summaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)Author
2010-12-17bitbake.conf: use bb.utils.whichBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-12-17bitbake/utils.py: Ensure the last lines of functions are printed in tracebacksRichard Purdie
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
2010-12-17bitbake/utils.py: Improve traceback to be more helpful/clear to usersRichard Purdie
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
2010-12-17Timeout UI runCommand calls if server doesn't respond (or dies)Bob Foerster
Signed-off-by: Bob Foerster <robert@erafx.com>
2010-12-17Move LAYERDIR expansion hack into DataSmartChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17parse: use bb.utils.whichChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17Update persist_data usage to the new APIChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17Revert "build: stop injecting os/bb into __builtins__"Chris Larson
Temporary revert, will resurrect. This reverts commit 566657c13de3fefb5043b71de953fb805799da95.
2010-12-17Fix logging level names for post-server-ui-splitChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17build: fix file mode for /dev/nullChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17bitbake/utils.py: Improve traceback to be more helpful/clear to usersRichard Purdie
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
2010-12-17bitbake/utils.py: Ensure the last lines of functions are printed in tracebacksRichard Purdie
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
2010-12-17build: stop injecting os/bb into __builtins__Chris Larson
For one, we should be able to avoid this via _context in bb.utils. For two, __builtins__ only has to exist for CPython, so this causes problems running bitbake under other python implementations (though we could change to doing setattr()s against the __builtin__ module, let's avoid this instead). Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17bin/bitbake: [HACK] Empty ui event queue on shutdownChris Larson
This is a workaround to ensure the user always sees any queued up log messages, even if the UI has shut down without showing them. This is far from ideal. The correct fix is to alter the UIs to shut down more cleanly, ensuring they display any remaining queued up events to the user before they exit. Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17Ensure LogHandler is set up for the server constructionChris Larson
If we add the LogHandler object in the server.main, rather than earlier, any log messages emitted during the construction of the server (and, therefore, during the construction of the cooker) will not be seen. This includes the error(s) which the cooker can emit when parsing the command line. With this, we again see the error if you try to use -g without specifying a recipe to operate against. Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17cooker: ensure parseCommandLine can get BBPKGSChris Larson
Without this, one can't use BBPKGS with bitbake -g, for example. Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-17Resurrect alternative UIsBob Foerster
The various alternative UIs have been updated to once again be functional with the latest bitbake internals. Each of the UIs still have much room for functional improvement. In particular, they have been updated to: - interact with the new process based server - handle the current set of events and notifications fired from the server and its associated subsystems Signed-off-by: Bob Foerster <robert@erafx.com>
2010-12-17utils: don't overwrite builtin dir() functionBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-12-17bzr: use utils.remove instead of os.systemBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-12-17build: use utils.remove instead of os.systemBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-12-16utils: drop unnecessary debug msgs from mkdirhierChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16Forcibly shut down the server if the UI is terminatedChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16server: ensure the cooker is stopped when shutting downChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16server: kill stdin/stdout/stderrChris Larson
This ensures that nothing run from the server can touch the console, in particular event handlers and python tasks, both of which can use bb.msg or the bitbake loggers to send output to the UI in a correct fashion instead. Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16server: fix interrupt handling for processBob Foerster
SIGINT is now blocked within the server context, thus allowing the UI to fully handle all user interaction. There is no longer a need to check for KeyboardInterrupt Exceptions anywhere within the server context. Signed-off-by: Bob Foerster <robert@erafx.com> Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16Show the user progress when loading the cacheBob Foerster
Signed-off-by: Bob Foerster <robert@erafx.com> Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16Run the server and UI in separate processesBob Foerster
This uses the python multiprocessing module, both to spawn the server process and for communication between the processes. Signed-off-by: Bob Foerster <robert@erafx.com> Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16cooker: fix remnant rq/rqdata issue from poky syncChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16Revert "build: kill stdout in python functions"Chris Larson
We have a better solution for this on the separate-ui-and-server branch, so this version is no longer necessary. This reverts commit 20929afdd87e5124891f121c3640aa9efa368c2c. Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-16runqueue: fix remnant rq/rqdata reference from poky syncChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15knotty: exit with 1 if we see a critical log messageChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15cooker: don't swallow recipe parse errorsChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15runqueue: resurrect the fix in 90c2b6cChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15build: fix log file displayChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15manual: document scmdata=keep for CVS, SVN and GitAndreas Oberritter
Signed-off-by: Andreas Oberritter <obi@opendreambox.org> Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15fetchers: Add parameter scmdata=keep to include .git/ and others in ↵Andreas Oberritter
generated tarballs. * Allows generating version information from SCMs during build. * Note that tar doesn't need to use --exclude '.git', because git checkout-index doesn't clone the repository. Signed-off-by: Andreas Oberritter <obi@opendreambox.org> Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15build: fix FuncFailed for the no logfile caseChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15process: handle OSErrors other than file not foundChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15build: write logfiles per task, not per functionChris Larson
Based on d14f9bf6 from poky, reworked for master and other cleanup. Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-15data_smart: add optional expansion to getVarFlag()Paul Eggleton
Add a parameter to getVarFlag() to auto-expand the value of the flag. This makes getVarFlag() more consistent with getVar(), and allows expansion of vardeps and vardepsexclude (which has been done in this commit). Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
2010-12-15build: fix copy/paste thinko regarding LC_ALLChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-14runqueue: fix check_stamp_task call in chck_stamp_fnChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-14runqueue: fix a get_task_id callChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-14build: ensure LogTee has a valid name propertyChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-14build: fix -D with shell functionsChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-14Fix PWD issue with new exec_func_shellChris Larson
The previous attempt was incorrect. The issue isn't that subprocess fails to set PWD, it's that PWD is in the metadata, inherited from the environment, and is re-exported, overwriting the actual accurate one in the shell environment with the old one from the metadata. So, ensure that PWD in the metadata is not exported. We can ditch this when the environment handling is reworked (e.g. poky's commit to do so). Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-13Ensure that the invalid UI error goes to stderrChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-13build: set PWD in the subprocess preexec hookChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-13Move the runq task start/failed messages to the UIChris Larson
Signed-off-by: Chris Larson <chris_larson@mentor.com>
2010-12-13bitbake/runqueue.py: Factor task skipping code into a functionRichard Purdie
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
generated by cgit 1.2.3-korg (git 2.39.0) at 2024-05-20 14:29:18 +0000