- Pacman-dev - lists.archlinux.org

[pacman-dev] [PATCH 1/3] Store a package info level flag if we fail to load data
by Dan McGee 30 Jul '11

30 Jul '11

If we are missing a local database file, we get repeated messages over and over telling us the same thing, rather than being sane and erroring only once. This package adds an INFRQ_ERROR level that is added to the mask if we encounter any errors on a local_db_read() operation, and short circuits future calls if found in the value. This fixes FS#25313. Note that this does not make any behavior changes other than suppressing error messages and repeated code calls to failure cases; we still have more to do in the "local database is hosed" department. Also make a small update to the wrong but unused flags set in be_package; using INFRQ_ALL there was not totally correct. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- lib/libalpm/be_local.c | 14 +++++++++++--- lib/libalpm/be_package.c | 4 ++-- lib/libalpm/db.h | 3 ++- 3 files changed, 15 insertions(+), 6 deletions(-) diff --git a/lib/libalpm/be_local.c b/lib/libalpm/be_local.c index 261ad87..73b533f 100644 --- a/lib/libalpm/be_local.c +++ b/lib/libalpm/be_local.c @@ -533,6 +533,13 @@ static int local_db_read(alpm_pkg_t *info, alpm_dbinfrq_t inforeq) /* already loaded all of this info, do nothing */ return 0; } + + if(info->infolevel & INFRQ_ERROR) { + /* We've encountered an error loading this package before. Don't attempt + * repeated reloads, just give up. */ + return -1; + } + _alpm_log(db->handle, ALPM_LOG_FUNCTION, "loading package data for %s : level=0x%x\n", info->name, inforeq); @@ -619,6 +626,7 @@ static int local_db_read(alpm_pkg_t *info, alpm_dbinfrq_t inforeq) } fclose(fp); fp = NULL; + info->infolevel |= INFRQ_DESC; } /* FILES */ @@ -673,6 +681,7 @@ static int local_db_read(alpm_pkg_t *info, alpm_dbinfrq_t inforeq) } fclose(fp); fp = NULL; + info->infolevel |= INFRQ_FILES; } /* INSTALL */ @@ -681,15 +690,14 @@ static int local_db_read(alpm_pkg_t *info, alpm_dbinfrq_t inforeq) if(access(path, F_OK) == 0) { info->scriptlet = 1; } + info->infolevel |= INFRQ_SCRIPTLET; } - /* internal */ - info->infolevel |= inforeq; - free(pkgpath); return 0; error: + info->infolevel |= INFRQ_ERROR; free(pkgpath); if(fp) { fclose(fp); diff --git a/lib/libalpm/be_package.c b/lib/libalpm/be_package.c index 0edaa5a..550f303 100644 --- a/lib/libalpm/be_package.c +++ b/lib/libalpm/be_package.c @@ -437,9 +437,9 @@ alpm_pkg_t *_alpm_pkg_load_internal(alpm_handle_t *handle, const char *pkgfile, _alpm_log(handle, ALPM_LOG_DEBUG, "sorting package filelist for %s\n", pkgfile); newpkg->files.files = files_msort(files, files_count); newpkg->files.count = files_count; - newpkg->infolevel = INFRQ_ALL; + newpkg->infolevel = INFRQ_BASE | INFRQ_DESC | INFRQ_FILES | INFRQ_SCRIPTLET; } else { - newpkg->infolevel = INFRQ_BASE | INFRQ_DESC; + newpkg->infolevel = INFRQ_BASE | INFRQ_DESC | INFRQ_SCRIPTLET; } return newpkg; diff --git a/lib/libalpm/db.h b/lib/libalpm/db.h index 5c7988b..a09c7e2 100644 --- a/lib/libalpm/db.h +++ b/lib/libalpm/db.h @@ -40,7 +40,8 @@ typedef enum _alpm_dbinfrq_t { INFRQ_SCRIPTLET = (1 << 3), INFRQ_DSIZE = (1 << 4), /* ALL should be info stored in the package or database */ - INFRQ_ALL = 0x1F + INFRQ_ALL = 0x1F, + INFRQ_ERROR = (1 << 31) } alpm_dbinfrq_t; /** Database status. Bitflags. */ -- 1.7.6

1 2

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.5.3-518-g804e250
by dan＠archlinux.org 27 Jul '11

27 Jul '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 804e2505cf4474791537a9f0ae8e9b6c274cb0c2 (commit) via c5d4c92ad411998b9d47d7bf89d84af547117be4 (commit) via d9875c5e6cbc0b59a4ba7c56e974da3d138f235d (commit) via 49d9426b6a3e6a019a0973ec73b0ff6cf402fa89 (commit) via 2b3405e01ba756a48ff374bf632f525ad44c47c1 (commit) via 94f61c5b29274df164b01105165945e445289ae7 (commit) via 9929a34a6d493c94bd176b66198bdbd734678043 (commit) via ccdb2fa800f3e89641d371056961fa1d654f363d (commit) via 03447ce39cd6a3adb7a56bb978368a5f11659374 (commit) via a7940e7419a6fb39eeafb0bfd053b90a64946d5c (commit) via 819f6750047cbdb7ecb7fa4724f90e532cd73446 (commit) via 00949db191d0f31add4919ae56ebdfef9813a90e (commit) via 06974ebf2f6674baf01da79be9658277cbf2650e (commit) from 98073afe5527aa8a08c7119b61d0251250168e82 (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 804e2505cf4474791537a9f0ae8e9b6c274cb0c2 Author: Pang Yan Han <pangyanhan(a)gmail.com> Date: Tue Jul 19 20:52:07 2011 +0800 pacman-key: Add --import and --import-trustdb Currently, pacman-key allows the user to import their keys using the --add option. However, no similar functionality exists for importing ownertrust values. The --import-trustdb option takes a list of directories and imports ownertrust values if the directories have a trustdb.gpg database. The --import option takes a list of directories and imports keys from pubring.gpg and ownertrust values from trustdb.gpg. Think of it as a combination of --add and --import-trustdb Signed-off-by: Pang Yan Han <pangyanhan(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit c5d4c92ad411998b9d47d7bf89d84af547117be4 Author: DJ Mills <danielmills1(a)gmail.com> Date: Wed Jul 20 14:28:39 2011 +1000 pacman-key: change GPG_PACMAN and GPG_NOKEYRING to arrays Allows the commands to safely handle any possible arguments Signed-off-by: DJ Mills <danielmills1(a)gmail.com> Allan: rebase patch Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit d9875c5e6cbc0b59a4ba7c56e974da3d138f235d Author: Allan McRae <allan(a)archlinux.org> Date: Wed Jul 20 14:19:53 2011 +1000 pacman-key: fix syntax error in -r arg parsing Previous fix did not work... Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 49d9426b6a3e6a019a0973ec73b0ff6cf402fa89 Author: Allan McRae <allan(a)archlinux.org> Date: Mon Jul 18 19:58:54 2011 +1000 makepkg: refactor checking source integrity Move the source integrity checking into its own function as the code was duplicated and is now more complicated with the separation of the two checks types. Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 2b3405e01ba756a48ff374bf632f525ad44c47c1 Author: Allan McRae <allan(a)archlinux.org> Date: Sat Jul 16 22:42:32 2011 +1000 makepkg: more control of skipping integrity checks Allows the skipping of all integrity checks (checksum and PGP) or either the checksum or PGP checks individually. Original-patch-by: Wieland Hoffman <theminew(a)googlemail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 94f61c5b29274df164b01105165945e445289ae7 Author: Wieland Hoffmann <themineo(a)googlemail.com> Date: Wed Jul 6 13:02:19 2011 +0200 makepkg: Add support for verifying pgp signatures Many projects provide signature files along with the source code archives. It's good to check these, too, when verifying the integrity of source code archives. Not everybody is using gpg so the verification can be disabled with --skippgpcheck. Additionally, only a warning is displayed when the key that signed the source file is unknown. Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 9929a34a6d493c94bd176b66198bdbd734678043 Author: Dan McGee <dan(a)archlinux.org> Date: Fri Jul 22 12:12:18 2011 -0500 Remove duplicate code shared between sync and upgrade Pacman did a great job of having almost (but not quite) duplicate code paths through the sync and upgrade code. We can use the same logic in both upgrade in sync once the targets are resolved, so extract a function and delete a bunch of code. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit ccdb2fa800f3e89641d371056961fa1d654f363d Author: Allan McRae <allan(a)archlinux.org> Date: Fri Jul 22 21:42:53 2011 +1000 makepkg: get package version with overrides When epoch, pkgver and/or pkgrel were overridden in a split package function, makepkg failed hard finding the real version for checking if packages were already built or trying to install packages. Fix the get_full_version function to deal with overrides and return the actual package version. Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 03447ce39cd6a3adb7a56bb978368a5f11659374 Author: Allan McRae <allan(a)archlinux.org> Date: Fri Jul 22 20:08:39 2011 +1000 makepkg: allow epoch to be overridden We can override pkgver and pkgrel so it is only logical to add epoch to that list Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit a7940e7419a6fb39eeafb0bfd053b90a64946d5c Author: Allan McRae <allan(a)archlinux.org> Date: Fri Jul 22 19:47:07 2011 +1000 makepkg: check arch overrides for required architecture Check any overrides of the "arch" variable contain the required architecture. Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 819f6750047cbdb7ecb7fa4724f90e532cd73446 Author: Allan McRae <allan(a)archlinux.org> Date: Fri Jul 22 16:30:41 2011 +1000 makepkg: check overrides for pkgrel and pkgver Enforce syntax checking for pkgrel and pkgver overrides in package functions. Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 00949db191d0f31add4919ae56ebdfef9813a90e Author: Allan McRae <allan(a)archlinux.org> Date: Fri Jul 22 12:37:25 2011 +1000 makepkg: pkgver and pkgrel can not have whitespace There is always someone who tries to break things (cough *Dave* cough...) Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 06974ebf2f6674baf01da79be9658277cbf2650e Author: Dan McGee <dan(a)archlinux.org> Date: Wed Jul 27 12:44:17 2011 -0500 contrib/pacsearch: skip non-matching lines This prevents some perl errors from popping up when pacman prints error or warning messages. Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: contrib/pacsearch.in | 2 + doc/PKGBUILD.5.txt | 2 +- doc/makepkg.8.txt | 8 ++- doc/pacman-key.8.txt | 7 ++ proto/PKGBUILD-split.proto | 1 + scripts/makepkg.sh.in | 239 ++++++++++++++++++++++++++++++++++---------- scripts/pacman-key.sh.in | 93 ++++++++++++----- src/pacman/pacman.h | 1 + src/pacman/sync.c | 19 ++-- src/pacman/upgrade.c | 114 +-------------------- 10 files changed, 287 insertions(+), 199 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH] Fix group selection entry for large inputs
by Dan McGee 25 Jul '11

25 Jul '11

Hardcoding anything always ends up burning you, and the arbitrary length of 64 here did just that. Add the ability to reallocate the readline buffer for longer inputs if necessary, and add other error checking as approprate. This also plugs one small memory leak of the group processing code selection array. Addresses FS#24253. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- src/pacman/sync.c | 11 ++++++++++- src/pacman/util.c | 38 +++++++++++++++++++++++++++++++++++--- 2 files changed, 45 insertions(+), 4 deletions(-) diff --git a/src/pacman/sync.c b/src/pacman/sync.c index c56934b..6962306 100644 --- a/src/pacman/sync.c +++ b/src/pacman/sync.c @@ -644,7 +644,15 @@ static int process_group(alpm_list_t *dbs, char *group) group); select_display(pkgs); char *array = malloc(count); - multiselect_question(array, count); + if(!array) { + ret = 1; + goto cleanup; + } + if(multiselect_question(array, count)) { + ret = 1; + free(array); + goto cleanup; + } int n = 0; for(i = pkgs; i; i = alpm_list_next(i)) { if(array[n++] == 0) @@ -657,6 +665,7 @@ static int process_group(alpm_list_t *dbs, char *group) goto cleanup; } } + free(array); } else { for(i = pkgs; i; i = alpm_list_next(i)) { pmpkg_t *pkg = alpm_list_getdata(i); diff --git a/src/pacman/util.c b/src/pacman/util.c index f695ffb..53dbd63 100644 --- a/src/pacman/util.c +++ b/src/pacman/util.c @@ -796,8 +796,9 @@ static int multiselect_parse(char *array, int count, char *response) int multiselect_question(char *array, int count) { - char response[64]; + char *response, *lastchar; FILE *stream; + size_t response_len = 64; if(config->noconfirm) { stream = stdout; @@ -806,19 +807,45 @@ int multiselect_question(char *array, int count) stream = stderr; } + response = malloc(response_len); + if(!response) { + return -1; + } + lastchar = response + response_len - 1; + /* sentinel byte to later see if we filled up the entire string */ + *lastchar = 1; + while(1) { memset(array, 1, count); fprintf(stream, "\n"); fprintf(stream, _("Enter a selection (default=all)")); fprintf(stream, ": "); + fflush(stream); if(config->noconfirm) { fprintf(stream, "\n"); break; } - if(fgets(response, sizeof(response), stdin)) { + if(fgets(response, response_len, stdin)) { + const size_t response_incr = 64; + /* handle buffer not being large enough to read full line case */ + while(*lastchar == '\0' && lastchar[-1] != '\n') { + response_len += response_incr; + response = realloc(response, response_len); + if(!response) { + return -1; + } + lastchar = response + response_len - 1; + /* sentinel byte */ + *lastchar = 1; + if(fgets(response + response_len - response_incr - 1, + response_incr + 1, stdin) == 0) { + free(response); + return -1; + } + } strtrim(response); if(strlen(response) > 0) { if(multiselect_parse(array, count, response) == -1) { @@ -826,9 +853,14 @@ int multiselect_question(char *array, int count) continue; } } + break; + } else { + free(response); + return -1; } - break; } + + free(response); return(0); } -- 1.7.6

1 0

[pacman-dev] [PATCH] libalpm: Set ret correctly in download_internal()
by Lukas Fleischer 25 Jul '11

25 Jul '11

Immediately jump to the cleanup code after setting the return code to -1 in case rename() fails. Otherwise, it will be reset to 0 right after we leave the if branch. Signed-off-by: Lukas Fleischer <archlinux(a)cryptocrack.de> --- Just spotted this one in maint. I realized the download_internal() code was completely refactored in master; the only reason I submit this is to fix this minor issue in case there will be another maintenance release before pacman hits the big 4.0.0. If there isn't going to be such a thing, you might ignore this or not. lib/libalpm/dload.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/lib/libalpm/dload.c b/lib/libalpm/dload.c index a98d84a..e3d30f2 100644 --- a/lib/libalpm/dload.c +++ b/lib/libalpm/dload.c @@ -308,6 +308,7 @@ static int download_internal(const char *url, const char *localpath, _alpm_log(PM_LOG_ERROR, _("could not rename %s to %s (%s)\n"), tempfile, destfile, strerror(errno)); ret = -1; + goto cleanup; } ret = 0; -- 1.7.6

2 1

[pacman-dev] package signing for pacman
by Shaun Kluzek 25 Jul '11

25 Jul '11

Hi, my name is Shaun Kluzek. I have been using linux now for about a decade and have spent the last two years using arch linux and previously using gentoo. I studied computer science in Melbourne Australia at RMIT University and program well in C. After reading "The real story behind Arch Linux package signing" by Dan Mcgee, I headed over to Alan McRae's TODO list for package signing, and spotted the pacman signature verification, acting upon this I built the gpgme documentation from the texi files and have spent the last two days reading and understanding it( gnu don't make this very easy ). As a result, I have successfully written a small C program that verifies a detached ascii armored signature with its accompanying file against users in your keyring. The code at this stage is test code and very messy, however, it loads in a detached sig say signed.asc, then loads in the signed file somefile and then verifies it, giving the validation back in a bit vector. The code needs to be rewritten and split into some nice funtions. I have obtained the pacman source code from git and had a brief look through it just prior to writing the libgpgme stuff. Once tidied up ( into funtions and descriptive error output ) I dont think it would be much problem adding it into pacman. How should I proceed ? Shaun Kluzek

2 1

[pacman-dev] [PATCH] makepkg: allow epoch to be overridden
by Allan McRae 22 Jul '11

22 Jul '11

We can override pkgver and pkgrel so it is only logical to add epoch to that list Signed-off-by: Allan McRae <allan(a)archlinux.org> --- scripts/makepkg.sh.in | 17 ++++++++++------- 1 files changed, 10 insertions(+), 7 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 4b8f167..88afe7b 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -44,9 +44,9 @@ startdir="$PWD" packaging_options=('strip' 'docs' 'libtool' 'emptydirs' 'zipman' 'purge' 'upx') other_options=('ccache' 'distcc' 'buildflags' 'makeflags') -splitpkg_overrides=('pkgver' 'pkgrel' 'pkgdesc' 'arch' 'license' 'groups' \ - 'depends' 'optdepends' 'provides' 'conflicts' 'replaces' \ - 'backup' 'options' 'install' 'changelog') +splitpkg_overrides=('pkgver' 'pkgrel' 'epoch' 'pkgdesc' 'arch' 'license' \ + 'groups' 'depends' 'optdepends' 'provides' 'conflicts' \ + 'replaces' 'backup' 'options' 'install' 'changelog') readonly -a packaging_options other_options splitpkg_overrides # Options @@ -1455,10 +1455,13 @@ check_sanity() { fi done || ret=1 - if [[ ! $epoch =~ ^[0-9]*$ ]]; then - error "$(gettext "%s must be an integer.")" "epoch" - ret=1 - fi + awk -F'=' '/^[[:space:]]*epoch=/ { $1=""; print $0 }' "$BUILDFILE" | + while read i; do + if [[ ! $i =~ ^[0-9]*$ ]]; then + error "$(gettext "%s must be an integer.")" "epoch" + return 1 + fi + done || ret=1 if [[ $arch != 'any' ]]; then if ! in_array $CARCH ${arch[@]}; then -- 1.7.6

2 7

[pacman-dev] [PATCH] makepkg: pkgver and pkgrel can not have whitespace
by Allan McRae 22 Jul '11

22 Jul '11

There is always someone who tries to break things (cough *Dave* cough...) Signed-off-by: Allan McRae <allan(a)archlinux.org> --- scripts/makepkg.sh.in | 10 ++++++---- 1 files changed, 6 insertions(+), 4 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 0682e4f..ec5b388 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -1438,12 +1438,14 @@ check_sanity() { error "$(gettext "%s is not allowed to start with a hyphen.")" "pkgbase" ret=1 fi - if [[ $pkgver =~ [:-] ]]; then - error "$(gettext "%s is not allowed to contain colons or hyphens.")" "pkgver" + + if [[ $pkgver =~ [[:space:]:-] ]]; then + error "$(gettext "%s is not allowed to contain colons, hyphens or whitespace.")" "pkgver" ret=1 fi - if [[ $pkgrel != ${pkgrel//-/} ]]; then - error "$(gettext "%s is not allowed to contain hyphens.")" "pkgrel" + + if [[ $pkgrel =~ [[:space:]-] ]]; then + error "$(gettext "%s is not allowed to contain hyphens or whitespace.")" "pkgrel" ret=1 fi -- 1.7.6

2 3

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.5.3-505-g98073af
by dan＠archlinux.org 21 Jul '11

21 Jul '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 98073afe5527aa8a08c7119b61d0251250168e82 (commit) via 768d3589a36c1a3c86cd91c4db401bc47737c47a (commit) via 2bd1687f51b51b21fa6a28a9d91d8509fb7c7c9f (commit) via 9f500f684d7db93c235d701be8d948402fd3b7ef (commit) via bb3dada8711fbb822513cd556167867b537f8986 (commit) via 058ee1737182c2d5e900e0feba57f0d6496e735e (commit) via 70d6fe66325e66a39e6c3d79ef79c6a9ad8225f0 (commit) from 61410814c2cb33055e0eb0276444b9073c3a1e7a (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 98073afe5527aa8a08c7119b61d0251250168e82 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Wed Jul 20 13:26:44 2011 -0400 pacman-key: refactor post parse opt check into a case This is a cleaner expression of the same information. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 768d3589a36c1a3c86cd91c4db401bc47737c47a Author: Dave Reisner <dreisner(a)archlinux.org> Date: Wed Jul 20 10:47:06 2011 -0400 pacman-key: s/UPDATEBD/UPDATEDB/ Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 2bd1687f51b51b21fa6a28a9d91d8509fb7c7c9f Author: Dave Reisner <dreisner(a)archlinux.org> Date: Tue Jul 19 14:25:16 2011 -0400 pacman-key: fix syntax error in -r arg parsing Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 9f500f684d7db93c235d701be8d948402fd3b7ef Author: Dave Reisner <d(a)falconindy.com> Date: Tue Jul 19 14:28:53 2011 -0400 pacman-key: return $ret, not errors fixes: /usr/bin/pacman-key: line 286: return: errors: numeric argument required Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit bb3dada8711fbb822513cd556167867b537f8986 Author: Dan McGee <dan(a)archlinux.org> Date: Tue Jul 19 04:47:29 2011 -0500 Convert package filelists to an array instead of linked list This accomplishes quite a few things with one rather invasive change. 1. Iteration is much more performant, due to a reduction in pointer chasing and linear item access. 2. Data structures are smaller- we no longer have the overhead of the linked list as the file struts are now laid out consecutively in memory. 3. Memory allocation has been massively reworked. Before, we would allocate three different pieces of memory per file item- the list struct, the file struct, and the copied filename. What this resulted in was massive fragmentation of memory when loading filelists since the memory allocator had to leave holes all over the place. The new situation here now removes the need for any list item allocation; allocates the file structs in contiguous memory (and reallocs as necessary), leaving only the strings as individually allocated. Tests using valgrind (massif) show some pretty significant memory reductions on the worst case `pacman -Ql > /dev/null` (366387 files on my machine): Before: Peak heap: 54,416,024 B Useful heap: 36,840,692 B Extra heap: 17,575,332 B After: Peak heap: 38,004,352 B Useful heap: 28,101,347 B Extra heap: 9,903,005 B Several small helper methods have been introduced, including a list to array conversion helper as well as a filelist merge sort that works directly on arrays. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 058ee1737182c2d5e900e0feba57f0d6496e735e Author: Dan McGee <dan(a)archlinux.org> Date: Tue Jul 19 05:02:06 2011 -0500 contrib: add paclog-pkglist to gitignore Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 70d6fe66325e66a39e6c3d79ef79c6a9ad8225f0 Author: Dan McGee <dan(a)archlinux.org> Date: Tue Jul 19 00:23:42 2011 -0500 Clean up my debug logger mess Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: contrib/.gitignore | 1 + lib/libalpm/alpm.h | 11 +++++- lib/libalpm/alpm_list.c | 32 +++++++++++++++++ lib/libalpm/alpm_list.h | 1 + lib/libalpm/be_local.c | 41 +++++++++++++++++----- lib/libalpm/be_package.c | 86 +++++++++++++++++++++++++++++++++++++++------- lib/libalpm/be_sync.c | 2 - lib/libalpm/conflict.c | 65 +++++++++++++++++++--------------- lib/libalpm/conflict.h | 4 +- lib/libalpm/diskspace.c | 24 ++++++++---- lib/libalpm/package.c | 46 ++++++++++++++----------- lib/libalpm/package.h | 8 ++-- lib/libalpm/remove.c | 25 ++++++++----- scripts/pacman-key.sh.in | 29 +++++++-------- src/pacman/package.c | 9 +++-- src/pacman/query.c | 19 ++++++---- 16 files changed, 277 insertions(+), 126 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH 0/6] Initial patches for better optdep support
by Benedikt Morbach 21 Jul '11

21 Jul '11

Hi, I've send this before, but it seems like it didn't reach the list. I apologize if I'm flooding your inbox and you got this patch series twice. The pacman codebase is completely new to me, so please forgive me the obvious mistakes that I overlooked and feel free to point them out and yell at me. This implements most features from https://wiki.archlinux.org/index.php/User:Allan/Pacman_OptDepends It probably isn't quite ready to be merged yet, but I want to gather some feedback on the stuff I've already done. What is there: - No regressions afaics - Only show uninstalled optdepends during install/upgrade - In package info, show [installed] after installed optdepends - In package info (local or -ii) show packages which optionally depend on the queried package - '-Qt' doesn't consider optdepends to be orphans, unless '--nooptdeps/-n' is given What is missing: - Show which of the packages listed by '-Qtn' is an optdepend and what optdeps on it - In package info display the description alongside the "reverse optdeps" - Recursive removal of unneeded optdeps - Anything listed under "Other Ideas" - Tests (mainly needed for the package removal stuff I think) - Docs with less sucky english ;-) -- Benedikt Benedikt Morbach (6): 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 --nooptdepends is specified Make package info show optional requirements doc/pacman.8.txt | 6 ++++ lib/libalpm/alpm.h | 14 ++++++++- lib/libalpm/be_local.c | 11 +++++- lib/libalpm/be_package.c | 5 ++- lib/libalpm/be_sync.c | 7 ++++- lib/libalpm/deps.c | 70 ++++++++++++++++++++++++++++++++++++++++++ lib/libalpm/deps.h | 3 ++ lib/libalpm/package.c | 39 ++++++++++++++++------- src/pacman/conf.h | 1 + src/pacman/package.c | 13 ++++++-- src/pacman/pacman.c | 3 ++ src/pacman/query.c | 16 ++++++++-- src/pacman/util.c | 75 +++++++++++++++++++++++++++++++++++++++------- src/pacman/util.h | 1 + src/util/pactree.c | 2 +- 15 files changed, 230 insertions(+), 36 deletions(-) -- 1.7.6

2 11

[pacman-dev] [PATCH] Change GPG_PACMAN and GPG_NOKEYRING to arrays
by DJ Mills 20 Jul '11

20 Jul '11

Allow the commands to safely handle any possible arguments. Signed-off-by: DJ Mills <danielmills1(a)gmail.com> --- scripts/pacman-key.sh.in | 48 +++++++++++++++++++++++----------------------- 1 files changed, 24 insertions(+), 24 deletions(-) diff --git a/scripts/pacman-key.sh.in b/scripts/pacman-key.sh.in index 833943c..5ad83c3 100644 --- a/scripts/pacman-key.sh.in +++ b/scripts/pacman-key.sh.in @@ -78,7 +78,7 @@ get_from() { reload_keyring() { local PACMAN_SHARE_DIR='@prefix@/share/pacman' - local GPG_NOKEYRING="gpg --batch --quiet --ignore-time-conflict --no-options --no-default-keyring --homedir ${PACMAN_KEYRING_DIR}" + local GPG_NOKEYRING=(gpg --batch --quiet --ignore-time-conflict --no-options --no-default-keyring --homedir "${PACMAN_KEYRING_DIR}") # Variable used for iterating on keyrings local key @@ -97,7 +97,7 @@ reload_keyring() { # Verify signatures of related files, if they exist if [[ -r "${ADDED_KEYS}" ]]; then msg "$(gettext "Verifying official keys file signature...")" - if ! ${GPG_PACMAN} --quiet --batch --verify "${ADDED_KEYS}.sig" 1>/dev/null; then + if ! "${GPG_PACMAN[@]}" --quiet --batch --verify "${ADDED_KEYS}.sig" 1>/dev/null; then error "$(gettext "The signature of file %s is not valid.")" "${ADDED_KEYS}" exit 1 fi @@ -105,7 +105,7 @@ reload_keyring() { if [[ -r "${DEPRECATED_KEYS}" ]]; then msg "$(gettext "Verifying deprecated keys file signature...")" - if ! ${GPG_PACMAN} --quiet --batch --verify "${DEPRECATED_KEYS}.sig" 1>/dev/null; then + if ! "${GPG_PACMAN[@]}" --quiet --batch --verify "${DEPRECATED_KEYS}.sig" 1>/dev/null; then error "$(gettext "The signature of file %s is not valid.")" "${DEPRECATED_KEYS}" exit 1 fi @@ -113,7 +113,7 @@ reload_keyring() { if [[ -r "${REMOVED_KEYS}" ]]; then msg "$(gettext "Verifying deleted keys file signature...")" - if ! ${GPG_PACMAN} --quiet --batch --verify "${REMOVED_KEYS}.sig"; then + if ! "${GPG_PACMAN[@]}" --quiet --batch --verify "${REMOVED_KEYS}.sig"; then error "$(gettext "The signature of file %s is not valid.")" "${REMOVED_KEYS}" exit 1 fi @@ -126,7 +126,7 @@ reload_keyring() { if [[ -r "${REMOVED_KEYS}" ]]; then while read key; do local key_values name - key_values=$(${GPG_PACMAN} --quiet --with-colons --list-key "${key}" | grep ^pub | cut -d: -f5,10 --output-delimiter=' ') + key_values=$("${GPG_PACMAN[@]}" --quiet --with-colons --list-key "${key}" | grep ^pub | cut -d: -f5,10 --output-delimiter=' ') if [[ -n $key_values ]]; then # The first word is the key_id key_id=${key_values%% *} @@ -146,7 +146,7 @@ reload_keyring() { # Remove the keys that must be kept from the set of keys that should be removed if [[ -n ${HOLD_KEYS} ]]; then for key in ${HOLD_KEYS}; do - key_id=$(${GPG_PACMAN} --quiet --with-colons --list-key "${key}" | grep ^pub | cut -d: -f5) + key_id=$("${GPG_PACMAN[@]}" --quiet --with-colons --list-key "${key}" | grep ^pub | cut -d: -f5) if [[ -n "${removed_ids[$key_id]}" ]]; then unset removed_ids[$key_id] fi @@ -157,22 +157,22 @@ reload_keyring() { # be updated automatically. if [[ -r "${ADDED_KEYS}" ]]; then msg "$(gettext "Appending official keys...")" - local add_keys=$(${GPG_NOKEYRING} --keyring "${ADDED_KEYS}" --with-colons --list-keys | grep ^pub | cut -d: -f5) + local add_keys=$("${GPG_NOKEYRING[@]}" --keyring "${ADDED_KEYS}" --with-colons --list-keys | grep ^pub | cut -d: -f5) for key_id in ${add_keys}; do # There is no point in adding a key that will be deleted right after if [[ -z "${removed_ids[$key_id]}" ]]; then - ${GPG_NOKEYRING} --keyring "${ADDED_KEYS}" --export "${key_id}" | ${GPG_PACMAN} --import + "${GPG_NOKEYRING[@]}" --keyring "${ADDED_KEYS}" --export "${key_id}" | "${GPG_PACMAN[@]}" --import fi done fi if [[ -r "${DEPRECATED_KEYS}" ]]; then msg "$(gettext "Appending deprecated keys...")" - local add_keys=$(${GPG_NOKEYRING} --keyring "${DEPRECATED_KEYS}" --with-colons --list-keys | grep ^pub | cut -d: -f5) + local add_keys=$("${GPG_NOKEYRING[@]}" --keyring "${DEPRECATED_KEYS}" --with-colons --list-keys | grep ^pub | cut -d: -f5) for key_id in ${add_keys}; do # There is no point in adding a key that will be deleted right after if [[ -z "${removed_ids[$key_id]}" ]]; then - ${GPG_NOKEYRING} --keyring "${DEPRECATED_KEYS}" --export "${key_id}" | ${GPG_PACMAN} --import + "${GPG_NOKEYRING[@]}" --keyring "${DEPRECATED_KEYS}" --export "${key_id}" | "${GPG_PACMAN[@]}" --import fi done fi @@ -182,13 +182,13 @@ reload_keyring() { msg "$(gettext "Removing deleted keys from keyring...")" for key_id in "${!removed_ids[@]}"; do echo " removing key $key_id - ${removed_ids[$key_id]}" - ${GPG_PACMAN} --quiet --batch --yes --delete-key "${key_id}" + "${GPG_PACMAN[@]}" --quiet --batch --yes --delete-key "${key_id}" done fi # Update trustdb, just to be sure msg "$(gettext "Updating trust database...")" - ${GPG_PACMAN} --batch --check-trustdb + "${GPG_PACMAN[@]}" --batch --check-trustdb } # PROGRAM START @@ -229,7 +229,7 @@ fi if [[ GPGDIR=$(get_from "$CONFIG" "GPGDir") == 0 ]]; then PACMAN_KEYRING_DIR="${GPGDIR}" fi -GPG_PACMAN="gpg --homedir ${PACMAN_KEYRING_DIR} --no-permission-warning" +GPG_PACMAN=(gpg --homedir "${PACMAN_KEYRING_DIR}" --no-permission-warning) # Try to create $PACMAN_KEYRING_DIR if non-existent # Check for simple existence rather than for a directory as someone may want @@ -247,29 +247,29 @@ shift case "${command}" in -a|--add) # If there is no extra parameter, gpg will read stdin - ${GPG_PACMAN} --quiet --batch --import "$@" + "${GPG_PACMAN[@]}" --quiet --batch --import "$@" ;; -d|--del) if (( $# == 0 )); then error "$(gettext "You need to specify at least one key identifier")" exit 1 fi - ${GPG_PACMAN} --quiet --batch --delete-key --yes "$@" + "${GPG_PACMAN[@]}" --quiet --batch --delete-key --yes "$@" ;; -u|--updatedb) - ${GPG_PACMAN} --batch --check-trustdb + "${GPG_PACMAN[@]}" --batch --check-trustdb ;; --reload) reload_keyring ;; -l|--list) - ${GPG_PACMAN} --batch --list-sigs "$@" + "${GPG_PACMAN[@]}" --batch --list-sigs "$@" ;; -f|--finger) - ${GPG_PACMAN} --batch --fingerprint "$@" + "${GPG_PACMAN[@]}" --batch --fingerprint "$@" ;; -e|--export) - ${GPG_PACMAN} --armor --export "$@" + "${GPG_PACMAN[@]}" --armor --export "$@" ;; -r|--receive) if (( $# < 2 )); then @@ -278,7 +278,7 @@ case "${command}" in fi keyserver="$1" shift - ${GPG_PACMAN} --keyserver "${keyserver}" --recv-keys "$@" + "${GPG_PACMAN[@]}" --keyserver "${keyserver}" --recv-keys "$@" ;; -t|--trust) if (( $# == 0 )); then @@ -287,8 +287,8 @@ case "${command}" in fi while (( $# > 0 )); do # Verify if the key exists in pacman's keyring - if ${GPG_PACMAN} --list-keys "$1" > /dev/null 2>&1; then - ${GPG_PACMAN} --edit-key "$1" + if "${GPG_PACMAN[@]}" --list-keys "$1" > /dev/null 2>&1; then + "${GPG_PACMAN[@]}" --edit-key "$1" else error "$(gettext "The key identified by %s doesn't exist")" "$1" exit 1 @@ -297,8 +297,8 @@ case "${command}" in done ;; --adv) - msg "$(gettext "Executing: %s ")$*" "${GPG_PACMAN}" - ${GPG_PACMAN} "$@" || ret=$? + msg "$(gettext "Executing: %s ")$*" ""${GPG_PACMAN[@]}"" + "${GPG_PACMAN[@]}" "$@" || ret=$? exit $ret ;; -h|--help) -- 1.7.6

2 1