Date: Sat, 10 May 2003 12:25:06 -0700 From: Terry Lambert <tlambert2@mindspring.com> To: David Schultz <das@FreeBSD.ORG> Cc: John Baldwin <jhb@FreeBSD.ORG> Subject: Re: [Bikeshed] sigacts locking Message-ID: <3EBD5212.225AC129@mindspring.com> References: <XFMail.20030509175046.jhb@FreeBSD.org> <20030510172609.GA29039@HAL9000.homeunix.com> <3EBD4214.73CB8B7C@mindspring.com> <20030510191352.GA225@HAL9000.homeunix.com>
next in thread | previous in thread | raw e-mail | index | archive | help
David Schultz wrote: > No, they're already wired kernel pages that are unwired and > unmapped when the process is swapped out. Moreover, there's > probably *more* KVA pressure with upages swapping, because each > tiny struct upages gets a 4K or larger page all to itself unless > it's swapped out. The wiring and unwiring doesn't change the fundamental nature of the page. I can think of a lot of useful things to put in that page, if you are willing to map it read-only into user space. The pid, uid, gid, ppid, etc. could all be made zero system call functions, for example. So I'm not as concerned about it being 4K in size. I'm more concerned with the idea of being able to migrate a process to another node in a cluster, and having all of the information needed to do this outside nominally kernel data structures, instead of having to do a lot of work to extricate the process from a given node's kernel data. (Well, you *did* wan us in the "Subject:"... 8-) 8-)). -- Terry
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3EBD5212.225AC129>