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:   Fri, 7 Dec 2018 07:25:02 -0800
From:   Christoph Hellwig <hch@infradead.org>
To:     Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc:     "Matwey V. Kornilov" <matwey@sai.msu.ru>,
        linux-media@vger.kernel.org, linux-kernel@vger.kernel.org,
        matwey.kornilov@gmail.com, tfiga@chromium.org,
        stern@rowland.harvard.edu, ezequiel@collabora.com,
        hdegoede@redhat.com, hverkuil@xs4all.nl, mchehab@kernel.org,
        rostedt@goodmis.org, mingo@redhat.com, isely@pobox.com,
        bhumirks@gmail.com, colin.king@canonical.com,
        kieran.bingham@ideasonboard.com, keiichiw@chromium.org,
        Christoph Hellwig <hch@infradead.org>
Subject: Re: [PATCH v5 2/2] media: usb: pwc: Don't use coherent DMA buffers
 for ISO transfer
Message-ID: <20181207152502.GA30455@infradead.org>
References: <20180821170629.18408-1-matwey@sai.msu.ru>
 <20180821170629.18408-3-matwey@sai.msu.ru>
 <2213616.rQm4DhIJ7U@avalon>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <2213616.rQm4DhIJ7U@avalon>
User-Agent: Mutt/1.9.2 (2017-12-15)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Folks, can you take a look at this tree and see if this is useful
for USB:

http://git.infradead.org/users/hch/misc.git/shortlog/refs/heads/dma-noncoherent-allocator

The idea is that you use dma_alloc_attrs with the DMA_ATTR_NON_CONSISTENT
now that I've made sure it is avaiable everywhere [1], and we can use
dma_sync_single_* on it.

The only special case USB will need are the HCD_LOCAL_MEM devices, for
which we must use dma_alloc_coherent (or dma_alloc_attrs without
DMA_ATTR_NON_CONSISTENT) and must skip the dma_sync_single_* calls,
so we'll probably need USB subsystem wrappers for those calls.

[1] except powerpc in this tree - I have another series to make powerpc
use the generic dma noncoherent code, which would cover it.