On 12/16/24 15:25, James Almer wrote:
On 12/16/2024 5:05 PM, Scott Theisen wrote:
On 12/15/24 22:25, James Almer wrote:
On 12/15/2024 1:14 AM, Scott Theisen wrote:
Originally from:
https://github.com/MythTV/mythtv/
commit/669955c6cb29196b4b5120451b5b998d67a65749
---
libavcodec/v4l2_buffers.c | 38
++++++++++++++++++++++++++++++++++++++
1 file changed, 38 insertions(+)
diff --git a/libavcodec/v4l2_buffers.c b/libavcodec/v4l2_buffers.c
index 23474ee143..56a8f0825c 100644
--- a/libavcodec/v4l2_buffers.c
+++ b/libavcodec/v4l2_buffers.c
@@ -210,6 +210,43 @@ static enum AVColorTransferCharacteristic
v4l2_get_color_trc(V4L2Buffer *buf)
return AVCOL_TRC_UNSPECIFIED;
}
+static void v4l2_get_interlacing(AVFrame *frame, V4L2Buffer *buf)
+{
+ enum v4l2_field field;
+ field = V4L2_TYPE_IS_MULTIPLANAR(buf->buf.type) ?
+ buf->context->format.fmt.pix_mp.field :
+ buf->context->format.fmt.pix.field;
+
+ if (field == V4L2_FIELD_INTERLACED || field ==
V4L2_FIELD_INTERLACED_TB) {
+ frame->flags |= AV_FRAME_FLAG_INTERLACED;
+ frame->flags |= AV_FRAME_FLAG_TOP_FIELD_FIRST;
+#if FF_API_INTERLACED_FRAME
+FF_DISABLE_DEPRECATION_WARNINGS
+ frame->interlaced_frame = 1;
+ frame->top_field_first = 1;
No need to set these two fields. It's done by the generic code.
OK, I'll remove those lines. Although it is not clear to me when
those will be set after libavcodec/pthread_frame.c calls
ff_decode_receive_frame_internal().
They are set in decode_receive_frame_internal(), libavcodec/decode.c,
not too long before the frame is ultimately returned to the caller.
I saw that decode_receive_frame_internal() calls
ff_decode_receive_frame_internal() and then sets the flags, but
ff_decode_receive_frame_internal() is called directly in
libavcodec/pthread_frame.c.
It is not clear to me what calls decode_receive_frame_internal() to set
the flags after the call to ff_decode_receive_frame_internal() in
libavcodec/pthread_frame.c. However, if it is called at some point to
set the flags and it works, it works, so it is not really that
important, just that is was not obvious to me when it would be called.
Regards,
Scott Theisen
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
