Commit Graph

1814 Commits

Author SHA1 Message Date
Alexis Laferrière 43654566ae Frontend: Intro emit-polyglot-ast action 2026-03-25 11:36:15 -07:00
Ryan Mansfield ae78773ad8 Swap the alias relationship between -language-mode and -swift-version.
As specified in SE-0441, -language-mode should appear in help output
while -swift-version should be the hidden alias.

Related to swiftlang/swift-driver#1894
2026-02-18 16:06:08 -05:00
Hamish Knight eb45f2396d Replace vfs::getRealFileSystem -> vfs::createPhysicalFileSystem
Calling `setCurrentWorkingDirectory` on `getRealFileSystem` sets the
working directory for the process itself. This is unsafe for clients
such as SourceKit which can process multiple concurrent requests,
and may be used as an in-process library in e.g sourcekit-lsp. Switch
to `createPhysicalFileSystem` instead, where setting the working
directory is done locally on the FileSystem itself.
2026-02-09 09:46:34 +00:00
Alastair Houghton e38c0652b6 Merge pull request #84906 from al45tair/eng/PR-101623384
[Backtracing] Add initial support for Windows.
2026-02-06 14:53:16 +00:00
Alastair Houghton 54385105a0 [Backtracing] We don't need to change the driver.
This change in the driver isn't necessary after all.

rdar://101623384
2026-02-06 08:53:16 +00:00
Ian Anderson 69fa949c01 [Triple] Support the new target triple "os" for firmware
Add support for the new Firmware llvm::Triple::OSType, and make a swift::DarwinPlatformKind to go with it.

rdar://165361368
2026-02-03 13:02:44 -08:00
Alastair Houghton a02fea9c7a [Frontend][Driver] Fix Windows paths for swiftCore.
For testing, we need to make sure that we pick up `libswiftCore.dll`
from the `bin` directory, rather than the copy in `lib`.  Otherwise,
when we use the interpreter mode, we'll `LoadLibrary` the one from
`lib`, but Windows found the one in `bin` already, so we end up with
two `libswiftCore.dll`s in the same process, which is bad.

rdar://101623384
2026-02-03 18:23:29 +00:00
Adrian Prantl 0ec2762f96 [driver] Implement -debug-module-path in the legacy driver
Not having suport for this option causes confusing and non-obvious failures in
the LLDB testsuite, and some people build LLDB without also compiling the driver.

rdar://168107340
2026-01-22 17:03:32 -08:00
Anthony Latsis 8572b7e38c Address llvm::StringSwitch deprecations in advance
There's a whole bunch of these, e.g.
- https://github.com/llvm/llvm-project/pull/163405
- https://github.com/llvm/llvm-project/pull/164276
- https://github.com/llvm/llvm-project/pull/165119
- https://github.com/llvm/llvm-project/pull/166016
- https://github.com/llvm/llvm-project/pull/166066
2025-12-02 17:13:17 +00:00
Dylan Sturgeon 4b4f9f18fc Add an option for symbol graph to support long module names. (#83782)
Currently symbol graphs are always written in files that contain 1 to 2
module names. It's possible for Swift module names to be very long, so
combining 2 of them in file name like `module1@module2...` in the same
path component means the name can be too long for some file systems. The
new option `-symbol-graph-shorten-output-names` changes the symbol graph
output files to use a MD5 hash of the module name(s) as the filename and
outputs an additional JSON file with the original names mapped to the
real filename. The module names JSON can be used to construct a VFS
overlay with the original naming scheme.

fix #83723

I considered using vfsoverlay, which seems like a viable solution, but
the vfsoverlay options don't seem to apply to any of the outputs from
the compiler. When I set an overlay to remap the symbol graph file
outputs, the remapped external paths aren't used so the root problem of
too long file names remains.
2025-11-06 19:30:44 -08:00
Becca Royal-Gordon 9454c0aaa4 Emit module selectors in swiftinterfaces
This support is currently opt-in and can be disabled by a blocklist.
2025-10-31 16:48:06 -07:00
Doug Gregor ce5b1a126d Merge pull request #84392 from rjmansfield/sil-ir-extra-outputs
Add frontend options to write SIL and LLVM IR as additional compilation output
2025-10-10 12:09:38 -07:00
Chirag Ramani 5179bc9609 Add IRPGO and CSIRPGO options to Swift (#84335)
This PR introduces three new instrumentation flags and plumbs them
through to IRGen:

1. `-ir-profile-generate` - enable IR-level instrumentation.
2. `-cs-profile-generate` - enable context-sensitive IR-level
instrumentation.
3. `-ir-profile-use` - IR-level PGO input profdata file to enable
profile-guided optimization (both IRPGO and CSIRPGO)

**Context:**
https://forums.swift.org/t/ir-level-pgo-instrumentation-in-swift/82123

**Swift-driver PR:** https://github.com/swiftlang/swift-driver/pull/1992

**Tests and validation:**
This PR includes ir level verification tests, also checks few edge-cases
when `-ir-profile-use` supplied profile is either missing or is an
invalid IR profile.

However, for argument validation, linking, and generating IR profiles
that can later be consumed by -cs-profile-generate, I’ll need
corresponding swift-driver changes. Those changes are being tracked in
https://github.com/swiftlang/swift-driver/pull/1992
2025-10-09 17:41:47 -07:00
Ryan Mansfield ba0ce8aea6 Add frontend options to write SIL and LLVM IR as additional compilation output.
This commit adds -sil-output-path and -ir-output-path frontend options that
allow generating SIL and LLVM IR files as supplementary outputs during normal
compilation.

These options can be useful for debugging and analysis tools
workflows that need access to intermediate compilation artifacts
without requiring separate compiler invocations.

Expected behaviour:

Primary File mode:
 - SIL: Generates one .sil file per source file
 - IR: Generates one .ll file per source file

Single-threaded WMO mode:
 - SIL: Generates one .sil file for the entire module
 - IR: Generates one .ll file for the entire module

Multi-threaded WMO mode:
 - SIL: Generates one .sil file for the entire module
 - IR: Generates separate .ll files per source file

File Maps with WMO:
 - Both SIL and IR outputs using first entry's naming, which is
   consistent with the behaviour of other supplementary outputs.

rdar://160297898
2025-10-06 15:45:49 -04:00
Doug Gregor ae8f9d8234 Add -print-static-build-config to print a static build conformance
Introduce the ability to form a `StaticBuildConfiguration` from
language options. Add a frontend option `-print-static-build-config`
to then print that static build configuration as JSON in a manner that
can be decoded into a `StaticBuildConfiguration`.

Most of the change here is in sinking the bridged ASTContext queries
of language options into a new BridgedLangOptions. The printing of the
static build configuration only has a LangOptions (not an ASTContext),
so this refactoring is required for printing.
2025-09-29 18:42:14 -07:00
Doug Gregor 2383d7ab2d Introduce "-internal" variant of bridging header import flags
The flags "-import-bridging-header" and "-import-pch" import a bridging
header, treating the contents as a public import. Introduce
"internal-" variants of both flags that provide the same semantics,
but are intended to treat the imported contents as if they came in
through an internal import. This is just plumbing of the options for
the moment.
2025-09-19 16:49:11 -07:00
Doug Gregor ffa6d65f12 [Options] Make -import-bridging-header the canonical spelling instead of -import-objc-header
This command-line option hasn't been Objective-C specific ever, really.
Make the language-independent spelling the primary one to make that
more obvious.
2025-09-19 16:49:11 -07:00
Doug Gregor 56f60635b1 Merge pull request #84328 from DougGregor/coroframealloc-backdeploy 2025-09-16 20:15:50 -07:00
Doug Gregor f7264e327f [IRGen] Only use a stub for swift_coroFrameAlloc when we need it
swift_coroFrameAlloc was introduced in the Swift 6.2 runtime. Give it
the appropriate availability in IRGen, so that it gets weak
availability when needed (per the deployment target). Then, only
create the stub function for calling into swift_coroFrameAlloc or
malloc (when the former isn't available) when we're back-deploying to
a runtime prior to Swift 6.2. This is a small code size/performance
win when allocating coroutine frames on Swift 6.2-or-newer platforms.

This has a side effect of fixing a bug in Embedded Swift, where the
swift_coroFrameAlloc was getting unconditionally set to have weak
external linkage despite behind defined in the same LLVM module
(because it comes from the standard library).

Fixes rdar://149695139 / issue #80947.
2025-09-16 10:51:12 -07:00
Yuta Saito 2b6ea81b9e Merge pull request #84285 from kateinoigakukun/yt/fix-tagged-funcptr
[Legacy Driver][wasm] Pass `--table-base` to reserve low function addresses
2025-09-16 08:26:04 +09:00
Yuta Saito e12642f917 [Legacy Driver][wasm] Pass --table-base to reserve low function addresses
WebAssembly does not have a reserved address space by default, so we
need to explicitly reserve low addresses for extra inhabitants for
enum types with pointer payloads. https://github.com/swiftlang/swift/pull/39300
added `--global-base` to reserve low data addresses, but we also need
to reserve low function addresses with `--table-base` for function
pointers because WebAssembly uses a separate address space for function
pointers.
2025-09-15 06:22:20 +00:00
Yuta Saito 6915a36895 [NFC][wasm] Fix incorrect comment about reserved low addresses 2025-09-15 06:21:35 +00:00
Chris McGee e71009b3ff Add symbol graph option for skipping inherited docs 2025-09-10 19:01:29 -04:00
Chris McGee 1380a3e2a9 Merge pull request #83044 from cmcgee1024/add_more_symbol_graph_opts_swiftc
Add pretty print and skip synthesized members to the frontend options
2025-09-09 15:24:27 -04:00
Chris McGee b8ab5d548d Set driver options for frontend too and propagate them in compiler invocation 2025-09-08 16:41:11 -04:00
Chris McGee 6a38bbe73d Qualify the skip synthesized members option with symbol graph for the compiler frontend
Rollback the additional usages of the new options and test cases
2025-08-07 15:44:33 -04:00
Chris McGee 26168f57cc Create tests for the two new symbol graph options
Add options to the driver and frontend options
2025-07-30 10:24:56 -04:00
Anthony Latsis 2920ea84d1 Address llvm::(Mutable)ArrayRef ctor deprecations
See:
- https://github.com/llvm/llvm-project/pull/146113
- https://github.com/llvm/llvm-project/pull/146011
2025-07-21 12:36:53 +01:00
Evan Wilde 6e9c01015d Merge pull request #82504 from etcwilde/ewilde/unversioned-clang-resource-dirs
Unversioned clang resource OS name
2025-07-02 11:20:59 -07:00
Slava Pestov 34253a8606 Merge pull request #82506 from slavapestov/warn-long-expression-type-checking
Sema: Fix the -warn-long-expression-type-checking flag
2025-06-27 01:10:25 -04:00
Allan Shortlidge 3e780ac69e Driver: Remove unused functions. 2025-06-25 22:23:46 -07:00
Slava Pestov 84d1d115e1 Sema: Change -solver-memory-threshold into a frontend flag for consistency with the others 2025-06-25 22:07:23 -04:00
Evan Wilde dbe7c13160 Unversioned clang resource OS name
The clang resource directory generally does not use a versioned form of
the platform name. This aligns the old driver behavior with the new
driver for FreeBSD support.

https://github.com/swiftlang/swift-driver/blob/a7f32bd005de1b998495ddc052bf78fc3cebdaa5/Sources/SwiftDriver/Jobs/GenericUnixToolchain%2BLinkerSupport.swift#L309
2025-06-25 14:26:05 -07:00
Evan Wilde 7f3fdb322d Merge pull request #82355 from etcwilde/ewilde/fbsd-cxx-interop-runtime
FreeBSD: Link C++ runtimes when Cxx interop is enabled
2025-06-23 14:21:02 -07:00
Anthony Latsis dd78cd6dbe ASTBridging: Bridge swift::PlatformKind directly 2025-06-20 16:46:03 +01:00
Artem Chikin 5eb85acad5 Merge pull request #82031 from artemcm/RemovePlaceholdersInDepScanner
[Dependency Scanning] Remove obsolete placeholder module concept
2025-06-18 12:40:30 -07:00
Evan Wilde 3376785630 Use C++ driver when C++ interop is enabled
This aligns the old driver with the behavior of the new driver.
When building with C++ interop enabled, it's important that we link a
C++ runtime, which is handled by the clang++ driver.
The new driver uses clang++ when linking with C++ enabled, either
through the c++ interoperability mode flag or the experimental C++
interop flag. The old driver only enabled it with the experimental C++
interop flag. This results in the C++ interop tests failing on FreeBSD
and a behavioral difference between what we are testing and what we are
shipping.
2025-06-17 10:42:57 -07:00
Artem Chikin a78ee29692 [Dependency Scanning] Remove obsolete placeholder module concept
This was used a long time ago for a design of a scanner which could rely on the client to specify that some modules *will be* present at a given location but are not yet during the scan. We have long ago determined that the scanner must have all modules available to it at the time of scan for soundness. This code has been stale for a couple of years and it is time to simplify things a bit by deleting it.
2025-06-12 08:32:25 -07:00
Artem Chikin 10cd7baef0 [Legacy Driver] Obsolete and remove batch compilation mode from the legacy driver
It is a maintenance burden and having the legacy driver exist in a simplified state reduces the possibility of things going wrong and hitting old bugs.
2025-06-06 09:51:00 -07:00
Ian Anderson a1d10ed6e7 [Frontend] -Isystem doesn't work everywhere -I does
We missed a spot where we need to pass -Isystem along with -I.

rdar://152267441
2025-06-02 15:06:32 -07:00
Ryan Mansfield cc2189bd05 Merge pull request #78791 from rjmansfield/fix-dwarf-version-legacy
[Driver] Fix -dwarf-version for Darwin toolchains in legacy driver.
2025-05-24 10:47:39 -04:00
Ben Barham 99489503e4 Merge pull request #81729 from bnbarham/add-def-iso-to-old-driver
[Driver] Add missing -default-isolation to the old driver
2025-05-23 12:10:24 -07:00
Ben Barham 0b7fbb6419 [Driver] Add missing -default-isolation to the old driver
Resolves rdar://151643763.
2025-05-22 19:07:44 -07:00
Doug Gregor abad2fae0f Make the optional feature StrictMemorySafety migratable
This feature is essentially self-migrating, but fit it into the
migration flow by marking it as migratable, adding
`-strict-memory-safety:migrate`, and introducing a test.
2025-05-22 11:07:13 +01:00
Hamish Knight 6eb76dfd15 [Driver] Avoid using response files for SourceKit
We only care about the frontend arguments here, creating a response
file is unnecessary.

rdar://98880399
2025-05-16 15:35:15 +01:00
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
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
Evan Wilde 96ea03f255 Merge pull request #80479 from etcwilde/ewilde/clang-linkers-choice
[Driver] Don't hardcode default linker on Linux
2025-04-03 17:32:33 -07:00
Doug Gregor e49afc8797 Merge pull request #80389 from 3405691582/nobtcfi
Add a build flavor to opt-out of BTCFI on OpenBSD.
2025-04-03 13:29:04 -07:00