From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from uplift.swm.pp.se (swm.pp.se [212.247.200.143]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by huchra.bufferbloat.net (Postfix) with ESMTPS id BE2F621F378 for ; Tue, 17 Nov 2015 04:01:56 -0800 (PST) Received: by uplift.swm.pp.se (Postfix, from userid 501) id 1EC8BA2; Tue, 17 Nov 2015 13:01:53 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=swm.pp.se; s=mail; t=1447761713; bh=7ufeOlddph5zySXR8oHd35hZozvoiXeRDOYhw22z7Y8=; h=Date:From:To:cc:Subject:In-Reply-To:References:From; b=xb+nAKnZ1q8t4kIu68wU9umtuAm48ZHQaC8Os9lCGz1Q3NM1QfL6N1Grb1RSl1FZY JFc6EKwSq29Lux3baEvkFOs/qD2gLNPzo3ZJz80HeazvxUIpO8Gw55mhkMu11ZnZZZ 535fv2fWfArEChcMnci04q7lccIyac/lN/CexCp8= Received: from localhost (localhost [127.0.0.1]) by uplift.swm.pp.se (Postfix) with ESMTP id 15FA5A1; Tue, 17 Nov 2015 13:01:53 +0100 (CET) Date: Tue, 17 Nov 2015 13:01:53 +0100 (CET) From: Mikael Abrahamsson To: Dave Taht In-Reply-To: Message-ID: References: User-Agent: Alpine 2.02 (DEB 1266 2009-07-14) Organization: People's Front Against WWW MIME-Version: 1.0 Content-Type: TEXT/PLAIN; format=flowed; charset=US-ASCII Cc: "cerowrt-devel@lists.bufferbloat.net" Subject: Re: [Cerowrt-devel] linksys vfp vs eabi? X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: Development issues regarding the cerowrt test router project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 17 Nov 2015 12:02:20 -0000 On Mon, 16 Nov 2015, Dave Taht wrote: > My build of the linksys 1200 code is coming up with the eabi which is > less efficient than the hard float alternative (armhf or vfp), Are > people building for the eabi? > > http://www.openwall.com/lists/musl/2015/10/15/2 is the only reference > I've found to this issue. I don't know if this is related, but I remember I found something that made the openssl benchmark run a lot slower, I found a patch that basically disabled some hardware function for most ARM architectures because of some vague bug. I reverted the patch for the Armada architecture and got back the speed... I believe it's this one I was mucking around with: https://dev.openwrt.org/ticket/19420 -- Mikael Abrahamsson email: swmike@swm.pp.se