Create Atomic<T> type alias
#130543
Create Atomic<T> type alias
#130543
Conversation
|
r? @Noratrieb rustbot has assigned @Noratrieb. Use |
rustbot
added
O-hermit
This comment was marked as outdated.
This comment was marked as outdated.
CAD97
force-pushed
the
generic-atomic
branch
from
bacfd33 to
cc104b1
Compare
| macro impl_atomic_primitive( | ||
| $Atom:ident $(<$T:ident>)? ($Primitive:ty), | ||
| size($size:literal), | ||
| align($align:literal) $(,)? |
There was a problem hiding this comment.
FYI the align is given to impl_atomic_primitive! because while it isn't actually used in this form, it's needed for the form where Atomic<T> is the struct definition. So it'll make the next diff a little cleaner (and here is kinda just residual from paring this back to just do an alias initially).
|
Alright that's the merge conflicts fixed. If I end up needing to continue playing catch-up on some of the touched files, I'll probably just drop files' touch-up to use |
This comment has been minimized.
This comment has been minimized.
|
r=me after fixing the import issues |
CAD97
force-pushed
the
generic-atomic
branch
2 times, most recently
from
67a79e8 to
fbf092e
Compare
|
Also found a similar conflict with the windows futex impl plumbing via local check, so that's also resolved by the amend (hopefully; my local checkout suddenly is complaining about duplicate rmeta for crate IIRC I don't have perms to do this but might as well try: @bors r=Noratrieb rollup=iffy |
|
@CAD97: 🔑 Insufficient privileges: Not in reviewers |
|
@CAD97: 🔑 Insufficient privileges: not in try users |
|
@bors r=Noratrieb rollup=iffy |
bors
added
S-waiting-on-bors
Create `Atomic<T>` type alias and use it in core/alloc/std where possible, ignoring test files for now. This is step one, creating the alias from `Atomic<T>` to `AtomicT`. The next step of flipping this and aliasing `AtomicT` to `Atomic<T>` will have a bigger impact, since `AtomicT` imports can be dropped once `Atomic::new` is a usable name. First commit is the true change. Second commit is mostly mechanical replacement of `AtomicT` type names with `Atomic<T>`. See [how this was done for `NonZero`](rust-lang#120257) for the rough blueprint I'm following. - ACP: rust-lang/libs-team#443 (comment) - Tracking issue: rust-lang#130539
This comment has been minimized.
This comment has been minimized.
|
whups |
bors
added
S-waiting-on-author
|
I
so it might take me a while to diagnose and fix whatever's going on. I did say rollup was iffy 😅 |
|
@CAD97 I usually give PRs a shot in a rollup even if they're iffy if they haven't ever been tried. |
in core/alloc/std only for now, and ignoring test files
CAD97
force-pushed
the
generic-atomic
branch
from
fbf092e to
00d1ee5
Compare
This comment has been minimized.
This comment has been minimized.
clippy's ref to unfreeze const lint can't see through type projections, so just avoid using a type projection inside of the type of ONCE_INIT.
|
Oh, I think I might see what the issue could be. The atomic within |
This comment has been minimized.
This comment has been minimized.
|
The job Click to see the possible cause of the failure (guessed by this bot) |
|
☔ The latest upstream changes (presumably #130724) made this pull request unmergeable. Please resolve the merge conflicts. |
alex-semenyuk
added
S-waiting-on-author
and use it in core/alloc/std where possible, ignoring test files for now.
This is step one, creating the alias from
Atomic<T>toAtomicT. The next step of flipping this and aliasingAtomicTtoAtomic<T>will have a bigger impact, sinceAtomicTimports can be dropped onceAtomic::newis a usable name.First commit is the true change. Second commit is mostly mechanical replacement of
AtomicTtype names withAtomic<T>.See how this was done for
NonZerofor the rough blueprint I'm following.Atomic<T>libs-team#443 (comment)Atomic#130539