mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Benjamin Gaignard <benjamin.gaignard@collabora.com>
To: Hans Verkuil <hverkuil@xs4all.nl>, mchehab@kernel.org
Cc: linux-kernel@vger.kernel.org, linux-media@vger.kernel.org,
	kernel@collabora.com
Subject: Re: [PATCH v16 7/8] media: v4l2: Add mem2mem helpers for DELETE_BUFS ioctl
Date: Tue, 16 Jan 2024 11:42:56 +0100	[thread overview]
Message-ID: <60c942d8-e0bd-4609-8fc4-1e80102ac051@collabora.com> (raw)
In-Reply-To: <6b0e4c6b-493c-4916-ab3c-deeeb725fdec@xs4all.nl>


Le 15/01/2024 à 17:50, Hans Verkuil a écrit :
> On 15/12/2023 10:08, Benjamin Gaignard wrote:
>> Create v4l2-mem2mem helpers for VIDIOC_DELETE_BUFS ioctl.
>>
>> Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com>
>> ---
>>   .../media/platform/verisilicon/hantro_drv.c   |  1 +
>>   .../media/platform/verisilicon/hantro_v4l2.c  |  1 +
>>   drivers/media/test-drivers/vim2m.c            |  2 ++
> The driver changes should be done in a separate patch.
>
>>   drivers/media/v4l2-core/v4l2-mem2mem.c        | 20 +++++++++++++++++++
>>   include/media/v4l2-mem2mem.h                  | 12 +++++++++++
>>   5 files changed, 36 insertions(+)
>>
>> diff --git a/drivers/media/platform/verisilicon/hantro_drv.c b/drivers/media/platform/verisilicon/hantro_drv.c
>> index db3df6cc4513..f6b0a676a740 100644
>> --- a/drivers/media/platform/verisilicon/hantro_drv.c
>> +++ b/drivers/media/platform/verisilicon/hantro_drv.c
>> @@ -248,6 +248,7 @@ queue_init(void *priv, struct vb2_queue *src_vq, struct vb2_queue *dst_vq)
>>   	dst_vq->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_COPY;
>>   	dst_vq->lock = &ctx->dev->vpu_mutex;
>>   	dst_vq->dev = ctx->dev->v4l2_dev.dev;
>> +	src_vq->supports_delete_bufs = true;
> Isn't this something that can be supported for both queues?

For me it isn't useful to support it on the both queues because
only capture queue will store unused buffers after a dynamic
resolution change. Output queue buffers are smaller and always
recycled even after a dynamic resolution change.

>
>>   
>>   	return vb2_queue_init(dst_vq);
>>   }
>> diff --git a/drivers/media/platform/verisilicon/hantro_v4l2.c b/drivers/media/platform/verisilicon/hantro_v4l2.c
>> index 941fa23c211a..34eab90e8a42 100644
>> --- a/drivers/media/platform/verisilicon/hantro_v4l2.c
>> +++ b/drivers/media/platform/verisilicon/hantro_v4l2.c
>> @@ -756,6 +756,7 @@ const struct v4l2_ioctl_ops hantro_ioctl_ops = {
>>   	.vidioc_dqbuf = v4l2_m2m_ioctl_dqbuf,
>>   	.vidioc_prepare_buf = v4l2_m2m_ioctl_prepare_buf,
>>   	.vidioc_create_bufs = v4l2_m2m_ioctl_create_bufs,
>> +	.vidioc_delete_bufs = v4l2_m2m_ioctl_delete_bufs,
>>   	.vidioc_expbuf = v4l2_m2m_ioctl_expbuf,
>>   
>>   	.vidioc_subscribe_event = v4l2_ctrl_subscribe_event,
>> diff --git a/drivers/media/test-drivers/vim2m.c b/drivers/media/test-drivers/vim2m.c
>> index 3e3b424b4860..17213ce42059 100644
>> --- a/drivers/media/test-drivers/vim2m.c
>> +++ b/drivers/media/test-drivers/vim2m.c
>> @@ -960,6 +960,7 @@ static const struct v4l2_ioctl_ops vim2m_ioctl_ops = {
>>   	.vidioc_dqbuf		= v4l2_m2m_ioctl_dqbuf,
>>   	.vidioc_prepare_buf	= v4l2_m2m_ioctl_prepare_buf,
>>   	.vidioc_create_bufs	= v4l2_m2m_ioctl_create_bufs,
>> +	.vidioc_delete_bufs	= v4l2_m2m_ioctl_delete_bufs,
>>   	.vidioc_expbuf		= v4l2_m2m_ioctl_expbuf,
>>   
>>   	.vidioc_streamon	= v4l2_m2m_ioctl_streamon,
>> @@ -1133,6 +1134,7 @@ static int queue_init(void *priv, struct vb2_queue *src_vq,
>>   	dst_vq->mem_ops = &vb2_vmalloc_memops;
>>   	dst_vq->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_COPY;
>>   	dst_vq->lock = &ctx->vb_mutex;
>> +	dst_vq->supports_delete_bufs = true;
> Same question.

I want to test something similar to what the real use case does.

>
>>   
>>   	return vb2_queue_init(dst_vq);
>>   }
>> diff --git a/drivers/media/v4l2-core/v4l2-mem2mem.c b/drivers/media/v4l2-core/v4l2-mem2mem.c
>> index 9e983176542b..dbc4711fc556 100644
>> --- a/drivers/media/v4l2-core/v4l2-mem2mem.c
>> +++ b/drivers/media/v4l2-core/v4l2-mem2mem.c
>> @@ -834,6 +834,17 @@ int v4l2_m2m_prepare_buf(struct file *file, struct v4l2_m2m_ctx *m2m_ctx,
>>   }
>>   EXPORT_SYMBOL_GPL(v4l2_m2m_prepare_buf);
>>   
>> +int v4l2_m2m_delete_bufs(struct file *file, struct v4l2_m2m_ctx *m2m_ctx,
>> +			 struct v4l2_delete_buffers *d)
>> +{
>> +	struct vb2_queue *vq;
>> +
>> +	vq = v4l2_m2m_get_vq(m2m_ctx, d->type);
> These 3 lines can be combined into one.
>
>> +
>> +	return vb2_delete_bufs(vq, d);
>> +}
>> +EXPORT_SYMBOL_GPL(v4l2_m2m_delete_bufs);
> I'm not sure we need to export this. Drivers should really just use the
> v4l2_m2m_ioctl_ variant below.

ok

regards,
Benjamin

>
>> +
>>   int v4l2_m2m_create_bufs(struct file *file, struct v4l2_m2m_ctx *m2m_ctx,
>>   			 struct v4l2_create_buffers *create)
>>   {
>> @@ -1380,6 +1391,15 @@ int v4l2_m2m_ioctl_create_bufs(struct file *file, void *priv,
>>   }
>>   EXPORT_SYMBOL_GPL(v4l2_m2m_ioctl_create_bufs);
>>   
>> +int v4l2_m2m_ioctl_delete_bufs(struct file *file, void *priv,
>> +			       struct v4l2_delete_buffers *d)
>> +{
>> +	struct v4l2_fh *fh = file->private_data;
>> +
>> +	return v4l2_m2m_delete_bufs(file, fh->m2m_ctx, d);
>> +}
>> +EXPORT_SYMBOL_GPL(v4l2_m2m_ioctl_delete_bufs);
>> +
>>   int v4l2_m2m_ioctl_querybuf(struct file *file, void *priv,
>>   				struct v4l2_buffer *buf)
>>   {
>> diff --git a/include/media/v4l2-mem2mem.h b/include/media/v4l2-mem2mem.h
>> index 7f1af1f7f912..5314952ad3d5 100644
>> --- a/include/media/v4l2-mem2mem.h
>> +++ b/include/media/v4l2-mem2mem.h
>> @@ -388,6 +388,16 @@ int v4l2_m2m_dqbuf(struct file *file, struct v4l2_m2m_ctx *m2m_ctx,
>>   int v4l2_m2m_prepare_buf(struct file *file, struct v4l2_m2m_ctx *m2m_ctx,
>>   			 struct v4l2_buffer *buf);
>>   
>> +/**
>> + * v4l2_m2m_delete_bufs() - delete buffers from the queue
>> + *
>> + * @file: pointer to struct &file
>> + * @m2m_ctx: m2m context assigned to the instance given by struct &v4l2_m2m_ctx
>> + * @d: pointer to struct &v4l2_delete_buffers
>> + */
>> +int v4l2_m2m_delete_bufs(struct file *file, struct v4l2_m2m_ctx *m2m_ctx,
>> +			 struct v4l2_delete_buffers *d);
>> +
>>   /**
>>    * v4l2_m2m_create_bufs() - create a source or destination buffer, depending
>>    * on the type
>> @@ -867,6 +877,8 @@ int v4l2_m2m_ioctl_reqbufs(struct file *file, void *priv,
>>   				struct v4l2_requestbuffers *rb);
>>   int v4l2_m2m_ioctl_create_bufs(struct file *file, void *fh,
>>   				struct v4l2_create_buffers *create);
>> +int v4l2_m2m_ioctl_delete_bufs(struct file *file, void *priv,
>> +			       struct v4l2_delete_buffers *d);
>>   int v4l2_m2m_ioctl_querybuf(struct file *file, void *fh,
>>   				struct v4l2_buffer *buf);
>>   int v4l2_m2m_ioctl_expbuf(struct file *file, void *fh,
> Regards,
>
> 	Hans
>

  reply	other threads:[~2024-01-16 10:42 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-15  9:08 [PATCH v16 0/8] Add DELETE_BUF ioctl Benjamin Gaignard
2023-12-15  9:08 ` [PATCH v16 1/8] videobuf2: Add min_reqbufs_allocation field to vb2_queue structure Benjamin Gaignard
2024-01-15  9:16   ` Hans Verkuil
2024-01-15 10:27   ` Hans Verkuil
2023-12-15  9:08 ` [PATCH v16 2/8] media: test-drivers: Set REQBUFS minimum number of buffers Benjamin Gaignard
2024-01-15 10:35   ` Hans Verkuil
2023-12-15  9:08 ` [PATCH v16 3/8] media: core: Rework how create_buf index returned value is computed Benjamin Gaignard
2024-01-15 11:00   ` Hans Verkuil
2024-01-15 11:40     ` Hans Verkuil
2024-01-15 12:11   ` Hans Verkuil
2024-01-15 14:52     ` Benjamin Gaignard
2024-01-15 15:17       ` Hans Verkuil
2024-01-15 16:04   ` Hans Verkuil
2023-12-15  9:08 ` [PATCH v16 4/8] media: core: Add bitmap manage bufs array entries Benjamin Gaignard
2024-01-15 12:21   ` Hans Verkuil
2024-01-15 14:51     ` Benjamin Gaignard
2024-01-15 15:14       ` Hans Verkuil
2023-12-15  9:08 ` [PATCH v16 5/8] media: core: Free range of buffers Benjamin Gaignard
2024-01-15 16:08   ` Hans Verkuil
2023-12-15  9:08 ` [PATCH v16 6/8] media: v4l2: Add DELETE_BUFS ioctl Benjamin Gaignard
2024-01-15 16:43   ` Hans Verkuil
2024-01-16  9:37     ` Benjamin Gaignard
2023-12-15  9:08 ` [PATCH v16 7/8] media: v4l2: Add mem2mem helpers for " Benjamin Gaignard
2024-01-15 16:50   ` Hans Verkuil
2024-01-16 10:42     ` Benjamin Gaignard [this message]
2023-12-15  9:08 ` [PATCH v16 8/8] media: test-drivers: Use helper " Benjamin Gaignard
2024-01-15 16:53   ` Hans Verkuil
2024-01-09  9:37 ` [PATCH v16 0/8] Add DELETE_BUF ioctl 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=60c942d8-e0bd-4609-8fc4-1e80102ac051@collabora.com \
    --to=benjamin.gaignard@collabora.com \
    --cc=hverkuil@xs4all.nl \
    --cc=kernel@collabora.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab@kernel.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

all inboxes | Powered by JetHome®