Test linking and running `no_std` binaries #138904

madsmtm · 2025-03-24T21:25:40Z

I looked around, but it seems that we do not have a test that tests a #![no_std] + #![no_main] binary. So now I've added one. Motivated by discussion in this Zulip thread.

r? @tgross35

try-job: arm-android
try-job: armhf-gnu
try-job: dist-ohos
try-job: x86_64-mingw-1

madsmtm · 2025-03-24T21:34:45Z

tests/ui/no_std/simple-runs.rs

+// Need to link libSystem on Apple platforms, otherwise the linker fails with:
+// > ld: dynamic executables or dylibs must link with libSystem.dylib
+//
+// With the new linker introduced in Xcode 15, the error is instead:
+// > Undefined symbols: "dyld_stub_binder", referenced from: <initial-undefines>
+//
+// This _can_ be worked around by raising the deployment target with
+// MACOSX_DEPLOYMENT_TARGET=13.0, though it's a bit hard to test that while
+// still allowing the test suite to support running with older Xcode versions.
+#[cfg_attr(target_vendor = "apple", link(name = "System"))]
+extern "C" {}


We could just always link libSystem internally in rustc, but I'm not sure that's a good idea, since as noted, it is possible to create a fully no_std binary that doesn't link libSystem, it just requires the user to do some work.

Maybe better would be to document the limitation somewhere (where?)? Note that libc does the linking internally too (linking to libc.dylib also links to libSystem.dylib), so any code that does #![no_std] but uses libc works without this.

folkertdev · 2025-03-24T21:34:54Z

For completeness, there is https://github.com./rust-lang/rust/blob/master/tests/run-make/thumb-none-qemu/example/src/main.rs so it's not completely untested, but coverage on more platforms is good of course.

madsmtm · 2025-03-24T21:38:50Z

tests/ui/no_std/simple-runs.rs

+
+//@ run-pass
+//@ compile-flags: -Cpanic=abort
+//@ ignore-wasm different `main` convention


Unsure if we need to add more to the ignores here? Is something like ignore-embedded a thing, and would it make sense?

I'm guessing run-pass already omits platforms that can't run standalone executables, cc @jieyouxu for a better answer.

See

rust/src/tools/compiletest/src/runtest.rs

Line 986 in 4fd8c04

let proc_res = match &*self.config.target {

, I believe in the general case run-pass intentionally does not try to be "smart" and omit platforms that can't run standalone executables. Generally tests will need //@ ignore-cross-compile if they need to run the test binary (but that only runs on host).

madsmtm · 2025-03-24T21:39:15Z

For completeness, there is https://github.com./rust-lang/rust/blob/master/tests/run-make/thumb-none-qemu/example/src/main.rs so it's not completely untested, but coverage on more platforms is good of course.

Ah nice. Do you think I should try to merge this test with that?

folkertdev · 2025-03-24T21:59:42Z

I don't think so, that test is very specifically for checking that the embedded ecosystem continues to work. It uses qemu to run on a simulated microcontroller. So instead additionally having this test for many more platforms in the ui tests seems independently useful.

Noratrieb · 2025-03-25T06:05:18Z

tests/ui/no_std/simple-runs.rs

+
+// # Linux
+//
+// Linking `libc` is required by crt1.o, otherwise the linker fails with:


it would be fun to have a _start lib-less linux binary as well

tgross35 · 2025-04-04T03:16:08Z

To have a no_main binary on Windows I believe you need mainCRTStartup as the entrypoint, plus a /subsystem linker arg and a handful of symbols need to be defined (https://nullprogram.com/blog/2023/02/15/ is a pretty good tutorial). I'm not sure if there is something that can be linked to make things "just work" with an unmangled main like MacOS and Linux, @ChrisDenton would likely know. We can still try this to see where it gets.

@bors try

Maybe it's worth adding a second test that links nothing, like Nora's suggestion? I'm not sure that is possible on MacOS but it is probably more interesting on Linux.

Test linking and running `no_std` binaries I looked around, but it seems that we do not have a test that tests a `#![no_std]` + `#![no_main]` binary. So now I've added one. Motivated by discussion in [this Zulip thread](https://rust-lang.zulipchat.com/#narrow/channel/219381-t-libs/topic/Provide.20.60__isPlatformVersionAtLeast.60.20in.20.60std.60.3F/with/507870028). r? `@tgross35` I haven't tested on Windows, so: try-job: x86_64-msvc-1 try-job: x86_64-msvc-2 And only tested lightly on Linux, so: try-job: x86_64-linux try-job: test-various

ChrisDenton · 2025-04-04T03:28:35Z

@tgross35 if you're using C main then it should just work, no extra steps needed, because you're using the C runtime.

tgross35 · 2025-04-04T03:54:47Z

@bors try

Test linking and running `no_std` binaries I looked around, but it seems that we do not have a test that tests a `#![no_std]` + `#![no_main]` binary. So now I've added one. Motivated by discussion in [this Zulip thread](https://rust-lang.zulipchat.com/#narrow/channel/219381-t-libs/topic/Provide.20.60__isPlatformVersionAtLeast.60.20in.20.60std.60.3F/with/507870028). r? `@tgross35` I haven't tested on Windows, so: try-job: x86_64-msvc-1 try-job: x86_64-msvc-2 And only tested lightly on Linux, so: try-job: x86_64-gnu try-job: test-various

bors · 2025-04-04T03:55:58Z

⌛ Trying commit f9770e7 with merge bb9c57a...

bors · 2025-04-04T06:28:18Z

☀️ Try build successful - checks-actions
Build commit: bb9c57a (bb9c57adf8785606340844eeae4898130594e2ff)

tgross35 · 2025-04-08T02:10:28Z

LGTM but may as well try a couple of the more exotic targets before putting this into the queue. r=me with a pass.

@bors try

Test linking and running `no_std` binaries I looked around, but it seems that we do not have a test that tests a `#![no_std]` + `#![no_main]` binary. So now I've added one. Motivated by discussion in [this Zulip thread](https://rust-lang.zulipchat.com/#narrow/channel/219381-t-libs/topic/Provide.20.60__isPlatformVersionAtLeast.60.20in.20.60std.60.3F/with/507870028). r? `@tgross35` try-job: arm-android try-job: armhf-gnu try-job: dist-ohos try-job: x86_64-mingw-1

bors · 2025-04-08T02:11:40Z

⌛ Trying commit f9770e7 with merge b40a57b...

bors · 2025-04-08T08:11:40Z

💥 Test timed out

jieyouxu · 2025-04-08T08:13:24Z

@bors try

Test linking and running `no_std` binaries I looked around, but it seems that we do not have a test that tests a `#![no_std]` + `#![no_main]` binary. So now I've added one. Motivated by discussion in [this Zulip thread](https://rust-lang.zulipchat.com/#narrow/channel/219381-t-libs/topic/Provide.20.60__isPlatformVersionAtLeast.60.20in.20.60std.60.3F/with/507870028). r? `@tgross35` try-job: arm-android try-job: armhf-gnu try-job: dist-ohos try-job: x86_64-mingw-1

bors · 2025-04-08T08:14:36Z

⌛ Trying commit f9770e7 with merge 893fcd4...

bors · 2025-04-08T11:21:10Z

☀️ Try build successful - checks-actions
Build commit: 893fcd4 (893fcd4da31eb770555aa598b389e89901312482)

tgross35 · 2025-04-08T16:06:36Z

@bors r+

bors · 2025-04-08T16:06:38Z

📌 Commit f9770e7 has been approved by tgross35

It is now in the queue for this repository.

Test linking and running `no_std` binaries I looked around, but it seems that we do not have a test that tests a `#![no_std]` + `#![no_main]` binary. So now I've added one. Motivated by discussion in [this Zulip thread](https://rust-lang.zulipchat.com/#narrow/channel/219381-t-libs/topic/Provide.20.60__isPlatformVersionAtLeast.60.20in.20.60std.60.3F/with/507870028). r? `@tgross35` try-job: arm-android try-job: armhf-gnu try-job: dist-ohos try-job: x86_64-mingw-1

Rollup of 18 pull requests Successful merges: - rust-lang#137412 (Ensure `swap_nonoverlapping` is really always untyped) - rust-lang#138167 (Small code improvement in rustdoc hidden stripper) - rust-lang#138605 (Clean up librustdoc::html::render to be better encapsulated) - rust-lang#138682 (Allow drivers to supply a list of extra symbols to intern) - rust-lang#138904 (Test linking and running `no_std` binaries) - rust-lang#139423 (Suppress missing field error when autoderef bottoms out in infer) - rust-lang#139449 (match ergonomics: replace `peel_off_references` with a recursive call) - rust-lang#139507 (compiletest: Trim whitespace from environment variable names) - rust-lang#139530 (Remove some dead or leftover code related to rustc-intrinsic abi removal) - rust-lang#139560 (fix title of offset_of_enum feature) - rust-lang#139563 (emit a better error message for using the macro incorrectly) - rust-lang#139568 (Don't use empty trait names) - rust-lang#139580 (Temporarily leave the review rotation) - rust-lang#139589 (saethlin is back from vacation) - rust-lang#139592 (rustdoc: Enable Markdown extensions when looking for doctests) - rust-lang#139599 (Tracking issue template: fine-grained information on style update status) - rust-lang#139600 (Update `compiler-builtins` to 0.1.153) - rust-lang#139606 (Update compiletest to Edition 2024) r? `@ghost` `@rustbot` modify labels: rollup

Test linking and running `no_std` binaries I looked around, but it seems that we do not have a test that tests a `#![no_std]` + `#![no_main]` binary. So now I've added one. Motivated by discussion in [this Zulip thread](https://rust-lang.zulipchat.com/#narrow/channel/219381-t-libs/topic/Provide.20.60__isPlatformVersionAtLeast.60.20in.20.60std.60.3F/with/507870028). r? ``@tgross35`` try-job: arm-android try-job: armhf-gnu try-job: dist-ohos try-job: x86_64-mingw-1

Rollup of 17 pull requests Successful merges: - rust-lang#137412 (Ensure `swap_nonoverlapping` is really always untyped) - rust-lang#138167 (Small code improvement in rustdoc hidden stripper) - rust-lang#138605 (Clean up librustdoc::html::render to be better encapsulated) - rust-lang#138682 (Allow drivers to supply a list of extra symbols to intern) - rust-lang#138904 (Test linking and running `no_std` binaries) - rust-lang#139423 (Suppress missing field error when autoderef bottoms out in infer) - rust-lang#139449 (match ergonomics: replace `peel_off_references` with a recursive call) - rust-lang#139507 (compiletest: Trim whitespace from environment variable names) - rust-lang#139530 (Remove some dead or leftover code related to rustc-intrinsic abi removal) - rust-lang#139560 (fix title of offset_of_enum feature) - rust-lang#139563 (emit a better error message for using the macro incorrectly) - rust-lang#139568 (Don't use empty trait names) - rust-lang#139580 (Temporarily leave the review rotation) - rust-lang#139589 (saethlin is back from vacation) - rust-lang#139592 (rustdoc: Enable Markdown extensions when looking for doctests) - rust-lang#139599 (Tracking issue template: fine-grained information on style update status) - rust-lang#139600 (Update `compiler-builtins` to 0.1.153) r? `@ghost` `@rustbot` modify labels: rollup

Rollup of 12 pull requests Successful merges: - rust-lang#137447 (add `core::intrinsics::simd::{simd_extract_dyn, simd_insert_dyn}`) - rust-lang#138182 (rustc_target: update x86_win64 to match the documented calling convention for f128) - rust-lang#138682 (Allow drivers to supply a list of extra symbols to intern) - rust-lang#138904 (Test linking and running `no_std` binaries) - rust-lang#138998 (Don't suggest the use of `impl Trait` in closure parameter) - rust-lang#139447 (doc changes: debug assertions -> overflow checks) - rust-lang#139469 (Introduce a `//@ needs-crate-type` compiletest directive) - rust-lang#139564 (Deeply normalize obligations in `BestObligation` folder) - rust-lang#139574 (bootstrap: improve `channel` handling) - rust-lang#139600 (Update `compiler-builtins` to 0.1.153) - rust-lang#139641 (Allow parenthesis around inferred array lengths) - rust-lang#139654 (Improve `AssocItem::descr`.) r? `@ghost` `@rustbot` modify labels: rollup

Rollup merge of rust-lang#138904 - madsmtm:apple-test-no-std, r=tgross35 Test linking and running `no_std` binaries I looked around, but it seems that we do not have a test that tests a `#![no_std]` + `#![no_main]` binary. So now I've added one. Motivated by discussion in [this Zulip thread](https://rust-lang.zulipchat.com/#narrow/channel/219381-t-libs/topic/Provide.20.60__isPlatformVersionAtLeast.60.20in.20.60std.60.3F/with/507870028). r? ```@tgross35``` try-job: arm-android try-job: armhf-gnu try-job: dist-ohos try-job: x86_64-mingw-1

rustbot assigned tgross35 Mar 24, 2025

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels Mar 24, 2025

madsmtm commented Mar 24, 2025

View reviewed changes

This comment has been minimized.

Sign in to view

Test linking and running no_std binaries

f9770e7

madsmtm force-pushed the apple-test-no-std branch from e74978d to f9770e7 Compare March 24, 2025 22:22

Noratrieb reviewed Mar 25, 2025

View reviewed changes

This comment was marked as outdated.

Sign in to view

tgross35 approved these changes Apr 8, 2025

View reviewed changes

bors mentioned this pull request Apr 8, 2025

Rollup of 19 pull requests #139508

Closed

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 Apr 8, 2025

Zalathar mentioned this pull request Apr 10, 2025

Rollup of 18 pull requests #139612

Closed

Zalathar mentioned this pull request Apr 10, 2025

Rollup of 17 pull requests #139613

Closed

Zalathar mentioned this pull request Apr 11, 2025

Rollup of 12 pull requests #139657

Merged

bors merged commit 9a9a078 into rust-lang:master Apr 11, 2025
7 checks passed

rustbot added this to the 1.88.0 milestone Apr 11, 2025

madsmtm deleted the apple-test-no-std branch April 11, 2025 14:43

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Test linking and running `no_std` binaries #138904

Test linking and running `no_std` binaries #138904

madsmtm commented Mar 24, 2025 •

edited by tgross35

Loading

madsmtm Mar 24, 2025

folkertdev commented Mar 24, 2025

madsmtm Mar 24, 2025

tgross35 Apr 4, 2025

jieyouxu Apr 4, 2025 •

edited

Loading

madsmtm commented Mar 24, 2025

This comment has been minimized.

folkertdev commented Mar 24, 2025 •

edited

Loading

Noratrieb Mar 25, 2025

tgross35 commented Apr 4, 2025

This comment was marked as outdated.

This comment was marked as outdated.

ChrisDenton commented Apr 4, 2025

tgross35 commented Apr 4, 2025

bors commented Apr 4, 2025

bors commented Apr 4, 2025

tgross35 commented Apr 8, 2025

bors commented Apr 8, 2025

bors commented Apr 8, 2025

jieyouxu commented Apr 8, 2025

bors commented Apr 8, 2025

bors commented Apr 8, 2025

tgross35 commented Apr 8, 2025

bors commented Apr 8, 2025

Test linking and running no_std binaries #138904

Test linking and running no_std binaries #138904

Conversation

madsmtm commented Mar 24, 2025 • edited by tgross35 Loading

madsmtm Mar 24, 2025

Choose a reason for hiding this comment

folkertdev commented Mar 24, 2025

madsmtm Mar 24, 2025

Choose a reason for hiding this comment

tgross35 Apr 4, 2025

Choose a reason for hiding this comment

jieyouxu Apr 4, 2025 • edited Loading

Choose a reason for hiding this comment

madsmtm commented Mar 24, 2025

This comment has been minimized.

folkertdev commented Mar 24, 2025 • edited Loading

Noratrieb Mar 25, 2025

Choose a reason for hiding this comment

tgross35 commented Apr 4, 2025

This comment was marked as outdated.

This comment was marked as outdated.

ChrisDenton commented Apr 4, 2025

tgross35 commented Apr 4, 2025

bors commented Apr 4, 2025

bors commented Apr 4, 2025

tgross35 commented Apr 8, 2025

bors commented Apr 8, 2025

bors commented Apr 8, 2025

jieyouxu commented Apr 8, 2025

bors commented Apr 8, 2025

bors commented Apr 8, 2025

tgross35 commented Apr 8, 2025

bors commented Apr 8, 2025

Test linking and running `no_std` binaries #138904

Test linking and running `no_std` binaries #138904

madsmtm commented Mar 24, 2025 •

edited by tgross35

Loading

jieyouxu Apr 4, 2025 •

edited

Loading

folkertdev commented Mar 24, 2025 •

edited

Loading