The build uses several python code generators. The package worked anyway
despite pkg_setup only running for USE=test, because it always found the
latest version, ignoring PYTHON_COMPAT entirely. This occurs because
cmake.eclass no longer forces the python_setup executable to be found.
Bug: https://bugs.gentoo.org/959154
Signed-off-by: Eli Schwartz <eschwartz@gentoo.org>
We appreciate the kind gift, but this is a leaf package with no tests,
and it's still stuck on py3.12 without a way to effectively test for new
impl support...
Signed-off-by: Eli Schwartz <eschwartz@gentoo.org>
To build, this needs an opencl runtime and also an opencl buildtime. :)
opencl provides these in different packages, so it's not sufficient to
depend on the former alone. Although the headers are also a build dep
for the runtime, they might be depcleaned and are never installed at all
when binpkgs are used.
Signed-off-by: Eli Schwartz <eschwartz@gentoo.org>
Since the initial round of the sec-keys.eclass patchset, an updated
version of Daiki Ueno's key appeared and tests failed. Bump the version
to demonstrate the update process.
Signed-off-by: Eli Schwartz <eschwartz@gentoo.org>
Bump EAPI 7 -> 8 as required by eclass. ;)
This is a nice demonstration of using the eclass. It showcases some
advanced usage, in particular, how to augment a single key within a
keyring published on upstream's website with an extended expiration date
from the openpgp.org keyserver.
The ebuild was previously updated in commit
1061fd37f9 with some explanatory text
about why we needed to do that. It wasn't a very good commit (sorry,
past self!) because it was kind of inscrutable whether the key "should"
have been included. That commit would have instead modified "manual" to
"openpgp" for daiki, had the eclass existed exactly a year ago.
Bug: https://gitlab.com/gnutls/web-pages/-/issues/6
Fixes: 1061fd37f9
Signed-off-by: Eli Schwartz <eschwartz@gentoo.org>
The current state of verify-sig support is a bit awkward. We rely on
validating distfiles against a known trusted keyring, but creating the
known trusted keyring is basically all manual verification. We somehow
decide an ascii armored key is good enough without any portage
assistance, then arrange to download it and trust it by Manifest hash.
How do we know when updating a key is actually safe?
This eclass handles the problem in a manner inspired in part by pacman.
We require an eclass variable that lists all permitted PGP fingerprints,
and the eclass is responsible checking that list against the keys we
will install. It comes with a mechanism for computing SRC_URI for a
couple of well known locations, or you can append your own in the
ebuild.
Key rotations, both expected and malicious, are easily detected by
checking the git log for changes to declared fingerprints in a bump. The
former can be rationalized in the commit message. So can the latter, but
in most cases those will be rejected during peer review.
Signed-off-by: Eli Schwartz <eschwartz@gentoo.org>
The 2nd one does nothing given only the one before eautoreconf
matters, seems forgot to remove when updated for wine.eclass
only in live.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>