Send patches - preferably formatted by git format-patch - to patches at archlinux32 dot org.
summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2007-11-06Maintain list tail pointers in the head nodeAaron Griffin
List head nodes contain null 'prev' pointer, which we can (ab)use to maintain a back reference to the tail pointer of the list. While list additions are not _significantly_ improved, they are still sped up. Original $ time pacman -Qo /usr/bin/wtpt /usr/bin/wtpt is owned by lcms 1.17-2 real 0m3.623s user 0m1.883s sys 0m1.473s New $ time pacman -Qo /usr/bin/wtpt /usr/bin/wtpt is owned by lcms 1.17-2 real 0m2.006s user 0m0.263s sys 0m1.627s Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
2007-11-05Comment out paths by default in pacman.confDan McGee
Now that all paths are defined to reasonable defaults at compile time by pacman, we shouldn't force users to have RootDir, DBPath, etc. in their default pacman.conf although these options are always available. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-05libalpm/db.c: change two warnings to errorsDan McGee
These two warnings really indicate failure, so the message they print should do so as well. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-05Fix mercurial complaining about directory already existingSebastian Nowicki
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-05Add LOGFILE as a define at compile time for pacmanDan McGee
It has always been a bit odd that logfile had to be specified in the config file, but no other paths did. Add LOGFILE as a preprocessor definition, and make a call to alpm_option_set_logfile() to set the default location so no logfile parameter is necessary in pacman.conf. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Remove the newline automatically added by alpm_logaction.Chantry Xavier
This way, _alpm_logaction behaves like _alpm_log, and gives more control. Signed-off-by: Chantry Xavier <shiningxc@gmail.com>
2007-11-04libalpm: use an lstat wrapper so we never dereference dir symlinksDan McGee
Linux lstat follows POSIX standards and dereferences a symlink pointing to a directory if there is a trailing slash. For purposes of libalpm, we don't want this so make a lstat wrapper that suppresses this behavior. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04libalpm: open the logstream on demandDan McGee
Don't open a stream to the logfile until necessary. This will allow us to catch any errors in opening the logfile instead of ignorning them. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Clean up usage of extern variablesDan McGee
Instead of declaring the extern variable in every *.c file, include it in the header file that makes sense. This means handle.h for the handle, and conf.h for the pacman side config object. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Readd scriptlet logging that got lost in an earlier commitDan McGee
I broke scriptlet logging with ad691001e20272b794d2ed574b556f520e3555c0. Readd more or less what was there before, although it still needs a lot of work including hopefully rewriting it to a new event subsystem and having it log to a seperate file. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04vercmp: add some usage instructionsDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Remove unneeded CXX macros from acinclude.m4Dan McGee
This follows the removal patch previously used for GCC & F77 macros. If we don't use it, dump it so we can speed up configure time. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Add some text at the top of acinclude.m4 so people know what is going onDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Remove unneeded F77 and GCJ libtool macros from acinclude.m4Dan McGee
We don't use Fortran or Java code in our project, so no need to waste time looking for this stuff. This gives a noticeable speedup to running the ./configure program as well. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Make building of pacman.static optionalDan McGee
Because building of pacman.static fails on some platforms, we should make it optional. It is enabled by default but can be disabled with the use of the --disable-pacman-static flag. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Add some more autoconf macros to filter our CFLAGS usageDan McGee
Hopefully these new autoconf macros, with a little magic, will allow us to compile with any compiler and still choose the options we have available to us. Tested locally with gcc 4.2.2 and gcc 3.4.6; the latter doesn't support two of the items we previously had hardcoded in our CFLAGS. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Use an autoconf macro to see if -fstack-protector is availableDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Revise configure.acDan McGee
Do a little cleanup of our configure script. Highlights: * Remove macros deemed unnecessary to call [1] * Change check for compiler to look for one that is C99 capable- this automatically adds the -std=gnu99 flag [1] Noted in the autoconf NEWS file, notably entries for 2.59d Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Update libtool files and macros to newest versionsDan McGee
It looks like some of the newer libtool m4 files offer improved support for the Darwin platform and possibly BSD, so bump our files to these new versions. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Update my TODODan McGee
2007-11-04Add a little const correctness fix to alpm_listDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Fix basename usage in pacman and utilitiesDan McGee
basename() is a rather untrusty function call on a lot of platforms as it does some weird and different things. To solve this, I added a mbasename fuction to pacman to take its place, and simply removed its usage in the utilities (it isn't worth dealing with there). Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Add support for arch='any'Roman Kyrylych
Signed-off-by: Roman Kyrylych <roman@archlinux.org> Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Remove final use of the math library from pacmanDan McGee
To round a value, we don't need floorf- we can just cast to an int. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-11-04Added optdepends support to pactestJeff Mickey
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-30abs: prefer csup to cvsupDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-30Pick best python program available for 'make check'Dan McGee
Use an autoconf macro to find us a python executable, preferring python2.5 if we can find it. From there, fall back to python2.4 and then python. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29When printing size_t, use %zdDan McGee
%d was used, which worked for Linux and FreeBSD. Not so for Darwin. The warning was probably spat out when compiling on x68_64 as well, but no developers use this architecture as their primary one. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29libalpm/handle.c: make realpath() call portableDan McGee
BSD didn't support the NULL second argument GNU extension, so do it the old fashioned way. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29pacman/query.c: add missing free()Dan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29Small updates to documentation.Scott Horowitz
Signed-off-by: Scott Horowitz <stonecrest@gmail.com> Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29Copy pmdelta_t objects in _alpm_pkg_dup()Dan McGee
This was forgotten in the original set of patches. CC: Nathan Jones <nathanj@insightbb.com> Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29Make general list copy functionDan McGee
Package dup needs to copy all members. Nathan had his implementation, but I generalized it to this new alpm_list function (and will use it in the next commit). CC: Nathan Jones <nathanj@insightbb.com> Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29simple pactest renaming : split conflict* into depconflict* and fileconflict*.Chantry Xavier
Signed-off-by: Chantry Xavier <shiningxc@gmail.com> Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29doxygen: preprocess SYMHIDDEN/SYMEXPORT macros so they don't show in docsDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29libalpm/delta: add const to most methodsDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29libalpm: introduce MALLOC and CALLOC macrosDan McGee
These macros take the place of the common 4 or 5 line blocks of code we had in most places that called malloc or calloc. This should reduce some code duplication and make memory allocation more standard in libalpm. Highlights: * Note that the MALLOC macro actually uses calloc, this is just for safety so that memory is initialized to 0. This can be easily changed in one place. * One malloc call was completely eliminated- it made more sense to do it on the stack. * The use of RET_ERR in public functions (mainly the alpm_*_new functions) was standardized, this makes sense so pm_errno is set. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-29Add some more debugging output to dep checking codeDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-28pactest: generate the filelist the correct wayDan McGee
Actually use python commands to generate the filelist instead of os.system calls that don't necessarily work everywhere. Noticed when running "make check" on FreeBSD where the tar program is actually bsdtar. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-28extend type001 pactest with a symlink.Chantry Xavier
Signed-off-by: Chantry Xavier <shiningxc@gmail.com>
2007-10-28makepkg: add explicit dir to find callsDan McGee
Some of the find calls did not have a path, which isn't valid in some versions of the find utility. Add a . for compatibility. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-26Integrate versionpkg into makepkgScott Horowitz
If PKGBUILDs use the correct svn/cvs/etc variables, determine the latest revision number and force this as the pkgver for building. Signed-off-by: Scott Horowitz <stonecrest@gmail.com> Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-26Clean up LDADD usageDan McGee
We had a lot of unnecessary overstatements of libraries to include on linking, and autoconf/automake takes care of this for us. This also helps some compilation issues on other platforms. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-26Fix mcheck detection and usageDan McGee
Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-26Explicitly cast void* to char* during printf callDan McGee
If we don't have an explicit cast, make fails during -Wall -Werror. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-26dependency error message fix in libalpm/remove.cNagy Gabor
Normally you must never see that error message. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-26Dependency error/log messages fixNagy Gabor
The old code used only the depend.name in messages, which might have not been informative. The new code uses the whole dependency string in %DEPENDS% format. (Dan: slight English clarification in one of the messages) Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-26makepkg: readd -p option that was lost in getopt conversionDan McGee
The -p option was accidentally dropped in commit 54b71f0427e87e6d525423df06f8a06f2b71c518. Readd it. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-25Add alpm_dep_get_string methodNagy Gabor
Public alpm_dep_get_string function is introduced, which converts a pmdepend_t structure to printable string in %DEPENDS% format. This function is now used in pacman to print dependency error messages. Signed-off-by: Dan McGee <dan@archlinux.org>
2007-10-24_alpm_depmiss_isin fixNagy Gabor
The old code used memcmp, which is not good for comparing strings: "pkgname"'\0''\0' should be equal to "pkgname"'\0''a' for example. The new code uses strcmp. Signed-off-by: Dan McGee <dan@archlinux.org>