January 2011 - Pacman-dev - lists.archlinux.org

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.4.2-256-g3863e48
by dan＠archlinux.org 21 Jan '11

21 Jan '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 3863e48788e2424c4f1469e72dabda6fba595c4c (commit) via c86ff120c870a2d0e0427567ae7449e23ad42954 (commit) via 01c8f39ab8db36ad7ac1c2078a8c1a5b7504192a (commit) via aff3e63c451ead18571a8abed56e911f71906fe6 (commit) via b40c8e292244c497ddb557428b8bd152ed7f0482 (commit) via c49f1980429f88cd0a73340df7c1e9d02e7026a7 (commit) via f65edb7f299537357354b5bd637353db6c7ddcd1 (commit) via a88cb03a58909304603a794e42070656fca6b626 (commit) from 665528d7bac99228849d1162f78c0817bc619eee (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 3863e48788e2424c4f1469e72dabda6fba595c4c Author: Dan McGee <dan(a)archlinux.org> Date: Fri Jan 21 10:11:59 2011 -0600 Add a likely_pkg hint argument to sync_db_read In most (all?) cases, we will process all files for a given sync database entry sequentially. The code currently does an _alpm_pkg_find() for every file in the database, but we had the "current" package readily available. Shift some local variables around a bit to expose this to sync_db_read() and use it if the package is the correct one. On my system, this cuts calls to _alpm_pkg_find() from 20,769 to 10,349 calls during a -Qu operation, and results in a ~30% speedup of the same operation (0.35 sec -> 0.27 sec). This benefit should be apparent anywhere we read in the full contents of the sync databases. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit c86ff120c870a2d0e0427567ae7449e23ad42954 Author: Dan McGee <dan(a)archlinux.org> Date: Wed Jan 19 12:20:32 2011 -0600 Improve splitname memory allocation We don't need to create a temporary copy of the string if we are smart with our pointer manipulation and string copying. This saves a bunch of string duplication during database parsing, both local and sync. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 01c8f39ab8db36ad7ac1c2078a8c1a5b7504192a Author: Dan McGee <dan(a)archlinux.org> Date: Wed Jan 19 12:00:39 2011 -0600 Improve depends string parsing Remove the need for an unconditional string duplication by using pointer arithmetic instead, and strndup() instead of an unspecified-length strdup(). This should reduce memory churn a fair amount as this is called pretty frequently during database loads. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit aff3e63c451ead18571a8abed56e911f71906fe6 Author: Dan McGee <dan(a)archlinux.org> Date: Wed Jan 19 12:00:39 2011 -0600 Add strndup fallback function to libalpm util The same fallback we are currently using in the pacman frontend. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit b40c8e292244c497ddb557428b8bd152ed7f0482 Author: Dan McGee <dan(a)archlinux.org> Date: Wed Jan 19 12:22:33 2011 -0600 Update valgrind suppressions file We haven't tweaked this in a while, but some of the old stuff seems to no longer be necessary and there are a few new things we should add. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit c49f1980429f88cd0a73340df7c1e9d02e7026a7 Author: Dan McGee <dan(a)archlinux.org> Date: Thu Jan 20 15:41:03 2011 -0600 Add a pactest for the situation in FS#7524 Confirming the current behavior. And yes, the error message is still no better than it was when this was reported 3.5 years ago. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit f65edb7f299537357354b5bd637353db6c7ddcd1 Author: Dan McGee <dan(a)archlinux.org> Date: Tue Jan 18 13:19:25 2011 -0600 Fix integrity check status when installing from file When installing packages from a file, the integrity check count stays at (0/x) complete. This ensures it is bumped to (x/x) at the end of the process. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit a88cb03a58909304603a794e42070656fca6b626 Author: Dan McGee <dan(a)archlinux.org> Date: Mon Jan 10 13:56:27 2011 -0600 makepkg: perform all sanity checks before erroring out It is pretty annoying to get one, fix it, and then get another. We should be able to continue on through most of the sanity checks in one go so the user gets all the error messages at once. Also ensure $pkgbase is defined by the time we call this function; previously we printed nothing where a package name should have been due to this oversight. Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: lib/libalpm/be_sync.c | 26 ++++++++++++-------- lib/libalpm/deps.c | 53 +++++++++++++++-------------------------- lib/libalpm/sync.c | 3 +- lib/libalpm/util.c | 49 +++++++++++++++++++++++++++----------- lib/libalpm/util.h | 5 ++++ scripts/makepkg.sh.in | 37 +++++++++++++++-------------- test/pacman/tests/sync407.py | 38 ++++++++++++++++++++++++++++++ valgrind.supp | 27 +++++++++++---------- 8 files changed, 148 insertions(+), 90 deletions(-) create mode 100644 test/pacman/tests/sync407.py hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH 1/3] Add strndup fallback function to libalpm util
by Dan McGee 20 Jan '11

20 Jan '11

The same fallback we are currently using in the pacman frontend. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- lib/libalpm/util.c | 22 ++++++++++++++++++++++ lib/libalpm/util.h | 5 +++++ 2 files changed, 27 insertions(+), 0 deletions(-) diff --git a/lib/libalpm/util.c b/lib/libalpm/util.c index 2eee5e4..81d950e 100644 --- a/lib/libalpm/util.c +++ b/lib/libalpm/util.c @@ -937,4 +937,26 @@ long _alpm_parsedate(const char *line) return(atol(line)); } +#ifndef HAVE_STRNDUP +/* A quick and dirty implementation derived from glibc */ +static size_t strnlen(const char *s, size_t max) +{ + register const char *p; + for(p = s; *p && max--; ++p); + return(p - s); +} + +char *strndup(const char *s, size_t n) +{ + size_t len = strnlen(s, n); + char *new = (char *) malloc(len + 1); + + if (new == NULL) + return NULL; + + new[len] = '\0'; + return (char *) memcpy(new, s, len); +} +#endif + /* vim: set ts=2 sw=2 noet: */ diff --git a/lib/libalpm/util.h b/lib/libalpm/util.h index be5c1d9..015e9bf 100644 --- a/lib/libalpm/util.h +++ b/lib/libalpm/util.h @@ -50,6 +50,7 @@ #define CALLOC(p, l, s, action) do { p = calloc(l, s); if(p == NULL) { ALLOC_FAIL(s); action; } } while(0) /* This strdup macro is NULL safe- copying NULL will yield NULL */ #define STRDUP(r, s, action) do { if(s != NULL) { r = strdup(s); if(r == NULL) { ALLOC_FAIL(strlen(s)); action; } } else { r = NULL; } } while(0) +#define STRNDUP(r, s, l, action) do { if(s != NULL) { r = strndup(s, l); if(r == NULL) { ALLOC_FAIL(strlen(s)); action; } } else { r = NULL; } } while(0) #define FREE(p) do { free(p); p = NULL; } while(0) @@ -101,6 +102,10 @@ long _alpm_parsedate(const char *line); char *strsep(char **, const char *); #endif +#ifndef HAVE_STRNDUP +char *strndup(const char *s, size_t n); +#endif + /* check exported library symbols with: nm -C -D <lib> */ #define SYMEXPORT __attribute__((visibility("default"))) #define SYMHIDDEN __attribute__((visibility("internal"))) -- 1.7.3.5

1 2

[pacman-dev] when upgrading display versions
by Divan Santana 20 Jan '11

20 Jan '11

Hi Pacman Devs, Hope this is the correct place to ask such a question... I'm new to Arch Linux and loving the Arch so far. I'm impressed and pleasantly surprised by pacman as a package management system. One thing that would be nice, and I think it's in line with The Arch Way (openness perhaps ;) is when updating to display(or be able to display) the versions of the packages being upgraded. I know one sees the new version being installed but not the currently installed version so one can quickly compare before proceeding/deciding to upgrade. I know one can easily achieve this with queries and small scripts(such as below) but it would really be quite nice to pass a pacman parameter and be able to see the output. I think it's quite important as Arch being a rolling distro sometimes you get major updates of packages and it would be nice to see straight away which are major updates and which are minor. Script to do this I use is: for package in `pacman -Qu | awk {'print $1'}`; do echo "`pacman -Q $package` to `pacman -Si $package | grep Version | awk {'print $NF'}`"; done Other package management programs + arguments that do this is: zypper -v up apt-get dist-upgrade -V Anyway it's a minor thing perhaps the best is to log a feature request at https://bugs.archlinux.org ? Better(and also in line with the Arch Way) would be contribute the code and not request it but unfortunately I can't help with that since I know so little programming. Greetings :) -- Divan Santana

3 3

[pacman-dev] Silly output with checking package integrity progress bar
by Dan McGee 20 Jan '11

20 Jan '11

Another bit of fallout: :: Retrieving packages from community... picard-0.12.1-5-x86_64 607.1K 347.1K/s 00:00:02 [##################################################] 100% pv-1.2.0-1-x86_64 26.8K 119.7K/s 00:00:00 [##################################################] 100% :: File pyqt-4.8.2-2-x86_64.pkg.tar.xz is corrupted. Do you want to delete it? [Y/n] y[##############------------------------------------] 29% (23/24) checking package integrity [##################################################] 100% error: failed to commit transaction (invalid or corrupted package) pyqt-4.8.2-2-x86_64.pkg.tar.xz is invalid or corrupted Errors occurred, no packages were upgraded. Notice how the bar overlapped with the meter since we never hit 100% and line wrapped. Perhaps we should hold all of these conversations until after the integrity check is complete, if possible? -Dan

1 0

[pacman-dev] [PATCH] Allow files entries to be read from a sync database
by Dan McGee 18 Jan '11

18 Jan '11

Nothing will use this yet, but we might as well read the info in while processing the database. The drawback would be vastly increased memory usage, so we might want to add a flag enabling/disabling the loading of this data. This will at least for now be unused by pacman, but a pkgfile.c program at least now has the option of using standard libalpm interfaces to search for file ownership. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- Thoughts from the peanut gallery? -Dan lib/libalpm/be_sync.c | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/lib/libalpm/be_sync.c b/lib/libalpm/be_sync.c index b11499c..58ff3c3 100644 --- a/lib/libalpm/be_sync.c +++ b/lib/libalpm/be_sync.c @@ -282,7 +282,7 @@ static int sync_db_read(pmdb_t *db, struct archive *archive, struct archive_entr } if(strcmp(filename, "desc") == 0 || strcmp(filename, "depends") == 0 - || strcmp(filename, "deltas") == 0) { + || strcmp(filename, "deltas") == 0 || strcmp(filename, "files") == 0) { while(_alpm_archive_fgets(archive, &buf) == ARCHIVE_OK) { char *line = _alpm_strtrim(buf.line); @@ -357,6 +357,8 @@ static int sync_db_read(pmdb_t *db, struct archive *archive, struct archive_entr READ_AND_STORE_ALL(pkg->provides); } else if(strcmp(line, "%DELTAS%") == 0) { READ_AND_STORE_ALL(pkg->deltas); + } else if(strcmp(line, "%FILES%") == 0) { + READ_AND_STORE_ALL(pkg->files); } } } else { -- 1.7.3.5

3 3

[pacman-dev] [PATCH] Fix integrity check status when installing from file
by Allan McRae 18 Jan '11

18 Jan '11

When installing packages from a file, the integrity check count stays at (0/x) complete. This ensures it is bumped to (x/x) at the end of the process. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- I'm not sure this is the best way to fix this issue... But given the loop is essentially doing nothing in this case, I do not see the point in updating the progress bar as we go. lib/libalpm/sync.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/lib/libalpm/sync.c b/lib/libalpm/sync.c index b1eab55..bbb319c 100644 --- a/lib/libalpm/sync.c +++ b/lib/libalpm/sync.c @@ -957,6 +957,7 @@ int _alpm_sync_commit(pmtrans_t *trans, pmdb_t *db_local, alpm_list_t **data) pmpkg_t *spkg = i->data; int percent = (current * 100) / numtargs; if(spkg->origin == PKG_FROM_FILE) { + current++; continue; /* pkg_load() has been already called, this package is valid */ } PROGRESS(trans, PM_TRANS_PROGRESS_INTEGRITY_START, "", percent, -- 1.7.3.5

3 4

[pacman-dev] New Guy Wants to Change Everything
by Westley Martínez 18 Jan '11

18 Jan '11

Hello I'm Anikom15 and I'm new. You might know me from the Arch Linux forums but whatever. I'm interested in helping develop pacman considering the fact I've been using the program for so long I figure I should give something back, and maybe get better at C coding, too. Okay, one question: why does the source code use tabs instead of spaces? Is there a specific reason as to why? In my experience spaces makes everything easier because tabs can mess up alignment if the editor isn't set properly. I know what you're thinking, "These new people they come in and wanna change everything and then leave". I don't wanna change anything unnecessarily, I'd just like an explanation. Now just between you and me I think some other conventions the code is using like return statements being written like a function call are stupid as well, but those don't have dire consequences like the tabs can potentially have. Sorry if I sound a little invasive, I don't collaborate much.

4 4

[pacman-dev] [PATCH] remove ":: " from messages
by Vojtěch Gondžala 18 Jan '11

18 Jan '11

Hi all, there is an simple patch for remove prefix "::" from messages, all magic chars are in function `info_msg' like "warning: : and "error: " prefixes in `pm_*printf'. It will be better for translators to, and it's first step prepare for colorize output support for pacman. -- Vojtěch "vogo" Gondžala

2 1

[pacman-dev] Using libalpm with Vala
by Rémy Oudompheng 14 Jan '11

14 Jan '11

Hello, I became recently interested in Vala, and I did several experiments to try interfacing it with libalpm. Vala is an object-oriented language created to be used mainly with glib2/gtk, and writing Vala bindings is especially lightweight : it is only a matter of writing a "dictionary" between C functions and methods. I think it might be a nice alternative to C when writing small command line or GTK utilities. I have put an example of how the C version of pactree would be written in Vala, at http://github.com/remyoudompheng/libalpm-vala Memory management is especially tricky, as I am new to Vala, and not very familiar with the internals of libalpm. So for the moment, only a very small set of functions are wrapped and this version of pactree does not segfault. Feel free to make any comments. Regards, Rémy.

1 0

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.4.2-248-g665528d
by dan＠archlinux.org 14 Jan '11

14 Jan '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 665528d7bac99228849d1162f78c0817bc619eee (commit) via 5615b716887960260016646c2d0f812806abe2ec (commit) via 30f53cfe8d3f07bc9f350387b55ddf8579dd75e3 (commit) via cae2bdafec93381284ba487010738d72d0992aab (commit) via 8e30a46adbe350a9741bbb9748334a4adc9742f6 (commit) via 9e8af82c97992ab15e4c182af84f06afd4885dc7 (commit) from dd265921559b187cba59e83bfdbfda11bcec2f1e (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 665528d7bac99228849d1162f78c0817bc619eee Author: Dan McGee <dan(a)archlinux.org> Date: Thu Jan 13 09:33:01 2011 -0600 repo-add: fix misguided conditional correction I tried to move things around here when testing and did a bit too much; the warning message always showed regardless of delta inclusion in the call. Fix it so we only warn if we have a filename, but the file couldn't be located. Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 5615b716887960260016646c2d0f812806abe2ec Merge: dd26592 30f53cf Author: Dan McGee <dan(a)archlinux.org> Date: Wed Jan 12 09:23:24 2011 -0600 Merge branch 'maint' Conflicts: lib/libalpm/be_files.c ----------------------------------------------------------------------- Summary of changes: scripts/repo-add.sh.in | 14 ++++++++------ src/pacman/pacman.c | 4 ++-- 2 files changed, 10 insertions(+), 8 deletions(-) hooks/post-receive -- The official pacman repository

1 0