xbps-src always ignore self deps in depends and in hostmakedepends
when not cross compiling. The old behaviour did only check for cross.
This now also ignores them in makedepends, but they are either way
disallowed in there, and returning an error would be correct, but
./xbps-src show-* is and never was suitable for linting, so ignore for now.
* $XBPS_CROSS_TRIPLET-pkg-config is our wrapper not a real executable
* In a build that have both freetype-config and pkg-config,
$XBPS_CROSS_BASE will be prepended twice with current system,
let's fix it.
this introduces a new build-style void-cross, which can be used
to write system crosstoolchain templates; this is to reduce the
amount of maintenance, resolve existing problems with the cross
toolchain templates and remove repeated code
this makes sure we don't have to worry about packages installing
stuff in lib32/lib64, it will be automagically symlinked
if something is still left over for whatever reason, or if the
opposite wordsize directory exists, that will be caught by pkglint
Also remove a few more private libraries that can be built in.
Co-authored-by: andry-dev <andry-dev@users.noreply.github.com>
Co-authored-by: Andrew J. Hesford <ajh@sideband.org>
Closes: #27300 [via git-merge-pr]
for now, these profiles set -mcpu to power8 and -mtune to power9
to match our ppc64le profiles, since there isn't really any native
harwdare to run these on (at least yet), and any packages will be
experimental for now anyway
As discussing in [1], on template with "subpackages" as multilines will
report false positive on some packages will never be built.
There're multiple problems here:
- expanded "subpackages" will have an empty line if it has a newline
inside template
- "sed" expression couldn't work with multilines "subpackages"
Let's not quote "$subpkgs" and "$subpackages" in "printf" to let the
shell do expansion and trim the empty lines for us. And rewrite the
"sed" expression to work with multilines "subpackages"
[1]: https://github.com/void-linux/void-packages/pull/26939#issuecomment-739098547
The file was created in chroot_prepare(), but the configuration
directory was being cleaned in chroot_sync_repodata(). This commit moves
the configuration to after the directories are created.
Fixes#25534
Outside of chroot, we calcucate XBPS_ALT_REPOSITORY only to throw it
away and calculate it again inside chroot. Normally, this is not problem
except when we're working with a Git's worktree, in which case, we don't
have the access to original gitdir inside chroot.
Let's bring that value from outside of chroot into it.
we don't need to specify -mcpu where compiler default is used,
preserve mtune though; do not -mno-altivec on ppc since altivec
is not default in the first place and doing this will just
break random things that explicitly enable altivec for runtime
checked code; do not use -maltivec on either ppc64le or ppc64
since it's implied by their -mcpu's (power8 for ppc64le which
is implicit, and 970 for ppc64 which is explicit)
also remove -mlong-double-64 for musl targets as that's compiler
default and it's not necessary to pass it; also remove -mabi=elfv2
for both ppc64le and ppc64 since it's implied default
Current wrappers hard-coded them at 0.42 which is very long ago.
Hence, those wrappers insert a versioned vapidir that no longer exists.
The build still work because valac will look into host's versioned vapidir
instead.
Let's check the version we're having and wrap that instead of changing
the hard-coded number from time to time.
- Remove .travis.yml, common/travis/docker.sh.
- Change references to travis to be more generic.
- Manual.md: Orient against using CI logs in broken= or nocross=.
