summaryrefslogtreecommitdiff
path: root/apt-pkg
Commit message (Collapse)AuthorAgeFilesLines
* Add the ?security patternJulian Andres Klode2024-02-132-0/+11
|
* Add a new ?phasing patternJulian Andres Klode2024-02-132-0/+12
| | | | | This selects all packages that are being kept back due to phasing on your system.
* Add public phased update APIJulian Andres Klode2024-02-137-110/+117
| | | | | | | | | | | | | | | | This moves the functions of the PhasedUpgrader class into various other classes so they can be publicly exposed. This introduces three new functions: pkgDepCache::PhasingApplied() tells you whether phasing should be applied to the package. pkgProblemResolver::KeepPhasedUpdates() keeps back updates that have phasing applied. pkgCache::VerIterator::IsSecurityUpdate() determines whether this version contains security fixes.
* For phasing, check if current version is a security update, not just ↵Julian Andres Klode2024-02-131-1/+0
| | | | | | | | | previous ones We only considered an update a security update if a previous update is a security update but not the update in question itself. LP: #2051181
* pkgcachegen: Use placement new to construct headerJulian Andres Klode2024-01-121-1/+1
| | | | | | | | | | | | | | | | | Avoid copying the header from a stack allocated object as this will copy uninitialized padding bytes into the cache, triggering valgrind errors which people then use as a strawman for unrelated errors on armhf. In an optimal world we should annotate the allocator however such that valgrind actually does treat those bytes as uninitialized and then supress warnings in the harmless places, such that when you then go and try to access it in a place that matters, you do get an error for uninitialized memory. Currently any access within the pool will be considered initialized which is clearly suboptimal. But this is very much a TBD topic and involves annotating the allocator everywhere.
* Merge branch 'fix/dontstorediffindex' into 'main'Julian Andres Klode2024-01-083-51/+36
|\ | | | | | | | | Do not store .diff_Index files in update See merge request apt-team/apt!316
| * Do not store .diff_Index files in updateDavid Kalnischkies2024-01-031-22/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | Nowadays we only download the index file if we have a non-current file on disk which we want to patch. If that is the case, any index file for patches we could have stored is by definition outdated, so storing those files just takes up disk space. At least, that is the case if we have a Release file – if we don't this commit introduces a needless redownload for such repositories but such repositories are an error by default and if they can't be bothered to provide a Release file its very unlikely they actually ship diffs, so adding detection code for this seems pointless at best.
| * Improve and test distclean implementationDavid Kalnischkies2024-01-032-29/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | The implementation as-is as various smaller/esoteric bugs and inconsistencies like apt-get not supporting them, the option -s being supported in code but not accepted on the command line, the regex not escaping the dot before the file extension and exposing more implementation details to public headers than we actually need. Also comes with a small test case to ensure it actually works. References: bd7c126e3fb1b94e76e0e632c657cea854586844
* | Do not silently ignore directories for reserved file namesJulian Andres Klode2023-12-131-3/+5
|/ | | | | | Files with reserved extensions like .list, .sources, .conf, and .pref should receive notices in their respective directories even if they are directories.
* Merge branch 'fix-959093' into 'main'Julian Andres Klode2023-12-082-13/+64
|\ | | | | | | | | Add 'dist-clean' command to remove packages and list files See merge request apt-team/apt!290
| * Add 'dist-clean' command to remove packages and list filesGábor Németh2023-11-222-13/+64
| | | | | | | | | | | | | | We assume all files in the 'listsdir' are candidates. Keep only files ending with Release, Release.gpg, and InRelease. Closes: #959093
* | Merge branch 'fix/unknownarchconflicts' into 'main'Julian Andres Klode2023-12-081-5/+15
|\ \ | | | | | | | | | | | | Have Grp.FindPreferredPkg return very foreign pkgs as last resort See merge request apt-team/apt!310
| * | Have Grp.FindPreferredPkg return very foreign pkgs as last resortDavid Kalnischkies2023-12-041-5/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Usually this method will return the package in the most preferred architecture (e.g. native) as that is usually what the user talks about and also information wise for our internal usage the most dense. Early on in parsing Packages files through it can happen that we encounter stanzas about packages in architectures we are not even configured to know about – we have to collect them anyhow as we might be requested to show info about them or they could be in the status file and we can't ignore stanzas in the status file… trouble is that this method used to not return anything if only such an architecture was present if we later discover other architectures which causes Provides and Conflicts which are added lazily on discovery of an architecture to not be added correctly. The result is like in the testcase that apt could be instructed to install a package without respecting its negative dependencies, which is bad even if its discovered by dpkg and refused. It does only happen with unknown architectures through which mostly happens if you are unlucky (amd64 users tend to be very lucky as that sorts early) and use flat-style repositories containing multiple architectures. Reported-By: Tianyu Chen (billchenchina) on IRC
* | | Merge branch 'main' into 'main'Julian Andres Klode2023-12-081-0/+3
|\ \ \ | |/ / |/| | | | | | | | apt-pkg/cacheset.cc: set ShowErrors to true when no version matched See merge request apt-team/apt!308
| * | apt-pkg/cacheset.cc: set ShowErrors to true when no version matchedTianyu Chen2023-12-081-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | Enforce helper.canNotGetVersion to show error if no version matched. Regression-of: 572810e9f321237873d1536c88991d7825c6f1db Closes: #1053887
* | | Merge branch 'fix-time-unit-comment' into 'main'Julian Andres Klode2023-11-201-1/+1
|\ \ \ | |/ / |/| | | | | | | | Fix incorrect time unit comment for PulseInterval See merge request apt-team/apt!304
| * | Fix incorrect time unit comment for PulseIntervalsid2023-07-271-1/+1
| | |
* | | Revert "Do not fail on systems running in FIPSmode."Julian Andres Klode2023-10-041-3/+0
| | | | | | | | | | | | This reverts commit 668451def296afeb0c358a7d80ff39dc546defab.
* | | Stop calculating Description-md5 if missingJulian Andres Klode2023-10-041-19/+1
| | | | | | | | | | | | | | | | | | | | | | | | This avoids the rabbit hole of md5 on FIPS systems, and repositories have moved to including the value as well. Also stop validating the field, this can be an arbitrary string as far as we are concerned.
* | | Downgrade unmerged-usr from error to two warningsJulian Andres Klode2023-09-201-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | One warning will be issued before the Y/n prompt, the other will be issued at the end after package installs have been attempted or if there were other failures, such that the last line you see is warnings about unmerged-usr I do not anticipate this to be the final version either, but there we go. Closes: #1052058
* | | Remove unnecessary explicit st_dev check for merged-usrJulian Andres Klode2023-09-161-2/+0
| | |
* | | Only accept installs of usrmerge on unmerged-usr systemsJulian Andres Klode2023-09-162-2/+33
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As of bookworm, merged-usr is mandatory, and people got caught in the crosshairs of the dpkg fsys-unmessusr debacle and inadvertently reverted back to an unmerged configuration and continue to remain on an unsupported system unknowingly. Help them by erroring out when they are installing packages on /, they are not in a chroot, and a usrmerge package is available.
* | | Compare SHA256 to check if versions are really the sameJulian Andres Klode2023-08-023-2/+41
|/ / | | | | | | | | | | | | | | | | | | | | | | | | If we know both SHA256, and they're different, the packages are. This approach stores the SHA256 only at runtime, avoiding the overhead of storing it on-disk, because when we update repositories we update all of them anyhow. Note that pkgCacheGenerator is hidden, so we can just modify its ABI, hooray. Closes: #931175 LP: #2029268
* | Fix snapshot crashesJulian Andres Klode2023-07-121-1/+3
| | | | | | | | | | We did not handle multiple components properly, add a contrib component to the test case.
* | Merge branch 'main' into 'main'Julian Andres Klode2023-07-111-0/+3
|\ \ | | | | | | | | | | | | Do not fail on systems running in FIPSmode. See merge request apt-team/apt!295
| * | Do not fail on systems running in FIPSmode.A. Maitland Bottoms2023-05-291-0/+3
| | | | | | | | | | | | | | | Initialize using gcrypt's GCRYCTL_NO_FIPS_MODE, available since gcrypt version 1.10.0, otherwise apt aborts on FIPS enabled systems.
* | | Merge branch 'pu/ubuntu-bug-2025462' into 'main'Julian Andres Klode2023-07-113-14/+55
|\ \ \ | | | | | | | | | | | | | | | | dist-upgrade: Revert phased updates using keeps only See merge request apt-team/apt!299
| * | | Do not mark updates for install that are still phasingJulian Andres Klode2023-07-071-4/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fixes an issue where phased updates gain new dependencies and cause them to be installed despite themselves not being installed. In the cause of investigation, it turned out that we also need to evaluate the candidate version at those early stage rather than the install version (which is only valid *after* MarkInstall). This does not fully resolve the problem: If an update pulls in a phased update, depends are still being installed. Resolving this while ensuring that phased updates cannot uninstall packages requires us to do a minimization of changes by trying to keep back each new install removal and then seeing if any dependency is being broken by it. This is more complex and will happen later.
| * | | dist-upgrade: Revert phased updates using keeps onlyJulian Andres Klode2023-07-053-10/+32
| |/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the bug, mutter was kept back due to phasing and the new gnome-shell depended on that, and was therefore kept back as well, however, gnome-shell-common was not broken, and apt decided to continue upgrading it by removing gnome-shell and the ubuntu desktop meta packages. This is potentially a regression of LP#1990586 where we added keep back calls to the start of the dist-upgrade to ensure that we do not mark stuff for upgrade in the first place that depends on phasing updates, however it was generally allowed by the resolver to also do those removals. To fix this, we need to resolve the update normally and then use ResolveByKeepInternal to keep back any changes broken by held back packages. However, doing so breaks test-bug-591882-conkeror because ResolveByKeep keeps back packages for broken Recommends as well, which is not something we generally want to do in a dist-upgrade after we already decided to upgrade it. To circumvent that issue, extend the pkgProblemResolver to allow a package to be policy broken, and mark all packages that already were already going to be policy broken to be allowed to be that, such that we don't try to undo their installs. LP: #2025462
* / / update: Add notice about missing Signed-By in deb822 sourcesJulian Andres Klode2023-06-274-2/+20
|/ / | | | | | | | | | | | | We want to gently steer users towards having Signed-By for each source such that we can retire a shared keyring across sources which improves resilience against configuration issues and incompetent malicious actors.
* | Rework the domain iteration in the host fallbackJulian Andres Klode2023-05-241-3/+18
| | | | | | | | | | | | Separate the determination of the next level domain into its own function and split out the "we found a result" into its own break for improved readability.
* | Ensure that Snapshots: no doesn't get overriden by host-name configJulian Andres Klode2023-05-241-7/+3
| |
* | Seed snapshot servers for well-known hostsJulian Andres Klode2023-05-242-9/+38
| | | | | | | | | | This will attempt to fallback to a per-server setting if we could not determine a value from the release file.
* | Merge branch 'pu/snapshot' into 'main'Julian Andres Klode2023-05-026-4/+112
|\ \ | | | | | | | | | | | | Add --snapshot and --update support See merge request apt-team/apt!291
| * | Initial support for snapshot servers, apt --snapshot optionJulian Andres Klode2023-05-026-4/+112
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Provide snapshot support for offical Debian and Ubuntu archives. There are two ways to enable snapshots for sources: 1. Add Snapshot: yes to your sources file ([snapshot=yes]). This will allow you to specify a snapshot to use when updating or installing using the --snapshot,-S option. 2. Add Snapshot: ID to your sources files to request a specific snapshot for this source. Snapshots are discovered using Label and Origin fields in the Release file of the main source, hence you need to have updated the source at least once before you can use snapshots. The Release file may also declare a snapshots server to use, similar to Changelogs, it can contain a Snapshots field with the values: 1. `Snapshots: https://example.com/@SNAPSHOTID@` where `@SNAPSHOTID@` is a placeholder that is replaced with the requested snapshot id 2. `Snapshots: no` to disable snapshot support for this source. Requesting snapshots for this source will result in a failure to load the source. The implementation adds a SHADOWED option to deb source entries, and marks the main entry as SHADOWED when a snapshot has been requested, which will cause it to be updated, but not included in the generated cache. The concern here was that we need to keep generating the shadowed entries because the cleanup in `apt update` deletes any files not queued for download, so we gotta keep downloading the main source. This design is not entirely optimal, but avoids the pitfalls of having to reimplement list cleanup. Gaps: - Ubuntu Pro repositories and PPAs are not yet supported.
* | | Keep "or group" when installing package to satisfy itJacob Kauffmann2023-05-021-1/+4
|/ /
* | Address statements of public domainJulian Andres Klode2023-03-0612-2/+54
| |
* | Merge branch 'pu/never-sections-matching' into 'main'Julian Andres Klode2023-03-061-2/+2
|\ \ | | | | | | | | | | | | Fix permissions && change section matching in config files to be more gitignore style rightmost match See merge request apt-team/apt!286
| * | Use a more gitignore style matchingJulian Andres Klode2023-02-271-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | Use a rightmost match for / so that if we end up with a Section: a/b/c, a 'c' matcher still matches. If the section does not contain any /, it can be matched using /pattern, e.g. /c only matches Section: c, but not Section: a/b/c.
* | | Do not store trusted=yes Release file unconditionallyDavid Kalnischkies2023-03-041-3/+6
| |/ |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A source marked with trusted=yes can still fail verification of the Release file, mostly for Date related issues, like being too new or too old, which have other options to force them in. The update code was not using the Release file (which was a InRelease file but failed verification – which was overridden by trusted=yes) as intended, but it marked it for storage, so that this "bad" Release file would end up being moved into lists/, which is bad as the indexes it refers to aren't updated while the next update run assumes that the indexes are in the state the Release file claims them to be in. Fixed simply by making the storage conditional on the usage as intended, which also resolves a second issue: The verification can also detect that a Release file we got is older than what we already have to avoid down- grade attacks. The more likely explanation is a slightly outdated mirror in a rotation/CDN through, so this gets the silent treatment to avoid scaring users by handling it as if we had got the same Release file we already have stored locally, removing the freshly received older file in the process alongside setting some variables. Those variables were already modified in the trusted=yes case though resulting in the stored Release file being removed instead. Not modifying the variables too early resolves this problem as well. Both seem to exist since at least 2015 as traces are visible in 448c38bdcd already, which shuffled lots of code around including the bad ones, but as we are in trusted=yes land, security is of no concern here, this "just" leads to failed pinning, hashsum mismatches and other strange problems in follow-up calls depending on how out of sync the Release file (if its still present) is with the rest of the trusted data. Reported-By: Dima Kogan <dkogan@debian.org> on IRC Tested-By: Dima Kogan <dkogan@debian.org>
* | Detect trimmed changelogs and pick online insteadDavid Kalnischkies2023-03-031-11/+56
|/ | | | | | | | | | | | | | | | | | | We only check the start of these lines to avoid hard coding the exact command and we pick 150 as maximum line length as the longest package name on my system is apparently 75 characters long. We could choose longer or shorter without much issue as over-length just means we mishandle the rest of the line as a new line and it should be really unlikely that a) lines are that long in this file and b) that such long lines contain one of our trigger sequences – but even if, all we do is start a download of an online file. Could be worse. This auto-detection can be avoided by setting Acquire::Changelogs::AlwaysOnline (or Origin specific sub options) to "true" if you always want the changelog from an online source. The reverse – setting it to "false" in the hope it would not get the changelog from an online source – was not and is still not possible. Closes: #1024457
* Merge branch 'feature/non-free-firmware' into 'main'Julian Andres Klode2023-02-272-12/+28
|\ | | | | | | | | Support transition to new non-free-firmware component See merge request apt-team/apt!282
| * Have values in Section config trees refer to them in all componentsDavid Kalnischkies2023-01-301-12/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Hard coding each and every component is not only boring but given that everyone is free to add or use more we end up in situations in which apt behaves differently for the same binary package just because metadata said it is in different components (e.g. non-free vs. non-free-firmware). It is also probably not what the casual user would expect. So we instead treat a value without a component as if it applies for all of them. The previous behaviour can be restored by prefixing the value with "<undefined>/" as in the component is not defined. In an ideal world we would probably use "*/foo" for the new default instead of changing the behaviour for "foo", but it seems rather unlikely that the old behaviour is actually desired. All existing values were duplicated for all (previously) known components in Debian and Ubuntu.
| * Teach apt-cdrom's scoring system about non-free-firmwareCyril Brulebois2023-01-301-0/+2
| | | | | | | | Closes: #1029751
* | Unhide virtual overloads in some cachefiltersDavid Kalnischkies2023-01-291-0/+2
| | | | | | | | | | | | | | | | | | The matchers are usually called via its base class, but if we would call them via the derivate class we would not be able to call the not explicitly "imported" overloads, which would be strange. Reported-By: gcc-13 -Woverloaded-virtual Gbp-Dch: Ignore
* | Drop redundant 'virtual' on function marked 'override'David Kalnischkies2023-01-291-32/+32
| | | | | | | | | | Reported-By: clang-tidy [cppcoreguidelines-explicit-virtual-functions] Gbp-Dch: Ignore
* | Include cstdint for uint16_t in our mmap wrappingDavid Kalnischkies2023-01-291-0/+1
| | | | | | | | | | | | | | In gcc-13 internal includes were reduced exposing our laziness. Reported-By: gcc-13 Gbp-Dch: Ignore
* | Avoid comparison of integers of different signedness for ETADavid Kalnischkies2023-01-291-1/+2
| | | | | | | | | | Reported-By: gcc -Wsign-compare Gbp-Dch: Ignore
* | Avoid triggering unused parameter warning in staticSIGWINCHDavid Kalnischkies2023-01-291-1/+1
|/ | | | | Reported-By: gcc -Wunused-parameter Gbp-Dch: Ignore
* make ?installed pattern match installed version only when narrowedJulian Andres Klode2023-01-131-0/+5
| | | | | | | | This is the correct behavior, but it was overlooked when aptitude patterns where ported. I remember wondering about this, but I checked the aptitude code and saw a check that CurrentVer != 0 or something and then apparently did not notice another implementation for version matching.
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2026-05-12 17:32:59 +0000