linux-kernel - Re: [PATCH v2 01/10] media: hantro: Fix H264 max frmsize supported on RK3288

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: <20191105085639.76a63a5a@collabora.com>
Date:   Tue, 5 Nov 2019 08:56:39 +0100
From:   Boris Brezillon <boris.brezillon@...labora.com>
To:     Hans Verkuil <hverkuil@...all.nl>
Cc:     Mauro Carvalho Chehab <mchehab+samsung@...nel.org>,
        Jonas Karlman <jonas@...boo.se>,
        Ezequiel Garcia <ezequiel@...labora.com>,
        Tomasz Figa <tfiga@...omium.org>,
        Philipp Zabel <p.zabel@...gutronix.de>,
        "linux-media@...r.kernel.org" <linux-media@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 01/10] media: hantro: Fix H264 max frmsize supported
 on RK3288

On Fri, 1 Nov 2019 09:36:55 +0100
Hans Verkuil <hverkuil@...all.nl> wrote:

> On 10/31/19 9:52 AM, Boris Brezillon wrote:
> > On Tue, 29 Oct 2019 01:24:47 +0000
> > Jonas Karlman <jonas@...boo.se> wrote:
> >   
> >> TRM specify supported image size 48x48 to 4096x2304 at step size 16 pixels,
> >> change frmsize max_width/max_height to match TRM at [1].
> >>
> >> This patch makes it possible to decode the 4096x2304 sample at [2].
> >>
> >> [1] http://www.t-firefly.com/download/firefly-rk3288/docs/TRM/rk3288-chapter-25-video-encoder-decoder-unit-(vcodec).pdf
> >> [2] https://4ksamples.com/puppies-bath-in-4k/
> >>
> >> Fixes: 760327930e10 ("media: hantro: Enable H264 decoding on rk3288")
> >> Signed-off-by: Jonas Karlman <jonas@...boo.se>  
> > 
> > Reviewed-by: Boris Brezillon <boris.brezillon@...labora.com>
> > Tested-by: Boris Brezillon <boris.brezillon@...labora.com>
> > 
> > Let's also add
> > 
> > Cc: <stable@...r.kernel.org>
> > 
> > just in case this patch doesn't make it to 5.4.
> > 
> >   
> >> ---
> >> Changes in v2:
> >>   - updated commit message with reference to TRM and sample video
> >> ---
> >>  drivers/staging/media/hantro/rk3288_vpu_hw.c | 4 ++--
> >>  1 file changed, 2 insertions(+), 2 deletions(-)
> >>
> >> diff --git a/drivers/staging/media/hantro/rk3288_vpu_hw.c b/drivers/staging/media/hantro/rk3288_vpu_hw.c
> >> index c0bdd6c02520..f8db6fcaad73 100644
> >> --- a/drivers/staging/media/hantro/rk3288_vpu_hw.c
> >> +++ b/drivers/staging/media/hantro/rk3288_vpu_hw.c
> >> @@ -67,10 +67,10 @@ static const struct hantro_fmt rk3288_vpu_dec_fmts[] = {
> >>  		.max_depth = 2,
> >>  		.frmsize = {
> >>  			.min_width = 48,
> >> -			.max_width = 3840,
> >> +			.max_width = 4096,
> >>  			.step_width = MB_DIM,
> >>  			.min_height = 48,
> >> -			.max_height = 2160,
> >> +			.max_height = 2304,
> >>  			.step_height = MB_DIM,  
> > 
> > Hans, Mauro, we were intending to have this fix merged in 5.4 or at
> > the very least be backported to the 5.4 stable branch at some point,
> > the problem is, this patch is based on media/master which contains the
> > s/MB_DIM_H264/MB_DIM/ change. I can send a new version based on
> > media/fixes, but that means Linus will have a conflict when merging the
> > media 5.5 PR in his tree. Are you fine dealing with this conflict
> > (letting Linus know about the expected resolution or backmerging the -rc
> > containing the fix in media/master so that he doesn't even have to deal
> > with it), or should we just let this patch go in media/master and
> > backport it later?  
> 
> Backport it later once it is merged in mainline.
> 
> This patch doesn't fix a bug, it is really an enhancement, so I think this
> can safely be delayed.

Okay, let's make things simple. Can we have patches 1 to 3 applied to
media/master? I'll take care of backporting those patches to 5.4 once
5.5-rc1 is out.

Thanks,

Boris