From mboxrd@z Thu Jan 1 00:00:00 1970 From: Daniel Vetter Subject: Re: [PATCH 05/12] drm/i915: Remove INTEL_INFO() Date: Sun, 26 Jan 2014 16:09:42 +0100 Message-ID: <20140126150942.GV9772@phenom.ffwll.local> References: <1389035849-20563-1-git-send-email-damien.lespiau@intel.com> <1389035849-20563-6-git-send-email-damien.lespiau@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mail-ee0-f42.google.com (mail-ee0-f42.google.com [74.125.83.42]) by gabe.freedesktop.org (Postfix) with ESMTP id A27D6FA240 for ; Sun, 26 Jan 2014 07:09:48 -0800 (PST) Received: by mail-ee0-f42.google.com with SMTP id e49so1787542eek.15 for ; Sun, 26 Jan 2014 07:09:46 -0800 (PST) Content-Disposition: inline In-Reply-To: <1389035849-20563-6-git-send-email-damien.lespiau@intel.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: intel-gfx-bounces@lists.freedesktop.org Errors-To: intel-gfx-bounces@lists.freedesktop.org To: Damien Lespiau Cc: intel-gfx@lists.freedesktop.org List-Id: intel-gfx@lists.freedesktop.org On Mon, Jan 06, 2014 at 07:17:22PM +0000, Damien Lespiau wrote: > Finally, we reach the goal of the last few patches: the removal of > INTEL_INFO() in favor of a direct dev_priv usage. > > Signed-off-by: Damien Lespiau Please pardon my incompetence, but I think I need to pull a heelturn on this one :( Apparently product groups noticed again that we have a unified driver across all platforms ever shipped and they don't like the space overhead that much. Since the usual approaches of splattering #ifdefs all over the code or (gasp!) editing all the sources files are complete no-goes for upstream we need something else. My idea would be to: - convert all HAS_FOO macros to used INTEL_INFO underneath. - allow hard-coding of INTEL_INFO structs. - enable LTE on i915 and let gcc's DCE path take care of all the dead code removal. Minimal overhead for upstream, still tidy code and a good chance to get product groups what they want. But it also means we'll need to keep INTEL_INFO around, hence why I'll nak patches 1-5. Patch 6 is merged with Paulo's r-b (it still applied). Ofc to_i915_dev is still really nice, so can I sheepishly ask you to resurrect that part of your patches and roll to_i915_dev a bit out over the tree? I think I'll owe you a few rounds of your $favourite_beverage the next time we meet for recompense, maybe at fosdem ... Cheers, Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch