Fix RISC-V Test Failures in ./x test for Multiple Codegen and Assembly Cases #143915

CaiWeiran · 2025-07-14T03:49:50Z

Resumption of #143792.
This PR addresses and resolves several test failures that occurred when running ./x test on the RISC-V architecture. The issues were due to platform-specific behavior, ABI differences, or code generation inconsistencies specific to RISC-V.

The following tests have been fixed:

assembly/dwarf-mixed-versions-lto.rs
codegen/const-vector.rs
codegen/enum/enum-aggregate.rs
codegen/simd/extract-insert-dyn.rs
codegen/transmute-scalar.rs
codegen/uninhabited-transparent-return-abi.rs

All changes have been tested locally with ./x test on a RISC-V target and now pass as expected.

Notes:

These fixes are scoped specifically to enable full test suite compliance for RISC-V targets.
No changes were made that affect other architectures.

Fix tests/assembly/dwarf-mixed-versions-lto.rs test failure on riscv64.

Fix tests/codegen/const-vector.rs test failure on riscv64.

Fix tests/codegen/enum/enum-aggregate.rs test failure on riscv64.

Fix tests/codegen/simd/extract-insert-dyn.rs test failure on riscv64.

Fix tests/codegen/transmute-scalar.rs test failure on riscv64.

Fix tests/codegen/uninhabited-transparent-return-abi.rs test failure on riscv64.

rustbot · 2025-07-14T03:49:54Z

r? @Mark-Simulacrum

rustbot has assigned @Mark-Simulacrum.
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

CaiWeiran · 2025-07-14T03:51:50Z

@Mark-Simulacrum I have merged different revisions into a single file.

Fix tidy check error.

Fix regex matching error.

tests/codegen/enum/enum-aggregate.rs

rustbot · 2025-07-15T19:52:27Z

Error: This team (llvm) cannot be pinged via this command; it may need to be added to triagebot.toml on the default branch.

Please file an issue on GitHub at triagebot if there's a problem with this bot, or reach out on #t-infra on Zulip.

Mark-Simulacrum · 2025-07-15T19:53:50Z

cc @nikic for whether there's a better pattern for us to follow in codegen tests to normalize/ignore the per-target ABI questions (E.g., signext above), or if we should instead retain test coverage very explicitly of them.

nikic · 2025-07-15T20:03:03Z

@Mark-Simulacrum Ignoring ABI attributes via {{.*}} seems fine. But I think that for tests like transmute-scalar.rs that aren't testing anything target specific, we should be specifying an explicit --target to avoid these issues. I feel like that might have been the intention for that test all along, seeing as how it uses minicore already.

CaiWeiran · 2025-07-16T01:17:37Z

@Mark-Simulacrum Ignoring ABI attributes via {{.*}} seems fine. But I think that for tests like transmute-scalar.rs that aren't testing anything target specific, we should be specifying an explicit --target to avoid these issues. I feel like that might have been the intention for that test all along, seeing as how it uses minicore already.

@nikic I'm not sure I understand. If I don't make any changes, running ./x test --target riscv64gc-unknown-linux-gnu tests/codegen/enum/enum-aggregate.rs still fails. So even when specifying --target, the test fails due to target-specific ABI differences.

nikic · 2025-07-16T07:59:37Z

@Mark-Simulacrum Ignoring ABI attributes via {{.*}} seems fine. But I think that for tests like transmute-scalar.rs that aren't testing anything target specific, we should be specifying an explicit --target to avoid these issues. I feel like that might have been the intention for that test all along, seeing as how it uses minicore already.

@nikic I'm not sure I understand. If I don't make any changes, running ./x test --target riscv64gc-unknown-linux-gnu tests/codegen/enum/enum-aggregate.rs still fails. So even when specifying --target, the test fails due to target-specific ABI differences.

The test should specify

//@ compile-flags: --target=x86_64-unknown-linux-gnu
//@ needs-llvm-components: x86

To always run against the x86 triple even if you are running on a riscv host.

CaiWeiran · 2025-07-16T08:32:16Z

@Mark-Simulacrum Ignoring ABI attributes via {{.*}} seems fine. But I think that for tests like transmute-scalar.rs that aren't testing anything target specific, we should be specifying an explicit --target to avoid these issues. I feel like that might have been the intention for that test all along, seeing as how it uses minicore already.

@nikic I'm not sure I understand. If I don't make any changes, running ./x test --target riscv64gc-unknown-linux-gnu tests/codegen/enum/enum-aggregate.rs still fails. So even when specifying --target, the test fails due to target-specific ABI differences.

The test should specify
//@ compile-flags: --target=x86_64-unknown-linux-gnu
//@ needs-llvm-components: x86
To always run against the x86 triple even if you are running on a riscv host.

@nikic It will fail the test if I don't install the x86_64-unknown-linux-gnu target on the RISC-V host.
When running tests on a RISC-V host, my goal is to test the RISC-V triple.
If a certain test is specifically for x86, would it be better to mark it with //@ only-x86?
Or alternatively, modify the matching rules like I did to adapt to different ISAs?

nikic · 2025-07-16T09:07:19Z

@CaiWeiran Using --target + needs-llvm-components + no_core ensures that the test can run on arbitrary hosts. It should not fail on RISCV if everything is implemented correctly.

Ignore cross-compile.

rustbot · 2025-07-18T01:09:28Z

This PR modifies run-make tests.

cc @jieyouxu

This reverts commit d464e57.

bors · 2025-07-22T23:02:23Z

☔ The latest upstream changes (presumably #144249) made this pull request unmergeable. Please resolve the merge conflicts.

Mark-Simulacrum · 2025-08-03T23:19:37Z

r=me with commits squashed, I don't think this needs more than one commit.

scottmcm · 2025-08-03T23:47:54Z

tests/codegen/transmute-scalar.rs

@@ -79,7 +80,7 @@ pub fn bool_to_fake_bool_signed(b: bool) -> FakeBoolSigned {
    unsafe { mem::transmute(b) }
 }

-// CHECK-LABEL: define{{.*}}i1 @fake_bool_signed_to_bool(i8 %b)
+// CHECK-LABEL: define{{.*}}i1 @fake_bool_signed_to_bool(i8 {{.*}}%b)


At least for things like this one that aren't about SIMD, the point of having the signature is basically just to make it easier to read the verifications and check that the argument has the expected name (as opposed to being called %0 because of the ABI or LocalKind or something).

So for things like that, I think ignoring extra flags in here is acceptable -- though admittedly slightly unusual because it's a -C no-prepopulate-passes test. (In optimized tests it's very common to have a wildcard for resiliency against LLVM adding new parameter attributes.)

Hmm, though if it's just an optional extension flag, consider using

Suggested change

// CHECK-LABEL: define{{.*}}i1 @fake_bool_signed_to_bool(i8 {{.*}}%b)

// CHECK-LABEL: define{{.*}}i1 @fake_bool_signed_to_bool(i8{{( zeroext)?}} %b)

because you know exactly what you're ignoring in this case, rather than allowing for arbitrary future attributes that might be added by the optimizer.

scottmcm · 2025-08-03T23:56:58Z

tests/codegen/transmute-scalar.rs

 pub extern "C" fn single_element_simd_to_scalar(b: S) -> i64 {
    unsafe { mem::transmute(b) }
 }

-// CHECK-LABEL: define{{.*}}<1 x i64> @scalar_to_single_element_simd(i64 %b)
+// CHECK-LABEL: define{{.*}}i64{{.*}} @scalar_to_single_element_simd(i64 %b)


I'm less sure what should be done for this case. Is it even BackendRepr::SimdVector in these cases if it's not a vector type in the ABI? Should these SIMD ones be split into another file from the scalar ones? Should it be more like {{ i64| <1 x i64>}} if those are the only two realistic possibilities for what will be here?

scottmcm · 2025-08-03T23:59:16Z

tests/codegen/simd/extract-insert-dyn.rs

 unsafe extern "C" fn dyn_simd_insert(x: i8x16, e: i8, idx: u32) -> i8x16 {
    simd_insert_dyn(x, idx, e)
 }

 // CHECK-LABEL: literal_dyn_simd_insert
-// CHECK: insertelement <16 x i8> %x, i8 %e, i32 7
+// CHECK: insertelement <16 x i8> %[[TEMP:.+]], i8 %e, i32 7


nit: if you're not going to consume the variable again later (by mentioning %[[TEMP]] somewhere) then I think you should just use a {{regex}} not a [[NAME:regex]].

scottmcm

Sorry I missed the ping here. I've dropped some thoughts, but nothing super-critical, so feel free to do with them what you will, since you already have an approval.

CaiWeiran added 6 commits July 14, 2025 11:41

Update dwarf-mixed-versions-lto.rs

3f6ce29

Fix tests/assembly/dwarf-mixed-versions-lto.rs test failure on riscv64.

Update const-vector.rs

cd818c7

Fix tests/codegen/const-vector.rs test failure on riscv64.

Update enum-aggregate.rs

cb80360

Fix tests/codegen/enum/enum-aggregate.rs test failure on riscv64.

Update extract-insert-dyn.rs

326780b

Fix tests/codegen/simd/extract-insert-dyn.rs test failure on riscv64.

Update transmute-scalar.rs

437b206

Fix tests/codegen/transmute-scalar.rs test failure on riscv64.

Update uninhabited-transparent-return-abi.rs

c162f6b

Fix tests/codegen/uninhabited-transparent-return-abi.rs test failure on riscv64.

rustbot assigned Mark-Simulacrum Jul 14, 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 Jul 14, 2025

This comment has been minimized.

Sign in to view

Update transmute-scalar.rs

9273462

Fix tidy check error.

CaiWeiran changed the title ~~Ci test~~ Fix RISC-V Test Failures in ./x test for Multiple Codegen and Assembly Cases Jul 14, 2025

This comment has been minimized.

Sign in to view

Update transmute-scalar.rs

77e17ff

Fix regex matching error.

Mark-Simulacrum reviewed Jul 15, 2025

View reviewed changes

tests/codegen/enum/enum-aggregate.rs Show resolved Hide resolved

Update rmake.rs

d464e57

Ignore cross-compile.

rustbot added the A-run-make Area: port run-make Makefiles to rmake.rs label Jul 18, 2025

Revert "Update rmake.rs"

2593e13

This reverts commit d464e57.

CaiWeiran mentioned this pull request Jul 28, 2025

Fix tests/codegen-llvm/transmute-scalar.rs test failure on riscv64 #144567

Open

scottmcm reviewed Aug 3, 2025

View reviewed changes

scottmcm reviewed Aug 4, 2025

View reviewed changes

	// CHECK-LABEL: define{{.}}i1 @fake_bool_signed_to_bool(i8 {{.}}%b)
	// CHECK-LABEL: define{{.*}}i1 @fake_bool_signed_to_bool(i8{{( zeroext)?}} %b)

Fix RISC-V Test Failures in ./x test for Multiple Codegen and Assembly Cases #143915

Are you sure you want to change the base?

Fix RISC-V Test Failures in ./x test for Multiple Codegen and Assembly Cases #143915

Conversation

CaiWeiran commented Jul 14, 2025

Notes:

Uh oh!

rustbot commented Jul 14, 2025

Uh oh!

CaiWeiran commented Jul 14, 2025

Uh oh!

This comment has been minimized.

This comment has been minimized.

Uh oh!

rustbot commented Jul 15, 2025

Uh oh!

Mark-Simulacrum commented Jul 15, 2025

Uh oh!

nikic commented Jul 15, 2025

Uh oh!

CaiWeiran commented Jul 16, 2025

Uh oh!

nikic commented Jul 16, 2025

Uh oh!

CaiWeiran commented Jul 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

nikic commented Jul 16, 2025

Uh oh!

rustbot commented Jul 18, 2025

Uh oh!

bors commented Jul 22, 2025

Uh oh!

Mark-Simulacrum commented Aug 3, 2025

Uh oh!

scottmcm Aug 3, 2025

Choose a reason for hiding this comment

Uh oh!

scottmcm Aug 4, 2025

Choose a reason for hiding this comment

Uh oh!

scottmcm Aug 3, 2025

Choose a reason for hiding this comment

Uh oh!

scottmcm Aug 3, 2025

Choose a reason for hiding this comment

Uh oh!

scottmcm left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

CaiWeiran commented Jul 16, 2025 •

edited

Loading