From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from [64.233.182.186] (helo=nf-out-0910.google.com) by linuxtogo.org with esmtp (Exim 4.63) (envelope-from ) id 1GxS04-0000xr-CG for openembedded-devel@lists.openembedded.org; Thu, 21 Dec 2006 18:46:36 +0100 Received: by nf-out-0910.google.com with SMTP id l24so2906278nfc for ; Thu, 21 Dec 2006 09:44:08 -0800 (PST) Received: by 10.49.91.6 with SMTP id t6mr11081169nfl.1166723047698; Thu, 21 Dec 2006 09:44:07 -0800 (PST) Received: from CUBE ( [85.202.124.214]) by mx.google.com with ESMTP id l22sm41007046nfc.2006.12.21.09.44.06; Thu, 21 Dec 2006 09:44:07 -0800 (PST) Date: Thu, 21 Dec 2006 19:44:02 +0200 From: Paul Sokolovsky X-Priority: 3 (Normal) Message-ID: <1151969252.20061221194402@gmail.com> To: Koen Kooi In-Reply-To: <458AC366.3030405@dominion.kabel.utwente.nl> References: <458AC366.3030405@dominion.kabel.utwente.nl> MIME-Version: 1.0 Subject: Re: [RFC] Enable tune-thumb.inc for all armv5t and armv6t machines X-BeenThere: openembedded-devel@lists.openembedded.org X-Mailman-Version: 2.1.9 Precedence: list Reply-To: openembedded-devel@lists.openembedded.org List-Id: Using the OpenEmbedded metadata to build Distributions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 21 Dec 2006 17:46:36 -0000 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello Koen, Thursday, December 21, 2006, 7:24:54 PM, you wrote: > Hi, > Today I seperated the thumb bits from > conf/machine/include/ipx4xx.conf into > conf/machine/include/. I'd like to enable for all >=v5t machines, but I'm unsure about > using positive or negative logic. What about armv4t? > * positive logic: might break for (soft)fpa distros that don't disable it, no problem for > EABI distros > * negative logic: no difference to status quo, but having negative logic in tune-* files > feels a bit wrong. It's very nice to see ("real") thumb support being added to OE in general! But I guess, it's too bold a step to switch to it be the default. There're usual risk management issues (who knows what bugs CPUs/software have regarding thumb), plus, thumb has known drawback of reduced performance. In ideal case, thumb usage would be set per-package, with most of libraries being normal code, and most apps - thumb (and in this case thumb can be indeed the default). > If you need numbers: > http://ewi546.ewi.utwente.nl/tmp/thumb/comparison.txt (package: > ()) Using ipk sizes is of course pretty heuristical ;-). IMHO, those sizes are not too much convincing at all, but I'm sure raw exe sizes are better, and striving for that ideal case is worthy task ;-) > what do you all think? > regards, > Koen -- Best regards, Paul mailto:pmiscml@gmail.com