https://code.google.com/p/chromium/codesearch#chromium/src/third_party/libvpx/so
urce/libvpx/vp8/decoder/decodeframe.c&l=634 reads
ref_buffer[xd->mode_info_context->mbmi.ref_frame][0] + recon_yoffset;
but ref_buffer[i] is only initialized for i=1..(MAX_REF_FRAMES-1), and yet e.g.
bear.webm triggers a value of 0 for
xd->mode_info_context->mbmi.ref_frame
Should the initialization loop at
https://code.google.com/p/chromium/codesearch#chromium/src/third_party/libvpx/so
urce/libvpx/vp8/decoder/decodeframe.c&l=527 simply start at i=0 instead of i=1,
or is there a deeper bug here (i.e. that 0 should never be an index)?
Original issue reported on code.google.com by fischman@chromium.org on 16 Apr 2014 at 6:36
Original issue reported on code.google.com by
fischman@chromium.org
on 16 Apr 2014 at 6:36