Re: [PATCH 2/9] [media] s5p-jpeg: Call jpeg_bound_align_image after qbuf

[Jacek Anaszewski]

Hi Thierry,

Thanks for the patch.

On 06/02/2017 06:02 PM, Thierry Escande wrote:
> From: Tony K Nadackal 
> 
> When queuing an OUTPUT buffer for decoder, s5p_jpeg_parse_hdr()
> function parses the input jpeg file and takes the width and height
> parameters from its header. These new width/height values will be used
> for the calculation of stride. HX_JPEG Hardware needs the width and
> height values aligned on a 16 bits boundary. This width/height alignment
> is handled in the s5p_jpeg_s_fmt_vid_cap() function during the S_FMT
> ioctl call.
> 
> But if user space calls the QBUF of OUTPUT buffer after the S_FMT of
> CAPTURE buffer, these aligned values will be replaced by the values in
> jpeg header.

I assume that you may want to avoid re-setting the capture buf format
when decoding a stream of JPEGs and you are certain that all of them
have the same subsampling. Nonetheless, please keep in mind that in case
of Exynos4x12 SoCs there is a risk of permanent decoder hangup if you'd
try to decode to a YUV with lower subsampling than the one of input
JPEG. s5p_jpeg_try_fmt_vid_cap() does a suitable adjustment to avoid the
problem.

I'd add a comment over this call to jpeg_bound_align_image() that
resigning from executing S_FMT on capture buf for each JPEG image
can result in a hardware hangup if forbidden decoding will be enforced.

> If the width/height values of jpeg are not aligned, the
> decoder output will be corrupted. So in this patch we call
> jpeg_bound_align_image() to align the width/height values of Capture
> buffer in s5p_jpeg_buf_queue().
> 
> Signed-off-by: Tony K Nadackal 
> Signed-off-by: Thierry Escande 
> ---
>  drivers/media/platform/s5p-jpeg/jpeg-core.c | 7 +++
>  1 file changed, 7 insertions(+)
> 
> diff --git a/drivers/media/platform/s5p-jpeg/jpeg-core.c 
> b/drivers/media/platform/s5p-jpeg/jpeg-core.c
> index 52dc794..6fb1ab4 100644
> --- a/drivers/media/platform/s5p-jpeg/jpeg-core.c
> +++ b/drivers/media/platform/s5p-jpeg/jpeg-core.c
> @@ -2523,6 +2523,13 @@ static void s5p_jpeg_buf_queue(struct vb2_buffer *vb)
>   q_data = >cap_q;
>   q_data->w = tmp.w;
>   q_data->h = tmp.h;
> +
> + jpeg_bound_align_image(ctx, _data->w, S5P_JPEG_MIN_WIDTH,
> +S5P_JPEG_MAX_WIDTH, q_data->fmt->h_align,
> +_data->h, S5P_JPEG_MIN_HEIGHT,
> +S5P_JPEG_MAX_HEIGHT, q_data->fmt->v_align
> +   );
> + q_data->size = q_data->w * q_data->h * q_data->fmt->depth >> 3;
>   }
>  
>   v4l2_m2m_buf_queue(ctx->fh.m2m_ctx, vbuf);
> 

-- 
Best regards,
Jacek Anaszewski

[PATCH 2/9] [media] s5p-jpeg: Call jpeg_bound_align_image after qbuf

[Thierry Escande]

From: Tony K Nadackal 

When queuing an OUTPUT buffer for decoder, s5p_jpeg_parse_hdr()
function parses the input jpeg file and takes the width and height
parameters from its header. These new width/height values will be used
for the calculation of stride. HX_JPEG Hardware needs the width and
height values aligned on a 16 bits boundary. This width/height alignment
is handled in the s5p_jpeg_s_fmt_vid_cap() function during the S_FMT
ioctl call.

But if user space calls the QBUF of OUTPUT buffer after the S_FMT of
CAPTURE buffer, these aligned values will be replaced by the values in
jpeg header. If the width/height values of jpeg are not aligned, the
decoder output will be corrupted. So in this patch we call
jpeg_bound_align_image() to align the width/height values of Capture
buffer in s5p_jpeg_buf_queue().

Signed-off-by: Tony K Nadackal 
Signed-off-by: Thierry Escande 
---
 drivers/media/platform/s5p-jpeg/jpeg-core.c | 7 +++
 1 file changed, 7 insertions(+)

diff --git a/drivers/media/platform/s5p-jpeg/jpeg-core.c 
b/drivers/media/platform/s5p-jpeg/jpeg-core.c
index 52dc794..6fb1ab4 100644
--- a/drivers/media/platform/s5p-jpeg/jpeg-core.c
+++ b/drivers/media/platform/s5p-jpeg/jpeg-core.c
@@ -2523,6 +2523,13 @@ static void s5p_jpeg_buf_queue(struct vb2_buffer *vb)
q_data = >cap_q;
q_data->w = tmp.w;
q_data->h = tmp.h;
+
+   jpeg_bound_align_image(ctx, _data->w, S5P_JPEG_MIN_WIDTH,
+  S5P_JPEG_MAX_WIDTH, q_data->fmt->h_align,
+  _data->h, S5P_JPEG_MIN_HEIGHT,
+  S5P_JPEG_MAX_HEIGHT, q_data->fmt->v_align
+ );
+   q_data->size = q_data->w * q_data->h * q_data->fmt->depth >> 3;
}
 
v4l2_m2m_buf_queue(ctx->fh.m2m_ctx, vbuf);
-- 
2.7.4