From owner-svn-src-all@FreeBSD.ORG Sat Dec 19 19:01:53 2009 Return-Path: Delivered-To: svn-src-all@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 9C7681065694; Sat, 19 Dec 2009 19:01:53 +0000 (UTC) (envelope-from ed@hoeg.nl) Received: from palm.hoeg.nl (mx0.hoeg.nl [IPv6:2001:7b8:613:100::211]) by mx1.freebsd.org (Postfix) with ESMTP id 4018E8FC1B; Sat, 19 Dec 2009 19:01:53 +0000 (UTC) Received: by palm.hoeg.nl (Postfix, from userid 1000) id 343431CECF; Sat, 19 Dec 2009 20:01:52 +0100 (CET) Date: Sat, 19 Dec 2009 20:01:52 +0100 From: Ed Schouten To: Kostik Belousov Message-ID: <20091219190152.GJ64905@hoeg.nl> References: <200912191842.nBJIgDgP040506@svn.freebsd.org> <20091219185457.GH2170@deviant.kiev.zoral.com.ua> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="fZfO+AAAmBD6aS/A" Content-Disposition: inline In-Reply-To: <20091219185457.GH2170@deviant.kiev.zoral.com.ua> User-Agent: Mutt/1.5.20 (2009-06-14) Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r200732 - in head/sys: fs/devfs kern sys X-BeenThere: svn-src-all@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: "SVN commit messages for the entire src tree \(except for " user" and " projects" \)" List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 19 Dec 2009 19:01:53 -0000 --fZfO+AAAmBD6aS/A Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable * Kostik Belousov wrote: > Note that struct devfs_dirent is available as vp->v_data, and > devfs_dirent' de_cdp member contains pointer to cdev_priv. I think this > would allow to remove s_ttydp. Well, there are lots of fields that could be derived. We could even remove s_ttyp and only use the vnode, because cdp_c.si_drv1 contains a pointer to the TTY. The reason why I didn't choose that approach, was because I'm a bit afraid we get these insanely long chains of pointer dereferences, of which I'm not entirely sure we do the locking properly. Even right now I'm not sure whether we pick up proctree_lock enough. s_ttydp isn't too strict about any form of locking, because its value is only used comparatively. --=20 Ed Schouten WWW: http://80386.nl/ --fZfO+AAAmBD6aS/A Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (FreeBSD) iEYEARECAAYFAkstIyAACgkQ52SDGA2eCwW+2QCdFgQyWZZU9OZ0bF3NMst6saQ/ vHIAnA/8vvu5SVGTyg6ndImpHoCTDq4t =3nhx -----END PGP SIGNATURE----- --fZfO+AAAmBD6aS/A--