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:   Tue, 18 Dec 2018 23:51:50 -0800
From:   Christoph Hellwig <hch@infradead.org>
To:     Tomasz Figa <tfiga@chromium.org>
Cc:     Christoph Hellwig <hch@infradead.org>,
        Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
        "Matwey V. Kornilov" <matwey@sai.msu.ru>,
        Linux Media Mailing List <linux-media@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        "Matwey V. Kornilov" <matwey.kornilov@gmail.com>,
        Alan Stern <stern@rowland.harvard.edu>,
        Ezequiel Garcia <ezequiel@collabora.com>, hdegoede@redhat.com,
        Hans Verkuil <hverkuil@xs4all.nl>,
        Mauro Carvalho Chehab <mchehab@kernel.org>,
        rostedt@goodmis.org, mingo@redhat.com,
        Mike Isely <isely@pobox.com>,
        Bhumika Goyal <bhumirks@gmail.com>,
        Colin King <colin.king@canonical.com>,
        Kieran Bingham <kieran.bingham@ideasonboard.com>,
        keiichiw@chromium.org
Subject: Re: [PATCH v5 2/2] media: usb: pwc: Don't use coherent DMA buffers
 for ISO transfer
Message-ID: <20181219075150.GA26656@infradead.org>
References: <20181212090917.GA30598@infradead.org>
 <CAAFQd5DhDOfk_2Dhq4MfJmoxpXP=Bm36HMZ55PSXxwkTAoCXSQ@mail.gmail.com>
 <20181212135440.GA6137@infradead.org>
 <CAAFQd5C4RbMxRP+ox+BDuApMusTD=WUD9Vs6aWL3u=HovuWUig@mail.gmail.com>
 <20181213140329.GA25339@infradead.org>
 <CAAFQd5BudF84jVaiy7KwevzBZnfYUZggDK=4W=g+Znf5VJjHsQ@mail.gmail.com>
 <20181214123624.GA5824@infradead.org>
 <CAAFQd5BnZHhNjOc6HKt=YVBVQFCcqN0RxAcfDp3S+gDKRvciqQ@mail.gmail.com>
 <20181218073847.GA4552@infradead.org>
 <CAAFQd5AT3ixnbZRm3TOjoWrk2UNH0bXqgR+Z8wyjMhr0xHtSOg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CAAFQd5AT3ixnbZRm3TOjoWrk2UNH0bXqgR+Z8wyjMhr0xHtSOg@mail.gmail.com>
User-Agent: Mutt/1.9.2 (2017-12-15)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Dec 18, 2018 at 06:48:03PM +0900, Tomasz Figa wrote:
> > So as I said you can call dma_alloc_attrs with DMA_ATTR_NON_CONSISTENT
> > in a loop with a suitably small chunk size, then stuff the results into
> > a scatterlist and map that again for the device share with if you don't
> > want a single contigous region.  You just have to either deal with
> > non-contigous access from the kernel or use vmap and the right vmap
> > cache flushing helpers.
> 
> The point is that you didn't have to do this small chunk loop without
> DMA_ATTR_NON_CONSISTENT, so it's at least inconsistent now and not
> sure why it could be better than just a loop of alloc_page().

You have to do it if you want to map the addresses for a second device.

> > I would advice against new non-consistent users until this series
> > goes through, mostly because dma_cache_sync is such an amazing bad
> > API.  Otherwise things will just work at the allocation side, you'll
> > just need to be careful to transfer ownership between the cpu and
> > the device(s) carefully using the dma_sync_* APIs.
> 
> Just to clarify, the actual code isn't very likely to surface any time
> soon. so I assume it would be after this series lands.
> 
> We will however need an API that can transparently handle both cases
> of contiguous (without IOMMU) and page-by-page allocations (with
> IOMMU) behind the scenes, like the current dma_alloc_attrs() without
> DMA_ATTR_NON_CONSISTENT.

Is the use case to then share the memory between multiples devices
or just for a single device?  The latter case is generally easy, the
former is rather more painful.