- Pacman-dev - lists.archlinux.org

[pacman-dev] [PATCH] Do not compare signed and unsigned types
by Allan McRae 28 Jun '10

28 Jun '10

The casting of nread is safe as it is tested to be >0 when it is initally assigned. It is also being implicitly case in the fwrite call in the line above. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- lib/libalpm/dload.c | 2 +- src/pacman/pacman.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/lib/libalpm/dload.c b/lib/libalpm/dload.c index c11148d..3185d2a 100644 --- a/lib/libalpm/dload.c +++ b/lib/libalpm/dload.c @@ -251,7 +251,7 @@ static int download_internal(const char *url, const char *localpath, check_stop(); size_t nwritten = 0; nwritten = fwrite(buffer, 1, nread, localf); - if((nwritten != nread) || ferror(localf)) { + if((nwritten != (size_t)nread) || ferror(localf)) { pm_errno = PM_ERR_RETRIEVE; _alpm_log(PM_LOG_ERROR, _("error writing to file '%s': %s\n"), tempfile, strerror(errno)); diff --git a/src/pacman/pacman.c b/src/pacman/pacman.c index ecc37c9..60f2e5e 100644 --- a/src/pacman/pacman.c +++ b/src/pacman/pacman.c @@ -1002,7 +1002,7 @@ static int _parseconfig(const char *file, const char *givensection, file, linenum, value); break; default: - for(int gindex = 0; gindex < globbuf.gl_pathc; gindex++) { + for(size_t gindex = 0; gindex < globbuf.gl_pathc; gindex++) { pm_printf(PM_LOG_DEBUG, "config file %s, line %d: including %s\n", file, linenum, globbuf.gl_pathv[gindex]); _parseconfig(globbuf.gl_pathv[gindex], section, db); -- 1.7.1

1 0

[pacman-dev] [PATCH] makepkg: Only check regular files in $srcdir check
by Thomas Bächler 28 Jun '10

28 Jun '10

The 'grep -R' in the $srcdir check would not only grep regular files, but also devices, symlinks (that might potentially point outside of $pkgdir), pipes and so on. Use find to ensure only regular files are examined. This should fix https://bugs.archlinux.org/task/19975 --- I haven't tested this yet (no time right now), but this should work and fix the bug. Please review and comment, I may be able to test it later. scripts/makepkg.sh.in | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index cbc344d..97bf5da 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -967,7 +967,7 @@ check_package() { done # check for references to the build directory - if grep -R "${srcdir}" "${pkgdir}" &>/dev/null; then + if find "${pkgdir}" -type f -exec grep -q "${srcdir}" {} +; then warning "$(gettext "Package contains reference to %s")" "\$srcdir" fi } -- 1.7.1

2 1

[pacman-dev] Dealing with asprintf failures
by Allan McRae 28 Jun '10

28 Jun '10

The remaining warnings generated with -D_FORTIFY_SOURCE=2 are usages of asprintf in src/pacman/util.c. e.g. util.c: In function ‘display_targets’: util.c:533:12: warning: ignoring return value of ‘asprintf’, declared with attribute warn_unused_result What is the best way to deal with these? One idea is: - asprintf(&str, "%s-%s [%.2f MB]", alpm_pkg_get_name(pkg), - alpm_pkg_get_version(pkg), mbsize); + if(asprintf(&str, "%s-%s [%.2f MB]", alpm_pkg_get_name(pkg), + alpm_pkg_get_version(pkg), mbsize) == -1) { + pm_fprintf(stderr, PM_LOG_ERROR, _("failed to allocate string\n")); + } This prints an error, but does not actually bail, on failure of asprintf. This is an improvement over the current code as we would know where the error occurred, but feels wrong not to exit. Note that several of the functions affected return void, so erroring out is not a simple "return -1". So, any ideas on the best way to approach this? Allan

3 10

[pacman-dev] [PATCH] Check return value of fgets calls
by Allan McRae 27 Jun '10

27 Jun '10

Prevents compiler warnings with -D_FORTIFY_SOURCE=2 Signed-off-by: Allan McRae <allan(a)archlinux.org> --- lib/libalpm/be_files.c | 8 ++++++-- lib/libalpm/trans.c | 3 +-- 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/lib/libalpm/be_files.c b/lib/libalpm/be_files.c index c26bef2..e2bbe28 100644 --- a/lib/libalpm/be_files.c +++ b/lib/libalpm/be_files.c @@ -657,7 +657,9 @@ int _alpm_db_read(pmdb_t *db, pmpkg_t *info, pmdbinfrq_t inforeq) goto error; } while(!feof(fp)) { - fgets(line, 255, fp); + if(fgets(line, 256, fp) == NULL) { + break; + } _alpm_strtrim(line); if(strcmp(line, "%DEPENDS%") == 0) { while(fgets(line, sline, fp) && strlen(_alpm_strtrim(line))) { @@ -693,7 +695,9 @@ int _alpm_db_read(pmdb_t *db, pmpkg_t *info, pmdbinfrq_t inforeq) snprintf(path, PATH_MAX, "%sdeltas", pkgpath); if((fp = fopen(path, "r"))) { while(!feof(fp)) { - fgets(line, 255, fp); + if(fgets(line, 256, fp) == NULL) { + break; + } _alpm_strtrim(line); if(strcmp(line, "%DELTAS%") == 0) { while(fgets(line, sline, fp) && strlen(_alpm_strtrim(line))) { diff --git a/lib/libalpm/trans.c b/lib/libalpm/trans.c index c042f09..6ebfa8f 100644 --- a/lib/libalpm/trans.c +++ b/lib/libalpm/trans.c @@ -324,8 +324,7 @@ static int grep(const char *fn, const char *needle) while(!feof(fp)) { char line[1024]; int sline = sizeof(line)-1; - fgets(line, sline, fp); - if(feof(fp)) { + if(fgets(line, sline, fp) == NULL || feof(fp)) { continue; } if(strstr(line, needle)) { -- 1.7.1

2 2

[pacman-dev] [PATCH] Check return value of fwrite when copying files
by Allan McRae 27 Jun '10

27 Jun '10

Check that writing to destination file actually occurs in _alpm_copyfile. Required adding a new error (PM_ERR_WRITE) as none of the others appeared appropriate. Prevents compiler warning when using -D_FORTIFY_SOURCE=2. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- lib/libalpm/alpm.h | 1 + lib/libalpm/util.c | 10 +++++++++- 2 files changed, 10 insertions(+), 1 deletions(-) diff --git a/lib/libalpm/alpm.h b/lib/libalpm/alpm.h index 3329132..0c01f21 100644 --- a/lib/libalpm/alpm.h +++ b/lib/libalpm/alpm.h @@ -522,6 +522,7 @@ enum _pmerrno_t { PM_ERR_FILE_CONFLICTS, /* Misc */ PM_ERR_RETRIEVE, + PM_ERR_WRITE, PM_ERR_INVALID_REGEX, /* External library errors */ PM_ERR_LIBARCHIVE, diff --git a/lib/libalpm/util.c b/lib/libalpm/util.c index 383568c..559d62e 100644 --- a/lib/libalpm/util.c +++ b/lib/libalpm/util.c @@ -143,7 +143,15 @@ int _alpm_copyfile(const char *src, const char *dest) /* do the actual file copy */ while((len = fread(buf, 1, CPBUFSIZE, in))) { - fwrite(buf, 1, len, out); + size_t nwritten = 0; + nwritten = fwrite(buf, 1, len, out); + if((nwritten != len) || ferror(out)) { + pm_errno = PM_ERR_WRITE; + _alpm_log(PM_LOG_ERROR, _("error writing to file '%s': %s\n"), + dest, strerror(errno)); + ret = -1; + goto cleanup; + } } /* chmod dest to permissions of src, as long as it is not a symlink */ -- 1.7.1

1 0

Re: [pacman-dev] [PATCH] redirect scriptlet stderr through alpm
by Jonathan Conder 24 Jun '10

24 Jun '10

Sorry, I forgot to rebase against master. Here's the real patch: Fixes FS#18770, and hopefully an occasional deadlock in my frontend as well. For simplicity it redirects all scriptlet output through SCRIPTLET_INFO, and all callbacks in the child process have been replaced for thread-safety. Signed-off-by: Jonathan Conder <j(a)skurvy.no-ip.org> --- lib/libalpm/util.c | 86 +++++++++++++++++++++++++++++---------------------- 1 files changed, 49 insertions(+), 37 deletions(-) diff --git a/lib/libalpm/util.c b/lib/libalpm/util.c index 32eaa44..62851c1 100644 --- a/lib/libalpm/util.c +++ b/lib/libalpm/util.c @@ -448,6 +448,7 @@ int _alpm_run_chroot(const char *root, const char *cmd) { char cwd[PATH_MAX]; pid_t pid; + int pipefd[2]; int restore_cwd = 0; int retval = 0; @@ -471,6 +472,12 @@ int _alpm_run_chroot(const char *root, const char *cmd) /* Flush open fds before fork() to avoid cloning buffers */ fflush(NULL); + if(pipe(pipefd) == -1) { + _alpm_log(PM_LOG_ERROR, _("could not create pipe (%s)\n"), strerror(errno)); + retval = 1; + goto cleanup; + } + /* fork- parent and child each have seperate code blocks below */ pid = fork(); if(pid == -1) { @@ -480,55 +487,60 @@ int _alpm_run_chroot(const char *root, const char *cmd) } if(pid == 0) { - FILE *pipe; /* this code runs for the child only (the actual chroot/exec) */ - _alpm_log(PM_LOG_DEBUG, "chrooting in %s\n", root); + close(1); + close(2); + close(pipefd[0]); + while(dup2(pipefd[1], 1) == -1 && errno == EINTR); + while(dup2(pipefd[1], 2) == -1 && errno == EINTR); + close(pipefd[1]); + if(chroot(root) != 0) { - _alpm_log(PM_LOG_ERROR, _("could not change the root directory (%s)\n"), - strerror(errno)); + printf(_("could not change the root directory (%s)\n"), strerror(errno)); exit(1); } if(chdir("/") != 0) { - _alpm_log(PM_LOG_ERROR, _("could not change directory to / (%s)\n"), - strerror(errno)); + printf(_("could not change directory to / (%s)\n"), strerror(errno)); exit(1); } umask(0022); - pipe = popen(cmd, "r"); - if(!pipe) { - _alpm_log(PM_LOG_ERROR, _("call to popen failed (%s)\n"), - strerror(errno)); - exit(1); - } - while(!feof(pipe)) { - char line[PATH_MAX]; - if(fgets(line, PATH_MAX, pipe) == NULL) - break; - alpm_logaction("%s", line); - EVENT(handle->trans, PM_TRANS_EVT_SCRIPTLET_INFO, line, NULL); - } - retval = pclose(pipe); - exit(WEXITSTATUS(retval)); + execl("/bin/sh", "sh", "-c", cmd, (char *) NULL); + printf(_("call to execl failed (%s)\n"), strerror(errno)); + exit(1); } else { /* this code runs for the parent only (wait on the child) */ - pid_t retpid; int status; - do { - retpid = waitpid(pid, &status, 0); - } while(retpid == -1 && errno == EINTR); - if(retpid == -1) { - _alpm_log(PM_LOG_ERROR, _("call to waitpid failed (%s)\n"), - strerror(errno)); - retval = 1; - goto cleanup; + FILE *pipe; + + close(pipefd[1]); + pipe = fdopen(pipefd[0], "r"); + if(pipe == NULL) { + close(pipefd[0]); } else { - /* check the return status, make sure it is 0 (success) */ - if(WIFEXITED(status)) { - _alpm_log(PM_LOG_DEBUG, "call to waitpid succeeded\n"); - if(WEXITSTATUS(status) != 0) { - _alpm_log(PM_LOG_ERROR, _("command failed to execute correctly\n")); - retval = 1; - } + while(!feof(pipe)) { + char line[PATH_MAX]; + if(fgets(line, PATH_MAX, pipe) == NULL) + break; + alpm_logaction("%s", line); + EVENT(handle->trans, PM_TRANS_EVT_SCRIPTLET_INFO, line, NULL); + } + fclose(pipe); + } + + while(waitpid(pid, &status, 0) == -1) { + if(errno != EINTR) { + _alpm_log(PM_LOG_ERROR, _("call to waitpid failed (%s)\n"), strerror(errno)); + retval = 1; + goto cleanup; + } + } + + /* check the return status, make sure it is 0 (success) */ + if(WIFEXITED(status)) { + _alpm_log(PM_LOG_DEBUG, "call to waitpid succeeded\n"); + if(WEXITSTATUS(status) != 0) { + _alpm_log(PM_LOG_ERROR, _("command failed to execute correctly\n")); + retval = 1; } } } -- 1.7.1

3 10

[pacman-dev] [PATCH] repo-add: try symlink, then hardlink, then copy for DB file
by Dan McGee 24 Jun '10

24 Jun '10

We were seeing some issues when trying to create our new database alias using symlinks on certain filesystems (see FS#19907). Have a fallback method in place where we first try a symlink, then a hard link, then just copy the database if all else fails. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- scripts/repo-add.sh.in | 5 ++++- 1 files changed, 4 insertions(+), 1 deletions(-) diff --git a/scripts/repo-add.sh.in b/scripts/repo-add.sh.in index 7bde600..2838f81 100644 --- a/scripts/repo-add.sh.in +++ b/scripts/repo-add.sh.in @@ -487,7 +487,10 @@ if (( success )); then [[ -f $REPO_DB_FILE ]] && mv -f "$REPO_DB_FILE" "${REPO_DB_FILE}.old" [[ -f $tmpdir/$filename ]] && mv "$tmpdir/$filename" "$REPO_DB_FILE" - ln -sf "$REPO_DB_FILE" "${REPO_DB_FILE%.tar.*}" + dblink="${REPO_DB_FILE%.tar.*}" + ln -sf "$REPO_DB_FILE" "$dblink" 2>/dev/null || \ + ln -f "$REPO_DB_FILE" "$dblink" 2>/dev/null || \ + cp "$REPO_DB_FILE" "$dblink" else msg "$(gettext "No packages modified, nothing to do.")" exit 1 -- 1.7.1

2 1

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.4.0-29-g630e750
by dan＠archlinux.org 24 Jun '10

24 Jun '10

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 630e7508ee44f2c41ba3f5d17c0c9ac15b15af3b (commit) via 71330a431040377ae9a37989fcfd43cee7596194 (commit) via c6a94a8d2204cfdec0f782f647903d597b4faf57 (commit) via 6f4f9c1b66ed859b3679d70e15c003ab4907b823 (commit) via 07a9effdd06345d09f09cdc92e23c937d8fa94b5 (commit) via 708f186f98a0c2094225aa94ac8a139ac3a9163e (commit) via 226c137245192444085d03a7f841b35afe99791c (commit) via 2222e9f8dfe6e149262f629571b3432718351d3f (commit) from 67d71ea932a18d27dbb87f4f364ae9159b4edc96 (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 630e7508ee44f2c41ba3f5d17c0c9ac15b15af3b Merge: 71330a4 6f4f9c1 Author: Dan McGee <dan(a)archlinux.org> Date: Thu Jun 24 08:32:33 2010 -0500 Merge branch 'maint' commit 71330a431040377ae9a37989fcfd43cee7596194 Author: Eric Bélanger <snowmaniscool(a)gmail.com> Date: Wed Jun 23 01:35:56 2010 -0400 makepkg: Place source packages symlinks in build dir when SRCPKGDEST is used Signed-off-by: Eric Bélanger <snowmaniscool(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit c6a94a8d2204cfdec0f782f647903d597b4faf57 Author: Andres P <aepd87(a)gmail.com> Date: Mon Jun 21 12:25:37 2010 -0430 makepkg: fix regression in split package function checking Commit 13748ca0529 inversed the nature of one test wherein the if clause would throw a fatal error if a legitimate package function was defined in PKGBUILD. Signed-off-by: Andres P <aepd87(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: contrib/bash_completion | 17 ++++++++++++----- scripts/makepkg.sh.in | 24 +++++++++++++++++++----- scripts/rankmirrors.sh.in | 12 ++++++------ 3 files changed, 37 insertions(+), 16 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [GIT] The official pacman repository branch, maint, updated. v3.4.0-9-g6f4f9c1
by dan＠archlinux.org 24 Jun '10

24 Jun '10

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, maint has been updated via 6f4f9c1b66ed859b3679d70e15c003ab4907b823 (commit) via 07a9effdd06345d09f09cdc92e23c937d8fa94b5 (commit) via 708f186f98a0c2094225aa94ac8a139ac3a9163e (commit) via 226c137245192444085d03a7f841b35afe99791c (commit) via 2222e9f8dfe6e149262f629571b3432718351d3f (commit) from c2cf6a14cf44400d0ef249b38d37eb04d3424bf4 (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 ----------------------------------------------------------------- ----------------------------------------------------------------------- Summary of changes: contrib/bash_completion | 17 ++++++++++++----- scripts/makepkg.sh.in | 10 +++++++--- scripts/rankmirrors.sh.in | 12 ++++++------ 3 files changed, 25 insertions(+), 14 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH] makepkg: Place source packages symlinks in build dir when SRCPKGDEST is used
by Eric Bélanger 23 Jun '10

23 Jun '10

Signed-off-by: Eric Bélanger <snowmaniscool(a)gmail.com> --- scripts/makepkg.sh.in | 12 +++++++++++- 1 files changed, 11 insertions(+), 1 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index dbc4047..0219015 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -153,7 +153,7 @@ clean_up() { # clean up dangling symlinks to packages for pkg in ${pkgname[@]}; do - for file in ${pkg}-*-*-${CARCH}${PKGEXT}; do + for file in ${pkg}-*-*-${CARCH}${PKGEXT} ${pkg}-*-*-${CARCH}${SRCEXT}; do if [[ -h $file && ! -e $file ]]; then rm -f $file fi @@ -1126,6 +1126,16 @@ create_srcpackage() { error "$(gettext "Failed to create source package file.")" exit 1 # TODO: error code fi + + if (( ! ret )) && [[ "$SRCPKGDEST" != "${startdir}" ]]; then + ln -sf "${pkg_file}" "${pkg_file/$SRCPKGDEST/$startdir}" + ret=$? + fi + + if (( ret )); then + warning "$(gettext "Failed to create symlink to source package file.")" + fi + cd "${startdir}" rm -rf "${srclinks}" } -- 1.7.1

2 1

[pacman-dev] [PATCH] makepkg: Place source packages symlinks in build dir when SRCPKGDEST is used
by Eric Bélanger 23 Jun '10

23 Jun '10

Signed-off-by: Eric Bélanger <snowmaniscool(a)gmail.com> --- scripts/makepkg.sh.in | 20 ++++++++++++++++++++ 1 files changed, 20 insertions(+), 0 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index dbc4047..83d656f 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -159,6 +159,16 @@ clean_up() { fi done done + + # clean up dangling symlinks to source packages + for pkg in ${pkgname[@]}; do + for file in ${pkg}-*-*-${CARCH}${SRCEXT}; do + if [[ -h $file && ! -e $file ]]; then + rm -f $file + fi + done + done + fi fi @@ -1126,6 +1136,16 @@ create_srcpackage() { error "$(gettext "Failed to create source package file.")" exit 1 # TODO: error code fi + + if (( ! ret )) && [[ "$SRCPKGDEST" != "${startdir}" ]]; then + ln -sf "${pkg_file}" "${pkg_file/$SRCPKGDEST/$startdir}" + ret=$? + fi + + if (( ret )); then + warning "$(gettext "Failed to create symlink to source package file.")" + fi + cd "${startdir}" rm -rf "${srclinks}" } -- 1.7.1

2 2

[pacman-dev] [PATCH] makepkg: prevent error trap activation in bash-3.2
by Allan McRae 23 Jun '10

23 Jun '10

Running "pacman -T foo" is expected to return a non-zero value when "foo" is not installed. This sets of the error trap in bash-3.2 but not bash 4.x. Work around this by disabling the error trap around this pacman call as we are manually checking the return value anyway. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- scripts/makepkg.sh.in | 10 +++++++--- 1 files changed, 7 insertions(+), 3 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 4f9f89b..2699f63 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -382,11 +382,15 @@ run_pacman() { } check_deps() { - (( $# > 0 )) || return + (( $# > 0 )) || return 0 + # Disable error trap in pacman subshell call as this breaks bash-3.2 compatibility + # Also, a non-zero return value is not unexpected and we are manually dealing them + set +E local ret=0 - pmout=$(run_pacman -T "$@") - ret=$? + pmout=$(run_pacman -T "$@") || ret=$? + set -E + if (( ret == 127 )); then #unresolved deps echo "$pmout" elif (( ret )); then -- 1.7.1

2 2

[pacman-dev] pacman shell scripts and bash3.2 compat - Was: [PATCH 04/11] makepkg: less code duplication in create_package
by Andres P 23 Jun '10

23 Jun '10

On Sun, Jun 20, 2010 at 6:09 AM, Allan McRae <allan(a)archlinux.org> wrote: > That is a bash4 command so will break cygwin compatibility (which currently > has bash-3.2). Resubmit with something like "tr [:lower:] [:upper:]" usage. > > (I know the ${i^^} is used elsewhere, but that slipped passed my notice and > will be being reverted. > bash_completion also has bash 4 only items... Please write a style guide for all shell scripts.... Andres P

2 3

[pacman-dev] [PATCH 1/3] rankmirror: pipe errors to stderr
by Andres P 23 Jun '10

23 Jun '10

If this is to be scripted with AIF or another tool, it needs to respect stderr. Signed-off-by: Andres P <aepd87(a)gmail.com> --- scripts/rankmirrors.sh.in | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/scripts/rankmirrors.sh.in b/scripts/rankmirrors.sh.in index 9b4e973..283fb46 100644 --- a/scripts/rankmirrors.sh.in +++ b/scripts/rankmirrors.sh.in @@ -49,7 +49,7 @@ version() { } err() { - echo "$1" + echo "$1" >&2 exit 1 } -- 1.7.1

3 6

Re: [pacman-dev] [arch-dev-public] Pacman Architecture option (was: [signoff] pacman 3.4.0-2)
by Dan McGee 22 Jun '10

22 Jun '10

On Tue, Jun 22, 2010 at 10:55 AM, Thomas Bächler <thomas(a)archlinux.org> wrote: > Am 22.06.2010 17:42, schrieb Dan McGee: >> This is the last thing I will say on this. Where the heck were you >> guys 11 months ago? Surely not following development, but you want to >> piss about it now. >> >> Discussion: >> http://mailman.archlinux.org/pipermail/pacman-dev/2009-July/008965.html >> http://mailman.archlinux.org/pipermail/pacman-dev/2009-July/008968.html >> http://mailman.archlinux.org/pipermail/pacman-dev/2009-July/008970.html >> >> Patches (with no feedback, mind you): >> http://mailman.archlinux.org/pipermail/pacman-dev/2009-August/009193.html > > I do read pacman-dev on occasion, but not regularly, and I don't follow > each patch thread or new feature. I heard that the Architecture option > was coming, but was never too concerned with the details (just random > happiness about it) - I didn't even know it would be an option, but > thought it was just default behaviour. > I even discussed the issue with Xavier I think, maybe it was Allan, it's > too long ago. > > If you read my emails, you will see that I did not complain a single > time, but rather wanted to know what the reasoning behind this decision > was (which I'd still like to know btw). OK, I lied about no more replies becuase I did forget to put this in my original email- the primary driving reason from my point of view is "If you do nothing, nothing changes". e.g. for anyone not adding this to their pacman.conf, they won't have to worry about this feature getting in the way. With that said, I do realize this would not inhibet most people to be turned on by default, but I think we saw this as a precautionary measure rather than something we should force on people. I've CC-ed to pacman-dev; if I end up being in the minority on this then I'll take a patch to make the default be "auto" (and we will then need to add an explicit "none" option and document it). > And while I think that this should be changed to the way I suggested > (which would be a trivial change), for the reasons I posted, I can live > with it either way.

2 1

[pacman-dev] bash-3 only error trap activation
by Allan McRae 22 Jun '10

22 Jun '10

Hi, When running makepkg in bash-3.2 I am getting a couple of error trap activations that we do not get in bash-4.1. Here is the problem area: check_deps() { (( $# > 0 )) || return <- HERE local ret=0 pmout=$(run_pacman -T "$@") <- HERE ret=$? The first one gets set off anytime "depends" or "makedepends" are empty and can be fixed by using "|| return 0", but the second is doing my head in... Of course, turning off the error trap around those commands makes it work, and that may be an OK approach given we are dealing with all error states below that. If you want a bash-3.2 package to test grab one from http://allanmcrae.com/packages/bash3-3.2.051-1.src.tar.gz . Test on a PKGBUILD with unsatisfied dependencies. Allan

2 7

[pacman-dev] [PATCH] Add check() function for testsuites.
by Jeff 22 Jun '10

22 Jun '10

If this is an accepted feature, I can resend the patch with manpage updates. Signed-off-by: Jeff C <jeff(a)kcaccess.com> --- etc/makepkg.conf.in | 5 +++-- scripts/makepkg.sh.in | 24 ++++++++++++++++++++++-- 2 files changed, 25 insertions(+), 4 deletions(-) diff --git a/etc/makepkg.conf.in b/etc/makepkg.conf.in index f0d1c44..b3c4b4c 100644 --- a/etc/makepkg.conf.in +++ b/etc/makepkg.conf.in @@ -58,7 +58,7 @@ BUILDENV=(fakeroot !distcc color !ccache) # These are default values for the options=() settings ######################################################################### # -# Default: OPTIONS=(strip docs libtool emptydirs zipman purge) +# Default: OPTIONS=(strip docs libtool emptydirs zipman purge !check) # A negated option will do the opposite of the comments below. # #-- strip: Strip symbols from binaries/libraries in STRIP_DIRS @@ -67,8 +67,9 @@ BUILDENV=(fakeroot !distcc color !ccache) #-- emptydirs: Leave empty directories in packages #-- zipman: Compress manual (man and info) pages in MAN_DIRS with gzip #-- purge: Remove files specified by PURGE_TARGETS +#-- check: Run the check() function if it exists. # -OPTIONS=(strip docs libtool emptydirs zipman purge) +OPTIONS=(strip docs libtool emptydirs zipman purge !check) #-- File integrity checks to use. Valid: md5, sha1, sha256, sha384, sha512 INTEGRITY_CHECK=(md5) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 678359f..2a659e6 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -44,7 +44,7 @@ startdir="$PWD" srcdir="$startdir/src" pkgdir="$startdir/pkg" -packaging_options=('strip' 'docs' 'libtool' 'emptydirs' 'zipman' 'purge') +packaging_options=('strip' 'docs' 'libtool' 'emptydirs' 'zipman' 'purge' 'check') other_options=('ccache' 'distcc' 'makeflags' 'force') splitpkg_overrides=('pkgver' 'pkgrel' 'pkgdesc' 'arch' 'license' 'groups' \ 'depends' 'optdepends' 'provides' 'conflicts' 'replaces' \ @@ -71,6 +71,7 @@ SOURCEONLY=0 IGNOREARCH=0 HOLDVER=0 BUILDFUNC=0 +CHECKFUNC=0 PKGFUNC=0 SPLITPKG=0 PKGLIST="" @@ -143,6 +144,9 @@ clean_up() { if (( BUILDFUNC )); then rm -f "${pkgbase}-${pkgver}-${pkgrel}-${CARCH}-build.log"* fi + if (( CHECKFUNC )); then + rm -f "${pkgbase}-${pkgver}-${pkgrel}-${CARCH}-check.log"* + fi if (( PKGFUNC )); then rm -f "${pkgbase}-${pkgver}-${pkgrel}-${CARCH}-package.log"* elif (( SPLITPKG )); then @@ -735,7 +739,11 @@ run_function() { fi msg "$(gettext "Starting %s()...")" "$pkgfunc" - cd "$srcdir" + if [[ -d "$srcdir"/"$pkgname"-"$pkgver" ]]; then + cd "$srcdir"/"$pkgname"-"$pkgver" + else + cd "$srcdir" + fi # ensure all necessary build variables are exported export CFLAGS CXXFLAGS LDFLAGS MAKEFLAGS CHOST @@ -798,6 +806,10 @@ run_build() { run_function "build" } +run_check() { + [[ $(check_option check) = "y" ]] && run_function "check" +} + run_package() { if [[ -z $1 ]]; then pkgfunc="package" @@ -1772,6 +1784,9 @@ fi if [[ $(type -t build) = "function" ]]; then BUILDFUNC=1 fi +if [[ $(type -t check) = "function" ]]; then + CHECKFUNC=1 +fi if [[ $(type -t package) = "function" ]]; then PKGFUNC=1 elif [[ $SPLITPKG -eq 0 && $(type -t package_${pkgname}) = "function" ]]; then @@ -1835,6 +1850,9 @@ if (( INFAKEROOT )); then if (( ! REPKG )); then if (( BUILDFUNC )); then run_build + if (( CHECKFUNC )); then + run_check + fi tidy_install fi else @@ -1960,6 +1978,7 @@ else if (( ! REPKG )); then devel_update (( BUILDFUNC )) && run_build + (( CHECKFUNC )) && run_check fi if (( ! SPLITPKG )); then if (( PKGFUNC )); then @@ -1991,6 +2010,7 @@ else if (( ! REPKG && ( PKGFUNC || SPLITPKG ) )); then devel_update (( BUILDFUNC )) && run_build + (( CHECKFUNC )) && run_check cd "$startdir" fi -- 1.7.1

2 3

[pacman-dev] [PATCH 1/2] makepkg: fix regression in split package function checking
by Andres P 21 Jun '10

21 Jun '10

Commit 13748ca0529 inversed the nature of one test wherein the if clause would throw a fatal error if a legitimate package function was defined in PKGBUILD. Signed-off-by: Andres P <aepd87(a)gmail.com> --- scripts/makepkg.sh.in | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index dbc4047..f2ec78e 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -1255,7 +1255,7 @@ check_sanity() { if (( ${#pkgname[@]} > 1 )); then for pkg in ${pkgname[@]}; do - if declare -f package_${pkg} >/dev/null; then + if ! declare -f package_${pkg} >/dev/null; then error "$(gettext "missing package function for split package '%s'")" "$pkg" return 1 fi -- 1.7.1

2 4

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.4.0-21-g67d71ea
by dan＠archlinux.org 21 Jun '10

21 Jun '10

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 67d71ea932a18d27dbb87f4f364ae9159b4edc96 (commit) via c2cf6a14cf44400d0ef249b38d37eb04d3424bf4 (commit) via 6c00ca8f234cfcbfde7e36b68d85f9dcab5716c8 (commit) via 7fc50d79508f30e74e5ea8d9c801549e7f52229d (commit) via dc817a2061699cd1f33ca93f0d93a1fbc2f33ea1 (commit) from 59c47aaf529df02ec1577fe727c3c84d13592666 (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 67d71ea932a18d27dbb87f4f364ae9159b4edc96 Merge: 59c47aa c2cf6a1 Author: Dan McGee <dan(a)archlinux.org> Date: Sun Jun 20 21:01:32 2010 -0500 Merge branch 'maint' ----------------------------------------------------------------------- Summary of changes: configure.ac | 6 +++--- lib/libalpm/sync.c | 6 +++++- scripts/makepkg.sh.in | 10 +++++----- test/pacman/tests/ignore006.py | 10 ++++++++++ 4 files changed, 23 insertions(+), 9 deletions(-) create mode 100644 test/pacman/tests/ignore006.py hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [GIT] The official pacman repository branch, maint, updated. v3.4.0-4-gc2cf6a1
by dan＠archlinux.org 21 Jun '10

21 Jun '10

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, maint has been updated via c2cf6a14cf44400d0ef249b38d37eb04d3424bf4 (commit) via 6c00ca8f234cfcbfde7e36b68d85f9dcab5716c8 (commit) from 7fc50d79508f30e74e5ea8d9c801549e7f52229d (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 ----------------------------------------------------------------- ----------------------------------------------------------------------- Summary of changes: lib/libalpm/sync.c | 6 +++++- scripts/makepkg.sh.in | 10 +++++----- test/pacman/tests/ignore006.py | 10 ++++++++++ 3 files changed, 20 insertions(+), 6 deletions(-) create mode 100644 test/pacman/tests/ignore006.py hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH] Handle sync target + ignore properly
by Dan McGee 21 Jun '10

21 Jun '10

Rather than say we can't find the target after saying "No, I guess I don't want to install this", we should make sure the ignored status gets passed all the way through. This fixes FS#19866. Pactest is also included that failed before due to the fact that we normally treat an unfound package as a reason to exit with a non-zero status. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- Candidate for maint, it should be pretty straightforward. -Dan lib/libalpm/sync.c | 6 +++++- test/pacman/tests/ignore006.py | 10 ++++++++++ 2 files changed, 15 insertions(+), 1 deletions(-) create mode 100644 test/pacman/tests/ignore006.py diff --git a/lib/libalpm/sync.c b/lib/libalpm/sync.c index 67afd39..68ee8dc 100644 --- a/lib/libalpm/sync.c +++ b/lib/libalpm/sync.c @@ -291,7 +291,11 @@ int _alpm_sync_target(alpm_list_t *dbs_sync, char *target) } if(!found) { - RET_ERR(PM_ERR_PKG_NOT_FOUND, -1); + /* pass through any 'found but ignored' errors */ + if(pm_errno != PM_ERR_PKG_IGNORED) { + pm_errno = PM_ERR_PKG_NOT_FOUND; + } + return(-1); } return(0); diff --git a/test/pacman/tests/ignore006.py b/test/pacman/tests/ignore006.py new file mode 100644 index 0000000..4c1c1ca --- /dev/null +++ b/test/pacman/tests/ignore006.py @@ -0,0 +1,10 @@ +self.description = "Sync with target in ignore list and say no" + +pkg = pmpkg("package1") +self.addpkg2db("sync", pkg) + +self.option["IgnorePkg"] = ["package1"] +self.args = "--ask=1 -S %s" % pkg.name + +self.addrule("PACMAN_RETCODE=0") +self.addrule("!PKG_EXIST=package1") -- 1.7.1

2 1

[pacman-dev] [PATCH] makepkg: revert bash4-ism
by Allan McRae 21 Jun '10

21 Jun '10

Commit 3d67d9b1 introduced a bash4 string manipulation. Revert that in order retain compatibility with bash-3.2 which is still widely used. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- This is for the maint branch scripts/makepkg.sh.in | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index c6bc738..3bc6019 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -1665,7 +1665,7 @@ if (( CLEANCACHE )); then echo -n "$(gettext " Are you sure you wish to do this? ")" echo -n "$(gettext "[y/N]")" read answer - answer="${answer^^}" + answer=$(echo $answer | tr '[:lower:]' '[:upper:]') if [[ $answer = $(gettext YES) || $answer = $(gettext Y) ]]; then rm "$SRCDEST"/* if (( $? )); then -- 1.7.1

3 7

[pacman-dev] [PATCH] makepkg: remove bash4 style output redirection
by Allan McRae 21 Jun '10

21 Jun '10

convert bash-4 "&>" style redirection to "2>&1" to maintain compatibility with bash-3.2 Signed-off-by: Allan McRae <allan(a)archlinux.org> --- scripts/makepkg.sh.in | 10 +++++----- 1 files changed, 5 insertions(+), 5 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 126c5f5..8e2a523 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -370,7 +370,7 @@ download_file() { run_pacman() { local ret=0 if (( ! ASROOT )) && [[ $1 != "-T" && $1 != "-Qq" ]]; then - if [ "$(type -p sudo)" ] && sudo -l $PACMAN &>/dev/null; then + if [ "$(type -p sudo)" ] && sudo -l $PACMAN >/dev/null 2>&1; then sudo $PACMAN $PACMAN_OPTS "$@" || ret=$? else su -c "$PACMAN $PACMAN_OPTS $*" || ret=$? @@ -421,7 +421,7 @@ handle_deps() { # avoid triggering the ERR trap local restoretrap=$(trap -p ERR) trap - ERR - source /etc/profile &>/dev/null + source /etc/profile >/dev/null 2>&1 eval $restoretrap return $R_DEPS_SATISFIED @@ -479,7 +479,7 @@ remove_deps() { download_sources() { msg "$(gettext "Retrieving Sources...")" - pushd "$SRCDEST" &>/dev/null + pushd "$SRCDEST" >/dev/null 2>&1 local netfile for netfile in "${source[@]}"; do @@ -515,7 +515,7 @@ download_sources() { ln -s "$SRCDEST/$file" "$srcdir/" done - popd &>/dev/null + popd >/dev/null 2>&1 } get_integlist() { @@ -976,7 +976,7 @@ check_package() { done # check for references to the build directory - if grep -R "${srcdir}" "${pkgdir}" &>/dev/null; then + if grep -R "${srcdir}" "${pkgdir}" >/dev/null 2>&1; then warning "$(gettext "Package contains reference to %s")" "\$srcdir" fi } -- 1.7.1

5 5

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.4.0-16-g59c47aa
by dan＠archlinux.org 21 Jun '10

21 Jun '10

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 59c47aaf529df02ec1577fe727c3c84d13592666 (commit) via d7dccd541962f0dd8bf323ae11633e595bfb4922 (commit) via 3e4d2c3aa65416487939148828afb385de2ee146 (commit) via f5f107674bd34b281681d73a32e4dde1f9d0856f (commit) via 1b93a116e73050d2a037843647b961432b2a0dc8 (commit) via 70418e48f6c0033a1bf942fec0bbc001d3d0d8f5 (commit) via c7a37d039bc4361077fdf77180b35b81bd8b12a7 (commit) via cd042640c66e73f3fe4f5299a36e4cca5384b34b (commit) via 13748ca052975cbf6354cfc4554f49a3d8ff46fe (commit) via 34229c562571d458446c9e4da197fcae251c2f6b (commit) from 09aae4b7a5891ad738754a22f956c83911d1d9eb (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 59c47aaf529df02ec1577fe727c3c84d13592666 Author: Allan McRae <allan(a)archlinux.org> Date: Sat Jun 19 18:55:08 2010 +1000 Clarify testing within conditional statements Follow the HACKING guidelines and always use != 0 or == 0 rather than negation within conditional statements to improve clarity. Most of these are !strcmp usages which is the example of what not to do in the HACKING document. Signed-off-by: Allan McRae <allan(a)archlinux.org> commit d7dccd541962f0dd8bf323ae11633e595bfb4922 Author: Andres P <aepd87(a)gmail.com> Date: Fri Jun 18 19:21:10 2010 -0430 makepkg: remove unnecessary use of sort pacman -Qq output is sorted according to the users LC_COLLATE setting as is needed for comm. Signed-off-by: Andres P <aepd87(a)gmail.com> [Allan: adjust commit message] Signed-off-by: Allan McRae <allan(a)archlinux.org> commit 3e4d2c3aa65416487939148828afb385de2ee146 Author: Andres P <aepd87(a)gmail.com> Date: Sat Jun 19 01:25:15 2010 -0430 libalpm: compare pkgname with strcoll Use strcoll to compare package names to provide output sorted according to a users LC_COLLATE settings. Signed-off-by: Andres P <aepd87(a)gmail.com> [Allan: added commit message] Signed-off-by: Allan McRae <allan(a)archlinux.org> commit f5f107674bd34b281681d73a32e4dde1f9d0856f Author: Andres P <aepd87(a)gmail.com> Date: Fri Jun 18 19:21:09 2010 -0430 makepkg: use parameter expansion instead of basename Signed-off-by: Andres P <aepd87(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> commit 1b93a116e73050d2a037843647b961432b2a0dc8 Author: Andres P <aepd87(a)gmail.com> Date: Thu Jun 17 08:14:51 2010 -0430 makepkg: fix inconsistent output when checking dependencies Instead of writing: ==> Checking Runtime Dependencies... ==> Checking Buildtime Dependencies... ==> Installing missing dependencies... Just make it homogeneous: ==> Checking runtime dependencies... ==> Checking buildtime dependencies... ==> Installing missing dependencies... Signed-off-by: Andres P <aepd87(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> commit 70418e48f6c0033a1bf942fec0bbc001d3d0d8f5 Author: Andres P <aepd87(a)gmail.com> Date: Thu Jun 17 08:14:49 2010 -0430 makepkg: merge in-line splitpkg handling logic Ease maintainace; the two parts that have been combined into a function were identical. Signed-off-by: Andres P <aepd87(a)gmail.com> [Allan: rename function] Signed-off-by: Allan McRae <allan(a)archlinux.org> commit c7a37d039bc4361077fdf77180b35b81bd8b12a7 Author: Andres P <aepd87(a)gmail.com> Date: Thu Jun 17 08:14:48 2010 -0430 makepkg: add to variables without expanding them Use foo+=" bar" instead of foo="${foo} bar" Signed-off-by: Andres P <aepd87(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> commit cd042640c66e73f3fe4f5299a36e4cca5384b34b Author: Andres P <aepd87(a)gmail.com> Date: Thu Jun 17 08:14:46 2010 -0430 makepkg: less code repetition for empty variable checking Signed-off-by: Andres P <aepd87(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> commit 13748ca052975cbf6354cfc4554f49a3d8ff46fe Author: Andres P <aepd87(a)gmail.com> Date: Thu Jun 17 08:14:42 2010 -0430 makepkg: use "declare -f" to test for function presence Signed-off-by: Andres P <aepd87(a)gmail.com> Signed-off-by: Allan McRae <allan(a)archlinux.org> commit 34229c562571d458446c9e4da197fcae251c2f6b Author: Andres P <aepd87(a)gmail.com> Date: Thu Jun 17 08:14:41 2010 -0430 makepkg: fix variable checks when writing pkginfo Regression caused by c71fe7db. Was checking for "optdepend" and "conflict" rather than "optdepends" and "conflicts" when populating .PKGINFO. Signed-off-by: Andres P <aepd87(a)gmail.com> [Allan: adjust commit message] Signed-off-by: Allan McRae <allan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: lib/libalpm/be_package.c | 34 +++++++------- lib/libalpm/conflict.c | 4 +- lib/libalpm/deps.c | 2 +- lib/libalpm/package.c | 2 +- lib/libalpm/trans.c | 4 +- lib/libalpm/util.c | 2 +- scripts/makepkg.sh.in | 107 ++++++++++++++++++++------------------------- src/pacman/callback.c | 2 +- src/pacman/package.c | 2 +- src/pacman/sync.c | 8 ++-- src/pacman/upgrade.c | 2 +- src/pacman/util.c | 6 +- src/util/testdb.c | 2 +- 13 files changed, 83 insertions(+), 94 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH] Move gpg pactests to new location
by Allan McRae 20 Jun '10

20 Jun '10

Signed-off-by: Allan McRae <allan(a)archlinux.org> --- For the gpg branch. Could be rebased into earlier commits... {pactest => test/pacman}/tests/sign001.py | 0 {pactest => test/pacman}/tests/sign002.py | 0 {pactest => test/pacman}/tests/smoke002.py | 0 3 files changed, 0 insertions(+), 0 deletions(-) rename {pactest => test/pacman}/tests/sign001.py (100%) rename {pactest => test/pacman}/tests/sign002.py (100%) rename {pactest => test/pacman}/tests/smoke002.py (100%) diff --git a/pactest/tests/sign001.py b/test/pacman/tests/sign001.py similarity index 100% rename from pactest/tests/sign001.py rename to test/pacman/tests/sign001.py diff --git a/pactest/tests/sign002.py b/test/pacman/tests/sign002.py similarity index 100% rename from pactest/tests/sign002.py rename to test/pacman/tests/sign002.py diff --git a/pactest/tests/smoke002.py b/test/pacman/tests/smoke002.py similarity index 100% rename from pactest/tests/smoke002.py rename to test/pacman/tests/smoke002.py -- 1.7.1

2 1