[FFe] update llvm-defaults toolchain to 11
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
llvm-defaults (Ubuntu) |
Expired
|
Undecided
|
Unassigned |
Bug Description
llvm-toolchain-11 was delayed a bit this cycle I didn't land earlier.
llvm-toolchain is a fast moving package, and was anticipated to be available and the default.
It is desired for defaults to point to the latest stable llvm-toolchain, even if very few packages use it in the archive.
$ reverse-depends -c main -b src:llvm-defaults
Reverse-
* boost1.71 (for clang)
* libreoffice (for clang)
* squid (for clang)
Reverse-
* ceph (for clang)
* libreoffice (for clang)
* mozjs78 (for llvm)
* mozjs78 (for clang)
* squid (for clang)
* v4l-utils (for clang)
If this doesn't land well, we can revert it.
tags: | added: block-proposed |
description: | updated |
summary: |
- [FFe] update llvm toolchain to 11 + [FFe] update llvm-defaults toolchain to 11 |
tags: | removed: block-proposed |
Reverse-Depends clang-format- linter
* arcanist-
* cargo (for clang)
What's the impact on rust ecosystem buildability from updating clang?
> If this doesn't land well, we can revert it.
We also have the option of updating any failing packages to hard-code an earlier version of llvm, instead.
So I'm broadly +1 but would like to have more surety regarding the impact on archive buildability, given the short window before release.