From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Collins Date: Sun, 20 Jun 2004 19:25:49 +0000 Subject: Re: New SILO (was Re: [SOLVED] Take 3: sparc64 CDs work?) Message-Id: <20040620192549.GC19634@phunnypharm.org> List-Id: References: <20040620170821.GA19634@phunnypharm.org> In-Reply-To: <20040620170821.GA19634@phunnypharm.org> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: sparclinux@vger.kernel.org On Sun, Jun 20, 2004 at 12:41:25PM -0700, David S. Miller wrote: > On Sun, 20 Jun 2004 13:08:22 -0400 > Ben Collins wrote: > > > I suspect that in the long run, I really need to make it so that the > > initrd gets lower memory precendence over the kernel. The kernel itself > > can be loaded anywhere for sparc64, even in 64-bit address ranges. > > > > But for now, I am about to release SILO 1.4.6 to atleast make this less > > of a mystery on systems where it fails. > > I have a better idea, remove the 32-bit limitation. > > Create a new HdrS version, add a new member at the end of the bootloader > header area "sparc_ramdisk_image64", set it to ~0 by default, if the > bootloader sets it to some non-~0 value, we use it instead of the > 32-bit limited sparc_ramdisk_image value. The downside is that I need to load and evaluate the kernel before loading the initrd in order for it to work. I was already considering this, but I need to make sure I don't get into a chicken and egg situation like I did with the feature to load the kernel to a non-zero location. -- Debian - http://www.debian.org/ Linux 1394 - http://www.linux1394.org/ Subversion - http://subversion.tigris.org/ WatchGuard - http://www.watchguard.com/