Send patches - preferably formatted by git format-patch - to patches at archlinux32 dot org.
summaryrefslogtreecommitdiff
path: root/lib
AgeCommit message (Collapse)Author
2011-08-16dload: zero out pm_errno in curl_download_internalDave Reisner
This reverts some hacky behavior from 5fc3ec and resets the handle's pm_errno where it should be reset -- prior to each download. This prevents a transaction with a download from being aborted when a package is successfully grabbed from a secondary server. Signed-off-by: Dave Reisner <dreisner@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Update message catalogsDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Update translations from TransifexDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Only check necessary signatures and checksumsDan McGee
The precedence goes as follows: signature > sha256sum > md5sum Add some logic and helper methods to check what we have available when loading a package, and then only check what is necessary to verify the package. This should speed up sync database verifies as we no longer will be doing both a checksum and a signature validation. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15libalpm: fix a remaining old syntax of RET_ERR() macroRémy Oudompheng
It would prevent compilation of pacman on FreeBSD, and possibly other systems. Signed-off-by: Rémy Oudompheng <remy@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15lib/sync: ignore download errors for as long as possibleDave Reisner
Previously, the behavior was such that if a sync operation required packages from multiple repos, a download error in the first repo would cause a hard repo, ignoring the remainder of the repositories. Change this behavior so that we do a better job of fetching as many packages as possible before aborting the transaction. There's a little bit of refactoring mixed in here to get rid of some useless variables. Since we now depend heavily on the value of handle->pm_errno being accurate the determine the function's return value, we clear it when the transaction state is set. Fixes FS#25532. Signed-off-by: Dave Reisner <dreisner@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Further fixes to replacement iterationDan McGee
A partial fix for this was in commit 7de92cb22, but this should fix the remaining cases. There are still several issues dealing with "provision as replacement" selection however. Addresses FS#25538 and FS#25527. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Parse conflicts/provides/replaces at database load timeDan McGee
We did this with depends way back in commit c244cfecf654d3 in 2007. We can do it with these fields as well. Of note is the inclusion of provides even though only '=' is supported- we'll parse other things, but no guarantees are given as to behavior, which is more or less similar to before since we only looked for the equals sign. Also of note is the non-inclusion of optdepends; this will likely be resolved down the road. The biggest benefactors of this change will be the resolving code that formerly had to parse and reparse several of these fields; it only happens once now at load time. This does lead to the disadvantage that we will now always be parsing this information up front even if we never need it in the split form, but as these are uncommon fields and our parser is quite efficient it shouldn't be a big concern. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Validate the sha256sum if availableDan McGee
Adjust load_internal() to check the sha256sum value if we have it. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15decode_signature: guess signature data length for efficiencyDan McGee
We may end up allocating 1 or 2 extra bytes this way, but it is worth it to simplify the method and not have to call base64_decode() a second time. Use the hueristic that base64 encoding produces 3 bytes of decoded data for every 4 bytes of encoded data. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Remove checksum access indirectionDan McGee
These items are never present in anything but sync databases, nor do we even try to load them from the local database. Remvoe the indirection meant to allow the caching layer to work since it will never do anything anyway. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15lib/sync: reset flag after rejecting a replaceDave Reisner
This prevents iteration through the remainder of the current tree, with pacman claiming that they're all replacements to the original replacement candidate. :: Synchronizing package databases... allanbrokeit is up to date testing is up to date core is up to date extra is up to date community-testing is up to date community is up to date :: Starting full system upgrade... :: Replace util-linux-git with core/util-linux? [Y/n] n :: Replace util-linux-git with core/vi? [Y/n] n :: Replace util-linux-git with core/vpnc? [Y/n] n :: Replace util-linux-git with core/wget? [Y/n] n :: Replace util-linux-git with core/which? [Y/n] n :: Replace util-linux-git with core/wireless-regdb? [Y/n] n :: Replace util-linux-git with core/wireless_tools? [Y/n] n :: Replace util-linux-git with core/wpa_actiond? [Y/n] n :: Replace util-linux-git with core/wpa_supplicant? [Y/n] n :: Replace util-linux-git with core/xfsprogs? [Y/n] n :: Replace util-linux-git with core/xinetd? [Y/n] n :: Replace util-linux-git with core/xz? [Y/n] n :: Replace util-linux-git with core/zd1211-firmware? [Y/n] n :: Replace util-linux-git with core/zlib? [Y/n] n there is nothing to do Signed-off-by: Dave Reisner <dreisner@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Load and allow access to sha256sumDan McGee
This adds a field in the package struct for this checksum type as well as allowing access via the API to it. The frontend is now able to display any read value. Note that this does not implement any use or verification of the value internally. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Add ALPM sha256sum routinesDan McGee
These mirror ones we already have for md5sums. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Add sha2 (sha256) routines from PolarSSLDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15pacman/package: show presence of signature in in -SiDave Reisner
adds a new API method: alpm_pkg_get_base64_sig [Dan: don't use a new header string in frontend] Signed-off-by: Dave Reisner <dreisner@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15alpm.h: fix inconsistency in function prototypeDave Reisner
Signed-off-by: Dave Reisner <dreisner@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15lib/dload: avoid deleting .part file on too-slow xferDave Reisner
Take this opportunity to refactor the if/then/else logic into a switch/case which is likely going to be needed to fine tune more exceptions in the future. Fixes FS#25531 Signed-off-by: Dave Reisner <dreisner@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-15Update base64 PolarSSL codeDan McGee
Also adjust our code using it for the size_t adjustments made by upstream. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-11Update MD5 routines with changes from PolarSSLDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-11Add -S --recursive operationDan McGee
This closely matches what we had before for -R --recursive. Basically, when specifying a target (e.g., pacman), we can now recursively pull all dependencies, regardless of version specifiers and whether they are already satisfied in the local database. This could be used to update pacman on a system with an old glibc, for example, as both pacman and glibc would get pulled into the transaction. This is most useful with --needed to prevent needless reinstalls as described in the man page changes. The end goal of this change is to wire it into SyncFirst and have it be the default mode of operation there, but that belongs in a separate changeset. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-11Dependency code style cleanupsDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-11stop progress callbacks after curl_easy_perform returnsDave Reisner
This prevents possible null dereferences in FTP transfers when the progress callback is touched during connection teardown. http://curl.haxx.se/mail/lib-2011-08/0128.html Signed-off-by: Dave Reisner <dreisner@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-11dload: remove unnecessary cast in alpm_load_payload_freeDave Reisner
Dan: make it compile, s/load/payload/. Signed-off-by: Dave Reisner <dreisner@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-09Parse replaces strings as dep strings with version specsDan McGee
This is done extremely crudely and is not very efficient, but it does push us down the path of being closer to right, as one additional test now passes. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-09Conflict comparison performance enhancementsDan McGee
* Add *_hash fields to conflict struct and populate them * Remove unnecessary backwards string comparisons Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-09Update string catalogs after string tweaksDan McGee
This also pulls in some early translations we had entered in Transifex in the last day so those would not be lost. The diffstat is huge and not very telling as usual, as all sorts of fuzzyness switches happened this time around for some reason. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-09Don't walk off front of string when stripping newlineDan McGee
If the string was zero-length to begin with, or consists of only newline characters, nothing stopped us from incrementing right off the front of the string. Ensure len stays above zero the whole time. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-09Fix compile error when curl is not usedDan McGee
Noticed in my PowerPC Linux VM: cc1: warnings being treated as errors dload.c:45: error: 'get_filename' defined but not used make[3]: *** [dload.lo] Error 1 Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-09Depend on name_hash being setDan McGee
This is a fairly valid assumption at this point, or at least as good of one as assuming packages all have names. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-08Add reason to corrupted package callbackDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-08Attempt to fix up some of the brokenness around failed package loadsDan McGee
This is a bit of a mess, due to the fact that we have a progress meter running. It is also ironic that we are in the midst of a method named "commit" when we haven't done a damn thing yet, and can still fail hard if either a checksum or signature is invalid or unrecognized. Adapt the former test_md5sum method to be invoked for any of the various failure types, which at least gives the user some indication of what packages are failing. A second patch will be needed to actually show worthwhile error codes, but this is going to involve modifying the actual data passed with the callback. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-08Update all translation filesDan McGee
This moves us toward staring translations for the 4.0.0 release, although this should not be interpreted as a string freeze by any means. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-08_alpm_access(): don't call gettext() in debug level loggersDan McGee
This is standard procedure elsewhere and cuts down on translations that won't be seen (and we don't want if we need English debug output anyway). Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-08Merge branch 'maint'Dan McGee
Conflicts: lib/libalpm/dload.c lib/libalpm/po/fi.po lib/libalpm/po/libalpm.pot po/de.po po/fi.po src/pacman/po/pacman.pot src/pacman/util.c
2011-08-08Store a package info level flag if we fail to load dataDan McGee
If we are missing a local database file, we get repeated messages over and over telling us the same thing, rather than being sane and erroring only once. This package adds an INFRQ_ERROR level that is added to the mask if we encounter any errors on a local_db_read() operation, and short circuits future calls if found in the value. This fixes FS#25313. Note that this does not make any behavior changes other than suppressing error messages and repeated code calls to failure cases; we still have more to do in the "local database is hosed" department. Also make a small update to the wrong but unused flags set in be_package; using INFRQ_ALL there was not totally correct. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-08Update translations from transifexDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-08Check return value of rename() callsDan McGee
We did a good job checking this in add.c, but not necessarily anywhere else. Fix this up by adding checks into dload.c, remove.c, and conf.c in the frontend. Also add loggers where appropriate and make the message syntax more consistent. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-08-02Don't trim whitespace when reading database entriesDan McGee
We don't write with extra or unknown whitespace, so there is little reason for us to trim it when reading either. This also fixes the hopefully never encountered "paths that start or end with spaces" issue, for which two pactests have been added. The tests also contain other evil characters that we have encountered before and handle just fine, but it doesn't hurt to ensure we don't break such support in the future. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-28Revamp signing checksDan McGee
This ensures we are actually making correct use of the information gpgme is returning to us. Marginal being allowed was obvious before, but Unknown should deal with trust level, and not the presence or lack thereof of a public key to validate the signature with. Return status and validity information in two separate values so check methods and the frontend can use them independently. For now, we treat expired keys as valid, while expired signatures are invalid. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-28Make free_groupcache() privateDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-25libalpm: Set ret correctly in download_internal()Lukas Fleischer
Immediately jump to the cleanup code after setting the return code to -1 in case rename() fails. Otherwise, it will be reset to 0 right after we leave the if branch. Signed-off-by: Lukas Fleischer <archlinux@cryptocrack.de> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-21Convert package filelists to an array instead of linked listDan McGee
This accomplishes quite a few things with one rather invasive change. 1. Iteration is much more performant, due to a reduction in pointer chasing and linear item access. 2. Data structures are smaller- we no longer have the overhead of the linked list as the file struts are now laid out consecutively in memory. 3. Memory allocation has been massively reworked. Before, we would allocate three different pieces of memory per file item- the list struct, the file struct, and the copied filename. What this resulted in was massive fragmentation of memory when loading filelists since the memory allocator had to leave holes all over the place. The new situation here now removes the need for any list item allocation; allocates the file structs in contiguous memory (and reallocs as necessary), leaving only the strings as individually allocated. Tests using valgrind (massif) show some pretty significant memory reductions on the worst case `pacman -Ql > /dev/null` (366387 files on my machine): Before: Peak heap: 54,416,024 B Useful heap: 36,840,692 B Extra heap: 17,575,332 B After: Peak heap: 38,004,352 B Useful heap: 28,101,347 B Extra heap: 9,903,005 B Several small helper methods have been introduced, including a list to array conversion helper as well as a filelist merge sort that works directly on arrays. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-19Clean up my debug logger messDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-18Fix test suite when GPGME is disabledDan McGee
As noted by Allan, we failed pretty hard if gpgme was compiled out. With these changes, only sign001.py fails. This can/will be fixed later once we beef up the test suite with more signing tests anyway. Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-18replace access() calls for debug info where applicableFlorian Pritz
Signed-off-by: Florian Pritz <bluewind@xinu.at> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-18signing.c: check if needed files are readableFlorian Pritz
If we can't read the keyring, gpgme will output confusing debug information and fail to verify the signature, so we should log some debug information. Signed-off-by: Florian Pritz <bluewind@xinu.at> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-18add _alpm_access() wrapperFlorian Pritz
This is a wrapper function for access() which logs some debug information and eases handling in case of split directory and filename. Signed-off-by: Florian Pritz <bluewind@xinu.at> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-18Fix compilation without gpgmeAllan McRae
Signed-off-by: Allan McRae <allan@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2011-07-18Handle removal of empty directories properlyDan McGee
This addresses FS#25141. We shouldn't remove every empty directory we come across during the removal process unless it is truly not known to any other package. This will prevent removal of essential directories such as '/var/lock/'. This is accomplished by first checking the empty/non-empty status of a directory, which was previously done implicitly by calling rmdir() and ignoring errors. We do this to avoid the next (new) check in most cases, which is to look at all local packages to see if the to-be-removed directory is present in another packages' filelist. If we do not find it anywhere, then we remove it, else we keep the file around. The pactest has been updated to test more cases, as well as finding a flaw in the original expected to fail case- we need separate DIR and FILE based EXIST rules. Signed-off-by: Dan McGee <dan@archlinux.org>