Github CI with clang failing

11 views
Skip to first unread message

Bram Moolenaar

unread,
Oct 11, 2021, 11:37:37 AM10/11/21
to vim...@googlegroups.com

It appears github actions using clang have recently started failing:

update-alternatives: using /usr/bin/clang-11 to provide /usr/bin/clang (clang) in manual mode
update-alternatives: error: alternative path /usr/bin/llvm-cov-11 doesn't exist

Does someone know how to fix that?
I assume there must have been an update in the environment provided by
github.

--
Don't drink and drive. You might hit a bump and spill your beer.

/// Bram Moolenaar -- Br...@Moolenaar.net -- http://www.Moolenaar.net \\\
/// \\\
\\\ sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ ///
\\\ help me help AIDS victims -- http://ICCF-Holland.org ///

Christian Brabandt

unread,
Oct 12, 2021, 6:08:04 AM10/12/21
to vim...@googlegroups.com

On Mo, 11 Okt 2021, Bram Moolenaar wrote:

>
> It appears github actions using clang have recently started failing:
>
> update-alternatives: using /usr/bin/clang-11 to provide /usr/bin/clang (clang) in manual mode
> update-alternatives: error: alternative path /usr/bin/llvm-cov-11 doesn't exist
>
> Does someone know how to fix that?
> I assume there must have been an update in the environment provided by
> github.

I don't know. Possibly some change in the dependencies of the clang-11
package.
It looks like the CI image does not install llvm-11 anymore, but this
package seems to provide /usr/bin/llvm-cov-11

Let's see if https://github.com/vim/vim/pull/8993 fixes it.

Best,
Christian
--
Letzte Worte des Briefbombenlegers:
"Wie war noch mal die Faxnummer?"

Dusan Popovic

unread,
Oct 12, 2021, 6:22:31 AM10/12/21
to vim...@googlegroups.com
On 12/10/21, Christian Brabandt wrote:
> Let's see if https://github.com/vim/vim/pull/8993 fixes it.

I'll apply last step of changes to my PR again and push at the evening
today so we can see the results. Alternatively current code should
produce exactly two errors, that's what it did before clang problems, if
you can trigger another run of tests with the existing PR code we can
reliably compare results.
signature.asc
Reply all
Reply to author
Forward
0 new messages