| Age | Commit message (Collapse) | Author |
|---|
|
Signed-off-by: Z. Liu <zhixu.liu@gmail.com>
Part-of: https://github.com/gentoo/gentoo/pull/44884
Closes: https://github.com/gentoo/gentoo/pull/44884
Signed-off-by: Sam James <sam@gentoo.org>
|
|
This causes problems with stale crate induced breakage that upstream
rust doesn't appear to be affected by. Notably this downgraded crates to
older versions than what were available at src_prepare time.
Removing this got me around this issue and I didn't encounter workspaces
needing this.
In this instance, we hit an issue w/ icu_properties being pinned when
upstream CI doesn't seem to be hitting that:
```
Locking 510 packages to latest compatible versions
...
Adding icu_properties v2.0.1 (available: v2.1.1)
```
That said, we seem to pass locked-deps in our bootstrap args, as upstream
CI does at https://github.com/rust-lang/rust/blob/503dce33e2e2a5d2fe978b2723ab2a994cc27472/src/ci/run.sh#L66.
[sam: Include an example.]
Bug: https://github.com/unicode-org/icu4x/issues/7190
Signed-off-by: Alfred Wingate <parona@protonmail.com>
Signed-off-by: Sam James <sam@gentoo.org>
|
|
Update done using:
```
git grep -l sys-libs/zlib dev-* | xargs sed -i -e s@sys-libs/zlib@virtual/zlib@g
git diff --name-only | xargs copybump
git diff --name-only | xargs grep -l PYTHON_COMPAT | xargs gpy-impl -@dead
pkgcheck scan --commits -c SourcingCheck,VisibilityCheck --exit error
```
Followed by manual revert in dev-python/zlib-ng where it accidentally
caught sys-libs/zlib-ng dependencies.
Signed-off-by: Michał Górny <mgorny@gentoo.org>
|
|
Signed-off-by: Sam James <sam@gentoo.org>
|
|
Signed-off-by: Sam James <sam@gentoo.org>
|
|
As discussed in bug #963657, I've changed RUST_PATCH_VER to use _p* rather
than -rN which makes it painful to revbump Rust for any other reason. I haven't
made any actual changes to fix the linked bug though.
Bug: https://bugs.gentoo.org/963657
Signed-off-by: Sam James <sam@gentoo.org>
|
|
etc/target-spec-json-schema.json now gets installed by Rust but gets
caught by the sanity check we have for cleaning up bash-completion (which
gets handled by dev-lang/rust-common). It's now expected to have something
left in etc/, so cleanup the rmdir accordingly to just kill the bash-completion
dir.
Note that this JSON file doesn't seem appropriate for rust-common because:
1) it's generated (so would require us to build something to install it);
2) per the commit introducing it:
> Because the schema is subject to change, you should always use the schema
> from the version of rustc which you are passing the target to.
>
> The existence and name of this schema is, just like the properties of the
> JSON specification, not stable and subject to change.
Closes: https://bugs.gentoo.org/964416
Signed-off-by: Sam James <sam@gentoo.org>
|
|
This came up in the context of bug #964416 where someone was confused
as to its purpose, so explain it.
Bug: https://bugs.gentoo.org/689562
Bug: https://bugs.gentoo.org/689160
Bug: https://bugs.gentoo.org/964416
Signed-off-by: Sam James <sam@gentoo.org>
|
|
Add a new `USE_EXPAND` variable called `RUST_SYSROOTS` which allows to
specify additional Rust sysroots to build. Rust sysroot is a directory
which contains the prebuilt standard libraries (in form of `*.rlib`
files) and utility binaries (e.g. `rust-objcopy`).
This variable is limited to bare metal targets. Sysroot for targets not
falling into that category contain a local copy of compiler runtime,
libc and unwinder, so crossdev is a better fit for them.
In this change, the only `RUST_SYSROOT` option is `wasm`, which enables
build of a sysroot for `wasm32-unknown-unknown` target and replaces the
former `wasm` USE flag.
Signed-off-by: Michal Rostecki <vadorovsky@protonmail.com>
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
Add reminder to ebuild to update the git repo and make a new tag as
required instead of using filesdir: that's how we end up exceeding the
limit; things are very rarely cleaned up here!
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
- Enable building with 1.89.0
- bootstrap.toml - `rust.description` -> `build.description`
See-also: https://github.com/rust-lang/rust/commit/e7f5bbe7438019ac6af84ce4f8cc04ab2df8c744
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
We already depend on the right LLVM targets but without =, but we need
that as it's automagic otherwise. I'm not completely convinced we need
it here, but my aim is to avoid breaking things and then improve more
later.
Bug: https://bugs.gentoo.org/735154
Bug: https://bugs.gentoo.org/767700
Bug: https://bugs.gentoo.org/768267
Signed-off-by: Sam James <sam@gentoo.org>
|
|
My previous fix didn't actually unpack tarballs at all.
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
When the patch tarballs were added the ebuilds were not updated to
ensure that only upstream-signed files are verified.
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
This is our first beta for 1.88.0, corresponding to beta.4 upstream.
It is also the first release where `beta` and `live` ebuilds get their
patches from gitweb rather than ${FILESDIR}.
Changes:
- in line with upstream changes, `config.toml` has been renamed to
`bootstrap.toml`.
- `*bootstrap.toml` has `change-id = "ignore"` added to suppress
warnings that are not useful to downstream.
- after discussion with upstream `x.py` verbosity reduced to `-v`.
- restrictions on building beta with only the last stable have been
dropped, these can be restored if required.
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
Rust upstream have identified issues with this patch relating to broken
x87 handling in LLVM.
Drop the patch and revbump all impacted revisions to ensure that this is
cleaned up.
Closes: https://www.github.com/gentoo/gentoo/pulls/40616
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
It's a good idea anyway but it also avoids the C23 issue. We should look
at unbundling the other bits but I'd like to avoid requiring libssh2
if we can help it unconditionally so punting on that for now w/
a comment.
Closes: https://bugs.gentoo.org/956044
Signed-off-by: Sam James <sam@gentoo.org>
|
|
Closes: https://bugs.gentoo.org/952925
Signed-off-by: blackteahamburger <blackteahamburger@outlook.com>
Closes: https://github.com/gentoo/gentoo/pull/41387
Signed-off-by: Sam James <sam@gentoo.org>
|
|
Signed-off-by: Z. Liu <zhixu.liu@gmail.com>
Closes: https://github.com/gentoo/gentoo/pull/41523
Signed-off-by: Sam James <sam@gentoo.org>
|
|
use-lld appeared first in commit bfba6ef328bbba327cae8918e795c11b89217672
refactored in commit befd1eb4ec78e19f0c8628fa75241ea3ed6ec792
Closes: https://bugs.gentoo.org/948027
Signed-off-by: Z. Liu <zhixu.liu@gmail.com>
Closes: https://github.com/gentoo/gentoo/pull/40772
Signed-off-by: Sam James <sam@gentoo.org>
|
|
... otherwise CC at the time of building dev-lang/rust is baked in,
which is problematic if e.g. CC=clang-19 is used and then it gets
depcleaned later (as reported on the gentoo-user ML).
Signed-off-by: Sam James <sam@gentoo.org>
|
|
By default VCS checkouts don't include commit info in the
version string to save on rebuilds.
We don't have to worry about incremental builds in Gentoo,
and having a date and commit for the live ebuild will greatly
assist in any troubleshooting we have to do (or bug reports
for upstream).
Signed-off-by: Matt Jolly <kangie@gentoo.org>
|
|
We already have a similar patch for versions before 1.85.0.
Change rust-lang/rust@9c37c14aa245, which got released in 1.85.0, made
us think that such a patch is not needed anymore, but that's not true.
Upstream didn't fix the existing individual targets yet.
Bug: https://bugs.gentoo.org/950275
Signed-off-by: Michal Rostecki <vadorovsky@protonmail.com>
Closes: https://github.com/gentoo/gentoo/pull/40752
Signed-off-by: Sam James <sam@gentoo.org>
|
|
1. target in ALL_LLVM_TARGETS has prefix llvm_targets_, so the check in
last commit failed always.
2. using associate array for the test of experimental target's existence
3. sed -i 's:LLVM_EXPERIMENTAL_TARGETS:RUST_EXPERIMENTAL_TARGETS:g' to
fix warning on var name collide with ALL_LLVM_EXPERIMENTAL_TARGETS in
llvm.org.eclass
Fixes: ff8e2b548c258e60463b1df0224beda29c7aacec
Signed-off-by: Z. Liu <zhixu.liu@gmail.com>
Closes: https://github.com/gentoo/gentoo/pull/40698
Signed-off-by: Sam James <sam@gentoo.org>
|
|
While currently unkeyworded, the intent is that eventually
`beta` will only ever be keyworded `testing` (i.e. not stabilised)
and will enable us to catch regressions on esoteric systems and
report them upstream before a stable release is cut (and we
have to do the same thing anyway, but with revbumps).
All `beta` channel releases will share a single `SLOT`, i.e.
1.85.0_beta20250208 has `SLOT` 1.85.0, the same as the eventual
`stable` release. Installing different `beta` releases side-by-side
is not supported via Portage (and it's unlikely that they will stay
in-tree for long once superseded). This enables straightforward
`ver_test` comparisons to reflect a particular 'version'.
`nightly` is a VCS sources ebuild template which will soon be the
only way to get `nightly` Rust via Portage (from source, anyway,
a followup commit will add a 'dev-lang/rust-bin' nightly binary
release, using upstream infrastructure). Nightly can only be built
by the `beta` channel release.
Signed-off-by: Matt Jolly <kangie@gentoo.org>
unkeyword betas
Signed-off-by: Arthur Zamarin <arthurzam@gentoo.org>
|
