- Pacman-dev - lists.archlinux.org

Re: [pacman-dev] Splitting packages in makepkg
by Miklos Vajna 17 Jun '08

17 Jun '08

On Tue, Jun 17, 2008 at 10:20:59PM +1000, Allan McRae <mcrae_allan(a)hotmail.com> wrote: > Just another thought. Why do we even need the subpkg array. Why not: > "pkgname=foo" for single packages and "pkgname=('foo' 'foo-bar' > 'foo-doc')" for split packages? I would have to check that you can do > that in bash... pkgname=("foo") is equivalent to pkgname="foo" in bash, at least ${pkgname[0]} will work for both, ${#pkgname[@]} will report 1 for both, etc.

1 0

[pacman-dev] FS#7982 - patch to makepkg to allow PKGBUILDs building more than one package
by Marc - A. Dahlhaus 17 Jun '08

17 Jun '08

Hello List, a few comments on the posted patch... @@ -1389,7 +1391,57 @@ if [ "$INFAKEROOT" = "1" ]; then tidy_install fi - create_package + if [ "$splitinstall" != "" ] ; then + backup_o=$backup + conflicts_o=$conflicts + depends_o=$depends + groups_o=$groups + install_o=$install + license_o=$license + pkgdesc_o=$pkgdesc + pkgname_o=$pkgname + pkgver_o=$pkgver + provides_o=$provides + replaces_o=$replaces + url_o=$url This is the right place to do this in my opinion, that way you have the global defined defaults for every slpit-package. To move this inside of the loop would kill the posibility to get the fuits from my next comment. + create_package I think this is wrong from the view to build sub-packages out of one compiled tree. build should be used to only build the sources and prepare everything which would be usefull for all subpackages. That way we could add a new param to makepkg to bould only a subset of the splitpackages by overriding the contents of the defined var splitinstall. + for it in "${splitinstall[@]}" ; do + if [ -d "$pkgdir" ]; then + msg "Removing existing pkg/ directory..." + rm -rf "$pkgdir" + fi + mkdir -p "$pkgdir" I like it that way and we can fix the issue spotted by *Allan McRae* (non working repackage param) if we take the way of my prevous comment and allow to rebuild only a subset of the splitpackages. I think we should add the splitpackage-name to repackage as additional param to makepkg's command line and use that to override the splitinstall value in case we want repackage. --8<-- thanks, Marc

7 8

[pacman-dev] Splitting packages in makepkg
by Allan McRae 17 Jun '08

17 Jun '08

NOTE: This is a discussion about the technical implementation of the ability to build split packages in makepkg. DO NOT reply if all you want to say is splitting packages is a bad thing. Just because makepkg can split packages does not mean Arch will start splitting everything up into binary/header packages. I can say with great certainty that that will never happen. I was going to wait until 3.2 was out the door to start implementing the ability to split packages but there has been enough activity on the bug tracker lately that I thought this discussion would be timely. There have been several proposals for building split packages with makepkg. See: http://bugs.archlinux.org/task/7144 http://bugs.archlinux.org/task/7982 http://bugs.archlinux.org/task/8187 The question is how best to implement this. I can not see any other way than having a "split" array listing the names of the split components, so the need for this is almost a given. With input from the bug reports, I see two possible ways to implement this (in the PKGBUILD): 1) The KDEMod style (FS#7982), where each package has its own install function. e.g. base_install() { # do install stuff } gui_install() { # override pkgname etc # do install stuff } 2) A "package" function with a case statement. e.g. package() { case "$1" in "base" ) # do install stuff ;; "gui" ) # override pkgname etc # do install stuff ;; esac } I am kind of torn here... I like the idea of having a single function but the syntax is slightly more complex. Although the multi function version is already in use, that should not be a limiting factor - we should be striving for technical elegance rather than accepting the current situation. The second point I would like to bring up is based on the suggestion in FS#8187. In it I suggested having separate subdirectories where each sub-package is installed into. This would be a benefit because it would not require the clearing of the pkg directory after each sub-package install as in the current KDEmod implementation. I see this would be a definite help when bugfixing a PKGBUILD as you can easily browse the installed files. It would also mean that we could keep the repackage option working. This could be implemented with either method by setting the pkgdir variable before calling whichever package function is chosen. But it would require the forcing of the use of $pkgdir rather than $startdir/pkg as that would fail. So people, what are you opinions on the ideas above? These are just the choices I can currently see so don't feel limited to only commenting on them. At this stage I would not worry about the actual naming scheme of functions etc, just the implementation. So please discuss (after going back and reading the top paragraph...). Allan

12 23

Re: [pacman-dev] Splitting packages in makepkg
by Sergej Pupykin 17 Jun '08

17 Jun '08

Modified idea: PKGBUILD ------------------ pkgname=foobar pkgver=1.0 pkgrel=1 subpackages=(dev doc) pkgdesc="base package desc" pkgdesc_dev="foo dev desc" .... build() { ..... make DESTDIR=$startdir/pkg install ... make DESTDIR=$startdir/pkg-dev install ... make DESTDIR=$startdir/pkg-doc install } --------------- According to subpackages array makepkg creates 3 dirs: - $startdir/pkg (as usually) - $startdir/pkg-dev - $startdir/pkg-doc PKGBUILDs build function install all stuff into them And then, makepkg creates 3 packages: $pkgname-...pkg.tar.gz (if $startdir/pkg is not empty) $pkgname-dev-...pkg.tar.gz (if $startdir/pkg-dev is not empty) $pkgname-doc-...pkg.tar.gz (if $startdir/pkg-doc is not empty) May be I am not too famillar with bash, but you can get pkgdesc_dev at least with construction like this: $ export pkgdesc_dev="foo dev desc" $ export pkgdescname=pkgdesc_dev $ echo `bash -c "echo \$\`echo $pkgdescname\`"` foo dev desc It can be easy modified for full package names as I described above: subpackages=(foo foo-dev foo-doc kmail kopete) or pkgname=(foo foo-dev foo-doc kmail kopete)

2 1

[pacman-dev] FS#7982 - V2: patch to makepkg to allow PKGBUILDs building more than one package
by silvio＠port1024.net 16 Jun '08

16 Jun '08

AWKWARD! - Now I have test it before! Hello Friends, this is version2 of my patch. Explanation is in the patch. Please review and discuss this patch. Bye, eS.eF.

1 1

[pacman-dev] FS#7982 - V2: patch to makepkg to allow PKGBUILDs building more than one package
by silvio＠port1024.net 16 Jun '08

16 Jun '08

Sorry, I was too fast with send, without checkin :-) Hello Friends, this is version2 of my patch. Explanation is in the patch. Please review and discuss this patch. Bye, eS.eF.

1 1

[pacman-dev] FS#7982 - V2: patch to makepkg to allow PKGBUILDs building more than one package
by silvio＠port1024.net 16 Jun '08

16 Jun '08

Hello Friends, this is version2 of my patch. Explanation is in the patch. Please review and discuss this patch. Bye, eS.eF.

1 1

[pacman-dev] [PATCH] Use access() instead of stat() when possible
by Dan McGee 16 Jun '08

16 Jun '08

We were using the stat() system call in quite a few places when we didn't actually need anything the stat struct returned- we were simply checking for file existence. access() will be more efficient in those cases. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- lib/libalpm/be_files.c | 5 ++--- lib/libalpm/trans.c | 8 ++++---- lib/libalpm/util.c | 12 ++++-------- src/pacman/package.c | 5 ++--- src/pacman/util.c | 6 ++---- src/util/testdb.c | 8 +++----- 6 files changed, 17 insertions(+), 27 deletions(-) diff --git a/lib/libalpm/be_files.c b/lib/libalpm/be_files.c index 12c60b2..f5d4e82 100644 --- a/lib/libalpm/be_files.c +++ b/lib/libalpm/be_files.c @@ -355,7 +355,6 @@ static char *get_pkgpath(pmdb_t *db, pmpkg_t *info) int _alpm_db_read(pmdb_t *db, pmpkg_t *info, pmdbinfrq_t inforeq) { FILE *fp = NULL; - struct stat buf; char path[PATH_MAX]; char line[513]; char *pkgpath = NULL; @@ -393,7 +392,7 @@ int _alpm_db_read(pmdb_t *db, pmpkg_t *info, pmdbinfrq_t inforeq) pkgpath = get_pkgpath(db, info); - if(stat(pkgpath, &buf)) { + if(access(pkgpath, F_OK)) { /* directory doesn't exist or can't be opened */ _alpm_log(PM_LOG_DEBUG, "cannot find '%s-%s' in db '%s'\n", info->name, info->version, db->treename); @@ -631,7 +630,7 @@ int _alpm_db_read(pmdb_t *db, pmpkg_t *info, pmdbinfrq_t inforeq) /* INSTALL */ if(inforeq & INFRQ_SCRIPTLET) { snprintf(path, PATH_MAX, "%sinstall", pkgpath); - if(!stat(path, &buf)) { + if(access(path, F_OK) == 0) { info->scriptlet = 1; } } diff --git a/lib/libalpm/trans.c b/lib/libalpm/trans.c index eb53e95..06084ae 100644 --- a/lib/libalpm/trans.c +++ b/lib/libalpm/trans.c @@ -25,6 +25,7 @@ #include <stdlib.h> #include <stdio.h> #include <string.h> +#include <unistd.h> #include <sys/wait.h> #include <sys/types.h> #include <sys/stat.h> @@ -451,7 +452,6 @@ int _alpm_runscriptlet(const char *root, const char *installfn, char tmpdir[PATH_MAX]; char cwd[PATH_MAX]; char *scriptpath; - struct stat buf; pid_t pid; int clean_tmpdir = 0; int restore_cwd = 0; @@ -459,14 +459,14 @@ int _alpm_runscriptlet(const char *root, const char *installfn, ALPM_LOG_FUNC; - if(stat(installfn, &buf)) { + if(access(installfn, R_OK)) { /* not found */ _alpm_log(PM_LOG_DEBUG, "scriptlet '%s' not found\n", installfn); return(0); } /* NOTE: popen will use the PARENT's /bin/sh, not the chroot's */ - if(stat("/bin/sh", &buf)) { + if(access("/bin/sh", X_OK)) { /* not found */ _alpm_log(PM_LOG_ERROR, _("No /bin/sh in parent environment, aborting scriptlet\n")); return(0); @@ -474,7 +474,7 @@ int _alpm_runscriptlet(const char *root, const char *installfn, /* creates a directory in $root/tmp/ for copying/extracting the scriptlet */ snprintf(tmpdir, PATH_MAX, "%stmp/", root); - if(stat(tmpdir, &buf)) { + if(access(tmpdir, F_OK) != 0) { _alpm_makepath_mode(tmpdir, 01777); } snprintf(tmpdir, PATH_MAX, "%stmp/alpm_XXXXXX", root); diff --git a/lib/libalpm/util.c b/lib/libalpm/util.c index 55bd46a..9199545 100644 --- a/lib/libalpm/util.c +++ b/lib/libalpm/util.c @@ -200,12 +200,11 @@ int _alpm_makepath_mode(const char *path, mode_t mode) str = orig; while((ptr = strsep(&str, "/"))) { if(strlen(ptr)) { - struct stat buf; /* we have another path component- append the newest component to * existing string and create one more level of dir structure */ strcat(incr, "/"); strcat(incr, ptr); - if(stat(incr, &buf)) { + if(access(incr, F_OK)) { if(mkdir(incr, mode)) { ret = 1; break; @@ -533,12 +532,11 @@ int _alpm_logaction(unsigned short usesyslog, FILE *f, const char *fmt, va_list int _alpm_ldconfig(const char *root) { char line[PATH_MAX]; - struct stat buf; snprintf(line, PATH_MAX, "%setc/ld.so.conf", root); - if(stat(line, &buf) == 0) { + if(access(line, F_OK) == 0) { snprintf(line, PATH_MAX, "%ssbin/ldconfig", root); - if(stat(line, &buf) == 0) { + if(access(line, X_OK) == 0) { char cmd[PATH_MAX]; snprintf(cmd, PATH_MAX, "%s -r %s", line, root); system(cmd); @@ -561,7 +559,6 @@ int _alpm_str_cmp(const void *s1, const void *s2) */ char *_alpm_filecache_find(const char* filename) { - struct stat buf; char path[PATH_MAX]; char *retpath; alpm_list_t *i; @@ -570,8 +567,7 @@ char *_alpm_filecache_find(const char* filename) for(i = alpm_option_get_cachedirs(); i; i = alpm_list_next(i)) { snprintf(path, PATH_MAX, "%s%s", (char*)alpm_list_getdata(i), filename); - if(stat(path, &buf) == 0) { - /* TODO maybe check to make sure it is readable? */ + if(access(path, R_OK) == 0) { retpath = strdup(path); _alpm_log(PM_LOG_DEBUG, "found cached pkg: %s\n", retpath); return(retpath); diff --git a/src/pacman/package.c b/src/pacman/package.c index 1698806..0680037 100644 --- a/src/pacman/package.c +++ b/src/pacman/package.c @@ -22,8 +22,8 @@ #include <stdlib.h> #include <stdio.h> #include <string.h> +#include <unistd.h> #include <limits.h> -#include <sys/stat.h> #include <wchar.h> #include <alpm.h> @@ -176,7 +176,6 @@ void dump_pkg_backups(pmpkg_t *pkg) if(alpm_pkg_get_backup(pkg)) { /* package has backup files, so print them */ for(i = alpm_pkg_get_backup(pkg); i; i = alpm_list_next(i)) { - struct stat buf; char path[PATH_MAX]; char *str = strdup(alpm_list_getdata(i)); char *ptr = index(str, '\t'); @@ -188,7 +187,7 @@ void dump_pkg_backups(pmpkg_t *pkg) ptr++; snprintf(path, PATH_MAX-1, "%s%s", root, str); /* if we find the file, calculate checksums, otherwise it is missing */ - if(!stat(path, &buf)) { + if(access(path, R_OK) == 0) { char *md5sum = alpm_get_md5sum(path); if(md5sum == NULL) { diff --git a/src/pacman/util.c b/src/pacman/util.c index e702886..d7ac9e3 100644 --- a/src/pacman/util.c +++ b/src/pacman/util.c @@ -22,7 +22,6 @@ #include <sys/types.h> #include <sys/ioctl.h> #include <sys/time.h> -#include <sys/stat.h> #include <stdio.h> #include <stdlib.h> @@ -125,7 +124,7 @@ int makepath(const char *path) * orig - a copy of path so we can safely butcher it with strsep * str - the current position in the path string (after the delimiter) * ptr - the original position of str after calling strsep - * incr - incrementally generated path for use in stat/mkdir call + * incr - incrementally generated path for use in access/mkdir call */ char *orig, *str, *ptr, *incr; mode_t oldmask = umask(0000); @@ -136,12 +135,11 @@ int makepath(const char *path) str = orig; while((ptr = strsep(&str, "/"))) { if(strlen(ptr)) { - struct stat buf; /* we have another path component- append the newest component to * existing string and create one more level of dir structure */ strcat(incr, "/"); strcat(incr, ptr); - if(stat(incr, &buf)) { + if(access(incr, F_OK)) { if(mkdir(incr, 0755)) { ret = 1; break; diff --git a/src/util/testdb.c b/src/util/testdb.c index f354eca..87bfcf9 100644 --- a/src/util/testdb.c +++ b/src/util/testdb.c @@ -23,7 +23,6 @@ #include <errno.h> #include <limits.h> #include <string.h> -#include <sys/stat.h> #include <dirent.h> #include <libgen.h> @@ -61,7 +60,6 @@ static int db_test(char *dbpath) { struct dirent *ent; char path[PATH_MAX]; - struct stat buf; int ret = 0; DIR *dir; @@ -77,17 +75,17 @@ static int db_test(char *dbpath) } /* check for desc, depends, and files */ snprintf(path, PATH_MAX, "%s/%s/desc", dbpath, ent->d_name); - if(stat(path, &buf)) { + if(access(path, F_OK)) { printf("%s: description file is missing\n", ent->d_name); ret++; } snprintf(path, PATH_MAX, "%s/%s/depends", dbpath, ent->d_name); - if(stat(path, &buf)) { + if(access(path, F_OK)) { printf("%s: dependency file is missing\n", ent->d_name); ret++; } snprintf(path, PATH_MAX, "%s/%s/files", dbpath, ent->d_name); - if(stat(path, &buf)) { + if(access(path, F_OK)) { printf("%s: file list is missing\n", ent->d_name); ret++; } -- 1.5.5.3

2 1

[pacman-dev] FS#7982 - patch to makepkg to allow PKGBUILDs building more than one package
by silvio＠port1024.net 15 Jun '08

15 Jun '08

Hello Freaks, This patch is for building more than one package with makepkg. Please review and discuss this patch! -- eS.eF.

6 7

Re: [pacman-dev] [PATCH] makepkg.sh.in is prepared for building of splittet packages
by Xavier 15 Jun '08

15 Jun '08

Xavier wrote: >> Also it is rather strange that there is no space between your name and >> mail address. Did you do something like the above : >> git repo-config user.name "Your Name" >> git repo-config user.email "me(a)example.com" > >> silvio(a)port1024.net wrote: >>> From: Silvio fricke<silvio(a)port1024.net> > > Very strange, because [1] show me that I have spaces between my name and > address. > Oops, seems to be a thunderbird issue when replying. >> I have only minor "appearance" issues to address here, I don't know what >> to think about the patch itself. Only that it does not seem too big and >> mostly concentrated in one place, which is good. > > Ok, fine! > >> So first, isn't the correct term "splitted"? >> It seems that you use both "split" and "splittet" which would then be >> incorrect. > > Hmm, english is not my native language and I need help in language > things. > But I have the suggestions from [2] used. > Yeah, use split. >> Also you should read this : >> http://archlinux.org/pacman/submitting-patches.html >> Especially this part : >> << >> Describe your patch. >> >> Before the actual diff begins, it helps if you describe the changes in >> the patch. This allows others to see what you intended so as to compare >> it to what was actually done, and allows better feedback. If you use >> git-format-patch to create your patch, then your commit message will be >> shown above the patch by default. >> >> > > Right you are! The next version of this patch has a more descriptive > description. > Good. Also have a look at Allan's suggestion. Seems like the whole patch would be nicer if it could use a loop.

1 0