Message-ID: <8bf06ce5c79271a326ee375f104e475c6e482f58.camel@mediatek.com>
Date: Mon, 25 Nov 2024 05:54:01 +0000
From: CK Hu (胡俊光) <ck.hu@...iatek.com>
To: Julien Stephan <jstephan@...libre.com>
CC: "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"linux-mediatek@...ts.infradead.org" <linux-mediatek@...ts.infradead.org>,
	"linux-media@...r.kernel.org" <linux-media@...r.kernel.org>,
	"devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
	"paul.elder@...asonboard.com" <paul.elder@...asonboard.com>,
	"mchehab@...nel.org" <mchehab@...nel.org>, "conor+dt@...nel.org"
	<conor+dt@...nel.org>, "robh@...nel.org" <robh@...nel.org>,
	Andy Hsieh (謝智皓) <Andy.Hsieh@...iatek.com>,
	"linux-arm-kernel@...ts.infradead.org"
	<linux-arm-kernel@...ts.infradead.org>, "matthias.bgg@...il.com"
	<matthias.bgg@...il.com>, "laurent.pinchart@...asonboard.com"
	<laurent.pinchart@...asonboard.com>, "krzk+dt@...nel.org"
	<krzk+dt@...nel.org>, "fsylvestre@...libre.com" <fsylvestre@...libre.com>,
	AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>,
	"pnguyen@...libre.com" <pnguyen@...libre.com>
Subject: Re: [PATCH v7 4/5] media: platform: mediatek: isp: add mediatek
 ISP3.0 camsv

On Fri, 2024-11-22 at 10:44 +0100, Julien Stephan wrote:
> External email : Please do not click links or open attachments until you have verified the sender or the content.
> 
> 
> Le ven. 22 nov. 2024 à 10:19, CK Hu (胡俊光) <ck.hu@...iatek.com> a écrit :
> > 
> > Hi, Julien:
> > 
> > On Thu, 2024-11-21 at 09:53 +0100, Julien Stephan wrote:
> > > External email : Please do not click links or open attachments until you have verified the sender or the content.
> > > 
> > > 
> > > From: Phi-bang Nguyen <pnguyen@...libre.com>
> > > 
> > > This driver provides a path to bypass the SoC ISP so that image data
> > > coming from the SENINF can go directly into memory without any image
> > > processing. This allows the use of an external ISP.
> > > 
> > > Signed-off-by: Phi-bang Nguyen <pnguyen@...libre.com>
> > > Signed-off-by: Florian Sylvestre <fsylvestre@...libre.com>
> > > [Paul Elder fix irq locking]
> > > Signed-off-by: Paul Elder <paul.elder@...asonboard.com>
> > > Co-developed-by: Laurent Pinchart <laurent.pinchart@...asonboard.com>
> > > Signed-off-by: Laurent Pinchart <laurent.pinchart@...asonboard.com>
> > > Co-developed-by: Julien Stephan <jstephan@...libre.com>
> > > Signed-off-by: Julien Stephan <jstephan@...libre.com>
> > > ---
> > 
> > [snip]
> > 
> > > +static void mtk_cam_vb2_buf_queue(struct vb2_buffer *vb)
> > > +{
> > > +       struct mtk_cam_dev *cam = vb2_get_drv_priv(vb->vb2_queue);
> > > +       struct mtk_cam_dev_buffer *buf = to_mtk_cam_dev_buffer(vb);
> > > +       unsigned long flags;
> > > +
> > > +       /* Add the buffer into the tracking list */
> > > +       spin_lock_irqsave(&cam->buf_list_lock, flags);
> > > +       if (list_empty(&cam->buf_list))
> > > +               (*cam->hw_functions->mtk_cam_update_buffers_add)(cam, buf);
> > > +
> > > +       list_add_tail(&buf->list, &cam->buf_list);
> > > +       (*cam->hw_functions->mtk_cam_fbc_inc)(cam);
> > 
> > I think fbc_inc should together with update_buffers_add.
> > After update_buffers_add then fbc_inc.
> > So squash fbc_inc into update_buffers_add and drop fbc_inc function.
> > 
> 
> No, this is not true.
> mtk_cam_update_buffers_add is used to indicate which buffer should be
> used for dma write. This is the first entry in the buf list.
> mtk_cam_fbc_inc is used to increase the number of available user space buffers.
> 
> If the buffer list is not empty and user space calls buf_queue again,
> we need to call mtk_cam_fbc_inc to increase the number of available
> user buffers, but we don't want to change the buffer for DMA write.
> 
> mtk_camsv30_update_buffers_add is called on irq to update the address
> to the next buffer (if available).

I think current design is OK.
I just think it could have more flexible design.
For example, reduce frame rate from 30 fps to 15 fps.
And this is a time sequence to reduce frame rate:

t = -5 ms, set buf1 address and fbc increase. (hardware index = 0, software index = 1)
t = 0 ms,
t = 33 ms, buf1 is done, do not set next buffer. (hardware index = 1, software index = 1)
t = 67 ms, set buf2 addres and fbc increase. (hardware index = 1, software index = 2)
t = 100 ms, buf2 is done, (hardware index = 2, software idex = 2)

If want to keep the 30 fps, when t = 33ms, set buf2 address and fbc increase.

If you want to keep current design, that's OK for me because now has no requirement for such advanced control.

But I have a new question, in the time sequence I list, does it has interrupt when t = 0 (the first vblank) ?
If t = 0 has no interrupt, I think t = 67 also has no interrupt and my design would not work.
It t = 0 has interrupt, I think software should think this buffer is not done.

> 
> Maybe the name mtk_camsv30_update_buffers_add is confusing then?
> What do you think about:
> - mtk_camsv30_update_buffers_add -> mtk_camsv30_update_buffers_address

mtk_camsv30_update_buffers_address is better.

> - mtk_cam_fbc_inc -> mtk_camsv30_buffer_add

mtk_cam_fbc_inc is better, it show counter increase.

Regards,
CK

> 
> Cheers
> Julien
> 
>  > Regards,
> > CK
> > 
> > > +       spin_unlock_irqrestore(&cam->buf_list_lock, flags);
> > > +}
> > > +
> > > 
> > 
> > ************* MEDIATEK Confidentiality Notice ********************
> > The information contained in this e-mail message (including any
> > attachments) may be confidential, proprietary, privileged, or otherwise
> > exempt from disclosure under applicable laws. It is intended to be
> > conveyed only to the designated recipient(s). Any use, dissemination,
> > distribution, printing, retaining or copying of this e-mail (including its
> > attachments) by unintended recipient(s) is strictly prohibited and may
> > be unlawful. If you are not an intended recipient of this e-mail, or believe
> > that you have received this e-mail in error, please notify the sender
> > immediately (by replying to this e-mail), delete any and all copies of
> > this e-mail (including any attachments) from your system, and do not
> > disclose the content of this e-mail to any other person. Thank you!

Powered by blists - more mailing lists