linux-kernel - Re: [PATCH v1.1 2/2] drm: rcar-du: Repair vblank for DRM page flips using the VSP1

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <1681295.5XGp1sIbmj@avalon>
Date:   Sun, 02 Jul 2017 17:23:24 +0300
From:   Laurent Pinchart <laurent.pinchart@...asonboard.com>
To:     Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>
Cc:     Kieran Bingham <kieran.bingham+renesas@...asonboard.com>,
        linux-renesas-soc@...r.kernel.org, linux-media@...r.kernel.org,
        dri-devel@...ts.freedesktop.org, David Airlie <airlied@...ux.ie>,
        Mauro Carvalho Chehab <mchehab@...nel.org>,
        open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v1.1 2/2] drm: rcar-du: Repair vblank for DRM page flips using the VSP1

Hi Morimoto-san,

On Friday 30 Jun 2017 08:32:04 Kuninori Morimoto wrote:
> Hi Kieran
> 
> > -static void rcar_du_vsp_complete(void *private)
> > +static void rcar_du_vsp_complete(void *private, bool completed)
> >  {
> >  	struct rcar_du_crtc *crtc = private;
> > 
> > -	rcar_du_crtc_finish_page_flip(crtc);
> > +	if (crtc->vblank_enable)
> > +		drm_crtc_handle_vblank(&crtc->crtc);
> > +
> > +	if (completed)
> > +		rcar_du_crtc_finish_page_flip(crtc);
> >  }
> 
> Here, this "vblank_enable" flag, timestamp will be update on
> drm_crtc_handle_vblank().
> 
> For example modetest Flip test, if we stop it by Ctrl+C, then, vblank_enable
> will be false, Then, vblank timestamp isn't updated on waiting method on
> drm_atomic_helper_wait_for_vblanks(). Thus we will have timeout error.

I've noticed this issue as well when testing Kieran's patch, and I will fix 
it.

> And, print complete is now indicated on VSP Frame End,
> in interlace input case, print complete will be indicated to user
> on each ODD, EVEN timing.
> 
> Before this patch, for example 1080i@...z, print complete indication
> happen in 30Hz.
> After this patch, in interlace case, indication coming 60Hz

Isn't this to be expected ? In 1080i@...z the frame rate is 60 frames per 
second, so shouldn't vertical blanking be reported at 60Hz ?

-- 
Regards,

Laurent Pinchart