Date: Fri, 18 Aug 2006 10:15:47 +0200 From: Divacky Roman <xdivac02@stud.fit.vutbr.cz> To: Peter Jeremy <peterjeremy@optushome.com.au> Cc: Astrodog <astrodog@gmail.com>, current@freebsd.org Subject: Re: HEADS-UP: starting to commit linuxolator (SoC 2006) changes... Message-ID: <20060818081546.GA40569@stud.fit.vutbr.cz> In-Reply-To: <20060818080607.GA732@turion.vk2pj.dyndns.org> References: <20060816132539.owwerbnw0okwc8wo@netchild.homeip.net> <20060817080533.GA845@turion.vk2pj.dyndns.org> <20060817122534.e57aqlbrwwogg8ko@netchild.homeip.net> <44E4454B.2080606@elischer.org> <20060817133721.h4cbucizcw8wc88k@netchild.homeip.net> <20060817140122.GA90642@stud.fit.vutbr.cz> <44E4881A.3050907@elischer.org> <20060817164352.GB96801@stud.fit.vutbr.cz> <2fd864e0608171556o6af3e1fdxa29c252f07c98444@mail.gmail.com> <20060818080607.GA732@turion.vk2pj.dyndns.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> IMHO, it is cleaner to explicitly delete it once it becomes non-functional. > Having dead code lying around just slows down compilation and misleads > people into thinking that it is functional. If someone wants to resurrect > it at a later date, they can always retrieve it from the Attic. what makes 2.6 and 2.4 incompatible is the way they treat "pid". in 2.4 times pid is something unique to a process, ie. you cannot have two processes with the same pid. in 2.6 this changed pid is number processes can share and now tid is used as "unique identifier of a process". netchild commited patch which lets this happen and it makes getpid/getppid/exit_group quite expensive (I have to retrieve osrelease for a given jail and compare the string).
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060818081546.GA40569>