Date: Mon, 3 Mar 2003 14:08:30 -0800 (PST) From: Nate Lawson <nate@root.org> To: Robert Watson <rwatson@FreeBSD.org> Cc: Jeff Roberson <jroberson@chesapeake.net>, src-committers@FreeBSD.org, cvs-src@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/kern vfs_default.c Message-ID: <Pine.BSF.4.21.0303031407010.78667-100000@root.org> In-Reply-To: <Pine.NEB.3.96L.1030303160011.34673F-100000@fledge.watson.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 3 Mar 2003, Robert Watson wrote: > On Mon, 3 Mar 2003, Nate Lawson wrote: > > On Mon, 3 Mar 2003, Jeff Roberson wrote: > > > Hey, good work here. Do you think there would be any way to use the > > > vnode's char name in the generic sync routine? There are a few other > > > places where we msleep that could use a more specific name as well. > > > > Interesting idea but how would you suggest compressing that to 6 chars? > > I assume "show lockedvnodes" will always be supreme? And when I just > > glanced at the implementation of "show lockedvnodes", I noticed: > > mtx_lock(&mountlist_mtx); How can this be correct? > > It's not. I've been meaning to remove the locking from ddb for ages, and > I think Bruce also had patches floating around at one point. > > BTW, we now get some fascinating failure modes if you do: > > call boot(0) > > in ddb, since it runs in an interrupt context using a serial console, and > when the shutdown sequence tries to malloc storage, it panics... Please do this! I've noticed this failure for a while but I think it's non-deterministic and there are other problems. I never got around to looking into it though. -Nate To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-src" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0303031407010.78667-100000>