std panicking implementation comments suggest we might make double-panic not abort

[CAD97]

rust/library/std/src/panicking.rs

Lines 707 to 714 in efb171e

	if panics > 1 || !can_unwind {
	// If a thread panics while it's already unwinding then we
	// have limited options. Currently our preference is to
	// just abort. In the future we may consider resuming
	// unwinding or otherwise exiting the thread cleanly.
	rtprintpanic!("thread panicked while panicking. aborting.\n");
	crate::sys::abort_internal();
	}

At this point, I think we've effectively guaranteed in practice that a panicking while already unwinding leads to a full-process abort, and changing this to attempt to unwind the inner panic or even just to abort the thread but not the whole process is likely far to breaking. We're even considering making unwinds from drops always abort, though AIUI this cannot replace the panic while unwinding immediate abort check.

Best guess at appropriate labels...

@rustbot modify labels +T-compiler +T-lang +C-cleanup +A-technical-debt

[peter-lyons-kehl]

Thank you. The unrealistic comment has been removed. See

rust/library/std/src/panicking.rs

Lines 814 to 820 in 6de928d

	if !can_unwind {
	// If a thread panics while running destructors or tries to unwind
	// through a nounwind function (e.g. extern "C") then we cannot continue
	// unwinding and have to abort immediately.
	rtprintpanic!("thread caused non-unwinding panic. aborting.\n");
	crate::sys::abort_internal();
	}

. I believe this issue may be closed.

Not related to this issue, but while on this topic: Yesterday I've read somewhere in the std source, or in std or Cargo (nightly) docs, that in Cargo.toml > profile.xxx, in addition to panic=abort and panic=unwind Rust may consider other panic strategies in the future (I don't remember exact wording). Today I've been searching for it, but I can't locate it anymore. Any idea where?