- Pacman-dev - lists.archlinux.org

Re: [pacman-dev] [arch-general] Package signing
by Linas 08 May '10

08 May '10

I wrote about this topic ~1 month ago. You don't need PKCis or distribute the keyrings themselves. GPG supports transitive trust. The pacman keyring would be installed by default trusting on whatever keys a pacman root signature has signed (there could also be a different master key for community developers). The basic idea here is that you are not trusting the repository, but the individuals themselves. The master key -which can be kept offline and is only used when a developer joins/part- provides a basic default (people we generally trust) but a power user could reconfigure it to not accept packages signed by Pierre, because he distrusts him :), or he can add additional trusted people (a much more likely scenario) by just adding that person key to its keyring. A developer asks you to try a beta package -> You need to do nothing since you already have its key. Fetching from personal repo from a developer -> You need to do nothing since you already have its key. That really smart guy from the forum whose advice you follow blindly. -> Add its key. Packages built by you -> Add your own key. Sysadmin -> Add its key to the computers he administers. Paranoic sysadmin -> Add its key and remove the developers one. Then resign all packages after verifying that they work. I think this is more or less what Florian Pritz proposes, althouugh I think he was thinking in terms of the user gpg, while I would go using a different, pacman gpg. Denis A. Altoé Falqueto wrote: > The central point is the management of the keyring. > > We could have a single key that all developers share, but this is > troublesome as it would require some trusted channel to exchange the > password (the first time and in the event of it changing). This > approach is what others big distributions do, but the have a different > structure than ours. > With this structure the developers never share the keys. You could even remove the master key and consider someone a developer when n developers signed him. But I find the master key method clearer. > We could have a keyring with all the trusted keys of the developers, > but this could bring some problems as the keyring should be very well > synchronized with the official one. Including a rule to upgrade the > keyring package before others, as pacman. This requires intervention > on user`s systems, so I think that it should be avoided, if possible. > There are two cases where the keyring could change due to distro changes. a) Add a new developer. b) Remove a developer. In case (a), pacman would find an unknown signature, an download [on user approval] the public key from the key servers. Case (b) is trickier, since no matter how it is done, I think there will always be a race condition on a roque developer. Luckily, the shouts of such case would make it difficult to have people holding outdated keys. I would probably make the key updating part of pacman -y to ensure that users are not using a revoked key. > My idea tries to give the best of both worlds. The management of > trusted devs would be made by a few admins and users would just be > exposed to an official Arch key that would change just a few times, if > it is not compromised. > Your approach is more insecure, since you are removing the developer key (end-to-end signing) and putting instead a generic one. With the other schema, in case a developer key was compromised, you would only need to untrust that key, not the whole master key which signs everything. Plus, your key is used continously, so it is more exposed (the signing server is compromised, it signs a fake request...). IMHO trusting the individuals also facilitates forking. A derivative distribution could share arch packages by trusting the same devvelopers. With a master key owning everything, the child distro would need to have a second key resigning everything. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

8 31

[pacman-dev] [PATCH 2/2] Switch Contributor line with Maintainer line.
by David Campbell 06 May '10

06 May '10

When someone is creating a new PKGBUILD he will most likely be the maintainer not a contributor. --- PKGBUILD-split.proto | 2 +- PKGBUILD.proto | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/PKGBUILD-split.proto b/PKGBUILD-split.proto index 53a3e8f..f549cfd 100644 --- a/PKGBUILD-split.proto +++ b/PKGBUILD-split.proto @@ -3,7 +3,7 @@ # see 'man PKGBUILD'. NOTE: Please fill out the license field for your package! # If it is unknown, then please put 'unknown'. -# Contributor: Your Name <youremail(a)domain.com> +# Maintainer: Your Name <youremail(a)domain.com> pkgname=('pkg1' 'pkg2') pkgbase="" pkgver=VERSION diff --git a/PKGBUILD.proto b/PKGBUILD.proto index 93da0fc..59b9acb 100644 --- a/PKGBUILD.proto +++ b/PKGBUILD.proto @@ -3,7 +3,7 @@ # NOTE: Please fill out the license field for your package! If it is unknown, # then please put 'unknown'. -# Contributor: Your Name <youremail(a)domain.com> +# Maintainer: Your Name <youremail(a)domain.com> pkgname=NAME pkgver=VERSION pkgrel=1 -- 1.7.0.3

2 1

[pacman-dev] [PATCH] compute package download size outside _alpm_sync_prepare
by Jonathan Conder 06 May '10

06 May '10

Thanks to Dan for pointing out a mistake I made. --- lib/libalpm/db.h | 3 ++- lib/libalpm/sync.c | 14 +++++++++++++- 2 files changed, 15 insertions(+), 2 deletions(-) diff --git a/lib/libalpm/db.h b/lib/libalpm/db.h index 9b78ad4..1851b5c 100644 --- a/lib/libalpm/db.h +++ b/lib/libalpm/db.h @@ -34,7 +34,8 @@ typedef enum _pmdbinfrq_t { INFRQ_FILES = (1 << 3), INFRQ_SCRIPTLET = (1 << 4), INFRQ_DELTAS = (1 << 5), - /* ALL should be sum of all above */ + INFRQ_DSIZE = (1 << 6), + /* ALL should be info stored in the package or database */ INFRQ_ALL = 0x3F } pmdbinfrq_t; diff --git a/lib/libalpm/sync.c b/lib/libalpm/sync.c index 4996920..13d39aa 100644 --- a/lib/libalpm/sync.c +++ b/lib/libalpm/sync.c @@ -356,6 +356,7 @@ static int compute_download_size(pmpkg_t *newpkg) off_t size = 0; if(newpkg->origin == PKG_FROM_FILE) { + newpkg->infolevel |= INFRQ_DSIZE; newpkg->download_size = 0; return(0); } @@ -392,6 +393,7 @@ static int compute_download_size(pmpkg_t *newpkg) _alpm_log(PM_LOG_DEBUG, "setting download size %jd for pkg %s\n", (intmax_t)size, alpm_pkg_get_name(newpkg)); + newpkg->infolevel |= INFRQ_DSIZE; newpkg->download_size = size; return(0); } @@ -649,6 +651,9 @@ cleanup: */ off_t SYMEXPORT alpm_pkg_download_size(pmpkg_t *newpkg) { + if(!(newpkg->infolevel & INFRQ_DSIZE)) { + compute_download_size(newpkg); + } return(newpkg->download_size); } @@ -850,7 +855,14 @@ int _alpm_sync_commit(pmtrans_t *trans, pmdb_t *db_local, alpm_list_t **data) if(files) { EVENT(trans, PM_TRANS_EVT_RETRIEVE_START, current->treename, NULL); - if(_alpm_download_files(files, current->servers, cachedir)) { + errors = _alpm_download_files(files, current->servers, cachedir); + + for(j = trans->add; j; j = j->next) { + j->infolevel &= ~INFRQ_DSIZE; + j->download_size = 0; + } + + if (errors) { _alpm_log(PM_LOG_WARNING, _("failed to retrieve some files from %s\n"), current->treename); if(pm_errno == 0) { -- 1.7.1

3 4

[pacman-dev] [PATCH] compute package download size outside _alpm_sync_prepare
by Jonathan Conder 05 May '10

05 May '10

Hi Here is a patch to resolve FS#18769. I created a new INFRQ/infolevel to describe package download size, so that it could be lazily calculated like most of the other package info. This differs slightly from most of the other infolevels, which correspond to information actually read by _alpm_db_read and alpm_pkg_load. However, I felt it better to do it this way than setting the download size to -1 for invalid, as this would make it the only field in pmpkg_t that needs to be initialized to something other than 0. It is also possible that the download size will be reported incorrectly if a package download size is read, and then that same package is downloaded with alpm_fetch_pkgurl. However, this is better than the present situation where download sizes aren't reported at all, and furthermore this is unlikely since alpm_fetch_pkgurl would probably not be used to download a package that is in the repositories. Thanks for your time, Jonathan

3 4

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.3.3-166-gd485c03
by dan＠archlinux.org 05 May '10

05 May '10

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 d485c0370f554758a28eff2cc32b2cc9c72a4604 (commit) via 9528c36452a6ca8b3718ef5895f31f46c16b17fa (commit) via f9fa822ccb0a077cf4cfa2cbbdac773187b88e20 (commit) via ac9dde072c56b9dbc3fe515dc3702c2c992f492f (commit) via 6b6eb6345bb3d9f59d967144968328c9a4b7873c (commit) via eb6af031ec83c4c0c1d7e12d1cc8f585197c025f (commit) via 4f80993933a3feee68758c27c58e9cf5d84bf20e (commit) via 2a6f3f06528149200e4fc8996c0c1e6d2bfa5480 (commit) via 8a6bbed98e11be3f619ef31e9e753d31081d11ee (commit) from 90aca75cb92c12419f2998cf54caa8ef5fe59f99 (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 d485c0370f554758a28eff2cc32b2cc9c72a4604 Author: Dan McGee <dan(a)archlinux.org> Date: Wed May 5 11:18:27 2010 -0500 Improve documentation of -k/--dbonly We had the long option wrong in some places and its behavior wasn't documented at all with regards to -U/--upgrade. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 9528c36452a6ca8b3718ef5895f31f46c16b17fa Author: Dan McGee <dan(a)archlinux.org> Date: Wed May 5 10:59:20 2010 -0500 Add two new pactests for --dbonly It caught me by surprise that: 1. These weren't being tested at all 2. The --dbonly combined with -U not only "works" but is also completely undocumented. It also has some weird behavior on install vs. upgrade that may need addressing. Add some tests which will hopefully provoke some discussion. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit f9fa822ccb0a077cf4cfa2cbbdac773187b88e20 Author: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Date: Tue Mar 16 19:40:29 2010 +0100 Remove unused 'z' option from getopt_long's optstring In addition, I permuted shortopts to make it more readable. Signed-off-by: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit ac9dde072c56b9dbc3fe515dc3702c2c992f492f Author: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Date: Wed May 20 21:46:23 2009 +0200 Introduce -D, --database The request of FS#12950 is implemented. On the backend side, I introduced a new function, alpm_db_set_pkgreason(), to modify the install reason of a package in the local database. On the front-end side, I introduced a new main operation, -D/--database, which has two options, --asdeps and --asexplicit. I documented this in pacman manual. I've created two pactests to test -D: database001.py and database002.py. Signed-off-by: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 6b6eb6345bb3d9f59d967144968328c9a4b7873c Author: Dan McGee <dan(a)archlinux.org> Date: Tue May 4 23:43:43 2010 -0500 Fix up the cross-compilation patch library lookup Don't explicitly add things to the list that might not need to be there, and get the fallback list of libraries correct. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit eb6af031ec83c4c0c1d7e12d1cc8f585197c025f Author: Dan McGee <dan(a)archlinux.org> Date: Tue May 4 23:42:01 2010 -0500 Build vercmp without needing link to libalpm Include the object file directly from the libalpm version comparison code as it is the only thing we need. This drops the dependency of vercmp on libalpm and all of the stuff we know it drags in. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 4f80993933a3feee68758c27c58e9cf5d84bf20e Author: Dan McGee <dan(a)archlinux.org> Date: Tue May 4 23:38:52 2010 -0500 Remove call to function logger It isn't really necessary here and it helps us get rid of some link pollution so we can have a slim vercmp binary. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 2a6f3f06528149200e4fc8996c0c1e6d2bfa5480 Author: Dan McGee <dan(a)archlinux.org> Date: Mon Apr 5 14:56:32 2010 -0500 Move vercmp code into a separate file This will facilitate using this object file on its own in the vercmp tool which will be done in a future commit. The net impact on the generated binaries should not be noticeable after this commit. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 8a6bbed98e11be3f619ef31e9e753d31081d11ee Author: Dan McGee <dan(a)archlinux.org> Date: Tue Apr 27 13:08:09 2010 -0500 Update PKGBUILD.proto to include a package() function Relevant after we deprecated `makepkg -R` without a package() function being present in the PKGBUILD. Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: PKGBUILD.proto | 3 + configure.ac | 4 +- doc/pacman.8.txt | 21 ++++- lib/libalpm/Makefile.am | 3 +- lib/libalpm/alpm.h | 16 ++-- lib/libalpm/db.c | 38 ++++++++ lib/libalpm/package.c | 171 ------------------------------------ lib/libalpm/version.c | 195 ++++++++++++++++++++++++++++++++++++++++++ pactest/tests/database001.py | 11 +++ pactest/tests/database002.py | 11 +++ pactest/tests/database010.py | 17 ++++ pactest/tests/database011.py | 17 ++++ pactest/tests/database012.py | 31 +++++++ src/pacman/Makefile.am | 1 + src/pacman/conf.h | 3 +- src/pacman/database.c | 90 +++++++++++++++++++ src/pacman/pacman.c | 24 ++++-- src/pacman/pacman.h | 2 + src/pacman/util.c | 2 + src/util/Makefile.am | 2 +- src/util/vercmp.c | 6 +- 21 files changed, 475 insertions(+), 193 deletions(-) create mode 100644 lib/libalpm/version.c create mode 100644 pactest/tests/database001.py create mode 100644 pactest/tests/database002.py create mode 100644 pactest/tests/database010.py create mode 100644 pactest/tests/database011.py create mode 100644 pactest/tests/database012.py create mode 100644 src/pacman/database.c hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH 1/2] Add two new pactests for --dbonly
by Dan McGee 05 May '10

05 May '10

It caught me by surprise that: 1. These weren't being tested at all 2. The --dbonly combined with -U not only "works" but is also completely undocumented. It also has some weird behavior on install vs. upgrade that may need addressing. Add some tests which will hopefully provoke some discussion. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- pactest/tests/database010.py | 17 +++++++++++++++++ pactest/tests/database011.py | 17 +++++++++++++++++ pactest/tests/database012.py | 31 +++++++++++++++++++++++++++++++ 3 files changed, 65 insertions(+), 0 deletions(-) create mode 100644 pactest/tests/database010.py create mode 100644 pactest/tests/database011.py create mode 100644 pactest/tests/database012.py diff --git a/pactest/tests/database010.py b/pactest/tests/database010.py new file mode 100644 index 0000000..662dc1b --- /dev/null +++ b/pactest/tests/database010.py @@ -0,0 +1,17 @@ +# TODO: these are labeled as database packages because they sure seem to me to +# be database-type operations. In their current implementation however they are +# calling -U and -R rather than -D. Obviously the tests will need to be updated +# if this changes. +self.description = "Remove a package with --dbonly, no files touched" + +p = pmpkg("dummy") +p.files = ["bin/dummy", + "usr/man/man1/dummy.1"] +self.addpkg2db("local", p) + +self.args = "-R --dbonly %s" % p.name + +self.addrule("PACMAN_RETCODE=0") +self.addrule("!PKG_EXIST=dummy") +for f in p.files: + self.addrule("FILE_EXIST=%s" % f) diff --git a/pactest/tests/database011.py b/pactest/tests/database011.py new file mode 100644 index 0000000..7d475d9 --- /dev/null +++ b/pactest/tests/database011.py @@ -0,0 +1,17 @@ +# TODO: these are labeled as database packages because they sure seem to me to +# be database-type operations. In their current implementation however they are +# calling -U and -R rather than -D. Obviously the tests will need to be updated +# if this changes. +self.description = "Install a package with --dbonly, no files touched" + +p = pmpkg("dummy") +p.files = ["bin/dummy", + "usr/man/man1/dummy.1"] +self.addpkg(p) + +self.args = "-U --dbonly %s" % p.filename() + +self.addrule("PACMAN_RETCODE=0") +self.addrule("PKG_EXIST=dummy") +for f in p.files: + self.addrule("!FILE_EXIST=%s" % f) diff --git a/pactest/tests/database012.py b/pactest/tests/database012.py new file mode 100644 index 0000000..a1f8698 --- /dev/null +++ b/pactest/tests/database012.py @@ -0,0 +1,31 @@ +# TODO: these are labeled as database packages because they sure seem to me to +# be database-type operations. In their current implementation however they are +# calling -U and -R rather than -D. Obviously the tests will need to be updated +# if this changes. +self.description = "Upgrade a package with --dbonly, no files touched" + +lp = pmpkg("dummy") +lp.files = ["bin/dummy", + "usr/man/man1/dummy.1"] +self.addpkg2db("local", lp) + +p = pmpkg("dummy", "2.0-1") +p.files = ["bin/dummy", + "bin/dummy2", + "usr/man/man1/dummy.1"] +self.addpkg(p) + +self.args = "-U --dbonly %s" % p.filename() + +self.addrule("PACMAN_RETCODE=0") +self.addrule("PKG_EXIST=dummy") +self.addrule("PKG_VERSION=dummy|2.0-1") +for f in lp.files: + self.addrule("!FILE_EXIST=%s" % f) +# TODO: I honestly think the above should NOT delete the original files, it +# should upgrade the DB entry without touching anything on the file system. +# E.g. this test should be the same as: +# pacman -R --dbonly dummy && pacman -U --dbonly dummy.pkg.tar.gz +#for f in lp.files: +# self.addrule("FILE_EXIST=%s" % f) +self.addrule("!FILE_EXIST=bin/dummy2") -- 1.7.1

1 1

[pacman-dev] [PATCH 1/4] Move vercmp code into a separate file
by Dan McGee 05 May '10

05 May '10

This will facilitate using this object file on its own in the vercmp tool which will be done in a future commit. The net impact on the generated binaries should not be noticeable after this commit. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- lib/libalpm/Makefile.am | 3 +- lib/libalpm/package.c | 171 ---------------------------------------- lib/libalpm/version.c | 198 +++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 200 insertions(+), 172 deletions(-) create mode 100644 lib/libalpm/version.c diff --git a/lib/libalpm/Makefile.am b/lib/libalpm/Makefile.am index 871855e..3473a73 100644 --- a/lib/libalpm/Makefile.am +++ b/lib/libalpm/Makefile.am @@ -43,7 +43,8 @@ libalpm_la_SOURCES = \ remove.h remove.c \ sync.h sync.c \ trans.h trans.c \ - util.h util.c + util.h util.c \ + version.c libalpm_la_LDFLAGS = -no-undefined -version-info $(LIB_VERSION_INFO) libalpm_la_LIBADD = $(LTLIBINTL) diff --git a/lib/libalpm/package.c b/lib/libalpm/package.c index ed6d71d..becbc60 100644 --- a/lib/libalpm/package.c +++ b/lib/libalpm/package.c @@ -604,177 +604,6 @@ alpm_list_t SYMEXPORT *alpm_pkg_compute_requiredby(pmpkg_t *pkg) return(reqs); } -/** Compare two version strings and determine which one is 'newer'. - * Returns a value comparable to the way strcmp works. Returns 1 - * if a is newer than b, 0 if a and b are the same version, or -1 - * if b is newer than a. - * - * This function has been adopted from the rpmvercmp function located - * at lib/rpmvercmp.c, and was most recently updated against rpm - * version 4.4.2.3. Small modifications have been made to make it more - * consistent with the libalpm coding style. - * - * Keep in mind that the pkgrel is only compared if it is available - * on both versions handed to this function. For example, comparing - * 1.5-1 and 1.5 will yield 0; comparing 1.5-1 and 1.5-2 will yield - * -1 as expected. This is mainly for supporting versioned dependencies - * that do not include the pkgrel. - */ -int SYMEXPORT alpm_pkg_vercmp(const char *a, const char *b) -{ - char oldch1, oldch2; - char *str1, *str2; - char *ptr1, *ptr2; - char *one, *two; - int rc; - int isnum; - int ret = 0; - - ALPM_LOG_FUNC; - - /* libalpm added code. ensure our strings are not null */ - if(!a) { - if(!b) return(0); - return(-1); - } - if(!b) return(1); - - /* easy comparison to see if versions are identical */ - if(strcmp(a, b) == 0) return(0); - - str1 = strdup(a); - str2 = strdup(b); - - one = str1; - two = str2; - - /* loop through each version segment of str1 and str2 and compare them */ - while(*one && *two) { - while(*one && !isalnum((int)*one)) one++; - while(*two && !isalnum((int)*two)) two++; - - /* If we ran to the end of either, we are finished with the loop */ - if(!(*one && *two)) break; - - ptr1 = one; - ptr2 = two; - - /* grab first completely alpha or completely numeric segment */ - /* leave one and two pointing to the start of the alpha or numeric */ - /* segment and walk ptr1 and ptr2 to end of segment */ - if(isdigit((int)*ptr1)) { - while(*ptr1 && isdigit((int)*ptr1)) ptr1++; - while(*ptr2 && isdigit((int)*ptr2)) ptr2++; - isnum = 1; - } else { - while(*ptr1 && isalpha((int)*ptr1)) ptr1++; - while(*ptr2 && isalpha((int)*ptr2)) ptr2++; - isnum = 0; - } - - /* save character at the end of the alpha or numeric segment */ - /* so that they can be restored after the comparison */ - oldch1 = *ptr1; - *ptr1 = '\0'; - oldch2 = *ptr2; - *ptr2 = '\0'; - - /* this cannot happen, as we previously tested to make sure that */ - /* the first string has a non-null segment */ - if (one == ptr1) { - ret = -1; /* arbitrary */ - goto cleanup; - } - - /* take care of the case where the two version segments are */ - /* different types: one numeric, the other alpha (i.e. empty) */ - /* numeric segments are always newer than alpha segments */ - /* XXX See patch #60884 (and details) from bugzilla #50977. */ - if (two == ptr2) { - ret = isnum ? 1 : -1; - goto cleanup; - } - - if (isnum) { - /* this used to be done by converting the digit segments */ - /* to ints using atoi() - it's changed because long */ - /* digit segments can overflow an int - this should fix that. */ - - /* throw away any leading zeros - it's a number, right? */ - while (*one == '0') one++; - while (*two == '0') two++; - - /* whichever number has more digits wins */ - if (strlen(one) > strlen(two)) { - ret = 1; - goto cleanup; - } - if (strlen(two) > strlen(one)) { - ret = -1; - goto cleanup; - } - } - - /* strcmp will return which one is greater - even if the two */ - /* segments are alpha or if they are numeric. don't return */ - /* if they are equal because there might be more segments to */ - /* compare */ - rc = strcmp(one, two); - if (rc) { - ret = rc < 1 ? -1 : 1; - goto cleanup; - } - - /* restore character that was replaced by null above */ - *ptr1 = oldch1; - one = ptr1; - *ptr2 = oldch2; - two = ptr2; - - /* libalpm added code. check if version strings have hit the pkgrel - * portion. depending on which strings have hit, take correct action. - * this is all based on the premise that we only have one dash in - * the version string, and it separates pkgver from pkgrel. */ - if(*ptr1 == '-' && *ptr2 == '-') { - /* no-op, continue comparing since we are equivalent throughout */ - } else if(*ptr1 == '-') { - /* ptr1 has hit the pkgrel and ptr2 has not. continue version - * comparison after stripping the pkgrel from ptr1. */ - *ptr1 = '\0'; - } else if(*ptr2 == '-') { - /* ptr2 has hit the pkgrel and ptr1 has not. continue version - * comparison after stripping the pkgrel from ptr2. */ - *ptr2 = '\0'; - } - } - - /* this catches the case where all numeric and alpha segments have */ - /* compared identically but the segment separating characters were */ - /* different */ - if ((!*one) && (!*two)) { - ret = 0; - goto cleanup; - } - - /* the final showdown. we never want a remaining alpha string to - * beat an empty string. the logic is a bit weird, but: - * - if one is empty and two is not an alpha, two is newer. - * - if one is an alpha, two is newer. - * - otherwise one is newer. - * */ - if ( ( !*one && !isalpha((int)*two) ) - || isalpha((int)*one) ) { - ret = -1; - } else { - ret = 1; - } - -cleanup: - free(str1); - free(str2); - return(ret); -} - /** @} */ pmpkg_t *_alpm_pkg_new(void) diff --git a/lib/libalpm/version.c b/lib/libalpm/version.c new file mode 100644 index 0000000..1c838f4 --- /dev/null +++ b/lib/libalpm/version.c @@ -0,0 +1,198 @@ +/* + * Copyright (c) 2006-2010 Pacman Development Team <pacman-dev(a)archlinux.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "config.h" + +#include <string.h> +#include <ctype.h> + +/* libalpm */ +#include "log.h" +#include "util.h" + +/** Compare two version strings and determine which one is 'newer'. + * Returns a value comparable to the way strcmp works. Returns 1 + * if a is newer than b, 0 if a and b are the same version, or -1 + * if b is newer than a. + * + * This function has been adopted from the rpmvercmp function located + * at lib/rpmvercmp.c, and was most recently updated against rpm + * version 4.4.2.3. Small modifications have been made to make it more + * consistent with the libalpm coding style. + * + * Keep in mind that the pkgrel is only compared if it is available + * on both versions handed to this function. For example, comparing + * 1.5-1 and 1.5 will yield 0; comparing 1.5-1 and 1.5-2 will yield + * -1 as expected. This is mainly for supporting versioned dependencies + * that do not include the pkgrel. + */ +int SYMEXPORT alpm_pkg_vercmp(const char *a, const char *b) +{ + char oldch1, oldch2; + char *str1, *str2; + char *ptr1, *ptr2; + char *one, *two; + int rc; + int isnum; + int ret = 0; + + ALPM_LOG_FUNC; + + /* libalpm added code. ensure our strings are not null */ + if(!a) { + if(!b) return(0); + return(-1); + } + if(!b) return(1); + + /* easy comparison to see if versions are identical */ + if(strcmp(a, b) == 0) return(0); + + str1 = strdup(a); + str2 = strdup(b); + + one = str1; + two = str2; + + /* loop through each version segment of str1 and str2 and compare them */ + while(*one && *two) { + while(*one && !isalnum((int)*one)) one++; + while(*two && !isalnum((int)*two)) two++; + + /* If we ran to the end of either, we are finished with the loop */ + if(!(*one && *two)) break; + + ptr1 = one; + ptr2 = two; + + /* grab first completely alpha or completely numeric segment */ + /* leave one and two pointing to the start of the alpha or numeric */ + /* segment and walk ptr1 and ptr2 to end of segment */ + if(isdigit((int)*ptr1)) { + while(*ptr1 && isdigit((int)*ptr1)) ptr1++; + while(*ptr2 && isdigit((int)*ptr2)) ptr2++; + isnum = 1; + } else { + while(*ptr1 && isalpha((int)*ptr1)) ptr1++; + while(*ptr2 && isalpha((int)*ptr2)) ptr2++; + isnum = 0; + } + + /* save character at the end of the alpha or numeric segment */ + /* so that they can be restored after the comparison */ + oldch1 = *ptr1; + *ptr1 = '\0'; + oldch2 = *ptr2; + *ptr2 = '\0'; + + /* this cannot happen, as we previously tested to make sure that */ + /* the first string has a non-null segment */ + if (one == ptr1) { + ret = -1; /* arbitrary */ + goto cleanup; + } + + /* take care of the case where the two version segments are */ + /* different types: one numeric, the other alpha (i.e. empty) */ + /* numeric segments are always newer than alpha segments */ + /* XXX See patch #60884 (and details) from bugzilla #50977. */ + if (two == ptr2) { + ret = isnum ? 1 : -1; + goto cleanup; + } + + if (isnum) { + /* this used to be done by converting the digit segments */ + /* to ints using atoi() - it's changed because long */ + /* digit segments can overflow an int - this should fix that. */ + + /* throw away any leading zeros - it's a number, right? */ + while (*one == '0') one++; + while (*two == '0') two++; + + /* whichever number has more digits wins */ + if (strlen(one) > strlen(two)) { + ret = 1; + goto cleanup; + } + if (strlen(two) > strlen(one)) { + ret = -1; + goto cleanup; + } + } + + /* strcmp will return which one is greater - even if the two */ + /* segments are alpha or if they are numeric. don't return */ + /* if they are equal because there might be more segments to */ + /* compare */ + rc = strcmp(one, two); + if (rc) { + ret = rc < 1 ? -1 : 1; + goto cleanup; + } + + /* restore character that was replaced by null above */ + *ptr1 = oldch1; + one = ptr1; + *ptr2 = oldch2; + two = ptr2; + + /* libalpm added code. check if version strings have hit the pkgrel + * portion. depending on which strings have hit, take correct action. + * this is all based on the premise that we only have one dash in + * the version string, and it separates pkgver from pkgrel. */ + if(*ptr1 == '-' && *ptr2 == '-') { + /* no-op, continue comparing since we are equivalent throughout */ + } else if(*ptr1 == '-') { + /* ptr1 has hit the pkgrel and ptr2 has not. continue version + * comparison after stripping the pkgrel from ptr1. */ + *ptr1 = '\0'; + } else if(*ptr2 == '-') { + /* ptr2 has hit the pkgrel and ptr1 has not. continue version + * comparison after stripping the pkgrel from ptr2. */ + *ptr2 = '\0'; + } + } + + /* this catches the case where all numeric and alpha segments have */ + /* compared identically but the segment separating characters were */ + /* different */ + if ((!*one) && (!*two)) { + ret = 0; + goto cleanup; + } + + /* the final showdown. we never want a remaining alpha string to + * beat an empty string. the logic is a bit weird, but: + * - if one is empty and two is not an alpha, two is newer. + * - if one is an alpha, two is newer. + * - otherwise one is newer. + * */ + if ( ( !*one && !isalpha((int)*two) ) + || isalpha((int)*one) ) { + ret = -1; + } else { + ret = 1; + } + +cleanup: + free(str1); + free(str2); + return(ret); +} + +/* vim: set ts=2 sw=2 noet: */ -- 1.7.1

3 6

[pacman-dev] An interesting error
by Roman Kyrylych 02 May '10

02 May '10

I've just got this when updating my linode server: $ sudo pacman -Syyu :: Synchronizing package databases... error: failed retrieving file 'testing.db.tar.gz' from ftp.archlinux.org : No address record testing 13.4K 733.6K/s 00:00:00 [#####################] 100% error: failed retrieving file 'core.db.tar.gz' from ftp.archlinux.org : No address record core 36.1K 1303.4K/s 00:00:00 [#####################] 100% error: failed retrieving file 'extra.db.tar.gz' from ftp.archlinux.org : No address record extra 449.8K 4.7M/s 00:00:00 [#####################] 100% error: failed retrieving file 'community-testing.db.tar.gz' from ftp.archlinux.org : No address record community-testing 1.2K 20.1M/s 00:00:00 [#####################] 100% error: failed retrieving file 'community.db.tar.gz' from ftp.archlinux.org : No address record community 373.4K 4.6M/s 00:00:00 [#####################] 100% :: Starting full system upgrade... resolving dependencies... looking for inter-conflicts... Targets (4): kernel26-firmware-2.6.33.3-2 kernel26-2.6.33.3-2 psmisc-22.11-1 tzdata-2010i-1 Total Download Size: 32.46 MB Total Installed Size: 114.00 MB Proceed with installation? [Y/n] n [roman@codeplace ~]$ sudo pacman -Su :: Starting full system upgrade... resolving dependencies... looking for inter-conflicts... Targets (4): kernel26-firmware-2.6.33.3-2 kernel26-2.6.33.3-2 psmisc-22.11-1 tzdata-2010i-1 Total Download Size: 32.46 MB Total Installed Size: 114.00 MB Proceed with installation? [Y/n] :: Retrieving packages from testing... error: failed retrieving file 'kernel26-firmware-2.6.33.3-2-x86_64.pkg.tar.gz' from ftp.archlinux.org : File unavailable (e.g., file not found, no access) kernel26-firmware-2... 1048.8K 664.4K/s 00:00:02 [#####################] 100% error: failed retrieving file 'kernel26-2.6.33.3-2-x86_64.pkg.tar.gz' from ftp.archlinux.org : File unavailable (e.g., file not found, no access) kernel26-2.6.33.3-2... 31.2M 5.1M/s 00:00:06 [#####################] 100% psmisc-22.11-1-x86_... 76.9K 153.5K/s 00:00:01 [#####################] 100% tzdata-2010i-1-x86_... 178.7K 90.6K/s 00:00:02 [#####################] 100% Can anyone explain what exactly happens here? -- Roman Kyrylych (Роман Кирилич)

1 1

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.3.3-157-g90aca75
by dan＠archlinux.org 26 Apr '10

26 Apr '10

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 90aca75cb92c12419f2998cf54caa8ef5fe59f99 (commit) via 753599b5046392477449ed550cd99c41631145ed (commit) via 652762488a1fbe70b073588dc5f88412ff952fb7 (commit) via 6995aed9aeae1e95d9de3b2f539f904dbde4e165 (commit) via 7608dd74d73b66595ad34c861491fc86be335e8c (commit) via 88254d762d68f93b2be04e686abbc126e2afcf11 (commit) via 7965345d435870a47073fffafa2a429744aa2deb (commit) from 21abae98cbd382d8fab4427109776d983cd07c9f (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 90aca75cb92c12419f2998cf54caa8ef5fe59f99 Author: Allan McRae <allan(a)archlinux.org> Date: Mon Mar 29 15:17:22 2010 +1000 makepkg: BSD find compatibility fix Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 753599b5046392477449ed550cd99c41631145ed Author: Christophe Chapuis <chris.chapuis(a)gmail.com> Date: Tue Apr 20 16:21:00 2010 +1000 contrib/pactree: generate reverse dependency trees Add an option to show the tree of packages which depend on a given package Signed-off-by: Allan McRae <allan(a)archlinux.org> commit 652762488a1fbe70b073588dc5f88412ff952fb7 Author: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Date: Sat Apr 24 16:48:14 2010 +0200 New pactests for the phonon/qt issue Original-work-by: Xavier Chantry <shiningxc(a)gmail.com> Signed-off-by: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 6995aed9aeae1e95d9de3b2f539f904dbde4e165 Author: Allan McRae <allan(a)archlinux.org> Date: Wed Apr 14 16:22:43 2010 +1000 makepkg: deprecate repackaging without a package function File permissions are not guaranteed to stay the same on exit from fakeroot, so repackaging may result in files with different permissions. This is avoided when using a package() function (or split packages) as the packaging step is rerun. Signed-off-by: Allan McRae <allan(a)archlinux.org> [Dan: touched up message for translation purposes] Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 7608dd74d73b66595ad34c861491fc86be335e8c Author: Serge Ziryukin <ftrvxmtrx(a)gmail.com> Date: Sun Apr 25 19:48:03 2010 +0300 check for valid optarg before using strdup Signed-off-by: Serge Ziryukin <ftrvxmtrx(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 88254d762d68f93b2be04e686abbc126e2afcf11 Author: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Date: Mon Mar 15 23:54:00 2010 +0100 Show --print and --print-format options with -Rh, -Sh and -Uh only http://mailman.archlinux.org/pipermail/pacman-dev/2010-March/010519.html Signed-off-by: Nagy Gabor <ngaba(a)bibl.u-szeged.hu> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 7965345d435870a47073fffafa2a429744aa2deb Author: Xavier Chantry <shiningxc(a)gmail.com> Date: Sat Nov 14 20:26:37 2009 +0100 More consistent printing of off_t and time_t time_t : %ld off_t : %jd and cast to intmax_t Signed-off-by: Xavier Chantry <shiningxc(a)gmail.com> Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: contrib/pactree | 31 ++++++++++++++++++++++++++----- lib/libalpm/be_files.c | 6 +++--- lib/libalpm/delta.c | 2 +- lib/libalpm/dload.c | 6 +++--- pactest/tests/sync405.py | 26 ++++++++++++++++++++++++++ pactest/tests/sync406.py | 31 +++++++++++++++++++++++++++++++ scripts/makepkg.sh.in | 22 ++++++++++++++++------ src/pacman/pacman.c | 13 ++++++++++--- 8 files changed, 116 insertions(+), 21 deletions(-) create mode 100644 pactest/tests/sync405.py create mode 100644 pactest/tests/sync406.py hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH] makepkg: handle multiple install and changelog files
by Allan McRae 26 Apr '10

26 Apr '10

The presence of all install and changelog files (multiple files may be used with package splitting) is checked for in check_sanity(). All install and changelog files are copied to the source location when using --source. The check for install and changelog file presence is removed in create_srcpackage() as this is redundant to the checks performed in check_sanity(). Moved install and changelog handling in create_srcpackage() to after source array files, as this is more logical and readily allows for the following. A check is made when creating a source package that a symlink to an install file has not already been added. This can occur if the install file is used multiple times or if it is listed in the source array. Fixes FS#18831, FS#18394 and partially fixes FS#16004 Signed-off-by: Allan McRae <allan(a)archlinux.org> --- I could split these into a set of smaller changes but they overlap to such a great extent it seems redundant. Also, the patch remains fairly small. scripts/makepkg.sh.in | 71 ++++++++++++++++++++++++++++++++---------------- 1 files changed, 47 insertions(+), 24 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index b306620..8ae41ea 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -1095,24 +1095,6 @@ create_srcpackage() { msg2 "$(gettext "Adding %s...")" "$BUILDSCRIPT" ln -s "${BUILDFILE}" "${srclinks}/${pkgbase}/${BUILDSCRIPT}" - if [[ -n $install ]]; then - if [[ -f $install ]]; then - msg2 "$(gettext "Adding install script...")" - ln -s "${startdir}/$install" "${srclinks}/${pkgbase}/" - else - error "$(gettext "Install scriptlet (%s) does not exist.")" "$install" - fi - fi - - if [[ -n $changelog ]]; then - if [[ -f $changelog ]]; then - msg2 "$(gettext "Adding package changelog...")" - ln -s "${startdir}/$changelog" "${srclinks}/${pkgbase}/" - else - error "$(gettext "Changelog file (%s) does not exist.")" "$changelog" - fi - fi - local netfile for netfile in "${source[@]}"; do local file=$(get_filename "$netfile") @@ -1125,6 +1107,32 @@ create_srcpackage() { fi done + local install_files=( $(grep "^[[:space:]]*install=" "$BUILDSCRIPT" | sed "s/install=//") ) + if [[ -n $install_files ]]; then + local file + for file in ${install_files[@]}; do + # evaluate any bash variables used + eval file=${file} + if [[ ! -f "${srclinks}/${pkgbase}/$file" ]]; then + msg2 "$(gettext "Adding install script (%s)...")" "$file" + ln -s "${startdir}/$file" "${srclinks}/${pkgbase}/" + fi + done + fi + + local changelog_files=( $(grep "^[[:space:]]*changelog=" "$BUILDSCRIPT" | sed "s/changelog=//") ) + if [[ -n $changelog_files ]]; then + local file + for file in ${changelog_files[@]}; do + # evaluate any bash variables used + eval file=${file} + if [[ ! -f "${srclinks}/${pkgbase}/$file" ]]; then + msg2 "$(gettext "Adding package changelog (%s)...")" "$file" + ln -s "${startdir}/$file" "${srclinks}/${pkgbase}/" + fi + done + fi + local TAR_OPT case "$SRCEXT" in *tar.gz) TAR_OPT="z" ;; @@ -1242,14 +1250,29 @@ check_sanity() { fi done - if [[ $install && ! -f $install ]]; then - error "$(gettext "Install scriptlet (%s) does not exist.")" "$install" - return 1 + local install_files=( $(grep "^[[:space:]]*install=" "$BUILDSCRIPT" | sed "s/install=//") ) + if [[ -n $install_files ]]; then + local file + for file in ${install_files[@]}; do + # evaluate any bash variables used + eval file=${file} + if [[ ! -f $file ]]; then + error "$(gettext "Install scriptlet (%s) does not exist.")" "$file" + return 1 + fi + done fi - if [[ -n $changelog && ! -f $changelog ]]; then - error "$(gettext "Changelog file (%s) does not exist.")" "$changelog" - return 1 + local changelog_files=( $(grep "^[[:space:]]*changelog=" "$BUILDSCRIPT" | sed "s/changelog=//") ) + if [[ -n $changelog_files ]]; then + for file in ${changelog_files[@]}; do + # evaluate any bash variables used + eval file=${file} + if [[ ! -f $file ]]; then + error "$(gettext "Changelog file (%s) does not exist.")" "$file" + return 1 + fi + done fi local valid_options=1 -- 1.7.0.6

1 0