Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1NMz1v-0002Pg-44 for qemu-devel@nongnu.org; Tue, 22 Dec 2009 02:19:39 -0500 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1NMz1t-0002PE-Q5 for qemu-devel@nongnu.org; Tue, 22 Dec 2009 02:19:38 -0500 Received: from [199.232.76.173] (port=40972 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1NMz1t-0002P8-Iy for qemu-devel@nongnu.org; Tue, 22 Dec 2009 02:19:37 -0500 Received: from hall.aurel32.net ([88.191.82.174]:49040) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1NMz1t-0007pL-9E for qemu-devel@nongnu.org; Tue, 22 Dec 2009 02:19:37 -0500 Date: Tue, 22 Dec 2009 08:19:30 +0100 From: Aurelien Jarno Subject: Re: [Qemu-devel] [PATCH 0/5] tcg conditional set, round 4 Message-ID: <20091222071930.GA31094@hall.aurel32.net> References: <200912202257.22503.paul@codesourcery.com> <4B2ED6D0.1010900@twiddle.net> <761ea48b0912210113p42608776q7677f79eec83f5a6@mail.gmail.com> <4B2FDA53.6040605@twiddle.net> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-15 Content-Disposition: inline In-Reply-To: <4B2FDA53.6040605@twiddle.net> Sender: Aurelien Jarno List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Richard Henderson Cc: Laurent Desnogues , Paul Brook , qemu-devel@nongnu.org On Mon, Dec 21, 2009 at 12:28:03PM -0800, Richard Henderson wrote: > On 12/21/2009 01:13 AM, Laurent Desnogues wrote: >> The question for the generalized movcond is how useful is it? >> Which front-ends would need it and would the cost to generate >> code for it on some (most?) back-ends be amortized? > > ... Any front end that has a conditional move instruction? > Sparcv9, Mips32, Alpha, ARM... > > That said, I think the *biggest* gains are to be had because with ^^^^^ This word is the main reason why I am not convinced at all by movcond. > movcond -- at least on some targets -- we can have one BB per TB, and > avoid any intermediate spilling of global registers back to memory. Memory load/stores are far more frequent, and also spill global registers back to memory. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurelien@aurel32.net http://www.aurel32.net