Only traverse attrs once while checking for coherence override attributes

[lqd]

In coherence, while checking for negative impls override attributes: only traverse the DefIds' attributes once.

This PR is an easy way to get back some of the small perf loss in #93175

[lqd]

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

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

[bors]

⌛ Trying commit 644c22d with merge 14b4e3475978414bd2231200da745dc72f659842...

[bors]

☀️ Try build successful - checks-actions
Build commit: 14b4e3475978414bd2231200da745dc72f659842 (14b4e3475978414bd2231200da745dc72f659842)

[rust-timer]

Queued 14b4e3475978414bd2231200da745dc72f659842 with parent a7f3757, future comparison URL.

[rust-timer]

Finished benchmarking commit (14b4e3475978414bd2231200da745dc72f659842): comparison url.

Summary: This benchmark run shows 7 relevant improvements 🎉 to instruction counts.

• Average relevant improvement: -1.1%
• Largest improvement in instruction counts: -1.4% on incr-full builds of stm32f4 check

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

[lqd]

Super quick way to get back some of the small losses in the negative impl PR #93175, so r? @spastorino to do what you want with it, especially when #93348 should be better than this

[spastorino]

@bors r+

Let's land this but I'm taking a different approach in another PR. Would be nice to see performance difference with that other approach too.

[bors]

📌 Commit 644c22d has been approved by spastorino

[bors]

⌛ Testing commit 644c22d with merge 312a799...

[bors]

☀️ Test successful - checks-actions
Approved by: spastorino
Pushing 312a799 to master...

[rust-timer]

Finished benchmarking commit (312a799): comparison url.

Summary: This benchmark run shows 7 relevant improvements 🎉 to instruction counts.

• Average relevant improvement: -1.1%
• Largest improvement in instruction counts: -1.4% on incr-full builds of stm32f4 check

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

@rustbot label: -perf-regression