From: Tomasz Figa <tfiga@chromium.org>
To: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Benjamin Gaignard <benjamin.gaignard@collabora.com>,
m.szyprowski@samsung.com, mchehab@kernel.org, ming.qian@nxp.com,
shijie.qin@nxp.com, eagle.zhou@nxp.com, bin.liu@mediatek.com,
matthias.bgg@gmail.com, angelogioacchino.delregno@collabora.com,
tiffany.lin@mediatek.com, andrew-ct.chen@mediatek.com,
yunfei.dong@mediatek.com, stanimir.k.varbanov@gmail.com,
quic_vgarodia@quicinc.com, agross@kernel.org,
andersson@kernel.org, konrad.dybcio@linaro.org,
ezequiel@vanguardiasur.com.ar, p.zabel@pengutronix.de,
daniel.almeida@collabora.com, hverkuil-cisco@xs4all.nl,
jernel@kernel.org, linux-media@vger.kernel.org,
linux-kernel@vger.kernel.org,
linux-arm-kernel@lists.infradead.org,
linux-mediatek@lists.infradead.org,
linux-arm-msm@vger.kernel.org,
linux-rockchip@lists.infradead.org, kernel@collabora.com
Subject: Re: [PATCH v2 2/8] media: videobuf2: Make bufs array dynamic allocated
Date: Fri, 19 May 2023 10:04:53 +0000 [thread overview]
Message-ID: <20230519100453.6o4eriisloiji32g@chromium.org> (raw)
In-Reply-To: <20230321181610.GE20234@pendragon.ideasonboard.com>
On Tue, Mar 21, 2023 at 08:16:10PM +0200, Laurent Pinchart wrote:
> Hi Benjamin,
>
> Thank you for the patch.
>
> On Tue, Mar 21, 2023 at 11:28:49AM +0100, Benjamin Gaignard wrote:
> > Instead of a static array change bufs to a dynamically allocated array.
> > This will allow to store more video buffer if needed.
> > Protect the array with a spinlock.
>
> I think I asked in the review of v1 why we couldn't use the kernel
> IDA/IDR APIs to allocate buffer IDs and register buffers, but I don't
> think I've received a reply. Wouldn't it be better than rolling out our
> own mechanism ?
>
+1, with a note that [1] suggests that IDR is deprecated and XArray[2]
should be used instead.
[1] https://docs.kernel.org/core-api/idr.html
[2] https://docs.kernel.org/core-api/xarray.html
Also from my quick look, XArray may be solving the locking concerns for us
automatically.
Best regards,
Tomasz
> > Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com>
> > ---
> > .../media/common/videobuf2/videobuf2-core.c | 8 +++
> > include/media/videobuf2-core.h | 49 ++++++++++++++++---
> > 2 files changed, 49 insertions(+), 8 deletions(-)
> >
> > diff --git a/drivers/media/common/videobuf2/videobuf2-core.c b/drivers/media/common/videobuf2/videobuf2-core.c
> > index 79e90e338846..ae9d72f4d181 100644
> > --- a/drivers/media/common/videobuf2/videobuf2-core.c
> > +++ b/drivers/media/common/videobuf2/videobuf2-core.c
> > @@ -2452,6 +2452,13 @@ int vb2_core_queue_init(struct vb2_queue *q)
> > mutex_init(&q->mmap_lock);
> > init_waitqueue_head(&q->done_wq);
> >
> > + q->max_num_bufs = 32;
> > + q->bufs = kmalloc_array(q->max_num_bufs, sizeof(*q->bufs), GFP_KERNEL | __GFP_ZERO);
> > + if (!q->bufs)
> > + return -ENOMEM;
> > +
> > + spin_lock_init(&q->bufs_lock);
> > +
> > q->memory = VB2_MEMORY_UNKNOWN;
> >
> > if (q->buf_struct_size == 0)
> > @@ -2479,6 +2486,7 @@ void vb2_core_queue_release(struct vb2_queue *q)
> > mutex_lock(&q->mmap_lock);
> > __vb2_queue_free(q, q->num_buffers);
> > mutex_unlock(&q->mmap_lock);
> > + kfree(q->bufs);
> > }
> > EXPORT_SYMBOL_GPL(vb2_core_queue_release);
> >
> > diff --git a/include/media/videobuf2-core.h b/include/media/videobuf2-core.h
> > index 5b1e3d801546..397dbf6e61e1 100644
> > --- a/include/media/videobuf2-core.h
> > +++ b/include/media/videobuf2-core.h
> > @@ -558,6 +558,8 @@ struct vb2_buf_ops {
> > * @dma_dir: DMA mapping direction.
> > * @bufs: videobuf2 buffer structures
> > * @num_buffers: number of allocated/used buffers
> > + * @bufs_lock: lock to protect bufs access
> > + * @max_num_bufs: max number of buffers storable in bufs
> > * @queued_list: list of buffers currently queued from userspace
> > * @queued_count: number of buffers queued and ready for streaming.
> > * @owned_by_drv_count: number of buffers owned by the driver
> > @@ -619,8 +621,10 @@ struct vb2_queue {
> > struct mutex mmap_lock;
> > unsigned int memory;
> > enum dma_data_direction dma_dir;
> > - struct vb2_buffer *bufs[VB2_MAX_FRAME];
> > + struct vb2_buffer **bufs;
> > unsigned int num_buffers;
> > + spinlock_t bufs_lock;
> > + size_t max_num_bufs;
> >
> > struct list_head queued_list;
> > unsigned int queued_count;
> > @@ -1239,9 +1243,16 @@ static inline void vb2_clear_last_buffer_dequeued(struct vb2_queue *q)
> > static inline struct vb2_buffer *vb2_get_buffer(struct vb2_queue *q,
> > unsigned int index)
> > {
> > - if (index < q->num_buffers)
> > - return q->bufs[index];
> > - return NULL;
> > + struct vb2_buffer *vb = NULL;
> > +
> > + spin_lock(&q->bufs_lock);
> > +
> > + if (index < q->max_num_bufs)
> > + vb = q->bufs[index];
> > +
> > + spin_unlock(&q->bufs_lock);
> > +
> > + return vb;
> > }
> >
> > /**
> > @@ -1251,12 +1262,30 @@ static inline struct vb2_buffer *vb2_get_buffer(struct vb2_queue *q,
> > */
> > static inline bool vb2_queue_add_buffer(struct vb2_queue *q, struct vb2_buffer *vb)
> > {
> > - if (vb->index < VB2_MAX_FRAME) {
> > + bool ret = false;
> > +
> > + spin_lock(&q->bufs_lock);
> > +
> > + if (vb->index >= q->max_num_bufs) {
> > + struct vb2_buffer **tmp;
> > +
> > + tmp = krealloc_array(q->bufs, q->max_num_bufs * 2, sizeof(*q->bufs), GFP_KERNEL);
> > + if (!tmp)
> > + goto realloc_failed;
> > +
> > + q->max_num_bufs *= 2;
> > + q->bufs = tmp;
> > + }
> > +
> > + if (vb->index < q->max_num_bufs) {
> > q->bufs[vb->index] = vb;
> > - return true;
> > + ret = true;
> > }
> >
> > - return false;
> > +realloc_failed:
> > + spin_unlock(&q->bufs_lock);
> > +
> > + return ret;
> > }
> >
> > /**
> > @@ -1266,8 +1295,12 @@ static inline bool vb2_queue_add_buffer(struct vb2_queue *q, struct vb2_buffer *
> > */
> > static inline void vb2_queue_remove_buffer(struct vb2_queue *q, struct vb2_buffer *vb)
> > {
> > - if (vb->index < VB2_MAX_FRAME)
> > + spin_lock(&q->bufs_lock);
> > +
> > + if (vb->index < q->max_num_bufs)
> > q->bufs[vb->index] = NULL;
> > +
> > + spin_unlock(&q->bufs_lock);
> > }
> >
> > /*
>
> --
> Regards,
>
> Laurent Pinchart
next prev parent reply other threads:[~2023-05-19 10:05 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-21 10:28 [PATCH v2 0/8] Add DELETE_BUF ioctl Benjamin Gaignard
2023-03-21 10:28 ` [PATCH v2 1/8] media: videobuf2: Access vb2_queue bufs array through helper functions Benjamin Gaignard
2023-03-21 17:42 ` kernel test robot
2023-05-18 10:37 ` Tomasz Figa
2023-03-21 10:28 ` [PATCH v2 2/8] media: videobuf2: Make bufs array dynamic allocated Benjamin Gaignard
2023-03-21 18:16 ` Laurent Pinchart
2023-05-19 10:04 ` Tomasz Figa [this message]
2023-03-24 5:01 ` Dan Carpenter
2023-03-24 8:11 ` Benjamin Gaignard
2023-03-24 8:31 ` Hans Verkuil
2023-03-24 8:48 ` Laurent Pinchart
2023-03-24 8:52 ` Hans Verkuil
2023-03-24 8:56 ` Benjamin Gaignard
2023-05-19 10:00 ` Tomasz Figa
2023-03-24 13:02 ` Dmitry Osipenko
2023-03-21 10:28 ` [PATCH v2 3/8] media: videobuf2: Add a module param to limit vb2 queue buffer storage Benjamin Gaignard
2023-03-21 17:01 ` kernel test robot
2023-03-21 19:15 ` kernel test robot
2023-03-22 6:22 ` [EXT] " Ming Qian
2023-05-19 10:19 ` Tomasz Figa
2023-05-30 17:38 ` Laurent Pinchart
2023-05-31 6:36 ` Hans Verkuil
2023-05-31 8:03 ` Laurent Pinchart
2023-05-31 8:30 ` Hans Verkuil
2023-05-31 12:39 ` Laurent Pinchart
2023-06-01 8:03 ` Benjamin Gaignard
2023-06-01 8:34 ` Laurent Pinchart
2023-06-08 10:24 ` Tomasz Figa
2023-06-08 10:42 ` Laurent Pinchart
2023-03-21 10:28 ` [PATCH v2 4/8] media: videobuf2: Stop define VB2_MAX_FRAME as global Benjamin Gaignard
2023-05-23 7:14 ` Tomasz Figa
2023-03-21 10:28 ` [PATCH v2 5/8] media: v4l2: Add DELETE_BUF ioctl Benjamin Gaignard
2023-03-21 20:06 ` kernel test robot
2023-05-23 8:23 ` Tomasz Figa
2023-03-21 10:28 ` [PATCH v2 6/8] media: v4l2: Add mem2mem helpers for " Benjamin Gaignard
2023-03-21 10:28 ` [PATCH v2 7/8] media: vim2m: Use v4l2-mem2mem helpers for VIDIOC_DELETE_BUF ioctl Benjamin Gaignard
2023-03-21 10:28 ` [PATCH v2 8/8] media: verisilicon: " Benjamin Gaignard
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=20230519100453.6o4eriisloiji32g@chromium.org \
--to=tfiga@chromium.org \
--cc=agross@kernel.org \
--cc=andersson@kernel.org \
--cc=andrew-ct.chen@mediatek.com \
--cc=angelogioacchino.delregno@collabora.com \
--cc=benjamin.gaignard@collabora.com \
--cc=bin.liu@mediatek.com \
--cc=daniel.almeida@collabora.com \
--cc=eagle.zhou@nxp.com \
--cc=ezequiel@vanguardiasur.com.ar \
--cc=hverkuil-cisco@xs4all.nl \
--cc=jernel@kernel.org \
--cc=kernel@collabora.com \
--cc=konrad.dybcio@linaro.org \
--cc=laurent.pinchart@ideasonboard.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-media@vger.kernel.org \
--cc=linux-mediatek@lists.infradead.org \
--cc=linux-rockchip@lists.infradead.org \
--cc=m.szyprowski@samsung.com \
--cc=matthias.bgg@gmail.com \
--cc=mchehab@kernel.org \
--cc=ming.qian@nxp.com \
--cc=p.zabel@pengutronix.de \
--cc=quic_vgarodia@quicinc.com \
--cc=shijie.qin@nxp.com \
--cc=stanimir.k.varbanov@gmail.com \
--cc=tiffany.lin@mediatek.com \
--cc=yunfei.dong@mediatek.com \
/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
all inboxes | Powered by JetHome®