CVE-2024-5197: Integer Overflow or Wraparound

There exists interger overflows in libvpx in versions prior to 1.14.1. Calling vpx_img_alloc() with a large value of the d_w, d_h, or align parameter may result in integer overflows in the calculations of buffer sizes and offsets and some fields of the returned vpx_image_t struct may be invalid. Calling vpx_img_wrap() with a large value of the d_w, d_h, or stride_align parameter may result in integer overflows in the calculations of buffer sizes and offsets and some fields of the returned vpx_image_t struct may be invalid. We recommend upgrading to version 1.14.1 or beyond