December 2013 - Pacman-dev - lists.archlinux.org

[pacman-dev] Bug in libalpm: sizeof(off_t)
by Jeremy Heiner 11 Dec '13

11 Dec '13

An app developer can sometimes choose how big they want off_t to be (by #defining _FILE_OFFSET_BITS). The default for i686 is 32 bits (a bit odd, that). Pacman's configure attempts to choose 64 bits, and succeeds in doing so for the Arch builds (but doesn't force that choice). Bad things happen when the app developer chooses differently than configure did, but not until runtime. For example, the app's .o files will think sizeof(alpm_file_t) == 12, whereas libalpm thinks it == 16. The compile and link phases don't catch this. One possible solution might be to "mv alpm.h alpm.h.in", and add something like: #if sizeof(off_t) != @???@ #error ??? #endif Not sure yet what to put in those "???" spots, but the basic idea is to prevent the compiler from generating any .o files that disagree with the ABI of /usr/lib/libalpm. Hope that all made sense, and that I'm not duplicating - I searched the mailing list and the bug database and found zero hits. Anyone see a better solution? Jeremy

4 22

[pacman-dev] [PATCH] remove useless continue
by Andrew Gregory 11 Dec '13

11 Dec '13

Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- lib/libalpm/sync.c | 1 - 1 file changed, 1 deletion(-) diff --git a/lib/libalpm/sync.c b/lib/libalpm/sync.c index 551f926..7dbb8a0 100644 --- a/lib/libalpm/sync.c +++ b/lib/libalpm/sync.c @@ -536,7 +536,6 @@ int _alpm_sync_prepare(alpm_handle_t *handle, alpm_list_t **data) trans->add = alpm_list_remove(trans->add, rsync, _alpm_pkg_cmp, NULL); /* rsync is not a transaction target anymore */ trans->unresolvable = alpm_list_add(trans->unresolvable, rsync); - continue; } alpm_list_free_inner(deps, (alpm_list_fn_free)_alpm_conflict_free); -- 1.8.5.1

1 0

[pacman-dev] [PATCH] db.c: require unique database names
by Andrew Gregory 11 Dec '13

11 Dec '13

Allowing multiple databases with the same name causes conflicts as they both point to the same database file but may use different servers, usages, or siglevels. Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- doc/pacman.conf.5.txt | 9 +++++---- lib/libalpm/db.c | 13 +++++++++++++ 2 files changed, 18 insertions(+), 4 deletions(-) diff --git a/doc/pacman.conf.5.txt b/doc/pacman.conf.5.txt index 8cd2d0e..9c3dffe 100644 --- a/doc/pacman.conf.5.txt +++ b/doc/pacman.conf.5.txt @@ -202,10 +202,11 @@ Repository Sections ------------------- Each repository section defines a section name and at least one location where the packages can be found. The section name is defined by the string within -square brackets (the two above are 'current' and 'custom'). Locations are -defined with the 'Server' directive and follow a URL naming structure. If you -want to use a local directory, you can specify the full path with a ``file://'' -prefix, as shown above. +square brackets (the two above are 'core' and 'custom'). Repository names +must be unique and the name 'local' is reserved for the database of installed +packages. Locations are defined with the 'Server' directive and follow a URL +naming structure. If you want to use a local directory, you can specify the +full path with a ``file://'' prefix, as shown above. A common way to define DB locations utilizes the 'Include' directive. For each repository defined in the configuration file, a single 'Include' directive can diff --git a/lib/libalpm/db.c b/lib/libalpm/db.c index 528b04b..2069a7b 100644 --- a/lib/libalpm/db.c +++ b/lib/libalpm/db.c @@ -46,6 +46,8 @@ alpm_db_t SYMEXPORT *alpm_register_syncdb(alpm_handle_t *handle, const char *treename, alpm_siglevel_t level) { + alpm_list_t *i; + /* Sanity checks */ CHECK_HANDLE(handle, return NULL); ASSERT(treename != NULL && strlen(treename) != 0, @@ -53,6 +55,17 @@ alpm_db_t SYMEXPORT *alpm_register_syncdb(alpm_handle_t *handle, /* Do not register a database if a transaction is on-going */ ASSERT(handle->trans == NULL, RET_ERR(handle, ALPM_ERR_TRANS_NOT_NULL, NULL)); + /* ensure database name is unique */ + if(strcmp(treename, "local") == 0) { + RET_ERR(handle, ALPM_ERR_DB_NOT_NULL, NULL); + } + for(i = handle->dbs_sync; i; i = i->next) { + alpm_db_t *d = i->data; + if(strcmp(treename, d->treename) == 0) { + RET_ERR(handle, ALPM_ERR_DB_NOT_NULL, NULL); + } + } + return _alpm_db_register_sync(handle, treename, level); } -- 1.8.4.2

2 1

[pacman-dev] [PATCH] alpm_sync_sysupgrade: skip packages being removed
by Andrew Gregory 11 Dec '13

11 Dec '13

Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- pacman can't run into this (yet), but other front-ends might. lib/libalpm/sync.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/lib/libalpm/sync.c b/lib/libalpm/sync.c index 551f926..af82dc5 100644 --- a/lib/libalpm/sync.c +++ b/lib/libalpm/sync.c @@ -209,6 +209,11 @@ int SYMEXPORT alpm_sync_sysupgrade(alpm_handle_t *handle, int enable_downgrade) for(i = _alpm_db_get_pkgcache(handle->db_local); i; i = i->next) { alpm_pkg_t *lpkg = i->data; + if(alpm_pkg_find(trans->remove, lpkg->name)) { + _alpm_log(handle, ALPM_LOG_DEBUG, "%s is marked for removal -- skipping\n", lpkg->name); + continue; + } + if(alpm_pkg_find(trans->add, lpkg->name)) { _alpm_log(handle, ALPM_LOG_DEBUG, "%s is already in the target list -- skipping\n", lpkg->name); continue; -- 1.8.4.2

2 1

Re: [pacman-dev] [PATCH 4/4] makepkg: git: update existing sources in srcdir without removing them first.
by Ido Rosen 10 Dec '13

10 Dec '13

I think I found a bug: If the repository URL/source array entry has changed but the directory name has not, the user would have to manually delete the cloned repository, otherwise the git fetch/checkout would be from the incorrect upstream repo. On Mon, Dec 9, 2013 at 3:31 PM, Lukáš Jirkovský <l.jirkovsky(a)gmail.com>wrote: > The local changes are discarded when updating. This matches the behaviour > when non-VCS sources are used. It also allows incremental builds. > > Signed-off-by: Lukáš Jirkovský <l.jirkovsky(a)gmail.com> > --- > scripts/makepkg.sh.in | 19 ++++++++++++++----- > 1 file changed, 14 insertions(+), 5 deletions(-) > > diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in > index 1421bec..99af551 100644 > --- a/scripts/makepkg.sh.in > +++ b/scripts/makepkg.sh.in > @@ -581,9 +581,18 @@ extract_git() { > > msg2 "$(gettext "Creating working copy of %s %s repo...")" > "${repo}" "git" > pushd "$srcdir" &>/dev/null > - rm -rf "${dir##*/}" > > - if ! git clone "$dir"; then > + local updating=false > + if [[ -d "${dir##*/}" ]]; then > + updating=true > + cd_safe "${dir##*/}" > + if ! git fetch; then > + error "$(gettext "Failure while updating working > copy of %s %s repo")" "${repo}" "git" > + plain "$(gettext "Aborting...")" > + exit 1 > + fi > + cd_safe "$srcdir" > + elif ! git clone "$dir"; then > error "$(gettext "Failure while creating working copy of > %s %s repo")" "${repo}" "git" > plain "$(gettext "Aborting...")" > exit 1 > @@ -591,7 +600,7 @@ extract_git() { > > cd_safe "${dir##*/}" > > - local ref > + local ref=origin/HEAD > if [[ -n $fragment ]]; then > case ${fragment%%=*} in > commit|tag) > @@ -607,8 +616,8 @@ extract_git() { > esac > fi > > - if [[ -n $ref ]]; then > - if ! git checkout -b makepkg $ref; then > + if [[ -n $ref ]] || ((updating)) ; then > + if ! git checkout --force -B makepkg $ref; then > error "$(gettext "Failure while creating working > copy of %s %s repo")" "${repo}" "git" > plain "$(gettext "Aborting...")" > exit 1 > -- > 1.8.5.1 > > >

2 1

[pacman-dev] Questions about how to show messages (warnings, etc)
by Olivier Brunel 04 Dec '13

04 Dec '13

Hi, So I'm working on changing the events callback as discussed earlier, and I'm running into some little issues (discrepancies really) when it comes to showing warnings/errors. As you know, pacman can show warnings/errors to the user; and when colors are used those prefixes will be shown in color. Except, not always. There seems to be some inconsistency when it comes to this, because when a warning/error needs to be shown, pacman can do one of two things: - show it right away - cache it and show it later, i.e. when the current progress will be done Now, when shown right away, prefixes are in color, and messages are shown on stderr. However, when delayed, there will be no colors (because unlike pm_vfprintf(), pm_vasprintf() doesn't add colors to the prefixes) *and* messages will be shown on stdout. Unless there's a reason for this I'm not aware of, I think that this should probably be fixed, as in have delayed messages use colors & show on stderr? As a side note, the FS#37711 that Andrew mentioned previously was about the lack of colors for creation of .pac* files, but it might actually have been a manifestation of this bug/discrepancy (i.e. when fixed, or if using --noprogressbar, one would have colors on those warnings already). -j

3 4