summaryrefslogtreecommitdiff
path: root/apt-pkg/edsp.cc
Commit message (Collapse)AuthorAgeFilesLines
* cleanup headers and especially #includes everywhereDavid Kalnischkies2014-03-131-4/+13
| | | | | | | | Beside being a bit cleaner it hopefully also resolves oddball problems I have with high levels of parallel jobs. Git-Dch: Ignore Reported-By: iwyu (include-what-you-use)
* warning: unused parameter ‘foo’ [-Wunused-parameter]David Kalnischkies2014-03-131-6/+4
| | | | | Reported-By: gcc -Wunused-parameter Git-Dch: Ignore
* apt-pkg/edsp.cc: do not use stderr in WriteSolution at allMichael Vogt2013-03-201-3/+0
|
* apt-pkg/edsp.cc: do not spam stderr in WriteSolutionMichael Vogt2013-03-201-0/+2
|
* * apt-pkg/edsp.cc:David Kalnischkies2012-10-131-2/+4
| | | | - include reinstall requests and already installed (= protected) packages in the install-request for external resolvers (Closes: #689331)
* * apt-pkg/cacheiterators.h:David Kalnischkies2012-06-111-8/+4
| | | - add an IsMultiArchImplicit() method for Dep- and PrvIterator
* merged lp:~donkult/apt/experimentalMichael Vogt2011-10-141-1/+8
|\
| * use forward declaration in headers if possible instead of includesDavid Kalnischkies2011-09-191-0/+7
|/
* reorder includes: add <config.h> if needed and include it at firstDavid Kalnischkies2011-09-131-2/+4
|
* * apt-pkg/pkgcache.h:David Kalnischkies2011-05-171-3/+3
| | | | - clean up mess with the "all" handling in MultiArch to fix LP: #733741 cleanly for everyone now
* rename option APT::Solver::Name to simply APT::SolverDavid Kalnischkies2011-05-171-1/+1
|
* send the first line of the error message to the error list and fail a bitDavid Kalnischkies2011-05-101-3/+11
| | | | more nicely and in order
* fix SubProgress to accept a Percent parameter to update the CurrentDavid Kalnischkies2011-05-101-5/+3
| | | | with the text as otherwise the update will be ignored
* implement proper progress report with OpProgressDavid Kalnischkies2011-05-091-19/+45
|
* implement correct error reportingDavid Kalnischkies2011-05-071-2/+12
|
* undo the temporary progress reporting disabling which slipped into last commitDavid Kalnischkies2011-05-061-4/+4
|
* fix arguments for MarkInstall so packages are really marked as automaticDavid Kalnischkies2011-05-031-5/+5
|
* maybe Pre-Depends are checked if they write them as Pre-Depends andDavid Kalnischkies2011-05-031-1/+1
| | | | not as PreDepends (doh!) …
* tell the resolver a package is set on hold if it was set by the userDavid Kalnischkies2011-05-031-1/+2
| | | | | to Keep which happens for example if a user decides to "remove" a not installed package to forbid that it's part of the solution
* implement external solver calling for upgrade and dist-upgrade, tooDavid Kalnischkies2011-05-031-34/+54
|
* doesn't execute autoremove marker setting if an external solver is calledDavid Kalnischkies2011-05-021-2/+18
| | | | | | | and instead rely on the Autoremove tagging to show us what could be done. (apt-internal-solver doesn't support this currently as it doesn't load the auto-information into the cache)
* refactor: move solver execution into his own EDSP methodDavid Kalnischkies2011-05-021-0/+42
|
* reduce the buffer size so we get a sort of realtime progress reportDavid Kalnischkies2011-05-021-1/+2
| | | | | and print the time of output at the front of the progress report so we can see the delay
* move the mapping generation to the top as the response reading isDavid Kalnischkies2011-05-021-5/+5
| | | | | currently waiting for the solver to complete and not non-blocking so we can generate the map while waiting for the solver
* implement optional Progress report in EDSPDavid Kalnischkies2011-05-021-2/+19
|
* reorganize WriteScenario to add a WriteLimitedScenario in which a scenarioDavid Kalnischkies2011-04-251-80/+175
| | | | can be limited to a subset of packages with only relevant dependencies
* use the version id instead of the mmap offset as APT-IDDavid Kalnischkies2011-04-191-4/+18
| | | | | | | This leads to a small performance decrease as we need to build this mapping now while interpreting the Response but a (buggy) solver can't point us to dangerous memory locations anymore this way and VersionCount remains useful for other mapping proposes
* Interpret Remove and Install lines in Responses correctlyDavid Kalnischkies2011-04-191-5/+12
|
* send the scenario through a pipe to the solver and get the solution backDavid Kalnischkies2011-04-021-2/+26
| | | | The solution is NOT interpreted so far.
* parse also the action flags Upgrade, Dist-Upgrade and alike from the requestDavid Kalnischkies2011-04-011-7/+37
|
* Read and apply install/remove requests correctlyDavid Kalnischkies2011-04-011-8/+93
|
* WriteRequest according to current EDSP draftDavid Kalnischkies2011-03-311-7/+17
|
* rename edspwriter to straight edsp in toplevel as it does more thanDavid Kalnischkies2011-03-311-0/+160
just writing stuff… it also reads and can work for both: - APT talking to an external solver - an external solver (understanding EDSP) talking to APT