use sort_unstable to sort primitive types

[matthiaskrgr]

It's not important to retain original order if we have &[1, 1, 2, 3] for example.

clippy::stable_sort_primitive

[rust-highfive]

r? @davidtwco

(rust_highfive has picked a reviewer for you, use r? to override)

[leonardo-m]

Can't sort perform by itself this optimization inside?

[matthiaskrgr]

Hmm, it may be possible to have a custom sort() impl for slices of primitives and use sort::quicksort() (unstable sort) instead of merge_sort() under the hood, but I don't know what kind of dangers this implies. I'm wondering what the libs team thinks.

sort
sort_unstable

[davidtwco]

LGTM, can make an issue if you want to look into performing this optimisation within sort.

[davidtwco]

@bors r+ rollup=never (might have a perf impact)

[bors]

📌 Commit b4935e0 has been approved by davidtwco

[matthiaskrgr]

Zulip thread: https://rust-lang.zulipchat.com/#narrow/stream/219381-t-libs/topic/default.20to.20unstable.20sort.20when.20sorting.20primitive.20type.20slices

May make an issue if this remains unanswered so the idea does not get lost :)

[bors]

⌛ Testing commit b4935e0 with merge 49888362e2fbf2738f473ef73621c78475d32d34...

[bors]

💥 Test timed out

[lqd]

@bors retry

[bors]

⌛ Testing commit b4935e0 with merge f8b002441bba6130859a9fe7a8b86c2d39196ca2...

[bors]

💔 Test failed - checks-azure

[matthiaskrgr]

failures:

---- [run-make] run-make-fulldeps/long-linker-command-lines stdout ----

error: make failed
status: exit code: 2
command: "make"
stdout:
------------------------------------------
DYLD_LIBRARY_PATH="/Users/runner/work/1/s/build/x86_64-apple-darwin/test/run-make-fulldeps/long-linker-command-lines/long-linker-command-lines:/Users/runner/work/1/s/build/x86_64-apple-darwin/stage2/lib:" '/Users/runner/work/1/s/build/x86_64-apple-darwin/stage2/bin/rustc' --out-dir /Users/runner/work/1/s/build/x86_64-apple-darwin/test/run-make-fulldeps/long-linker-command-lines/long-linker-command-lines -L /Users/runner/work/1/s/build/x86_64-apple-darwin/test/run-make-fulldeps/long-linker-command-lines/long-linker-command-lines  foo.rs -g -O
RUSTC="/Users/runner/work/1/s/build/x86_64-apple-darwin/stage2/bin/rustc" DYLD_LIBRARY_PATH="/Users/runner/work/1/s/build/x86_64-apple-darwin/test/run-make-fulldeps/long-linker-command-lines/long-linker-command-lines:/Users/runner/work/1/s/build/x86_64-apple-darwin/stage2/lib/rustlib/x86_64-apple-darwin/lib:" /Users/runner/work/1/s/build/x86_64-apple-darwin/test/run-make-fulldeps/long-linker-command-lines/long-linker-command-lines/foo
attempt: 100
attempt: 200
attempt: 300

------------------------------------------
stderr:
------------------------------------------
thread 'main' panicked at 'status: exit code: 1
stdout:

stderr:
error: couldn't read /Users/runner/work/1/s/build/x86_64-apple-darwin/test/run-make-fulldeps/long-linker-command-lines/long-linker-command-lines/bar.rs: stream did not contain valid UTF-8

error: aborting due to previous error

', foo.rs:76:13
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
make: *** [all] Error 101

Hm, since this test did not fail previously (when the build timed out) I assume this is spurious?

[Mark-Simulacrum]

Seems quite likely, though I am a bit surprised by spurious failures in that test.

@bors retry run-make-fulldeps/long-linker-command-lines

[bors]

⌛ Testing commit b4935e0 with merge 285c7137e0fdfdd91de86ed9d178e295fe2b8569...

[bors]

💥 Test timed out

[davidtwco]

This is one unlucky PR..
@bors retry

[bors]

⌛ Testing commit b4935e0 with merge 8d505c0c5330ef924a4583158ab3a50d432c13d1...

[bors]

💔 Test failed - checks-actions

[davidtwco]

---- workspaces::relative_rustc stdout ----
running `/checkout/obj/build/x86_64-unknown-linux-gnu/stage2-tools/x86_64-unknown-linux-gnu/release/cargo build`
running `/checkout/obj/build/x86_64-unknown-linux-gnu/stage2-tools/x86_64-unknown-linux-gnu/release/cargo build`
thread 'workspaces::relative_rustc' panicked at '
Expected: execs
    but: exited with exit code: 101
--- stdout

--- stderr
error: could not execute process `/checkout/obj/build/x86_64-unknown-linux-gnu/stage2-tools/x86_64-unknown-linux-gnu/cit/t1927/lib/./foo -vV` (never executed)

Caused by:
  Text file busy (os error 26)
', src/tools/cargo/crates/cargo-test-support/src/lib.rs:832:13

Another spurious failure?

[matthiaskrgr]

Hmm, I rebased and retried again locally and didn't see any of the these test failures.

[Mark-Simulacrum]

Okay, let's try again. @bors retry

[bors]

⌛ Testing commit b4935e0 with merge 9b41541...

[bors]

☀️ Test successful - checks-actions, checks-azure
Approved by: davidtwco
Pushing 9b41541 to master...