| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
Store "vulnerable" and "upgrade" packages in a table, and use that
data to determine which packages cannot be upgraded, and which
packages actually cause upgrades
svn path=/trunk/gentoolkit/; revision=648
http://git.overlays.gentoo.org/gitweb/?p=proj/gentoolkit.git;a=commit;h=b6a2a23926d54ccfa9a1ce331c1bc97dbe2c73d1
|
|
|
|
|
|
|
|
|
|
|
|
| |
This allows to differentiate between situations where
the system is unaffected and unexistance of an upgrade path.
Previously, the glsa-check would treat GLSAs that had no
upgrade path (such as mask glsas) as not affecting the system.
svn path=/trunk/gentoolkit/; revision=647
http://git.overlays.gentoo.org/gitweb/?p=proj/gentoolkit.git;a=commit;h=ef38a394c5c2f5901173a53914705730850f9b3f
|
|
|
|
|
|
|
|
| |
StringIO
svn path=/trunk/gentoolkit/; revision=645
http://git.overlays.gentoo.org/gitweb/?p=proj/gentoolkit.git;a=commit;h=88c18ffe936e602c45dcaa7500961623c49697ca
|
|
|
|
|
| |
This helps to ensure consistent results, regardless of whether we're
using Python 2 or Python 3.
|
| |
|
| |
|
|
|
|
| |
Also, remove deprecated code from doebuild (with hardcoded pipe fd).
|
|
|
|
|
|
|
| |
By using allocated file descriptors for keys in fd_pipes, we naturally
avoid interference with callers such as FileDigester and MergeProcess.
See the _setup_pipes docstring for more benefits of this allocation
approach.
|
| |
|
|
|
|
|
| |
Special thanks to Brian Harring, author of the mirror-dist program from
which emirrordist is derived.
|
| |
|
| |
|
|
|
|
| |
It got disabled in commit 06d31ef00da24352a6614f20bccfc892d2120ed9.
|
|
|
|
| |
It got disabled by commit 439151bb5b0bd2ac4c9d2dd218fe377ce0b988e6.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
before the patch:
$ repoman full
RepoMan scours the neighborhood...
metadata.missing 1
dev-haskell/qthaskellc/metadata.xml
SRC_URI.mirror 1
dev-haskell/qthaskellc/qthaskellc-1.1.4.ebuild: 'http://download.berlios.de/' found in thirdpartymirrors
after the patch:
$ repoman full
RepoMan scours the neighborhood...
metadata.missing 1
dev-haskell/qthaskellc/metadata.xml
SRC_URI.mirror 1
dev-haskell/qthaskellc/qthaskellc-1.1.4.ebuild: 'http://download.berlios.de/' found in thirdpartymirrors: 'mirror://berlios'
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
We want the effective ACCEPT_KEYWORDS to correspond to the current
profile's arch (rather than using wildcards).
|
| |
|
| |
|
|
|
|
|
|
|
| |
Buggy code in python's multiprocessing/process.py closes sys.stdin and
reassigns it to open(os.devnull), but fails to update the corresponding
__stdin__ reference. So, detect that case and handle it appropriately.
The buggy code is visible in http://hg.python.org/lookup/r73708.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
This prevents us from trying to hardlink duplicate splitdebug files,
avoiding ln "File exists" failure as reported in bug #445336.
|
|
|
|
|
|
| |
This reverts commit bba20c632e79426cd006dd2b732fcaed079ac43a.
The sort needs to be inside the subshell in order for it to
work as intended. A correct patch will follow.
|
|
|
|
|
| |
This prevents us from trying to hardlink duplicate splitdebug files,
avoiding ln "File exists" failure as reported in bug #445336.
|
| |
|
|
|
|
|
|
|
|
| |
* changes dyn_spec to use a sources_dir in the sandbox
* remove deprecated Buildroot from specfile
* call rpmbuild with --target instead of Buildroot
* grab machine_name from $CHOST and use as --target in rpmbuild
* use dir in sandbox instead of /usr/src/rpm/RPMS
|
|
|
|
|
| |
The raw metadata is considered a private low-level internal. This
makes the Package class consistent with the _pkg_str class.
|
| |
|
|
|
|
| |
subshell.
|
|
|
|
|
|
| |
No functional changes here.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
|
|
|
|
| |
package_manager_build_group().
|
| |
|
|
|
|
|
| |
This fixes a regression since the -a option was removed in
commit e672b5ccbbbe38eda5c733fc799a20d92a8bdf75.
|
| |
|
| |
|
|
|
|
|
|
|
| |
Since strip creates a new inode, we need to know the initial set of
inodes in advance, so that we can avoid interference due to trying
to strip the same (hardlinked) file multiple times in parallel.
See bug #421099.
|
| |
|