rustc_codegen_llvm: make sse4.2 imply crc32 for LLVM 14

[durin42]

This fixes compiling things like the snap crate after
https://reviews.llvm.org/D105462. I added a test that verifies the
additional attribute gets specified, and confirmed that I can build
cargo with both LLVM 13 and 14 with this change applied.

r? @nagisa cc @nikic

[rust-highfive]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @nagisa (or someone else) soon.

Please see the contribution instructions for more information.

[durin42]

Added an assembly test that covers the -Ctarget-feature=+sse4.2 case.

[nagisa]

@bors try @rust-timer queue

r=me once the commits are squashed and the perf run does not indicate any significant negative trend.

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit 28e0274cb84cda89e3f085b55060d877eeb42c21 with merge 232f9135f71fd7a09bc04b440e3cb7a1966d361a...

[bors]

☀️ Try build successful - checks-actions
Build commit: 232f9135f71fd7a09bc04b440e3cb7a1966d361a (232f9135f71fd7a09bc04b440e3cb7a1966d361a)

[rust-timer]

Queued 232f9135f71fd7a09bc04b440e3cb7a1966d361a with parent d1d8145, future comparison URL.

[rust-timer]

Finished benchmarking commit (232f9135f71fd7a09bc04b440e3cb7a1966d361a): comparison url.

Summary: This benchmark run did not return any relevant changes.

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR led to changes in compiler perf.

@bors rollup=never
@rustbot label: +S-waiting-on-review -S-waiting-on-perf -perf-regression

[durin42]

Now that the perf run is done, rebased and squashed.

[nagisa]

@bors r+

[bors]

📌 Commit 1c779647259894dbc52b3e571ff8642c21347a35 has been approved by nagisa

[bors]

⌛ Testing commit 1c779647259894dbc52b3e571ff8642c21347a35 with merge 2c1f37364fd5b287f5dfb5c04e6d29cef63df49b...

[bors]

💔 Test failed - checks-actions

[durin42]

Do you have a preference between those options?

…

On Sat, Sep 18, 2021, 17:06 Simonas Kazlauskas ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In src/test/assembly/x86_64-sse_crc.rs <#88981 (comment)>: > @@ -0,0 +1,12 @@ +// assembly-output: emit-asm +// compile-flags: --target=x86_64-unknown-linux-gnu --crate-type staticlib -Ctarget-feature=+sse4.2 +// needs-llvm-components: x86 If you want this to work, you will need to either: 1. Make this a #[no_core] test (allowing this test to run on any architecture); 2. Make this a // only-x86 test instead (removing the --target flag and the needs-llvm-components annotation). — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#88981 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAE6LJZBGSZ4OEDVWJLCATUCT5L7ANCNFSM5EC563RQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

[nagisa]

The only-x86 approach is definitely more straightforward, so its probably going to be better for you to adjust it that way.

That said, I personally prefer the no_core option. It allows for better cross-architecture coverage when developing on machines that execute a different instruction set.

[durin42]

I got a little lost trying to figure out the no_core version, so I did the only-x86_64 flavor with an amend+rebase.

[nagisa]

@bors r+

[bors]

📌 Commit 4185b76 has been approved by nagisa

[bors]

⌛ Testing commit 4185b76 with merge 840acd3...

[bors]

☀️ Test successful - checks-actions
Approved by: nagisa
Pushing 840acd3 to master...

[rust-timer]

Finished benchmarking commit (840acd3): comparison url.

Summary: This benchmark run did not return any relevant changes.

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

@rustbot label: -perf-regression