summaryrefslogtreecommitdiffstats
path: root/pym/_emerge/Scheduler.py
Commit message (Collapse)AuthorAgeFilesLines
* When portage updates itself and the new version either inheritsZac Medico2010-08-311-3/+6
| | | | | | | | | git.eclass or has 9999 in its version, trigger the 'use copy of current sources for install, then restart portage' routine. This is required since with the live git portage-9999 ebuild we can't use the version to detect incompatible changes to ebuild.sh. This issue was noticed due to recent versions of ebuild.sh expecting PORTAGE_BZIP2_COMMAND to be set, causing incompatibility with installed portage-9999 instances.
* Rename the portage.sets module to portage._sets since it will be usefulZac Medico2010-08-301-2/+2
| | | | | in the upcoming 2.1.9 branch which will not have sets support but will still have the code in private and disabled form.
* Remove unused imports found by pylint.Arfrever Frehtes Taifersar Arahesis2010-08-291-2/+0
|
* Add a TODO note about replacing BlockerDB with an incremental depgraphZac Medico2010-08-291-0/+5
| | | | for solving bug #279623.
* Add a Scheduler._running_tasks() set in order to simplify theZac Medico2010-08-281-2/+5
| | | | _is_work_scheduled() method.
* Inside Scheduler._dependent_on_scheduled_merges(), don't traverse childrenZac Medico2010-08-281-1/+5
| | | | of uninstall nodes since those aren't dependencies in the usual sense.
* Handle zlib.error when reading compressed build logs via GzipFile.Zac Medico2010-08-221-0/+4
|
* Bug #324191 - Add support for FEATURES=compress-build-logs. The causesZac Medico2010-08-211-27/+13
| | | | | | | all build logs to be compressed while they are being written. Log file names have an extension that is appropriate for the compression type. Currently, only gzip(1) compression is supported, so build logs will have a '.gz' extension when this feature is enabled.
* When accessing pkgsettings.features inside Scheduler._generate_digests(),Zac Medico2010-08-211-0/+8
| | | | | ensure that we are using global features settings rather than those from package.env.
* Move environment sanity check to the Scheduler and do it if thereZac Medico2010-08-181-0/+36
| | | | are any source packages in the merge list.
* Fix breakage in REPLACING_VERSIONS support for emerge.Zac Medico2010-08-171-0/+6
|
* Remove unused attributes from classes.Zac Medico2010-08-161-6/+6
|
* Remove PollScheduler._schedule_waitpid() since this case isn't triggeredZac Medico2010-08-151-2/+0
| | | | | | anymore, now that PollScheduler._unregister() has be fixed to discard unhandled events (so that reallocation of file descriptors no longer triggers erroneous delivery of stale events).
* Add a PollScheduler._schedule_waitpid() method for SubProcess._wait()Zac Medico2010-08-141-0/+2
| | | | | to use as an alternative to hanging up the scheduler in a blocking waitpid call.
* Allow flags from IUSE_IMPLICIT in REQUIRED_USE. Move checks to resolution time.Sebastian Luther2010-08-101-62/+1
|
* Add REQUIRED_USE check as pre-merge check.Sebastian Luther2010-08-101-0/+61
|
* In action_unmerge(), create a Schuduler instance for calls to unmerge(),Zac Medico2010-08-061-0/+1
| | | | | in order to cause redirection of ebuild phase output to logs as required for options such as --quiet.
* Make Scheduler._schedule_tasks() use a loop to ensure that _merge_wait_queueZac Medico2010-07-071-24/+26
| | | | is emptied whenever possible.
* Add a Scheduler._is_work_scheduled() method, to handle the logic involvedZac Medico2010-07-071-7/+8
| | | | in bug #326787.
* Bug #326787 - Fix intermittent failure to satisify dependencies in theZac Medico2010-07-071-1/+2
| | | | | scheduler, caused by failure to account for _merge_wait_queue inside _choose_pkg.
* Only count "merge" nodes when determining whether or not to triggerZac Medico2010-05-161-1/+1
| | | | parallel-fetch.
* Use keep_going local var.Michał Górny2010-05-031-1/+1
| | | | Use keep_going local var instead of checking self.myopts once again.
* Use tbz2.unpackinfo() to unpack the xpak when running pkg_pretend, since weZac Medico2010-04-071-7/+4
| | | | need evironment.bz2 to get unpacked here.
* In _run_pkg_pretend(), write ebuild bytes from tbz2.getfile() in binary modeZac Medico2010-04-071-1/+1
| | | | for python3 compatibility.
* Eliminate redundant code in Scheduler._run_pkg_pretend() and remove unusedZac Medico2010-04-071-44/+19
| | | | quiet_config variables.
* Handle pkg_pretend output properlySebastian Luther2010-04-071-15/+45
|
* Implement emerge part of pkg_pretendSebastian Luther2010-04-021-1/+83
|
* Remove all svn $Id keywords.Zac Medico2010-03-241-1/+0
|
* Only create implicit libc deps when the version changes.Zac Medico2010-03-091-0/+10
| | | | svn path=/main/trunk/; revision=15810
* Disable implicit libc deps for ROOT != "/" since it's probably not needed.Zac Medico2010-03-091-2/+5
| | | | svn path=/main/trunk/; revision=15809
* Add --debug output for the scheduler digraph.Zac Medico2010-03-091-0/+4
| | | | svn path=/main/trunk/; revision=15790
* If _implicit_libc_deps() finds both a new-style virtual and an old-styleZac Medico2010-03-091-8/+32
| | | | | | PROVIDE virtual, use the new-style virtual. svn path=/main/trunk/; revision=15789
* Fix typo in parenthesis from previous commit.Zac Medico2010-03-091-1/+1
| | | | svn path=/main/trunk/; revision=15784
* Bug #303567 - Create implicit dependencies on libc, in order to ensure thatZac Medico2010-03-091-0/+28
| | | | | | libc is installed as early as possible. svn path=/main/trunk/; revision=15783
* Add a note in _find_system_deps about needing the system set added to theZac Medico2010-03-011-0/+3
| | | | | | graph. svn path=/main/trunk/; revision=15510
* Move portage.digestcheck to portage.package.ebuild.digestcheck.Zac Medico2010-02-251-1/+2
| | | | svn path=/main/trunk/; revision=15458
* Fix broken digestgen import.Zac Medico2010-02-251-1/+1
| | | | svn path=/main/trunk/; revision=15457
* Move digestgen to portage.util.digestgen.Zac Medico2010-02-251-1/+2
| | | | svn path=/main/trunk/; revision=15454
* Make the first two positional arguments of digestgen() into keyword argumentsZac Medico2010-02-231-1/+1
| | | | | | | since the first one isn't really needed (when this code is in stable we can start showing a deprecation warning when the redundant argument is used). svn path=/main/trunk/; revision=15438
* Bug #295715 - Implement a register_success_hook function to complementZac Medico2009-12-061-3/+9
| | | | | | | register_die_hook, and also add code to ensure the die hooks are called even when the failure is on the python side (like for file collisions). svn path=/main/trunk/; revision=14936
* Use portage.StringIO to avoid duplicate import fallback code.Zac Medico2009-12-061-8/+1
| | | | svn path=/main/trunk/; revision=14931
* Add setup phases to the tail of the merge queue instead of the front, sinceZac Medico2009-11-121-1/+1
| | | | | | | | otherwise when merging binary packages we sometimes tend to get a large number of setup phases flooding the queue and preventing queued installs from running. svn path=/main/trunk/; revision=14816
* Bug #291200 - Add a --quiet-build option to redirect all build output to logsZac Medico2009-11-061-1/+2
| | | | | | alone, and do not display it on stdout. svn path=/main/trunk/; revision=14783
* Factor out duplicate "These are the packages that would be merged",Zac Medico2009-10-231-25/+0
| | | | | | "Calculating dependencies", and spinner cleanup code. svn path=/main/trunk/; revision=14704
* Add a --unordered-display option for use with --tree. This allows the displayZac Medico2009-10-231-3/+8
| | | | | | | to be optimized differently since the merge order is not preserved. Thanks to Sebastian Mingramm (few) for the initial patch. svn path=/main/trunk/; revision=14703
* Bug #266454 - Make @world an all-inclusive set once again, like it was priorZac Medico2009-10-161-1/+1
| | | | | | | | to portage-2.2_rc* releases. In addition to @system, @world now includes a @selected set which represents user-selected "world" packages and sets that saved in /var/lib/portage/world{,sets}. svn path=/main/trunk/; revision=14614
* Bug #288906 - Fix Scheduler._dblink_elog() to always append messages to theZac Medico2009-10-141-14/+19
| | | | | | build log if available (regardless whether or not message goes to stdout). svn path=/main/trunk/; revision=14600
* Bug #272000 - When --keep-going is enabled, don't completely bail out atZac Medico2009-10-091-6/+8
| | | | | | | the beginning due to corrupt manifest(s). Thanks to Sebastian Mingramm (few) for this patch. svn path=/main/trunk/; revision=14523
* Bug #265267 - When disabling --jobs due to interactive package(s), hintZac Medico2009-10-061-1/+4
| | | | | | about --accept-properties=-interactive option. svn path=/main/trunk/; revision=14507
* When restarting emerge, don't trust sys.argv[0] since eselect-python mayZac Medico2009-10-061-1/+3
| | | | | | modify it. svn path=/main/trunk/; revision=14499