Commit Graph

4649 Commits

Author SHA1 Message Date
Egor Zhdan
571bd8e9d9 [cxx-interop] Remove a workaround for CoreGraphics module interface
CoreGraphics can now be rebuilt with C++ interop enabled, which makes this workaround obsolete.

rdar://150211857
2025-04-29 22:02:47 +01:00
Pavel Yaskevich
1ba077d922 [Feature] Switch migration postfix from :adoption to :migrate 2025-04-28 13:08:24 -07:00
Pavel Yaskevich
085078dd8a [Feature] Rename Feature APIs from adoption to migration 2025-04-28 11:52:46 -07:00
Arnold Schwaighofer
7d72959028 Merge pull request #81037 from aschwaighofer/irgen_coroFrameAlloc_cleanup
IRGen: Centralize logic that decides whether the platform suppports typed malloc
2025-04-24 13:21:16 -07:00
Arnold Schwaighofer
3f8c743aed IRGen: Centralize logic that decides whether the platform suppports typed malloc
NFC intended
2025-04-23 10:43:06 -07:00
QuietMisdreavus
f7bf701ed9 parse symbol availability allow/block-list args together (#80968) 2025-04-22 11:46:32 -06:00
Doug Gregor
5bd782e178 Merge pull request #80874 from DougGregor/new-diagnostics-formatter-without-sourceloc
[Diagnostics] Use the swift-syntax formatting for invalid source locations, too
2025-04-18 20:04:53 -07:00
Doug Gregor
235242e8b3 [Diagnostics] Use the swift-syntax formatting for invalid source locations, too
The diagnostics formatter from swift-syntax previously only handled
fully-formed diagnostics anchored at a particular syntax node.
Therefore, the compiler would fall back to the existing LLVM-based
diagnostic formatter for diagnostics that had no source location.

Adopt new API in the swift-syntax diagnostics formatter that renders a
diagnostic message without requiring source location information, so
that we consistently use the swift-syntax formatter when it is
selected (which is the default).
2025-04-18 15:36:43 -07:00
Becca Royal-Gordon
8f75878455 Forbid @_borrowed in @abi
It has indirect effects on the accessors, so it shouldn’t matter, but we can defensively redirect the query to the API counterpart anyway.

This was the last `InferredInABIAttr` attribute, so we can now remove all of the infrastructure involved in supporting attribute inference.
2025-04-18 14:50:01 -07:00
Pavel Yaskevich
bc0d47484f Merge pull request #80833 from xedin/print-supported-features
[Frontend] Add a way to print features supported by the compiler
2025-04-18 00:24:47 -07:00
Pavel Yaskevich
18703d64d6 [Frontend] Rework -emit-supported-features mode into -emit-supported-arguments
The "featues" part was never actually implemented and Swift Driver
is replying on information about arguments, so instead of removing
this mode, let's scope it down to "arguments" to be deprecated in
the future.
2025-04-17 12:33:23 -07:00
Pavel Yaskevich
55bd906906 [Frontend] Add -print-supported-features option
This is a replacement for `-emit-supported-features` that prints
all of the upcoming/experimental features supported by the compiler
with some additional meta information in JSON format to stdout.
2025-04-16 16:34:47 -07:00
Doug Gregor
d72763f404 Merge pull request #80814 from DougGregor/diag-group-docs-on-swift-org
Point at diagnostic group documentation on docs.swift.org
2025-04-16 13:44:56 -07:00
Doug Gregor
5b5a3e3d07 Point at diagnostic group documentation on docs.swift.org 2025-04-14 20:08:24 -07:00
QuietMisdreavus
41120da702 [SymbolGraphGen] add flags to filter platforms out of availability metadata (#80778)
* add option to filter availability metadata in symbol graphs

* filter out platform-specific availability in the stdlib docs

rdar://144379124
2025-04-14 15:20:22 -07:00
Alastair Houghton
94b054479b [Concurrency] Remove -executor-factory option and replace with magic type.
We decided that using a magic typealias to set the executor factory was better
than using a compiler option. Remove the `-executor-factory` option, and replace
by looking up the `DefaultExecutorFactory` type, first in the main module, and
then if that fails in Concurrency.

rdar://149058236
2025-04-11 13:55:38 +01:00
Steven Wu
2dd49f5209 Merge pull request #80474 from cachemeifyoucan/eng/PR-148465899
[Caching][Macro] Make macro plugin options cacheable
2025-04-10 16:57:00 -07:00
nate-chandler
2e1494df9f Merge pull request #80701 from nate-chandler/rdar148941214
[CoroutineAccessors] Use yield_once_2 on Darwin and Linux.
2025-04-10 14:46:05 -07:00
Steven Wu
3d38d0dd56 [Caching][Macro] Make macro plugin options cacheable
Currently, the macro plugin options are included as cache key and the
absolute path of the plugin executable and library will affect cache
hit, even the plugin itself is identical.

Using the new option `-resolved-plugin-validation` flag, the macro
plugin paths are remapped just like the other paths during dependency
scanning. `swift-frontend` will unmap to its original path during the
compilation, make sure the content hasn't changed, and load the plugin.
It also hands few other corner cases for macro plugins:

* Make sure the plugin options in the swift module is prefix mapped.
* Make sure the remarks of the macro loading is not cached, as the
  mesasge includes the absolute path of the plugin, and is not
  cacheable.

rdar://148465899
2025-04-10 12:10:00 -07:00
Nate Chandler
056fbc44c9 [CoroutineAccessors] Control ABI via flag. 2025-04-10 07:41:37 -07:00
Allan Shortlidge
abac42b759 Serialization: Encode custom availability domains.
When serializing `@available` attributes, if the attribute applies to a custom
domain include enough information to deserialize the reference to that domain.

Resolves rdar://138441265.
2025-04-08 08:53:42 -07:00
Michael Gottesman
3ff9463957 Make Feature a struct enum so we can put methods on it.
Just noticed this as I was looking at making other changes.
2025-04-05 10:08:29 +01:00
Pavel Yaskevich
8bf8c24bf9 Merge pull request #80503 from xedin/extensible-attr
[AST/Sema] Add `@extensible` attribute on `enum` declarations
2025-04-04 09:17:56 -07:00
Pavel Yaskevich
bf19481ab6 [Frontend/Serialization] Remove ExtensibleEnums experimental flag
For now the semantics provided by `@extensible` keyword on per-enum
basis. We might return this as an upcoming feature in the future with
a way to opt-out.
2025-04-03 16:30:23 -07:00
Steven Wu
80f726a815 [ExplicitModuleBuild] Don't leak chained bridging header in objc header
When generating an objc header from the swift module when a bridging
header is used, make sure to use the original bridging header, not
the chained bridging header. This also avoids incorrectly generated a
header include when no actual bridging header is used, just a chained
bridging header that is coming from a dependency.

rdar://148446465
2025-04-02 11:21:17 -07:00
Allan Shortlidge
645628f3e4 Frontend: Reinstate suppression of warnings during interface type-checking.
We're not ready to start emitting warnings when type-checking modules from
interface. It is causing performance regressions and spurious diagnostics to be
emitted.

Reverts a small part of https://github.com/swiftlang/swift/pull/80360.

Resolves rdar://148257136.
2025-03-31 20:41:24 -07:00
Alastair Houghton
b1c345f1be Merge pull request #80266 from al45tair/custom-executors-take2
[Concurrency] Provide a Swift interface for custom main and global executors.
2025-03-31 09:53:48 +01:00
Doug Gregor
b182c96bd7 Print diagnostic group names by default
Print diagnostic groups as part of the LLVM printer in the same manner as the
Swift one does, always. Make `-print-diagnostic-groups` an inert option, since we
always print diagnostic group names with the `[#GroupName]` syntax.

As part of this, we no longer render the diagnostic group name as part
of the diagnostic *text*, instead leaving it up to the diagnostic
renderer to handle the category appropriately. Update all of the tests
that were depending on `-print-diagnostic-groups` putting it into the
text to instead use the `{{documentation-file=<file name>}}`
diagnostic verification syntax.
2025-03-29 15:40:56 -07:00
Doug Gregor
e88f8995e1 [Diagnostics] Eliminate educational notes in favor of diagnostic groups
We've been converging the implementations of educational notes and
diagnostic groups, where both provide category information in
diagnostics (e.g., `[#StrictMemorySafety]`) and corresponding
short-form documentation files. The diagnostic group model is more
useful in a few ways:

* It provides warnings-as-errors control for warnings in the group
* It is easier to associate a diagnostic with a group with
GROUPED_ERROR/GROUPED_WARNING than it is to have a separate diagnostic
ID -> mapping.
* It is easier to see our progress on diagnostic-group coverage
* It provides an easy name to use for diagnostic purposes.

Collapse the educational-notes infrastructure into diagnostic groups,
migrating all of the existing educational notes into new groups.
Simplify the code paths that dealt with multiple educational notes to
have a single, possibly-missing "category documentation URL", which is
how we're treating this.
2025-03-29 15:40:35 -07:00
Doug Gregor
08b27804f0 Merge pull request #80378 from DougGregor/remove-educational-note-markdown-rendering
[Diagnostics] Remove rendering of educational notes to the terminal
2025-03-28 20:12:44 -07:00
Doug Gregor
6e436a536b Merge pull request #80377 from DougGregor/diagnostics-link-to-github 2025-03-28 20:03:03 -07:00
Doug Gregor
7d569b989d [Diagnostics] Remove rendering of educational notes to the terminal
We're moving over to a model where we provide direct links to educational notes /
diagnostic group notes whenever relevant. Rendering the Markdown from these
files to the terminal is less relevant with this approach, so remove it from the
compiler.
2025-03-28 14:12:27 -07:00
Doug Gregor
2c822634dd Point educational notes at education notes / diagnostic groups markdown on Github
Rather than pointing at a Markdown file in the toolchain directory by default,
which won't render well immediately for most users, point at the sources on
GitHub, which will render it properly and will still work from (e.g.)
CI systems where the toolchain content might not be accessible.

Toolchain-aware diagnostic renders can look in the installed toolchain
location (share/doc/swift/diagnostics) for the markdown files for that
specific version of the markdown files.
2025-03-28 13:57:04 -07:00
Alastair Houghton
47fa71787f Revert "Merge pull request #80224 from glessard/revert-79789-custom-executors"
This reverts commit 06f6358067, reversing
changes made to 033f6679e8.
2025-03-28 10:15:07 +00:00
Becca Royal-Gordon
ce8cc1780a Ignore bad @_hasStorage in module interfaces
An objcImpl bug previously caused `@_hasStorage` to be emitted inside some extensions in module interfaces. An earlier commit in this PR created an error for this, but for backwards compatibility, it would actually be better to simply ignore the attribute in module interfaces. Modify TypeCheckStorage to emit a warning, not an error, in this situation.

Additionally, modify the module interface loader to show warnings when you verify a module interface, but not for other module interface uses (like compiling or importing one). The assumption here is that if you’re verifying a module interface, you’re either the author of the module that created it or you’re investigating a problem with it, and in either case you’d like to be told about minor defects in case they’re related.

Fixes rdar://144811653 thoroughly.
2025-03-27 18:25:11 -07:00
Becca Royal-Gordon
132f49108d Check attributes in @abi attr
This commit compares the attributes on the decl inside the `@abi` attribute to those in the decl it’s attached to, diagnosing ABI-incompatible differences. It also rejects many attributes that don’t need to be specified in the `@abi` attribute, such as ObjC-ness, access control, or ABI-neutral traits like `@discardableResult`, so developers know to remove them.
2025-03-26 10:47:57 -07:00
Alexis Laferrière
a4aaa1063a Merge pull request #80098 from xymus/upstream-swiftmodule-blocklist
Frontend: Upstream blocklist check when ignoring adjacent swiftmodule files
2025-03-26 10:28:30 -07:00
Kuba Mracek
1476c04761 Fix a 'misleading indentation' warning in CompilerInvocation.cpp 2025-03-25 13:43:36 -07:00
Kuba (Brecka) Mracek
41ae31647a Merge pull request #76045 from kubamracek/mergeable-traps
Add -Xfrontend -mergeable-traps as a way to emit mergeable traps
2025-03-25 07:13:18 -07:00
Kuba Mracek
b08f630c23 Fix incorrect computation of Opts.MergeableTraps 2025-03-24 14:51:43 -07:00
Artem Chikin
c36ae0d0d6 Merge pull request #79818 from artemcm/SeparateSDKExplicitModules
[Dependency Scanning] Add support for placing explicitly-built SDK modules into a separate module cache
2025-03-24 12:27:06 -06:00
Jan Svoboda
0aeba05790 [Frontend] Use new LockFileManager APIs (#80123) 2025-03-24 10:14:24 -07:00
Pavel Yaskevich
8f4220c6fa Merge pull request #80219 from xedin/rdar-145768557
[Concurrency] SE-0466: Replace `UnspecifiedMeansMainActorIsolated` flag with `-default-isolation`
2025-03-24 09:05:07 -07:00
Kuba Mracek
ed5e89a501 Also respect -mergeable-traps when merging cond_fails in SILOptimizer 2025-03-24 09:04:44 -07:00
Kuba Mracek
164b61c83d Add -Xfrontend -mergeable-traps as a way to emit mergeable traps 2025-03-24 09:04:11 -07:00
Andrew Trick
d97a8cec64 Merge pull request #80223 from atrick/enable-addressable
LifetimeDependence: enable addressable dependencies
2025-03-24 08:30:01 -07:00
Doug Gregor
c69cb76e62 Merge pull request #80228 from DougGregor/serialized-category-documentation-url
[Serialized diagnostics] Emit category documentation URL
2025-03-24 07:45:01 -07:00
Pavel Yaskevich
2221c140d2 [Frontend] SE-0466: Add -default-isolation frontend that accepts MainActor and nonisolated 2025-03-23 22:04:32 -07:00
Andrew Trick
afb0de5d3f LifetimeDependence: enable addressable dependencies 2025-03-23 19:00:42 -07:00
Hamish Knight
e2718f3526 Merge pull request #80162 from hamishknight/inspector-gadget
Remove `IDEInspectionFileRequest`
2025-03-23 00:56:18 +00:00