| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Fixes installation of cross headers for ebuilds using EAPI 5+.
|
|
|
|
|
|
| |
set_java_env is now deprecated. The new function is better because any
variable in the environment file can be resolved and it even allows
subshells for more dynamic content.
|
|
|
|
|
| |
This is useful for multilib JVM ebuilds. revdep-rebuild currently
supports multiple SEARCH_DIRS_MASK entries.
|
|
|
|
| |
The old version just sucked. Closes bug #159439.
|
|
|
|
| |
Signed-off-by: Tomáš Chvátal <scarabeus@gentoo.org>
|
|
|
|
| |
skipping rebuild of some docs.
|
| |
|
| |
|
|
|
|
| |
webkit flag is on by default to build as upstream intended.
|
| |
|
|
|
|
| |
Bug: 534194
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
End-users want to experiment with clang and -O4, so let them.
Bug: http://bugs.gentoo.org/462488
|
|
|
|
| |
Fixes bug 586538.
|
|
|
|
| |
Bug: 583740
|
| |
|
| |
|
|
|
|
| |
Signed-off-by: Johannes Huber <johu@gentoo.org>
|
|
|
|
| |
Signed-off-by: Johannes Huber <johu@gentoo.org>
|
|
|
|
| |
Signed-off-by: Johannes Huber <johu@gentoo.org>
|
|
|
|
|
| |
Improve the names of some variables to match what upstream docs call
things.
|
|
|
|
| |
Signed-off-by: Johannes Huber <johu@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
The php-ext-pecl eclasses are based mainly on the php-ext-source
eclasses. Now that we have a new revision php-ext-source-r3.eclass,
this new revision of php-ext-pecl inherits that. As a result, all of
the changes affecting that revision also affect this one. A migration
guide for users can be found on the wiki:
https://wiki.gentoo.org/wiki/Project:PHP/Php-ext-source-r3_migration_guide
Gentoo-Bug: 512184
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is a new revision of the php-ext-source eclass that supports
EAPI=6 (only) and cleans up some of the existing code. The list of
user-facing changes is,
* Support only EAPI=6.
* PATCHES array/variable support.
* DOCS array support (bug 512184).
* Renamed my_conf and PHPSAPILIST variables.
Some refactoring was done, but not in a way that consumers should
notice. A migration guide can be found on the wiki:
https://wiki.gentoo.org/wiki/Project:PHP/Php-ext-source-r3_migration_guide
In addition, a number of improvements were suggested by Michał Górny
on the gentoo-dev mailing list:
https://archives.gentoo.org/gentoo-dev/message/9159cba2eabceb96d8ae16ced6631f1e
Gentoo-Bug: 512184
|
| |
|
| |
|
|
|
|
|
|
| |
Also drop mozconfig-v6.41.eclass as this ebuild was the last consumer
Package-Manager: portage-2.2.28
|
|
|
|
|
|
|
|
|
|
| |
This ebuild is and has for quite some time been the last of the
consumers of mozconfig-3.eclass. Although the package is masked
there may still be a need for it to allow migration of user
data. This commit moves all relevant portions of the eclass
into the ebuild and expunges the eclass from the repo.
Package-Manager: portage-2.2.28
|
|
|
|
|
| |
Some of the L10N-migrated code did not work properly when generating
langpacks as part of the build process. This commit fixes those issues.
|
| |
|
| |
|
|
|
|
| |
Gentoo-Bug: 587582
|
|
|
|
| |
Bug: https://bugs.gentoo.org/588198
|
|
|
|
| |
Gentoo-bug: 587998
|
|
|
|
|
|
|
|
| |
This used to "work" because gcc-version relied on gcc-config, even if GCC was
not the active compiler. gcc-version has since been fixed to always use CC
instead.
Gentoo-bug: 587998
|
|
|
|
|
|
|
|
|
| |
HOST_CC and HOST_CXX are used to build tools like nsinstall which are needed
to install later on. Until now, these values were undefined and for whatever
reason were never autodetected to match the CHOST based tools. This commit
should resolve that issue.
Bug: http://bugs.gentoo.org/565124
|
|
|
|
|
|
| |
Bug: http://bugs.gentoo.org/587334
Package-Manager: portage-2.2.28
|
|
|
|
|
|
|
|
|
|
| |
mozconfig
mozilla-48 and newer have dropped some configuration options that are no longer used
in the build system, and due to the way the new python configure operates, any options
specified that are not recognized are fatal. Moving these options out of mozcoreconf
allows the older mozconfig's to retain the same configuration while also allowing newer
mozconfig's to drop them as appropriate.
|
| |
|
|
|
|
|
|
| |
1) 15.08.x has left the tree
2) Prepare for a future kde4-l10n without USE=minimal
Version block makes sure of no ancient kde4-l10n anyway
|
|
|
|
| |
is defined
|
|
|
|
|
| |
KDE_GCC_MINIMAL will now only be exported in packages that require it, as it
triggers an expensive pkg_pretend check whenever it is defined.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Exporting pkg_pretend is expensive, even if the exported function is a noop.
As kde5_pkg_pretend is currently only used to check if the active GCC version
is recent enough, only export that phase if it will have something to do.
The number of situations in which KDE_GCC_MINIMAL is exported will be
substantially reduced in a future commit. This will result in a large time
saving at the beginning of the emerge process as we will no longer see hundreds
of lines of:
>>> Running pre-merge checks for kde-foo/bar-123
>>> Running pre-merge checks for kde-foo/baz-123
|
|
|
|
|
|
|
|
| |
This avoids bugs where too many arguments are passed by mistake, causing the
last one to be ignored.
For example, the subslot operator added in d1bc5ac3f0e42df0f0255c8b39d7df16c5d4176b
was never applied because too many '' were passed.
|
| |
|
|
|
|
|
| |
Newer versions offer a configure flag so we don't have to manually patch
in the default setting.
|
|
|
|
| |
Bug: https://bugs.gentoo.org/577704
|
| |
|