From owner-freebsd-hackers Tue Aug 26 06:30:25 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id GAA10019 for hackers-outgoing; Tue, 26 Aug 1997 06:30:25 -0700 (PDT) Received: from eddie.mit.edu (EDDIE.MIT.EDU [18.62.0.6]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id GAA10001 for ; Tue, 26 Aug 1997 06:30:12 -0700 (PDT) Received: from ccomp.inode.com by eddie.mit.edu id aa09318; 26 Aug 97 9:22 EDT Received: (from gdk@localhost) by ccomp.inode.com (8.6.9/1.3) id JAA02677; Tue, 26 Aug 1997 09:09:01 -0400 From: Gary Kendall Message-Id: <199708261309.JAA02677@ccomp.inode.com> Subject: Re: Sizing mem > 64MB at boot-time To: joerg_wunsch@uriah.heep.sax.DE Date: Tue, 26 Aug 1997 09:09:01 -0400 (EDT) Cc: freebsd-hackers@freebsd.ORG In-Reply-To: <19970826080352.YY43772@uriah.heep.sax.de> from J Wunsch at "Aug 26, 97 08:03:52 am" X-Mailer: ELM [version 2.4ME+ PL17 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-hackers@freebsd.ORG X-Loop: FreeBSD.org Precedence: bulk It wasn't too long ago that J Wunsch said: > I recently wrote the `cdboot' code, but things like these are still > highly puzzling for me. In particular, when it comes to gas and gas's > unsupport for 16-bit code. > Bless you, sir. Misery loves company. ;-) > I sometimes went the route that i've been stuffing the code into a > .bin file, and ran symdeb.exe inside pcemu to analyze the 16-bit code. > Ugh... make mine hemlock, please. :-\ > Btw., the most recent bootblocks are probably too full for this > anyway. :-( > Well, I'd just like to see it work under the 2.0 system that I'm running now. I've got 2.2.2 sitting here on CD, but it's going on a machine that I haven't even built yet. After I build the machine, and install the OS, I'll worry about the new boot blocks being too big. ;-) Maybe, by then, I'll be "inspired" enough to write a 3-stage boot with a /boot file and maybe even a system monitor and diagnostics package. > Btw^2, cdboot contains a simple malloc implementation that operates on > the `heap' that is also available for the realmode BIOS. This was > required for me to read the CD-ROM root directory (which can vary in > size). Maybe this solves your storage pool problem. > I'll take a look, but I've got a real problem with having to allocate a 20-byte buffer of what should be "local" storage on a heap. This may be the only way to go if you want the data accessable in both real and protected mode, but what's killing me is that the code I wrote only needs to use that storage in real mode! The returned value (memsize in 1KB pages) is in the %ebx register during the transition from real_to_prot(), and moved into %eax afterward, where the compiler expects to find it. > I think there's a section in the handbook explaining the boot process. > This should also mention the memory regions that are used in the > bootstrap. > I'm guessing there's a copy of the handbook at www.freebsd.org... am I right? (Don't bother to answer... I'm on my way there now. ;-) > -- > cheers, J"org > > joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE > Never trust an operating system you don't have sources for. ;-) > Thanks, J"org. Regards, Gary Kendall %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % The difficult is done immediately, the impossible takes a little longer. % %----------------------------------------------------------------------------% % gdk@ccomp.inode.com or gkendall@eddie.mit.edu % %----------------------------------------------------------------------------% % Creative Computing * 96 Forest Street * Danvers, MA 01923 * 508-777-3784 % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%