Re: kern/57705: rump doesn't have a way to interrupt rump_sys_* by a signal

[Taylor R Campbell <riastradh%NetBSD.org@localhost>]

The following reply was made to PR kern/57705; it has been noted by GNATS.

From: Taylor R Campbell <riastradh%NetBSD.org@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: netbsd-bugs%NetBSD.org@localhost
Subject: Re: kern/57705: rump doesn't have a way to interrupt rump_sys_* by a signal
Date: Sun, 19 Nov 2023 04:09:30 +0000

 Oops... that patch isn't going to work reliably: potential lock order
 reversal and races.  Need to find another approach -- probably have to
 arrange psignal to take the lock that's serializing cv_wait_sig, which
 means finding some way to safely expose that lock through
 l->l_sched.info.