| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
Closes: https://bugs.gentoo.org/650170
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
|
|
|
|
|
|
| |
With the latest changes (slot operators) we need at least EAPI 5 when
inheriting toolchain.eclass. Thus, in case of EAPI=4, die early with a
more descriptive error message.
|
|
|
|
|
|
|
|
|
|
|
|
| |
We should annotate a number of critical library dependencies of the
compiler with := operators to ensure that the compiler gets rebuilt in
time.
TODO: We should also check for FEATURES=preserve-libs / print a warning
before attempting to upgrade a library with soname change. Otherwise,
users might be off to a rocky ride.
Closes: https://bugs.gentoo.org/642316
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
CROSSCOMPILE_OPTS is a USE_EXPAND of a single item: headers-only.
Convert it to a global USE flag instead.
Mechanical ebuild rename done as:
$ sed -e 's@crosscompile_opts_headers-only@headers-only@g' \
-i $(git grep -l headers-only)
'headers-only' flag is used by crossdev to bootstrap stage1 compiler
before libc is available.
crossdev switched to USE=headers-only in =sys-devel/crossdev-20171230.
Bug: https://bugs.gentoo.org/642712
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
| |
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
| |
This eclass is inherited by ebuilds in sys-devel/{gcc,kgcc64,gcc-apple},
each which make use of different IUSE flags. This causes problems with
`use X` constructions when X is not in the IUSE flags. At EAPI=4 this
simply emitted a warning, while at EAPI=5 this is an error. We update
the eclass to make use of use_if_iuse and similar constructions where
necessary to bring the eclass into compliance with EAPI=5.
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
|
|
|
|
| |
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
|
|
|
| |
gcc-6 and above no longer needs uclibc specific patches, so we don't die if
the patchset is not available. We do, however, still apply it if UCLIBC_VER
is defined in the ebuild to future proof the code in case we need to reintroduce
the patchset in the future.
|
|
|
|
|
|
| |
>=gcc6, bug 618908
Closes: https://bugs.gentoo.org/618908
|
|
|
|
| |
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.10, Repoman-2.3.3
Closes: https://github.com/gentoo/gentoo/pull/5741
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
GCC_FILESDIR support was broken by portage-2.3.6 commit:
https://github.com/gentoo/portage/commit/0637c95545ab7dc0bb5d091de221b9e99e2cf5b2
("portage.package.ebuild: Use a fake FILESDIR to catch invalid accesses")
Unfortunately it was applied to older EAPIs without fixing all
the users first.
The change yanks GCC_FILESDIR override support and hardcodes
FILESDIR usage (as all the ebuilds except 'gcc' are broken anyway).
Bug: https://bugs.gentoo.org/629776
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On CTARGET="hppa2.0-unknown-linux-gnu" the following code
local VAR="CFLAGS_"${CTARGET//-/_}
CXXFLAGS=${!VAR}
produces incorrect substitution:
CFLAGS_hppa2.0_unknown_linux_gnu
which caused 'crossdev hppa2.0-unknown-linux-gnu' to fail.
The change adds '.' mangling as well.
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
| |
|
| |
|
|
|
|
|
|
|
| |
sys-devel/gcc-7.1.0 requires external dev-libs/boehm-gc, the internal
copy got removed [1].
[1] https://gcc.gnu.org/viewcvs/gcc?view=revision&revision=242985
|
|
|
|
| |
Package-Manager: Portage-2.3.5, Repoman-2.3.2
|
| |
|
|
|
|
| |
Signed-off-by: Robin H. Johnson <robbat2@gentoo.org>
|
| |
|
| |
|
|
|
|
| |
Bug: 531610
|
|
|
|
|
|
|
|
|
| |
Directory prefixify part 4.
LIBPATH, etc. now have EPREFIX prepended. The latter need to be
quoted.
Bug: 531610
|
|
|
|
|
|
|
|
|
| |
Directory prefixify part 3.
Raw directories are prepended by EPREFIX. Directories passed to
ebuild helpers are EPREFIX stripped.
Bug: 531610
|
|
|
|
|
|
| |
Directory prefixify part 2.
Bug: 531610
|
|
|
|
|
|
|
|
|
| |
Directory prefixify part 1.
In addition, E/D and E/ROOT has trailing slashes. No need to write
an additional slash.
Bug: 531610
|
|
|
|
|
|
|
|
|
|
| |
In Prefix, PATH should also be preserved, resulting in a mouthful of
`env -i PATH=${PATH} ROOT=${ROOT}`. The origin commit introducing
env -i was for "cleanup". Dropping env -i is cleaner.
Reference: https://gitweb.gentoo.org/repo/gentoo/historical.git/commit/?id=a8a64d1886cc5c9f975353b97237f2900a3ee88c
Reference: https://gitweb.gentoo.org/repo/gentoo/historical.git/commit/?id=7d9f89700a2ee50674c0d871ec6d21b468ac6206
Bug: 531610
|
|
|
|
|
| |
/usr/sbin is in PATH, avoid writing
${EPREFIX}/usr/sbin/fix_libtool_files.sh.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The USE=multislot as defined partially by toolchain.eclass and partially
by sys-devel/gcc was used for two purposes:
- enabling build-time (only) blockers on old gcc versions -- which do
not make any sense because they are build-time only and there is no
technical reason for two gcc version ranges not to be installed at
the same time, both at build time and at run time,
- changing behavior of post-install wrt conditional gcc-config calls.
Both cases are invalid use of USE flags, considering that the flag does
not affect the installed files in any way.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
unmerging active version #529608
If the user is unmerging the current version, try and automatically
switch to a better version. We blindly pick the latest one available
for the target in question since there's not really a better metric
to figure out which one the user actually would want.
|
| |
|
| |
|
|
|
|
|
|
| |
I'm not entirely happy with this logic, but as other related bugs are
addressed, this will probably evolve with them. This at least unblocks
people trying to use gcc-5 cross-compilers.
|
| |
|
| |
|
|
|
|
|
|
|
| |
We have dev-lang/go provide `go` and `gofmt`, and having gcc include
these in its BINPATH causes gcc-config to clobber the go package with
symlinks. Drop the gcc versions for now (they're copies from the go
package anyways) until we have a better solution.
|
|
|
|
|
|
|
|
| |
When gcc installs programs into BINPATH, it sometimes links them so that
the $CHOST-xxx and xxx progs are the same. The logic to standardize the
naming in BINPATH then triggers a warning as they're the same file.
Manually delete the target before running the mv to avoid that business.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
Signed-off-by: Justin Lecher <jlec@gentoo.org>
|
|
|
|
| |
stack-protector-strong behavior
|