From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chris Wilson Subject: Re: [PATCH 13/13 v4] drm/i915/intel_i2c: reuse GMBUS2 value read in polling loop Date: Wed, 28 Mar 2012 12:44:59 +0100 Message-ID: <1332935113_126617@CP5-2952> References: <1332873382-29373-1-git-send-email-djkurtz@chromium.org> <1332873382-29373-14-git-send-email-djkurtz@chromium.org> <1332874994_117123@CP5-2952> Return-path: In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Daniel Kurtz Cc: Daniel Vetter , Keith Packard , David Airlie , dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org, Benson Leung , Yufeng Shen List-Id: dri-devel@lists.freedesktop.org On Wed, 28 Mar 2012 19:39:17 +0800, Daniel Kurtz wrote: > On Wed, Mar 28, 2012 at 3:02 AM, Chris Wilson wrote: > > Might as well shave this read as well. > > Do you know why POSTING_READ() was there in the first place? > As far as I can tell, these are used to ensure memory barriers are > inserted between a group of writes, and subsequent reads to memory > mapped io registers. > However, the normal I915_READ() and I915_WRITE() macros already call > readl() / writel(), which already have an explicit mb(). > So, can we just get rid of all of them, or am I missing something? They can go. They were there just as paranoia to make sure the writes were flushed before any timing delays and across loops. Once the code settled I never took the liberty of removing them. -Chris -- Chris Wilson, Intel Open Source Technology Centre