Date: Fri, 8 Jan 2010 18:03:39 +0100 From: Ulrich =?utf-8?B?U3DDtnJsZWlu?= <uqs@spoerlein.net> To: Julian Elischer <julian@elischer.org> Cc: Pieter de Goeje <pieter@degoeje.nl>, shrivatsan <shrivatsan@gmail.com>, freebsd-hackers@freebsd.org Subject: Re: Question regarding memory disks Message-ID: <20100108170338.GC52442@acme.spoerlein.net> In-Reply-To: <4B452A5D.4000208@elischer.org> References: <5a13b8941001061349m701d17fbl489ec8cf883e8c3c@mail.gmail.com> <201001070017.36855.pieter@degoeje.nl> <4B452A5D.4000208@elischer.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 06.01.2010 at 16:27:09 -0800, Julian Elischer wrote: > Pieter de Goeje wrote: > > Because when you "erase" something, all it does is unlink (delete the > > reference to) the data. So there is currently no way the memory disk can free > > the memory associated with the data. That is also why you should normally use > > swap backed memory disks instead, or use tmpfs. These can return memory to > > the system. > > > > The ability of the filesystem to mark certain blocks as "erased" is important > > not only for memory disks but also for solid state drives. It is a feature > > UFS2 is currently lacking unfortunately. > > but is being worked on Will this automagically work for md(4) backed UFS2 file-systems? Is this also being worked on? Regards, Uli
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100108170338.GC52442>