DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com D8F9EC054905
Message-ID: <1492522793.27392.55.camel@redhat.com>
Subject: Re: [RfC PATCH] drm: fourcc byteorder: brings header file comments
 in line with reality.
From: Gerd Hoffmann <kraxel@redhat.com>
To: Pekka Paalanen <ppaalanen@gmail.com>
Cc: dri-devel@lists.freedesktop.org,
        open list <linux-kernel@vger.kernel.org>,
        amd-gfx@lists.freedesktop.org, Daniel Vetter <daniel.vetter@intel.com>,
        Ilia Mirkin <imirkin@alum.mit.edu>
Date: Tue, 18 Apr 2017 15:39:53 +0200
In-Reply-To: <20170418141827.11634103@eldfell>
References: <20170410101202.19229-1-kraxel@redhat.com>
         <20170410161214.305f5daf@eldfell> <1491833847.30990.77.camel@redhat.com>
         <20170410180941.43922e25@eldfell> <1492509617.27392.19.camel@redhat.com>
         <20170418141827.11634103@eldfell>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8BIT
Mime-Version: 1.0
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1375
Lines: 35

  Hi,

> > Historical note:  RHEL-6.9 (gnome 2) works fine.  Not of much interest
> > here, it drives the qemu stdvga with offb, not bochs-drm.
> 
> I suppose this proves the virtual machine itself is correct about
> framebuffer endianess? Except you are running it on a little-endian
> host machine I presume...

Yes, little endian host, qemu interprets the framebuffer as
PIXMAN_b8g8r8x8.  Which should be correct for a xrgb bigendian
framebuffer as pixman formats are native endian.

> > More interesting:  RHEL-7.3 (gnome 3.14) works fine too.  kernel 3.10,
> > but drm drivers updated to roughly 4.6 level.  Runs bochs-drm.  mesa
> > 11.2.2.  glamour not used.
> > 
> > Most recent:  Fedora 25 (gnome 3.22) looks mostly ok, but there are
> > rendering glitches, for example in the gnome activities screen (the one
> > you get when you press the windows key).  kernel 4.10, mesa 13.0.4.
> > glamor not used, but I think gnome-shell uses opengl (via llvmpipe) for
> > compositing.
> 
> I believe glitches are irrelevant for this topic, what we are
> interested in is if the colors are right or byte-swapped (also mind
> alpha/blue etc. swaps).

Well, I mean color glitches.  But it isn't consistent.  As if some
operations operate with the correct byteorder and some don't.
alpha/blue being swapped is a problem in some areas.

https://www.kraxel.org/tmp/

cheers,
  Gerd