Looks like 1.59.0 enters a busy loop or infinite recursion when compiling certain crates (in particular, clap_derive and structopt-derive) on s390x: https://bugzilla.redhat.com/show_bug.cgi?id=2058803
The patched version seems to work fine. Thanks!
The patched version seems to work fine. Thanks!
The changes to add package notes break some packages: https://pagure.io/fedora-rust/rust2rpm/issue/181
The changes to add package notes break some packages: https://pagure.io/fedora-rust/rust2rpm/issue/181
Is this a release or a post-release snapshot? The commit message / changelog entry suggest the former, while the version string suggests the latter.
Fixes "rpmdev-bumpspec --new", should also fix the-new-hotness patch / scratch builld creation.
I don't know what caused this issue, but I got hundreds of lines of scriptlet errors when upgrading to kernel-devel from this update:
Installing : kernel-devel-5.16.8-200.fc35.x86_64 13/27
Running scriptlet: kernel-devel-5.16.8-200.fc35.x86_64 13/27
hardlink: cannot link ./Documentation/kbuild/Kconfig.recursion-issue-01 to /usr/src/kernels/5.16.5-200.fc35.x86_64/./Documentation/kbuild/Kconfig.recursion-issue-01.hardlink-temporary: File exists
hardlink: cannot link ./Documentation/kbuild/Kconfig.select-break to /usr/src/kernels/5.16.5-200.fc35.x86_64/./Documentation/kbuild/Kconfig.select-break.hardlink-temporary: File exists
hardlink: cannot link ./Documentation/userspace-api/media/Makefile to /usr/src/kernels/5.16.5-200.fc35.x86_64/./Documentation/userspace-api/media/Makefile.hardlink-temporary: File exists
(and thousands more lines like these)
Looks like they reference the wrong kernel version (5.16.5, should be 5.16.8?)
Full log from the failed scriptlet here: https://gist.github.com/decathorpe/55a821a98b6cc39a1c1fca7cede78325
Prepared and built tons of updates with this today, worked like a charm. The new argcomplete-based bash completions are also very nice :+1:
Works fine. Default and performance regression tests also pass on all my systems:
Though it seems tmpfs is noticably slower with any 5.16 kernel ... or maybe I'm imagining things.
Works fine. Default and performance regression tests also pass on all my systems:
This is suffering the same broken dependency problem as the rawhide update: https://bugzilla.redhat.com/show_bug.cgi?id=2047225
You will probably need to disable CUDA support in these packages.
This is suffering the same broken dependency problem as the rawhide update: https://bugzilla.redhat.com/show_bug.cgi?id=2047225
You will probably need to disable CUDA support in these packages.
python3-azure-cli-core from this update is not installable due to this unsatisfied dependency: "(python3.10dist(humanfriendly) >= 10 with python3.10dist(humanfriendly) < 11)"
It doesn't really matter. Only "security" updates are handled slightly differently.
This update has been in testing for three weeks (and in f34-stable for two weeks), and nobody else seems to have hit this problem. I'll push it to stable (since I'll build 1.18.6 tomorrow anyway).
Looks like 1.59.0 enters a busy loop or infinite recursion when compiling certain crates (in particular, clap_derive and structopt-derive) on s390x: https://bugzilla.redhat.com/show_bug.cgi?id=2058803