Cygwin support in rustc #140154

Berrysoft · 2025-04-22T08:29:40Z

This PR builds host rustc targeting cygwin.

Currently supported:

rustc
rustdoc
rustfmt
clippy

Blocking:

cargo: blocked by Add cygwin support socket2#568
rust-analyzer: needs cargo update, fixed upstream

$ rustc --version --verbose
rustc 1.88.0-dev
binary: rustc
commit-hash: unknown
commit-date: unknown
host: x86_64-pc-cygwin
release: 1.88.0-dev
LLVM version: 20.1.4

rustbot · 2025-04-22T08:29:47Z

r? @compiler-errors

rustbot has assigned @compiler-errors.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

src/bootstrap/src/core/builder/mod.rs

jeremyd2019 · 2025-04-29T18:23:32Z

I'm looking forward to being able to try this, and I bet @tyan0 and others from Cygwin are too!

jeremyd2019 · 2025-05-08T22:57:28Z

The build script of rustc_llvm is modified a little because llvm-config of MSYS2 is broken:

This should be fixed now (llvm/llvm-project#136599), and _GNU_SOURCE is defined in llvm-config --cppflags and friends (llvm/llvm-project#138329):

$ llvm-config --cxxflags
-I/usr/include -std=c++17   -fno-exceptions -funwind-tables -D_GNU_SOURCE -D__STDC_CONSTANT_MACROS -D__STDC_FORMAT_MACROS -D__STDC_LIMIT_MACROS

Berrysoft · 2025-05-14T07:18:46Z

r? @jieyouxu

Co-authored-by: Ookiineko <[email protected]>

compiler/rustc_session/src/filesearch.rs

rustbot · 2025-05-16T09:47:54Z

The Miri subtree was changed

cc @rust-lang/miri

jieyouxu

Seems fine, I'll r+ after PR CI is green.

jieyouxu · 2025-05-16T11:24:52Z

@bors r+

bors · 2025-05-16T11:24:56Z

📌 Commit 12c8fb8 has been approved by jieyouxu

It is now in the queue for this repository.

src/tools/miri/src/shims/native_lib.rs

RalfJung · 2025-05-16T21:17:51Z

compiler/rustc_session/src/filesearch.rs

+                let fname_ptr = info.dli_fname.as_ptr();
+                #[cfg(not(target_os = "cygwin"))]
+                let fname_ptr = {
+                    if info.dli_fname.is_null() {


Hm, this can be null? I can't find that in the docs. Miri seems to assume it's never null...

(This is pre-existing, not new in this PR.)

https://man7.org/linux/man-pages/man3/dladdr.3.html

If the address specified in addr could not be matched to a shared
object, then these functions return 0. In this case, an error
message is not available via dlerror(3).

My read of that is that either it finds a shared object, in which case dli_fname is valid, or else it will return 0 (but not tell you why in dlerror).

dladdr can return null, yes. That's handled above. I am talking about dli_fname being null, which is different and which should not be possible according to the docs.

I opened a PR for this: #141239

Cygwin support in rustc This PR builds host rustc targeting cygwin. - [x] rust-lang/stacker#122 - [x] nagisa/rust_libloading#173 - [x] Detegr/rust-ctrlc#131 - [x] rust-random/getrandom#654 - [x] msys2/MSYS2-packages#5350 - [x] rust-lang#140886 - [x] rust-lang#140921 - [x] rust-lang#140973 Currently supported: * rustc * rustdoc * rustfmt * clippy Blocking: * cargo: blocked by rust-lang/socket2#568 * rust-analyzer: needs `cargo update`, fixed upstream ``` $ rustc --version --verbose rustc 1.88.0-dev binary: rustc commit-hash: unknown commit-date: unknown host: x86_64-pc-cygwin release: 1.88.0-dev LLVM version: 20.1.4 ```

Rollup of 6 pull requests Successful merges: - rust-lang#127013 (Add `f16` formatting and parsing) - rust-lang#140154 (Cygwin support in rustc) - rust-lang#140628 (std: stop using TLS in signal handler) - rust-lang#140966 (Remove #![feature(let_chains)] from library and src/librustdoc) - rust-lang#140994 (replace `cc_detect::cc2ar` with `cc::try_get_archiver`) - rust-lang#141127 (bump windows crate for compiler,bootstrap and tools) r? `@ghost` `@rustbot` modify labels: rollup

Cygwin support in rustc This PR builds host rustc targeting cygwin. - [x] rust-lang/stacker#122 - [x] nagisa/rust_libloading#173 - [x] Detegr/rust-ctrlc#131 - [x] rust-random/getrandom#654 - [x] msys2/MSYS2-packages#5350 - [x] rust-lang#140886 - [x] rust-lang#140921 - [x] rust-lang#140973 Currently supported: * rustc * rustdoc * rustfmt * clippy Blocking: * cargo: blocked by rust-lang/socket2#568 * rust-analyzer: needs `cargo update`, fixed upstream ``` $ rustc --version --verbose rustc 1.88.0-dev binary: rustc commit-hash: unknown commit-date: unknown host: x86_64-pc-cygwin release: 1.88.0-dev LLVM version: 20.1.4 ```

Rollup of 10 pull requests Successful merges: - rust-lang#127013 (Add `f16` formatting and parsing) - rust-lang#138940 (Stabilize the avx512 target features) - rust-lang#140154 (Cygwin support in rustc) - rust-lang#140490 (split `asm!` parsing and validation) - rust-lang#140628 (std: stop using TLS in signal handler) - rust-lang#140746 (name resolution for guard patterns) - rust-lang#140926 (Return value of coroutine_layout fn changed to Result with LayoutError) - rust-lang#141127 (bump windows crate for compiler,bootstrap and tools) - rust-lang#141214 (Miri subtree update) - rust-lang#141218 (gvn: avoid creating overlapping assignments) r? `@ghost` `@rustbot` modify labels: rollup

bors · 2025-05-18T16:44:08Z

⌛ Testing commit 12c8fb8 with merge 27c4e28...

Cygwin support in rustc This PR builds host rustc targeting cygwin. - [x] rust-lang/stacker#122 - [x] nagisa/rust_libloading#173 - [x] Detegr/rust-ctrlc#131 - [x] rust-random/getrandom#654 - [x] msys2/MSYS2-packages#5350 - [x] rust-lang#140886 - [x] rust-lang#140921 - [x] rust-lang#140973 Currently supported: * rustc * rustdoc * rustfmt * clippy Blocking: * cargo: blocked by rust-lang/socket2#568 * rust-analyzer: needs `cargo update`, fixed upstream ``` $ rustc --version --verbose rustc 1.88.0-dev binary: rustc commit-hash: unknown commit-date: unknown host: x86_64-pc-cygwin release: 1.88.0-dev LLVM version: 20.1.4 ```

fmease · 2025-05-18T16:45:02Z

@bors retry (yield to rollup, sry)

Cygwin support in rustc This PR builds host rustc targeting cygwin. - [x] rust-lang/stacker#122 - [x] nagisa/rust_libloading#173 - [x] Detegr/rust-ctrlc#131 - [x] rust-random/getrandom#654 - [x] msys2/MSYS2-packages#5350 - [x] rust-lang#140886 - [x] rust-lang#140921 - [x] rust-lang#140973 Currently supported: * rustc * rustdoc * rustfmt * clippy Blocking: * cargo: blocked by rust-lang/socket2#568 * rust-analyzer: needs `cargo update`, fixed upstream ``` $ rustc --version --verbose rustc 1.88.0-dev binary: rustc commit-hash: unknown commit-date: unknown host: x86_64-pc-cygwin release: 1.88.0-dev LLVM version: 20.1.4 ```

Rollup of 6 pull requests Successful merges: - rust-lang#127013 (Add `f16` formatting and parsing) - rust-lang#130887 (Safer implementation of RepeatN) - rust-lang#140154 (Cygwin support in rustc) - rust-lang#140874 (make `rustc_attr_parsing` less dominant in the rustc crate graph) - rust-lang#141094 (add regression test for rust-lang#101650) - rust-lang#141110 ([std] fix the presentation of `split_off_mut` and `split_off` documentation) r? `@ghost` `@rustbot` modify labels: rollup

bors · 2025-05-18T23:53:08Z

⌛ Testing commit 12c8fb8 with merge e42bbfe...

bors · 2025-05-19T03:04:09Z

☀️ Test successful - checks-actions
Approved by: jieyouxu
Pushing e42bbfe to master...

github-actions · 2025-05-19T03:07:06Z

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing 4d051fb (parent) -> e42bbfe (this PR)

Test differences

No test diffs found

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard e42bbfe1f7c26f8760a99c4b1f27d33aba1040bb --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

dist-aarch64-msvc: 8392.8s -> 9988.1s (19.0%)
dist-x86_64-msvc-alt: 7229.5s -> 8453.0s (16.9%)
x86_64-apple-2: 4440.8s -> 5014.1s (12.9%)
x86_64-apple-1: 6936.2s -> 7747.4s (11.7%)
dist-arm-linux: 4717.1s -> 5118.8s (8.5%)
dist-apple-various: 6435.1s -> 6031.4s (-6.3%)
dist-aarch64-apple: 5038.3s -> 5350.2s (6.2%)
dist-x86_64-mingw: 7570.2s -> 7985.9s (5.5%)
i686-msvc-1: 9029.3s -> 9483.5s (5.0%)
x86_64-gnu-stable: 6518.7s -> 6823.8s (4.7%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

rust-timer · 2025-05-19T04:35:19Z

Finished benchmarking commit (e42bbfe): comparison URL.

Overall result: ❌✅ regressions and improvements - no action needed

@rustbot label: -perf-regression

Instruction count

This is the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.2%	[0.2%, 0.2%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.4%	[-0.4%, -0.4%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (secondary 0.4%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	1.2%	[0.7%, 1.8%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.4%	[-0.4%, -0.4%]	2
All ❌✅ (primary)	-	-	0

Cycles

Results (secondary -0.4%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.5%	[0.4%, 0.6%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.7%	[-1.1%, -0.4%]	6
All ❌✅ (primary)	-	-	0

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 775.275s -> 774.961s (-0.04%)
Artifact size: 365.57 MiB -> 365.57 MiB (-0.00%)

jeremyd2019 · 2025-05-20T23:10:27Z

src/bootstrap/src/utils/shared_helpers.rs

@@ -46,7 +46,7 @@ pub fn dylib_path() -> Vec<std::path::PathBuf> {
 /// Given an executable called `name`, return the filename for the
 /// executable for a particular target.
 pub fn exe(name: &str, target: &str) -> String {
-    if target.contains("windows") || target.contains("cygwin") {
+    if target.contains("windows") {


I'm confused by this. I just had to re-discover this change because bootstrap panicked not finding rustc-main when it was named rustc-main.exe, cross-compiling from Windows to Cygwin. While Cygwin does have 'magic' that will let you access a file X.exe by the name X, some other host wouldn't know to do that. What was the problem that reverting this change addressed, and is there maybe some other way to fix it?

The rustc wrapper of bootstrap failed to determine the correct parameters. The wrapper is executed like /foo/rustc /bar/rustc --baz.

I think there should be other solutions, but it's only a bootstrap tool. So rely on the cygwin magic is fine.

I thought that bootstrap tool was the general way to cross-compile rust (build != host), and that should work for build != cygwin host = cygwin

Well it might be complex then... Although the bootstrap tool should provide a way to cross-compile, it's not that common to cross compile from Windows to a Unix platform (Cygwin). Yet it would be better if it can support that case.

Linux to Cygwin should work though? I might try that and see.

OK, I see what the problem is - /proc/self/exe's target is without the .exe extension, so the comparison in bootstrap's rustc wrapper would not match on a Cygwin host if this function returned the name with .exe appended. However, on a non-Cygwin host, bootstrap needs to know that executables for a Cygwin target will have a .exe extension. I'm not sure how to reconcile these two different uses of exe function.

Linux to Cygwin should work though?

The great pioneer @Ookiineko succeeded compiling rustc from Fedora to Cygwin.

rustbot assigned compiler-errors Apr 22, 2025

This comment was marked as outdated.

Sign in to view

jieyouxu reviewed Apr 22, 2025

View reviewed changes

src/bootstrap/src/core/builder/mod.rs Show resolved Hide resolved

jieyouxu added the O-cygwin Target: *-pc-cygwin label Apr 25, 2025

This comment was marked as outdated.

Sign in to view

Berrysoft force-pushed the cygwin-host branch 2 times, most recently from 7ee303f to d23fa03 Compare May 13, 2025 16:35

Berrysoft changed the title ~~Experimental cygwin support in rustc~~ Cygwin support in rustc May 14, 2025

rustbot assigned jieyouxu and unassigned compiler-errors May 14, 2025

Berrysoft and others added 2 commits May 15, 2025 23:34

Experimental cygwin support in rustc

49553be

Co-authored-by: Ookiineko <[email protected]>

Fix exe() to make rustc wrapper happy

12fda86

Berrysoft force-pushed the cygwin-host branch from d23fa03 to 12fda86 Compare May 15, 2025 15:34

Berrysoft requested a review from jieyouxu May 15, 2025 15:34

jeremyd2019 reviewed May 15, 2025

View reviewed changes

compiler/rustc_session/src/filesearch.rs Outdated Show resolved Hide resolved

Berrysoft added 2 commits May 16, 2025 09:41

Simplify current_dll_path for Cygwin

5e048e0

Fix fname for miri

12c8fb8

jieyouxu approved these changes May 16, 2025

View reviewed changes

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels May 16, 2025

RalfJung reviewed May 16, 2025

View reviewed changes

src/tools/miri/src/shims/native_lib.rs Show resolved Hide resolved

RalfJung reviewed May 16, 2025

View reviewed changes

fmease mentioned this pull request May 18, 2025

Rollup of 6 pull requests #141208

Closed

fmease mentioned this pull request May 18, 2025

Rollup of 10 pull requests #141228

Closed

fmease mentioned this pull request May 18, 2025

Rollup of 6 pull requests #141233

Closed

bors added the merged-by-bors This PR was explicitly merged by bors. label May 19, 2025

bors merged commit e42bbfe into rust-lang:master May 19, 2025
7 checks passed

rustbot added this to the 1.89.0 milestone May 19, 2025

Berrysoft deleted the cygwin-host branch May 19, 2025 03:06

jeremyd2019 reviewed May 20, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Cygwin support in rustc #140154

Cygwin support in rustc #140154

Berrysoft commented Apr 22, 2025 •

edited

Loading

rustbot commented Apr 22, 2025

This comment was marked as outdated.

jeremyd2019 commented Apr 29, 2025

jeremyd2019 commented May 8, 2025 •

edited

Loading

This comment was marked as outdated.

Berrysoft commented May 14, 2025

rustbot commented May 16, 2025

jieyouxu left a comment

jieyouxu commented May 16, 2025

bors commented May 16, 2025

RalfJung May 16, 2025

jeremyd2019 May 16, 2025 •

edited

Loading

RalfJung May 17, 2025

RalfJung May 19, 2025

bors commented May 18, 2025

fmease commented May 18, 2025

bors commented May 18, 2025

bors commented May 19, 2025

github-actions bot commented May 19, 2025

rust-timer commented May 19, 2025

jeremyd2019 May 20, 2025

Berrysoft May 21, 2025

jeremyd2019 May 21, 2025

Berrysoft May 21, 2025

jeremyd2019 May 21, 2025

jeremyd2019 May 21, 2025

Berrysoft May 21, 2025

Cygwin support in rustc #140154

Cygwin support in rustc #140154

Conversation

Berrysoft commented Apr 22, 2025 • edited Loading

rustbot commented Apr 22, 2025

This comment was marked as outdated.

jeremyd2019 commented Apr 29, 2025

jeremyd2019 commented May 8, 2025 • edited Loading

This comment was marked as outdated.

Berrysoft commented May 14, 2025

rustbot commented May 16, 2025

jieyouxu left a comment

Choose a reason for hiding this comment

jieyouxu commented May 16, 2025

bors commented May 16, 2025

Choose a reason for hiding this comment

jeremyd2019 May 16, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bors commented May 18, 2025

fmease commented May 18, 2025

bors commented May 18, 2025

bors commented May 19, 2025

github-actions bot commented May 19, 2025

Test differences

Job duration changes

rust-timer commented May 19, 2025

Overall result: ❌✅ regressions and improvements - no action needed

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Berrysoft commented Apr 22, 2025 •

edited

Loading

jeremyd2019 commented May 8, 2025 •

edited

Loading

jeremyd2019 May 16, 2025 •

edited

Loading