February 2008 - Pacman-dev - lists.archlinux.org

[pacman-dev] Fwd: [BUG] Serious problems with file relocation on upgrade
by Nagy Gabor 11 Feb '08

11 Feb '08

I think we should fix/discuss this before 3.1.2 release, because this is quite critical imho. Part I ----- Továbbított levél Nagy Gabor <ngaba(a)bibl.u-szeged.hu> címről ----- Dátum: Sat, 9 Feb 2008 18:07:28 +0100 Feladó: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Viszontválasz cím: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Tartalom: [BUG] Serious problems with file relocation on upgrade Címzett: Xavier <shiningxc(a)gmail.com> Hi! For more info see the attached patch. Bye ----- Vége a továbbított üzenetnek ----- ---------------------------------------------------- SZTE Egyetemi Könyvtár - http://www.bibl.u-szeged.hu This mail sent through IMP: http://horde.org/imp/

1 0

[pacman-dev] Differences between pacman and pacman-g2
by ciol 11 Feb '08

11 Feb '08

Hi, what are the differences between pacman and pacman-g2? Thanks.

3 4

[pacman-dev] memleak caused by "don't duplicate package in cache"
by Xavier 10 Feb '08

10 Feb '08

Previously, all installed packages were duplicated when added to the cache. in libalpm/add.c : 801 >-if(_alpm_db_add_pkgincache(db, newpkg) == -1) { And the _alpm_db_add_pkgincache function in cache.c duplicated the newpkg. I wanted to avoid this duplication with the following commit : http://projects.archlinux.org/git/?p=pacman.git;a=commit;h=8240da6cb3ff95ad… I simply removed the pkg_dup in that function. Since I removed a pkg_dup, I had to remove the corresponding pkg_free call also. Previously, the packages in the targets list were freed in the _alpm_trans_free function. And the duplicated package in the cache were freed in _alpm_db_free_pkgcache. Since the cache is freed only when alpm is released, which happens at the very end, I let the free there. So I removed it from _alpm_trans_free, which is called when a transaction is released. Now the ugly part begins. The targets are only added to the cache when they are installed, not when they are removed (obviously). That explains the condition I added in trans_free : + } else if (trans->type == PM_TRANS_TYPE_REMOVE || + trans->type == PM_TRANS_TYPE_REMOVEUPGRADE) { alpm_list_free_inner(trans->packages, (alpm_list_fn_free)_alpm_pkg_free); So, when a whole install transaction goes fine, without errors, everything works well. All targets are added to the cache, and are then all freed correctly by the _alpm_db_free_pkgcache function. However, when the transaction doesn't go correctly, the targets are loaded in memory, but they are not installed, and so not freed! All targets stay in memory, and cause a big memleak. For example, a conflict could be detected, and the transaction is aborted (as in sync400). But it might even be possible that a transaction gets aborted in the middle. Maybe because it ran out of memory, or because the user interrupted it, or whatever. So, for solving this issue, maybe we would need a way to track down which packages were loaded in the cache. That should allow us to free everything correctly, which doesn't seem possible currently. We could then free in the trans_free function all targets that are not in the pkg cache. And the remaining targets that are in the pkg cache will be freed when the cache is freed.

2 2

[pacman-dev] Thoughts of a 3.1.2 release
by Dan McGee 08 Feb '08

08 Feb '08

Here is the current git shortlog since the last release: $ git shortlog v3.1.1..maint Alex Merry (1): updatesync: incorrect package deletion logic Chantry Xavier (9): workaround for a gettext string starting with --. two string fixes. makepkg : improve determination of svn revision. Clarify the "cancel current operation" message. repo-remove: print an error in case of an empty db Clarify the "failed to add target" errors. add 'force' to PKGINFO, have repo-add respect it. Add new ChangeLog.proto file. Don't follow symlinks with -Qo. Dan McGee (7): makepkg: add check to ensure we have non-URL files in build dir Remove frontend translation of "debug:" message doc/makepkg.8: add missing '-' to manpage option makepkg: ensure binaries in /opt/* are stripped Add deprecation warnings to gensync and updatesync doc/makepkg: document --log, address locale issues pacman/util.c: add mdirname function Roman Kyrylych (1): Remove /bin/true from install.proto Given this list of changes, I think it would be prudent to think about a 3.1.2 release soon, although it is not something we need to rush out as there is nothing pressing like the provides logic prior to 3.1.1. Here are some issues I think we should address before prepping for release: http://bugs.archlinux.org/task/9362 - Any opinions here? Put them in the bug report. http://bugs.archlinux.org/task/9242 - Still not completely sure if this is easily solvable http://bugs.archlinux.org/task/9230 - Deals with makepkg/versionpkg stuff, should be an easy fix Please add any (constructive) comments you may have, and quote the relevant parts of this email. *Please* keep it limited to things that would make it into 3.1.2- if it is clearly something that can't fit in a maint release, I won't even bother responding. A second email will follow this one with some things that Xavier has brought up he thinks need addressing. Some of these are relevant for 3.1.2, others may be a bit further in the future. -Dan

3 3

[pacman-dev] GZip compression ratios
by Travis Willard 08 Feb '08

08 Feb '08

Just doing an update, and I noticed this: Targets: coreutils-6.10-1 bash-3.2.033-2 cairo-1.4.14-1 gtk2-2.12.7-1 gcc-libs-4.2.3-3 mpfr-2.3.1-1 gcc-4.2.3-1 mesa-7.0.3rc1-1 acroread-8.1.2-1 automake-1.10.1-2 kernel26-2.6.24-2 catalyst-8.01-3 curl-7.18.0-1 devtools-0.5-1 dhcpcd-3.2.1-1 e2fsprogs-1.40.5-1 faac-1.26-1 filesystem-2007.11-6 fakeroot-1.9.2-1 findutils-4.2.32-1 flex-2.5.33-4 librsvg-2.20.0-1 libxmu-1.0.4-1 gimp-2.4.4-1 git-1.5.4-1 gpgme-1.1.6-1 groff-1.19.2-4 kde-common-3.5.8-3 lm_sensors-3.0.1-1 kdebase-3.5.8-3 lftp-3.6.3-1 libgnomecups-0.2.3-1 libxml2-2.6.31-1 libgnomeprint-2.18.3-1 libgnomeprintui-2.18.2-1 libidl2-0.8.10-1 libmysqlclient-5.0.51-2 libtasn1-1.3-1 libtool-1.5.26-1 lua-5.1.3-1 lvm2-2.02.33-1 man-1.6f-2 man-pages-2.77-1 mysql-clients-5.0.51-3 nasm-2.01-1 orbit2-2.14.12-1 pcre-7.6-2 php-5.2.5-3 pil-1.1.6-4 postgresql-libs-8.3.0-1 sdl_mixer-1.2.8-2 shared-mime-info-0.23-1 snownews-1.5.8-1 spamassassin-3.2.4-2 sqlite3-3.5.5-1 sudo-1.6.9p12-1 tk-8.5.0-2 wget-1.11-1 which-2.19-2 xfsprogs-2.9.5-1 xine-lib-1.1.10-1 xorg-server-1.4.0.90-6 Total Download Size: 194.03 MB Total Installed Size: 194.39 MB Does gzip compression really give us so little as to only save 0.3 MB over uncompressed for so many packages? I figured the difference would be a bit more - does it suck for compressing binary data? Should we consider moving to a different, more compressed archive format? Or am I just making a mountain out of a molehill here? -- Travis

2 1

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.1.1-61-gbfc024e
by Dan McGee 08 Feb '08

08 Feb '08

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "The official pacman repository". The branch, master has been updated via bfc024eab3a1d26ba37620175be815cb2eb59606 (commit) via 77c3cf9790472592902183354ea5481cffc01f58 (commit) via 7dc37109b096b484fc569060cc896b5127419775 (commit) via 14ee1be1ef21e09a84f01dc9623e8513218709d4 (commit) via 11fe18479eeb7bb97062a2922da13ba816597007 (commit) via 0c2206f542ce6df2606586d43f190cd5a423fb13 (commit) via 05d56349588665da7ed01eaa3337be7ece35fd2f (commit) via e63366ae5e701d8e9ae33144f68e8786b092a468 (commit) from e81dec9b8c65f7c882f5f447cdc117783259f4a4 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit bfc024eab3a1d26ba37620175be815cb2eb59606 Merge: 77c3cf9790472592902183354ea5481cffc01f58 14ee1be1ef21e09a84f01dc9623e8513218709d4 Author: Dan McGee <dan(a)archlinux.org> Date: Thu Feb 7 20:16:16 2008 -0600 Merge branch 'maint' commit 77c3cf9790472592902183354ea5481cffc01f58 Author: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Date: Wed Jan 30 13:55:49 2008 +0100 Remove alpm_sync_get_newversion function sync->newversion shouldn't be public at all, and internally we access it directly. (Before pmsyncpkg_t clean-up the analogue of this field [type] was needed in replaces computation.) Signed-off-by: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Signed-off-by: Chantry Xavier <shiningxc(a)gmail.com> commit 7dc37109b096b484fc569060cc896b5127419775 Author: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Date: Mon Jan 28 01:44:23 2008 +0100 alpm_sync_sysupgrade split This patch indroduces a new public alpm_sync_newversion, which scans for new version of a package in sync repos. Hopefully this will reduce code duplication in the future: * check-for-pacman-new-version from front-end can be easier * -Qu refactoring Signed-off-by: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Signed-off-by: Chantry Xavier <shiningxc(a)gmail.com> commit 05d56349588665da7ed01eaa3337be7ece35fd2f Author: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Date: Sun Nov 18 19:05:47 2007 +0100 libalpm/package.c : add _alpm_pkgname_pkg_cmp function. _alpm_pkgname_pkg_cmp(pkgname, pkg) returns true iff pkg's name is pkgname. This is useful if you want to remove a package from pmpkg_t* list, and you want to search for package name. This allows cleaning the -Ru code a bit, by removing the need of a dummy pkg. Signed-off-by: Chantry Xavier <shiningxc(a)gmail.com> commit e63366ae5e701d8e9ae33144f68e8786b092a468 Author: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Date: Sun Nov 18 18:45:46 2007 +0100 New remove option : -u / --unneeded (FS#6505). With --unneeded option 'pacman -R' doesn't stop in case of dependency error; it removes the needed-dependency targets from the target-list instead. See also: http://archlinux.org/pipermail/pacman-dev/2007-October/009653.html . The patch also adds a new causingpkg field to pmdepmissing_t which indicates the to-be-removed package which would cause a dependency break. This is needed, because miss->depend.name may be a provision. miss->causingpkg will be useful in -R dependency error messages too. [Xavier: renamed inducer to causingpkg, removed the _alpm_pkgname_pkg_cmp helper function as requested by Aaron. This might be added by a further commit. Other small cleanups, updated manpage and bash completion.] Signed-off-by: Chantry Xavier <shiningxc(a)gmail.com> ----------------------------------------------------------------------- Summary of changes: contrib/bash_completion | 2 + doc/pacman.8.txt | 5 ++ lib/libalpm/add.c | 10 ++-- lib/libalpm/alpm.h | 6 ++- lib/libalpm/deps.c | 22 +++++++-- lib/libalpm/deps.h | 4 +- lib/libalpm/package.c | 6 ++ lib/libalpm/package.h | 1 + lib/libalpm/remove.c | 24 ++++++++- lib/libalpm/sync.c | 117 ++++++++++++++++++++++++------------------- pactest/pmdb.py | 13 ++++- pactest/tests/remove049.py | 19 +++++++ pactest/tests/upgrade045.py | 16 ++++++ pactest/util.py | 6 ++ scripts/makepkg.sh.in | 7 +++ src/pacman/pacman.c | 3 + 16 files changed, 193 insertions(+), 68 deletions(-) create mode 100644 pactest/tests/remove049.py create mode 100644 pactest/tests/upgrade045.py hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [GIT] The official pacman repository branch, maint, updated. v3.1.1-21-g14ee1be
by Dan McGee 08 Feb '08

08 Feb '08

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "The official pacman repository". The branch, maint has been updated via 14ee1be1ef21e09a84f01dc9623e8513218709d4 (commit) via 11fe18479eeb7bb97062a2922da13ba816597007 (commit) via 0c2206f542ce6df2606586d43f190cd5a423fb13 (commit) from b29838c8250125a22016c16180e941d059d55539 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- ----------------------------------------------------------------------- Summary of changes: lib/libalpm/add.c | 10 ++++------ pactest/pmdb.py | 13 +++++++++++-- pactest/tests/upgrade045.py | 16 ++++++++++++++++ pactest/util.py | 6 ++++++ scripts/makepkg.sh.in | 7 +++++++ 5 files changed, 44 insertions(+), 8 deletions(-) create mode 100644 pactest/tests/upgrade045.py hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] xdelta status
by Nathan Jones 07 Feb '08

07 Feb '08

Today I read a post[1] by Xavier about how deltas cannot currently be used. As I say in the thread, I will work on adding delta creation support to repo-add in the next week. Is there anything else that needs to be changed before this can actually be used? [1] http://bbs.archlinux.org/viewtopic.php?id=43426

3 6

[pacman-dev] Pacman TODO from Xavier
by Dan McGee 06 Feb '08

06 Feb '08

Xavier pointed out a few pending items to me with pacman, and I thought I'd respond to his list here on the ML. Thoughts welcome from anyone else. > * -Qo and symlink : > http://www.archlinux.org/pipermail/pacman-dev/2008-February/011061.html > http://shining.toofishes.net/gitweb/gitweb.cgi?p=pacman.git;a=commit;h=cd43… Merged into maint tonight. > * pacman.conf clean up : > http://www.archlinux.org/pipermail/pacman-dev/2008-January/010996.html Replied to this message on the ML. > * ChangeLog.proto : > http://archlinux.org/pipermail/arch-dev-public/2008-January/004020.html > http://shining.toofishes.net/gitweb/gitweb.cgi?p=pacman.git;a=commitdiff;h=… Merged. > * install contrib files > http://www.archlinux.org/pipermail/pacman-dev/2008-January/010782.html How about an AUR package? pacman-contrib? Someone might even maintain it in community. > * ignore targets not found : > point 2) there : > http://www.archlinux.org/pipermail/pacman-dev/2007-September/009412.html 3.2 issue for sure. Interesting, but I like programs to fail hard with invalid input rather than gloss it over. When you script programs, it can be very misleading if things don't fail when you expect them to. > * broken database with two duplicated entries : > for example : > var/lib/pacman/local/xorg-xauth-1.0.1-1/ > var/lib/pacman/local/xorg-xauth-1.0.2-1/ > > If we want to do this check in testdb, don't we need db_splitname or something? > If we do it in libalpm, where exactly? I punted this one to Xavier because I didn't exactly know how to do it. :) Can't we just load the entire local pkgcache and then scan for duplicate names? I was originally thinking of doing this outside of libalpm, and at some point we can add some consistency checks into the library if necessary that ensure no pkgname is duplicated inside a single repo. > * segfault : FS#9235 > > * memleak caused by 8240da6cb3ff95ad480efe3e1876104024398fae > how to track whether a package has been added to the pkgcache or not. > The target list is currently never freed. only the pkgcache is. > If a package from the target list hasn't been added to the cache (because of a > failure, conflict or whatever), it won't be freed. Xavier- if you could bring this up in a separate email thread, that would probably be best. This is an issue where people are just going to have to sit down and try and work out a system for being sure of when/where a package needs to be freed. > * -Ru (unneeded) branch > http://shining.toofishes.net/gitweb/gitweb.cgi?p=pacman.git;a=shortlog;h=un… Mostly because I just haven't merged it yet. I'll try to get a good look at this soon and see if it is good to go for master. > * xdelta support > maybe we need a mail about the status on this, what needs to be done: > http://bbs.archlinux.org/viewtopic.php?id=43426 > (waiting on Nathan) xdelta support in repo-add would probably be our next step- we may need to refactor the common code out of repo-add and makepkg into some kind of makepkg lib, which I am not completely sure how to address. > * upgrade refactoring > status mail on this too? What does this refer to? The idea of our add/remove package lists? If so, it is still an idea brewing in a lot of our heads but no one has made any huge advances on it yet. Of course, it is not going to me an easy or clean change either. -Dan

4 4

[pacman-dev] pacman.conf little clean up
by Xavier 06 Feb '08

06 Feb '08

1) Why do we have this comment everywhere in the conf file : # Add your preferred servers here, they will be used first In my opinion, all these lines could disappear. It's recommended that you keep all your repositories in sync. That means using the same mirror for all official repo. And since all official repo already have this by default : Include = /etc/pacman.d/mirrorlist Only that file needs to be edited, and this will affect all official repo automatically. Advanced users who know what they are doing might not want to use the same mirror everywhere for specific reasons. But I don't think they need a comment everywhere explaining how to do it. But that comment might rather belong to the mirrorlist file. 2) For similar reasons, I don't think the following comment is needed : # REPOSITORIES # - can be defined here or included from another file I doubt many people use that. pacman.conf currently has only two parts : a few options, and the repositories. I don't think there is a need to split these. And again, people who do it probably know what they are doing. 3) I find pacman.conf more readable with more new lines (especially when read without syntax highlighting), so I added a few of these as well, to better separate the different parts, like the comments and the commented lines. Anyway, all these things are not important at all, and might be stupid ideas. It's just random thoughts that came to my mind after seeing an user messed up again configuring his pacman.conf. --- pacman.conf 2008-01-25 19:34:04.000000000 +0100 +++ pacman.conf.new 2008-01-25 18:03:29.000000000 +0100 @@ -6,19 +6,22 @@ # # GENERAL OPTIONS # + [options] + # The following paths are commented out with their default values listed. # If you wish to use different paths, uncomment and update the paths. + #RootDir = / #DBPath = /var/lib/pacman/ #CacheDir = /var/cache/pacman/pkg/ #LogFile = /var/log/pacman.log + HoldPkg = pacman glibc #XferCommand = /usr/bin/wget --passive-ftp -c -O %o %u # # REPOSITORIES -# - can be defined here or included from another file # - pacman will search repositories in the order defined here # - local/custom mirrors can be added here or in separate files # - repositories listed first will take precedence when packages @@ -34,31 +37,28 @@ # # Testing is disabled by default. To enable, uncomment the following -# two lines. You can add preferred servers immediately after the header, -# and they will be used before the default mirrors. +# two lines. + #[testing] #Include = /etc/pacman.d/mirrorlist [core] -# Add your preferred servers here, they will be used first Include = /etc/pacman.d/mirrorlist [extra] -# Add your preferred servers here, they will be used first Include = /etc/pacman.d/mirrorlist [community] -# Add your preferred servers here, they will be used first Include = /etc/pacman.d/mirrorlist # Unstable is disabled by default. To enable, uncomment the following -# two lines. You can add preferred servers immediately after the header, -# and they will be used before the default mirrors. +# two lines. + #[unstable] #Include = /etc/pacman.d/mirrorlist # An example of a custom package repository. See the pacman manpage for # tips on creating your own repositories. + #[custom] #Server = file:///home/custompkgs -

3 3