From mboxrd@z Thu Jan 1 00:00:00 1970 From: Carl-Daniel Hailfinger Subject: Re: [Qemu-devel] [PATCH 0/3] Add BIOS splash image support Date: Wed, 17 Dec 2008 18:12:11 +0100 Message-ID: <494932EB.3070304@gmx.net> References: <1229440810-12394-1-git-send-email-Laurent.Vivier@bull.net> <49480F6D.2010302@codemonkey.ws> <20081217110319.GD13455@redhat.com> <200812171318.58650.paul@codesourcery.com> <49490282.3080800@siemens.com> <20081217135512.GJ13794@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Blue Swirl , bochs developers , Paul Brook , kvm developers To: qemu-devel@nongnu.org Return-path: Received: from mail.gmx.net ([213.165.64.20]:43219 "HELO mail.gmx.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1750853AbYLQRMP (ORCPT ); Wed, 17 Dec 2008 12:12:15 -0500 In-Reply-To: <20081217135512.GJ13794@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 17.12.2008 14:55, Gleb Natapov wrote: > On Wed, Dec 17, 2008 at 02:45:38PM +0100, Jan Kiszka wrote: > >> Paul Brook wrote: >> >>>>> Modern BIOSes have splash screens. I don't see why our BIOS shouldn't >>>>> have one too. >>>>> >>>> Crap PC BIOSes have splash screens because they're horribly slow >>>> and otherwise printing lots of irrelevant scary junk at users. The >>>> best BIOS 'splash' screen is one which never appears unless there >>>> is a boot failure, and gets control to the OS as quickly as possible. >>>> IMHO a better goal is reducing the time until the OS / bootloader is >>>> able to take over all management of the display. >>>> >>> I agree. The qemu bios init process takes almost no time. >>> >>> The only reason it takes a noticeable amount of time is that we have a >>> deliberate delay in there to allow the user to access the boot menu. >>> I'm not convinced this menu is actually very useful in practice, It's >>> something that should probably be delegated to your management utility and/or >>> be optional. >>> >> Yes, please. I hate this artificial delay, specifically as I have to do >> a lot of short boot tests where this contributes noticeably to their >> execution time. But, as usual, it didn't hurt enough to make me hack a >> patch yet. I'm willing to do this if we can agree on how it should be done. >> >> > Add "-bootmenu" option to qmeu and pass it to BIOS via fw confing interface? > If you pick NVRAM (RTC battery data storage) for this config item, traditional BIOS and coreboot and EFI and OFW can pick it up without problems. Plus, it is the traditional way to store splash screen settings. Regards, Carl-Daniel -- http://www.hailfinger.org/