Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Wed, 22 Jan 2020 10:39:37 +0900
From:   Sergey Senozhatsky <senozhatsky@chromium.org>
To:     Hans Verkuil <hverkuil@xs4all.nl>
Cc:     Sergey Senozhatsky <senozhatsky@chromium.org>,
        Hans Verkuil <hans.verkuil@cisco.com>,
        Tomasz Figa <tfiga@chromium.org>,
        Mauro Carvalho Chehab <mchehab@kernel.org>,
        Kyungmin Park <kyungmin.park@samsung.com>,
        Marek Szyprowski <m.szyprowski@samsung.com>,
        Sakari Ailus <sakari.ailus@iki.fi>,
        Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
        Pawel Osciak <posciak@chromium.org>,
        linux-media@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [RFC][PATCH 02/15] videobuf2: handle V4L2 buffer cache flags
Message-ID: <20200122013937.GC149602@google.com>
References: <20191217032034.54897-1-senozhatsky@chromium.org>
 <20191217032034.54897-3-senozhatsky@chromium.org>
 <ada2381c-2c1c-17c3-c190-48439ae1657a@xs4all.nl>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <ada2381c-2c1c-17c3-c190-48439ae1657a@xs4all.nl>
User-Agent: Mutt/1.10.1 (2018-07-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi,

Sorry for the delayed.

On (20/01/10 11:24), Hans Verkuil wrote:
> On 12/17/19 4:20 AM, Sergey Senozhatsky wrote:
> > Set video buffer cache management flags corresponding to V4L2 cache
> > flags.
> > 
> > Both ->prepare() and ->finish() cache management hints should be
> > passed during this stage (buffer preparation), because there is no
> > other way for user-space to skip ->finish() cache flush.
> > 
> > There are two possible alternative approaches:
> > - The first one is to move cache sync from ->finish() to dqbuf().
> >   But this breaks some drivers, that need to fix-up buffers before
> >   dequeueing them.
> > 
> > - The second one is to move ->finish() call from ->done() to dqbuf.
> 
> Please combine this patch with patch 13/15!

OK.

[..]
> >  }
> >  
> > +static void set_buffer_cache_hints(struct vb2_queue *q,
> > +				   struct vb2_buffer *vb,
> > +				   struct v4l2_buffer *b)
> > +{
> > +	vb->need_cache_sync_on_prepare = 1;
> > +
> > +	if (q->dma_dir != DMA_TO_DEVICE)
> 
> What should be done when dma_dir == DMA_BIDIRECTIONAL?

Well, ->need_cache_sync_on_prepare/->need_cache_sync_on_finish are set,
by default. If the queue supports user-space cache hints (driver has
set ->allow_cache_hints), then user-space can override cache behavior.
We probably cannot enforce any other behavior here. Am I missing
something?

> > +		vb->need_cache_sync_on_finish = 1;
> > +	else
> > +		vb->need_cache_sync_on_finish = 0;
> > +
> > +	if (!q->allow_cache_hints)
> > +		return;
> > +
> > +	if (b->flags & V4L2_BUF_FLAG_NO_CACHE_INVALIDATE)
> > +		vb->need_cache_sync_on_finish = 0;
> > +
> > +	if (b->flags & V4L2_BUF_FLAG_NO_CACHE_CLEAN)
> > +		vb->need_cache_sync_on_prepare = 0;
> > +}

	-ss