Add minimal x86_64-lynx-lynxos178 support. #138011

tnewsome-lynx · 2025-03-04T17:13:19Z

Add minimal x86_64-lynx-lynxos178 support. It's possible to build no_std
programs with this compiler.

Tier 3 Target Policy

A tier 3 target must have a designated developer or developers (the "target
maintainers") on record to be CCed when issues arise regarding the target. (The
mechanism to track and CC such developers may evolve over time.)

Tim Newsome (@tnewsome-lynx) will be the designated developer for
x86_64-lynx-lynxos178 support.

Targets must use naming consistent with any existing targets; for instance, a
target for the same CPU or OS as an existing Rust target should use the same
name for that CPU or OS. Targets should normally use the same names and naming
conventions as used elsewhere in the broader ecosystem beyond Rust (such as in
other toolchains), unless they have a very good reason to diverge. Changing the
name of a target can be highly disruptive, especially once the target reaches a
higher tier, so getting the name right is important even for a tier 3 target.

I believe the target is named appropriately.

Target names should not introduce undue confusion or ambiguity unless
absolutely necessary to maintain ecosystem compatibility. For example, if the
name of the target makes people extremely likely to form incorrect beliefs about
what it targets, the name should be changed or augmented to disambiguate it.

The target name is not confusing.

If possible, use only letters, numbers, dashes and underscores for the name.
Periods (.) are known to cause issues in Cargo.

Done.

Tier 3 targets may have unusual requirements to build or use, but must not
create legal issues or impose onerous legal terms for the Rust project or for
Rust developers or users.
The target must not introduce license incompatibilities.
Anything added to the Rust repository must be under the standard Rust license
(MIT OR Apache-2.0).

All this new code is licensed under the Apache-2.0 license.

The target must not cause the Rust tools or libraries built for any other host
(even when supporting cross-compilation to the target) to depend on any new
dependency less permissive than the Rust licensing policy. This applies whether
the dependency is a Rust crate that would require adding new license exceptions
(as specified by the tidy tool in the rust-lang/rust repository), or whether the
dependency is a native library or binary. In other words, the introduction of
the target must not cause a user installing or running a version of Rust or the
Rust tools to be subject to any new license requirements.

Done.

Compiling, linking, and emitting functional binaries, libraries, or other code
for the target (whether hosted on the target itself or cross-compiling from
another target) must not depend on proprietary (non-FOSS) libraries. Host tools
built for the target itself may depend on the ordinary runtime libraries
supplied by the platform and commonly used by other applications built for the
target, but those libraries must not be required for code generation for the
target; cross-compilation to the target must not require such libraries at all.
For instance, rustc built for the target may depend on a common proprietary C
runtime library or console output library, but must not depend on a proprietary
code generation library or code optimization library. Rust's license permits
such combinations, but the Rust project has no interest in maintaining such
combinations within the scope of Rust itself, even at tier 3.

I think we're in the clear here. We do link against some static libraries that
are proprietary (like libm and libc), but those are not used to generate code.
E.g. the VxWorks target requires wr-c++ to be installed, which is not
publically available.

"onerous" here is an intentionally subjective term. At a minimum, "onerous"
legal/licensing terms include but are not limited to: non-disclosure
requirements, non-compete requirements, contributor license agreements (CLAs) or
equivalent, "non-commercial"/"research-only"/etc terms, requirements conditional
on the employer or employment of any particular Rust developers, revocable
terms, any requirements that create liability for the Rust project or its
developers or users, or any requirements that adversely affect the livelihood or
prospects of the Rust project or its developers or users.

Our intention is to allow anyone with access to LynxOS CDK to use Rust for it.

Neither this policy nor any decisions made regarding targets shall create any
binding agreement or estoppel by any party. If any member of an approving Rust
team serves as one of the maintainers of a target, or has any legal or
employment requirement (explicit or implicit) that might affect their decisions
regarding a target, they must recuse themselves from any approval decisions
regarding the target's tier status, though they may otherwise participate in
discussions.
This requirement does not prevent part or all of this policy from being cited
in an explicit contract or work agreement (e.g. to implement or maintain support
for a target). This requirement exists to ensure that a developer or team
responsible for reviewing and approving a target does not face any legal threats
or obligations that would prevent them from freely exercising their judgment in
such approval, even if such judgment involves subjective matters or goes beyond
the letter of these requirements.

No problem.

Tier 3 targets should attempt to implement as much of the standard libraries
as possible and appropriate (core for most targets, alloc for targets that can
support dynamic memory allocation, std for targets with an operating system or
equivalent layer of system-provided functionality), but may leave some code
unimplemented (either unavailable or stubbed out as appropriate), whether
because the target makes it impossible to implement or challenging to implement.
The authors of pull requests are not obligated to avoid calling any portions of
the standard library on the basis of a tier 3 target not implementing those
portions.

With this first PR, only core is supported. I am working on support for the std
library and intend to submit that once all the tests are passing.

The target must provide documentation for the Rust community explaining how to
build for the target, using cross-compilation if possible. If the target
supports running binaries, or running tests (even if they do not pass), the
documentation must explain how to run such binaries or tests for the target,
using emulation if possible or dedicated hardware if necessary.

This is documented in src/doc/rustc/src/platform-support/lynxos178.md.

Tier 3 targets must not impose burden on the authors of pull requests, or
other developers in the community, to maintain the target. In particular, do not
post comments (automated or manual) on a PR that derail or suggest a block on
the PR based on a tier 3 target. Do not send automated messages or notifications
(via any medium, including via @) to a PR author or others involved with a PR
regarding a tier 3 target, unless they have opted into such messages.
Backlinks such as those generated by the issue/PR tracker when linking to an
issue or PR are not considered a violation of this policy, within reason.
However, such messages (even on a separate repository) must not generate
notifications to anyone involved with a PR who has not requested such
notifications.

Understood.

Patches adding or updating tier 3 targets must not break any existing tier 2
or tier 1 target, and must not knowingly break another tier 3 target without
approval of either the compiler team or the maintainers of the other tier 3
target.
In particular, this may come up when working on closely related targets, such
as variations of the same architecture with different features. Avoid
introducing unconditional uses of features that another variation of the target
may not have; use conditional compilation or runtime detection, as appropriate,
to let each target run code supported by that target.

As far as I know this change does not affect any other targets.

Tier 3 targets must be able to produce assembly using at least one of rustc's
supported backends from any host target. (Having support in a fork of the
backend is not sufficient, it must be upstream.)

Many targets produce assembly for x86_64 so that also works for LynxOS-178.

rustbot · 2025-03-04T17:13:25Z

r? @nnethercote

rustbot has assigned @nnethercote.
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

rustbot · 2025-03-04T17:13:27Z

These commits modify compiler targets.
(See the Target Tier Policy.)

Some changes occurred in src/doc/rustc/src/platform-support

cc @Noratrieb

Noratrieb

For the target name, x86_64-unknown-lynxos_178-elf is fine, but is there any reason to have the elf in the end? It seems unnecessary, unless LynxOS also supports other binaries (we could also write lynx in the vendor field but thats not very important).

compiler/rustc_target/src/spec/targets/x86_64_unknown_lynxos_178_elf.rs

src/doc/rustc/src/platform-support/lynxos_178.md

compiler/rustc_target/src/spec/base/lynxos_178.rs

Noratrieb · 2025-03-04T18:13:17Z

compiler/rustc_target/src/spec/base/lynxos_178.rs

+pub(crate) fn opts() -> TargetOptions {
+    let linker_flavor = LinkerFlavor::Gnu(Cc::No, Lld::No);
+    let other_linker_flavor = LinkerFlavor::Gnu(Cc::No, Lld::Yes);
+    let env_prefix = env::var("ENV_PREFIX").unwrap_or("ENV_PREFIX".into());


Requiring ENV_PREFIX to be set won't work when just loading targets without the intention of compiling got LynxOS. You either need a dummy fallback or preferably not have it at all.
How is your toolchain expected to be used that introduces these requirements? I assume you have the same env var for C code already today.

There is a dummy fallback (ENV_PREFIX in this case). Although maybe I should use "<ENV_PREFIX not set>".

The way C development works is you source setup.sh which sets ENV_PREFIX and GCC_DIR. Then various Makefiles use those environment variables to execute the right compiler and find the right libraries.

Right, I overlooked that. This is a very unconventional approach that isn't used in any other place here, but it seems unavoidable here, so it would be nice if we could avoid this whole situtation^^, but if there's no other way we'll have to do it I guess (other than requiring the user to set -Clinker as mentioned by Björn in the other comment thread).

tnewsome-lynx · 2025-03-04T20:35:23Z

I'll be looking more into the CI failure, but locally it's passing:

> ./x test --target x86_64-unknown-linux-gnu assembly
...
test result: ok. 499 passed; 0 failed; 36 ignored; 0 measured; 0 filtered out; finished in 3.74s

tnewsome-lynx · 2025-03-04T20:36:50Z

@Noratrieb Thanks for the review. Do you prefer I keep the history clean and force push, or do you prefer I make new git commits for your feedback?

Noratrieb · 2025-03-04T21:21:03Z

We want to squash it together before merging, but before that it's up to you. Force pushing is fine.

jieyouxu · 2025-03-05T00:12:43Z

While still ironing out the review process... https://forge.rust-lang.org/compiler/proposals-and-stabilization.html#targets now says adding new target needs r+ from compiler-leads.

So I guess cc @davidtwco and @wesleywiser for this proposal.

bors · 2025-03-11T11:31:50Z

☔ The latest upstream changes (presumably #135651) made this pull request unmergeable. Please resolve the merge conflicts.

tnewsome-lynx · 2025-03-11T19:13:03Z

☔ The latest upstream changes (presumably #135651) made this pull request unmergeable. Please resolve the merge conflicts.

Rebased. Thanks.

jieyouxu · 2025-03-11T22:30:14Z

r? compiler_leads

bjorn3 · 2025-03-12T10:07:07Z

compiler/rustc_target/src/spec/base/lynxos_178.rs

+        crt_static_respected: true,
+        panic_strategy: PanicStrategy::Abort,
+        // Don't rely on the path to find the correct ld.
+        linker: Some(cow_format!("{cdk_path}/bin/ld")),


Should this perhaps use gcc as linker driver instead? That should allow avoiding all (or at least most) pre and post link args. And instead of reading ENV_PREFIX and GCC_DIR you can have the user pass -Clinker=/path/to/gcc like we normally require for cross-compiling. Though I don't have a strong opinion on the latter.

I'm not familiar enough with rustc to know what you mean by using "gcc as linker driver." I guess that's different from the linker_flavor TargetOption?

Use of ENV_PREFIX and GCC_DIR is baked pretty far into the LynxOS-178 development workflow. There's no way ld can find the libraries in ENV_PREFIX without it being set, and the same script that sets ENV_PREFIX also sets GCC_DIR.

The way we generally do linkage on unix-like platforms is by invoking cc with our libraries and object files. This will ensure that all the C standard libraries are found and we don't have to worry about them ourselves.

I assume that using cc (or some other GCC binary name) on LynxOS-178 still requires setting these environment variables? Or will that magically pick everything up?

I see... I'll experiment with calling gcc instead of ld. It does seem to magically pick up the right thing, and I'm not sure what that magic depends on.

Changing to using gcc makes the arguments much simpler, as @bjorn3 mentioned. I think I can get away without relying on ENV_PREFIX at all, because the LynxOS-178 setup also changes the PATH. This is good because now we can find the LynxOS-178 linker, and bad because it masks the linker for the host. I think I know how to fix this, but it's going to take a bit more talking to people and experimenting.

But I am still working on this PR, and hopefully will have something simpler for you all next week.

bjorn3 · 2025-03-12T10:07:40Z

compiler/rustc_target/src/spec/base/lynxos_178.rs

+macro_rules! cow_format {
+    ($arg:expr) => {
+        // Cow::from works with &str but not with &mut str.
+        Cow::from(immutable(Box::leak(format!($arg).into_boxed_str())))


Suggested change

Cow::from(immutable(Box::leak(format!($arg).into_boxed_str())))

Cow::Borrowed(Box::leak(format!($arg).into_boxed_str()))

Done. Thanks.

bjorn3 · 2025-03-12T10:24:52Z

The PR title says x86_64-lynxos_178-unknown-elf, but the source code says x86_64-unknown-lynxos_178-elf. I presume the latter is correct?

tnewsome-lynx · 2025-03-12T15:35:20Z

The PR title says x86_64-lynxos_178-unknown-elf, but the source code says x86_64-unknown-lynxos_178-elf. I presume the latter is correct?

Yes. Fixing that. Good eye.

davidtwco

r=me with a rebase

@tnewsome-lynx

It's possible to build no_std programs with this compiler. > A tier 3 target must have a designated developer or developers (the "target maintainers") on record to be CCed when issues arise regarding the target. (The mechanism to track and CC such developers may evolve over time.) Tim Newsome (@tnewsome-lynx) will be the designated developer for x86_64-lynx-lynxos178 support. > Targets must use naming consistent with any existing targets; for instance, a target for the same CPU or OS as an existing Rust target should use the same name for that CPU or OS. Targets should normally use the same names and naming conventions as used elsewhere in the broader ecosystem beyond Rust (such as in other toolchains), unless they have a very good reason to diverge. Changing the name of a target can be highly disruptive, especially once the target reaches a higher tier, so getting the name right is important even for a tier 3 target. I believe the target is named appropriately. > Target names should not introduce undue confusion or ambiguity unless absolutely necessary to maintain ecosystem compatibility. For example, if the name of the target makes people extremely likely to form incorrect beliefs about what it targets, the name should be changed or augmented to disambiguate it. The target name is not confusing. > If possible, use only letters, numbers, dashes and underscores for the name. Periods (.) are known to cause issues in Cargo. Done. > Tier 3 targets may have unusual requirements to build or use, but must not create legal issues or impose onerous legal terms for the Rust project or for Rust developers or users. > The target must not introduce license incompatibilities. > Anything added to the Rust repository must be under the standard Rust license (MIT OR Apache-2.0). All this new code is licensed under the Apache-2.0 license. > The target must not cause the Rust tools or libraries built for any other host (even when supporting cross-compilation to the target) to depend on any new dependency less permissive than the Rust licensing policy. This applies whether the dependency is a Rust crate that would require adding new license exceptions (as specified by the tidy tool in the rust-lang/rust repository), or whether the dependency is a native library or binary. In other words, the introduction of the target must not cause a user installing or running a version of Rust or the Rust tools to be subject to any new license requirements. Done. > Compiling, linking, and emitting functional binaries, libraries, or other code for the target (whether hosted on the target itself or cross-compiling from another target) must not depend on proprietary (non-FOSS) libraries. Host tools built for the target itself may depend on the ordinary runtime libraries supplied by the platform and commonly used by other applications built for the target, but those libraries must not be required for code generation for the target; cross-compilation to the target must not require such libraries at all. For instance, rustc built for the target may depend on a common proprietary C runtime library or console output library, but must not depend on a proprietary code generation library or code optimization library. Rust's license permits such combinations, but the Rust project has no interest in maintaining such combinations within the scope of Rust itself, even at tier 3. I think we're in the clear here. We do link against some static libraries that are proprietary (like libm and libc), but those are not used to generate code. E.g. the VxWorks target requires `wr-c++` to be installed, which is not publically available. > "onerous" here is an intentionally subjective term. At a minimum, "onerous" legal/licensing terms include but are not limited to: non-disclosure requirements, non-compete requirements, contributor license agreements (CLAs) or equivalent, "non-commercial"/"research-only"/etc terms, requirements conditional on the employer or employment of any particular Rust developers, revocable terms, any requirements that create liability for the Rust project or its developers or users, or any requirements that adversely affect the livelihood or prospects of the Rust project or its developers or users. Our intention is to allow anyone with access to LynxOS CDK to use Rust for it. > Neither this policy nor any decisions made regarding targets shall create any binding agreement or estoppel by any party. If any member of an approving Rust team serves as one of the maintainers of a target, or has any legal or employment requirement (explicit or implicit) that might affect their decisions regarding a target, they must recuse themselves from any approval decisions regarding the target's tier status, though they may otherwise participate in discussions. > This requirement does not prevent part or all of this policy from being cited in an explicit contract or work agreement (e.g. to implement or maintain support for a target). This requirement exists to ensure that a developer or team responsible for reviewing and approving a target does not face any legal threats or obligations that would prevent them from freely exercising their judgment in such approval, even if such judgment involves subjective matters or goes beyond the letter of these requirements. No problem. > Tier 3 targets should attempt to implement as much of the standard libraries as possible and appropriate (core for most targets, alloc for targets that can support dynamic memory allocation, std for targets with an operating system or equivalent layer of system-provided functionality), but may leave some code unimplemented (either unavailable or stubbed out as appropriate), whether because the target makes it impossible to implement or challenging to implement. The authors of pull requests are not obligated to avoid calling any portions of the standard library on the basis of a tier 3 target not implementing those portions. With this first PR, only core is supported. I am working on support for the std library and intend to submit that once all the tests are passing. > The target must provide documentation for the Rust community explaining how to build for the target, using cross-compilation if possible. If the target supports running binaries, or running tests (even if they do not pass), the documentation must explain how to run such binaries or tests for the target, using emulation if possible or dedicated hardware if necessary. This is documented in `src/doc/rustc/src/platform-support/lynxos_178.md`. > Tier 3 targets must not impose burden on the authors of pull requests, or other developers in the community, to maintain the target. In particular, do not post comments (automated or manual) on a PR that derail or suggest a block on the PR based on a tier 3 target. Do not send automated messages or notifications (via any medium, including via @) to a PR author or others involved with a PR regarding a tier 3 target, unless they have opted into such messages. > Backlinks such as those generated by the issue/PR tracker when linking to an issue or PR are not considered a violation of this policy, within reason. However, such messages (even on a separate repository) must not generate notifications to anyone involved with a PR who has not requested such notifications. Understood. > Patches adding or updating tier 3 targets must not break any existing tier 2 or tier 1 target, and must not knowingly break another tier 3 target without approval of either the compiler team or the maintainers of the other tier 3 target. > In particular, this may come up when working on closely related targets, such as variations of the same architecture with different features. Avoid introducing unconditional uses of features that another variation of the target may not have; use conditional compilation or runtime detection, as appropriate, to let each target run code supported by that target. As far as I know this change does not affect any other targets. > Tier 3 targets must be able to produce assembly using at least one of rustc's supported backends from any host target. (Having support in a fork of the backend is not sufficient, it must be upstream.) Many targets produce assembly for x86_64 so that also works for LynxOS-178.

tnewsome-lynx · 2025-03-24T16:47:16Z

There are a few changes over the last couple of weeks. The main one is that I've changed what Lynx tools are available to rustc, so the rustc code is not a lot simpler than it used to be. As part of this, I also replaced "lynxos_178" with "lynxos178" everywhere to match the triple we use when building gcc for LynxOS-178.

tnewsome-lynx · 2025-03-31T16:25:01Z

@Noratrieb Can you take a look at this again? I think it's good to go now.

rustbot assigned nnethercote Mar 4, 2025

tnewsome-lynx force-pushed the lynxos_178-nostd branch from cd3db29 to 4777c64 Compare March 4, 2025 17:17

Noratrieb requested changes Mar 4, 2025

View reviewed changes

Noratrieb assigned Noratrieb and unassigned nnethercote Mar 4, 2025

tnewsome-lynx force-pushed the lynxos_178-nostd branch 3 times, most recently from 6982378 to 2d6c4fd Compare March 6, 2025 22:00

tnewsome-lynx requested a review from Noratrieb March 6, 2025 22:01

tnewsome-lynx force-pushed the lynxos_178-nostd branch from 2d6c4fd to 4a066a0 Compare March 11, 2025 19:12

rustbot assigned davidtwco and unassigned Noratrieb Mar 11, 2025

bjorn3 reviewed Mar 12, 2025

View reviewed changes

tnewsome-lynx force-pushed the lynxos_178-nostd branch from 4a066a0 to 04130a9 Compare March 12, 2025 15:35

tnewsome-lynx changed the title ~~Add minimal x86_64-lynxos_178-unknown-elf support.~~ Add minimal x86_64-unknown-lynxos_178-elf support. Mar 12, 2025

tnewsome-lynx force-pushed the lynxos_178-nostd branch from 04130a9 to 7f56516 Compare March 12, 2025 16:08

tnewsome-lynx force-pushed the lynxos_178-nostd branch from 7f56516 to 3b74771 Compare March 21, 2025 16:17

tnewsome-lynx changed the title ~~Add minimal x86_64-unknown-lynxos_178-elf support.~~ Add minimal x86_64-lynx-lynxos178 support. Mar 21, 2025

tnewsome-lynx marked this pull request as draft March 21, 2025 16:24

davidtwco approved these changes Mar 24, 2025

View reviewed changes

tnewsome-lynx force-pushed the lynxos_178-nostd branch from 3b74771 to f2c72cd Compare March 24, 2025 16:40

tnewsome-lynx marked this pull request as ready for review March 24, 2025 16:41

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add minimal x86_64-lynx-lynxos178 support. #138011

Add minimal x86_64-lynx-lynxos178 support. #138011

tnewsome-lynx commented Mar 4, 2025 •

edited

Loading

rustbot commented Mar 4, 2025

rustbot commented Mar 4, 2025

Noratrieb left a comment

Noratrieb Mar 4, 2025

tnewsome-lynx Mar 4, 2025

Noratrieb Mar 12, 2025 •

edited

Loading

tnewsome-lynx commented Mar 4, 2025

tnewsome-lynx commented Mar 4, 2025

Noratrieb commented Mar 4, 2025

jieyouxu commented Mar 5, 2025

bors commented Mar 11, 2025

tnewsome-lynx commented Mar 11, 2025

jieyouxu commented Mar 11, 2025

bjorn3 Mar 12, 2025 •

edited

Loading

tnewsome-lynx Mar 12, 2025

Noratrieb Mar 12, 2025

tnewsome-lynx Mar 12, 2025

tnewsome-lynx Mar 14, 2025

bjorn3 Mar 12, 2025

tnewsome-lynx Mar 12, 2025

bjorn3 commented Mar 12, 2025

tnewsome-lynx commented Mar 12, 2025

davidtwco left a comment

tnewsome-lynx commented Mar 24, 2025

tnewsome-lynx commented Mar 31, 2025

	Cow::from(immutable(Box::leak(format!($arg).into_boxed_str())))
	Cow::Borrowed(Box::leak(format!($arg).into_boxed_str()))

Add minimal x86_64-lynx-lynxos178 support. #138011

Are you sure you want to change the base?

Add minimal x86_64-lynx-lynxos178 support. #138011

Conversation

tnewsome-lynx commented Mar 4, 2025 • edited Loading

Tier 3 Target Policy

rustbot commented Mar 4, 2025

rustbot commented Mar 4, 2025

Noratrieb left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Noratrieb Mar 12, 2025 • edited Loading

Choose a reason for hiding this comment

tnewsome-lynx commented Mar 4, 2025

tnewsome-lynx commented Mar 4, 2025

Noratrieb commented Mar 4, 2025

jieyouxu commented Mar 5, 2025

bors commented Mar 11, 2025

tnewsome-lynx commented Mar 11, 2025

jieyouxu commented Mar 11, 2025

bjorn3 Mar 12, 2025 • edited Loading

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

bjorn3 commented Mar 12, 2025

tnewsome-lynx commented Mar 12, 2025

davidtwco left a comment

Choose a reason for hiding this comment

tnewsome-lynx commented Mar 24, 2025

tnewsome-lynx commented Mar 31, 2025

tnewsome-lynx commented Mar 4, 2025 •

edited

Loading

Noratrieb Mar 12, 2025 •

edited

Loading

bjorn3 Mar 12, 2025 •

edited

Loading