Auto merge of #61922 - tmandry:moar-generator-optimization, r=matthewjasper
Don't store locals that have been moved from in generators
This avoids reserving storage in generators for locals that are moved
out of (and not re-initialized) prior to yield points. Fixes #59123.
This adds a new dataflow analysis, `RequiresStorage`, to determine whether the storage of a local can be destroyed without being observed by the program. The rules are:
1. StorageLive(x) => mark x live
2. StorageDead(x) => mark x dead
3. If a local is moved from, _and has never had its address taken_, mark it dead
4. If (any part of) a local is initialized, mark it live'
This is used to determine whether to save a local in the generator object at all, as well as which locals can be overlapped in the generator layout.
Here's the size in bytes of all testcases included in the change, before and after the change:
async fn test |Size before |Size after
-----------------|------------|----------
single | 1028 | 1028
single_with_noop | 2056 | 1032
joined | 5132 | 3084
joined_with_noop | 8208 | 3084
generator test |Size before |Size after
----------------------------|------------|----------
move_before_yield | 1028 | 1028
move_before_yield_with_noop | 2056 | 1032
overlap_move_points | 3080 | 2056
## Future work
Note that there is a possible extension to this optimization, which modifies rule 3 to read: "If a local is moved from, _**and either has never had its address taken, or is Freeze and has never been mutably borrowed**_, mark it dead." This was discussed at length in #59123 and then #61849. Because this would cause some behavior to be UB which was not UB before, it's a step that needs to be taken carefully.
A more immediate priority for me is inlining `std::mem::size_of_val(&x)` so it becomes apparent that the address of `x` is not taken. This way, using `size_of_val` to look at the size of your inner futures does not affect the size of your outer future.
cc @cramertj @eddyb @Matthias247 @nikomatsakis @RalfJung @Zoxc
Auto merge of #62279 - pietroalbini:fix-azure-crlf, r=alexcrichton
ci: explicitly disable CRLF conversion on Windows
The Azure image enables CRLF conversion on Windows builders, but that caused regressions both in our test suite (the miri test suite broke) and in the ecosystem, since we started shipping install scripts with CRLF endings instead of the old LF. The [Godbolt Compiler Explorer](https://godbolt.org/) is one such case of breakage.
This adds a step to the build explicitly disabling the conversion before the repository is checked out.
Pietro Albini [Mon, 1 Jul 2019 19:57:08 +0000 (21:57 +0200)]
ci: explicitly disable CRLF conversion on Windows
The Azure image enables CRLF conversion on Windows builders, but that
caused regressions both in our test suite (the miri test suite broke)
and in the ecosystem, since we started shipping install scripts with
CRLF endings instead of the old LF. The Godbolt Compiler Explorer is one
such case of breakage.
This adds a step to the build explicitly disabling the conversion before
the repository is checked out.
Igor Matuszewski [Sun, 30 Jun 2019 23:36:43 +0000 (01:36 +0200)]
Update RLS to disable spurious client_find_definitions test
Since switching CI to Azure Pipelines it seems that this test seems
to fail more consistently, so let's disable that for now. It helps
that we have less than a week before release - we disallow PRs that
break the tools to land in this period, so this makes landing critical
PRs smoother now.
Auto merge of #62253 - Centril:rollup-115uuuq, r=Centril
Rollup of 8 pull requests
Successful merges:
- #62062 (Use a more efficient iteration order for forward dataflow)
- #62063 (Use a more efficient iteration order for backward dataflow)
- #62224 (rustdoc: remove unused derives and variants)
- #62228 (Extend the #[must_use] lint to boxed types)
- #62235 (Extend the `#[must_use]` lint to arrays)
- #62239 (Fix a typo)
- #62241 (Always parse 'async unsafe fn' + properly ban in 2015)
- #62248 (before_exec actually will only get deprecated with 1.37)
Auto merge of #61682 - Centril:stabilize-type_alias_enum_variants, r=petrochenkov
Stabilize `type_alias_enum_variants` in Rust 1.37.0
Stabilize `#![feature(type_alias_enum_variants)]` which allows type-relative resolution with highest priority to `enum` variants in both expression and pattern contexts. For example, you may now write:
```rust
enum Option<T> {
None,
Some(T),
}
type OptAlias<T> = Option<T>;
fn work_on_alias(x: Option<u8>) -> u8 {
match x {
OptAlias::Some(y) => y + 1,
OptAlias::None => 0,
}
}
```
Rollup merge of #62241 - Centril:fix-async-unsafe-order, r=petrochenkov
Always parse 'async unsafe fn' + properly ban in 2015
Parse `async unsafe fn` not `unsafe async fn` in implementations. We also take the opportunity to properly ban `async fn` in Rust 2015 when they are inside implementations.
Rollup merge of #62239 - lcolaholicl:lcolaholicl-patch-1, r=kennytm
Fix a typo
The definition of 京 seems to be capital, but not capitol.
[reference](https://en.wiktionary.org/wiki/%E4%BA%AC#Etymology_1)
[another reference](https://jisho.org/word/%E4%BA%AC)
Rollup merge of #62224 - euclio:remove-derives, r=GuillaumeGomez
rustdoc: remove unused derives and variants
Though many structs in rustdoc derive `RustcEncodable` and `RustcDecodable`, the impls do not appear to be used by the crate or its dependents. Removing them revealed some enum variants that are never constructed, too.
Rollup merge of #62063 - ecstatic-morse:dataflow-backward-order, r=nagisa
Use a more efficient iteration order for backward dataflow
This applies the same basic principle as #62062 to the reverse dataflow analysis used to compute liveness information. It is functionally equivalent, except that post-order is used instead of reverse post-order.
In the long-term, `BitDenotation` should probably be extended to support both forward and backward dataflow, but there's some more work needed to get to that point.
Rollup merge of #62062 - ecstatic-morse:dataflow-order, r=nagisa
Use a more efficient iteration order for forward dataflow
Currently, dataflow begins by visiting each block in order of ID (`BasicBlock(0)`, `BasicBlock(1)`, etc.). This PR changes that initial iteration to reverse post-order (see [this blog post](https://eli.thegreenplace.net/2015/directed-graph-traversal-orderings-and-applications-to-data-flow-analysis/#data-flow-analysis) for more info). This ensures that the effects of all predecessors will be applied before a basic block is visited if the CFG has no back-edges, and should result in less total iterations even when back-edges exist. This should not change the results of dataflow analysis.
The current ordering for basic blocks may be pretty close to RPO already--`BasicBlock(0)` is already the start block--in which case the cost of doing the traversal up front will outweigh the efficiency gains.
A perf run is needed to check this.
bors [Sun, 30 Jun 2019 12:15:14 +0000 (12:15 +0000)]
Auto merge of #59722 - Zoxc:query-cache, r=eddyb
Clean up query cache code
There are a couple of queries for which we do not promote saved results, which have been removed from caching here. This was likely the cause of the regression in https://github.com/rust-lang/rust/pull/57293#issuecomment-476421298.
bors [Sun, 30 Jun 2019 08:42:58 +0000 (08:42 +0000)]
Auto merge of #61252 - indygreg:upgrade-musl, r=alexcrichton
Update musl to 1.1.22
This is the latest available version. I noticed Rust wasn't using the
latest version when I attempted to have Cargo link object files produced
outside of Cargo / Rust's toolchain and was getting missing symbol
errors. Those missing symbols were added in 1.1.21 and 1.1.22.
I'm not fully sure of the implications of this change or how
comprehensive the test coverage is. Upstream changes in 1.1.21 and
1.1.22 can be found at
https://git.musl-libc.org/cgit/musl/tree/WHATSNEW?h=v1.1.22&id=e97681d6f2c44bf5fa9ecdd30607cb63c780062e#n1989.
bors [Sat, 29 Jun 2019 11:45:11 +0000 (11:45 +0000)]
Auto merge of #62226 - Centril:rollup-rcy1alx, r=Centril
Rollup of 7 pull requests
Successful merges:
- #61199 (Revert "Set test flag when rustdoc is running with --test option" )
- #61755 (Add `--pass $mode` to compiletest through `./x.py`)
- #61818 (Issue #60709 test)
- #62023 (publish_toolstate: don't use 'new' from inside the loop)
- #62104 (Inform the query system about properties of queries at compile time)
- #62163 (Avoid mem::uninitialized() in std::sys::unix)
- #62204 (doc(libcore) Fix CS)
Rollup merge of #62023 - RalfJung:miri-toolstate, r=kennytm
publish_toolstate: don't use 'new' from inside the loop
I think I made a mistake in https://github.com/rust-lang/rust/pull/61938 by using `new` outside the inner loop. This PR fixes that.
However, no issue got created at all for https://github.com/rust-lang/rust/pull/62003#issuecomment-504356964, and I don't know why that is. The script should be [triggered by Traivs](https://github.com/rust-lang/rust/blob/56a12b2ad058f22f1ef090713df15598525ba4a4/.travis.yml#L240), and I can find no trace of that in [the build log](https://travis-ci.com/rust-lang/rust/jobs/209880042).
projects / rust.git / log