- Pacman-dev - lists.archlinux.org

[pacman-dev] [WIP] Fix for overzealous package removal with unmet dependencies
by Allan McRae 16 Sep '12

16 Sep '12

Here goes a WIP patch for FS#30649. It fixes one the sync140.py testcase just sent to the mailing list, but not sync141.py - it works only when the root cause of the excess dependency removal is first in the upgraded package list. If anyone has an idea for how to get it fully fixed, I would appreciate the help. Allan diff --git a/lib/libalpm/sync.c b/lib/libalpm/sync.c index 841e5d6..65a8f52 100644 --- a/lib/libalpm/sync.c +++ b/lib/libalpm/sync.c @@ -406,14 +406,29 @@ int _alpm_sync_prepare(alpm_handle_t *handle, alpm_list_t **data) /* Resolve packages in the transaction one at a time, in addition building up a list of packages which could not be resolved. */ - for(i = trans->add; i; i = i->next) { + + i = trans->add; + while(i != NULL) { alpm_pkg_t *pkg = i->data; if(_alpm_resolvedeps(handle, localpkgs, pkg, trans->add, &resolved, remove, data) == -1) { unresolvable = alpm_list_add(unresolvable, pkg); + + /* remove the unresolvable package from the upgrade + list and recalculate localpkgs to allow the a local + version to satisfy future deps */ + + i = i->next; + trans->add = alpm_list_remove(trans->add, pkg, + _alpm_pkg_cmp, NULL); + localpkgs = alpm_list_diff( + _alpm_db_get_pkgcache(handle->db_local), + trans->add, _alpm_pkg_cmp); + } else { + /* [resolved] now additionally contains [pkg] and all of its + dependencies not already on the list */ + i = i->next; } - /* Else, [resolved] now additionally contains [pkg] and all of its - dependencies not already on the list */ } alpm_list_free(localpkgs); alpm_list_free(remove);

1 0

[pacman-dev] [PATCH 2/2] pactest: variant of package removal due to unmet dependencies
by Allan McRae 16 Sep '12

16 Sep '12

Duplicate of sync140.py but with the alphabetical ordering of the packages changed. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- test/pacman/tests/sync141.py | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) create mode 100644 test/pacman/tests/sync141.py diff --git a/test/pacman/tests/sync141.py b/test/pacman/tests/sync141.py new file mode 100644 index 0000000..b2c996e --- /dev/null +++ b/test/pacman/tests/sync141.py @@ -0,0 +1,28 @@ +self.description = "Sysupgrade with ignored package prevent other upgrade" + +lp1 = pmpkg("c_glibc", "1.0-1") +lp2 = pmpkg("b_gcc-libs", "1.0-1") +lp2.depends = ["c_glibc>=1.0-1"] +lp3 = pmpkg("a_pcre", "1.0-1") +lp3.depends = ["b_gcc-libs"] + +for p in lp1, lp2, lp3: + self.addpkg2db("local", p) + +sp1 = pmpkg("c_glibc", "1.0-2") +sp2 = pmpkg("b_gcc-libs", "1.0-2") +sp2.depends = ["c_glibc>=1.0-2"] +sp3 = pmpkg("a_pcre", "1.0-2") +sp3.depends = ["b_gcc-libs"] + +for p in sp1, sp2, sp3: + self.addpkg2db("sync", p) + +self.args = "-Su --ignore %s --ask=32" % sp1.name + +self.addrule("PACMAN_RETCODE=0") +self.addrule("PKG_VERSION=c_glibc|1.0-1") +self.addrule("PKG_VERSION=b_gcc-libs|1.0-1") +self.addrule("PKG_VERSION=a_pcre|1.0-2") + +self.expectfailure = True -- 1.7.12

1 0

[pacman-dev] [PATCH 1/2] pactest: removal of packages with unmet dependencies
by Allan McRae 16 Sep '12

16 Sep '12

Provide a package for removing packages with unmet dependencies. Currently pacman removes too many packages from the transaction (FS#30649). Signed-off-by: Allan McRae <allan(a)archlinux.org> --- test/pacman/tests/sync140.py | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) create mode 100644 test/pacman/tests/sync140.py diff --git a/test/pacman/tests/sync140.py b/test/pacman/tests/sync140.py new file mode 100644 index 0000000..d20da98 --- /dev/null +++ b/test/pacman/tests/sync140.py @@ -0,0 +1,28 @@ +self.description = "Sysupgrade with ignored package prevent other upgrade" + +lp1 = pmpkg("glibc", "1.0-1") +lp2 = pmpkg("gcc-libs", "1.0-1") +lp2.depends = ["glibc>=1.0-1"] +lp3 = pmpkg("pcre", "1.0-1") +lp3.depends = ["gcc-libs"] + +for p in lp1, lp2, lp3: + self.addpkg2db("local", p) + +sp1 = pmpkg("glibc", "1.0-2") +sp2 = pmpkg("gcc-libs", "1.0-2") +sp2.depends = ["glibc>=1.0-2"] +sp3 = pmpkg("pcre", "1.0-2") +sp3.depends = ["gcc-libs"] + +for p in sp1, sp2, sp3: + self.addpkg2db("sync", p) + +self.args = "-Su --ignore %s --ask=32" % sp1.name + +self.addrule("PACMAN_RETCODE=0") +self.addrule("PKG_VERSION=glibc|1.0-1") +self.addrule("PKG_VERSION=gcc-libs|1.0-1") +self.addrule("PKG_VERSION=pcre|1.0-2") + +self.expectfailure = True -- 1.7.12

1 0

[pacman-dev] [PATCH 1/2] makepkg: Bazaar VCS support.
by Gary van der Merwe 13 Sep '12

13 Sep '12

This patch adds Bazaar VCS support to makepkg. Fragments can be any bzr revision spec. eg: http://example.com/project/#revno:62 or just http://example.com/project/#62 For more info, see `bzr help revisionspec`. Writing a pkgver func is easy with `bzr version-info`, eg: pkgver() { cd $srcdir/$pkgname bzr version-info --custom --template="{revno}\n" } --- doc/PKGBUILD.5.txt | 5 ++++- scripts/makepkg.sh.in | 60 ++++++++++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 61 insertions(+), 4 deletions(-) diff --git a/doc/PKGBUILD.5.txt b/doc/PKGBUILD.5.txt index 47f9e88..c0f96c2 100644 --- a/doc/PKGBUILD.5.txt +++ b/doc/PKGBUILD.5.txt @@ -405,7 +405,7 @@ Using VCS Sources[[VCS]] ------------------------ Building a developmental version of a package using sources from a version control system (VCS) is enabled by specifying the source in the form -`source=('folder::url#fragment')`. Currently makepkg supports the `git`, `hg` and +`source=('folder::url#fragment')`. Currently makepkg supports the `git`, `hg`, `bzr` and `svn` protocols. The source URL is divided into three components: @@ -433,6 +433,9 @@ The source URL is divided into three components: *hg*;; branch, revision, tag + *bzr*;; + revisionspec (see `bzr help revisionspec` for details) + *svn*;; revision diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index dd4066b..3460824 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -200,7 +200,7 @@ get_filepath() { local proto="$(get_protocol "$1")" case $proto in - git*|hg*|svn*) + git*|hg*|svn*|bzr*) if [[ -d "$startdir/$file" ]]; then file="$startdir/$file" elif [[ -d "$SRCDEST/$file" ]]; then @@ -236,7 +236,7 @@ get_filename() { local proto=$(get_protocol "$netfile") case $proto in - git*|hg*|svn*) + git*|hg*|svn*|bzr*) filename=${netfile##*/} filename=${filename%%#*} # fall-through @@ -572,6 +572,57 @@ download_svn() { popd &>/dev/null } +download_bzr() { + local netfile=$1 + + local url=$(get_url "$netfile") + url=${url##*bzr+} + url=${url%%#*} + + local fragment=${netfile##*#} + if [[ $fragment = "$netfile" ]]; then + unset fragment + local revisionspec="-r-1" + local displaylocation="$url" + else + local revisionspec="-r$fragment" + local displaylocation="$url -r $fragment" + fi + + local dir=$(get_filepath "$netfile") + [[ -z "$dir" ]] && dir="$SRCDEST/$(get_filename "$netfile")" + + if [[ ! -d "$dir" ]]; then + msg2 "$(gettext "Branching %s ...")" "${displaylocation}" + if ! bzr branch "$url" "$dir" $revisionspec --no-tree; then + error "$(gettext "Failure while branching %s")" "${displaylocation}" + plain "$(gettext "Aborting...")" + exit 1 + fi + elif (( ! HOLDVER )); then + msg2 "$(gettext "Pulling %s ...")" "${displaylocation}" + cd_safe "$dir" + if ! bzr pull "$url" $revisionspec --overwrite; then + # only warn on failure to allow offline builds + warning "$(gettext "Failure while pulling %s")" "${displaylocation}" + fi + fi + + msg2 "$(gettext "Creating working copy of %s %s repo...")" "${dir}" "bzr" + pushd "$srcdir" &>/dev/null + rm -rf "${dir##*/}" + + if ! bzr checkout "$dir" --lightweight; then + error "$(gettext "Failure while creating working copy of %s %s repo")" "${dir}" "bzr" + plain "$(gettext "Aborting...")" + exit 1 + fi + + cd_safe "${dir##*/}" + + popd &>/dev/null +} + download_sources() { msg "$(gettext "Retrieving Sources...")" @@ -602,6 +653,9 @@ download_sources() { svn*) (( GET_VCS )) && download_svn "$netfile" ;; + bzr*) + (( GET_VCS )) && download_bzr "$netfile" + ;; *) error "$(gettext "Unknown download protocol: %s")" "$proto" plain "$(gettext "Aborting...")" @@ -974,7 +1028,7 @@ generate_checksums() { proto="$(get_protocol "$netfile")" case $proto in - git*|hg*|svn*) + git*|hg*|svn*|bzr*) sum="SKIP" ;; *) -- 1.7.11.4

2 10

[pacman-dev] [PATCH] Remove leading "./" from filenames in mtree file
by Allan McRae 09 Sep '12

09 Sep '12

A properly formatted mtree file may stick a leading "./" in front of file names in the root directory. Strip this if present. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- Sending for completeness. The full mtree patchset can be seen at https://patchwork.archlinux.org/user/bundle/8/?state=* The whole patchset requires libarchives mtree generation to be fixed using this patch: https://github.com/allanbrokeit/libarchive/commit/32675a90 src/pacman/check.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/src/pacman/check.c b/src/pacman/check.c index d325a99..2ccf51d 100644 --- a/src/pacman/check.c +++ b/src/pacman/check.c @@ -242,6 +242,11 @@ int check_pkg_full(alpm_pkg_t *pkg) mode_t type; size_t file_errors = 0; + /* strip leading "./" from path entries */ + if(path[0] == '.' && path[1] == '/') { + path += 2; + } + if(strcmp(path, ".INSTALL") == 0) { char filename[PATH_MAX]; snprintf(filename, PATH_MAX, "%slocal/%s-%s/install", -- 1.7.12

1 0

[pacman-dev] [PATCH] makepkg: fix order of --help output
by Allan McRae 09 Sep '12

09 Sep '12

Small alphabeticalization issue in options passable to pacman. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- scripts/makepkg.sh.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index aed6b8b..6de7b4f 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -2261,8 +2261,8 @@ usage() { echo printf -- "$(gettext "These options can be passed to %s:")\n" "pacman" echo - printf -- "$(gettext " --noconfirm Do not ask for confirmation when resolving dependencies")\n" printf -- "$(gettext " --asdeps Install packages as non-explicitly installed")\n" + printf -- "$(gettext " --noconfirm Do not ask for confirmation when resolving dependencies")\n" printf -- "$(gettext " --noprogressbar Do not show a progress bar when downloading files")\n" echo printf -- "$(gettext "If %s is not specified, %s will look for '%s'")\n" "-p" "makepkg" "$BUILDSCRIPT" -- 1.7.12

1 0

[pacman-dev] [PATCH] Allow querying directory ownership
by Andrew Gregory 08 Sep '12

08 Sep '12

The restriction of not checking the ownership of a directory is unnecessary given that all the package filelists contain this information. Remove this restriction, with the expectation that you might get multiple packages returned for a given directory. Additionally attempt to minimise the number of files getting through to the slow realpath call. Original-work-by: Dan McGee <dan(a)archlinux.org> Original-work-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- Applies on top of my other -Qo patches queued in Allan's working-split/query-owner branch. src/pacman/query.c | 61 +++++++++++++++++++++++++++++++++++++++++------------- 1 file changed, 47 insertions(+), 14 deletions(-) diff --git a/src/pacman/query.c b/src/pacman/query.c index 92219e8..06505eb 100644 --- a/src/pacman/query.c +++ b/src/pacman/query.c @@ -97,6 +97,7 @@ static int query_fileowner(alpm_list_t *targets) size_t rootlen; alpm_list_t *t; alpm_db_t *db_local; + alpm_list_t *packages; /* This code is here for safety only */ if(targets == NULL) { @@ -129,13 +130,14 @@ static int query_fileowner(alpm_list_t *targets) } db_local = alpm_get_localdb(config->handle); + packages = alpm_db_get_pkgcache(db_local); for(t = targets; t; t = alpm_list_next(t)) { char *filename = NULL, *dname = NULL, *rpath = NULL; const char *bname; struct stat buf; alpm_list_t *i; - size_t len; + size_t len, is_dir, bname_len, pbname_len; int found = 0; if((filename = strdup(t->data)) == NULL) { @@ -163,15 +165,20 @@ static int query_fileowner(alpm_list_t *targets) } } - if(S_ISDIR(buf.st_mode)) { - pm_printf(ALPM_LOG_ERROR, - _("cannot determine ownership of directory '%s'\n"), filename); - goto targcleanup; + if((is_dir = S_ISDIR(buf.st_mode))) { + /* the entire filename is safe to resolve if we know it points to a dir, + * and it's necessary in case it ends in . or .. */ + dname = realpath(filename, NULL); + bname = mbasename(dname); + rpath = mdirname(dname); + } else { + bname = mbasename(filename); + dname = mdirname(filename); + rpath = realpath(dname, NULL); } - bname = mbasename(filename); - dname = mdirname(filename); - rpath = realpath(dname, NULL); + bname_len = strlen(bname); + pbname_len = bname_len + (is_dir ? 1 : 0); if(!dname || !rpath) { pm_printf(ALPM_LOG_ERROR, _("cannot determine real path for '%s': %s\n"), @@ -179,7 +186,7 @@ static int query_fileowner(alpm_list_t *targets) goto targcleanup; } - for(i = alpm_db_get_pkgcache(db_local); i && !found; i = alpm_list_next(i)) { + for(i = packages; i && (!found || is_dir); i = alpm_list_next(i)) { alpm_pkg_t *info = i->data; alpm_filelist_t *filelist = alpm_pkg_get_files(info); size_t j; @@ -188,22 +195,48 @@ static int query_fileowner(alpm_list_t *targets) const alpm_file_t *file = filelist->files + j; char *ppath, *pdname; const char *pkgfile = file->name; + size_t pkgfile_len = strlen(pkgfile); + + /* make sure pkgfile and target are of the same type */ + if(!is_dir != !(pkgfile[pkgfile_len - 1] == '/')) { + continue; + } + + /* make sure pkgfile is long enough */ + if(pkgfile_len < pbname_len) { + continue; + } - /* avoid the costly realpath usage if the basenames don't match */ - if(strcmp(mbasename(pkgfile), bname) != 0) { + /* make sure pbname_len takes us to the start of a path component */ + if(pbname_len != pkgfile_len && pkgfile[pkgfile_len - pbname_len - 1] != '/') { + continue; + } + + /* compare basename with bname */ + if(strncmp(pkgfile + pkgfile_len - pbname_len, bname, bname_len) != 0) { continue; } /* concatenate our file and the root path */ - if(rootlen + 1 + strlen(pkgfile) > PATH_MAX) { + if(rootlen + 1 + pkgfile_len > PATH_MAX) { path[rootlen] = '\0'; /* reset path for error message */ pm_printf(ALPM_LOG_ERROR, _("path too long: %s%s\n"), path, pkgfile); continue; } strcpy(path + rootlen, pkgfile); - pdname = mdirname(path); - ppath = realpath(pdname, NULL); + /* check that the file exists on disk and is the correct type */ + if(lstat(path, &buf) == -1 || !is_dir != !S_ISDIR(buf.st_mode)) { + continue; + } + + if(is_dir) { + pdname = realpath(path, NULL); + ppath = mdirname(pdname); + } else { + pdname = mdirname(path); + ppath = realpath(pdname, NULL); + } free(pdname); if(!ppath) { -- 1.7.12

3 5

[pacman-dev] [PATCH] pacman.8: remove redundant section
by Allan McRae 05 Sep '12

05 Sep '12

Signed-off-by: Allan McRae <allan(a)archlinux.org> --- doc/pacman.8.txt | 3 --- 1 file changed, 3 deletions(-) diff --git a/doc/pacman.8.txt b/doc/pacman.8.txt index 0ad8188..24c2243 100644 --- a/doc/pacman.8.txt +++ b/doc/pacman.8.txt @@ -401,9 +401,6 @@ system upgrade and install/upgrade the foo package in the same operation. will force a refresh of all package lists even if they appear to be up to date. -*\--needed*:: - Do not reinstall the targets that are already up to date. - Handling Config Files[[HCF]] ---------------------------- -- 1.7.12

1 0

[pacman-dev] [PATCH 0/4] Miscellaneous query_fileowner cleanup
by Andrew Gregory 29 Aug '12

29 Aug '12

These patches apply on top of my previous query.c patches sitting in Allan's working-split/query-owner branch. Also available under the dirowner_split branch of my repo at https://bitbucket.org/andrewgregory/pacman

2 10

[pacman-dev] [RFC] Generation of a .SRCINFO file for source packages
by canyonknight＠gmail.com 29 Aug '12

29 Aug '12

Hello all, I had the idea of adding a .SRCINFO file to source packages generated by makepkg. I did some research to see if this had been previously proposed and where the discussion went. I've collected my thoughts on such an addition for your consideration. Previous proposal: The idea of generating a .SRCINFO file was previously proposed on the mailing list [1]. No format was ever agreed upon and discussion died off when no fixes were sent based off of developer suggestions. I've included my own proposed format to fuel discussion and maybe reach an ideal format. Proposed format with example (using "boost" PKGBUILD [2]): - Every variable (pkgver, depends, etc) is preceded by ($pkgname). - Every "global" (defined outside a package_* function) variable is explicitly listed for each $pkgname. - Every array (arch, depends, etc) has one line for each array element # Generated by makepkg 4.0.3 # Mon Aug 27 01:23:45 UTC 2012 (boost) pkgbase = boost (boost) pkgname = boost (boost) pkgver = 1.50.0 (boost) pkgrel = 2 (boost) epoch = 0 (boost) pkgdesc = Free peer-reviewed portable C++ source libraries - Development (boost) arch = i686 (boost) arch = x86_64 (boost) url = http://www.boost.org/ (boost) depends = boost-libs=1.50.0 (boost) optdepends = python: for python bindings (boost) optdepends = python2: for python2 bindings (boost) optdepends = boost-build: to use boost jam for building your project. (boost) makedepends = icu (boost) makedepends = python (boost) makedepends = python2 (boost) makedepends = bzip2 (boost) makedepends = zlib (boost) makedepends = openmpi (boost) source = http://downloads.sourceforge.net/boost/boost_1_50_0.tar.gz (boost) license = custom (boost-libs) pkgbase = boost (boost-libs) pkgname = boost-libs (boost-libs) pkgver = 1.50.0 (boost-libs) pkgrel = 2 (boost-libs) epoch = 0 (boost-libs) pkgdesc = Free peer-reviewed portable C++ source libraries - Runtime (boost-libs) arch = i686 (boost-libs) arch = x86_64 (boost-libs) url = http://www.boost.org/ (boost-libs) depends = gcc-libs (boost-libs) depends = bzip2 (boost-libs) depends = zlib (boost-libs) depends = icu (boost-libs) optdepends = openmpi: for mpi support (boost-libs) makedepends = icu (boost-libs) makedepends = python (boost-libs) makedepends = python2 (boost-libs) makedepends = bzip2 (boost-libs) makedepends = zlib (boost-libs) makedepends = openmpi (boost-libs) source = http://downloads.sourceforge.net/boost/boost_1_50_0.tar.gz (boost-libs) license = custom Advantages of proposed format: - Don't need a bash parser to determine PKGBUILD variables in a source package. This would make things a lot easier for projects (including AUR) that depend on information from PKGBUILD variables. - Previously proposed implementation could run into issues for a package named "global", this proposal avoids need for "global" label. Disadvantages of proposed format: - Adds additional dot file to every generated source package. - Compared to the previously proposed implementation, this implementation results in more lines (and a bit of redundancy) in each .SRCINFO file due to the avoidance of any "global" variables. Thanks for the consideration. I am more than happy to answer any questions relating to anything above. Regards, Jason [1] http://mailman.archlinux.org/pipermail/pacman-dev/2010-August/011461.html [2] https://projects.archlinux.org/svntogit/packages.git/plain/trunk/PKGBUILD?h…

3 5