Date: Wed, 21 Jan 2015 17:38:29 +0100 From: Hans Petter Selasky <hps@selasky.org> To: Adrian Chadd <adrian@freebsd.org>, "freebsd-arch@freebsd.org" <freebsd-arch@freebsd.org> Subject: Re: please test: i386 bootloader changes Message-ID: <54BFD605.7090603@selasky.org> In-Reply-To: <CAJ-Vmo=AD5cJYsO1EEdfawusxVNVrJO_TTvQT5=_aGEF8JBq3A@mail.gmail.com> References: <CAJ-Vmo=AD5cJYsO1EEdfawusxVNVrJO_TTvQT5=_aGEF8JBq3A@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 01/21/15 17:30, Adrian Chadd wrote: > Hi, > > Here's something that I've been poking for a little while. > > The SeaBIOS emulation in the Chromebook C720 implements the E820 > memory map by putting a hole between 15 and 16MiB. That's some really > legacy PC-AT stuff right there. Unfortunately that stops > kernel.GENERIC from loading as loader thinks there's only < 14MiB of > extended RAM available. > > This patch: > > https://people.freebsd.org/~adrian/c720/20150121-seabios-loader-changes-2.diff > > updates it so if the first extended segment from the E820 call is too > small, it tries the E801 call. It also fixes a bug in the E801 call > that would treat the lower and upper memory regions returned as > contiguous when they may indeed not be. > > Now - here be dragons. It's possible that fixing this stuff for one > buggy bios makes freebsd's loader unhappy on other buggy bioses. So > I'd really appreciate if it people would try this out on 32/64 bit x86 > platforms and report back success/failures. So far nothing I have here > complains, but I don't have /that/ much hardware. > > Thanks, > Hi, What is the simplest way to install the loader changes? Can it be tested by replacing "/boot/loader" or do I need to update any boot partition code using "gpart bootcode" ? --HPS
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?54BFD605.7090603>