Date: Tue, 4 Sep 2001 13:03:43 -0400 (EDT) From: Robert Watson <rwatson@FreeBSD.org> To: arch@FreeBSD.org Subject: ucred -> cred, cr_<e>[ug]id Message-ID: <Pine.NEB.3.96L.1010904130008.54921D-100000@fledge.watson.org>
next in thread | raw e-mail | index | archive | help
Earlier this year, I merged the 'pcred' and 'ucred' structures into a single 'ucred' structure, recognizing that in practice our use of pcred and ucred was such that the costs of the change were low, but the clarity associated with the change was relatively high. This simplified things from a variety of perspective, including reference counting and synchronization. This is a change also performed on a number of other UNIX platforms, including Solaris and IRIX. However, on those platforms, the single structure is called 'cred' and not 'ucred'. As such, I'd like to propose that we do the following: o Replace all references to struct ucred with struct cred o Repo-copy src/sys/ucred.h to src/sys/cred.h I don't feel all that strongly about this change, but do think it would make sense given the use of the structure, and the evolution of other platforms making similar classes of changes. On a similar note, when I merged ucred and pcred, I maintained cr_uid as the effective uid: other platforms typically stick an 'e' in front to get cr_euid, making it consistent with other variables in the structure. If I were to make the 'cred' change, I'd also improve consistency by adding the 'e' to the effective uid and gid structures. Any strong reasons not to do this? Robert N M Watson FreeBSD Core Team, TrustedBSD Project robert@fledge.watson.org NAI Labs, Safeport Network Services To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.NEB.3.96L.1010904130008.54921D-100000>