Open hsyrja opened 7 years ago
e.g. in gstreamer
[48642.979908][ctx 0x02000000] pcm_sample_bit_depth_luma_minus1 = 0 [48642.979909][ctx 0x02000000] pcm_sample_bit_depth_chroma_minus1 = 0 ... [48642.979952][ctx 0x02000000] slice_parsing_fields = 15246
however in FFmpeg
[48794.782034][ctx 0x02000000] pcm_sample_bit_depth_luma_minus1 = 255 [48794.782034][ctx 0x02000000] pcm_sample_bit_depth_chroma_minus1 = 255 ... [48794.782054][ctx 0x02000000] slice_parsing_fields = 910
According to the tracer log, I think this is a FFmpeg specific issue.
See notes on https://trac.ffmpeg.org/ticket/6642. Maybe something is going wrong around the broken link handling for the initial part of the stream? (The non-failure with gstreamer seems to come from throwing away everything up to the first IDR frame.)
This is how I’ve seen it behaving, always when the stream starts, when the source is live (could start any point of the stream) with a good luck it may start successfully and then it works fine as long as you want.
On 19 Sep 2017, at 15:50, Mark Thompson notifications@github.com wrote:
See notes on https://trac.ffmpeg.org/ticket/6642 https://trac.ffmpeg.org/ticket/6642. Maybe something is going wrong around the broken link handling for the initial part of the stream? (The non-failure with gstreamer seems to come from throwing away everything up to the first IDR frame.)
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/01org/intel-vaapi-driver/issues/264#issuecomment-330528776, or mute the thread https://github.com/notifications/unsubscribe-auth/AGKiWwXQhI1rdPUbyVXqeRt9HDaE00e8ks5sj7ijgaJpZM4PPoU-.
GPU hang occurred when hw accelerated decoding is issued by ffmpeg and video codec is Hevc encoded.
ffmpeg command (decode only) ffmpeg -hwaccel vaapi -hwaccel_device /dev/dri/renderD128 -hwaccel_output_format vaapi -I hang.vid -f null -
Tested in Skylake as wel Kabyake i7.
Here is the input video for reproduce the hang. hang.zip
Behaviour is random hang won't happen every time when it's happen it is always in ffmpeg start.
I don't know if this is intel-vaapi-driver issue or FFmpeg issue so I will file bug to both projects.