about summary refs log tree commit diff
path: root/compiler/rustc_session/src/config.rs
AgeCommit message (Collapse)AuthorLines
2023-11-08Rollup merge of #117650 - saethlin:inline-me-please, r=davidtwcoMatthias Krüger-4/+18
Add -Zcross-crate-inline-threshold=yes ``@thomcc`` says this would be useful for > seeing if it makes a difference in some code if i do it when building the sysroot, since -Zbuild-std + lto helps more than it seems like it should And I've changed the possible values as a reference to ``@Manishearth`` saying > LLVM's inlining heuristic is "yes".
2023-11-07Add -Zcross-crate-inline-threshold=yesBen Kimock-4/+18
2023-11-08rustc: minor changes suggested by clippy perf lints.Nicholas Nethercote-1/+1
2023-11-03Auto merge of #117507 - nnethercote:rustc_span, r=Nilstriebbors-3/+2
`rustc_span` cleanups Just some things I found while looking over this crate. r? `@oli-obk`
2023-11-02Minimize `pub` usage in `source_map.rs`.Nicholas Nethercote-3/+2
Most notably, this commit changes the `pub use crate::*;` in that file to `use crate::*;`. This requires a lot of `use` items in other crates to be adjusted, because everything defined within `rustc_span::*` was also available via `rustc_span::source_map::*`, which is bizarre. The commit also removes `SourceMap::span_to_relative_line_string`, which is unused.
2023-11-02Remove support for alias `-Z symbol-mangling-version`Zalathar-21/+12
2023-10-30Make `Cfg` and `CheckCfg` non-generic.Nicholas Nethercote-21/+10
They now only ever contains symbols.
2023-10-30Change cfg parsers to produce symbols instead of strings.Nicholas Nethercote-31/+1
2023-10-28Change `Cfg<T>` to an `FxIndexSet`.Nicholas Nethercote-2/+5
Despite what I claimed in an earlier commit, the ordering does matter to some degree. Using `FxIndexSet` prevents changes to the error message order in `tests/ui/check-cfg/mix.rs`.
2023-10-28Clean up config mess.Nicholas Nethercote-25/+22
`parse_cfgspecs` and `parse_check_cfg` run very early, before the main interner is running. They each use a short-lived interner and convert all interned symbols to strings in their output data structures. Once the main interner starts up, these data structures get converted into new data structures that are identical except with the strings converted to symbols. All is not obvious from the current code, which is a mess, particularly with inconsistent naming that obscures the parallel string/symbol data structures. This commit clean things up a lot. - The existing `CheckCfg` type is generic, allowing both `CheckCfg<String>` and `CheckCfg<Symbol>` forms. This is really useful, but it defaults to `String`. The commit removes the default so we have to use `CheckCfg<String>` and `CheckCfg<Symbol>` explicitly, which makes things clearer. - Introduces `Cfg`, which is generic over `String` and `Symbol`, similar to `CheckCfg`. - Renames some things. - `parse_cfgspecs` -> `parse_cfg` - `CfgSpecs` -> `Cfg<String>`, plus it's used in more places, rather than the underlying `FxHashSet` type. - `CrateConfig` -> `Cfg<Symbol>`. - `CrateCheckConfig` -> `CheckCfg<Symbol>` - Adds some comments explaining the string-to-symbol conversions. - `to_crate_check_config`, which converts `CheckCfg<String>` to `CheckCfg<Symbol>`, is inlined and removed and combined with the overly-general `CheckCfg::map_data` to produce `CheckCfg::<String>::intern`. - `build_configuration` now does the `Cfg<String>`-to-`Cfg<Symbol>` conversion, so callers don't need to, which removes the need for `to_crate_config`. The diff for two of the fields in `Config` is a good example of the improved clarity: ``` - pub crate_cfg: FxHashSet<(String, Option<String>)>, - pub crate_check_cfg: CheckCfg, + pub crate_cfg: Cfg<String>, + pub crate_check_cfg: CheckCfg<String>, ``` Compare that with the diff for the corresponding fields in `ParseSess`, and the relationship to `Config` is much clearer than before: ``` - pub config: CrateConfig, - pub check_config: CrateCheckConfig, + pub config: Cfg<Symbol>, + pub check_config: CheckCfg<Symbol>, ```
2023-10-26The value of `-Cinstrument-coverage=` doesn't need to be `Option`Zalathar-7/+5
Not using this flag is identical to passing `-Cinstrument-coverage=off`, so there's no need to distinguish between `None` and `Some(Off)`.
2023-10-25Remove support for alias `-Z instrument-coverage`Zalathar-22/+17
This flag was stabilized in rustc 1.60.0 as `-C instrument-coverage`, but the old unstable flag was kept around as an alias to ease migration.
2023-10-24Introduce `-C instrument-coverage=branch` to gate branch coverageArpad Borsos-1/+7
This flag has to be used in combination with `-Zunstable-options`, and is added in advance of adding branch coverage instrumentation.
2023-10-23Rollup merge of #116960 - lqd:applied-member-constraints-scope, r=matthewjasperMatthias Krüger-7/+2
Location-insensitive polonius: consider a loan escaping if an SCC has member constraints applied only The location-insensitive analysis considered loans to escape if there were member constraints, which makes *some* sense for scopes and matches the scopes that NLL computes on all the tests. However, polonius and NLLs differ on the fuzzed case #116657, where an SCC has member constraints but no applied ones (and is kinda surprising). The existing UI tests with member constraints impacting scopes all have some constraint applied. This PR changes the location-insensitive analysis to consider a loan to escape if there are applied member constraints, and for extra paranoia/insurance via fuzzing and crater: actually checks the constraint's min choice is indeed a universal region as we expect. (This could be turned into a `debug_assert` and early return as a slight optimization after these periods of verification) The 4 UI tests where member constraints are meaningful for computing scopes still pass obviously, and this also fixes #116657. r? `@matthewjasper`
2023-10-20slight Default cleanup for optionRémy Rakic-7/+2
2023-10-17[RFC 3127 - Trim Paths]: Condition remapped filepath on remap scopesUrgau-3/+19
2023-10-17[RFC 3127 - Trim Paths]: Add unstable option and parsingUrgau-2/+29
2023-10-14Add `Config::hash_untracked_state` callbackAlex Macleod-5/+6
2023-10-13Make `needs_analysis` true for `PpHirMode::Typed`.Nicholas Nethercote-1/+1
This avoids the need for a bespoke `tcx.analysis()` call.
2023-10-13Remove PpAstTreeMode.Nicholas Nethercote-14/+9
It's simpler to distinguish the two AST modes directly in `PpMode`.
2023-10-11Auto merge of #113218 - lqd:polonius-scopes, r=jackh726bors-0/+34
Compute NLL loan scopes using the polonius model For a *location-insensitive* analysis (that is, without expressiveness improvements for users yet), this PR implements loans going out of scope using reachability and liveness, rather than checking if the issuing region's values contain a given CFG point. This is equivalent to NLL scopes and computes the same data. r? `@matthewjasper` A couple of notes: - there are some assumptions about SCC representatives, placeholders, free regions, and member constraints that I believe hold, and they're documented in the code - this passes all the UI tests with `-Zpolonius=next` -- the perf is [not terrible](https://github.com/rust-lang/rust/pull/112432#issuecomment-1749685862) and there are a bunch of ways to improve it in the future. - there's a fixme left, hopefully Matthew you know a clean way to get the information it mentions.
2023-10-08consistency check for self-contained linking components CLI optionsRémy Rakic-0/+24
emit an error if components are both enabled and disabled on the CLI
2023-10-08implement opt out `-Clink-self-contained=-linker`Rémy Rakic-12/+30
record both enabled and disabled components so that they can be merged with the ones that the target spec will define
2023-10-08move single component parsing to dedicated functionRémy Rakic-6/+6
this will prevent parsing when expecting more than a single component to be parsed, and prepare for the symetric variant-to-name function to be added
2023-10-08move `LinkSelfContainedComponents` to `rustc_target`Rémy Rakic-35/+1
2023-10-08prepare stabilization of modern linker-flavorsRémy Rakic-3/+2
fix a few comments
2023-10-08linker: Remove `-Zgcc-ld` optionVadim Petrochenkov-4/+3
It is subsumed by `-Clinker-flavor=*-lld-cc -Clink-self-contained=+linker` options now
2023-10-06Remove unused `FileName::CfgSpec`.Nicholas Nethercote-1/+0
2023-10-04extend Polonius options helpersRémy Rakic-1/+6
2023-10-04introduce `Polonius` enum for `-Zpolonius`Rémy Rakic-0/+29
this allows to opt into using the legacy version or the in-tree prototype
2023-09-22Auto merge of #116001 - fmease:validate-crate-name-extern-cli-opt, r=est31bors-0/+13
[breaking change] Validate crate name in `--extern` [MCP 650] Reject non-ASCII-identifier crate names passed to the CLI option `--extern` (`rustc`, `rustdoc`). Implements [MCP 650](https://github.com/rust-lang/compiler-team/issues/650) (except that we only allow ASCII identifiers not arbitrary Rust identifiers). Fixes #113035. [As mentioned on Zulip](https://rust-lang.zulipchat.com/#narrow/stream/233931-t-compiler.2Fmajor-changes/topic/Disallow.20non-identifier-valid.20--extern.20cr.E2.80.A6.20compiler-team.23650/near/376826988), doing a crater run probably doesn't make sense since it wouldn't yield anything. Most users don't interact with `rustc` directly but only ever through Cargo which always passes a valid crate name to `--extern` when it invokes `rustc` and `rustdoc`. In any case, the user wouldn't be able to use such a crate name in the source code anyway. Note that I'm not using [`rustc_session::output::validate_crate_name`](https://doc.rust-lang.org/nightly/nightly-rustc/rustc_session/output/fn.validate_crate_name.html) (used for `--crate-name` and `#![crate_name]`) since the latter doesn't reject non-ASCII crate names and ones that start with a digit. As an aside, I've also thought about getting rid of `validate_crate_name` entirely in a separate PR (with another MCP) in favor of `is_ascii_ident` to reject more weird `--crate-name`s, `#![crate_name]`s and file names but I think that would lead to a lot of actual breakage, namely because of file names starting with a digit. In `tests/ui` 9 tests would be impacted for example. CC `@estebank` r? `@est31`
2023-09-20Validate crate name in CLI option --externLeón Orell Valerian Liehr-0/+13
2023-09-17Auto merge of #114750 - Enselic:metadata-dep-info, r=compiler-errorsbors-1/+11
Make `.rmeta` file in `dep-info` have correct name (`lib` prefix) Since `filename_for_metadata()` and `OutputFilenames::path(OutputType::Metadata)` had different logic for the name of the metadata file, the `.d` file contained a file name different from the actual name used. Share the logic to fix the out-of-sync name. Without this fix, the `.d` file contained dash-separated_something-extra.rmeta: dash-separated.rs instead of libdash_separated_something-extra.rmeta: dash-separated.rs which is the name of the file that is actually written by the compiler. Worth noting: It took me several iterations to get all tests to pass, so I am relatively confident that this PR does not break anything. Closes #68839
2023-09-13Auto merge of #115735 - bjorn3:better_list_crate_metadata, r=wesleywiserbors-1/+1
Extend rustc -Zls This makes it show a lot more things and thus a lot more useful.
2023-09-11Auto merge of #115656 - cjgillot:default-relative-spans, r=davidtwcobors-6/+0
Enable incremental-relative-spans by default. This was enabled on nightly in https://github.com/rust-lang/rust/pull/84762. It has been a while, without obvious bugs. It's time to enable it by default for incremental runs.
2023-09-10Show lib features in -Zls and allow configuring which things are shownbjorn3-1/+1
2023-09-10Auto merge of #115713 - chenyukang:yukang-fix-115680-rustdoc-arg-check, ↵bors-5/+13
r=compiler-errors Abort if check nightly options failed on stable Fixes #115680 Also, if there are multiple unstable options passing on stable compiler, printing multiple same `note` and `help` seems noisy.
2023-09-10Abort if check nightly options failed on stableyukang-5/+13
2023-09-08debuginfo: add compiler option to allow compressed debuginfo sectionsAugie Fackler-5/+35
LLVM already supports emitting compressed debuginfo. In debuginfo=full builds, the debug section is often a large amount of data, and it typically compresses very well (3x is not unreasonable.) We add a new knob to allow debuginfo to be compressed when the matching LLVM functionality is present. Like clang, if a known-but-disabled compression mechanism is requested, we disable compression and emit uncompressed debuginfo sections. The API is different enough on older LLVMs we just pretend the support is missing on LLVM older than 16.
2023-09-07Enable incremental-relative-spans by default.Camille GILLOT-6/+0
2023-09-07Stabilize `PATH` option for `--print KIND=PATH`Urgau-6/+0
Description of the `PATH` option: > A filepath may optionally be specified for each requested information > kind, in the format `--print KIND=PATH`, just like for `--emit`. When > a path is specified, information will be written there instead of to > stdout.
2023-09-03Make `.rmeta` file in `dep-info` have correct name (`lib` prefix)Martin Nordholts-1/+11
Since `filename_for_metadata()` and `OutputFilenames::path(OutputType::Metadata)` had different logic for the name of the metadata file, the `.d` file contained a file name different from the actual name used. Share the logic to fix the out-of-sync name. Closes 68839.
2023-08-18Add the relocation_model to the cfgLuca Barbato-1/+11
This way is possible to write inline assembly code aware of it.
2023-07-30inline format!() args up to and including rustc_middleMatthias Krüger-5/+3
2023-07-27Rollup merge of #114139 - Urgau:make-print-with-path-unstable, r=jackh726Jubilee-0/+6
Make `--print` with path unstable https://github.com/rust-lang/rust/pull/113780 should have gone through an MCP+FCP but wasn't, but instead of reverting the original PR, this PR just make that new option unstable. [Zulip discussion](https://rust-lang.zulipchat.com/#narrow/stream/238009-t-compiler.2Fmeetings/topic/.5Bweekly.5D.202023-07-27/near/379199738) cc `@dtolnay`
2023-07-27Make `--print KIND=PATH` unstableUrgau-0/+6
https://github.com/rust-lang/rust/pull/113780 should have gone through an MCP+FCP but wasn't, but instead of reverting the original PR, this PR just make that new option unstable.
2023-07-26replace atty crate with std's isTerminalklensy-1/+2
2023-07-23match on chars instead of &strs for .split() or .strip_prefix()Matthias Krüger-2/+2
2023-07-21Revert "Auto merge of #113166 - moulins:ref-niches-initial, r=oli-obk"David Tolnay-2/+0
This reverts commit 557359f92512ca88b62a602ebda291f17a953002, reversing changes made to 1e6c09a803fd543a98bfbe1624d697a55300a786.
2023-07-21Auto merge of #113166 - moulins:ref-niches-initial, r=oli-obkbors-0/+2
Prototype: Add unstable `-Z reference-niches` option MCP: rust-lang/compiler-team#641 Relevant RFC: rust-lang/rfcs#3204 This prototype adds a new `-Z reference-niches` option, controlling the range of valid bit-patterns for reference types (`&T` and `&mut T`), thereby enabling new enum niching opportunities. Like `-Z randomize-layout`, this setting is crate-local; as such, references to built-in types (primitives, tuples, ...) are not affected. The possible settings are (here, `MAX` denotes the all-1 bit-pattern): | `-Z reference-niches=` | Valid range | |:---:|:---:| | `null` (the default) | `1..=MAX` | | `size` | `1..=(MAX- size)` | | `align` | `align..=MAX.align_down_to(align)` | | `size,align` | `align..=(MAX-size).align_down_to(align)` | ------ This is very WIP, and I'm not sure the approach I've taken here is the best one, but stage 1 tests pass locally; I believe this is in a good enough state to unleash this upon unsuspecting 3rd-party code, and see what breaks.
generated by cgit-pink 1.4.1-3-g733a5 (git 2.37.1) at 2026-02-28 10:37:55 +0000