From: Sergey Senozhatsky <senozhatsky@chromium.org>
To: Tomasz Figa <tfiga@chromium.org>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>,
Hans Verkuil <hverkuil@xs4all.nl>,
Hans Verkuil <hans.verkuil@cisco.com>,
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 Mailing List <linux-media@vger.kernel.org>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [RFC][PATCH 13/15] videobuf2: do not sync buffers for DMABUF queues
Date: Tue, 28 Jan 2020 16:34:55 +0900 [thread overview]
Message-ID: <20200128073455.GC115889@google.com> (raw)
In-Reply-To: <CAAFQd5Dx50nt=yOCj-ufDDey=Zur_F5yE34a9OQP630A1opBSQ@mail.gmail.com>
On (20/01/28 16:22), Tomasz Figa wrote:
[..]
> > > > Then, if q->allow_cache_hint is set, then default to a cache sync (cache clean)
> > > > in the prepare for OUTPUT buffers and a cache sync (cache invalidate) in the
> > > > finish for CAPTURE buffers.
> > >
> > > We alter default cache sync behaviour based both on queue ->memory
> > > type and queue ->dma_dir. Shall both of those cases depend on
> > > ->allow_cache_hints, or q->memory can be independent?
> > >
> > > static void set_buffer_cache_hints(struct vb2_queue *q,
> > > struct vb2_buffer *vb,
> > > struct v4l2_buffer *b)
> > > {
> > > if (!q->allow_cache_hints)
> > > return;
> > >
> > > /*
> > > * DMA exporter should take care of cache syncs, so we can avoid
> > > * explicit ->prepare()/->finish() syncs. For other ->memory types
> > > * we always need ->prepare() or/and ->finish() cache sync.
> > > */
> > > if (q->memory == VB2_MEMORY_DMABUF) {
> > > vb->need_cache_sync_on_finish = 0;
> > > vb->need_cache_sync_on_prepare = 0;
> > > return;
> > > }
> >
> > I personally would prefer this to be above ->allow_cache_hints check,
> > IOW to be independent. Because we need to skip flush/invalidation for
> > DMABUF anyway, that's what allocators do in ->prepare() and ->finish()
> > currently.
>
> I think it's even more than personal preference. We shouldn't even
> attempt to sync imported DMA-bufs, so the code above may result in
> incorrect behavior.
Sure. Allocators test buf->db_attach in prepare()/finish(). This patchset
removes that logic, because we move it to the core code. But if the
conclusion will be to clear ->need_cache_sync_on_finish/prepare only
when ->allow_cache_hints was set, then buf->db_attach bail out must stay
in allocators. And this is where I have preferences :)
-ss
next prev parent reply other threads:[~2020-01-28 7:35 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-17 3:20 [RFC][PATCH 00/15] Implement V4L2_BUF_FLAG_NO_CACHE_* flags Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 01/15] videobuf2: add cache management members Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 02/15] videobuf2: handle V4L2 buffer cache flags Sergey Senozhatsky
2020-01-10 10:24 ` Hans Verkuil
2020-01-22 1:39 ` Sergey Senozhatsky
2020-01-22 2:53 ` Sergey Senozhatsky
2020-01-28 4:35 ` Tomasz Figa
2019-12-17 3:20 ` [RFC][PATCH 03/15] videobuf2: add V4L2_FLAG_MEMORY_NON_CONSISTENT flag Sergey Senozhatsky
2020-01-10 9:36 ` Hans Verkuil
2020-01-10 9:46 ` Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 04/15] videobuf2: add queue memory consistency parameter Sergey Senozhatsky
2020-01-10 9:47 ` Hans Verkuil
2020-01-22 2:05 ` Sergey Senozhatsky
2020-01-23 11:02 ` Hans Verkuil
2020-01-24 2:04 ` Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 05/15] videobuf2: handle V4L2_FLAG_MEMORY_NON_CONSISTENT in REQBUFS Sergey Senozhatsky
2020-01-10 9:55 ` Hans Verkuil
2020-01-22 2:18 ` Sergey Senozhatsky
2020-01-22 3:48 ` Sergey Senozhatsky
2020-01-23 11:08 ` Hans Verkuil
2020-01-28 4:45 ` Tomasz Figa
2020-01-28 8:38 ` Hans Verkuil
2019-12-17 3:20 ` [RFC][PATCH 06/15] videobuf2: handle V4L2_FLAG_MEMORY_NON_CONSISTENT in CREATE_BUFS Sergey Senozhatsky
2020-01-10 9:59 ` Hans Verkuil
2020-01-23 3:41 ` Sergey Senozhatsky
2020-01-23 11:41 ` Hans Verkuil
2020-01-24 1:28 ` Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 07/15] videobuf2: factor out planes prepare/finish functions Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 08/15] videobuf2: do not sync caches when we are allowed not to Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 09/15] videobuf2: check ->synced flag in prepare() and finish() Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 10/15] videobuf2: let user-space know when driver supports cache hints Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 11/15] videobuf2: add begin/end cpu_access callbacks to dma-contig Sergey Senozhatsky
2019-12-17 3:20 ` [RFC][PATCH 12/15] videobuf2: add begin/end cpu_access callbacks to dma-sg Sergey Senozhatsky
2020-01-10 10:13 ` Hans Verkuil
2020-01-22 6:37 ` Sergey Senozhatsky
2020-01-28 4:38 ` Tomasz Figa
2020-01-28 8:36 ` Hans Verkuil
2020-01-30 11:02 ` Tomasz Figa
2020-01-30 12:18 ` Hans Verkuil
2020-02-03 10:04 ` Tomasz Figa
2020-02-04 2:50 ` Sergey Senozhatsky
2020-02-06 8:51 ` Tomasz Figa
2019-12-17 3:20 ` [RFC][PATCH 13/15] videobuf2: do not sync buffers for DMABUF queues Sergey Senozhatsky
2020-01-10 10:30 ` Hans Verkuil
2020-01-22 5:05 ` Sergey Senozhatsky
2020-01-23 11:35 ` Hans Verkuil
2020-01-24 2:25 ` Sergey Senozhatsky
2020-01-24 7:32 ` Sergey Senozhatsky
2020-01-28 7:22 ` Tomasz Figa
2020-01-28 7:34 ` Sergey Senozhatsky [this message]
2020-01-28 7:19 ` Tomasz Figa
2020-01-28 8:47 ` Hans Verkuil
2019-12-17 3:20 ` [RFC][PATCH 14/15] videobuf2: don't test db_attach in dma-contig prepare and finish Sergey Senozhatsky
2020-01-10 10:32 ` Hans Verkuil
2019-12-17 3:20 ` [RFC][PATCH 15/15] videobuf2: don't test db_attach in dma-sg " Sergey Senozhatsky
2020-01-08 2:27 ` [RFC][PATCH 00/15] Implement V4L2_BUF_FLAG_NO_CACHE_* flags Sergey Senozhatsky
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200128073455.GC115889@google.com \
--to=senozhatsky@chromium.org \
--cc=hans.verkuil@cisco.com \
--cc=hverkuil@xs4all.nl \
--cc=kyungmin.park@samsung.com \
--cc=laurent.pinchart@ideasonboard.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-media@vger.kernel.org \
--cc=m.szyprowski@samsung.com \
--cc=mchehab@kernel.org \
--cc=posciak@chromium.org \
--cc=sakari.ailus@iki.fi \
--cc=tfiga@chromium.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).