Re: [PATCH RESEND v9 18/18] media: platform: Add jpeg enc feature

From: Tomasz Figa
Date: Tue Jul 07 2020 - 09:35:36 EST


On Tue, Jul 7, 2020 at 8:47 AM Xia Jiang <xia.jiang@xxxxxxxxxxxx> wrote:
>
> On Tue, 2020-06-30 at 16:53 +0000, Tomasz Figa wrote:
> > Hi Xia,
> >
> > On Tue, Jun 30, 2020 at 10:56:21AM +0800, Xia Jiang wrote:
> > > On Thu, 2020-06-11 at 18:46 +0000, Tomasz Figa wrote:
> > > > Hi Xia,
> > > >
> > > > On Thu, Jun 04, 2020 at 05:05:53PM +0800, Xia Jiang wrote:
> > [snip]
> > > > > +static void mtk_jpeg_enc_device_run(void *priv)
> > > > > +{
> > > > > + struct mtk_jpeg_ctx *ctx = priv;
> > > > > + struct mtk_jpeg_dev *jpeg = ctx->jpeg;
> > > > > + struct vb2_v4l2_buffer *src_buf, *dst_buf;
> > > > > + enum vb2_buffer_state buf_state = VB2_BUF_STATE_ERROR;
> > > > > + unsigned long flags;
> > > > > + struct mtk_jpeg_src_buf *jpeg_src_buf;
> > > > > + struct mtk_jpeg_enc_bs enc_bs;
> > > > > + int ret;
> > > > > +
> > > > > + src_buf = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
> > > > > + dst_buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
> > > > > + jpeg_src_buf = mtk_jpeg_vb2_to_srcbuf(&src_buf->vb2_buf);
> > > > > +
> > > > > + ret = pm_runtime_get_sync(jpeg->dev);
> > > > > + if (ret < 0)
> > > > > + goto enc_end;
> > > > > +
> > > > > + spin_lock_irqsave(&jpeg->hw_lock, flags);
> > > > > +
> > > > > + /*
> > > > > + * Resetting the hardware every frame is to ensure that all the
> > > > > + * registers are cleared. This is a hardware requirement.
> > > > > + */
> > > > > + mtk_jpeg_enc_reset(jpeg->reg_base);
> > > > > +
> > > > > + mtk_jpeg_set_enc_dst(ctx, jpeg->reg_base, &dst_buf->vb2_buf, &enc_bs);
> > > > > + mtk_jpeg_set_enc_src(ctx, jpeg->reg_base, &src_buf->vb2_buf);
> > > > > + mtk_jpeg_enc_set_config(jpeg->reg_base, ctx->out_q.fmt->hw_format,
> > > > > + ctx->enable_exif, ctx->enc_quality,
> > > > > + ctx->restart_interval);
> > > > > + mtk_jpeg_enc_start(jpeg->reg_base);
> > > >
> > > > Could we just move the above 5 functions into one function inside
> > > > mtk_jpeg_enc_hw.c that takes mtk_jpeg_dev pointer as its argument, let's
> > > > say mtk_jpeg_enc_hw_run() and simply program all the data to the registers
> > > > directly, without the extra level of abstractions?
> > > I can move the 5 functions into one function(mtk_jpeg_enc_hw_run()), but
> > > this function will be very long, because it contains computation code
> > > such as setting dst addr, blk_num, quality.
> > > In v4, you have adviced the following architecture:
> > > How about the following model, as used by many other drivers:
> > >
> > > mtk_jpeg_enc_set_src()
> > > {
> > > // Set any registers related to source format and buffer
> > > }
> > >
> > > mtk_jpeg_enc_set_dst()
> > > {
> > > // Set any registers related to destination format and buffer
> > > }
> > >
> > > mtk_jpeg_enc_set_params()
> > > {
> > > // Set any registers related to additional encoding parameters
> > > }
> > >
> > > mtk_jpeg_enc_device_run(enc, ctx)
> > > {
> > > mtk_jpeg_enc_set_src(enc, src_buf, src_fmt);
> > > mtk_jpeg_enc_set_dst(enc, dst_buf, dst_fmt);
> > > mtk_jpeg_enc_set_params(enc, ctx);
> > > // Trigger the hardware run
> > > }
> > > I think that this architecture is more clear(mtk_jpeg_enc_set_config is
> > > equivalent to mtk_jpeg_enc_set_params).
> > > Should I keep the original architecture or move 5 functions into
> > > mtk_jpeg_enc_hw_run?
> >
> > Sounds good to me.
> >
> > My biggest issue with the code that it ends up introducing one more
> > level of abstraction, but with the approach you suggested, the arguments
> > just accept standard structs, which avoids that problem.
> >
> > [snip]
> > > > > +
> > > > > + ctx->fh.ctrl_handler = &ctx->ctrl_hdl;
> > > > > + ctx->colorspace = V4L2_COLORSPACE_JPEG,
> > > > > + ctx->ycbcr_enc = V4L2_YCBCR_ENC_DEFAULT;
> > > > > + ctx->quantization = V4L2_QUANTIZATION_DEFAULT;
> > > > > + ctx->xfer_func = V4L2_XFER_FUNC_DEFAULT;
> > > >
> > > > Since we already have a v4l2_pix_format_mplane struct which has fields for
> > > > the above 4 values, could we just store them there?
> > > >
> > > > Also, I don't see this driver handling the colorspaces in any way, but it
> > > > seems to allow changing them from the userspace. This is incorrect, because
> > > > the userspace has no way to know that the colorspace is not handled.
> > > > Instead, the try_fmt implementation should always override the
> > > > userspace-provided colorspace configuration with the ones that the driver
> > > > assumes.
> > > >
> > > > > + pix_mp->width = MTK_JPEG_MIN_WIDTH;
> > > > > + pix_mp->height = MTK_JPEG_MIN_HEIGHT;
> > > > > +
> > > > > + q->fmt = mtk_jpeg_find_format(V4L2_PIX_FMT_YUYV,
> > > > > + MTK_JPEG_FMT_FLAG_ENC_OUTPUT);
> > > > > + vidioc_try_fmt(container_of(pix_mp, struct v4l2_format,
> > > > > + fmt.pix_mp), q->fmt);
> > > > > + q->w = pix_mp->width;
> > > > > + q->h = pix_mp->height;
> > > > > + q->crop_rect.width = pix_mp->width;
> > > > > + q->crop_rect.height = pix_mp->height;
> > > > > + q->sizeimage[0] = pix_mp->plane_fmt[0].sizeimage;
> > > > > + q->bytesperline[0] = pix_mp->plane_fmt[0].bytesperline;
> > > >
> > > > Actually, do we need this custom mtk_jpeg_q_data struct? Why couldn't we
> > > > just keep the same values inside the standard v4l2_pix_format_mplane
> > > > struct?
> > > I think that we need mtk_jpeg_q_data struct.If we delete it, how can we
> > > know these values(w, h, sizeimage, bytesperline, mtk_jpeg_fmt) belong to
> > > output or capture(output and capture's sizeimages are different, width
> > > and height are differnt too for jpeg dec )?We have
> > > s_fmt_vid_out_mplane/cap_mplane function to set these values.
> > >
> > > But we can use standard v4l2_pix_format_mplane struct replacing the w, h
> > > bytesperline, sizeimage in mtk_jpeg_q_data struct like this:
> > > struct mtk_jpeg_q_data{
> > > struct mtk_jpeg_fmt *fmt;
> > > struct v4l2_pix_format_mplane pix_mp;
> > > struct v4l2_rect enc_crop_rect
> > > }
> > > Then delete ctx->colorspace ctx->ycbcr_enc ctx->quantization
> > > ctx->xfer_func, becuase v4l2_pix_format_mplane in q_data has contained
> > > them and assign them for out_q and cap_q separately.
> > >
> > > WDYT?
> >
> > Sounds good to me. I was considering just making it like
> >
> > struct mtk_jpeg_ctx {
> > struct mtk_jpeg_fmt *src_fmt;
> > struct v4l2_pix_format_mplane src_pix_mp;
> > struct v4l2_rect src_crop;
> >
> > struct mtk_jpeg_fmt *dst_fmt;
> > struct v4l2_pix_format_mplane dst_pix_mp;
> > struct v4l2_rect dst_crop;
> > };
> >
> > but I like your suggestion as well, as long as custom data structures
> > are not used to store standard information.
>
> Dear Tomasz,
>
> I used the structure like below:
> struct mtk_jpeg_q_data{
> struct mtk_jpeg_fmt *fmt;
> struct v4l2_pix_format_mplane pix_mp;
> struct v4l2_rect enc_crop_rect
> }
> Then delete ctx->colorspace ctx->ycbcr_enc ctx->quantization
> ctx->xfer_func, becuase v4l2_pix_format_mplane in q_data has contained
> them and assign them for out_q and cap_q separately.
>
> Then the v4l2_compliance test will fail, the fail log as below:
> Format ioctls:
> test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK
> test VIDIOC_G/S_PARM: OK (Not Supported)
> test VIDIOC_G_FBUF: OK (Not Supported)
> test VIDIOC_G_FMT: OK
> test VIDIOC_TRY_FMT: OK
> fail: v4l2-test-formats.cpp(836):
> fmt_cap.g_colorspace() != col
> test VIDIOC_S_FMT: FAIL
> test VIDIOC_G_SLICED_VBI_CAP: OK (Not Supported)
> test Cropping: OK
> test Composing: OK (Not Supported)
> test Scaling: OK (Not Supported)
>
> The source code of v4l2-test-formats.cpp as below:
>
> static int testM2MFormats(struct node *node)
> {
> cv4l_fmt fmt_out;
> cv4l_fmt fmt;
> cv4l_fmt fmt_cap;
> __u32 cap_type = node->g_type();
> __u32 out_type = v4l_type_invert(cap_type);
> __u32 col, ycbcr_enc, quant, xfer_func;
>
> fail_on_test(node->g_fmt(fmt_out, out_type));
> node->g_fmt(fmt_cap, cap_type);
> fail_on_test(fmt_cap.g_colorspace() != fmt_out.g_colorspace());
> fail_on_test(fmt_cap.g_ycbcr_enc() != fmt_out.g_ycbcr_enc());
> fail_on_test(fmt_cap.g_quantization() != fmt_out.g_quantization());
> fail_on_test(fmt_cap.g_xfer_func() != fmt_out.g_xfer_func());
> col = fmt_out.g_colorspace() == V4L2_COLORSPACE_SMPTE170M ?
> V4L2_COLORSPACE_REC709 : V4L2_COLORSPACE_SMPTE170M;
> ycbcr_enc = fmt_out.g_ycbcr_enc() == V4L2_YCBCR_ENC_601 ?
> V4L2_YCBCR_ENC_709 : V4L2_YCBCR_ENC_601;
> quant = fmt_out.g_quantization() == V4L2_QUANTIZATION_LIM_RANGE ?
> V4L2_QUANTIZATION_FULL_RANGE : V4L2_QUANTIZATION_LIM_RANGE;
> xfer_func = fmt_out.g_xfer_func() == V4L2_XFER_FUNC_SRGB ?
> V4L2_XFER_FUNC_709 : V4L2_XFER_FUNC_SRGB;
> fmt_out.s_colorspace(col);
> fmt_out.s_xfer_func(xfer_func);
> fmt_out.s_ycbcr_enc(ycbcr_enc);
> fmt_out.s_quantization(quant);
> node->s_fmt(fmt_out);
> fail_on_test(fmt_out.g_colorspace() != col);
> fail_on_test(fmt_out.g_xfer_func() != xfer_func);
> fail_on_test(fmt_out.g_ycbcr_enc() != ycbcr_enc);
> fail_on_test(fmt_out.g_quantization() != quant);
> node->g_fmt(fmt_cap);
> fail_on_test(fmt_cap.g_colorspace() != col); // line 836
> fail_on_test(fmt_cap.g_xfer_func() != xfer_func);
> fail_on_test(fmt_cap.g_ycbcr_enc() != ycbcr_enc);
> fail_on_test(fmt_cap.g_quantization() != quant);
> }
>
> It needs that cap's colorspace equals out's colorspace when userspace
> just set out's colorspace and then get cap's colorspace. However, cap's
> colorspace which getted from driver equals V4L2_COLORSPACE_JPEG, because
> the code in g_fmt() like this:
> pix_mp->colorspace = q_data->pix_mp.colorspace;
> pix_mp->ycbcr_enc = q_data->pix_mp.ycbcr_enc;
> pix_mp->xfer_func = q_data->pix_mp.xfer_func;
> pix_mp->quantization = q_data->pix_mp.quantization;
>
> How should I handle this case? Should I store them(colorspace,
> sfer_func, ycbcr_enc, quatization) in ctx as the orinal desin? Then I
> can get them from g_fmt() like this:
> pix_mp->colorspace = ctx->colorspace;
> pix_mp->ycbcr_enc = ctx->ycbcr_enc;
> pix_mp->xfer_func = ctx->xfer_func;
> pix_mp->quantization = ctx->quantization;

Why would there be any other colorspace accepted? I suppose that the
hardware only supports the JPEG color space, so it shouldn't accept
any other colorspace in TRY_FMT (and thus S_FMT) anyway.

Still, for correctness, I would suggest propagating the colorspace
(and related) information from OUTPUT format to CAPTURE format in
S_FMT(OUTPUT).

Best regards,
Tomasz

>
> Best Regards,
> Xia Jiang
> > [snip]
> > > > > @@ -1042,8 +1619,12 @@ static int mtk_jpeg_probe(struct platform_device *pdev)
> > > > > return jpeg_irq;
> > > > > }
> > > > >
> > > > > - ret = devm_request_irq(&pdev->dev, jpeg_irq, mtk_jpeg_dec_irq, 0,
> > > > > - pdev->name, jpeg);
> > > > > + if (jpeg->variant->is_encoder)
> > > > > + ret = devm_request_irq(&pdev->dev, jpeg_irq, mtk_jpeg_enc_irq,
> > > > > + 0, pdev->name, jpeg);
> > > > > + else
> > > > > + ret = devm_request_irq(&pdev->dev, jpeg_irq, mtk_jpeg_dec_irq,
> > > > > + 0, pdev->name, jpeg);
> > > >
> > > > Rather than having "is_encoder" in the variant struct, would it make more
> > > > sense to have "irq_handler" instead? That would avoid the explicit if.
> > > Do you mean to delete "is_encoder"? It is used 8 times in the
> > > driver.Should I move them all to the match data?
> >
> > Yes. It would make the code linear and the varability between the
> > decoder and encoder would be self-contained in the variant struct.
> >
> > Best regards,
> > Tomasz
>