Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <9f5cac46e38897f54c5246bf400e8888324abad8.camel@gmail.com>
Subject: Re: [PATCH v3 07/11] iio: core: Add new DMABUF interface
 infrastructure
From:   Nuno =?ISO-8859-1?Q?S=E1?= <noname.nuno@gmail.com>
To:     Paul Cercueil <paul@crapouillou.net>,
        Jonathan Cameron <jic23@kernel.org>,
        Lars-Peter Clausen <lars@metafoo.de>,
        Vinod Koul <vkoul@kernel.org>,
        Michael Hennerich <Michael.Hennerich@analog.com>,
        Sumit Semwal <sumit.semwal@linaro.org>,
        Christian =?ISO-8859-1?Q?K=F6nig?= <christian.koenig@amd.com>
Cc:     linux-iio@vger.kernel.org, linux-kernel@vger.kernel.org,
        dri-devel@lists.freedesktop.org, linaro-mm-sig@lists.linaro.org,
        dmaengine@vger.kernel.org, linux-media@vger.kernel.org
Date:   Tue, 04 Apr 2023 10:21:27 +0200
In-Reply-To: <2dac030470ffe74b6d21a1e6510afcefaf58cd6a.camel@crapouillou.net>
References: <20230403154800.215924-1-paul@crapouillou.net>
         <20230403154800.215924-8-paul@crapouillou.net>
         <798e1ff0651da8e4b113d30bf8cec2a7a0e6898f.camel@gmail.com>
         <2dac030470ffe74b6d21a1e6510afcefaf58cd6a.camel@crapouillou.net>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.46.4 
MIME-Version: 1.0
Precedence: bulk

On Tue, 2023-04-04 at 09:55 +0200, Paul Cercueil wrote:
> Hi Nuno,
>=20
> Le mardi 04 avril 2023 =C3=A0 09:32 +0200, Nuno S=C3=A1 a =C3=A9crit=C2=
=A0:
> > On Mon, 2023-04-03 at 17:47 +0200, Paul Cercueil wrote:
> > > Add the necessary infrastructure to the IIO core to support a new
> > > optional DMABUF based interface.
> > >=20
> > > With this new interface, DMABUF objects (externally created) can be
> > > attached to a IIO buffer, and subsequently used for data transfer.
> > >=20
> > > A userspace application can then use this interface to share DMABUF
> > > objects between several interfaces, allowing it to transfer data in
> > > a
> > > zero-copy fashion, for instance between IIO and the USB stack.
> > >=20
> > > The userspace application can also memory-map the DMABUF objects,
> > > and
> > > access the sample data directly. The advantage of doing this vs.
> > > the
> > > read() interface is that it avoids an extra copy of the data
> > > between
> > > the
> > > kernel and userspace. This is particularly userful for high-speed
> > > devices which produce several megabytes or even gigabytes of data
> > > per
> > > second.
> > >=20
> > > As part of the interface, 3 new IOCTLs have been added:
> > >=20
> > > IIO_BUFFER_DMABUF_ATTACH_IOCTL(int fd):
> > > =C2=A0Attach the DMABUF object identified by the given file descripto=
r
> > > to
> > > the
> > > =C2=A0buffer.
> > >=20
> > > IIO_BUFFER_DMABUF_DETACH_IOCTL(int fd):
> > > =C2=A0Detach the DMABUF object identified by the given file descripto=
r
> > > from
> > > =C2=A0the buffer. Note that closing the IIO buffer's file descriptor
> > > will
> > > =C2=A0automatically detach all previously attached DMABUF objects.
> > >=20
> > > IIO_BUFFER_DMABUF_ENQUEUE_IOCTL(struct iio_dmabuf *):
> > > =C2=A0Request a data transfer to/from the given DMABUF object. Its fi=
le
> > > =C2=A0descriptor, as well as the transfer size and flags are provided=
 in
> > > the
> > > =C2=A0"iio_dmabuf" structure.
> > >=20
> > > These three IOCTLs have to be performed on the IIO buffer's file
> > > descriptor, obtained using the IIO_BUFFER_GET_FD_IOCTL() ioctl.
> > >=20
> > > Signed-off-by: Paul Cercueil <paul@crapouillou.net>
> > >=20
> > > ---
> > > v2: Only allow the new IOCTLs on the buffer FD created with
> > > =C2=A0=C2=A0=C2=A0 IIO_BUFFER_GET_FD_IOCTL().
> > >=20
> > > v3: - Get rid of the old IOCTLs. The IIO subsystem does not create
> > > or
> > > =C2=A0=C2=A0=C2=A0 manage DMABUFs anymore, and only attaches/detaches=
 externally
> > > =C2=A0=C2=A0=C2=A0 created DMABUFs.
> > > =C2=A0=C2=A0=C2=A0 - Add IIO_BUFFER_DMABUF_CYCLIC to the supported fl=
ags.
> > > ---
> > > =C2=A0drivers/iio/industrialio-buffer.c | 402
> > > ++++++++++++++++++++++++++++++
> > > =C2=A0include/linux/iio/buffer_impl.h=C2=A0=C2=A0 |=C2=A0 22 ++
> > > =C2=A0include/uapi/linux/iio/buffer.h=C2=A0=C2=A0 |=C2=A0 22 ++
> > > =C2=A03 files changed, 446 insertions(+)
> > >=20
> > > diff --git a/drivers/iio/industrialio-buffer.c
> > > b/drivers/iio/industrialio-buffer.c
> > > index 80c78bd6bbef..5d88e098b3e7 100644
> > > --- a/drivers/iio/industrialio-buffer.c
> > > +++ b/drivers/iio/industrialio-buffer.c
> > > @@ -13,10 +13,14 @@
> > > =C2=A0#include <linux/kernel.h>
> > > =C2=A0#include <linux/export.h>
> > > =C2=A0#include <linux/device.h>
> > > +#include <linux/dma-buf.h>
> > > +#include <linux/dma-fence.h>
> > > +#include <linux/dma-resv.h>
> > > =C2=A0#include <linux/file.h>
> > > =C2=A0#include <linux/fs.h>
> > > =C2=A0#include <linux/cdev.h>
> > > =C2=A0#include <linux/slab.h>
> > > +#include <linux/mm.h>
> > > =C2=A0#include <linux/poll.h>
> > > =C2=A0#include <linux/sched/signal.h>
> > > =C2=A0
> > > @@ -28,11 +32,41 @@
> > > =C2=A0#include <linux/iio/buffer.h>
> > > =C2=A0#include <linux/iio/buffer_impl.h>
> > > =C2=A0
> > > +#define DMABUF_ENQUEUE_TIMEOUT_MS 5000
> > > +
> > > +struct iio_dma_fence;
> > > +
> > > +struct iio_dmabuf_priv {
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct list_head entry;
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct kref ref;
> > > +
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct iio_buffer *buffer;
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct iio_dma_buffer_bloc=
k *block;
> > > +
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0u64 context;
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0spinlock_t lock;
> > > +
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct dma_buf_attachment =
*attach;
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct iio_dma_fence *fenc=
e;
> > > +};
> > > +
> > > +struct iio_dma_fence {
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct dma_fence base;
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct iio_dmabuf_priv *pr=
iv;
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0struct sg_table *sgt;
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0enum dma_data_direction di=
r;
> > > +};
> > > +
> > > =C2=A0static const char * const iio_endian_prefix[] =3D {
> > > =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0[IIO_BE] =3D "be",
> > > =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0[IIO_LE] =3D "le",
> > > =C2=A0};
> > > =C2=A0
> > > +static inline struct iio_dma_fence *to_iio_dma_fence(struct
> > > dma_fence *fence)
> > > +{
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0return container_of(fence,=
 struct iio_dma_fence, base);
> > > +}
> > > +
> >=20
> > Kind of a nitpick but I only see this being used once so I would
> > maybe
> > use plain 'container_of()' as you are already doing for:
> >=20
> > ... =3D container_of(ref, struct iio_dmabuf_priv, ref);
> >=20
> > So I would at least advocate for consistency. I would also probably
> > ditch the inline but I guess that is more a matter of
> > style/preference.
>=20
> Yep, at least it should be consistent.
>=20
> >=20
> > > =C2=A0static bool iio_buffer_is_active(struct iio_buffer *buf)
> > > =C2=A0{
> > > =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0return !list_empty(&b=
uf->buffer_list);
> > > @@ -329,6 +363,7 @@ void iio_buffer_init(struct iio_buffer *buffer)
> > > =C2=A0{
> > >=20
> >=20
> > ...
> >=20
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0priv =3D attach->importer_=
priv;
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0list_del_init(&priv->entry=
);
> > > +
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0iio_buffer_dmabuf_put(atta=
ch);
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0iio_buffer_dmabuf_put(atta=
ch);
> > > +
> >=20
> > Is this intended? Looks suspicious...
>=20
> It is intended, yes. You want to release the dma_buf_attachment that's
> created in iio_buffer_attach_dmabuf(), and you need to call
> iio_buffer_find_attachment() to get a pointer to it, which also gets a
> second reference - so it needs to unref twice.
>=20

I see..

...


> >=20
> > > +out_dmabuf_put:
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0dma_buf_put(dmabuf);
> > > +
> > > +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0return ret;
> > > +}
> > >=20
> >=20
> > Hmmm, what about the legacy buffer? We should also support this
> > interface using it, right? Otherwise, using one of the new IOCTL in
> > iio_device_buffer_ioctl() (or /dev/iio:device0) will error out.
>=20
> According to Jonathan the old chardev route is deprecated, and it's
> fine not to support the IOCTL there.
>=20

Oh, alright then... Better that way indeed!

- Nuno S=C3=A1