From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chris Wilson Subject: Re: [Intel-gfx] [PATCH 05/21] drm/i915: Check eDP power when doing aux channel communications Date: Fri, 30 Sep 2011 19:23:44 +0100 Message-ID: References: <20110923085243.6e4b7b4c@jbarnes-x220> <1317344993-24945-1-git-send-email-keithp@keithp.com> <1317344993-24945-6-git-send-email-keithp@keithp.com> <20110930170242.GE2859@phenom.ffwll.local> Return-path: In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Keith Packard , Daniel Vetter Cc: Dave Airlie , intel-gfx@lists.freedesktop.org, linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org List-Id: dri-devel@lists.freedesktop.org On Fri, 30 Sep 2011 11:01:06 -0700, Keith Packard wrote: Non-text part: multipart/signed > On Fri, 30 Sep 2011 19:02:42 +0200, Daniel Vetter wrote: > > > Use pp_control instead of re-reading? > > Could, but you'll note a later patch eliminates both pp_status and > pp_control local variables, so I didn't bother to clean this up when > refactoring. > > > dp_aux_ch does the low-level io for the below, so either this one or the > > below three hunks look a bit redundant. > > Yeah, probably not necessary. I just added checks everywhere I could > think of to try and figure out where power was not being applied when > needed. > > Should I bother to include this patch in the series at all? It's purely > for diagnostics to make sure the panel is powered during all aux channel > transactions. Yes, adding the extra checks made my day. :) Modesetting bugs are the hardest to reproduce so every additional check (with a reasonable false positive rate) is worth their weight in gold when it comes to diagnosing misbehaving hardware. And the checks improve the maintainability of the code. -Chris -- Chris Wilson, Intel Open Source Technology Centre