| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Narrow scope of merge locks to improve performance.
Instead of locking the DB for the entire package merge, just lock it
when we actually need to do so. Also add locks around conf_mem_file
updating and pkg_* phases.
Locking in pkg_* phases can be turned off with
FEATURES="no-ebuild-locks" if you use ebuilds that are careful not
to mess with each other during theses phases. The default is to leave
this locking enabled.
Given this new locking, I've improved the scheduler to run merge jobs
in parallel.
Time required for merging 348 packages with --usepkgonly:
- Before patch: 29m50s
- After patch: 10m2s
- After patch w/o locks: 7m9s
Change-Id: I63588c4cc59fa6fe2f8327ea1e4a9e71b241d4fe
Review URL: http://gerrit.chromium.org/gerrit/498
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
force-mirror: Only fetch files from configured mirrors, ignoring
SRC_URI, except when mirror is in the ebuild(5) RESTRICT variable.
BUG=chromium-os:13221
TEST=Test that ebuilds now fail with FEATURES=force-mirror if files are
not in the configured mirror.
Change-Id: I8484d5af8dff97b431398030b33c024ff1295ba0
Review URL: http://codereview.chromium.org/6677171
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
a workaround for bug #288863 in the stable branch.
|
|
|
|
|
| |
disable warnings about "missing repo_name entries" for the stable
branch.
|
|
|
|
|
| |
enabled by default now, but it may be in the future, depending on feedback
from unknown-features-warn.
|
|
|
|
| |
minimize the diff between branches.
|
| |
|
| |
|
|
|
|
|
| |
merge and unmerge.
* Add FEATURES=prelink-checksums in order to trigger the old behavior.
|
|
|
|
|
|
|
| |
all build logs to be compressed while they are being written. Log file
names have an extension that is appropriate for the compression type.
Currently, only gzip(1) compression is supported, so build logs will
have a '.gz' extension when this feature is enabled.
|
| |
|
|
|
|
| |
(for bug #303567), promote os-headers before it.
|
|
|
|
| |
This caused Portage to wrongly claim some FEATURES not to be valid.
|
|
|
|
|
| |
Triggered by FEATURES="lafilefixing" (enabled by default)
Includes a QA warning for invalid .la files.
|
|
|
|
| |
values can be disabled.
|
| |
|
|
|
|
|
|
|
| |
Support loading per-package bashrc files from the /etc/portage/env
directory. The invidual files should reside in a directory matching
${CATEGORY}, with names matching ${PN}, ${PN}:${SLOT}, ${P} and/or ${PF}
(which are applied in that order).
|
|
|
|
| |
says he wants it to work via make.conf, requiring redesign.
|
| |
|
| |
|
|
|
|
|
| |
variables. Thanks to Alec Warner <antarus@g.o> for submitting the
initial PORTAGE_USERNAME patch from chromiumos-overlay.
|
|
|
|
| |
svn path=/main/trunk/; revision=15362
|
|
|
|
| |
svn path=/main/trunk/; revision=15282
|
|
|
|
| |
svn path=/main/trunk/; revision=15221
|
|
|
|
|
|
| |
target_root variable group
svn path=/main/trunk/; revision=14199
|
|
|
|
|
|
|
|
|
| |
pym/portage/const.py: introduce WORLD_SETS_FILE variable
pym/portage/sets/files.py: use new WORLD_SETS_FILE var
pym/_emerge/unmerge.py: likewise
svn path=/main/trunk/; revision=14197
|
|
|
|
|
|
|
|
|
|
| |
grouped into use with config_root, target_root or nothing, made all
variables that are used with config_root or target_root not start with a
leading slash, such that they can be directly used with os.path.join
(which they should because of xxxx_root).
svn path=/main/trunk/; revision=14192
|
|
|
|
|
|
| |
see-if-I-can-commit test)
svn path=/main/trunk/; revision=13824
|
|
|
|
|
|
|
| |
lstrip(os.sep) is not needed for os.path.join() calls. Thanks to Fabian
Groffen <grobian@g.o> for this patch.
svn path=/main/trunk/; revision=13821
|
|
|
|
|
|
|
|
|
|
|
|
| |
will work as expected. The ACCEPT_LICENSE variable is now treated as a lazily
evaluated incremental, so that * can be used to match all licenses without
every having to explicitly expand it to all licenses. It should now behave
as approved by the council in response to the "RFC: ACCEPT_LICENSE default
value (GLEP 23)" discussion:
http://archives.gentoo.org/gentoo-dev/msg_d5c1e7285399ebc27a74bdd02cb4d037.xml
svn path=/main/trunk/; revision=13670
|
|
|
|
| |
svn path=/main/trunk/; revision=11550
|
|
|
|
|
|
|
| |
and src_configure (and bump EAPI to 2_pre3). Thanks to Ciaran McCreesh for
the suggestion.
svn path=/main/trunk/; revision=11454
|
|
|
|
|
|
| |
EBUILD_PHASES.
svn path=/main/trunk/; revision=11394
|
|
|
|
|
|
|
|
|
|
|
| |
the correct phase function.
* Add --exclude-init-phases option to save_ebuild_env() for filtering
out unnecessary functions and variables when creating environment.bz2.
* Add support for default() and default_* functions that die in all
ebuild phases so that improper use of these functions is quickly
discovered.
svn path=/main/trunk/; revision=11346
|
|
|
|
|
|
|
| |
new EAPI=2_pre2. Thanks to Arfrever Frehtes Taifersar Arahesis for the
initial patch which I've made a few adjustments to.
svn path=/main/trunk/; revision=11327
|
|
|
|
|
|
| |
by install_mask() during that phase.
svn path=/main/trunk/; revision=10198
|
|
|
|
|
|
| |
running 'sys-apps/portage' instance to create a set of protected atoms.
svn path=/main/trunk/; revision=10184
|
|
|
|
|
|
|
|
|
|
| |
functions available for types such as RMD160. Now it's possible
to generate a Manifest when python-2.5 is built without ssl
support. MANIFEST1_REQUIRED_HASH and MANIFEST2_REQUIRED_HASH
are the only two functions that are strictly required now (MD5
and SHA1).
svn path=/main/trunk/; revision=8521
|
|
|
|
| |
svn path=/main/trunk/; revision=8253
|
|
|
|
| |
svn path=/main/trunk/; revision=8016
|
|
|
|
|
|
| |
it works properly with both 0 and 1.
svn path=/main/trunk/; revision=7909
|
|
|
|
|
|
| |
inside fakeroot when a non-root user runs the ebuild command. Thanks to swegener for the initial patch.
svn path=/main/trunk/; revision=7379
|