- Pacman-dev - lists.archlinux.org

[pacman-dev] [PATCH] Add config var to limit repo connection timeout
by Nathan Aclander 07 Apr '19

07 Apr '19

New ConnectionTimeout variable sets the amount of time trying to connect to a repository, before timing out and trying the next one. Signed-off-by: Nathan Aclander <nathan.aclander(a)gmail.com> --- doc/pacman.conf.5.asciidoc | 3 +++ etc/pacman.conf.in | 1 + lib/libalpm/alpm.h | 3 +++ lib/libalpm/dload.c | 2 +- lib/libalpm/handle.c | 7 +++++++ lib/libalpm/handle.h | 1 + src/pacman/conf.c | 7 +++++++ src/pacman/conf.h | 1 + 8 files changed, 24 insertions(+), 1 deletion(-) diff --git a/doc/pacman.conf.5.asciidoc b/doc/pacman.conf.5.asciidoc index b297e332..dee5139e 100644 --- a/doc/pacman.conf.5.asciidoc +++ b/doc/pacman.conf.5.asciidoc @@ -205,6 +205,9 @@ Options Disable defaults for low speed limit and timeout on downloads. Use this if you have issues downloading files with proxy and/or security gateway. +*ConnectionTimeout*:: + Set the amount of time trying to connect to a repository, before timnig + out and trying the next one. Repository Sections ------------------- diff --git a/etc/pacman.conf.in b/etc/pacman.conf.in index 7446944f..899ad18f 100644 --- a/etc/pacman.conf.in +++ b/etc/pacman.conf.in @@ -34,6 +34,7 @@ Architecture = auto #TotalDownload CheckSpace #VerbosePkgLists +#ConnectionTimeout = 10 # PGP signature checking #SigLevel = Optional diff --git a/lib/libalpm/alpm.h b/lib/libalpm/alpm.h index ffb2ad96..e0bf12ca 100644 --- a/lib/libalpm/alpm.h +++ b/lib/libalpm/alpm.h @@ -890,6 +890,9 @@ int alpm_option_set_remote_file_siglevel(alpm_handle_t *handle, int level); int alpm_option_set_disable_dl_timeout(alpm_handle_t *handle, unsigned short disable_dl_timeout); +/** Sets the timeout for the curl connect phase. */ +int alpm_option_set_connecttimeout(alpm_handle_t *handle, unsigned int timeout); + /** @} */ /** @addtogroup alpm_api_databases Database Functions diff --git a/lib/libalpm/dload.c b/lib/libalpm/dload.c index 05813c40..4e0703b4 100644 --- a/lib/libalpm/dload.c +++ b/lib/libalpm/dload.c @@ -258,7 +258,7 @@ static void curl_set_handle_opts(struct dload_payload *payload, curl_easy_reset(curl); curl_easy_setopt(curl, CURLOPT_URL, payload->fileurl); curl_easy_setopt(curl, CURLOPT_ERRORBUFFER, error_buffer); - curl_easy_setopt(curl, CURLOPT_CONNECTTIMEOUT, 10L); + curl_easy_setopt(curl, CURLOPT_CONNECTTIMEOUT, handle->connecttimeout); curl_easy_setopt(curl, CURLOPT_MAXREDIRS, 10L); curl_easy_setopt(curl, CURLOPT_FILETIME, 1L); curl_easy_setopt(curl, CURLOPT_NOPROGRESS, 0L); diff --git a/lib/libalpm/handle.c b/lib/libalpm/handle.c index 476779c4..92bc9038 100644 --- a/lib/libalpm/handle.c +++ b/lib/libalpm/handle.c @@ -556,6 +556,13 @@ int SYMEXPORT alpm_option_set_usesyslog(alpm_handle_t *handle, int usesyslog) return 0; } +int SYMEXPORT alpm_option_set_connecttimeout(alpm_handle_t *handle, unsigned int timeout) +{ + CHECK_HANDLE(handle, return -1); + handle->connecttimeout = timeout; + return 0; +} + static int _alpm_option_strlist_add(alpm_handle_t *handle, alpm_list_t **list, const char *str) { char *dup; diff --git a/lib/libalpm/handle.h b/lib/libalpm/handle.h index 14d20bbe..9defbc1b 100644 --- a/lib/libalpm/handle.h +++ b/lib/libalpm/handle.h @@ -103,6 +103,7 @@ struct __alpm_handle_t { upgrade operations */ int remotefilesiglevel; /* Signature verification level for remote file upgrade operations */ + unsigned int connecttimeout; /* timeout for the connect phase */ /* error code */ alpm_errno_t pm_errno; diff --git a/src/pacman/conf.c b/src/pacman/conf.c index 3b79fbc7..2cdc157e 100644 --- a/src/pacman/conf.c +++ b/src/pacman/conf.c @@ -513,6 +513,8 @@ static int _parse_options(const char *key, char *value, setrepeatingoption(value, "IgnoreGroup", &(config->ignoregrp)); } else if(strcmp(key, "HoldPkg") == 0) { setrepeatingoption(value, "HoldPkg", &(config->holdpkg)); + } else if(strcmp(key, "ConnectionTimeout") == 0) { + config->connecttimeout = atoi(value); } else if(strcmp(key, "CacheDir") == 0) { setrepeatingoption(value, "CacheDir", &(config->cachedirs)); } else if(strcmp(key, "HookDir") == 0) { @@ -739,6 +741,10 @@ static int setup_libalpm(void) alpm_option_set_totaldlcb(handle, cb_dl_total); } + if(!config->connecttimeout) { + alpm_option_set_connecttimeout(handle, 10L); + } + alpm_option_set_arch(handle, config->arch); alpm_option_set_checkspace(handle, config->checkspace); alpm_option_set_usesyslog(handle, config->usesyslog); @@ -749,6 +755,7 @@ static int setup_libalpm(void) alpm_option_set_noextracts(handle, config->noextract); alpm_option_set_disable_dl_timeout(handle, config->disable_dl_timeout); + alpm_option_set_connecttimeout(handle, config->connecttimeout); for(i = config->assumeinstalled; i; i = i->next) { char *entry = i->data; diff --git a/src/pacman/conf.h b/src/pacman/conf.h index f45ed436..e1f9934e 100644 --- a/src/pacman/conf.h +++ b/src/pacman/conf.h @@ -118,6 +118,7 @@ typedef struct __config_t { /* select -Sc behavior */ unsigned short cleanmethod; alpm_list_t *holdpkg; + unsigned int connecttimeout; alpm_list_t *ignorepkg; alpm_list_t *ignoregrp; alpm_list_t *assumeinstalled; -- 2.21.0

4 6

[pacman-dev] [PATCH] Style: Removing unecessary indentation level.
by Leonardo Bras 06 Apr '19

06 Apr '19

Signed-off-by: Leonardo Bras <leobras.c(a)gmail.com> --- src/pacman/sync.c | 84 ++++++++++++++++++++++++----------------------- 1 file changed, 43 insertions(+), 41 deletions(-) diff --git a/src/pacman/sync.c b/src/pacman/sync.c index 7dc69079..9e35053f 100644 --- a/src/pacman/sync.c +++ b/src/pacman/sync.c @@ -816,49 +816,51 @@ int sync_prepare_execute(void) goto cleanup; } - if(alpm_trans_commit(config->handle, &data) == -1) { - alpm_errno_t err = alpm_errno(config->handle); - pm_printf(ALPM_LOG_ERROR, _("failed to commit transaction (%s)\n"), - alpm_strerror(err)); - switch(err) { - case ALPM_ERR_FILE_CONFLICTS: - for(i = data; i; i = alpm_list_next(i)) { - alpm_fileconflict_t *conflict = i->data; - switch(conflict->type) { - case ALPM_FILECONFLICT_TARGET: - printf(_("%s exists in both '%s' and '%s'\n"), - conflict->file, conflict->target, conflict->ctarget); - break; - case ALPM_FILECONFLICT_FILESYSTEM: - if(conflict->ctarget[0]) { - printf(_("%s: %s exists in filesystem (owned by %s)\n"), - conflict->target, conflict->file, conflict->ctarget); - } else { - printf(_("%s: %s exists in filesystem\n"), - conflict->target, conflict->file); - } - break; - } - alpm_fileconflict_free(conflict); - } - break; - case ALPM_ERR_PKG_INVALID: - case ALPM_ERR_PKG_INVALID_CHECKSUM: - case ALPM_ERR_PKG_INVALID_SIG: - for(i = data; i; i = alpm_list_next(i)) { - char *filename = i->data; - printf(_("%s is invalid or corrupted\n"), filename); - free(filename); - } - break; - default: - break; - } - /* TODO: stderr? */ - printf(_("Errors occurred, no packages were upgraded.\n")); - retval = 1; + if(alpm_trans_commit(config->handle, &data) != -1) goto cleanup; + + alpm_errno_t err = alpm_errno(config->handle); + pm_printf(ALPM_LOG_ERROR, _("failed to commit transaction (%s)\n"), + alpm_strerror(err)); + switch(err) { + case ALPM_ERR_FILE_CONFLICTS: + for(i = data; i; i = alpm_list_next(i)) { + alpm_fileconflict_t *conflict = i->data; + switch(conflict->type) { + case ALPM_FILECONFLICT_TARGET: + printf(_("%s exists in both '%s' and '%s'\n"), + conflict->file, conflict->target, conflict->ctarget); + break; + case ALPM_FILECONFLICT_FILESYSTEM: + if(conflict->ctarget[0]) { + printf(_("%s: %s exists in filesystem (owned by %s)\n"), + conflict->target, conflict->file, conflict->ctarget); + } else { + printf(_("%s: %s exists in filesystem\n"), + conflict->target, conflict->file); + } + break; + } + alpm_fileconflict_free(conflict); + } + break; + case ALPM_ERR_PKG_INVALID: + case ALPM_ERR_PKG_INVALID_CHECKSUM: + case ALPM_ERR_PKG_INVALID_SIG: + for(i = data; i; i = alpm_list_next(i)) { + char *filename = i->data; + printf(_("%s is invalid or corrupted\n"), filename); + free(filename); + } + break; + default: + break; } + /* TODO: stderr? */ + printf(_("Errors occurred, no packages were upgraded.\n")); + retval = 1; + goto cleanup; + /* Step 4: release transaction resources */ cleanup: -- 2.21.0

2 1

[pacman-dev] [PATCH] libmakepkg: fix migration to schema.sh for integsums
by Eli Schwartz 31 Mar '19

31 Mar '19

One of the callers was changed to use known_hash_algos, one was not. Signed-off-by: Eli Schwartz <eschwartz(a)archlinux.org> --- scripts/libmakepkg/srcinfo.sh.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/libmakepkg/srcinfo.sh.in b/scripts/libmakepkg/srcinfo.sh.in index b9eeb817..76cfb787 100644 --- a/scripts/libmakepkg/srcinfo.sh.in +++ b/scripts/libmakepkg/srcinfo.sh.in @@ -90,7 +90,7 @@ srcinfo_write_global() { local multivalued=(arch groups license checkdepends makedepends depends optdepends provides conflicts replaces noextract options backup - source validpgpkeys {md5,sha{1,224,256,384,512}}sums) + source validpgpkeys "${known_hash_algos[@]/%/sums}") srcinfo_open_section 'pkgbase' "${pkgbase:-$pkgname}" srcinfo_write_section_details '' -- 2.21.0

1 0

[pacman-dev] [PATCH] libalpm: write {check, make}depends to localdb when installing package
by Denis Ollier 29 Mar '19

29 Mar '19

It is a follow up of commits 0994893 and f9eb2aa. Signed-off-by: Denis Ollier <larchunix(a)gmail.com> --- lib/libalpm/be_local.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/lib/libalpm/be_local.c b/lib/libalpm/be_local.c index 5a03f257..a4b72776 100644 --- a/lib/libalpm/be_local.c +++ b/lib/libalpm/be_local.c @@ -1033,6 +1033,8 @@ int _alpm_local_db_write(alpm_db_t *db, alpm_pkg_t *info, int inforeq) write_deps(fp, "%REPLACES%", info->replaces); write_deps(fp, "%DEPENDS%", info->depends); write_deps(fp, "%OPTDEPENDS%", info->optdepends); + write_deps(fp, "%MAKEDEPENDS%", info->makedepends); + write_deps(fp, "%CHECKDEPENDS%", info->checkdepends); write_deps(fp, "%CONFLICTS%", info->conflicts); write_deps(fp, "%PROVIDES%", info->provides); -- 2.21.0

3 2

[pacman-dev] should repo-add handle *.links.db?
by Erich Eckner 27 Mar '19

27 Mar '19

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Hi, I'm looking into creating links databases for archlinux32 and saw (thanks, Eli) that they are created asynchronously by systemd timers here: https://git.archlinux.org/infrastructure.git/tree/roles/sogrep/files I was wondering, if it wouldn't be better to create them during repo-add (and repo-remove) at a similar point, where *.files.db entries are created - - ignoring deltas (can we ignore deltas already?), this would be somewhere down in db_write_entry() in repo-add - but even including deltas, it should not need changes at too many places. However, before I start trying to patch repo-add and repo-remove to maintain a *.links.db, I wanted to ask if there are any objections to include this functionality there. Cheers, Erich -----BEGIN PGP SIGNATURE----- iQIzBAEBCAAdFiEE3p92iMrPBP64GmxZCu7JB1Xae1oFAlyaLrQACgkQCu7JB1Xa e1o26g/+PU80nwWnMcT/zwhzWuuTAP8hwIlb1dvovfFp5vkq9XjPC/A3yApfSg6m nUPb7Vjq9w3KGoKCMTwBWYO6jXLjLhvcB2gSBmhFlblDrMgGTnABrjNdfcaSUwcB hqF7cYDhSDJuUb+SyMBYKLoGDZL/Rxye/BAItiqBXfgitbxAcPWbNyZCKk3DYOBz /WHCHn3hMHJndxC7FJ9hfpGukES9Um2EW4bB8DGrz44AJwzyA8EyDO9e86QtR64z PX6qC2qEuAix55Ggvk2GILErPaea/5IuwhmyrZy6UtTEDysDfdkMSKWqV+r8wHJY 61fO7RGZnJH6F4h6YEpTncdEI4jnuIuR7FrheWauV83jMJ4E/X4nvNTNBlPDZUIt DC8Nj1nsSrTeD7HZmi6QiKHhPLpA5ZiE0ls+jdb/CcE0RaB+kqRtMRShanKDNiJi E4lhLO7nj6frNtpWhU3Iqb2O7azampIpL7VyCHDAfkuWPrIHfMGSRnwWmG2DW0G/ UTxHqDKqVFTDwBF2h6+4+J/xFO2T1uqm5OrAnR9ZsjGG6UMZEbsNvO0KxFiOdlYc UFPJJlwVn7OS28H94010tRU73LfE1xwZqRKiWxiR2NckjB9fvOf5lPUV5neZfeS1 O9yS9omfa9QzVH6dD/wDJArnt4RGj8mHBEXa3vwOZGEfXaF+waQ= =ny8G -----END PGP SIGNATURE-----

3 7

[pacman-dev] Man page translations
by Mario Blättermann 25 Mar '19

25 Mar '19

Hello, I've just read at [1] that nobody is working on man page translations for Pacman, but translations would be welcome anyway. That's great, because there are only a few developers who are really willing to consider localized man pages as an important part of a good user experience. I've already translated the current man pages based on the latest release into German, and they are currently maintained in the manpages-de project [2]. But this external approach has a few disadvantages: There's some delay between upstream release of the packages where we get the original man pages from, then translation, release of a tarball and finally packaging for the currently upported distributions (Archlinux and Debian unstable, support for Mageia is under construction). Moreover, we translate only into *one* language... It would be better to maintain the localized manpages in Pacman's git repo itself. We don't have that many options to setup i18n for man pages. In my opinion po4a [3] is the best software which can handle such things. We use it also in the manpages-de project. Well, we use it there with Groff and Mdoc, but it is also able to work with Asciidoc (says the documentation, I haven't tested it). I could setup a po4a toolchain in Pacman which takes the asciidoc files, puts the contents in pot templates and merges then the resuting po files back to the asciidoc source format. I've also some basic experience with Meson, but I'm afraid I won't be able to get it working with the GNU Autotools. The translations itself can be maintained at Transifex, as usual, but of course in separate templates for each asiidoc source file. What do you think about this idea? If it's OK for you, then I need some playground for my changes. I consider to fork the git repo at Github. @Allan, it is OK to fork your personal repo, or should I use the main repo at git.archlinux.org? Processing merge requests is maybe easier within Github. [1] https://www.archlinux.org/pacman/translation-help.html [2] https://salsa.debian.org/manpages-de-team/manpages-de [3]https://po4a.org/index.php Best Regards, Mario

1 0

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v5.1.1-160-gd37e6d40
by Allan McRae 25 Mar '19

25 Mar '19

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 d37e6d40a19dadf41a223dcb1fde36c6fa622958 (commit) via 7a598d0845fda08e195e1a709297c761b68d79f4 (commit) from 9f5412d809a50a7ad18013222582d5b943216ae1 (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 d37e6d40a19dadf41a223dcb1fde36c6fa622958 Author: Eli Schwartz <eschwartz(a)archlinux.org> Date: Thu Mar 7 23:22:49 2019 -0500 bash-completion: support file redirection completions The current completions don't properly handle redirection operators, and attempt to complete command completions rather than completing filenames to redirect to. bash-completion provides both _get_comp_words_by_ref and a higher-level wrapper _init_completion, but the latter provides handling of redirection operators, so switch to using that. Signed-off-by: Eli Schwartz <eschwartz(a)archlinux.org> Signed-off-by: Allan McRae <allan(a)archlinux.org> commit 7a598d0845fda08e195e1a709297c761b68d79f4 Author: Allan McRae <allan(a)archlinux.org> Date: Wed Mar 20 09:58:16 2019 +1000 Use bug tracker to track bugs The mailing list requires subscription. So does the bug tracker, but that is more obvious. Signed-off-by: Allan McRae <allan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: doc/footer.asciidoc | 5 +++-- scripts/completion/bash_completion.in | 18 +++++++++--------- 2 files changed, 12 insertions(+), 11 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] ask permission for forking pacman file manager
by dimension breaker 23 Mar '19

23 Mar '19

introduce me Purnomo Hadi I am the Chief Developer at Dracos-linux, one of the founders of LFS Indonesia. founder and developer kolomonggo linux dracOS linux and kolomonggo linux is a LFS based OS https://dracos-linux.org/ the purpose of sending this email is for forking pacman package manager to be a custom file manager with the name we will use. I am as the developer of dracOs linux and columnonggo linux request permission from the Pacman Development Team Thank you for your attention sory for my bad english -- dracOs linux - dev team -- dracOs linux - dev team

3 2

[pacman-dev] Version comparison algorithm.
by Richard Dodd 20 Mar '19

20 Mar '19

I'm writing to ask about the version comparison function `rpmvercmp` in `libalpm/version.cL83`. The algorithm is complex and I'm trying to understand its behavior. I want to hash using the package name and version as a key, and so I need a hash function where `i1 == i2 => hash(i1) == hash(i2)` according to the version comparison operation. I'll describe how the algorithm behaves and then ask my questions. The algorithm works on a byte string and uses ascii comparison rules (no unicode). - First, split the input up into blocks of *alpha*, *digit* or *non-alphanum*. - For each pair of blocks - If the types are different, then *non-alphanum* is newer than *numeric*, which is newer than *alpha* - If the types are the same, then the rules are - For *non-alphanum*, compare lengths, longer is newer, equal lengths are equal segments (so *--* and *::* are the same) - For *alpha* just do a lexicographic comparison (so *b* is newer than *a* etc.) - For *numeric*, do a numeric comparison. (this can be done by skipping leading zeros, then comparing length, then comparing lexicographically, to avoid overflow on integer conversion) - If one input is longer than the other, and all sections so far have been equal, then if the next section of the longer is *alpha*, it is older, and if it is *numeric* it is newer. (so "1a" is older than "1", "a1" is newer than "a"). - If the inputs have the same number of sections that are all equal, then they are equal. Questions: 1. Is the algorithm correctly described here. 2. This should mean that if I hash length for *non-numeric*, the string with stripped zeros for *numeric*, and just the string for *alpha*, the has law should be upheld. Is this correct? Thanks Rich

4 4

[pacman-dev] [PATCH] Use bug tracker to track bugs
by Allan McRae 19 Mar '19

19 Mar '19

The mailing list requires subscription. So does the bug tracker, but that is more obvious. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- doc/footer.asciidoc | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/doc/footer.asciidoc b/doc/footer.asciidoc index 5812aa72..cdaba40c 100644 --- a/doc/footer.asciidoc +++ b/doc/footer.asciidoc @@ -6,8 +6,9 @@ information on pacman and its related tools. Bugs ---- Bugs? You must be kidding; there are no bugs in this software. But if we -happen to be wrong, send us an email with as much detail as possible to -mailto:pacman-dev@archlinux.org[]. +happen to be wrong, submit a bug report with as much detail as possible +at the https://bugs.archlinux.org/index.php?project=3[Arch Linux Bug Tracker] +in the Pacman section. Authors -- 2.21.0

1 0