enable output names without "(skipped)" marker.
Example: $ bitbake-layers show-recipes -r -b
ace
backport-iwlwifi
core-image-rt
core-image-rt-sdk
core-image-tiny
Signed-off-by: Yeoh Ee Peng <ee.peng.yeoh@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Currently, show-recipes will show recipes from all configured layers.
Assume, meta-intel layer was added to conf/bblayers.conf.
Example of default $ bitbake-layers show-recipes:
core-image-rt:
meta-intel unknown (skipped)
meta unknown (skipped)
Add -l/--layer to enable showing recipes from user selected layer.
Example: $ bitbake-layers show-recipes -l meta-intel
core-image-rt:
meta-intel unknown (skipped)
Signed-off-by: Yeoh Ee Peng <ee.peng.yeoh@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Currently, show-recipes will show all recipes available (both
recipes with different version and recipes provided by more
than one layer).
Example of default $ bitbake-layers show-recipes:
core-image-rt:
meta-intel unknown (skipped)
meta unknown (skipped)
yajl:
meta-oe 2.1.0
meta-oe 1.0.12
Add -r/--recipes-only to enable showing recipes only. This
provide a focus view on unique recipes available.
Example of $ bitbake-layers show-recipes -r:
core-image-rt (skipped)
yajl
Signed-off-by: Yeoh Ee Peng <ee.peng.yeoh@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Not all servers and repositories have this property set
which results in failures like this when actual svn checkout
command succeeded:
svn: warning: W200017: Property 'svn:externals' not found on ''
svn: E200000: A problem occurred; see other errors for details
Signed-off-by: Mikko Rapeli <mikko.rapeli@bmw.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Reworks the hash equivalence server to address performance issues that
were encountered with the REST mechanism used previously, particularly
during the heavy request load encountered during signature generation.
Notable changes are:
1) The server protocol is no longer HTTP based. Instead, it uses a
simpler JSON over a streaming protocol link. This protocol has much
lower overhead than HTTP since it eliminates the HTTP headers.
2) The hash equivalence server can either bind to a TCP port, or a Unix
domain socket. Unix domain sockets are more efficient for local
communication, and so are preferred if the user enables hash
equivalence only for the local build. The arguments to the
'bitbake-hashserve' command have been updated accordingly.
3) The value to which BB_HASHSERVE should be set to enable a local hash
equivalence server is changed to "auto" instead of "localhost:0". The
latter didn't make sense when the local server was using a Unix
domain socket.
4) Clients are expected to keep a persistent connection to the server
instead of creating a new connection each time a request is made for
optimal performance.
5) Most of the client logic has been moved to the hashserve module in
bitbake. This makes it easier to share the client code.
6) A new bitbake command has been added called 'bitbake-hashclient'.
This command can be used to query a hash equivalence server, including
fetching the statistics and running a performance stress test.
7) The table indexes in the SQLite database have been updated to
optimize hash lookups. This change is backward compatible, as the
database will delete the old indexes first if they exist.
8) The server has been reworked to use python async to maximize
performance with persistently connected clients. This requires Python
3.5 or later.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
We only have hash equivalence for setscene tasks so only query the server
for those, reducing the number of connections needed.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This reverts commit 94c0c7f15c7a6244a8576ed948ffc21afb96ba82.
This ignores the layer priority, making the issue much worse.
E.g. I'm seeing a lot of failures caused by missing users, because
base-passwd bbappends applied in unexpected order caused different
passwd.master to be found in re-ordered FILESPATH.
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Nowadays bitbake applies overrides dynamically, not at a single
specific point in time during parsing.
Signed-off-by: Jacob Kroon <jacob.kroon@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The effects of _append/_prepend/_remove are applied when a variable
is expanded, not after parsing has completed.
Signed-off-by: Jacob Kroon <jacob.kroon@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
Richard Purdie
Ross Burton
Peter Kjellerstedt
Gavin Li
Volker Vogelhuber
Christopher Larson
Ivan Efimov
Alejandro Enedino Hernandez Samaniego
David Reyna
Joshua Watt
Jacob Kroon
Robert Yang
Yeoh Ee Peng
Mikko Rapeli
Martin Jansa