Date: Sun, 10 Jun 2007 15:37:22 +0200 From: Attilio Rao <attilio@FreeBSD.org> To: attilio@FreeBSD.org Cc: Randall Stewart <rrs@cisco.com>, freebsd-current@freebsd.org Subject: Re: Don't know what I was doing but... Message-ID: <466BFE92.6000804@FreeBSD.org> In-Reply-To: <466BFE2A.7020908@FreeBSD.org> References: <466BEA69.1010708@cisco.com> <466BFE2A.7020908@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Attilio Rao wrote: > Randall Stewart wrote: >> Hi all: >> >> I have a fairly recent current.. like yesterday I think.. and >> I just saw this tumble out (I usually run with witness and invariants >> on in my laptop now :-D) >> >> lock order reversal: >> 1st 0xc0b79f70 sched lock (sched lock) @ kern/sched_4bsd.c:846 >> 2nd 0xc0bec658 descriptor tables (descriptor tables) @ >> i386/i386/sys_machdep.c:363 > > Mm, since dt_lock is a leaf mutex, the opposite case seems quite > impossible, so I expect this would be the result of some other bug... if > next cvsup doesn't fix it, just let me know and I will look into this. More precisely, I'd like to better study effects recent thread_lock introduction had on witness rules. Attilio
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?466BFE92.6000804>