From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:57363) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1TiCKA-0007jr-Of for qemu-devel@nongnu.org; Mon, 10 Dec 2012 17:59:47 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1TiCK7-0000UO-Lq for qemu-devel@nongnu.org; Mon, 10 Dec 2012 17:59:46 -0500 Received: from cantor2.suse.de ([195.135.220.15]:48337 helo=mx2.suse.de) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1TiCK7-0000U1-Bi for qemu-devel@nongnu.org; Mon, 10 Dec 2012 17:59:43 -0500 From: =?UTF-8?q?Andreas=20F=C3=A4rber?= Date: Mon, 10 Dec 2012 23:59:30 +0100 Message-Id: <1355180372-6525-1-git-send-email-afaerber@suse.de> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: [Qemu-devel] [RFC qom-cpu v2 0/2] target-i386: X86CPU subclasses List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: blauwirbel@gmail.com, imammedo@redhat.com, ehabkost@redhat.com, anthony@codemonkey.ws, =?UTF-8?q?Andreas=20F=C3=A4rber?= Hello, Had some long train rides today, so here's a new throw at X86CPU subclass= es: After having initially proposed a "minimal" conversion by reusing the list/array of CPU definitions, public opinion seems to have shifted from preferring initfns to preferring class_inits. Combined with per-type Type= Infos this allows to create custom hierarchies to share data among CPU families= . Thereby I do what I have long tried to avoid: touching almost every line = of x86 CPU definitions with risk of merge conflicts. This conversion was don= e 100% manually and does not touch the feature flags apart from the assignm= ent and semicolon, hopefully manageable for review; otherwise suggestions wel= come. Although in this series only the CPU properties that I had prepared earli= er for this purpose get used, CPU feature properties still make sense for management tools to inspect and modify bits of a CPU just created. This mini-series is based on my latest cleanup proposals. Regards, Andreas v1-> v2: * Instead of turning x86_def_t into X86CPUInfo to initialize classes, drop it completely and register types manually with customizable TypeIn= fos * Use new list facilities for printing -cpu ? models * Adopt new name scheme suggested by Eduardo and ideas from my alpha seri= es * Keep short names in -cpu ? output for alignment reasons * Merge cpu_x86_init() into cpu.c:cpu_x86_register() * Append patch showing Haswell as subclass of SandyBridge Cc: Eduardo Habkost Cc: Igor Mammedov Cc: Blue Swirl Cc: Anthony Liguori Andreas F=C3=A4rber (2): target-i386: Convert CPU definitions into X86CPU subclasses target-i386: Turn Haswell into subclass of SandyBridge target-i386/cpu-qom.h | 17 + target-i386/cpu.c | 1475 +++++++++++++++++++++++++++++--------------= ------ target-i386/helper.c | 24 - 3 Dateien ge=C3=A4ndert, 898 Zeilen hinzugef=C3=BCgt(+), 618 Zeilen entf= ernt(-) --=20 1.7.10.4