Drop panic hook after running tests

[Duckilicious]

Issue: #119223
Previously we left the panic hook we allocated
on main termination. Doing so makes Valgrind
report it as a reachable unfreed block.
In order to fix that use panic::take_hook() before examining test results.

Example backtrace:

==146594== 16 bytes in 1 blocks are still reachable in loss record 1 of 1
==146594==    at 0x4A390C5: malloc (vg_replace_malloc.c:442)
==146594==    by 0x151336: alloc (alloc.rs:98)
==146594==    by 0x151336: alloc_impl (alloc.rs:181)
==146594==    by 0x151336: allocate (alloc.rs:241)
==146594==    by 0x151336: exchange_malloc (alloc.rs:330)
==146594==    by 0x151336: new<test::test_main::{closure_env#0}> (boxed.rs:217)
==146594==    by 0x151336: test::test_main (lib.rs:124)
==146594==    by 0x1522F9: test::test_main_static (lib.rs:160)
==146594==    by 0x11E102: reachable_block_with_cargo_test::main (lib.rs:1)
==146594==    by 0x11EABA: core::ops::function::FnOnce::call_once (function.rs:250)
==146594==    by 0x11E76D: std::sys_common::backtrace::__rust_begin_short_backtrace (backtrace.rs:154)
==146594==    by 0x11DFC0: std::rt::lang_start::{{closure}} (rt.rs:166)
==146594==    by 0x177D3A: call_once<(), (dyn core::ops::function::Fn<(), Output=i32> + core::marker::Sync + core::panic::unwind_safe::RefUnwindSafe)> (function.rs:284)
==146594==    by 0x177D3A: do_call<&(dyn core::ops::function::Fn<(), Output=i32> + core::marker::Sync + core::panic::unwind_safe::RefUnwindSafe), i32> (panicking.rs:504)
==146594==    by 0x177D3A: try<i32, &(dyn core::ops::function::Fn<(), Output=i32> + core::marker::Sync + core::panic::unwind_safe::RefUnwindSafe)> (panicking.rs:468)
==146594==    by 0x177D3A: catch_unwind<&(dyn core::ops::function::Fn<(), Output=i32> + core::marker::Sync + core::panic::unwind_safe::RefUnwindSafe), i32> (panic.rs:142)
==146594==    by 0x177D3A: {closure#2} (rt.rs:148)
==146594==    by 0x177D3A: do_call<std::rt::lang_start_internal::{closure_env#2}, isize> (panicking.rs:504)
==146594==    by 0x177D3A: try<isize, std::rt::lang_start_internal::{closure_env#2}> (panicking.rs:468)
==146594==    by 0x177D3A: catch_unwind<std::rt::lang_start_internal::{closure_env#2}, isize> (panic.rs:142)
==146594==    by 0x177D3A: std::rt::lang_start_internal (rt.rs:148)
==146594==    by 0x11DF99: std::rt::lang_start (rt.rs:165)

[rustbot]

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

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

• @rustbot author: the review is finished, PR author should check the comments and take action accordingly
• @rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

[Duckilicious]

Before:

valgrind target/debug/deps/reachable_block_with_cargo_test-049d9bdb11a1a464
==80325== Memcheck, a memory error detector
==80325== Copyright (C) 2002-2022, and GNU GPL'd, by Julian Seward et al.
==80325== Using Valgrind-3.22.0 and LibVEX; rerun with -h for copyright info
==80325== Command: target/debug/deps/reachable_block_with_cargo_test-049d9bdb11a1a464
==80325==

running 1 test
test tests::it_works ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.09s

==80325==
==80325== HEAP SUMMARY:
==80325==     in use at exit: 16 bytes in 1 blocks
==80325==   total heap usage: 446 allocs, 445 frees, 61,882 bytes allocated
==80325==
==80325== LEAK SUMMARY:
==80325==    definitely lost: 0 bytes in 0 blocks
==80325==    indirectly lost: 0 bytes in 0 blocks
==80325==      possibly lost: 0 bytes in 0 blocks
==80325==    still reachable: 16 bytes in 1 blocks
==80325==         suppressed: 0 bytes in 0 blocks
==80325== Rerun with --leak-check=full to see details of leaked memory
==80325==
==80325== For lists of detected and suppressed errors, rerun with: -s
==80325== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

After:

❯ valgrind target/debug/deps/reachable_block_with_cargo_test-47711fac179ab84c
==80182== Memcheck, a memory error detector
==80182== Copyright (C) 2002-2022, and GNU GPL'd, by Julian Seward et al.
==80182== Using Valgrind-3.22.0 and LibVEX; rerun with -h for copyright info
==80182== Command: target/debug/deps/reachable_block_with_cargo_test-47711fac179ab84c
==80182==

running 1 test
test tests::it_works ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.11s

==80182==
==80182== HEAP SUMMARY:
==80182==     in use at exit: 0 bytes in 0 blocks
==80182==   total heap usage: 445 allocs, 445 frees, 61,893 bytes allocated
==80182==
==80182== All heap blocks were freed -- no leaks are possible
==80182==
==80182== For lists of detected and suppressed errors, rerun with: -s
==80182== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

[cuviper]

Thanks!

@bors r+ rollup

[bors]

📌 Commit 92ebf60 has been approved by cuviper

It is now in the queue for this repository.