- Pacman-dev - lists.archlinux.org

[pacman-dev] [PATCH 00/15] Full optdep support with all the bells and whistles
by Benedikt Morbach 11 Sep '11

11 Sep '11

Hi there! This should implement everything that's listed on this page: https://wiki.archlinux.org/index.php/User:Allan/Pacman_OptDepends I know that this is probably 4.1 material and will most likely need quite a bit of polish, but I'd like to gather some comments now before the bit-rot creeps in. If you like, you can also get the code at https://github.com/moben/pacman on the optdep branch. Please take a look and comment. --- Benedikt Benedikt Morbach (15): Split optdep into alpm_depend_t and description Hook new optdepend structures up Only display uninstalled optdepends during install/upgrade Show optdep install status in package info optdepends are not orphans unless --optdeps is specified Make package info show optional requirements Make recursive removal consider optdepends Show list of optrequires with -Qtdn Add config option for showing all optdeps again Warn on optdep removal Add flag to recurse through optdepends Show status on optdepends that are part of transaction Add option to install all optdepends by default Add option to ask which optdeps to install Make both HandleOptdeps settings work together doc/pacman.8.txt | 11 +++ doc/pacman.conf.5.txt | 14 ++++ etc/pacman.conf.in | 3 + lib/libalpm/add.c | 4 +- lib/libalpm/alpm.h | 25 ++++++- lib/libalpm/be_local.c | 12 +++- lib/libalpm/be_package.c | 5 +- lib/libalpm/be_sync.c | 8 ++- lib/libalpm/deps.c | 127 ++++++++++++++++++++++++++++++---- lib/libalpm/deps.h | 5 +- lib/libalpm/error.c | 2 + lib/libalpm/package.c | 40 ++++++++--- lib/libalpm/remove.c | 22 +++++-- lib/libalpm/sync.c | 2 +- lib/libalpm/trans.c | 10 ++- src/pacman/conf.c | 28 ++++++++ src/pacman/conf.h | 10 +++ src/pacman/package.c | 24 +++++-- src/pacman/pacman.c | 7 ++ src/pacman/query.c | 27 ++++++- src/pacman/remove.c | 35 ++++++++-- src/pacman/sync.c | 125 ++++++++++++++++++++++++++++++++-- src/pacman/upgrade.c | 2 +- src/pacman/util.c | 172 +++++++++++++++++++++++++++++++++++++++------ src/pacman/util.h | 5 +- src/util/pactree.c | 2 +- src/util/testdb.c | 2 +- 27 files changed, 635 insertions(+), 94 deletions(-) -- 1.7.6.1

1 15

[pacman-dev] pacman -Qs first-run performance
by Philipp 11 Sep '11

11 Sep '11

Hi there, pacman works very well for me, with one exception: searching for installed packages. For me a -Ss takes about 5 seconds, a -Qs 25 seconds. I also noticed that this seems to be true for the first search run only, the following ones are less than a second, no matter whether it's a repo or local search. But if you want to just quickly find out a thing, 25 seconds or so is a long time. Anyway, I had a very brief look at the code and am far from understanding it, but I think libalpm/db.c handles the search, the package names, descriptions, etc. are stored in a linked list of structs, the whole thing is cached in memory only and regex wizardry is used for the search. If that's true, the bottleneck I experience is the caching. 5 seconds for -Ss is acceptable for me, but I wonder whether there's a reasonably easy way to improve the 25 seconds of -Qs. My knowledge of pacman internals is non-existant and my C skills are minimal, so I don't think I can be a lot of help with this one, but maybe there's something else I can be of help with instead. Best regards, Philipp

5 13

[pacman-dev] [PATCH] makepkg: unset errexit when sourcing /etc/profile
by Dave Reisner 11 Sep '11

11 Sep '11

This is a fix for a bash3 specific bug, where a file sourced by /etc/profile would exit non-zero and make its way back up to makepkg, forcing it to exit after package installation. Along with unsetting the ERR handler, temporarily unset errexit to avoid this. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> --- scripts/makepkg.sh.in | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 5981603..a130557 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -474,11 +474,13 @@ handle_deps() { fi # we might need the new system environment - # avoid triggering the ERR trap + # avoid triggering the ERR trap and exiting + set +x local restoretrap=$(trap -p ERR) trap - ERR source /etc/profile &>/dev/null eval $restoretrap + set -x return $R_DEPS_SATISFIED } -- 1.7.6.1

1 1

[pacman-dev] [PATCH] paccache: escape . in trimming of diskspace string
by Dave Reisner 11 Sep '11

11 Sep '11

Before: ==> finished dry run: 2 candidates (diskspace saved: 7. MiB) After: ==> finished dry run: 2 candidates (diskspace saved: 7.8 MiB) Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> --- contrib/paccache.in | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/contrib/paccache.in b/contrib/paccache.in index b191ed9..8a8cad4 100755 --- a/contrib/paccache.in +++ b/contrib/paccache.in @@ -121,7 +121,7 @@ size_to_human() { } sizestr = sprintf("%.2f", size) - sub(/.?0+$/, "", sizestr) + sub(/\.?0+$/, "", sizestr) printf("%s %s", sizestr, suffix[count]) }' } -- 1.7.6.1

1 0

[pacman-dev] [PATCH 1/2] Move download callback static vars into function
by Dan McGee 10 Sep '11

10 Sep '11

Signed-off-by: Dan McGee <dan(a)archlinux.org> --- src/pacman/callback.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/src/pacman/callback.c b/src/pacman/callback.c index 0d9b57b..793facd 100644 --- a/src/pacman/callback.c +++ b/src/pacman/callback.c @@ -37,11 +37,8 @@ #include "conf.h" /* download progress bar */ -static double rate_last; -static off_t xfered_last; static off_t list_xfered = 0.0; static off_t list_total = 0.0; -static struct timeval initial_time; /* delayed output during progress bar */ static int on_progress = 0; @@ -510,6 +507,9 @@ void cb_dl_total(off_t total) /* callback to handle display of download progress */ void cb_dl_progress(const char *filename, off_t file_xfered, off_t file_total) { + static double rate_last; + static off_t xfered_last; + static struct timeval initial_time; int infolen; int filenamelen; char *fname, *p; -- 1.7.6.1

1 1

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v4.0.0rc1-175-gdb70c9d
by dan＠archlinux.org 09 Sep '11

09 Sep '11

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 db70c9da153c4767bfd96a15d40e086184bd5402 (commit) via a03313f3f66e9ac03392cb255bc6c4e1958e14c2 (commit) via 905ae640cfb0787e9bdfbcacf21c62e64cc27512 (commit) via 4ed12aec30be0a36fff2937039c904526aeb3b79 (commit) via 17c3de3e4f6b03953e22d203d5891b37ffdf941c (commit) via 3d9f961d1367bf1669332047e667f98b9d35531e (commit) via 8ffa2b24a5e3d63d6f34a1257ed67904ec051e3d (commit) via b961ebe16ffb75bb947a193daa9e9fe639b6403d (commit) via 6317db84295430c2d2d164f81ff723b379b33035 (commit) via 3e08614fda144393ab1ca23dcf08f2cd2f604cf9 (commit) from 0e79802c0ac8453376d8c0f99629f5a3b499f571 (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 db70c9da153c4767bfd96a15d40e086184bd5402 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Wed Sep 7 22:22:25 2011 -0400 diskspace: fix memory leak on root mount not found Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit a03313f3f66e9ac03392cb255bc6c4e1958e14c2 Author: Dan McGee <dan(a)archlinux.org> Date: Thu Sep 8 14:14:51 2011 -0500 Print 'loading packages' message on -U operations This will be the first thing printed when doing an upgrade. Currently there is no output at all until we start resolving dependencies, which can be a while in if specifying very large targets on the command line, in which case it is nice to let the user know we are doing something. Addresses FS#25822 in the most KISS way possible. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 905ae640cfb0787e9bdfbcacf21c62e64cc27512 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Tue Sep 6 14:40:29 2011 -0400 makepkg: use more awk'ish syntax in sanity checks This simplifies the flow a bit, making the pipeline a little easier to grok. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit 4ed12aec30be0a36fff2937039c904526aeb3b79 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Tue Sep 6 14:17:15 2011 -0400 makepkg: avoid for loop in deleting manpage hardlinks find can do this all on its own and remain portable. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit 17c3de3e4f6b03953e22d203d5891b37ffdf941c Author: Dave Reisner <dreisner(a)archlinux.org> Date: Tue Sep 6 14:11:15 2011 -0400 makepkg: act on function return value, not output Correcting a typo, as this function will never output anything. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit 3d9f961d1367bf1669332047e667f98b9d35531e Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Sep 4 12:34:57 2011 -0400 makepkg: refactor check_pgpsigs output - display associated warnings on same line as pass/fail msg, to be more consistent with checksum verification output - properly error on a revoked key (matching pacman's behavior) Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit 8ffa2b24a5e3d63d6f34a1257ed67904ec051e3d Author: Dan McGee <dan(a)archlinux.org> Date: Wed Sep 7 20:43:49 2011 -0500 Use more correct integer types in diskspace checks This adjusts type usage to match POSIX provided types from <sys/types.h> rather than assuming everything will fit in a long or unsigned long. Use fsblkcnt_t (unsigned) and blkcnt_t (signed) as appropriate. These are affected the same way off_t is on 32 bit platforms, where the types are extende to 64 bits if large file support is enabled. Because most numbers here are block counts, this isn't near as pressing as using a 32-bit variable for file sizes where anything over 2GiB can burn you; we likely can support files at least 512 but mainly 4096 times larger. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit b961ebe16ffb75bb947a193daa9e9fe639b6403d Author: Dan McGee <dan(a)archlinux.org> Date: Wed Sep 7 20:41:09 2011 -0500 query check: use provided filelist count instead of keeping track We don't need to keep track of how many files are in a package now that said value is provided to us. It also makes more sense to use size_t here for types rather than the (hopefully never too short) int. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 6317db84295430c2d2d164f81ff723b379b33035 Author: Dan McGee <dan(a)archlinux.org> Date: Wed Sep 7 20:33:23 2011 -0500 Remove unnecessary cast Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 3e08614fda144393ab1ca23dcf08f2cd2f604cf9 Author: Dan McGee <dan(a)archlinux.org> Date: Wed Sep 7 15:02:52 2011 -0500 Ensure PackageRequired works as expected Changes in commit dc3336c277 caused this to stop working as expected for sync packages, due to the way the logic is structured. Ensure we always enter the signature code if the bitflag is flipped on to check signatures for packages. Rename 'use_sig' to 'has_sig' for clarity. Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: lib/libalpm/be_package.c | 13 ++++++----- lib/libalpm/diskspace.c | 33 +++++++++++++++------------ lib/libalpm/diskspace.h | 7 ++++- scripts/makepkg.sh.in | 54 +++++++++++++++++++++++----------------------- src/pacman/package.c | 2 +- src/pacman/query.c | 12 +++++----- src/pacman/upgrade.c | 1 + 7 files changed, 65 insertions(+), 57 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH 1/3] make compute_download_size consider .part files
by Dave Reisner 08 Sep '11

08 Sep '11

Check for the existance of a partial download of a package file before jumping to delta calculations. Currently, if there were 10MiB remaining in a 100MiB the values passed to the front end do not reflect this. The frontend accounts for this change in the verbose target display by using alpm_pkg_download_size() instead of alpm_pkg_get_size(). We also change the label in the final column from a vague "Size" to a more descriptive "Download Size" Refactored from an old patch originally by Dan. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> --- lib/libalpm/sync.c | 29 ++++++++++++++++++++++++----- src/pacman/util.c | 4 ++-- 2 files changed, 26 insertions(+), 7 deletions(-) diff --git a/lib/libalpm/sync.c b/lib/libalpm/sync.c index e2562c0..471bbe7 100644 --- a/lib/libalpm/sync.c +++ b/lib/libalpm/sync.c @@ -284,7 +284,8 @@ alpm_list_t SYMEXPORT *alpm_find_group_pkgs(alpm_list_t *dbs, */ static int compute_download_size(alpm_pkg_t *newpkg) { - char *fpath; + const char *fname; + char *fpath, *fnamepart = NULL; off_t size = 0; alpm_handle_t *handle = newpkg->handle; @@ -295,11 +296,26 @@ static int compute_download_size(alpm_pkg_t *newpkg) } ASSERT(newpkg->filename != NULL, RET_ERR(handle, ALPM_ERR_PKG_INVALID_NAME, -1)); - fpath = _alpm_filecache_find(handle, newpkg->filename); + fname = newpkg->filename; + fpath = _alpm_filecache_find(handle, fname); + /* downloaded file exists, so there's nothing to grab */ if(fpath) { - FREE(fpath); size = 0; + goto finish; + } + + CALLOC(fnamepart, strlen(fname) + 6, sizeof(char), return -1); + sprintf(fnamepart, "%s.part", fname); + fpath = _alpm_filecache_find(handle, fnamepart); + if(fpath) { + struct stat st; + if(stat(fpath, &st) == 0) { + /* subtract the size of the .part file */ + _alpm_log(handle, ALPM_LOG_DEBUG, "using (package - .part) size\n"); + size = newpkg->size - st.st_size; + size = size < 0 ? 0 : size; + } } else if(handle->usedelta) { off_t dltsize; @@ -315,15 +331,18 @@ static int compute_download_size(alpm_pkg_t *newpkg) alpm_list_free(newpkg->delta_path); newpkg->delta_path = NULL; } - } else { - size = newpkg->size; } +finish: _alpm_log(handle, ALPM_LOG_DEBUG, "setting download size %jd for pkg %s\n", (intmax_t)size, newpkg->name); newpkg->infolevel |= INFRQ_DSIZE; newpkg->download_size = size; + + FREE(fpath); + FREE(fnamepart); + return 0; } diff --git a/src/pacman/util.c b/src/pacman/util.c index 594186f..6963afb 100644 --- a/src/pacman/util.c +++ b/src/pacman/util.c @@ -748,7 +748,7 @@ static alpm_list_t *create_verbose_header(int install) pm_asprintf(&str, "%s", _("New Version")); res = alpm_list_add(res, str); } - pm_asprintf(&str, "%s", _("Size")); + pm_asprintf(&str, "%s", _("Download Size")); res = alpm_list_add(res, str); return res; @@ -779,7 +779,7 @@ static alpm_list_t *create_verbose_row(alpm_pkg_t *pkg, int install) ret = alpm_list_add(ret, str); /* and size */ - size = humanize_size(alpm_pkg_get_size(pkg), 'M', &label); + size = humanize_size(alpm_pkg_download_size(pkg), 'M', &label); pm_asprintf(&str, "%.2f %s", size, label); ret = alpm_list_add(ret, str); -- 1.7.6.1

2 3

[pacman-dev] [PATCH] Ensure PackageRequired works as expected
by Dan McGee 07 Sep '11

07 Sep '11

Changes in commit dc3336c277 caused this to stop working as expected for sync packages, due to the way the logic is structured. Ensure we always enter the signature code if the bitflag is flipped on to check signatures for packages. Rename 'use_sig' to 'has_sig' for clarity. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- lib/libalpm/be_package.c | 13 +++++++------ 1 files changed, 7 insertions(+), 6 deletions(-) diff --git a/lib/libalpm/be_package.c b/lib/libalpm/be_package.c index 98a1240..8b035ca 100644 --- a/lib/libalpm/be_package.c +++ b/lib/libalpm/be_package.c @@ -286,7 +286,7 @@ static alpm_file_t *files_msort(alpm_file_t *files, size_t n) alpm_pkg_t *_alpm_pkg_load_internal(alpm_handle_t *handle, const char *pkgfile, alpm_pkg_t *syncpkg, int full, alpm_siglevel_t level) { - int ret, use_sig, config = 0; + int ret, has_sig, config = 0; struct archive *archive; struct archive_entry *entry; alpm_pkg_t *newpkg = NULL; @@ -312,20 +312,20 @@ alpm_pkg_t *_alpm_pkg_load_internal(alpm_handle_t *handle, const char *pkgfile, } /* can we get away with skipping checksums? */ - use_sig = 0; + has_sig = 0; if(level & ALPM_SIG_PACKAGE) { if(syncpkg && syncpkg->base64_sig) { - use_sig = 1; + has_sig = 1; } else { char *sigpath = _alpm_sigpath(handle, pkgfile); if(sigpath && !_alpm_access(handle, NULL, sigpath, R_OK)) { - use_sig = 1; + has_sig = 1; } free(sigpath); } } - if(syncpkg && !use_sig) { + if(syncpkg && !has_sig) { if(syncpkg->md5sum && !syncpkg->sha256sum) { _alpm_log(handle, ALPM_LOG_DEBUG, "md5sum: %s\n", syncpkg->md5sum); _alpm_log(handle, ALPM_LOG_DEBUG, "checking md5sum for %s\n", pkgfile); @@ -345,7 +345,8 @@ alpm_pkg_t *_alpm_pkg_load_internal(alpm_handle_t *handle, const char *pkgfile, } } - if(use_sig) { + /* even if we don't have a sig, run the check code if level tells us to */ + if(has_sig || level & ALPM_SIG_PACKAGE) { const char *sig = syncpkg ? syncpkg->base64_sig : NULL; _alpm_log(handle, ALPM_LOG_DEBUG, "sig data: %s\n", sig ? sig : "<from .sig>"); if(_alpm_check_pgp_helper(handle, pkgfile, sig, -- 1.7.6.1

1 0

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v4.0.0rc1-165-g0e79802
by dan＠archlinux.org 06 Sep '11

06 Sep '11

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 0e79802c0ac8453376d8c0f99629f5a3b499f571 (commit) via 8d0ff3d7dc4ea925a1fd94ae022a16a14945aa3f (commit) from 29ad9e0a0a93a60a1118eb0e8cb6a14df147320f (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 0e79802c0ac8453376d8c0f99629f5a3b499f571 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sat Sep 3 22:24:27 2011 -0400 makepkg: use globs in place of regex We seem to enjoy using bash regex capabilities, but never referencing the result with BASH_REMATCH. Replace almost all regexes with equivalent globs which are faster and functionally equivalent in these cases. This enables the extglob shopt. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 8d0ff3d7dc4ea925a1fd94ae022a16a14945aa3f Author: Dan McGee <dan(a)archlinux.org> Date: Tue Sep 6 11:44:29 2011 -0500 dload: use intmax_t when printing off_t This works for both 32-bit and 64-bit platforms. Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: lib/libalpm/dload.c | 6 ++++-- scripts/makepkg.sh.in | 16 +++++++++------- 2 files changed, 13 insertions(+), 9 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v4.0.0rc1-163-g29ad9e0
by dan＠archlinux.org 06 Sep '11

06 Sep '11

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 29ad9e0a0a93a60a1118eb0e8cb6a14df147320f (commit) via 7ed54a99401f7a16edcba27725e92fb7474be240 (commit) via a4e0d3e9302aa16b966ea5d84926664748ba9ae2 (commit) via 6c236277a3d5b444a30790702b8114b6658696a2 (commit) via 7054e37126cc476b88da7b57a6efa658d4765cf0 (commit) via 400942fedf35d4fc6b584b8f1552f1ec649136fe (commit) via 7ea1ea88bbca64514d5aad64fe199191f0f2eea1 (commit) from 7c956d5d4bdf89bfece1244baa56a92f83bd9161 (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 29ad9e0a0a93a60a1118eb0e8cb6a14df147320f Author: Dave Reisner <dreisner(a)archlinux.org> Date: Tue Sep 6 09:29:52 2011 -0400 makepkg: unset variables as per !{make,build}flags Don't just set the flag variables to zero length strings, actually unset them from the environment. This fixes issues with broken gnu Makefies that use ?= for assigment, where the presence of a var is enough to make this condition avoid assignment. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 7ed54a99401f7a16edcba27725e92fb7474be240 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sat Sep 3 22:23:21 2011 -0400 rankmirrors: properly sort resulting times - Properly read each sorted line into a new array, instead of breaking on every word. - LC_COLLATE should apply to the sort portion of the pipeline, not the printing. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit a4e0d3e9302aa16b966ea5d84926664748ba9ae2 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Sep 4 11:58:45 2011 -0400 dload: abstract dload_interrupted reasons This gives us some amount of room to grow in case we ever find another reason that we might return with an error from the progress callback. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit 6c236277a3d5b444a30790702b8114b6658696a2 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Sep 4 11:47:40 2011 -0400 dload: improve debug output We lost some of this output in the fetch->curl conversion, but I also noticed in FS#25852 that we just lack some of this useful information along the way. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit 7054e37126cc476b88da7b57a6efa658d4765cf0 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Mon Sep 5 13:16:07 2011 -0400 sync: add missing newline in warning message Dan: fix the other missing one too. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 400942fedf35d4fc6b584b8f1552f1ec649136fe Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Sep 4 14:59:17 2011 -0400 avoid blowing out the filecache list when using tmpdir _alpm_filecache_setup() destroys the list of cachedirs when it finds no writeable directories in the config. This put us in an awkward situation where _alpm_filecache_find() would locate a downloaded file in a r/o cachedir, but then fail to install it after _alpm_filecache_setup() is called (with a NULL argument). Change this behavior to merely prepend the temporary directory to the list of available cachedirs. Dan exposed it in e07547ee4ed4, as now a package can be found in a directory we may not be able to actually store packages in. Reported-by: Rémy Oudompheng <remy(a)archlinux.org> Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 7ea1ea88bbca64514d5aad64fe199191f0f2eea1 Author: Dan McGee <dan(a)archlinux.org> Date: Fri Sep 2 22:07:06 2011 -0500 More package operations cleanup Neither deltas nor filename attributes are ever present in the local database, so we can remove all of the indirection for accessing these attributes. Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: lib/libalpm/be_local.c | 14 -------------- lib/libalpm/delta.c | 5 +---- lib/libalpm/dload.c | 27 ++++++++++++++++++++++----- lib/libalpm/package.c | 11 ++++------- lib/libalpm/package.h | 7 ------- lib/libalpm/signing.c | 2 +- lib/libalpm/sync.c | 18 ++++++------------ lib/libalpm/util.c | 6 ++---- scripts/makepkg.sh.in | 6 ++---- scripts/rankmirrors.sh.in | 3 ++- src/pacman/query.c | 2 +- src/pacman/sync.c | 2 +- 12 files changed, 42 insertions(+), 61 deletions(-) hooks/post-receive -- The official pacman repository

1 0