Date: Sun, 19 Oct 2003 22:51:33 +0100 From: David Malone <dwmalone@maths.tcd.ie> To: Robert Watson <rwatson@FreeBSD.org> Cc: Kris Kennaway <kris@obsecurity.org> Subject: Re: cvs commit: src/sys/dev/streams streams.c src/sys/kern kern_descrip.c kern_event.c sys_pipe.c uipc_syscalls.c vfs_syscalls.c src/sys/opencrypto cryptodev.c Message-ID: <200310192251.aa71195@salmon.maths.tcd.ie> In-Reply-To: Your message of "Sun, 19 Oct 2003 17:08:09 EDT." <Pine.NEB.3.96L.1031019165954.4181F-100000@fledge.watson.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> For example, some of the file descriptor handling during execve(), in > kqueue_register(), unp_externalize(), the /dev/fd code, et al. We need to > either pick to have the file descriptor locks in the lock order before > Giant, or after Giant, and right now we're in limbo, so Witness generates > lock order warnings... The code I've been looking at is mainly syscall code, and this grab Giant first and then grab the file descriptor lock. I don't plan to change the lock order here, but rather avoid grabbing Giant where it isn't needed any longer. I guess this may indirectly help if I can reduce the number of places where Giant and the file descriptor lock are held at the same time. David.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200310192251.aa71195>