Date: Sat, 24 Apr 1999 03:33:36 -0500 (CDT) From: Anthony Kimball <alk@pobox.com> To: jhermes@infoglobe.com Cc: hackers@freebsd.org Subject: signal permissions Message-ID: <14113.32330.554950.935946@avalon.east>
next in thread | raw e-mail | index | archive | help
: > : >Here's a trial balloon: Anyone who can write to an executable file : >should be permitted to signal a derived process. : > : I dunno, I like to keep process space (who it runs as) separate from file : space (how it exists). : Implied permissions from one space being applied to another always reduces : flexibility, right? I don't see how, in this case. It increases practical flexibility. We're not talking about orthogonality. But your criticism suggests another which you left implicit, but which perhaps motivated yours in some fundamental way: Overloading is almost always sucky. Perhaps a flag would be better. I can only think of 6 cases worth implementing: exec'ing/any uid/gid can sighup/sigterm, where any uid can sighup/sigterm == any gid can sighup/sigterm. That takes 3 flag bits to support. Blech. Oh, for capability tickets! To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14113.32330.554950.935946>