From owner-freebsd-arch Thu Mar 22 11:47:16 2001 Delivered-To: freebsd-arch@freebsd.org Received: from smtp04.primenet.com (smtp04.primenet.com [206.165.6.134]) by hub.freebsd.org (Postfix) with ESMTP id D215937B71D for ; Thu, 22 Mar 2001 11:47:10 -0800 (PST) (envelope-from tlambert@usr06.primenet.com) Received: (from daemon@localhost) by smtp04.primenet.com (8.9.3/8.9.3) id MAA04925 for ; Thu, 22 Mar 2001 12:41:02 -0700 (MST) Received: from usr06.primenet.com(206.165.6.206) via SMTP by smtp04.primenet.com, id smtpdAAAKOaGHj; Thu Mar 22 12:40:50 2001 Received: (from tlambert@localhost) by usr06.primenet.com (8.8.5/8.8.5) id MAA15517 for arch@FreeBSD.ORG; Thu, 22 Mar 2001 12:46:55 -0700 (MST) From: Terry Lambert Message-Id: <200103221946.MAA15517@usr06.primenet.com> Subject: 4MB pages To: arch@FreeBSD.ORG Date: Thu, 22 Mar 2001 19:46:30 +0000 (GMT) X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-arch@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I know that this is going to look like I should have posted it on the "-arch-questions" list, but there isn't one, and it involves future work (I think), so here goes... Has anyone tried playing with 4MB pages for doing other things? Specifically, has anyone played with them for mappings for all the text pages in an executable, or a large contiguous data mapping for something like a DVDROM or anonymous regions? I'm interested in playing around with this, and it seems to me the only place this flag is ever set is in the initial kernel mapping and mmap of physical devices? This is what pmap_bootstrap() and pmap_object_init_pt() leads me to believe... is this correct? That would imply that the DVD case would "just work", if the right dead chicken were waved over it... start, stop, size, etc. parameters to mmap() (would like to know if it needs a special kick in the pants). What about the /dev/zero case (to grab anonymous memory via mmap())? It seems that this would be incredibly useful for CPU emulation for something like running foreign code by mapping it into a big page in an emulator, and emulating the instruction set against the code, in memory. Will /dev/zero "just work" because it's a device (OBJT_DEVICE)? Will ALL devices "just work"? I can't tell what this would do to the pagability -- pmap_enter and family seem to want to panic, so would this only be useful on a machine with a huge amount of RAM, or only for using one page at a time as a window, if it were too large? As usual, I'm playing with 4.x, not 5.x, so feel free to tell me that this has already been thought of, and therefore is not an issue for -arch. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message